EPA/600/R-94/160
September 1993
BIOREMEDIATION OF HAZARDOUS WASTES
Research, Development, and Field Evaluations
1993
Biosystems Technology Development Program
Office of Research and Development
U.S. Environmental Protection Agency
U.S. Environmental Protection Agency
Ada, OK; Athens, GA; Cincinnati, OH; Gulf Breeze, FL;
and Research Triangle Park, NC
J&X Printed on paper that contains at least
0^7 SO percent recycled liber.
-------
Disclaimer
The information in this document has been funded wholly or in part by the U.S. Environmental
Protection Agency (EPA) and has been reviewed in accordance with EPA's peer and administrative
review policies and approved for presentation and publication. Mention of trade names or commer-
cial products does not constitute endorsement or recommendation for use.
>
-------
Intentionally Blank Page
-------
Contents (continued)
Page
Section 3; Field Research 43
Field Demonstration of a Constitutive TCE-Degrading Bacterium for the Bioremediation of TCE .... 44
Malcolm S. Shields, Michael Reagin, Robert Gerger, Rhonda Schaubhut, Robert Campbell,
Charles Somerville, and R Hap Pritchard
Bioremediation of TCE: Monitoring the Fate and Effects of a Microorganism Used in a Field
Bioaugmentation Study 47
M.S. Shields, R. Snyder, M. Reagin, R. Gerger, R. Campbell, C. Somerville, and RHap Pritchard
Factors Determining the Effectiveness of Microbial Inoculation in Soils and Sediments: Effectiveness
of Encapsulation 51
Jian-Er Lin, James G. Mueller, and P. Hap Pritchard
Combining Treatability Studies and Site Characterization for Rational Design of In Situ Bioremediation
Using Nitrate as an Electron Acceptor 54
S.R. Hutchins, D.H. Kampbell, M.L. Cook, P.M. Pfeffer, R.L Cosby, J.T. Wilson, B. Newell, J.A
Johnson, V. Ravi, and J.K. Rumery
Rate and Extent of Natural Anaerobic Bioremediation of BTEX Compounds in Ground Water Plumes 58
Morton A. Bariaz, Michael B. Shafer, Robert C. Borden, and John T. Wilson
Field Treatment of BTEX in Vadose Soils Using Hydrofracturing, Vacuum Extraction, and Biofiiters . . 61
DoJIoff F. Bishop, Wendy Davis-Hoover, and Rakesh Govind
Section 4: Pilot-Scale Research 63
Treatment of PCP-Contaminated Soils by Washing with Ethanol/Water Followed by
Anaerobic Treatment 64
Amid P. Khodadoust, Julie A. Wagner, Makram T. Suidan, and Steven I. Safferman
Preliminary Evaluation of Attachment Media for Gas Phase Biofiiters 67
Francis L. Smith, George A. Serial, Paul J. Smith, Makram T. Suidan, Pratim Biswas, and
Richard C. Brenner
Approaches to the Development of Comparative Genotoxicity Risk Assessment Methods
for Evaluating Hazardous Waste Control Technologies . . . 70
Larry D. Claxton
Development and Evaluation of Composting Techniques for Treatment of Soils Contaminated with
Hazardous Waste 71
John A Glaser, Carl L. Potter, Edward D. Kennedy, Jeffrey J. McCormack,
Joseph B. Farrell, and Michael Najar
Engineering Optimization of Slurry Bioreactors for Treating Hazardous Waste in Soil and Sediments 72
John Glaser, Paul McCauley, Majid Dosani, Jennifer Piatt, Edward Opatken, and Diana Roush
Biotreatability of a Vadose Zone Soil Contaminated with Dioctyi Phthalate 73
Don H. Kampbell, Dennis D. Fine, and Jerry W. Anderson
Section 5: Process Research 75
Methanogenic Degradation Kinetics of Nitrogen and Sulfur Containing Heterocyclic Aromatic
Compounds in Aquifer-Derived Microcosms 77
E. Michael Godsy, Donald F. Goerlitz, and Dunja Grbic-Galic
Anaerobic Degradation of Halogenated and Nonhalogenated Phenolic Compounds 80
M.M. Haggblom, M.D. Rivera, L.Y. Young, andJ.E. Rogers
iv
-------
Contents (continued)
Page
Anaerobic Biodegradation of 5-Chlorovanillate as a Model Substrate for the Bioremediation of
Paper-Milling Waste 81
Barbara R. Sharak Genthner
Bioremediation of Soils and Sediments Contaminated with Aromatic Amines 83
Eric J. Weber, David L Spidle, and Kevin A. Thorn
Anaerobic Biotransformation of Munitions Wastes 85
Deborah J. Roberts, Stephen Funk, Don L Crawford, and Ronald L Crawford
Cometabolic Biodegradation of 2,4-DinitrQtoluene Using Ethanol as a Primary Substrate 88
Jiayang Cheng, Makram T. Suidan, and Albert D. Venosa
Effects of Metals on Anaerobic Treatment Processes 90
W. Jack Jones and In Chut Kong
Fate of Highly Chlorinated Dibenzo-p-Dioxins and Dibenzofurans in Anaerobic Soils and Sediments . . 93
Peter Adriaens, Dunja Grbic-Galld, and Gregory D. Sayles
Bioreactor Treatment of Nitrate Contamination in Ground Water. Studies on the Sulfur-Mediated
Biological Denitrification Process 96
Michael S. Davidson and Thomas Cormack
Chemical Interactions and pH Profiles in Microbial Biofilms 99
Joseph R.V. Flora, Makram T. Suidan, Pratim Biswas, and Gregory D. Sayles
Characterization of Biofilter Microbial Populations 101
Alec Breen, Alan Rope, John C. Loper, and P.R. Sferra
Fundamental Studies in the Development of the Gas Phase Biofilter 103
Rakesh Govind, Vivek Utgikpr, Yonggui Shan, Wang Zhao, Madan Parvatiyar, Stephan
Junginzer, and Dolloff F. Bishop
Sequential Anaerobic/Aerobic Treatmen* of Contaminated Soils and Sediments 106
Grace M. Ldpez, Gregory D. Sayles, Karen Buhier, Dolloff F. Bishop, In S. Kim,
Guanrong You, Petra Klostermann, Margaret J. Kupferle, and Douglas S. Upton
PCB Biodegradation During Aerobic Treatment of Sludge from the French Limited NPL Site 108
J.W. Anderson, T. Smith, and J.T. Wilson
Innovative Bioremediation Strategies for Creosote: Geographic Diversity of PAH Degradation
Capabilities at Wood-Treating Sites 110
James G. Mueller, Suzanne E. Lantz, Richard Devereux, Deborah L Santavy,
and P. Hap Pritchard
Section 6: Development of Computer-Based Assessment Systems 115
Design of an Expert System to Select an Appropriate Bioremediation Technique 116
Raymond C. Loehr and Greg E. Schmidt
A Data Visualization System for Bioremediation Analysis . 118
Lewis A. Rossman, Kevin Savage, and John Franco
Section 7: Hazardous Substance Research Center Program 119
Development of a Knowledge-Based Bioremediation Adviser 121
Shu-Chi Chang, Peter Adriaens, Iris D. Tommelein, and Timothy M. Vogel
v
-------
Contents (continued)
Page
Use of Composting Technologies to Treat a TNT-Contaminated Soil 123
James H. Johnson, Jr., Mohammed Moshin, Lily Wan, and Abdul Shafagatti
Effect of Pore-Scale Hydrodynamics on Bulk Reaction Rates 124
Bruce B. Dykaar and Peter K. Kitanidis
In Situ Bioremediation Using a Recirculation Well 125
M.M. Lang, L. Semprini, P.V. Roberts, and Perry L McCarty (presenter)
Chlorinated Aliphatic Hydrocarbon Biodegradation by Methanotrophic Bacteria 126
Laurence H. Smith, Tomas Henrysson, and Perry L McCarty
Anaerobic Biodegradation of BTEX Compounds at Seal Beach, California 127
Harold A Ball, Martin Reinhard, Eva O. Orwin, and Perry L McCarty (presenter)
Biotransformation of Indole and Quinole Under Denitrifying Conditions 128
J.N.P. Black, R.M. Kauffman, D. Denney, D. Grbic-Galic, and Perry L. McCarty (presenter)
Cometaboh'sm of TCE by Nitrifying Bacteria 129
Michael Hyman, Daniel Arp, Sterling Russell, Roger Ely, and Ken Williamson
Assessing the Effect of Environmental Conditions on Chlorophenol Reductive Dechlorination
Pathways and Kinetics «• 131
Sandra Woods, Sheryl Stuart, David Nicholson, Teresa Lemmon, James Ingle>
and John Westall
Spatial Distribution of Nonaqueous Phase Liquid in Sand Cores Using X-ray Computed Tomography 134
John L. Holmes, R. Lee Peyton, and Tissa H. Illangasekare
Scale-Up Implications of Respirometrically Determined Microbial Kinetic Parameters 135
P.J. Sturman, R.R. Sharp, J.B. DeBar, P.S. Stewart, AS. Cunningham, and J.H. Wolfram
Dissipation of Polycyclic Aromatic Hydrocarbons in the Rhizosphere 136
M. Katherine Banks and A. Paul Schwab
Effect of Irreversible Sorption on Bioavailability 137
Amy T. Kan, Gongmin Fu, Mason B. Tomson, and Calvin H. Ward
The Effect of Population and Substrate Interactions on SBR Design Optimization 138
B.C. Baltzis, G.A. Lewandowski, S. Dikshitulu, and KW. Wang
vi
-------
Introduction
The U.S. Environmental Protection Agency (EPA) is
responsible for protecting public health and the environ-
ment from the adverse effects of pollutants.- EPA's
authority to develop regulations and to conduct environ-
mental health research is derived from major federal
laws passed in the last 20 years that mandate broad
programs to protect public health and the environment.
The Clean Air Act, the Safe Drinking Water Act, the
Clean Water Act, the Toxic Substances Control Act, the
Federal Insecticide, Fungicide, and Rodenticide Act, the
Resource Conservation and Recovery Act (RCRA), and
the Comprehensive Environmental Response, Com-
pensation, and Liability Act (CERCLA, known as Super-
fund) all require that EPA develop regulatory programs
to protect public health and the environment
For the control and cleanup of hazardous wastes, the
Superfund law gives EPA broad authority to respond
directly to releases of hazardous materials that endan-
ger public health or the environment. Also, the Super-
fund Amendments and Reauthorization Act of 1980
(SARA) expands EPA's authority in research and devel-
opment, training, health assessments, responding to
community right-to-know issues, and facilitating public
participation. EPA's Office of Research and Develop-
ment (ORD) conducts basic and applied research in
health and ecological effects, hazardous wastes, and
the development and demonstration of remediation
technologies. Technologies are designed to provide ef-
ficient cost-effective alternatives for cleaning up the
complex mixtures of pollutants found at Superfund sites
or at other locations, such as oil spills. As the technolo-
gies advance, ORD transfers information on their use to
groups that apply technologies at specific sites.
Some of the most promising new technologies for solv-
ing hazardous waste problems involve the use of biore-
mediation, an engineered process that relies on
microorganisms such as bacteria or fungi to transform
hazardous chemicals into less-toxic or nontoxic com-
pounds. These microorganisms have a wide range of
abilities to metabolize different chemicals. Scientists can
tailor'a technology to the pollutants at specific sites and
in specific media (e.g., contaminated aquifers, waste
lagoons, contaminated soils) by using organisms in the
treatment system that break down a particular poilutant
under specific conditions.
Bioremediation Is an attractive option because it is a
natural process, and the residues Jrom the biological
processes (e.g., carbon dioxide and water) are usually
geochemically cycled into the environment as harmless
products. These processes also are carefully monitored
to reduce the possibility of a product or a process being
more toxic than the original pollutant. Another advan-
tage of biological treatments—particularly in situ treat-
ment of soils, sludges, and ground water—is that they
can be less expensive and less disruptive than options
frequently used, to remediate hazardous wastes, such
as excavation followed by Incineration or landfilling. Fi-
nally, bioremediation holds another clear advantage
over many technologies relying on physical or chemical
processes: instead of merely transferring contaminants
from one medium to another, biological treatment can
degrade the target chemical.
Until recently, the use of bioremediation was limited by
the lack of a thorough understanding of biodegradation
processes, their appropriate applications, their control
and enhancement in environmental matrices, and the
engineering techniques required for broad application of
the technology. The Agency recognized that, along with
a basic understanding of biological processes, compre-
hensive mechanistic process control, engineering de-
sign, and cost data are necessary for the acceptance
and use of bioremediation by the technical and regula-
tory communities. In response to these needs, ORD
developed an integrated Bioremediation Research Pro-
gram to advance the understanding, development, and
application of bioremediation solutions to hazardous
waste problems threatening human health and the en-
vironment. The program was designed to strike a bal-
ance between basic research activities leading to a
fundamental understanding of biologjcal degradation
processes and engineering activities leading to practical
scientific applications of the technology. The Bioreme-
1
-------
diat'on Research Program is made up of three major
research components: the Biosystems Technology De-
velopment Program, the In Situ Application Program,
and the Bioremediation Field Initiative.
EPA's bioremediation research efforts have produced
significant results in the laboratory, at the pilot scale, and
in the field. The many accomplishments include aquifer
restoration, soil cleanup, process characterization, and
technology transfer. This symposium was held to pre-
sent and discuss recent developments in bioremediation
research undertaken during 1992-93 under the Biosys-
tems Technology Development Program.
2
-------
Executive Summary
In 1987, the U.S. Environmental Protection Agency's
(EPA's) Office of Research and Development (ORD)
initiated the Biosystems Technology Development Pro-
gram to anticipate and address research needs in man-
aging our nation's hazardous waste. Because the
Agency believes that bioremediation offers an attractive
alternative to conventional methods of cleaning up haz-
ardous waste, it has developed a strategic plan for its
acceptance and use by the technical and regulatory
communities. The Agency's strategic plan under this
program is centered on site-directed bioremediation re-
search to expedite the development and use of relevant
technology.
Related bioremediation studies are being carried out at
five EPA Hazardous Substance Research Centers
(HSRCs) under the direction of ORD's Office of Explora-
tory Research (OER). EPA was authorized to establish
these centers by provisions in the 1906 amendments to
the Superfund law calling for research into all aspects
of the "manufacture, use, transportation, disposal, and
management of hazardous substances."
In May 1993, ORD hosted the sixth annual Symposium
on Bioremediation of Hazardous Wastes: Research, De-
velopment, and Field Evaluations, in Dallas, Texas. The
symposium was held in cooperation with EPA's Region
6 offices, marking the first time the annual gathering
convened west of the Mississippi. More than 400 people
attended, including leading bioremediation researchers
and field personnel from federal, state, and local agen-
cies as well as representatives from industry and aca-
demia. The 32 papers delivered at the conference
highlighted recent program achievements and research
projects aimed at bringing bioremediation into more
widespread use. The symposium's poster session was
expanded from the previous year to include presenta-
tions from the five EPA Hazardous Substance Research
Centers in addition to presentations on research being
carried out under the Biosystems Technology Develop-
ment Program.
In this document, abstracts of paper and poster presen-
tations from the Biosystems Technology Development
Program are organized within the six key research and
program areas. Taken as a whole, these topic areas
represent a comprehensive approach to bioremediation
of hazardous waste sites:
1. Bioremediation Field Initiative. This initiative was
instituted in 1990 to collect and disseminate perform-
ance data on bioremediation techniques from field
application experiences. The Agency assists the re-
gions and states in conducting field tests and in
carrying out independent evaluations of site clean-
ups using bioremediation. Through this initiative,
tests are under way at Superfund sites, Resource
Conservation and Recovery Act corrective action fa-
cilities, and Underground Storage Tank sites. Six
paper presentations were devoted to this key pro-
gram area, covering field evaluations at sites using
bioventing, biochemical techniques, and bioremedia-
tion under a variety of aerobic and anaerobic condi-
tions.
2. Performance Evaluation. Performance evaluation
of various bioremediation technologies involves as-
sessing the extent and rate of cleanup for particular
bioremediation methods as well as monitoring the
environmental fate and effects of compounds and
their by-products. Because attempts to remediate a
contaminated site can result in the production of
additional compounds, an important aspect of per-
formance evaluation involves assessing the potential
health effects of processes. Six papers and one
poster were presented concerning the performance
of various bioremediation approaches as well as
risks related to public health.
3. Field Research. Once a bioremediation approach
has proven effective in a bforeactor, for instance, it
must be monitored and evaluated at a field site. The
objective of this level of research is to demonstrate
that the particular bioremediation process performs
as expected In the field. For most bioremediation
technologies, certain key factors concerning applica-
bility (e.g., cost effectiveness) cannot be thoroughly
evaluated until the approach is scaled up and field
tested. Five paper and several poster presentations
3
-------
provided information on recent or ongoing field re-
search.
4. Pilot-Scald Research. Pilot-scale research provides
information on the operation and control of bioreme-
diation technologies and the management of proc-
ess-related residuals and emissions. As such, it is a
necessary step in anticipation of full-scale applica-
tion of a technology. Given the expanding base of
experience with various bioremediation methods, the
need for pilot-scale research is increasing. Two pa-
pers and numerous posters were presented con-
cerning research based on microcosms of field sites.
5. Process Research. Process research involves iso-
lating and identifying microorganisms that carry out
biodegradation processes and the environmental
factors affecting those processes. Such research is
fundamental to the development of new biosystems
for treatment of environmental pollutants (n surface
waters, sediments, soils, and subsurface materials.
Thirteen papers and numerous poster presentations
addressed this critical area.
6. Development of Computer-Based Assessment
Systems. Potential applications of computer tech-
nology promise to make assessments related to the
bioremediation of hazardous waste sites faster and
more comprehensive. For instance, a data base
comprising research information from both the labo-
ratory and the field would provide an important tool
for method evaluation. Similarly, software capable of
modeling the possible outcomes of various remedia-
tion approaches would be indispensable for site
characterization. Two poster presentations ad-
dressed research on systems development.
The Biosystems Technology Development Program
draws on ORD scientists who possess unique skills and
expertise in biodegradation, toxicology, engineering,
modeling, biological and analytical chemistty, and mo-
lecular biology. These scientists work out of the following
laboratories and organizations, all of which are institu-
tional participants in tfie program:
• Environmental Research Laboratory-Ada, Oklahoma
• Environmental Research Laboratory-Athens, Georgia
• Center for Environmental Research Information-
Cincinnati, Ohio
• Risk Reduction Engineering Laboratory-Cincinnati,
Ohio
• Environmental Research Laboratory-Gulf Breeze,
Florida
•. Health Effects Research Laboratory-Research
Triangle Park, North Carolina
The last section of this document gathers abstracts of
poster presentations on bioremediation research per-
formed as part of the HSRC program. The scientists and
engineers involved in ttiis program conduct EPA re-
search sponsored by the following centers:
• Northeast Hazardous Substance Research Center
(Regions 1 and 2)
• Great Lakes and Mid-Atlantic Hazardous Substance
Research Center (Regions 3 and 5)
• South/Southwest Hazardous Substance Research
Center (Regions 4 and 6)
• Great Plains and Rocky Mountain Hazardous Sub-
stance Research Center (Regions 7 and 8)
• Western Region Hazardous Substance Research
Center (Regions 9 and 10)
4
-------
Section 1
Bioremediation Field Initiative
The Bioremediation Field Initiative is one of the major components of EPA's
Bioremediation Research Program. The initiative was undertaken in 1990 to ex-
pand the nation's field experience in bioremediation techniques. Objectives of the
initiative are to more fully document the performance of full-scale applications of
bioremediation; to provide technical assistance to site managecs; and to provide
regular information on treatability studies, design, and operation of bioremediation
projects, including cost and performance data. The initiative is currently tracking
bioremediation activities at more than 130 Superfund sites, Resource Conservation
and Recovery Act corrective action facilities, and Underground Storage Tank sites
nationwide.
Researchers conducted post-remediation monitoring of an aquifer at an industrial
site in Denver, Colorado, that includes heterogeneous geological material. Relying
on core sampling and modeling, researchers studied the likelihood of benzene,
toluene, ethylbenzene, and xylene reappearing in ground water following imple-
mentation of a bioremediation scheme to clean up used crankcase oil that had
leaked from a holding tank.
At a Montana Superfund site, researchers continued gathering data for a third year
on previous and ongoing efforts to bioremediate wood preserving wastes. This
full-scale bioremediation initiative involves both in situ and ex situ systems.
Two sti'dies are assessing the application of bioventing approaches on jet-fuel
contaminated soil at Air Force installations where structures and utilities cannot be
disturbed. Bioventing is an inexpensive in situ process that involves providing air
to oxygen-deprived microbes in unsaturated contaminated soil. Bioventing carried
out in conjunction witti soil warming is being assessed at a site in Alaska, while the
influence of air flow rates on large volumes of contaminated soil is being studied
at an Air Force base in Utah.
Preliminary results from an active wood treating operation in Mississippi indicate
that the metabolic activity of wood degrading fungi in the soil holds promise for
remediating pentachlorophenol and other difficult-to-degrade pollutants.
5
-------
Retrospective Performance Evaluation on In Situ Bloremediatlon: Site
Characterization
John T. Wilson and Don H. Kampbell
Robert S. Kerr.Environmental Research Laboratory, U.S. Environmental Protection Agency, Ada, OK
Bioremediation is difficult to assess in heterogeneous
geological material. Often, oily phase material is asso-
ciated with fine textured material with low hydraulic con-
ductivity. Remedial fluids tend to pass around the fine
textured material. Because the flux of nutrients and an
electron acceptor through the fine textured material is
small, little opportunity exists for bioremediation, and
significant concentrations of contaminants can remain in
subsurface material.
These relationships are illustrated in a case history from
an industrial site in Denver, Colorado (1). A temporary
holding tank under a garage leaked used crankcase oil,
diesel fuel, gasoline, and other materials into a shallow
water table aquifer. Figure 1 shows the relationship
between the garage, the work pit containing the leaking
holding tank, and the approximate area of the spill.
Remedial Effort
Remediation involved removal of separate oily phases,
in situ bioremediation with hydrogen peroxide and min-
eral nutrients, and bioventing. Ground water flow under
ambient conditions was to the north or northeast. The
flow of water during the remediation paralleled the natu-
ral gradient. Water was produced from a recovery well
on the northeast side of the spill RW-1 in Figure 1. The
flow from the well was split. Part of the flow was
amended with hydrogen peroxide and nutrients and re-
charged to the aquifer in a nutrient recharge gallery on
the south side of the spill. The remainder of the flow was
delivered to a ground water recharge gallery located to
the south of the nutrient recharge gallery. From 3 to 6
gpm was delivered to the nutrient recharge gallery, and
4 to 8 gpm was delivered to the ground water recharge
gallery for a total flow of 9 to 11 gpm. The system was
operated from October 1989 to March 1992. At a flow of
10 gpm, from 10 to 15 pore volumes would have been
exchanged in the area between the nutrient recharge
gallery and the recovery well.
RW-1
Hydrocarbon Release MW-3»®
61-B B1 « 61"J
• • 61-A •
\ • MW-2A
61E
61F
Work Pit
Existing
Service
Build Nig
MW-B
MW-1
Nutrient
Recharge
Gallery
Ground-water
Recharge
Gallery
i .... j
20 meters
Figure 1, Infrastructure at anjn situ bioremediation project in
Denver, Colorado. A holding tank In a work pit under
a garage leaked petroleum hydrocarbons to the water
table aquifer. Ground water was pumped from a re-
covery well (RW-1) and filtered through activated
carbon. The flow was split; part was amended with
hydrogen peroxide and mineral nutrients, and re-
charged (n a nutrient recharge gallery. The remainder
was recharged In a ground water recharge gallery.
The system was designed to sweep hydrogen perox-
ide and nutrients under the service building. MW-1,
-2A, and -3 are monitoring wells. 61A through 61J are
boreholes for cores.
6
-------
Ground Water Remediation
Table 1 compares the reduction in concentration of ben-
zene and total benzene, toluene, ethylbenzene, and
xylene (BTEX) compounds In ground water that was
achieved by in situ bioremediation at the site in Denver.
As shown in Figure 1, monitoring wells MW-1 and MW-8
are in areas with oily phase hydrocarbons. Well MW-2A
is just outside the region with oily phase hydrocarbons.
Weil MW-3 is a significant distance from the region with
oily phase hydrocarbons; it sampled the plume of con-
taminated ground water that moved away from the spill.
Table 1. Reduction In Concentration of Hydrocarbon
Contaminants in Ground Water Achieved by hi Situ
Bioremediation
Banana Total BTEX
Well
Before
During
After Before
(ng/llter)
During
After
MW-1
220
<1
<1
2,030
164
<6
MW-8
180
16
1,800
331
34
MW-2A
?
11
0.8
?
1,200
13
MW-3
11
5
2
1,200
820
46
RW-1
<1
2
<1 "
<1
2
<1
Prior to remediation, concentrations in wells MW-1 and
MW-8 were equivalent. Well MW-1 was closest to the
nutrient recharge gallery, and fre aquifer surrounding
MW-1 was completely remediated; BTEX compounds
were undetectable in ground water. Jn well MW-8, imme-
diately adjacent to the point of release, the concentration
of benzene was reduced at least one order of magni-
tude, and the concentration of benzene and BTEX com-
pounds in well MW-3 also were reduced an order of
magnitude.
Water in the monitoring wells and the recirculation well
contained low concentrations of contaminants by March
1992. Active remediation was terminated and the site
entered a period of post-remediation monitoring.
Remediation of Aquifer Solids
in June 1992, core samples were taken from the aquifer
to determine the extent of hydrocarbon contamination
remaining, and to determine whether a plume of con-
tamination could return once active remediation ceased.
The site was cored along a transect downgradient of the
release. The transect extended laterally from clean ma-
terial, through part of the spill, into clean material on the
other side (Figure 1). In each borehole, continuous
cores extended vertically from clean material above the
spill, and through the spiil to clean material below. The
cores were extracted and analyzed for total petroleum
hydrocarbons (TPH) and for the concentration of individ-
ual BTEX compounds.
The relationship among the land surface, the watBr ta-
ble, the region containing hydrocarbons, and the bed-
rock is presented in Figure 2. Significant amounts of
hydrocarbons remain within a narrow interval approxi-
mately 2.0 feet thick, near the water table. The total
saturated thickness of the aquifer was approximately 20
feet At the time of sampling, the elevation of the water
table was 5280.5 feet above mean sea level (AMSL) and
all the hydrocarbons were below the water table.
ABC D E F
5300 j i i i i i. 5300
/S '"v\aay
- 5290
Source ol Hydrocarbons Sj
Water Tabla .2
£.5280 52BO g
g N Residual Hydrocarbon ^
I 1
Ul s
5270 - - 5270 >
Sandy Aquifer.
5260 : 5260
Bedrock
Figure 2. Crosa section showing the vertical relationship of the
land surface, water table, and residual hydrocarbon
remaining after bioremediation, and the lower confin-
ing layer of the aquifer. The cross section runs
through core boreholes depicted In Figure 1.
The highest concentrations of hydrocarbons at the site
in Denver were obtained in samples from the borehole
(D), which was closest to the work pit. Table 2 presents
the vertical distribution of BTEX compounds and TPH in
borehole D. The material in the interior of the spill had
higher proportions of BTEX compounds. Table 3 makes
the same comparison at the most contaminated depth
interval along the transect. Material closer to the spill
had higher concentrations of TPH and greater relative
proportions of BTEX compounds.
Apparently at the Denver site, a cortex of material that
has been physically and biologically weathered sur-
rounds a central core of material that has not been
depleted of BTEX compounds. The concentration of an
individual petroleum hydrocarbon in solution in ground
water in contact with an oily phase hydrocarbon can be
predicted by Raoulfs law. The solution concentration in
water should be proportional to the mole fraction of the
hydrocarbon in the oily phase.
5290
7
-------
Table 2. Vertical Extant of Total BTEX Compounds and TPHs
at Borehole D, the Most Contaminated Borehole In
the Transect (Figure 1)
Color and
Elevation TPH BTEX Benzene Texture
(feet AMSL) —(mg/kg>-
281.14 to
5280.31
<44
<1
<0.2
Brown sand
5280.31 to
5279.97
227
5.1
<0.2
Brown sand
5279.97 to
5279.56
860
101
<0.2
Black sand
5279.56 to
5279.14
1176
206
4.3
Black sand
5279.14 to
5278.97
294
27
0.66
Black sand
5270.97 to
5278.64
273
7.4
0.26
Black sand
5278.94 to
5278.22
<34
<1
<0.2
Black sand
5278.22 to
5277.14
<24
<1
<0.2
Brown to
yellow sand
Table 3. Lateral Distribution of Total BTEX Compounds and
TPHs along the Transect (Figure 1) at the Most
Contaminated Depth Interval
Borehole
TPH
BTEX
—(mg/kg)—
Benzene
B
167
0.8
<0.2
C
156
3.5
<0.2
D
1,
,176
260
4,3
E
156
3.5
0.06
Assume that the weathered material is weathered be-
cause it is in effective contact with moving ground water,
which supplied nutrients and electron acceptors, and the
unweathered residual is not weathered because it was
not in effective contact and the supply of nutrients and
electron acceptors was inadequate. If partitioning be-
tween, moving ground water and the weathered oily
residual controls the concentration of hydrocarbons in
the water, the tenfold reduction in concentration of ben-
zene and BTEX compounds shown in the weathered
core material (Table 3) would produce the tenfold reduc-
tion in concentration of benzene and BTEX compounds
shown in the monitoring wells (Table 2).
inputs to Model
Raoulfs law was used to estimate the concentration of
BTEX compounds in ground water in contact with the
weathered residual. This information was used to imple-
ment an exposure model intended to evaluate the risk
of the contamination remaining after remediation. See
lllangasekare et al. for the results of the mathematical
modeling (2).
References
1. Nelson, C., R.J. Hicks, and S.D. Andrews. 1993.
In-situ bioremediation: an integrated system ap-
proach. In: Flathman, P.E., D.E. Jerger, and J.H.
Exner, eds. Bioremediation: Field Experiences. In
Press. Lewis Publishers, Chelsa, Ml.
2. lllangasekare, T.H., D.C. Szlag, and J.T. Wilson.
1993. Retrospective performance evaluation on in
situ bioremediation: modeling and risk assessment.
Proceedings of U.S. EPA Symposium on Bioreme-
diation of Hazardous Wastes: Research, Develop-
ment, and Field Evaluations. Dallas, TX, May 4-6,
1993.
8
-------
Retrospective Performance Evaluation on in Situ Bioremediation:
Modeling and Risk Assessment
Tissa H. Illangasekare and David C. Szlag
Civil, Environmental, and Architectural Engineering, University of Colorado, Boulder, CO
John T. Wilson
Robert S. Kerr Environmental Research Laboratory, U.S. Environmental Protection Agency, Ada, OK
Conventional methods for determining the extent of
cleanup at a bioremediation site can often be mislead-
ing. Monitoring wells may show very low or zero levels
of contaminants after active bioremediation, but then the
levels may increase over time. In most cases, regulatory
authorities require a direct measure of the residual
nonaqueous phase liquids (NAPLs) after bioremediation
in addition to monitoring well data. Often the relative
composition of the oily phase is assumed to remain
constant during bioremediation. This assumption is
conservative and generally leads to target levels of total
petroleum hydrocarbons (TPH) concentrations on the
order of 10 to 100 mg/kg aquifer material. Many biore-
mediation schemes, however, may preferentially de-
grade the compounds of regulatory concern leaving
relatively high TPH levels in the soil that pose a minimal
risk. This modeling study focuses on developing a meth-
odology to evaluate the possible risk, if any, associated
with benzene, toluene, ethylbenzene, and xylene
(BTEX) sources left in soils after the implementation of
a bioremediation scheme. This developed methodology
will assist in providing answers to the following ques-
tions: 1) Will BTEX reappear in ground water? 2) How
long will it take the plume to reappear? 3) What concen-
tration level may be expected? The results of the case
study also will assist in providing a technical basis for
monitoring schedules, locating compliance wells, and
constructing rational criteria for site closure.
Site Description and Characterization
The type and extent of contamination beneath a service
building at an industrial site in Denver, Colorado, is
documented in a companion paper by Wilson and
Kampbell (1). In brief, used hydrocarbons (crankcase oil
and gasoline) were released from an underground hold-
ing tank. These hydrocarbons exist as light non-aque-
ous phase liquids (LNAPLs) and have formed a smear
zone approximately 1 to 2 ft thick near the water table.
Soil was sampled at approximately 10-ft intervals trans-
verse to the original NAPL plume in order to determine
the remaining concentrations of TPH and BTEX. Tests
were conducted to determine the hydraulic conductivity
in the vicinity of the contaminant source.
Hydraulic conductivity data were collected using a pump
test, laboratory parameter tests, and bail tests. The
pump test provides a conductivity value that is repre-
sentative over the aquifer cross section but may not be
representative of the aquifer material containing en-
trapped LNAPL. Laboratory parameter tests conducted
on 28 soil core samples showed hydraulic conductivity
varies by two orders of magnitude across the sria with
some highly permeable channels evident. Relatively low
permeabilities are found in the soils that are visibly
stained with hydrocarbons and are closest to the water
table. These data and observations of the samples dur-
ing coring reveal the presence of layering in the aquifer.
Model Selection
The following observations at the site and in the labora-
tory indicated the need for three-dimensional simulation:
1) visual inspection of aquifer material indicated the
presence of coarse gravel lenses, clayey sands, and
sands of varying gradation; 2) LNAPL plumes are inher-
ently three-dimensional, forming thin pancakelike
plumes in the capillary fringe and just beneath the water
table; 3) LNAPL can become entrapped in coarse
lenses, which act as preferential flow channels, well
beneath the water table; and 4) solute plumes are not
vertically homogeneous and biological activity will not be
uniformly distributed vertically.
9
-------
Bypassing of the LNAPL plume by nutrients and an
electron acceptor, attributable to a lowered hydraulic
conductivity of the central part, resulted in high TPH and
BTEX levels in some cores. Coupled with tfie clear need
for a three-dimensional model are other criteria such as
availability, ease of use, reliability, and cost. MODFLOW,
a three-dimensional ground water flow model developed
by the U.S. Geological Survey (USGS), was selected to
simulate the ground water flow. Solute transport is simu-
lated with a three-dimensional random walk called
RAND3D.
Modeling Approach
i
The problem domain was modeled as a rectangular area
300 ft long and 200 ft wide. Two wells outside of the
modeled domain were used as reference head locations
for general head boundaries. The NAPL contaminant
zone covers approximately 1,600 ft2 area and a soil
depth of approximately 1.7 ft. In order to accurately
assess the mass flux from the LNAPL contaminant
source, three layers were chosen in the model. The
upper two layers are 1 ft thick and the bottom layer is
18 ft thick. The LNAPL organic contaminant is confined
to the upper two layers.
Data for the period June 8,1989, to April 1,1990, were
used for calibration. The goodness of fit between the
model and measured well data was characterized by a
mean residual and standard deviation of the mean re-
sidual. The best fit model was obtained by assigning the
pump test average hydraulic conductivity to the bottom
layer, which carried the majority of water.
The main focus of this modeling effort is to determine
how much contaminant mass will be transported from
the remaining residual and if it will generate a plume of
regulatory concern. A monitoring well screened only in
the upper two layers would detect the highest contami-
nant concentrations. A pumping well screened over the
entire aquifer thickness also is being considered. In this
case, however, dilution will play a major role in reducing
the maximum concentrations. Two significant assump-
tions are used in the solute transport modeling: 1) the
concentration of BTEX in the source zone remains con-
stant, and 2) water flowing from the contaminated cells
is in equilibrium with the residual NAPL. Using the heads
generated by MODFLOW, the ground water velocity
through each source cell was calculated. The known
NAPL BTEX concentration then was used to calculate
the equilibrium concentration and, consequently, the
mass flux. Estimated benzene mass fluxes were con-
verted into particle inputs for each layer. The particle
tracking model was used to simulate solute transport
using the velocity field generated from MODFLOW.
Results and Discussion
Any simulation of solute transport requires specification
of the contaminant source. For the case of a NAPL spill,
the source function will consist of a continuous mass flux
of solute from the residual NAPL phase to the aqueous
phase. Many researchers have shown that equilibrium
is quickly reached in spill scenarios if the ground water
velocity is low and the "residence time" of the water in
contact with NAPL is "long enough." From a regulatory
standpoint the assumption of equilibrium is conserva-
tive, as greater mass fluxes cannot be achieved. The
water flux can vary significantly in the source zone often
giving a misleading indication that the contaminant
transport is rate limited. This primary problem has been
a focus of our work. Preferential flow paths often de-
velop within the source zone in areas with low BTEX and
TPH concentrations, allowing water flow to bypass the
more highly contaminated areas. Laboratory determina-
tion of the hydraulic conductivity in the samples contain-
ing high amounts of TPH confirms this observation.
A key result of the modeling study is that the solute
plume emanating from a NAPL source is not homoge-
neous. In general, fre solute plume will consist of sub-
plumes over different depth intervals, of widely different
concentrations, and moving at different velocities. A
regulatory question posed earlier in this paper concerns
how long the compliance wells should be monitored.
The answer when all the subplumes have reached
steady-state. The plumes in the middle and bottom layer
have reached or are close to equilibrium by 330 days.
The plume in the top layer, which incidently has the
highest benzene concentration, has not reached equi-
librium at 420 days.
The design of the compliance wells will have tremen-
dous impact on the actual sampled concentration. If the
wells are bailed or pumped so that the well volume is
completely mixed, significant dilution will occur. The ex-
isting monitoring wells at the site are screened over the
top 5 ft of the aqurfer. The maximum concentration
achieved in the well screened over a 5-ft interval
reaches a steady-state concentration of 26 ppb. If that
well is screened over the entire saturated thickness, a
concentration of 15 ppb is achieved. Even greater dilu-
tion will occur if the well is pumped.
Several operational considerations for risk assessment
and compliance well monitoring can be made from the
modeling study: 1) a benzene plume will reestablish
itself at the site, but it will be three orders of magnitude
lower than the federal maximum contaminant level
(MCL); new standards may be set and the risk from this
plume may be deemed significant; 2) the local hydraulic
conductivity plays a significant role in determining the
contaminant mass flux and in creating subplumes of
different concentration and velocity; 3) compliance well
monitoring will have to be continued past August 1993
10
-------
so that solute plumes in all levels will reach steady-state;
4) retardation coefficients and effective porosity data
would significantly improve the time of arrival estimate
of the solute plume; and 5) compliance well design
should be carefully considered when sampling a three-
dimensional plume, as the well design can lead to sig-
nificant contaminant dilution.
Conclusions
A modeling methodology for the retrospective evaluation
of bioremediated aquifers contaminated with organic
chemicals was developed. The primary hypothesis on
which the methodology was based is that during the
spill, the NAPL contaminant becomes entrapped prefer-
entially in coarse formations in the saturated zone and
fine formations in the unsaturated zone. This hypothesis
is supported by laboratory (2) and field data. Flow chan-
nels created by naturally occurring aquifer soil hetero-
geneities as well as macroscale entrapment of the NAPL
also will produce preferential paths for the treating
agents. The proposed methodology requires that these
local heterogeneities in the contaminant zone of the spill
be captured. The standard pump tests that provide the
regional values for transmissivity will not have the ade-
quate resolution to capture these spill-site-scale hetero-
geneities. Even though the hydraulic conductivity values
determined In the laboratory on distorted soil samples
were used in. this study, a more appropriate charac-
terization method would be well-designed bail tests (or
slug tests), which capture the local layered heterogenei-
ties more accurately. These local hydraulic conductivity
values allow us to obtain the velocity field in the contami-
nant zone and to subsequently determine the contami-
nant mass flux. Solute breakthrough curves determined
by this method then can be used to conduct risk analysis
and to provide a rational basis for post-remediation well
monitoring.
References
1. Wilson, J.T., and D.H. Kampbell. 1993. Retrospec-
tive performance evaluation on in situ bioremedia-
tion; site characterization. Proceedings of U.S. EPA
Symposium on Bioremediation of Hazardous
Wastes: Research, Development, and Field Evalu-
ations, Dallas, TX, May 4-6, 1993.
2. Illangasekare, T.H., D.C. Szlag, J. Campbell, J. Ram-
sey, M. Al-Sherida, and D.D. Refole. 1991. Effect of
heterogeneities and preferential flow on distribution
and recovery of oily wastes in aquifers. Proceeding
of Conference on Hazardous Waste Research. Kan-
sas State University, Manhattan, KS.
Acknowledgments
The support of the U.S. Environmental Protection
Agency through the Hazardous Substance Research
Center at Kansas State University (agreement R-
815709) is gratefully acknowledged. We also would like
to thank Lisa Weers of the Colorado Department of
Health for her assistance.
11
-------
Evaluation of Full-Scale In Situ and Ex Situ Bioremediation of
Creosote Wastes In Soils and Ground Water
Ronald C. Sims
Utah State University, Logan, UT
John E Matthews, Scott G. Huling, Bert E. Bledsoe, and Mary E Randolph
U.S. Environmental Protection Agency, Ada, OK
Daniel Pope
Dynamac Corporation, Ada, OK
Objectives of the bioremediation field initiative are to:
1) document more fully the performance of full-scale
bioremediation field applications in terms of treatment
effectiveness, operational reliability, and cost; 2) provide
technical assistance to EPA and state agency site man-
agers who are overseeing or considering the use of
bioremediation; and 3) develop a treatability database
that will be available through EPA's Alternative Treat-
ment Technology Information Clearinghouse (ATTIC).
The performance evaluation project in libtay, Montana,
mainly focuses on the first objective.
The Champion International Superfund Site in Libby,
Montann, was nominated by the Robert S. Kerr Environ-
mental Research Laboratory as a candidate site for
performance evaluation. Two forms of wood pre-
servative were used at the site: creosote, containing
polycyclic aromatic hydrocarbons (PAHs) and pen-
tachlorophenol (PCP). PAHs are currently the primary
components of concern at the site. The performance
evaluation project, now in the third year, is under the
direction of Dr. Ronald Sims of Utah State University.
The bioremediation performance evaluation consists of
three phases: 1) summarizing previous and current re-
mediation activities; 2) identifying critical site charac-
terization and treatment parameters necessary to
evaluate bioremediation performance for each of the
bioremediation treatment units; and 3) evaluating biore-
mediation performance based on the information ob-
tained.
Three biological treatment processes are addressed in
the bioremediation performance evaluation: 1) above-
ground fixed film bioreactor for treatment of extracted
ground water from the upper aquifer, 2) surface soil
bioremediation in a prepared-bed, lined land treatment
unit (LTU); and 3) in situ bioremediation of the upper
aquifer at the site. Each biological treatment process will
be evaluated with regard to design, performance, and
monitoring activities. Figure 1 illustrates the relative lo-
cations of the three treatment' processes at the site.
Biological Treatment Processes
The upper aquifer above-ground treatment unit is pro-
vided for biological treatment of extracted ground water
for removal of PAHs and PCP prior to reinjection via an
infiltration trench. The subsequent biological treatment
consists of two fixed-film reactors operated in series.
The first reactor has been used for roughing purposes,
while the second has been used for polishing and reoxy-
genation of the effluent prior to reinjection. The system
was commissioned in February 1990.
An evaluation of the system components for equaliza-
tion and for biotreatment is being conducted. Equaliza-
tion system components include four ground water
extraction wells and an equalization tank consisting of a
cylindrical horizontal flow tank with a nominal hydraulic
residence time of 6 hours at a flow rate of 10 gpm. The
bioreactor treatment system components include nutri-
ent amendment, influent pumping, bioreactor vessels,
aeration, heating, and effluent pumping. Figure 2 illus-
trates the components of the above-ground treatment
system for extracted ground water.
The LTUs have been used for bioremediation of con-
taminated soil from three primary sources, including
tank farm, butt dip, and waste pit areas. Contaminated
12
-------
3004
,"9500
9501
t 3007
Infiltration
Tranen
Ground Water Flow
LTU
LTU
• Monitoring Wells I
¦ Injection Walls I
Figure 1. Relative locations of the three treatment processes
at the site, Including the bioraaetof building for treat-
ment at extracted ground water, LTUs, and monitor-
ing and Injection wells (or In situ treatment
soil was excavated and moved to one central location,
the waste pit After being pretreated in the waste pit
area, soil is further treated in one of two prepared-bed,
lined LTUs. Total estimated contaminated soil volume for
treatment is 45,000 yd3 (uncompacted). Contaminated
soil cleanup goals (dry-weight basis) are: 1) 88 mg/kg
total (sum of 10) carcinogenic PAHs; 2) 8 mg/kg naph-
thalene; 3) 0 mg/kg phenanthrene; 4) 7.3 mg/kg pyrene;
5) 37 mg/kg PCP; and 6) £0.001 mg/kg 2,3,7,8-dioxin
equivalent
The LTU comprises two adjacent 1-acre cells. Compo-
nents of the soil bioremediation system for each LTU cell
include the treatment zone, liner system, and leachate
collection system. Each cell is lined with low-permeabil-
ity materials to minimize leachate infiltration from the
unit When reduction of contaminant concentrations in
all lifts placed in the LTU has reached the cleanup goals
specified in the Record of Decision (ROD), a protective
cover will be installed over the total 2-acre unit and
maintained in such a way as to minimize surface infiltra-
tion, erosion, and direct contact.
Contaminated soil is applied and treated in lifts (approxi-
mately 9 in. thick) in the designated LTU. Degradation
rates, volume of soil to be treated, initial contaminant
Ground Water
from Extraction Walls
Steam —
Heat Exchanger
Nutrient .
Suppfy Tank
Heal
Exchanger
Fluid
Glower
To Rock Filer and t
Infiltration System i
Figure 2. Components of the above-ground treatment system
for bioremediation of extracted ground water.
concentration, degradation period, and LTU size deter-
mine the time required for remediation "of a given lift.
Based on an estimated 45-day timeframe for remedia-
tion of each applied lift to acceptable contaminant levels,
an estimated 45,000 yd3 of contaminated soil, and a
2-acre total LTU surface area, the projected time to
complete soil remediation is 8 to 10 years.
The pilot in situ bioremediation system for the upper
aquifer area involves the addition of hydrogen peroxide
and inorganic nutrients to stimulate the growth of
contaminant-specific microbes. The hydrogen peroxide
injection system was designed to maintain a concentra-
tion of approximately 100 mg/L of hydrogen peroxide.
Injection flow rate is approximately 100 gpm into three
injection clusters. Inorganic nutrients in the form of po-
tassium tripolyphosphate and ammonium chloride are
continuously added to achieve concentrations in the
injection water of 2.4 mg/L and 1 mg/L of nitrogen and
phosphorus, respectively.
The ROD calls for cleanup levels in the upper aquifer of
40 parts per trillion (ppt) for total carcinogenic PAHs, 400
ppt for total noncarcinogenic PAHs, 1.05 mg/L for PCP,
5 ng/L for benzene, 50 ng/L for arsenic, and a human
health threat no greater than 10*5 for ground water con-
centrations of other organic and/or inorganic com-
pounds.
Performance Evaluation Activities
Performance evaluation of the upper aquifer above-
ground treatment system involves evaluation of the
bioreactor system. Treatment evaluation is focused on
13
-------
characterizing performance with regard to system capa-
bility to remove PAHs and PCP from the ground water,
optimizing operation within the bioreactors, and investi-
gating the fate of contaminants in the bioreactors under
environmental operating conditions at the site. The
above-ground treatment system was sampled during
1991 and 1992 for chemical, physical, and biological
parameters. In addition, a hydraulically equivalent pilot-
scale reactor was constructed and operated to evaluate
abiotic reactions of chemicals present in the water
phase within the bioreactors. The information generated
from the sampling and monitoring of the full-scale reac-
tor and from the operation of the pilot-scale reactor will
be combined with data provided by Champion Interna-
tional to provide an in-depth evaluation of performance.
Performance of the soil bioremediatfon system in the
LTUs involves evaluation of the reduction in concentra-
tion of PAHs and PCP with time and with deptti within
the LTUs. The primary purpose of the LTU soii sampling
program being carried out in this project is to determine
the statistical significance and extent of contaminated
soil treatment at this site. A quantitative expression of
data variability is necessary to determine an accurate
estimate of biodegradation of these contaminants at
field scale. Such an expression will allow data generated
to be used by others to help estimate biodegradation
potential of similar types of waste under similar condi-
tions at other sites.
In most soils and disturbed soil materials, physical and
chemical properties are not distributed homogeneously
throughout the volume of the soil material. The variability
of these properties may range from 1 to more than 100
percent of the mean value within relatively small areas.
Chemical properties, including contaminants, often have
the highest variability. A first approximation of the total
variance in monitoring data can be defined by the fol-
lowing equation:
V, - VJk + Va/k • n
where:
k = the number of samples
n «the number of analyses per sample
k*n = the total number of analyses
Vt = the total variance
Va = the analytical variance
V, = the sample variance
In general, sampling efforts to minimize V, will result in
the best precision. Analytical procedures frequency
achieve precision levels (VVk'n) of 1 to 10 percent,
while soil sampling variation (Vs) may be greater than
35 percent. Sampling designs that will reduce the mag-
nitude of Vs should be employed where possible. There-
fore, the sampling procedures used in ttiis evaluation
were designed to minimize Vs and provide repre-
sentative information about the transformation of PAHs
and PCP witfiin the LTUs.
The LTUs were sampled in May, June, July, and Sep-
tember of 1991, and September of 1992. Field-scale
investigations concerning PAH and PCP concentrations
were supported by laboratory mass-balance investiga-
tions of radiolabeled compounds for determination of
mineralization as well as humification potential for target
contaminants.
Performance evaluation of the in situ bioremediation
system has focused on characterization of the water
phase, the solid phase (aquifer materials), and oil asso-
ciated with the aquifer solid material. The aquifer was
sampled during 1991 and 1992. An evaluation of the
water phase has included measurement of dissolved
oxygen (DO) concentrations, the inorganic chemicals
iron and manganese to evaluate potential abiotic de-
mand for injected hydrogen peroxide, and the concen-
trations of PAHs and PCP. An evaluation of the aquifer
solid phase has included PAH and PCP concentrations
in treated and background areas at the site. Laboratory
mass-balance experiments using radiolabeled target
compounds have been used in conjunction with field-
scale measurements to provide additional information
concerning biotic reactions (mineralization) and poten-
tial abiotic reactions (poisoned controls).
Performance Evaluation Reports
Separate reports are being prepared that will address
each of the three biological treatment systems at the
site: 1) above-ground system for extracted ground
water; 2) soil bioremediation in prepared-bed LTUs; and
3) in situ treatment Information generated from full-
scale characterization and monitoring, from pilot-scale
studies, and from laboratory treatability studies will be
combined with information supplied by Champion Inter-
national to provide an integrated evaluation of bioreme-
diation performance at the Libby, Montana, site that can
be used to evaluate and select rational approaches for
characterization, implementation, and monitoring of
bioremediation at other sites.
14
-------
An Evaluation of Concurrent Bioventing of Jet Fuel and Several Soil Warming
Methods: A Field Study at Eielson Air Force Base, Alaska
Gregory D. Sayles and Richard C. Brenner
U.S. Environmental Protection Agency, Risk Reduction Engineering Laboratory, Cincinnati, OH
Robert E. Hinchee
Battelle Laboratories, Columbus Division, Columbus, OH
Catherine M. Vogel
U.S. Air Force Armstrong Laboratories, Tyndali Air Force Base, FL
Ross N. Miller
U.S. Air Force Center for Environmental Excellence, Brooks Air Force Base, TX
Bioventing is the process of supplying oxygen in situ to
oxygen-deprived soil microbes by forcing air through
unsaturated contaminated soil at low flow rates (1).
Unlike soil venting or soil vacuum extraction technolo-
gies, bioventing attempts to stimulate biodegradative
activity while minimizing stripping of volatile organics,
thereby destroying the toxic compounds in the ground.
Since the bioventing equipment (air injection/withdrawal
wells, air blower, and soil gas monitoring wells) is rela-
tively noninvasive, bioventing technology is especially
valuable for treating contaminated soils in areas where
structures and utilities cannot be disturbed, such as at
the Hill Air Force Base (AFB) site.
Through the Bioremediation Held Initiative, a coopera-
tive program between EPA's Office of Research and
Development and Office of Solid Waste and Emergency
Response, EPA's Risk Reduction Engineering Labora-
tory began a 2.5-year field study of in situ bioventing in
the summer of 1991 in collaboration with the U.S. Air
Force at Eielson AFB near Fairbanks, Alaska. The site
has JP-4 jet fuel-contaminated unsaturated soil where a
spill has occurred in association with a fuel distribution
network. The contractor operating the project is Battelle
Laboratories, Columbus, OH. With the pilot-scale expe-
rience gained in these studies and others, bioventing
should be available in the near future as a reliable,
inexpensive, and unobtrusive means of treating large
quantities of organically contaminated soils.
Methodology
Site history, characterization, installation, and monitor-
ing were summarized previously (2). Figure 1 shows a
plan view of the project. Briefly, four test plots have been
established, all receiving relatively uniform injection of
air. The following four test plots are being used to evalu-
ate three soil warming methods:
• Passive Warming. Enhanced solar warming in late
spring, summer, and early fall using clear plastic cov-
ering over the plot, and passive heat retention the
remainder of the year by applying insulation on the
surface of the plot.
• Active Warming. Warming by applying heated water
from soaker hoses 2 ft below the surface. Water is
applied at roughly 35°C and at an overall rate to the
plot of roughly 1 gal/min. Five parallel hoses 10 ft
apart deliver the warm water. The surface is covered
with insulation year-round.
• Buried Heat Tape Warming. Warming by heat tape
buried at a depth of 3 ft and distributed throughout
the plot, 5 ft apart
• Contaminated Control. Contaminated soil vented with
injected air with no artificial method of heating,
The passively heated, actively heated, and control test
plots were installed in summer 1991, and the heat tape
plot was installed in September 1992.
15
-------
\
N
Jw
• IN"
Taxiway
¦ i
n a $¦» o S i o
Active
Warning
System
1 Passiv?
Active
ro
Trailer
Pumphouse
Taxiway
it
0 25' 50'
Scale
a. Ground Water Monitoring Well
» Air Injection/Withdrawal Wall
a • Three-level Soil Gas Probe
T - Three-level Thermocouple Probe
• *¦ Air Injection/Withdrawal Wall
(currently not In use)
Figure 1. Plan view of the joint U.S. EPA and U.S. Air Force Woven ting activities at Elelson AFB, near Fairbanks, Ala-ska. The
"surface" plot Is the test plot containing heat tape.
Periodically, in situ respirometry tests (3) are conducted
to measure the in situ oxygen uptake rates by the mi-
croorganisms. These tests allow estimation of the biode-
gradation rate as a function of time and, therefore, as a
function of ambient temperature and soil warming tech-
nique. Quarterly comprehensive and monthly abbrevi-
ated in situ respiration tests are planned.
Final soil hydrocarbon analyses will be conducted in late
1993 and compared with the initial soil analysis to docu-
ment actual hydrocarbon loss due to bioventing.
Results
Evaluation of Soil Warming Methods. Table 1 shows soii
temperatures in each plot averaged over the plot and
over a 3-month period (a season). Each warming
method calls for maintaining temperatures higher tiian
the unheated control despite, for example, mean mini-
mum ambient air temperatures in January of about
Table 1. Temperatures In Each Plot as a Function of Season
Temperature ('C)
Season
Active
Passive
Heat Tape
Control
Autumn 1991
10.8
5.7
NA
5.7
Winter 1992
10.1
0.5
NA
-0.2
Spring 1992
17.4
0.7
NA
-0.5
Summer 1992
18.1
15.6
NA
9.6
Autumn 1992
17.0
7.1
8.0
3.8
Winter 1993
15.9
1.3
10.0
0.2
Note- Temperatures are averaged over the plot and over the length
of the season.
NA = Not available, prior to installation.
-20°C. The active and heat tape warming methods in-
volve maintaining summerlike temperatures during the
winter.
16
-------
Biodegradation Rates. Soil warming is only worthwhile
if the resulting elevated temperature provides enhanced
rates of biodegradation. Figure 2 shows the rate of
biodegradation from all test plots, as measured by in situ
respirometry, as a function of seasonal averaged tem-
perature. The trend of higher rates at higher tempera-
tures is clear. To clarify this relationship, an Arrheniiis
plot was constructed (Figure 3) plotting iog10(rate) vs.
inverse absolute temperature. If an Arrhenius relation-
ship exists, ttie data should be linear in this plot The
resulting linear fit gives: log10[rate(mg/kg/day)] =
-4030.(1/T) +14.6, regression constant R=0.88, where
temperature, T, has units of K. The resulting Arrhenius
relationship is:
Rate (mg/kg/day) = 3.98 x 1014e'92eQ/T
This function is in agreement witti previous studies (4).
Soil Sampling. Soil samples for total petroleum hydro-
carbon measurements will be taken at the completion of
the project and compared to the initial soil analysis to
determine net loss of hydrocarbons by bioventing in
each test plot
10
0 5 10 15
Temperature fC)
Figure 2. Rate of petroleum biodegradation aa a function of
temperature averaged over the test plot and averaged
over the season (I.e., averaged over winter 1992).
1 1
&
0.1
340
350 360
1/(T«mp. V.E+05) (1/deg K)
370
Figure 3. Rate at petroleum biodegradation as a function of
Inverse temperature. Data are compiled from all test
plots. Line Is the linear regression fit
References
1. Hoeppel, R.E., R.E. Hinchee, and M.F. Arthur. 1991.
Bioventing soils contaminated with petroleum hydro-
carbons. J. Indust Miqrobiol. 8:141.
2. Sayles, G.D., R.C. Brenner, R.E. Hinchee, C.M. Vo-
gel, and R.N. Miller. 1992. Optimizing bioventing in
1 shallow vadose zones and cold climates: Eielson
AFB bioremediation of a JP-4 spill. Symposium on
Bioremediation of Hazardous Wastes, May 5-6,
1992, Chicago, IL. EPA/600/R-92/126.
3. Ong, S.K., R.E. Hinchee, R. Hoeppel, and R.
Schultz. 1991. In situ respirometry for determining
aerobic degradation rates. In: R.E. Hinchee and R.F.
Olfenbuttel, eds., In Situ Bioreclamation. pp. 541-
545. Butterworth-Heinemann, Boston.
4. Miller, R.N., R.E. Hinchee, and C.M. Vogel. 1991. A
field-scale investigation of petroleum hydrocarbon
biodegradation in the vadose zone enhanced by soil
venting at Tyndall AFB, Florida. In: R.E. Hinchee and
R.F. Olfenbuttel, eds., In Situ Bioreclamation. pp.
283-302. Butterworth-Heinemann, Boston.
17
-------
Documenting Bioventing of Jet Fuel to Great Depths: A Field Study at
Hill Air Force Base, Utah
Gregory D. Saylss and Richard C. Brenner
U.S. Environmental Protection Agency, Risk Reduction Engineering Laboratory, Cincinnati, OH
Robert E. Hinchee
Battelle Laboratories, Columbus Division, Columbus, OH
Robert Elliott
Hill Air Force Base, UT
Bioventing is the process of supplying oxygen in situ to
oxygen-deprived soil microbes by forcing air through
unsaturated contaminated soil at low flow rates (1).
Unlike soil venting or soil vacuum extraction technolo-
gies, bioventing attempts to stimulate biodegradative
activity while minimizing stripping of volatile organics,
thus destroying the toxic compounds in the ground.
Bioventing technology is especially valuable for treating
contaminated soils in areas where structures and utili-
ties cannot be disturbed, because bioventing equipment
(air injection/withdrawal wells, air biower, and soil gas
monitoring wells) is relatively non-invasive.
Through the Bioremediation Field Initiative, a coopera-
tive program between EPA's Office of Research and
Development and Office of Solid Waste and Emergency
Response, EPA's Risk Reduction Engineering Labora-
tory began a 2-year field study of in situ bioventing in
the summer of 1991 in collaboration with the U.S. Air
Force at Hill Air Force Base (AFB) near Salt Lake City,
Utah. The site has JP-4 jet fuel-contaminated unsatu-
rated soil where a spill has occurred in association with
a fuel distribution network. The contractor operating the
project is Battelle Laboratories, Columbus, Ohio. With
the pilot-scale experience gained in these studies and
others, bioventing should be available in the near future
as a reliable, Inexpensive, and unobtrusive means of
treating large quantities of organically contaminated soils.
The objectives of this project are to increase our under-
standing of bioventing used on large volumes of soil
and to determine the influence of air flow rate on the
biodegradation and volatilization rates of the organic
contaminants.
* See previous report (2) for additional details.
Methodology and Results*
Site Description/Installation. The site is contaminated
with JP-4 from depths of approximately 35 ft to perched
water at roughly 95 ft Here, bioventing, if successful,
will stimulate biodegradation of the fuel plume under
roads, underground utilities, and buildings without dis-
turbing these structures. A plan view of the installation
is shown in Figure 1. The single air injection well, con-
tinuously screened from 30 to 95 ft below grade, is
indicated; "CW" wells are soil gas "cluster wells" where
independent soil gas samples can be taken at 10-ft
intervals from 10 to 90 ft deep; "SMP" wells are shallow
"soil gas monitoring points" at a depth of 2 ft below grade
that were installed recently (September 1992) to assist
in determining the volatilization rate. No data from these
wells are available at this time.
Air Injection. From August 1991 until October 1992, air
was injected into the vadose zone at a rate of about 70
ftVmin. Since October 1993, the injection rate has been
about 45 frVmin. Biodegradation and volatilization rates
as a function of air flow rate will be measured. An optimal
air injection that maximizes overall biodegradation rate
and minimizes total volatilization rate will be sought
Soil Gas Composition. Quarterly soil gas measurements
during venting are conducted. Soil gas oxygen, carbon
dioxide, and total hydrocarbons are measured at each
depth in all wells. The radius to which the air injection
provides oxygen, and thus the area in which biodegra-
dation is occurring, depends on the air flow rate. This is
demonstrated by the data shown in Figure 2 where soil
gas oxygen as a function of distance from the injection
well is plotted for the two flow rates, 70 and 45 tfVmin.
18
-------
• SMP ¦ Suffeca Monitoring Point
O CW ¦ Soy Vtpor Oust*r WM .
(t.5) *TPH In Ground Water (9/91) 11 1
A-A' - Cross Section Traca 50 met
Figure 1. Plan view of the Joint U.S. EPA and U.S. Air Force
bioventing activities at Hill AFB, near Salt Lake City,
Utah. CWs are cluster soil gaa monitoring wells;
SMPs are shallow soil gas monitoring points.
The oxygen levels at great distance are strongly influ-
enced by the injection rate.
In Situ Respiration Tests. For each flow rate used, an in
situ respirometry test (3) is conducted to evaluate the in
situ biodegradation rate. Rates are measured at each
soil gas monitoring location. Table 1 shows rates at three
well locations, averaged over depth, from tests con-
ducted in September 1991 and 1992. The air flow rate
was about 70 fflmin for both measurements. The lower
20
is
O 10
s
0
0
50
100
150
200
250
Distance from Air Injection (ft)
Figure 2. Soil gas oxygen concentrations as a function of dis-
tance from the Injection well for two air Injection
rates, 70 and 45 frVmin (cfm). Oxygen concentrations
are averages of the oxygen level at the 70 and 80 ft
depths In a particular well.
Table 1. Rates of Biodegradation, Averaged over the Depth
Noted, at Three Wells
Rate (mg/kg/day)
Wall
Depths (ft)
September
1991
September
1992
CW-1
20-90
0.97
0.30
CW-2
60-90
0.59
0.36
CW-3
10-90
0.56
0.32
rates after a year of venting suggest that bioventing is
removing petroleum hydrocarbons from the site at a
significant rate.
Inert Tracer Gas Study. An Inert gas tracer study was
conducted in November and December of 1993 to
evaluate how uniformly the injected air is distributed at
the site. Maintaining a constant flow rate, the air was
supplemented with helium, giving a helium feed compo-
sition of 5.6 percent. The data are still being analyzed.
Figure 3, however, shows the helium concentration at
well CW-5, 70 ft deep, as a function of time. Well CW-5
is 50 ft from the injection well. As noted in the figure, the
mean time for injected air to arrive to this sampling point
was 118 hr.
Soil Sampling. Final soil hydrocarbon analyses will be
conducted in summer 1993 and compared with the initial
soil analysis to document actual hydrocarbon loss due
to bioventing.
1.4
Median time » 118 hr
1.0
£ 0.8
0.4
0.2
0.0
0
50
100
150
200
250
Time (hr)
Figure 3. Helium concentration in soil gas during the Inert
tracer gas study at well C-5, 70 ft below the grade
sampling port, versus time after commencement of
injection.
19
-------
References
1. Hoeppel, R.E., R.E. Hinchee, and M.F. Arthur. 1991.
Bioventing soils contaminated with petroleum hydro-
carbons. J. Indust. Microbiol. 8:141.
2. Sayles, Q.D., R.C. Brenner, R.E. Hinchee, C.M. Vo-
gel, and R.N. Miller. 1992. Optimizing bioventing in
deep vadose zones and moderate climates: Hill AFB
bioremediation of a JP-4 spill. Symposium on Biore-
medlation of Hazardous Wastes, May 5-6,1992, Chi-
cago, IL. EPA/600/R-92/126.
3. Ong, S.K., R.E. Hinchee, R. Hoeppel, and R.
Schultz. 1991. In situ respirometry for determining
aerobic degradation rates. In: Hinchee, R.E., and
Olfenbuttel, R.F., eds., In Situ Bioreclamation. pp.
541-545. Butterworth-Heinemann, Boston.
»
20
-------
Application of Wood-Degrading Fungi to Treat Contaminated Soils
John A. Glaser
Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
Richard T. Lamar, Mark W. Davis, and Diane M. Dietrich
Forest Products Laboratory, U.S. Department of Agriculture, Madison, Wl
Past investigations of soil treatment systems using lig-
nin-degrading fungi have largely been confined to labo-
ratory or bench-scale studies (1,2). Recently, a project
consisting of two phases, a treatability study in 1991 and
a demonstration in 1992, was conducted at an aban-
doned wood-treating site in Mississippi to evaluate fun-
gal treatment effectiveness under field conditions. The
study site in Brookhaven, Mississippi, located 60 miles
south of Jackson, was identified as a removal action site
for EPA Region 4. While the wood-treating facility was
in operation, two process liquid lagoons were drained
and excavated^ The sludge was mounded above the
ground surface in a Resource Conservation and Recov-
ery Act (RCRA) hazardous waste treatment unit The
excavated material provided the contaminated soil for
both phases of the project The demonstration phase
was undertaken as a Superfund Innovative Technology
Evaluation (SITE) Program demonstration project.
The fungal treatment studies reported on in this paper
were conducted at Brookhaven because the site char-
acteristics were suitable for conducting field investiga-
tions, not to promote consideration of fungal treatment
as one of the treatment options for the site. Treatability
results are presented in this paper. Analysis of the dem-
onstration phase results was not complete at the time of
this writing and will be reported at a later date.
Methodology
The treatability study was designed to evaluate the abil-
ity of three different fungal species to degrade pen-
tachlorophenol (PCP) in soil. The soil pile was sampled
and analyzed for PCP and creosote components (i.e.,
polycyclic aromatic hydrocarbons [PAHs]) prior to devel-
oping the test site. Analysis of the laboratory results
identified sections of the pile with PCP concentrations
of less than 700 mg/kg. These sections were used to
supply the contaminated soil for the treatability study.
A test location was constructed on an uncontaminated
portion of the wood-treating site. The base for the test
plots was formed by changing the elevation with the
addition of clean soil to promote better drainage condi-
tions. Soil beds measuring 3mx3m(10ftx10ft) were
constructed of galvanized sheet metal. A leachate col-
lection system was installed to direct the liquid dis-
charge from all test plots to a central location for testing
and treatment After installation of the leachate system,
25 cm (10 in.) of clean sand was layered into each test
plot followed by a 25-cm (10 in.) lift of contaminated soil.
The contaminated soil was sized through a 2.5-cm (1 -
in.) mesh screen using a Read Screen All shaker screen
having a 8.4-m3/hr (10-yd3/hr) capacity. The soil was
deposited in separate piles on a polyethylene tarp. Fur-
ther homogenization was accomplished by the mixing of
different portions of screened soil. The mixed soil was
ttien applied to the treatment plots using a front-end
loader. Wood chips were added to the soil plots to
provide a substrate that could sustain growth of the
fungi.
Inoculum was developed jointly with tfie L.F. Lambert
Spawn Co. of Coatesville, Pennsylvania. The prepared
inoculum and inoculum carrier were shipped to the site
by refrigerated transportation. A total of 10 plots were
used in the study. The experimental design (Table 1)
consisted of a randomized complete block (RCB) with-
out replication and a balanced incomplete block (BIB)
with treatments replicated four times. Six of the plots
were allocated to the RCB design and four to the BIB
design. The BIB plots were subdivided into 1.5 m x
1.5 m (5 ft x 5 ft) subplots.
After inoculation with fungi, each plot was irrigated and
tilled with a garden rototiller. The tiller was cleaned as it
was moved between the plots to prevent cross contami-
nation between treatments and controls. Soil moisture
was monitored on a daily basis throughout the study and
21
-------
Table 1. Feasibility Study Experimental Design and PCP
Transformation Results'
PCP PCP
Trans* Initial
formed Cone.
Treatment* Inoculum (dry (mgfrg)
Regime Fungus/Control Loading wgt.)
1
P. chrysosporium
5%
15%
576
2
P. chrysosporium
10%
67%
1,017
3
P. sordida
10%
89%
673
4
P. chrysosporium/
T. hirsuta
5%
5%
23%
615
5
Inoculum Carrier
Control
10%
14%
687
6
No Treatment
Control
—
15%
737
7
T. hirsuta
10%
55%
334
8
P. chrysosporium
13%
52%
333
9
P. chrysosporium
10% (Day 0)
2% (Day 14)
55%
360
10
Wood Chip Control
—
0%
471
¦ The extent of transformation indicated in this table is developed
from the 42-day treatability study conducted from September 18 to
November 13,1991.
maintained at a minimum of 20 percent. Ambient and
soil plot temperatures were recorded throughout the
study daily. Plot tilling was scheduled on a weekly basis
for the duration of the study. A time series analysis of
treatment performance was accomplished by sampling
the plots before application of the treatments; immedi-
ately after treatment application; and then after 1, 2, 4,
and 8 weeks of operation.
Results
The treatability study was conducted over a 2-month
period between September 18 and November 13,1991.
The greatest removal of PCP (Table 1) was achieved in
the plot inoculated with P. sordida. Over the 42-day
study period, this treatment regime produced nearly 90
percent transformation of PCP from the contaminated
soil initially having a pH of 3.8 and a PCP concentration
of 673 mg/kg.
Removal data for the creosote constituents (PAHs) are
presented in Table 2 for the treatment using P. sordida.
Concentration decreases of the three- and four-ring
PAHs were consistently greater for the fungal treatment
than for the controls. Larger ring PAHs persisted in both
the treatment and control plots.
Table 2. Creosote Constituent Transformation Results*
% Decrease
Compound
PAH Initial
Cone,
(mg/kg)
No
Treatment
Plot
Carrier
Plot
P. sordida
Plot
AcBnapthene
429
49
68
95
Ruorene
225
75
57
95
Phenanthrene
941
69
49
90
Anthracene
684
57
48
85
Fluoran thane
972
23
42
72
Pyrene
572
10
22
52
Benzo[a]anthracsne
74
11
13
24
Chryssne
90
6
14
33
1 Initial concentrations were based on soil samples taken 1 day after
treatment application. Each value for initial concentration and per-
centage decrease Is the mean of 24 observations.
Summary and Conclusions
The superior performance of P. sordida in biotransform-
ing PCP parallels previous field study experience with
P. chrysosporium at a Wisconsin site. At Brookhaven,
however, P. sordida achieved the greatest percentage
removal. P sordida is an organism that resides in the
soil under normal conditions. The soil conditions en-
countered at the Brookhaven site are not optimal for the
cultivation of microorganisms, and the concentrations of
PCP found in the Brookhaven soil are excessive for
sustaining suitable growth of many microorganisms. All
of these factors strongly suggest that fungal treatment
using P. sordida has promise for processing PCP and
other difficult-to-degrade pollutants. Future challenges
to making the technology more cost effective are devel-
oping improved and cheaper inoculation techniques and
extending application of the technology to other organic
pollutant classes.
References
1. Lamar, R.T., J.A. Glaser, and T.K. Kirk. 1990. Fate of
pentachlorophenol (PCP) in sterile soils inoculated
with white-rot basidiomycete phanerochaete chryso-
sporium Mineralization, volatilization, and depletion
of PCP. Soil Biol. Biochem. 22: 433-440.
2. Lamar, R.T., and D. Dietrich. 1990. In-situ depletion
of pentachlorophenol from contaminated soil by
phanerochaete spp. App. Environ. Microbiol. 56:
3093-3100.
22
-------
Section 2
Performance Evaluation
In an, effort to evaluate the performance of various bioremediation technologies,
researchers assess the extent and rate of cleanup for particular bioremediation
methods. They also study the environmental fate and effects of compounds and -
their by-products, since remediation efforts at a contaminated site can produce
intermediate compounds that can themselves be hazardous. Thus, another impor-
tant aspect of performance evaluation projects involves assessing the risk of
potential health effects and identifying bioremediation approaches that best protect
public health.
Recent studies at an EPA test site demonstrated the potential applicability of using
indigenous phenol-utilizers in situ to degrade high concentrations of trichio-
roethyiene in an aquifer. Similarly, preliminary results from studies of a sand aquifer
at a Superfund site abutting Lake Michigan mostly support earlier evidence of
anaerobic transformation of chlorinated aliphatic compounds. Another study re-
viewed efforts to enhance the biodegradation capabilities of indigenous microor-
ganisms ttirough the introduction of an inoculant
In other developments, the stage has been set for a study that will examine
approaches for addressing limiting factors (e.g., co-contaminants such as heavy
metals, and oil and grease) in the anaerobic dechlorination of polychlorinated
biphenyis in soils and sediments. Another study is comparing both the effectiveness
and related impacts of bioventing and air sparging for remediating a site contami-
nated by an aviation gasoline spill.
Also, recently gathered data further demonstrate tfiat short-term bioassays, which
have been used widely as a tool to monitor especially complex environments, can
be used to evaluate, among other things, the loss and/or production of toxic
materials during bioremediation.
The symposium's poster session included results from experiments in which mice
were exposed to microbial products. The tests indicated that such exposures may
alter the balance of protective intestinal microbiota in mice.
23
-------
Field Evaluation of Phenol for Cometabolism of Chlorinated Solvents
Gary D. Hopkins, Lewis Semprini, and Perry L. McCarty
Western Region Hazardous Substance Research Center, Department of Civil Engineering Stanford University,
Stanford, CA
Aerobic microorganisms grown on phenol or toluene
can initiate the cometabolic oxidation of chlorinated ali-
phatic hydrocarbons (CAHs) to stable nontoxic end
products (1,2,3,4). Such microorganisms have good po-
tential for bioremediating aquifers contaminated with
CAHs and their anaerobic and abiotic transformation
products. Recent In situ studies at the Moffett Field test
site demonstrated that 80 percent of the trichlo-
roethylene (TCE) added could be degraded in the 2 m
biostimulated zone while injecting 12.5 mg/L phenol, 35
mg/L dissolved oxygen (DO), and 40 (ig/L TCE (6). The
objective of the current work is to determine how effec-
tive the phenol-utilizers were at degrading TCE over
higher concentrations ranging from 62.5 to 1,000 |ig/L.
Methodology
The methodology for the in situ tests was similar to that
used in our previous studies (5,6). The experiments
consisted of a series of stimulus-response tests; the
stimulus being the injection of ground water amended
with the chemicals of interest, and the response being
the concentration history of the chemicals at the moni-
toring locations. Experiments were performed under the
induced gradient conditions of injection and extraction.
In this study, TCE was injected along with phenol and
DO. Injection of TCE permitted the TCE concentration
effect to be studied in greater detail. Phenol was pulse
injected at 100 mg/L for 1 hr in an 8-hr pulse cycle,
resulting in a time averaged concentration of 12.5 mg/L
during the first 1,000 hr of the test The initial TCE
concentration was 62 (ig/L. The TCE concentration was
gradually raised by doubling the concentration after ef-
fective transformation had been achieved at the lower
concentration. After 1 week of operation, the injected
TCE concentration was raised to 125 ng/L, then to 250
|ig/L after another week, 500 |ig/L during the sub-
sequent week, and then 1,000 jig/L. The time averaged
phenol injection concentration was then raised to 19
mg/L for 1 week, and then increased again to 25 mg/L,
or twice the initial concentration, while maintaining a
TCE concentration at 1,000 ^ig/L. Bromide was added
as a conservative tracer as a basis of comparison for
determining transformation extents.
Results
Figure 1 shows the normalized concentration break-
through of TCE at the three observation wells SSE1,
SSE2, and SSE3, located 1, 2.2, and 3.8 m from the
injection well. The normalized concentration represents
the observed concentration divided by the injection con-
centration at the time of the observation. The smoothed
lines through the raw data are running averages. The
depiction of data in running averages is especially help-
ful for well SSE1 where the pulsed addition of phenol
induced competitive inhibition causing fluctuations in
TCE concentration. The damping effect of transport and
sorption combined with the lack of phenol results in
much smaller fluctuations at the SSE2 and SSE3 welis.
TCE Injection Concantration ((jg/L)
r? 0.6 1 T
400 600
Time (hr)
1000
Figure 1. Normalized breakthrough of TCE at observation wells
during biostimulatlon at 12.5 mgA. phenol. TCE Injec-
tion concentrations range from 68 to 1,000 (ig/L.
24
-------
ForTCE injection concentrations ranging from 62.5 |ig/L
to 500 jxg/L, the extent of biodegradation is similar de-
spite the increases in injection concentration. Bromide
tracer tests conducted over this period show complete
breakthrough of the conservative tracer, C/C0 = 1. Ap-
proximately 70 and 85 percent TCE removal was ob-
served at wells SSE1 and SSE2, respectively. Upon
increasing the TCE injection concentration to 1,000
jig/L, removals were lower.
Whiie injecting 1,000 jig/L TCE, the phenol concentra-
tion was increased to 19 mg/L at 1,008 hr and then to
25 mg/L at 1,176 hr. The results are illustrated in Figure
2. TCE removal increased following each phenol in-
crease. Since no controls are available, it is not possible
to conclude that the degree of improvement results from
phenol concentration increases or other factors such as
increased population of phenol-oxidizing bacteria.
Phenol concentration was measured in the injection
stream and at all monitoring locations. Phenol was fre-
quently detected at SSE1, but the concentrations found
were generally less than 0.5 mg/L while 12.5 mg/L was
injected. Detections at SSE2 and SSE3, however, were
infrequent. The measurement method at the field site
had a visible response to phenol concentration as low
as 10 |ag/L, but the quantifiable detection limit was about
25 iig/L. The phenol concentrations at the SSE2 and
SSE3 monitoring locations were generally too tow to
give a visible response, and thus, are presumed to be
below 20 |ig/L, and probably below 10 |ig/L Thus, phe-
nol removal was excellent and probably greater than
99.9 percent in this system.
The TCE results are summarized in Table 1. The per-
centage removals listed are based on average values at
the end of the period following the changes in concen-
tration. Removal up to 89 percent was observed at the
SSE3 well, with 12.5 mg/L phenol added. Removals
Plwnol Infection Concentration (mg/L)
12-Smg/L | 19myL
Table 1. Average Removal Efficiencies for TCE, at Various
Well Locations, and Transformation Yields
25nWl.
Phenol
Added
mg/L
TCE
Added
ngfl-
Percentage Removal
Transforma-
tion Yield
g TCE/g
phenol
SSE1
SSE2
SSE3
12.5
62
60
78
89
0.0044
12.5
125
66
S2
87
0.0087
12.5
250
70
82
88
0.018
12.5
500
68
84
68
0.035
12.5
1,000
78
85
90
0.082
19
1,000
75
32
85
0.045
25
1,000
78
85
90
0.036
800 900 1000 1100 1200 1300 1400
TVna (hr)
Figure 2. TCE concentration response resulting from Increased
phenol addition.
were similar with up to 500 |ig/L TCE injected. This
indicates the removal was first order with respect to TCE
concentration. Upon increasing the TCE concentration
to 1,000 |ig/L, the removals decreased to 77 percent
with 12.5 mg/L phenol injected. This lower percentage
removal may result because TCE concentration is
nearer the K, value resulting in deviations from first-or-
der kinetics, TCE transformation product toxicity is be-
ginning to have a measurable effect, or sufficient
reducing power from phenol oxidation is not available.
Transformation yields also are presented in Table 1. The
highest yield observed in the field was 0.062 g TCE/g
phenol. This yield was obtained while injecting 12.5
mg/L phenol and 1,000 ng/L TCE. Lower yields are
obtained at lower TCE concentrations or at higher phe-
nol injection concentrations.
Summary and Conclusions
An indigenous phenol-utilizing population effectively de-
graded TCE up to 1,000 jig/L. At the highest phenol
injection concentration, up to 90 percent of the 1,000
jig/L TCE added was degraded in a 3 m biostimulated
zone. With injection of 12.5 mg/L phenol, first-order
removal within the test zone of 85 to 90 percent was
obtained with TCE concentrations up to 500 |ig/L. A
phenol concentration of 25 mg/L was required to obtain
similar removal efficiency with a TCE injection con-
centration of 1,000 jig/ll. These results suggest in
situ biodegradation of TCE with phenol to be quite
promising.
Future studies at the site will explore a range of contami-
nants including vinyl chloride, chioroform, and 1,1-di-
chloroethyiene. The injection of a noncompetitive
source of reducing power, such as formate, and its effect
on removal efficiency will be explored. Studies with tolu-
ene as a growth substrate also will be performed for a
comparison with phenol.
25
-------
References
1. Nelson, M.J.K., S.O. Montgomery, E.J. O'Neill, and
RH. Pritchard. 1986. Aerobic metabolism of trichlo-
roethylene by a bacterial isolate. Appi. Environ. Mi-
" crobiol. 52:383-384.
2. Nelson, M.J.K., S.O Montgomery, W.R. Mahaffey,
and RH. Pritchard. 1987. Biodegradation of trichlo-
roethylene and involvement of an aromatic biodegra-
dative pathway. Appl. Environ. Microbiol.
53:949-954.
3. Nelson, M.J.K., S.O Montgomery, and RH Pritchard.
1988. Trichloroethylene metabolism by microorgan-
isms ttiat degrade aromatic compounds. Appl. Envi-
ron. Microbiol. 54: 604-606.
4. Wackett, L.R, and D.T. Gibson. 1988. Degradation
of trichloroethylene by toluene dioxygenase in
whole-cell studies with Pseudomonas putida F1.
Appl. Environ. Microbiol. 54:1703-1708.
5. Semprini, L, P.V. Roberts, G.D Hopkins, and P.L.
McCarty. 1992. Afield evaluation of in situ biodegra-
dation of chlorinated ethenes: Part 2, Results of
biostimulation and biotransformation experiments.
Ground Water 28:715-727.
6. Hopkins, G.D., L. Semprini, and P.L. McCarty. 1992.
Microcosm and in situ field studies of enhanced
biotransformation of trichloroethylene by phenol-util-
izing microorganisms. Submitted for publication.
26
-------
Innovative Bioremediation Strategies For Creosote:
Characterization and Use of Inocula
James G< Mueller and Suzanne E. Lantz
SBP Technologies, Inc., Gulf Breeze, FL
Jian-Er Lin
Technical Resources, Inc., Gulf Breeze, FL
P. Hap Pritchard
Gulf Breeze Environmental Research Laboratory, U.S. Environmental Protection Agency, Gulf Breeze, FL
New microorganisms possessing novel biodegradative
abilities continue to be discovered at a rapid rate. For
example, bacteria have been isolated ttiat use high-mo-
lecular-weight polycyclic aromatic hydrocarbons (HMW
PAHs), such as those found in creosote, coat tar, and
crude oil, as sole sources of carbon and energy for
growth (1,2). In addition, microorganisms such as My-
cobacterium sp. strain PYR-1 (Dr. Carl Cemiglia, U.S.
Food and Drug Administration [FDA] National Center for
Toxicoiogical Research, Jefferson, Arkansas), which co-
metabolize mixtures of HMW PAHs and structurally re-
lated compounds, have been described (3,4).
Concomitantly, our knowledge .of physicochemical fac-
tors limiting biological activity in the field (e.g., bioavail-
ability) has increased significantly, while innovative
methods for assessing and monitoring field performance
have been developed.
Successful combination of these technological ad-
vances into functional bioremediation strategies could
result in reliable, cost-efficient remedial tools for full-
scale remediation of soil and water contaminated by
compounds notoriously difficult to treat biologically. De-
spite the significant application potential of these newly
described microorganisms, however, the most common,
and clearly the most successful, uses of bioremediation
in the field remain focused on the treatment of materials
contaminated by readily biodegradable organics (i.e.,
refined petroleum products). One of the primary reasons
for this rather narrow application spectrum relates to the
general inability to effectively employ inoculant microor-
ganisms in the field. Thus, as the practical applicability
of bioremediation technologies are expanded to include
more persistent chemicals, such as those found in creo-
sote (i.e., HMW PAHs), new inoculation procedures and
implementation strategies need to be developed.
Expected advantages of augmented or otherwise im-
proved microbiology of contaminated environments in-
clude: 1) accelerated rates of biodegradation (e.g.,
shorter treatment time); 2) furthered extent of contami-
nant removal ("cleaner" end product); 3) reduced prob-
ability of undesirable "side-reactions" (e.g., production
of more toxic intermediates by indigenous microbes);
and 4) ensured consistency and reliability of biodegra-
dative performance in the field thus yielding more
cost-effective treatment technologies. The overall effec-
tiveness of microbial modification, however, depends
largely on the type of bioremediation technology being
employed. Hence, unique tools for modifying and/or
managing microbial systems are being pursued for a
variety of bioremediation application strategies, includ-
ing bioreactor operations, solid-phase (tandfarming),
and in situ applications (5,6).
A review of the scientific literature and experience in the
bioremediation industry offer the following scenarios
where, in theory, the use of inoculants might be war-
ranted. In the most straightforward scenario, the inability
of indigenous microflora to degrade or transform a per-
sistent chemical is known. In this case, the addition of
relevant catabolic machinery, in conjunction with requi-
site chemical resistances (e.g., heavy metals) or physi-
cal tolerances (e.g., low temperature), may be
accomplished via inoculation. Asecond scenario consid-
ers incidents of sudden exposure (e.g., oil spill) where it
is assumed that indigenous microflora have not had time
to adapt biochemically to a chemical insult. If such is the
case, then the addition of "trained" microbes may facili-
27
-------
tate a more rapid increase in degrader biomass, hence
potentially accelerating the biodegradation process.
Lastly, when indigenous microorganisms effect incom-
plete or partial catabolism, inoculants can be used to
provide alternative catabolic pathways. Of these possi-
bilities, it appears that the most valid use of inoculants
is to increase the rate of biodegradation of the targeted
substrates. Hence, the efforts of this study have focused
on the development of effective inoculation strategies to
amplify the number of rare, or relatively scarce, micro-
organisms with requisite resistances and/or tolerances.
Few reports describe the use of microbial inoculants to
positively influence the biodegradation of targeted com-
pounds in the field under solid-phase conditions, but
most studies have been conducted under laboratory
conditions. Alternatively, the most successful use of in-
oculants, beyond the laboratory- or bench-scale levels,
has been in bioreactor operations. Thus, the present
status of inoculation technologies is limited by a general
inability to employ them, in a consistent and reliable
fashion, to treat soil or ground water outside of a con-
trolled environment, such as a laboratory vessel or a
bioreactor in the field.
When inoculant strains are applied to soil or water in the
open environment rarely is a measurable effect clearly
associated with ttieir presence. Because inoculation is
often performed in conjunction with other efforts to mod-
ify the chemical and/or physical environment, the com-
parative data necessary to clearly elucidate the
response as directly attributable to the inoculant per se
are not available. From more of a microbiological per-
spective, other potential reasons, for this lack of a dis-
cernible response include: 1) inoculant microorganisms
may not successfully compete with indigenous mi-
croflora for essential micro-sites or ecological niches
essential for their survival; 2) inoculant bacteria are
particularly prone to predation; and 3) inoculants expe-
rience rapid die-off because of desiccation and other
physiological shocks (e.g., temperature changes, os-
motic changes).
In an effort to overcome factors known to inhibit inocu-
lation procedures for bioremediation applications, a va-
riety of microbial encapsulation and cell immobilization
technologies (7,8) similar to some of those previously
described (9,10) have been developed. Data from in-
house laboratory studies using bacteria active toward
HMW PAHs have shown that these technologies pro-
vide an effective means of inoculant storage and distri-
bution (Lin et al., unpublished data). Furthermore, newly
developed encapsulated cell technologies provide an
efficient means of applying high numbers of viable, cat-
abolically active inoculant cells (>1 x 1010 bacterial
cells/g inoculant) to soil and facilitating their timed re-
lease (Lin et al., unpublished data).
Encapsulation technologies currently are being used to
improve the solid-phase bioremediation of soils con-
taminated with organic wood preservatives (i.e., creo-
sote and pentachlorophenol). The importance of the
encapsulated cells for these applications is to: 1) ensure
the consistent presence of catabolically relevant, active
biomass; 2) provide for slow-release of essential nutri-
ents and electron acceptor; and 3) offer an ecological
niche conducive to microbial growth, proliferation, and
catabolism. Results to date have identified effective en-
capsulation and immobilization technologies for various
PAH- and pesticide-degrading bacteria. Depending on
the desired end points, tfiis strategy may offer a viable
approach for remediating soils contaminated with creo-
sote and similar substances. Similarly, immobilized cells
in liquid bioreactor systems (aboveground or in situ
bioreactors) are being tested for their ability to treat
ground water impacted by related compounds. In addi-
tion, research is being conducted to determine the ef-
fectiveness of co-encapsulating microorganisms (e.g.,
HMW PAH-degraders) with nutrients, electron ac-
ceptors, and/or electron donors.
Acknowledgments
Integrated in situ bioremediation systems for creosote,
and for HMW PAHs in general, have been developed in
collaboration with Drs. Eduard Alesi and Marc Sick (lEG,
Technologies, Inc. and GfS, Germany). Other in situ
designs have been developed in collaboration witti Drs.
John Cherry (University of "Waterloo-Canada) and Mal-
colm Shields (University of West Florida). Dr. Carl
Cerniglia, U.S. FDA National Center for Toxicological
Research, Jefferson, Arkansas, has provided several
HMW PAH-degraders for testing.
This work was performed as part of a Cooperative Re-
search and Development Agreement between the Gulf
Breeze Environmental Research Laboratory and SBP
Technologies, Inc. (Stone Mountain, GA) as defined
under the Federal Technology Transfer Act, 1986 (con-
tract no. FTTA-003).
References
1. Mueller J.G., P.J. Chapman, B.O. Blattmann, and
P.H. Pritchard. 1989. Action of a fluoranthene-utiliz-
ing bacterial community on polycyclic aromatic hy-
drocarbon components of creosote. Appl. Environ.
Microbiol. 55:3085-3090.
2. Mueller, J.G., P.J. Chapman, B.O. Blattmann, and
P.H. Pritchard. 1990. Isolation and characterization
of a fluoranthene-utilizing strain of Pseudomonas
paucimobilis. Appl. Environ. Microbiol. 56:1079-
1086.
3. Grosser, R.J., D. Warshawsky, and J.R. Vestal.
1991. Indigenous and enhanced mineralization of
28
-------
pyrene, benzo[a]pyrene and carbazole in soils.
Appl. Environ. Microbiol. 57:3462-3469.
4. Kelly, I., and C.E. Cerniglia. 1991. The metabolism
of fluoranthene by a species of Mycobacterium. J.
lnd. Microbiol. 7:19-26,
5. U.S. Patent Office. 1993. Microbial degradation of
trichloroethylene, dichloroethylenes and aromatic
pollutants. M.S. Shields et al. (Patent Pending).
6. U.S. Patent Office. 1992. Arrangement for cleaning
contaminated ground water. B. Bernhardt, U.S. Pat-
ent No. 5,143,606.
7. Lin, J-E., J.G. Mueller, and P.H. Prftchard. 1992.
Use of encapsulated microorganisms as inoculants
for bioremediation. Amer. Cham. Soc. Special Ses-
sion on Bioremediation of Soils and Sediments.
September 21-23, 1992, Atlanta, GA.
8. Lin, J.-E., H.Y. Wang, and R.F. Hickey. 1991. Use
of coimmobiiized biological systems to degrade
toxic organic compounds. Biotechnol. Bioengineer.
38.
9. European Patent Office. 1989. Encapsulation
Method. C.A. Baker, AA Brooks, R.Z. Greenley,
and J.M.S. Henis. Publication No. 0 320 463.
10. Stormo, K.E., and R.L. Crawford. 1992. Preparation
of encapsulated microbial cells for environmental
applications. Appl. Environ. Microbiol. 58:727-730.
29
-------
Decontamination of PCB-Contamlnated Sediments Through the Use of
Bloremedlatlon Technologies
John F. Quensen, III, Stephen A. Boyd, and James M. Tiedje
Michigan State University, East Lansing, Ml
John E. Rogers
Environmental Research Laboratory, U.S. Environmental Protection Agency, Athens, GA
Polychlorinated biphenyls (PCBs) can be reductively
dechlorinated in anaerobic environments such as sedi-
ments. This process was first suggested by PCB conge-
ner distribution patterns observed in sediment core
samples from the upper Hudson River (1). Compared to
both surficial sediments and the Arodor 1242 originally
input to these sediments, chromatograms for deeper
sediment samples showed a depletion of the more heav-
ily chlorinated, later eluting congeners and a corre-
sponding relative increase in lesser chlorinated, early
eluting congeners. This transformation process has
since been demonstrated to fee the result of anaerobic
microbial activity (2), and has been observed in PCB-con-
taminated sediments from a number of other sites (3,4).
The discovery of the anaerobic dechlorination of PCBs
has stimulated interest in developing a sequential an-
aerobic/aerobic biotreatment process for their destruc-
tion. Although aerobic degradation of PCBs is generally
limited to congeners with four or fewer chlorines, the
anaerobic process is able to dechlorinate more highly
substituted congeners, producing products that are
aerobically degradable. Indeed, ail products of the an-
aerobic dechlorination of Aroclor 1254 (5) have been
shown to be aerobically degradable by one or more
strains of aerobic bacteria (6). Also, the high proportions
of mono- and dichlorinated biphenyls that can accumu-
late as a result of anaerobic PCB dechlorination can
serve to induce PCB-degrading enzymes in aerobic
microorganisms. Although more highly chlorinated con-
geners can be aerobically cometabolized, they are not
inducing substrates.
To develop a sequential anaerobic/aerobic biotreatment
process for PCBs, more about the factors controlling the
anaerobic dechlorination process needs to be under-
stood. The extent of in situ PCB dechlorination that has
occurred varies among sites (Table 1). This variation is
likely attributable to both the congener specificity of the
predominant PCB-dechlorinating microorganisms pre-
sent and certain environmental characteristics. Recog-
nizing that environmental characteristics may have
changed since the time of in situ dechlorination, assays
were conducted to determine the ability of each sedi-
ment to currently support dechlorination. Sediments
from each site were slurried with reduced anaerobic
mineral medium and inoculated with Hudson River mi-
croorganisms, and Aroclor 1242 (500 pg/g sediment)
was added. The amount of chlorine removed from me
PCBs in 16 weeks was taken as a measure of the
sediment's current ability to support PCB dechlorination.
This measure correlated with the extent of in situ
dechlorination, except for the Lagoon (A and B), Silver
Lake, and River Raisin E sediments. Conditions at the
first three sites may have become unfavorable for
dechlorination after in situ dechlorination had already
begun. The exceptionally high total PCB concentration
in the River Raisin E sediments made the detection of
assay dechlorination difficult for analyt'jal reasons.
Correlations observed between the extent of assay
dechlorination for the other samples and several site
characteristics suggested that co-contaminants, notably
heavy metals and oil and grease, may be important
limiting factors (Table 1). The preliminary results of labo-
ratory experiments designed to better define the impact
of these factors on the dechlorination process are sum-
marized below. Research to be undertaken under a
cooperative agreement between Michigan State Univer-
sity and EPA will address ways potentially limiting fac-
tors, including heavy metals and oil and grease as
co-contaminants, can be overcome in a biotreatment
process.
Related Research
Oil and Grease
Oil and grease in sediments probably act as a separate
phase to which PCBs may partition (7). High concentra-
30
-------
Table 1. Extent of Dechlorination and Selected Environmental Parameters (or PCB-Contamlnated Sediments
Sediment
In situ
Dechlorination
Assay
Dechlorination
Total PCBs
(ug/g)
Oil A
Grease
(M8/9)"
Total Metals
(M/g)b
Acushnet 17
5
0.3
20
10,730
2,796
Acushnet 19
26
0.5
1,112
11,939
2,790
Hudson H7
74
1.6
74
626 '
1,143
Lagoon A
37
0
800
17,750
1,934
Lagoon B
38
0
1,113
38,338
3,352
Raisin D
9
0.8
88
7,157
3,919
Rami E (0-10 cm)
7
0
25,093
3,807
1,152
Raisin E (10-25 cm)
3
0
14,993
3,542
542
Saginaw 8R (10-20 cm)
7
0
323
13,092
5,427
Sheboygan R212
41
0.5
418
1,083
295
Sheboygan R28
63
1.1
395
955
292
Sheboygan R9
67
1.8
319
581
244
Silver L F3
32
0
694
62,856
9,135
* 01 and grease concentrations were determined gravimetrically and corrected tor PCB concentration.
b Total concentration of cadmium, chromiiOT, copper, nickal, lead, and zinc.
Note: In situ dechlorination is defined as the percent decrease in mats and para chlorines that has occurred naturally in
the environment Assay dechlorination is defined as the tig atoms of chlorine removed from PCBs in a 16-week laboratory
incubation after addition of Aroclor 1242 (500 jig/g sediment).
tions of oil and grease therefore might be expected to
limit dechlorination by decreasing the availability of the
PCBs to the microorganisms. This hypothesis is being
tested using vacuum pump oil and Aroclor 1242 added
to sediment slurries inoculated with Hudson River micro-
organisms.
A dose response between oil concentration and the
extent of dechlorination after 8 weeks of incubation has
been observed, but dechlorination still occurred even
with the addition of 40 mg oil/g sediment, the highest
level tested (Table 2). This highest concentration is com-
parable to the level of oil and grease in the most heavily
contaminated sediments examined. The extent of
dechlorination during the first 8 weeks of incubation is
still approximately one-half of ttiat when no oil is added.
Continued monitoring of this experiment for 24 weeks is
proposed, but the results so far suggest that oil as a
sorptive phase is not extremely important in preventing
PCB dechlorination.
Qualitative differences between oils in environmental
samples (e.g., additives and polycyclic aromatic hydro-
carbons content) might be important in inhibiting PCB
dechlorination. Vacuum pump oil was chosen for the
above experiment because it does not contain poten-
tially toxic or inhibitory additives. Oils found in environ-
mental samples also are being characterized and their
effect on PCB dechlorination is being determined.
Table 2. Extent of Aroclor 1242 Dechlorination
Avg. m It p Cl's
Treatment*
Mean
S.D.
0
0.85
0.11
5
1.10
0.06
10
1.16
0.02
20
1.26
0.04
40
1.30
o.cr
1242
1.8
* mg oil/g sediment
Note: Dechlorination is as Indicated by the average number of chlo-
rines remaining in the meta and/or para positions, after 8 weeks of
incubation by Hudson River microorganisms with various levels of
vacuum pump oil added to the non-PCB-contamlnated Hudson River
sediments used in the dechlorination assay. No dechlorination from
the ortho positions was indicated. Addition of 40 mg ofl/g sediment
reduced the extent of dechlorination by half.
Metals
Additional experiments are directly testing the effects of
the more abundant metals found in the contaminated
sediments examined. Various concentrations of metal
salts (5 to 500 g/g sediment of copper, lead, zinc, or
chromium) were added to pre-incubated anaerobic sedi-
ment slurries. The slurries then were autoclaved and
inoculated with Hudson River microorganisms, and Aro-
31
-------
clor 1242 was added. These experiments have been
sampled over 8 weeks and preliminary analysis sug-
gests that zinc is the most important metal for limiting
dechlorination. Monitoring PCB dechlorination in these
experiments will continue for 24 weeks.
In a related experiment Silver Lake sediments were
found not to support dechlorination after oil and grease
have been removed by solvent extraction, leaving the
metals behind. Additional experiments are being con-
ducted to ensure that the extraction process itself does
not adversely affect the ability of the sediments to sup-
port PCB dechlorination.
Future Research Objectives and Approach
i
The basic objective of this research to be undertaken
under a cooperative agreement between Michigan State
and EPA is to find ways of overcoming potential environ-
mental limitations on PCB dechlorination. The general
approach will be to select soils or sediments based on
the probable factor(s) limiting PCB dechlorination, and
then experimentally evaluate means of overcoming
those limiting factors. The initial focus will be on over-
coming limitations attributable to high levels of oil and
grease, low carbon content, low bioavailability, and high
redox potential (in soils). These limiting factors may be
overcome by aerobically biodegrading the oil and
grease, adding nutrients, using surfactants, and chemi-
cally or biologically providing more reducing conditions.
Subsequently, an attempt will be made to overcome
limitations attributable to metal toxicity by chelation, ex-
traction, precipitation, and bioleaching. Research aimed
at overcoming limitations relating to the congener speci-
ficity of the microorganisms involved currently is being
investigated in a separate project funded by General
Electric. Together, these efforts represent a comprehen-
sive evaluation of factors limiting PCB dechlorination
and approaches for overcoming these limitations.
References
1. Brown,* J.F., R.E. Wagner, D.L. Bedard, M.J. Bren-
nan, J.C. Carnahan, RJ. May, and T.J. Tofflemire.
1984. PCB transformations in upper Hudson sedi-
ments. Northeast. Environ. Sci. 3:167-179.
2. Quensen, J.F., III, J.M. Tledje, and S.A. Boyd.
1988. Reductive dechlorination of polychlorinated
biphenyls by anaerobic microorganisms from sedi-
ments. Science 242:752-754.
3. Brown, J.F, D.L. Bedard, M.J. Brennan, J.C. Carna-
han, H. Feng, and R.E. Wagner. 1987. Polychlori-
nated biphenyl dechlorination in aquatic sediments.
Science 236:709-712.
4. Brown, J.F., R,E. Wagner, H. Feng, D.L Bedard, M.J.
Brennan, J.C. Carnahan, and R.J. May. 1987. Envi-
ronmental dechlorination of PCBs. Environ. Toxicol.
Chem. 6:579-593.
5. Quensen, J.F., III, S.A. Boyd, and J.M. Tledje. 1990.
Dechlorination of four commercial polychlorinated
biphenyl mixtures (Aroclors) by anaerobic microor-
ganisms from sediments. Appl. Environ. Microbiol.
56:2360-2369.
6. Bedard, D.L, R.E. Wagner, MJ. Brennan, M.L.
Haberl, and J.F. Brown, Jr. 1987. Extensive degra-
dation of Arociors and environmentally transformed
polychlorinated biphenyls by Atcaligenes eutrophus
H850. Appl. Environ. Microbiol. 53:1094-1102.
7. Boyd, S.A. and S. Sun. 1990. Residual petroleum
and polychlorinated oils as sorptive phases for or-
ganic contaminants in soils. Environ. Sci. Techncl.
24:142-144.
32
-------
Natural Anaerobic Bioremedlatlon of TCE at the
St Joseph, Michigan, Superfund Site
Peter K. Kitanidis and Lewis Semprini
Western Region Hazardous Substance Research Center, Stanford University, Stanford, CA
Don H. Kampbell and John T. Wilson
Robert S, Kerr Environmental Research Laboratory, U.S. Environmental Protection Agency, Ada, OK
In a sand aquifer near the town of St. Joseph, Michigan,
trlchloroethylene (TCE), 1,2-cis-dichloroethylene (c-
DCE), 1,2-trans-dichloroethylene (t-DCE), and vinyl
chloride (VC) are present at ground water concentra-
tions from 10 to 100 mg/L. In the summer of 1991, a few
hundred ground water samples were collected along
three transects located near the areas of highest con-
centrations. The study was originally undertaken to
evaluate the potential of in situ treatment by stimulating
the growth of a native population of methanotrophic
bacteria. The data suggested that natural anaerobic
degradation of TCE has been taking place. An analysis
of these measurements was undertaken to gain a
greater understanding of the distribution of chlorinated
aliphatic compounds several years after the contamina-
tion and of the natural mixing and transformation proc-
esses. "Piis presentation summarizes the methodology
and the results of ttiis analysis.
The geologic formation of interest is an unconfined aqui-
fer consisting of a layer of unconsolidated fine sand with
some silt. The aquifer is nearly homogeneous, having
been formed by eolian sorting of glacial deposits. The
thickness of the sand layer Is variable, ranging from 12
to 32, as is the elevation of the base of the aquifer that
undulates over a range of about 16 m. The aquifer
overlays a lacustrine clay unit and at some points the
transition from sand to clay is gradual.
The site is bounded by Lake Michigan to the west and
Hickory Creek to the east. The hydrology of the sandy
aquifer is relatively simple and is dominated by a high
recharge rate. The ground water drains into the lake and
ttie creek, which'act as constant-head boundaries, at a
nearly steady rate. A hydrologic divide separates the
part of the aquifer that drains into the lake from the part
that drains into the creek. The aquifer contamination
probably took place from TCE that leaked from lagoons
located over the hydrologic divide. As a result, one
plume was formed that moves toward Lake Michigan and
one was formed that moves toward Hickory Creek (1).
Halogenated aliphatic compounds in the subsurface can
be transformed biologically in ttie absence of oxygen (2)
and, in fact, have been shown to do so under a variety
of environmental conditions. Common anaerobic elec-
tron acceptors and the associated microbial process, in
the order of their redox potential, are nitrate (denitrifica-
tion); Mn(IV) (manganese reduction); Fe(lll) (iron reduc-
tion); sulfate (sulfate reduction); and carbon dioxide
(methanogenesis). Many of the chlorinated aliphatic
compounds are highly oxidized and have redox poten-
tials above those of common electron acceptors (3).
Generally, the less halogenated components are trans-
formed more slowly; for example, VC is transformed at
a rate over two orders of magnitude lower than that of
TCE. The reduction reaction rates also are faster the
lower the redox state (3,4).
Fortuitous anaerobic transformation of chlorinated
ethenes, such as tetrachloroethylene (PCE) and TCE
involving substitution of chlorine by hydrogen, has been
observed in several field studies (1). The detailed chemi-
cal characterization of the St. Joseph site that will be
presented supports previous laboratory and field find-
ings. The characterization permits zones to be identified
where transformations are occurring, and permits flux
estimates of the contaminants and transformation prod-
ucts. The characterization is a joint effort of Allied-Signal
Corporation, Engineering Science, U-.S. EPA Region 5,
U.S. EPA Robert S. Kerr Environmental Research Labo-
ratory (RSKERL), and Stanford University.
Data Analysis
Three transects were completed with 17 slotted auger
borings. Transects 1 and 2 span the width of the plume
and Transect 3 is located roughly in the center of the
33
-------
plume. Onsite gas chromatography analysis was per-
formed for TCE and its anaerobic transformation prod-
ucts. The purpose of the onsite analysis was to guide
the field personnel in the selection of the next boring
location, so that the width of the plume and the center
of highest concentration could be found using a succes-
sion of boreholes. Samples for volatile organic com-
pounds (VOC) analysis were shipped on ice to RSKERL
for measurement of solute concentration using EPA
methods.
The data from the water samples were then statistically
analyzed to construct contour lines of equal concentra-
tion and to estimate flux rates. Contouring and averag-
ing require the interpolation from the data Of the
concentration of the solutes on a fine regular mesh.
Univariate statistical methods of data analysis, which
treat the data independently of their location in space,
are not applicable. For example, to compute the total
mass, assigning equal weights to all measurements
would not be reasonable because it is common to have
more measurements near the center of the plume than
elsewhere; instead, one should assign weights that are
representative of the area of influence of each measure-
ment and account for the shape of the concentration
surface. Linear geostatistical methods account for spa-
tial variability and are practical tools, but they are not
accurate when the distribution of estimation errors is not
described adequately by the mean and the mean square
value. Estimation errors, however, can vary over orders
of magnitude, depending on the proximity of measure-
ments to "hot spots." Also, they are highly skewed (i.e.,
asymmetrically distributed).
The methodology used produces point estimates (i.e.,
representative values), as well as confidence intervals
(i.e., error bars that indicate the range of possible val-
ues), which make it possible to evaluate the accuracy of
estimated concentrations and masses. The approach
accounts for the skewness in the distribution of estima-
tion errors by adding only one parameter to those used
in linear geostatistics (variograms or generalized covari-
ances). This parameter can be determined from the
data. The resulting nonlinear estimation method is not
substantially more difficult to use than linear geostatistics.
Preliminary Results
The detailed chemical characterization of the St. Joseph
site that will be presented will mostly support the pre-
vious laboratory and field findings about anaerobic
transformation of chlorinated aliphatic compounds. A
remarkable result from the transect data was that the
concentration of the chlorinated aliphatics compounds
was found to vary significantly with depth. This feature
was not predictable from the data previously collected
from nearby monitoring wells.
Perhaps the most interesting results are the concentra-
tion distributions obtained from the two-dimensional
contour analysis in the two transects that are roughly
perpendicular to the ground water flow. Complete trans-
formation of TCE to ethene was associated with
methanogenic conditions in the aquifer. VC and ethene
concentrations turned out to be correlated with the high-
est methane concentrations. Zones of high TCE concen-
tration tended to be associated with zones of depressed
methane concentrations. The cause for this absence of
methane where TCE concentration is the highest is not
evident It is possible that the high TCE values are
actively depressing methanogenesis. Another possibility
is that TCE remains in zones where active methano-
genesis is not occurring, for example, because of an
absence of appropriate bacteria, electron donors, or
redox concStions. c-DCE represents most of the DCE
present, with t-DCE and 1,1-dichloroethylene present in
small amounts. The spatial distributions of the DCE
isomers, however, are similar, indicating similar forma-
tion process(es). c-DCE was found at a slightly greater
depth tfian TCE, and VC and ethene were located even
deeper, and associated with high methane concentrations.
The concentration of sulfate and ammonia as inorganic
constituents showed some, correlation with the concen-
tration of the chlorinated aliphatic compounds. Sulfate
tended to be present at shallower depths and decreased
in concentration with depth as methane concentrations
increased. Ammonia also decreased with depth, as
methane concentration increased, possibly because of
the uptake associated with biological growth. The data
suggest that c-DCE was formed and tended to persist
in transition zones from sulfate reduction to methano-
genic conditions. Active VC formation and transforma-
tion to ethene occurred under methanogenic conditions
that were observed at greater depth.
Flux estimates were made based on the contour aver-
age concentrations and an estimated ground water ve-
locity of 25 m/yr. The total flux of chlorinated aliphatics
plus ethene was 200 and 490 kg/yr for transects 1
(downgradient) and 2 (upgradient) respectively. On a
mole basis, the fluxes differ by a factor of about 2.2.
c-DCE represents the greatest mole-flux in both tran-
sects followed by TCE, VC, and ethene. The fact that
c-DCE has a greater mole flux than its parent TCE
indicates that significant anaerobic transformation has
already taken place. Ethene represents approximately 8
to 22 percent of the total mole flux, indicating significant
dehalogenation to a nontoxic end product. A greater
amount of contaminant and transformation products
were moving across the upgradient location compared
to the downgradient location. The methane flux estimate
is a factor of four greater at transect 2 than at 1, sug-
gesting more highly anaerobic conditions upgradient.
The differences between the two transects, however,
34
-------
might be because of the transient nature of transport of
contaminants as they move toward Lake Michigan.
Estimates were made of the chemical oxygen demand
(COD) reduction required to dehalogenate the chlorin-
ated aliphatic compounds and produce the methane (1).
The COD reduction required was 1,180 kg/yr and 310
kg/yr for transects 2 and 1, respectively. Up to 15 per-
cent of the COD reduction was associated with reduc-
tion of chlorinated aliphatics, which is high considering
that these transformations are presumed fortuitous in
nature. The results suggest that it might be possible to
manipulate the conditions at the site to enhance anaero-
bic transformation of TCE to nontoxic end products.
References
1. McCarty, P.L, and J.T. Wilson. 1992. Natural anaero-
bic treatment of a TCE plume, St. Joseph, Michigan,
NPL Site. In: Bioremediation of Hazardous Wastes,
pp. 47-50. U.S. Environmental Protection Agency.
EPA/600/R-92/126.
2. Bouwer, E.J., B.E. Rittmann, and P.L. McCarty. 1981.
Anaerobic degradation of halogenated 1- and 2-
carbon organic compounds. Environ. Sci, Technol.
15(5):596-599.
3. Vogel, T.M., C.S. Criddle, and P.L. McCarty. 1987.
Transformations of halogenated aliphatic com-
pounds. Environ. Sci. Technol. 21:722-736.
4. Criddle, C.S., and P.L. McCarty. 1991. Electrolytic
mode! system for reductive dehalogenation in aque-
ous environments. Environ. Sci. Technol. 25:973-
978.
35
-------
Comparison of Bioventing and Air Sparging For In Situ Bioremediation of Fuels
Don H. Kampbell
U.S. Environmental Protection Agency, Ada, OK
Christopher J. Griffin
Solar Universal Technology, Inc., Traverse City, Ml
Frank A. Blaha -
U.S. Coast Guard, Cleveland, OH
Bioremediation pilot-scale subsurface venting and
sparging systems were operated at a low aeration rate
at an aviation gasoline spill site. Bioventing removed 99
percent of vadose zone contamination in 8 months with
minimal surface emissions. The biosparging process is
presently operating and has removed one-third of oily
phase residue below the water table in 1 year. The
ground water plume has been cleansed of benzene,
toluene, ethylbenzane, and xylene (BTEX) components
by sparging.
Introduction
The failure in 1969 of a buried transfer pipe flange within
an underground storage area of a U.S. Coast Guard Air
Station resulted in an aviation gasoline (Avgas) spill of
about 35,000 gallons. The Avgas migrated downward
and laterally to form a plume below a surface area of
260 ft wide and 1,200 ft long. The subsurface to 45 ft
was a fairly uniform beach sand with the present water
table near 15 ft Water table fluctuations over the years,
as much as 5 ft, formed an oily phase smear of Avgas
contamination with about 30 percent in the vadose zone,
just above the present water table.
Venting and sparging are in situ bioremediation proc-
esses that can provide air flow to vaporize and transport
volatile organic pollutants upward to more amenable
media for mineralization. Microbial degradation proc-
esses also use the oxygen provided by the air. Injection
wells can be installed at spill sites to emit air just above
the water table for venting and below the water table for
sparging. Laboratory soil microcosm studies using sur-
face soil from the site showed that once acclimation
occurred the degradation of Avgas vapor was rapid and
complete (1).
The objective of the study was to demonstrate by pilot-
scale process units that venting and sparging can effec-
tively bioremediate an aviation gasoline contaminated
subsurface. Since this report has not been subjected to
EPA review, however, official endorsement should not
be inferred.
Experimental Design
Turf was established on a 75 x 45 ft rectangular area
overlying the plume of contamination. An initial nutrient
solution of 64 lb nitrogen, 5 lb potassium, and 13 lb
phosphorus was applied for dispersion throughout the
unsaturated subsurface. Aeration injection wells
screened acrrss the water table were placed 10 ft apart
in a 3 x 5 ft grid. A blower rate of 5 cfm was used, which
was estimated to be equivalent to a subsurface air
retention time of 24 hours.
Vadose zone soil gas samples were obtained with 1/2-
in. diameter stainless steel tubing clusters set at depths
of 3.2, 6.5, 9.7, and 13.0 ft A portable Bacharach TLV
combustible gas meter was used to measure subsur-
face Avgas vapors. An inverted stainless bowl with a
nipple outlet at the top was used as a collection device
for surface emissions. Vented air was removed from the
bowl canopy at the same rate of entrance as determined
by a propane dilution test and passed through a car-
tridge trap. The trapped hydrocarbons were analyzed by
gas chromatography.
Vertical profile core samples were obtained using a
drilling rig with hollow stem augers and a piston barrel
sampler. Core samples were placed in glass jars and
capped immediately. Within 15 minutes, jars were un-
capped one at a time, and a plug aliquot was removed
and preserved for later analysis of total petroleum hy-
36
-------
drocarbons. Avgas vapors In the plug-evacuated space
were quickly measured using a real-time core assay
method (2). Ground water samples were collected from
installed monitoring wells and analyzed by EPA standard
methods.
One year after installation of the venting process, the
apparatus was revised into a sparging process. Eight
2-in. diameter polyvinyl chloride (PVC) sparge wells
were installed to a depth of 10 ft below the water table
in a 2 x 4 grid 10 ft apart A minimum radius of influence
of 15 ft was determined by pressure changes at dis-
tances away from a sparge well (3). A blower rate of 10
cfm was used.
Results
Venting operations started during October 1990. Gaso-
line hydrocarbon vapor in the vadose zone initially in-
creased to near 5,000 mg/L, then within 3 weeks
decreased 80 percent (Figure 1). A gradual decline con-
tinued until the operation was shut down in January
1991 because the turf root zone was frozen. Venting
was restored, 3 months later, after the soil had thawed.
Vapor concentrations increased to near 1,000 mg/L then
decreased rapidly. After 8 months of operation, soil gas
concentrations were 50 mg/L or less. Surface emissions
during mid-term operation of the venting system were
less than 10 jag/L, compared to 3.2 ft depth soil gas
levels of 163 mg/L. Bio mass degradation activity from
tfie turf rhizosphere was greater by a factor of 10 tfian
at the same depth of a barren soil control that was not
acclimated to gasoline vapor.
Vertical profile fuel carbons for October 1990 and Octo-
ber 1991 are shown in Figure 2. The water table was at
the same level of 15.3 ft for both time periods. A reduc-
tion of oily phase residue in the vadose zone occurred
Blov anting Blov en ting Blovontlng Replicate
Startup Restartup
Startup
10
0
6
4
2
0
0
e
Time (Months since 10/1/90)
12
10/1/90)
Figure 1. Soil gas hydrocarbons at 4 in depth for injection-only
plot
600
. - 9/90
• - 10/91
598 '
1
LL
GW Table
e
1
>
3
HI
596 ¦
594
3000
4000
1000
2000
0
Fuat Carbon, mg/Kg Cora Material
Figure 2. Vertical profile oHy phase residua in bioventlng of
north piot
in excess of 99 percent, while reduction below the water
table was only about 22 percent.
Sparge wells were installed and operation started in
November 1991. Initial gasoline vapor concentrations
exceeded 6,000 mg/L in the vadose zone (Figure 1),
then after 5 months sparging were near 1,000 mg/L.
After 1 year, the levels were 50 mg/L or less. Ground
water monitoring well samples were collected and ana-
lyzed after 7 months of sparge operation (Table 1).
Sparging effectively reduced contaminants dissolved in
the ground water. Vertical profile core samples were
collected and analyzed in October 1991 and September
1992. Heterogeneity obtained from vertical profile sam-
pling was partially compensated by averaging replicates
to obtain a trend shift in Avgas contamination. Real-time
core assays are shown in Table 2. Assay levels of control
profiles had a 4 percent i9dudion during the 12-month
period, while sparging reduction was 80 percent. Oily
phase residue as total fuel carbon was reduced 39
percent by sparging during the 12-month period (Table
3). The data suggested that inaccessibility of gasoline
globule contact with air flow had restricted vaporization
and transport upward.
Table 1. Ground Water Quality after Seven Months of
Blosparging
Total Fuel
Monitoring Well Depth Benzene Xylenes Carbon
Well :
-ft ng/L
Control
16
9.9
19
2880
17.5
228
992
4490
20.5
70
38
956
22
57
7.7
783
Sparge Plot
15
1.9
5.3
559
10
>1
5.0
>6
19.5
>1
>1
>6
21
1
>1
>6
37
-------
labia 2. Cora Assay of Avgas Vapor during Drilling
Operations
Ona Yaar
Initial Later Reduction
Rapllcata Average over 40 in.
Profile, mg Avgas/ft %
Control 4,070 (3 reps) 3,920 (2 reps) 4
Sparge Plot 5,980 (3 reps) 1,220 (4 reps) 80
Table 3. Total Fuel Carbon in Vertical Profile
One Year
Initial Later Reduction
Replicate Average, mg/ft2
Surface Area %
Sparge Pfot 139,540 (4 reps) 85.230 (5 reps) 39
Conclusion
The vadose zone contaminated with aviation gasoline
was satisfactorily bioremediated by venting in 8 months.
Surface emissions of gasoline vapor during process
operation were minimal. Sparging of the ground water
plume at a low aeration rate has reduced oily phase
residue by at least one-third in a 12-month period. The
sparging system has been operational for 17 months
and will continue.
References
1. Kampbell, D.H., and J.T. Wilson. 1991. Bioventing to
treat fuel spills from underground storage tanks. J.
Haz. Materials 28:75-80.
2. Kampbell, D.H., and M.L. Cook. 1992. Core assay
method for fuel contamination during drilling opera-
tions. Proceedings of Subsurface Restoration Con-
ference, Dallas, TX, pp. 139.
3. Griffin, C.J., J.M. Armstrong, and R.H. Douglass.
1991. Engineering design aspects of an in-situ soil
vapor remediation sparging system. Proceedings of
1991 International Conference on On-Site and In-
Situ Bioremediation, San Diego, CA, pp. 517.
38
-------
Determining the Genotoxicity of Bioremediation Products
Larry D. Ciaxton, S. Elizabeth George, Robert W. Chadwick, and Virginia S. Houk
Health Effects Research Laboratory, U.S. Environmental Protection Agency,
Research Triangle Park, NC
LE. Rudd and J.J. Perry
North Carolina State University, Raleigh, NC
John T. Wilson
Robert S. Kerr Environmental Research Laboratory, U.S. Environmental Protection Agency, Ada, OK
The presence of toxicants in the environment can be
determined either by direct analytical chemistry meas-
urements of known toxic substances or by using bioas-
says. The wide variety of man-jnade and naturally
occurring toxicants in the environment, including car-
cinogens and mutagens, is well documented. Carcino-
gens (substances that initiate the cancer process) and
mutagens (substances that cause heritable effects in
ceils and organisms) cause "delayed effects." That is,
the response to exposures of genotoxic carcinogens
and mutagens (collectively called genotoxicants) may
not be seen for many years. In contrast to other types
of toxic interactions, the action of many genotoxicants is
not limited theoretically by threshold limits. In other
words, any measurable amount of the substance to
which an organism is exposed can potentially cause an
adverse effect The development of methods to detect,
measure, and evaluate environmental genotoxicants,
therefore, has been one of the primary focuses of the
EPA's Health Effects Research Laboratory (HERL). Be-
cause environmental samples often contain mixtures of
both identified and unidentified genotoxicants, a need to
evaluate environmental samples using bioassays ex-
ists. Potential health effects can be assessed through
the bioassay of samples brought to the laboratory from
the environment of interest or through in situ bioassays
using either indigenous species or introduced indicator
organisms. One potential advantage of biological moni-
toring is that the lack of hazard identification by bioassay
can lower the level of concern and perhaps help to avoid
costly analytical studies. If major toxicants are known to
be within the complex mixtures, however, chemical
measurements may provide the more cost-effective and
rapid approach.
The development of short-term bioassays for genotoxic-
ity in the 1970s led to the use of these assays for
evaluating complex environmental mixtures. The ge-
netic assay that has demonstrated the greatest utility is
the Salmonella typhimurium mutagenicity assay. Many
genetic assay systems have now been applied to under-
standing the toxicity of complex environmental situ-
ations. Using some of the studies done at HERL with the
salmonella assay (including the semiautomated spiral
assay) and the prophage induction assay, this paper
illustrates how genetic bioassays can be used to under-
stand the effect of bioremediation on genotoxicity.
Laboratory-Scale Studies Examining
Bioremediation Products
In order to evaluate the potential for both the degrada-
tion and production of genotoxicants during the biore-
mediation of crude oils, biodegradation products of three
crude oils metabolized by one of two species of fungi
were monitored by analytical chemistry methods and by
use of ttie spiral salmonella assay. The two fungi used,
Cunninghamella elegans and Penicillium zonatum,
grow with crude oil as a single carbon source. The three
crude oils were chosen to represent complex environ-
mental substances of no, moderate, and high mu-
tagenicity. Oils were incubated with respective fungi for
16 days, and aliquots taken at specified time points were
bioassayed for mutagenicity. When the most mutagenic
crude oil was degraded by either fungi, its mutagenicity
was significantly decreased. The mutagenicity of the
moderately mutagenic oil did not change significantly
during the 16-day period. The nonmutagenic oil (from
Cook Inlet, Alaska), however, showed a mutagenic re-
sponse after degradation. In ail cases, "weathered" con-
39
-------
trol samples (incubated without the fungus), showed
little to no change in mutagenicity.
This laboratory experiment illustrates the type of com-
plexity inherent in understanding the toxicity of bioreme-
diated substances and the usefulness of bioassays.
Most of the naturally occurring mutagens within the
mutagenic oils have not been identified. Likewise, the
mutagens produced or concentrated in the nonmu-
tagenic oil during fungal degradation have not been
identified. Chemical analysis would not be expected to
provide an appropriate approximation of relative geno-
toxicity for these samples. Bioassays, however, shown
to screen for the destruction and creation of genotoxi-
cants, are known to be useful in helping to choose
treatment alternatives, and can (on some occasions) be
used to evaluate relative toxicity.
Evaluation of Contaminated Sites
Spills of crude and processed oils and fuels are among
the most common. In a recently initiated study, soil
samples were taken at a Wichita, Kansas, location that
had experienced an accidental pipeline leak of a refined
hydrocarbon product Samples were taken from "highly"
and "moderately" contaminated sites as well as from a
"clean" site. The core samples were divided into four
sections (two above the water table and two below) for
extraction and testing using the salmonella mutagenicity
assay. Mutagenicity was demonstrated for all three
sites, both above and below the water table. After biore-
mediation of the site, core samples will be taken so that
mutagenicity before and after remediation efforts can be
compared.
Evaluating the Effects of Multiple
Pollutants
2,6-Dinitrotoluene (DNT) represents a class of com-
pounds that can be found in hazardous waste sites
contaminated with munition chemicals. The organo-
chiorine pentachlorophenol (PCP) is a wood preserv-
ative found within many sites. The activation of DNT to
genotoxic metabolites involves enzymes in both the liver
and the intestinal flora. PCP both has bactericidal activ-
ity and can induce hepatic mixed function oxidase enzy-
matic activity in the liver. Studies were done therefore to
determine the effect of PCP on the production of mu-
tagenic metabolites from DNT. Results of the study indi-
cated that PCP accelerates the biotransformation of
DNT to genotoxic metabolites and potentiates the for-
mation of DNT-induced DNA adducts in the liver.
These types of studies, which typically would be done
only for the most commonly co-located pollutants, will
help risk managers to understand which unique combi-
nations of pollutants must be closely monitored.
Summary
Although bioassays have been used for many years to
evaluate ecotoxicity in complex environmental.situ-
ations, only in the last decade have major efforts using
short-term bioassay for potential human health effects
been applied to actual environmental situations. These
studies illustrate that short-term bioassays can be ap-
plied to bioremediatiori research to identify toxic pollut-
ants, the loss and/or production of toxic materials during
bioremediation, the relative toxicity of contaminants
treated by different methodologies, and methods of
identifying and evaluating how mixtures of chemicals
affect toxicological processes.
40
-------
Modulation Effects of Biotechnology Microbial Agents Following Pulmonary
Exposure of Mice
S. Elizabeth George, Michael J. Kohan, John P. Creason, and Larry D. Claxton
Health Effects Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC
Endotoxin-resistant C3H/HeJ and endotoxin-sensitive
CD-1 mice were challenged intranasally with environ-
mental and clinical pseudomonads for evaluation of
modulation effects (alteration of intestinal microbiota,
translocation). At time intervals, the intestinal tract,
spleen, liver, and mesenteric lymph nodes (MLN) were
removed, homogenized, and plated for the enumeration
of the dosed strain and/or resident microbiota. Environ-
mental Pseudomonas maltophilia strain BC6, and P.
aeruginosa strains BC16, BC18, and AC869, as well as
clinical P. aeruginosa isolates DG1 and PA01, were
detectable in the intestinal tract 14 days following treat-
ment In CD-1 mice, P. cepacia strain AC1100 impacted
the small intestinal anaerobic count, lactobaciili, and
obligately anaerobic Gram-negative rods. Both strains
AC869 and AC1100 modified the cecal obligately an-
aerobic Gram-negative rods and lactobaciili during the
course of the experiment Strain BC6 had an overall
treatment effect on the cecal lactose negative enteric
rods in C3H/HeJ mice. Strain BC18 increased the small
intestinal lactose negative enteric rods, DG1 elevated
the cecal lactose fermenting enteric rods, and BC16
increased the cecal anaerobic count. Strains BC16,
AC869, and PA01 translocated to the MLN, spleen, and
liver during the experimental time. Strain BC18 was
detectable in the liver at 3 hours following treatment, and
strains BC17 and DG1 were recovered from the MLN
and liver. No translocation was observed for strain BC6.
Therefore, pulmonary treatment of mice with P. aerugi-
nosa, P. maltophilia, or P. cepacia may alter the balance
of the protective intestinal microbiota, which may cause
further negative health effects such as multiplication of
harbored pathogens, invasion by opportunistic patho-
gens, or translocation of bacteria to other organs.
41
-------
Section 3
Field Research
Field research is essential for evaluating the performance of full-scale bioremedia-
tion processes and for conducting testing on technologies that are appropriate for
scaled-up application. For example, problems associated with the use of bacteria
used in the laboratory include optimizing the activity of the organism under site
conditions and defining the risks associated with the introduction of a non-native
microorganism. The objective of this level of research is to demonstrate that the
particular bioremediation process performs as expected in the field. Researchers
at the symposium provided results from field experiments both as papers and
poster presentations.
Having demonstrated the ability of a nonrecombinant bacterium (Pseudomonas
cepacia) to degrade trichloroethylene (TCE) in a vapor phase bioreactor, re-
searchers currently are seeking an understanding of the microbial parameters
affected by the reactor's design and operation. Ultimately, researchers hope to
extend the range of substances degraded by the microorganism, since TGE in
aquifers is often found with other pollutants. In anticipation of an in situ bioreme-
diation experiment a related study is examining the likely fate of the introduced
organism and its effect on indigenous microbial populations.
In research on inoculation of contaminated soils, the utility of three encapsulation
technologies as means of storing, distributing, and introducing microorganisms to
a site was studied. In another effort to apply laboratory knowledge at the field-scale
level, researchers used a rational design approach tc measure the relative resi-
dence time and concentration of electron acceptor at a Kansas site contaminated
with refined petroleum.
At a different petroleum-spill site, the natural anaerobic bioremediation of toluene,
ethylbenzene, and xylene isomers was found to compare favorably with their transfor-
mation in microcosm stucfies. Both field monitoring and laboratory studies, however,
indicated benzene is not biotransformed under the particular site's conditions.
The symposium's poster session included information on a plan to field test
systems for bioremediating vadose-zone soils contaminated with jet fuel. Re-
searchers tentatively plan to test the systems, which involve hydrofracturing and
biofiltration, at an Air Force base in Ohio.
Preceding page blank
-------
Field Demonstration of a Constitutive TCE-Degrading Bacterium for the
Bioremediation of TCE
Malcolm S. Shields, Michael Reagin, Robert Gerger, Rhonda Schaubhut, arid Robert Campbell
Center for Environmental Diagnostics and Bioremediation, Department of Biology, University of West Florida,
Pensacola, FL
Charles Somerville
Technical Resources Inc., Gulf Breeze, FL
R Hap Pritchard
Gulf Breeze Environmental Research Laboratory, U.S. Environmental Protection Agency, Gulf Breeze, FL
The degree to which trichloroethyiene (TCE) has teen
recognized as a significant environmental pollutant is
reflected by the amount of research into methods for its
remediation. Despite the demonstrated environmental
hazards, its industrial use continues apace because few
alternatives exist. TCE owes its environmental behavior
partly to its physical properties (i.e., high density and
water solubility and low chemical reactivity), and partly
to its biological recalcitrance. Both contribute to its no-
toriety as a persistent point source pollutant, despite
numerous reports of both anaerobic and aerobic bacte-
rial transformation capabilities. Aerobic bacteria are
more rapid TCE metabolizers, but only in a cooxidative
fashion. TCE serves as a cooxidative substrate for vari-
ous bacterial oxygenases, but not as an inducer of them.
These bacteria require co-inducers that include toluene
(7,9,10,11), phenol (5,8,10), methane (4,6), ammonia
(1), isoprene (2), 2,4-dichlorophenoxyacetic acid (2,4-
D)(5), and propane (15).
Background
Our research has centered on the miocrobioiogy of P.
cepacia G4, which expresses a unique toluene ortho-
monooxygenase (Tom) in response to various aromatic
inducers. Tom carries out the cooxidative metabolism
of TCE by this strain (12,13). In this study, a non-
recombinant derivative of G4, called G4 PR1 has been
developed that constitutively expresses Tom, and con-
sequently degrades TCE without the need for a co-in-
ducer (14). This paper deals with the characterization,
alteration, and application of this constitutive derivative.
Results and Conclusions
Bloreactor
A fixed film bioreactor was investigated for the exploita-
tion of PR1 ,for the degradation of air-entrained TCE. A
reactor that would receive air-entrained TCE along with
a continuous inflow of nutrient medium (trickle feed) was
constructed. Several packing materials were tested for
their ability to support growth of a PR 1 biofiim capable
of TCE degradation: cellfte pellets (diatomaceous earth
pellets manufactured by Manville Corp.), gravel, glass
beads, activated charcoal, sari J, and crushed oyster
shell. The material exhibiting the greatest degree of TCE
removal per gram of colonized material was crushed
oyster shell. In addition to performance criteria, the cost
and buffering capacity (calcium carbonate) of the oyster
shell matrix contributed to its selection for further biore-
actor research.
All column designs tested thus far involved the metered
application of nutrients (yeast extract peptone, and glu-
cose or minimal medium with lactic acid, phthalate,
starch, or phenol) to a glass or stainless steel column,
packed with oyster shell and colonized with PR1. Air-en-
trained TCE is passed through the column (bottom to
top). Strain G4 (the nonconstitutive parent of PR1) had
no effect on TCE in a closed bioreactor operated with a
100 percent atmospheric refeed, and no additional nu-
trient was input beyond initial colonization. PR1, present
at £1.4 x 108 bacteria/g of the support material, com-
pletely removed the 80 |aM TCE (10 mg/L air) present in
the gas phase under the same conditions, over a 20-hr.
period.
44
-------
A 3-L column assembled in this manner, but allowing the
continuous application of nutrients and air with TCE,
was capable of achieving the complete removal of 7 to
10 [iM of air-entrained TCE introduced at a rate of 4
mL/min over a 96-hr period.
A similar design of a 26-liter column reactor (Figure 1)
allowed for nutrient (lactate) liquid recycle and a continu-
ous input of freshly grown cells from a lactate-fed che-
mostat during the inoculation phase. Two separate tests
of this reactor system at 0 percent refeed were con-
ducted. One resulted in a 92 percent removal of TCE at
an average input concentration of 12.6 |iM (2 mg/L) in
the air phase over a 72-hr test
rrfflq
d)
400|lLtoW
26 Liter Capacity
-Glaa* Column
-Tdton
-Stainless Steel
-0.1 nM
13 ppb
out <
TCE/Air
Recycle
1 l/min
400tiL/mln
Mass Balance
Carbon In: 6.4 mjfrnin
02 in: 82 mgihiin
TCE In: 13 ug'rnin Nutrients
3
S
Atmosprtera
Recycle
1 Umin
» 10
HL"
W
(2)
58% O,
ZuMTCE
lOOmUmin
TCE
-1 nM
131 ppb
400
3uM
Spent Nutrients
16 g Lactata/L
0.36Xbg
CH3CHOHCOOH
Figure 1. Schematic of 26-llter column reactor.
In another test, TCE was introduced at appreciably
lower levels and with a significant nutrient/liquid re-
cycle. After a 48-hr inoculation period, air-entrained TCE
(-1nM [131 jigrtJ) was introduced at 100 mUmin total
flow rate (c.a. 58 percent oxygen). Under these condi-
tions, approximately 90 percent of the TCE present in
the gas phase was continuously removed during a 4-day
test period (Figure 2).
Tom expression in liquid culture or as a biofilm attached
to an oyster shell surface may be monitored directly by
the rate of TCE degradation, or more indirectly by the
rate of trifluoromethyl-phenol (TFMP) conversion to the
50
Time (hrs)
Figurt! 2. IVventy-six liter Wofliter containing crushed oyster
shell and colonized with Pseudomonas cepacia G4
PR1. Air-entrained TCE was delivered from a com-
pressed gas cylinder and mixed with oxygen to bal-
ance nutrient requirements.
nonmetabolizable trifluoroheptadienoic acid (TFHA) (a
reaction requiring both Tom and catechol-2,3-dioxy-
genase [C230]). The 8 nmol TFHA/min/mg protein by
uninduced PR1 was similar to the 6.5 nmol/min/mg
protein produced by phenol induced G4. Likewise an -1
nmol/min/mg protein rate of uninduced TCE removal by
PR1 grown in batch culture is approximately one-fourth
of the average daily TCE degradation rate attained with
phenol induced (chemostat grown) G4 (3).
Nonrecombinant Expansion of Cataboiic
Range and Resistance to Toxic Intermediates
TCE in aquifers is found frequently in combination with
other pollutants. Chloro-aromatics like chlorobenzene
and chlorophenol are of considerable concern because
of their conversion via Tom to toxic chlorocatechols,
which are not metabolized by C230. PR1 metabolizes
chlorobenzene to 2-chlorophenol, and 2-chlorophenol to
3-chlorocatechol, but no further. In an effort to extend
the range of pollutants degraded by PR1 and offer some
protection to metabolic suicide with chloroaromatics,
plasmid pRO101 (a tetracycline derivative of pJP4,
specifying the degradation of 2,4-D) was introduced into
PR1. In so doing, the transconjugant PR1 (pRO101)
was created that expresses enzymes for ortrto-fission
(i.e., catechol-1,2-dioxygenase [C120]) of the resultant
3-chlorocatechol, in addition to the nonproductive meta-
cleavage mechanism of the C230 of PR1, which will
accept 3-chlorocatechol as a substrate, but in so doing
inactivates itself. In PR1 this results in the buildup of
chlorocatechol that is highly toxic and consequently in-
45
-------
hibitory to cellular metabolism, including its ability to
degrade TCE. PR1 (pR0101) is capable of growth on
2,4-D, chlorobenzene, or 2-chlorophenol as the sole
carbon sources and also is effective in degrading TCE
at approximately 20 jim, following a 24-hr pre-exposure
to 2-chlorophenol at 1 mM. PR1 is not able to use these
carbon sources, nor is it able to degrade TCE following
exposure to these levels of chlorobenzene or 2-chlo-
rophenol. Thus, the established substrate range of PR1
(pRO101) includes 10 aromatic substrates and 5
aliphatics.
Conclusions
.The use of a particular P. cepacia strain constitutive for
the production of Tom was investigated in a vapor phase
bioreactor. Though several toluene oxygenases capable
of TCE oxidation have been cloned to E. coli and thus
no longer require aromatic induction (7,16,18, and our
torn A, torn B clone pMS64), this use is the first reported
bioreactor application of a nonrecombinant bacterium
(i.e., PR1) capable of doing so. The capacity of this
organism to cause the degradation of TCE in a vapor
phase bioreactor under laboratory conditions has been
demonstrated at TCE concentrations well above those
expected from field air strippers. Current efforts are
directed at understanding the microbiological parame-
ters affected by the reactor design and operation. Future
considerations will include the organisms ability to me-
tabolize TCE in an in situ bioreactor.
The substrate range of Tom has been established to
include TCE, vinyl chloride, cis- and f/ans-1,2-dichlo-
roethylene, 1,1-dichloroethylene, benzene, toluene,
phenol, orffto-xylene, and ortho-, meta-, and para-ere-
sol. P. cepacia PR1 carrying pRO101 is capable of using
these same substrates in addition to 2,4-D, chloroben-
zene, and 2-chlorophenol. Naphthalene has been
shown to undergo a single hydroxylation to alpha-
naphthol. The efficacy for their treatment in a bioreactor
or in situ system remains to be demonstrated. The intro-
duction of pROlOl into PR1 resulted in expansion of the
substrate range of this organism to include chloroben-
zene and 2-chlorophenol (determination of the utilization
of other chloroaromatics is pending) and an increase in
its ability to degrade TCE during prolonged exposure to
these chloroaromatics.
References
1. Arciero, D.T., M. Vannelli, M. Logan, and A.B.
Hooper. 1989. Biochem. Biophys. Res. Commun.
159:640-643.
2. Ewers, J., D. Freier-Schroder, and H-J. Knack-
muss. 1990. Arch. Microbiol. 154:410-413.
3. Folsom, B.R., and P.J. Chapman. 1991. Appl. En-
viron. Microbiol. 57:1602-1608.
4. Fox, B.G., J.G. Borneman, LP. Wackett, and J.D.
Lipscomb. 1990. Biochemistry 29:6419-6427.
5. Harker, A.R., and Y. Kim. 1990. Appl. Environ. Mi-
crobiol. 56:1179-1181.
6. Henry, S.M., and D. Grbid-Galid. 1991. Appl. Envi-
ron. Microbiol. 57236-244.
7. Kaphammer, B.J., J.J. Kukor, and R.H. Olsen.
1990. Abstr K-145, p. 243. Abstr. 90th Annu. Meet.
A. Soc. Microbiol.
8. Montgomery, S.O., M.S. Shields, P.J. Chapman,
and P.H. Pritchard. 1989. Abstr. K-68, p. 256. Abstr.
89th Annu. Meet. Am. Soc. Microbiol.
9. Nelson, M.J.K., S.O. Montgomery, and P.H.
Pritchard. 1988. Appl. Environ. Microbiol. 54:604-
606.
10. Nelson, M.J.K., S.O. Montgomery, E.J. O'Neill, and
P.H. Pritchard. 1986. Appl. Environ. Microbiol.
52:383-384.
11. Nelson, M.J.K., S.O. Montgomery, W.R. Mahaffey,
and P.H. Pritchard. 1987. Appl. Environ. Microbiol.
53:949-954.
12. Shields, M.S., S.O. Montgomery, S.M. Cuskey, P.J.
Chapman, and P.H. Pritchard. 1991. Appl. Environ.
Microbiol. 57:1935-1941.
13. Shields, M.S., S.O. Montgomery, P.J. Chapman,
S.M. Cuskey, and P.H. Pritchard. 1989. Appl. Envi-
ron. Microbiol. 55:1624-1629.
14. Shields, M.S., and M.R. Reagin. 1992. Appl. Envi-
ron. Microbiol.
15. Wackett, L.P., GA Brusseau, S.R. Householder,
and R.S. Hanson. 1989. Appl. Environ. Microbiol.
55:2960-2964.
16. Winter, R.B., K.-M. Yen, and B.D. Ensley. 1989.
Bio/Technology 7:282-285.
17. Worsey, M J., and P.A. Williams. 1975. J. Bacteriol.
124:7-13.
18. Zylstra, GJ., L.P. Wackett, and D.T. Gibson. 1989.
Appl. Environ. Microbiol. 55:3162-3166.
46
-------
Bioremediation of TCE: Monitoring the Fate and Effects of a Microorganism Used
In a Field Bioaugmentatlon Study
M.S. Shields, R. Snyder, M. Fteagin, R. Gerger, and R. Campbell
Center for Environmental Diagnostics and Bioremediation, Department of Biology, University of West Florida,
Pensacola, FL
C. Somerviile
Technical Resources Inc., Gulf Breeze, FL
P. Hap Pritchard
Gulf Breeze Environmental Research Laboratory, U.S. Environmental Protection Agency, Gulf Breeze, FL
The development of a constitutive trichloroethylene
(TCE)-degrading Pseudomonas cepacia provides us
with a unique opportunity to study several microbiologi-
cal aspects of bioremediation that many believe to be
altogether overlooked. Systems for the utilization of
such organisms range from contained above-ground
bioreactors to more passive in situ processes. Implicit in
the understanding of this organism's overall effective-
ness in biodegrading a target pollutant like TCE is an
exploration of the microbial behavior of such a labora-
tory construct under anticipated operational conditions.
These issues are more readily addressed in a contained
bioreactor than in an environmental application. Prob-
lems associated with the use of laboratory bacteria in
field releases include optimizing the activity of the or-
ganism under environmental conditions and defining the
risk associated with the introduction of a non-native or
genetically altered microorganism. The use of a co-oxi-
dative bacterial pathway does not permit direct selection
for the organism because the pollutant cannot be used
as a carbon and energy source. Therefore, nutrients
must be added to the contaminated aquifer to feed the
TCE degrader. As a result, a significant shift in the
downstream aquifer microbial community is anticipated.
The primary purpose of this research is to address not
only the fate of the introduced altered bacterium and the
specific genetic elements involved, but also the extent
to which this treatment technology may affect the native
microbial populations during an in situ bioremediation
experiment. The anticipated application of this organism
in situ will involve the addition of TCE and nutrients to
an aquifer engineered to contain a bacterial treatment
system.
Two suites of parameters are critical to assessing the
impact of a microorganism released into the environment
1. Persistence and activity of the introduced cells and
their genetic information beyond the spatial and tem-
poral scales of their intended activity.
2. Response of indigenous populations of bacteria and
bacteriovores to the added biomass, nutrient enrich-
ment, contaminant, and supplemental substrate me-
tabolites.
Persistence and Activity
P. cepacia G4 PR1 is a nonrecombinant derivative of P.
cepacia G4 that constitutively expresses toluene ortho-
monooxygenase (Tom), which in turn co-oxidizes TCE
(3). Several methods have come to the forefront of
molecular ecology that present the opportunity to track
specific bacterial cells regardless of the ability to culture
or isolate them. Prior to its release, the identification of
DNA sequences unique to an organism will permit track-
ing of that organism by detection of those sequences in
nucleic acid samples extracted from the environment.
Our approach centers around determination of the nu-
cleotide sequence of three regions of PR1 DNA:
1. The Tom and catechol-2,3-dioxygenase (C230)
genes responsible for the oxidation of several aro-
matic chemicals and TCE located on the large plas-
mid of G4 (1,2).
2. The site of Tn 5 insertion in G4.
3. A gene encoding an oxygenase of unknown function,
cloned from P. cepacia G4, which allows Escherichia
47
-------
co//to produce indigo from indole (a trait not demon-
strable in G4).
Since the Tom pathway is believed to be broadly distrib-
uted, its cloned genes might not provide a specific DNA
probe for PR1. Likewise, Tn 5 is expected to occur
among the aquifer bacteria, though not likely in conjunc-
tion with Tom and C230 genes. The junction site of Tn
5 insertion in G4, however, should be unique over a
2Q-to-50 base pair sequence. Another "primer" se-
quence can be obtained from the published sequence
of Tn 5, and the two rain be used in concert as polym-
erase chain reaction (PCR) primers. It also would be
possible to use internal Tn 5 sequences, or PR1 se-
quences, just outside the site of insertion to provide
"nested" primers. Because the restriction enzyme map
of this region of DNA is known, it will be possible to verify
the identity of any amplified fragment by analysis of
restriction fragment lengths. Finally, the probe available
from the indole oxygenase (indigo-producing E. coli
clone) would provide additional confirmation.
Evidence for Tom Being Plasmid Encoded
G4 and its derivatives contain two plasmids of approxi-
mately 50 and 112 to (designated pG4S and pG4L,
respectively, for those found in strain G4; pPR1S and
pPRILfor those in strain PR1). Following 100 genera-
tions of nonselective growth, G4 PR1(pRO101) isolates
were discovered that could no longer grow on phenol as
a sole carbon source and were unable to degrade TCE.
These isolates were found to have retained plasmids
pRO101 and pPRlS, but not pPR1L.
Strain G4 (Km* and phenol inducible for TCE degrada-
tion [TCE4*1]) was mated to a Kmf, TCE' derivative of
PR1: PR1 (pRO101, pPRlS, pPR1L"). AKrtftranscon-
jugant was isolated that carried pG4L, pPRlS, and
pR0101. This strain was found to have regained the
ability to degrade phenol. TCE degradation, however,
occured only after phenol induction: TCE^.
The TCE* strain PR1 (pRO101, pPR1S, pPR1L") was
rendered resistant to nalidixic acid (Nal) and rifampicin
(Rif) and mated to PR1 (pPR1S, pPR1L) (NaP, Rif).
The 112-b pPR1L was transferred from the Nal8, Rif*
strain to the phenol", Nal', Rif PR1 (pRO101, pPR1S,
pPR1L") strain to create the Nalr, Rif strain PR1
(pROlOl, pPR1S, pPR1L). PR1 (pRO101, pPR1S,
pPR1L) also was able to degrade phenol and TCE.
However, PR1 (pROlOl, pPR1S, pPR1L) degraded
TCE constitutively (TCE*5). This is persuasive evidence
that Tom and C230 as well as their transcriptional con-
trollers are located on the large indigenous plasmid.
Cloning
The genes responsible for Tom and C230 have been
cloned from PR1 into a 2.9-kb E. coli cloning vector,
pGEM3Z, as an 11-kb Eco Rl fragment to create the
recombinant plasmid: pMS64. E. coli JM109 (pMS64)
constitutively degraded 20 M TCE, produced o-cresol
from toluene and 3-methylcatechol from m-cresol and
o-cresol (all phenotypes associated with Tom activity in
G4). In addition, C230 activity (encoded by the Tom
pathway in G4) was detectable in this recombinant £
coli. Various subclones of this Eco Rl fragment have
allowed us to define an approximately 6-kb fragment
that encodes Tom activity (Figure 1).
The TOM Plasmid and Probe Development
Hybridization of the 11-kb Eco Rl fragment of pMS64
with total DNA isolated from G4 (pG4S, pG4L) and its
derived strains—PR1 (pROlOl, pPR1S), PR1
(pRO101, pPR1S, pPR1L)—took place only with the
large, approximately 112-kb plasmid (i.e., pG4L or
pPR1L) in each rase. We propose that pG4L will serve
as the archetype for a new class of catabolic plasmid
known as TOM, which encodes an ortfto-hydroxylation
pathway for the degradation of benzene, toluene, o-
xylene, cresols, and phenol.
This approximately 112-kb toluene degradative plasmid,
TOM, is very different from the much better known TOL
plasmid (archetype: pWWO) (4), which encodes the
oxidation of toluene through a series of oxygenases and
dehydrogenases to benzoate and then catechol. Tom is
(so far) unique to G4and TOM. Its existence now poses
some provocative possibilities for environmental reme-
diation of several priority aromatic and aliphatic pollut-
ants not only from the point of view of their constitutive
degradation, but also from the observation that TOM
appears to be a broad host range, self-transmissible
plasmid (a kanamycin-resistant derivative was trans-
ferred to E. coli where it is currently stably maintained).
The cloning and expression of toe genes for Tom and
C230 (i.e., torn A and torn B) in E. coli allow study of the
effects of expression of these genes in a well-defined
genetic and physiological background. In addition, their
availability allowed the construction of function-specific
probes.
Phenotypic Tracking
In addition to the genetic probes, G4 is capable of being
isolated by making use of its ability to utilize pthalate in
ttie presence of kanamycin. Background numbers of
organisms from the aquifer material using this selective
medium will be determined. In addition, large numbers
of culturable bacteria can be screened for their ability to
constitutively transform a fluorinated cresol via the Tom
pathway (1). This degree of overlap between the genetic
and phenotypic analyses should provide an adequate
detection capability to monitor fate of not only the altered
organism in the environment, but more importantly, the
48
-------
0 1 z 3 4 S 6 7 8 9 10 11 12 13 14kb
J_J I I I
torn A
Hpal
Tan CZ30 E S
+ +
EKBH HK K S
- +
- +
B pGEM4Z B
0 pGEM4Z B
' I
E pGEM4Z B
B pG£M4Z B
i
E pG£M4Z H
B H B HN SI NB
H
L
H
L
B
L
H
< Kom
s M
8 8«nHf
N NT*
e ecqh
II I 11 1
ii l n rl i r |
| torn B
pMSG4
pMSS9
pMS70
E
j pMS7S
pMS76
rE pMS71
pMS73
- +
+ -
pGEM4Z
E pG EM4Z * *
K
L_
S
L
J pMS78
pMS79
+ -
+ +
f pGEM4Z f
^pGBWZ^j^
K
L.
(SfT,,a,) pG£M4Z J
+ + J - 1
N
N
L_
(Hpal)
(Hp»0
pMS80
E
j pMS81
pMS82
E
_j pM583
Figure 1. Eco Rl fragment that arte odes Tom activity.
49
-------
fate of the genetic information and its capacity for phe-
notypic expression.
Response of Native Populations
Monitoring the response of native bacteria and bac-
terivorous protists to the implementation of this technol-
ogy will involve a variety of parameters aimed at defining
numerical and functional responses of bacteria and pro-
tists. Total bacterial numbers will be determined by epi-
fluorescence direct counts using the fluorochrome DAPI
(5). Determination of viable bacterial counts will be
made using a low nutrient concentration agar in addition
to the selective plating for G4 described above. Bacterial
activity measurements will be carried out with the incor-
poration of ^ thymidine into DNA and 14C leucine into
protein in a dual label technique (6). Inorganic nutrient
analyses will be conducted by standard techniques.
Oxygen levels and oxygen consumption will be deter-
mined by Winkler titration. Total dissolved organic carb-
on (Shimadzu TOC) and the carbon-to-nitrogen ratio of
particulate material (Carla-Erba CHN) in the ground
water will be determined. The ability of G4 to compete
with and integrate into the communities of existing mi-
crobial organisms in the aquifer material also will be
tested by determining the ability of G4 to colonize
biofilms developed from aquifer material. This ability will
be tracked with fluorescence techniques, using scan-
ning confocal laser microscopy.
Previous studies of contaminated ground water sites
have indicated a significant numerical response of bac-
terivorous protist populations as part of overall microbial
response to carbon enrichment from contaminants (7,
8). The ecological roles of these protists affect the dy-
namics of bacterial response to organic substrates by
limiting bacterial numbers below saturation levels and
providing feedback stimulation of bacterial metabolism
and growth through excretion of nutrients and labile
organic substrates. Protists also might play an important
role in the containment and elimination of introduced
bacterial cells. Total bacterivorous protist numbers will
be determined using a modified MPN (most probable
number) technique (9), as well as by direct fluorescent
counts.
References
1. Shields, M.S., S.O. Montgomery, S.M. Cuskey, P.J.
Chapman, and P.H. Pritchard. 1991. Appl. Environ.
Microbiol. 57:1935-1941.
2. Shields, M.S., S.O. Montgomery, P.J. Chapman,
S.M. Cuskey, and P.H. Pritchard. 1989. Appl. Envi-
ron. Microbiol. 55:1624-1629.
3. Shields, M.S., and M.R. Reagin. 1992. Appl. Environ.
Microbiol. 58:3977-3983.
4. Worsey, M.J., and P.A. Williams. 1975. J. Bacterid.
124:7-13.
5. Porter, K.G., and Y.S. Feig. 1980. Limnol. Oceanogr.
25:943-948.
6. Chin-Leo, G., and D.L. Kirchman. 1988. Appl. Envi-
ron. Microbiol. 54:1934-1939.
7. Madsen, E.L., J.L Sinclair, and W.C. Ghiorse. 1991.
Science 252:830-833.
8. Sinclair, J.L 1991. Proceedings of the First Interna-
tional Symposium on Microbiology of the Deep Sub-
surface. pp. 3-39 to 3-45.
9. Sinclair, J.L., and W.C. Ghiorse. 1987. Appl. Environ.
Microbioi. 53:1157-1163.
50
-------
Factors Determining the Effectiveness of Microbial inoculation In Soils and
Sediments: Effectiveness of Encapsulation
Jian-Er Lin
Technical Resources, Inc., Gulf Breeze, FL
James G. Mueller
SBP Technologies, Inc., Gulf Breeze, FL
P. Hap Pritchard
Gulf Breeze Environmental Research Laboratory, U. S. Environmental Protection Agency, Gulf Breeze, FL
Extensive studies have resulted in identification of many
microorganisms with unique degradative ability. The po-
tential for inoculating contaminated soils and sediments
with the identified microorganisms to effect the degrada-
tion of pollutants has been proposed (1). If inoculation
is defined as the process of introducing microorganisms
into a site in which the inoculated microorganisms sur-
vive and significantly affect the fate of a target chemical,
then few scientifically documented examples, if any,
exist where this process has been successful on a
significant scale. In most cases to date, bioremediation
has depended on the abilities of naturally existing micro-
bial communities to degrade toxic wastes under environ-
mental conditions that have been managed to enhance
their activities. This approach has succeeded for certain
types of contaminants, such as petroleum products,
low-molecular-weight polycyclic aromatic hydrocarbons
(PAHs), and nonchlorinated solvents. As other complex
chemicals (e.g., chlorinated solvents, pesticides, high-
molecular-weight [HMW] PAHs, and PCBs) come more
into focus for biotreatment, however, the sophisticated
approaches to bioremediation wilt undoubtedly include
inoculation procedures to introduce unique and special-
ized metabolic capacities into a contaminated matrix.
Furthermore, inoculation also is important in enhancing
the consistency and controllability of a bioremediation
process.
Effectiveness of microbial inoculation in soils and sedi-
ments for bioremediation and pollution control may be
determined by the following factors: 1) concentration of
active inoculants; 2) interaction between added micro-
organisms and indigenous populations; 3) nutrient (in-
cluding electron acceptor) supplies to the target micro-
organism; 4) availability of target compounds to the
added microorganism; and 5) effects of heterogenous
matrices on the biodegradation process. Manipulation of
these factors has become a critical issue in an inocula-
tion practice.
Forttiis study, the use of cell encapsulation technologies
was proposed to overcome some of the existing difficul-
ties associated with inoculation for bioremediation. To
understand the effect of encapsulated microbial inocu-
lants on a biodegradation process, several encapsula-
tion technologies were established or evaluated. The
use of ttiese technologies for biodegradation of HMW
' PAHs and pesticides in soil-associated systems was
explored. The outline of this work follows.
Technology Development
Several encapsulation technologies (2-4) have been de-
veloped for this study. These technologies are similar to
or are modifications of those previously described (5-7)
for using identified microorganisms for bioremediation
and pollution prevention in soils and sediments (Table
1). These include polyvinyl alcohol (PVA) capsules,
vermicuiite formulation, and poiyurethane pellets, repre-
senting controlled-release, solid-particle-supported,
and polymer-immobilized inoculants, respectively. Four
microbial strains—strain CRE 7, Pseudomonas pauci-
mobilis EPA 505, Mycobacterium sp. PYR-1, and Alcali-
genes eutrophus AEO 106 (pRO 101)—were
encapsulated with ttiese technologies. Various addi-
tives, such as adsorbents, solid nutrients, and densifica-
tion agents, were also included in the capsules when
required. In addition, this study also established several
processes, such as a soil-slurry reactor system, solid-
51
-------
state remediation (landfarming or composting), and a
washing-water bioreactor system that used the devel-
oped encapsulated microorganisms. Three kinds of en-
capsulated inocuiants were used in different treatment
processes depending on the following factors: 1) solu-
bility of target compounds, 2) treatment purposes (reme-
diation or prevention), and 3) properties of treated
matrices and treatment processes (Table 1).
Experiments and Results
Experiments were performed to determine the following
characteristics of the developed encapsulated microor-
ganisms: 1) viability of encapsulated microorganisms
under actual storage and use conditions; 2) activity of
encapsulated microorganisms under a defined condition
and in microcosms containing samples from contami-
nated sites, and 3) degradation kinetics in conjunction
with a treatment system. The treated samples included
phenanthrene, fluoranthene, creosote, and 2,4-dichlo-
rophenoxyacetic acid-contaminated soils.
Experiments showed that the encapsulation technolo-
gies maintained a high viability of inocuiants for at least
2 months, while the non-encapsulated inocuiants lost
their viability within 2 to 3 weeks. When the encapsu-
lated inocuiants were used for biodegradation in various
test systems, they resulted in the degradation of target
compounds. Moreover, polymer imrtiobilization of cells
could facilitate the reuse of inocuiants in continuous-flow
systems and provide microenvironmental conditions
suitable for their metabolic activity. Co-encapsulation of
various additives with cells added another advantage
that enhanced the degradative activity over the non-en-
capsulated inocuiants.
Summary
Encapsulation represents a practical means of storage,
distribution, and application of identified microorganisms
for inoculation in soils. The optimization and application
of encapsulation technologies in a variety of contami-
nated systems need to be investigated further Various
factors influencing the effectiveness of encapsulated
inocuiants in a contaminated site need to be identified.
References
1. Pritehard, P.H. 1992. Use of inoculation in bioreme-
diation. Curr. Opin. in Biotech. 3:232-243.
2. Lin, J.-E., J.G. Mueller, and P.H. Pritehard. 1992. Use
of encapsulated microorganisms as inocuiants for
bioremediation. Abstract Book for the ACS Sympo-
sium on Emerging Technologies for Hazardous
Waste Management, pp. 126-128, Sept 21-23,
1992, Atlanta, GA.
3. Lin, J.-E., H.Y. Wang, and R.F. Hickey. 1991. Use of
co-immobilized biological systems to degrade toxic
Table 1. Summary of Developed Encapsulation Technologies
Encapsulation Microorganisms
Technology Properties Potential Applications Encapsulated Examples of Test
Polyvinyl Alcohol
(PVA) Capsules
capsules are water soluble
and their dissolution
controllable
additives co-encapsulated
maintain viability at 4°C for
>2 months
as carriers to deliver
inocuiants to a site
timed release of
inocuiants and additives
for remediation and
pollution prevention
Strain CRE 7
(Phenanthrene degrader)
P. pauclrrubiUs EPA 505
(Fluoranthene degrader)
Mycobacterium sp.
PYR-1 (HMW PAHs
degrader)
A. outrophus AE0106
(pRO 101) (2,4-D
degrader)
soil slurry reactor (3,8)
solid-state remediation
(landfarming or
composting) (3,8)
controlled release in
soil for preventing
pollution by pesticides
0)
Vermicullte
Formulation
Polyurethane Pellets
powdered solid support
supplement of medium
supports growth of
inocuiants (no additions]
fermentation needed)
maintains viability at room
temperature for >2 months
immobilize inocuiants and
additives (nutrients,
adsorbents, and densificatlon
agents) in the polymer matrix
as carriers to deliver
inocuiants to a site
forms nuclei of inocuiants
repeatedly use the
inoculant
remove contaminants
rapidly from the mobile
phase
provide micro-
environmental conditions
suitable for the
inoculanfs activity
same as above
P. paucimobilis EPA 505
(fluoranthene degrader)
• soil slurry reactor (8)
• solid-state remediation
(9)
bioreactors in
conjunction with soil
washing and ground
water cleanup (8)
52
-------
organic compounds. Biotechnol. Bioengineer.
38:273-279.
4. Lin, J.-E, and H.Y. Wang. 1991. Degradation of pen-
tachlorophenol by non-immobilized, immobilized,
and co-immobilized Arthrobacter cells. J. Ferm. Bio-
engineer..4:311-314.
5. European Patent Office. 1989. Encapsulation
Method. C.A. Baker, A.A. Brooks, R.Z. Greenley, and
J.M.S. Heins. Publication No. 0 320 483.
6. Graham-Weiss, L, M. Bennett, and A.S. Paau. 1987.
Production of bacterial inoculants by direct fermen-
tation on nutrient-supplemented vermiculite. Appl.
Environ. Microbiol. 53:2138-2140.
7. O'Reiily, K.T., and R.L. Crawford. 1989. Degradation
of pentachiorophenoi by polyurethane-immobilized
Flavobacterium cells. Appl. Environ. Microbiol.
92113-2118.
8. Lin, J.-E. et al. 1993. Unpublished data.
9. Lin, J.-E..J.G. Mueller, and P.H. Pritchard. 1993. Use
of encapsulated microorganisms to prevent pollution
by pesticides. HMCRI National R&D Symposium.
53
-------
Combining Treatability Studies and Site Characterization for Rational Design of
In Situ Bloremediation Using Nitrate as an Electron Acceptor
S.R. Hutchins, D.H. Kampbell, M.L. Cook, F.M. Pfeffer, R.L. Cosby, and J.T. Wilson
Robert S. Kerr Environmental Research Laboratory, U.S. Environmental Protection Agency, Ada, OK
B. Newell
Robert S. Kerr Environmental Research Laboratory (ManTech), U.S. Environmental Protection Agency, Ada, OK
J.A. Johnson, V. Ravi, and J.K. Rurnery
Robert S. Kerr Environmental Research Laboratory (Dynamac), U.S. Environmental Protection Agency, Ada, OK
Rational design relates laboratory treatability studies at
field scale to the distribution of contaminants and to the
residence time of remedial fluids. The electron acceptor
is usually the limiting factor in bioremediation. Ideally,
the electron acceptor should not be depleted as water
or air moves across the region contaminated with oily
phase material. When all of the contaminated mass
receives adequate supplies of electron acceptor, the
course of remediation should parallel that established in
the laboratory study. If regions of the contaminated
mass are not adequately supplied, the course of reme-
diation at field scale is not predicted in any straightfor-
ward way from the laboratory study. Rational design
compares the residence time and concentration of elec-
tron acceptor at field scale to the demand demonstrated
for the electron acceptor in the laboratory to ensure that,
the engineered implementation of in situ bioremediation
is adequate.
This approach will be used to predict, a priori, the course
of remediation of a spill of refined petroleum products
from a pipeline near Park City, Kansas. Figure 1 relates
the lithology to the vertical extent of hydrocarbon con-
tamination. A system of 530 shallow injection wells have
been installed above the spill. They are arranged in a
grid on 6-m centers. These wells will circulate ground
water through the spill. Additional site information has
been published by Kennedy and Hutchins (1).
The site has teen selected for an evaluation of the
relative contribution of the various components of in situ
bioremediation. Nitrate has been selected as the pri-
mary electron acceptor. The demonstration also evalu-
ates a potential synergism between nitrate and low
concentrations of oxygen. The water circulation system
Injection Well
Depth
(feet)
Total Petroleum
Hydrocarbon
(mg/kg)
3000 6000
Water
Sand and Gravel ?§£!
ligill Bedrock
Figure 1. Relationship between the lithology, water table, and
vertical interval contaminated with oily phase hydro-
carbon at a pipeline spill undergoing in situ bioreme-
diation.
isolates the site into six cells of roughly equivalent size
(Figure 2). Ground water without additional electron ac-
ceptors will be circulated to Cell #4, ground water
amended with 8 mg/L nitrate-nitrogen will be circulated
to Cell #2, and ground water amended with 8 mg/L
nitrate-nitrogen and 1 mg/L oxygen will be circulated to
Cell #3. Cells #1, #5, and #6 will not receive ground
54
-------
Pipeline
Key Description
TPH Concentration
(Kg/m2)
$1 1-3
100 meters
Figure 2. Relationship between treatment cells, monitoring wells, boreholes, and water circu-
lation walls at a pipeline spill undergoing In situ bioremedlation. Open circles are
pumped wells, doaed circles are passive monitoring wells. Recharge was distributed
evenly to the ceils. The infiltration wells are too numerous to display.
water and will serve as controls for Cells #4, #2, and #3,
respectively.
This study has three components: 1) a treatability study
to determine the rate of nitrate consumption in core
material; 2) a bromide tracer study of the vertical flow of
water at field scale, and 3) a modeling exercise validated
by a bromide tracer study to evaluate the horizontal flow
of water.
Treatability Study
Continuous core samples were acquired from boreholes
60N and 600 in Cell #2. The cores were collected in
vertical intervals of 0.5 to 1.0 ft from 9 to 25 ft below land
surface. Subsamples were extracted and analyzed for
benzene, toluene, ethylbenzene, and xylenes (BTEX)
by gas chromatography (GC)/mass spectrometry (MS),
and for total petroleum hydrocarbons (TPH) by GC using
JP-4 jet fuel as a reference standard.
Microcosms were prepared in an anaerobic glovebox
using 60-mL serum bottles with 75 g of aquifer solids
and sterile diluted spring water, and amended with am-
monium and phosphate salts as described previously
(2). No exogenous carbon source was added. Each
microcosm then was spiked with 2.21 mg nitrate as
nitrogen as a concentrated aqueous solution to yield
initial concentrations of 50 to 80 mg/L nitrate-nitrogen.
For each core sample, three viable microcosms and one
control were prepared. Controls contained 500 mg/L
sodium azide and 250 mg/L mercuric chloride as
biocides. Microcosms were sealed without headspace
using Teflon-faced septa and incubated in an anaerobic
glovebox at 20°C. Periodically, 2-mL samples were ob-
tained from each microcosm and the volume was re-
placed with sterile glass beads. Samples were analyzed
for nitrate, nitrite, ammonia, phosphate, and pH. Those
microcosms exhibiting complete removal were respiked
with nitrate.
Rates of nitrate depletion were plotted for each micro-
cosm. The maximum zero-order rate constants were
calculated from the initial linear portions of the graphs
(Table 1). The data indicate that, despite variability in the
actual rate of nitrate removal, the entire core profile
generally exhibited denitrifying activity, and that no sig-
nificant effect of depth on the rate of removal occurred.
55
-------
Table 1. Rate and Extant of Nitrate Removal in Cora Material
from Locations 60N and 800
NO3-N
Removal Nltrate-N
Core
Depth
BTEX
TPH
Rate
Removed
Sample
(ft)
(mg/kg) (mg/kg)
(mg/L/day)
(mg/kg)
60N-C
16.0-16.4
1.6
40
4.37±0.46
69.5
60N-B
16.4-17.2
1.4
112
1.3010.13
22.8
60N-A
17.2-18.0
2.2
256
0.8310.18
23.1
60N-E
18.0-19.4
1.4
95
1.2110.46
26.9
60N-D
19.4-20.0
11.3
567
1.2110.23
24.8
60N-H*
20.0-21.2
32.7
695
0.5010.46
26.0
60N-G
21.2-22.0
249
3760
2.5911.39
25.3
60N-F
22.0-22.5
264
3620
1.0510.46
18.8
60N-J
22.5-23.6
331
4530
1.1610.07
28.1
BON-1
23.6-25.0
3.1
38
0.6310.11
36.8
60N-L
25.0-25.4
0.6
8
2.0910.09
29.3
60N-K
25.4-26.2
0.4
14
0.6910.31
- 23.9
SOO-E
8.1-8.6
0.1
18
0.7310.06
29.5
600-0
8.6-9.0
0.1
12
1.0610.55
29.1
60O-C
9.0-9.8
0.1
14
1.7410.01
29.3
600-a
9.8-10.6
0.1
15
1.5910.15
42.5
60O-H
11.5-12.3
0.3
18
2.1410.42
29.3
60O-G
12.3-13.1
0.2
16
0.1510.08
15.6
60O-F
13.1-14.0
1.4
57
4.8410.77
57.7
60O-L
14.0-14.4
1.1
84
5.6110.54
102.3
600-K
14.4-14.8
2.5
245
7.0610.31
28.9
600-J
14.8-15.6
4.1
389
5.3210.43
28.4
60O-I
15.6-16.5
8.8
1060
3.5310.71
102.3
60O-O
16.8-17.2
17.9
179
3.3410.91
102.3
600-N
17.2-18.0
9.1
334
2.11 ±0.34
45.5
60O-M*
18.0-19.0
31.9
1090
2.9010.37
34.5
60Ofl
19.0-20.0
45.7
1420
3.4710.21
43.2
600-Q
20.0-21.0
127
3160
1.9411.15
28.7
60O-P
21.0-22.0
132
3110
0.3210.09
21.5 .
60O-T
22.0-210
13.5
219
0.5310.03
28.9
60O-S
23.0-24.0
0.5
9
4.6910.12
68.9
60O-V
24.0-24.9
0.5
9
3.01±0.58
72.8
60O-U
24.9-25.3
0.2
7
2.3310.56
53.1
* Water table
Further, no correlation with either BTEX concentrations
or concentration of TPH was evident.
The average nitrate removal rates were 1.47 ± 1.09
mg/LVday at borehole 60N and 2.85 ± 1.87 mg/Uday at
borehole 600. Taking 2.85 mg/Uday as an estimate of
the nitrate consumption rate, the circulated concentra-
tions of 8 mg/L would be consumed within 67 hours at
field scale.
Tracer Test of Vertical Flow
The pipeline spill is within alluvial deposits of the Little
Arkansas River. A clay and silt layer extends 10 to 15 ft
below land surface. Below the clay, a sand extends
down to about 35 ft below grade. The deeper regions of
the sand are more coarse and have higher hydraulic
conductivities. The water table averages about 17 to 20
ft below land surface. The oily phase spill extends from
the bottom of the clay layer through the water table to
about 20 ft below land surface.
A bromide tracer study was conducted in Cells #2, #3,.
and #4. Ground water was produced from recovery well
REC-1, amended with 45 mg/L bromide, and injected
back into the cells through the infiltration wells installed
just below the first clay layer. Breakthrough of the tracer
was monitored in two sets of cluster wells that were
installed in Cell #3, the most contaminated cell. Cluster
G was installed in the center of the cell, and cluster M
was installed at the downgradient margin of the cell,
close to the second recovery well, REC-2 (Figure 2).
Both sets of cluster wells showed the same pattern:
breakthrough in the second and third well were appro-
priately staggered, suggesting the flow was primarily
vertical in the depth interval from 20 to 30 ft (Figure 3).
Shortly after breakthrough at the third level, tracer broke
through at the fourth level. This breakthrough would be
expected if flow at the 30 to 35 ft level was primarily
horizontal toward the recovery well REC-2 and the bro-
Partt City Bromide ResuRs 5/1-17/92
G Cluster Wells
40
35
30
i25
E
& 20
15
10
5
0
40
35
30
3.25
E
m 20
15
10
5
0
0 50 100 150 200 250 300 350 400
Hours into Study
Figure 3. Breakthrough of a bromide tracer In stacked cluster
wells. Cluster wells G2 and M2 from 20 to 25 feet, G3
and M3 from 25 to 30 feet, G4 and M4 from 30 to 35
feet, and G5 and M5 from 35 to 40 feet below grade.
Compare Figure 1. The shallow wells sampled the
sand containing the water table aquifer, but G5 and
M5 also sampled water from the sand and gravel unit
below the deep day layer.
G3
04
G5
0 50 100 150 200 250 300 350 400
Hours into Study
M Cluster Wells
M2
M3
M4
MS
-------
mide originated from water that had penetrated to this
level from upgradient.
The average flow of water pumped to each cell was 120
gal/min, which averaged 40,000 ft3 in surface area. This
flow would produce 0.578 ft of recharge each day, at a
porosity of 0.3, this would advance the tracer 1.93 ft/day.
In cluster 60-G, the tracer moved 5 ft through the second
well in 50 hours, and 10 ft to the bottom of well G-3 in
100 hours. In cluster M, the tracer moved 5 ft to the
bottom of well M2 in 100 hours, and 10 ft to the bottom
of M3 in 200 hours. The vertical component of flow was
22 ft/day in duster G and 1.1 ft/day in cluster M. The
vertical rate of advance of the tracer is close to that
predicted from recharge and is consistent with the hy-
pothesis that flow is primarily vertical in wells 60-G2,
60-G3, 60-M2, and 60-M3.
Tracer Test and Computer Modeling of
Horizontal Row
Ground water was pumped from well REC-1 and circu-
lated to the cells at a total flow of 360 gal/min (see Figure
2). A flow of 120 gpm was delivered to each cell. To
maintain hydraulic control, ground water was pumped
from well REC-2, located just below Cell #3, at 200
gal/min, and discharged after treatment to surface flow.
Flow was modeled in two dimensions using the public
domain code RESSQC, which is a module of WHPA,
version 2.0, a modular semfanalytical model for the
delineation of wellhead protection areas (U.S. EPA, Of-
fice of Ground Water Protection, 1991).
The model was run forward in time until it approached
an equilibrium. It predicts two investing properties of
the engineering design. The ground water recirculation
well recruits water from Cell #4, which is much less
impacted, and to a lesser extent from Cell #2, which is
marginally impacted; however, most of the water is un-
contaminated water recruited from the aquifer (cf. Figure
2 and Figure 4). REC-2 captures water from heavily
contaminated Cell #3, and part of Cell #2, and relatively
little uncontaminated water from the aquifer. When the
demonstration goes on line, it will operate as a combi-
nation of pump-and-treat and in situ bioremediation.
The predictions of the models were evaluated by com-
paring the predicted concentration of bromide tracer
discharged from REC-2 to the actual concentrations.
Over a 20-day tracer test, considerable agreement was
found between the prediction of the model and the
concentrations of tracer in the ground water from the
discharge well (Figure 5).
120 Day Capture
Zone
Call 3
REC-1
120 Day Capture
Zona
REC-1
Figure 4. Capture zones of circulated water and uncontami-
nated water predicted for ground water circulation
wells.
• Observed
Predicted
Figure 5.
10 15
Time (days)
Correspondence between recovery of bromide tracer
In circulation well REC-2 and the predictions of the
model.
Conclusions
If nitrate survives 67 hours in water moving 2.2 ft/day, it
shouid move at least 6.1 vertical feet after recharge.
This should be adequate to carry it below the oil con-
taminated layer (cf. Figure 1), and no limitation should
occur on the supply of nfaate.
References
1. Kennedy, L., and S.R. Hutchins. 1992. Applied geo-
logic, microbiological, and engineering constraints of
in situ bioremediation. Remediation J. 3:83-107.
2. Hutchins, S.R. 1992. Use of nitrate to bioremediate
a pipeline spill at Park City, Kansas: Projecting from
a treatability study to full-scale remediation. In: Biore-
mediation of Hazardous Wastes. Office of Research
and Development U.S. Environmental Protection
Agency. EPA/600/R-92/126.
57
-------
Rate and Extent of Natural Anaerobic Bloremediation of BTEX Compounds in
Ground Water Piumes
Morton A. Barlaz, Michael B. Shafer, Robert C. Borden
North Carolina State University, Raleigh, NC
John T. Wilson
Robert S. Kerr Environmental Research Laboratory, U.S. Environmental Protection Agency, Ada, OK
A combined field and laboratory study has been con-
ducted on the natural bioremediation of alkylbenzenes
released from a petroleum spill at the Sleeping Bear
Dunes National Lakeshore near Empire, Michigan. The
water table aquifer at the site is in highly transmissive
glacial outwash. In December 1989, three underground
storage tanks were excavated and removed from the
site. Soil surrounding the tanks smelted of gasoline,
indicating a leak had occurred. In February 1991, a
detailed site investigation was initiated to define the
horizontal and vertical extent of soil and ground water
contamination and the extent of natural bioremediation
of the dissolved benzene, toluene, ethylbenzene, and
xylene (BTEX) isomers. This investigation included a
detailed hydrogeologic characterization, a soil gas sur-
vey, and vertical coring to define the area containing oily
phase hydrocarbons. Ground water was monitored at
multiple depths over time for BTEX, electron acceptors,
donors, and nutrients. Laboratory microcosm experi-
ments were conducted to determine the anaerobic
biodegradatlon of BTEX under controlled conditions.
Field Characterization
Ground water flow is from the contaminant source area
toward the Platte River, approximately 70 ft away.
Ground water flow at the site is complicated by seasonal
fluctuations in the water table because of discharge from
the Platte River to the aquifer. Contaminant residence
time in the aquifer ranges from 5 to 53 weeks. Ambient
concentrations of oxygen, nitrate-nitrogen, and sulfate
are 2.4, 15.3, and 20 mg/L, respectively.
Dissolved BTEX declines rapidly along a streamline
passing through the most contaminated zone. In the
most contaminated interval immediately beneath the
source, dissolved benzene, total BTEX, methane, ni-
trate, sulfate, oxygen, and iron II are 3.12, 51.8, 0.31,
4.6, 8.5, 0.3, and 6.3 mg/L, respectively. Seventy feet
downgradient, benzene, total BTEX, nitrate, and sulfate
have declined to maximum concentrations of 0.45, 2.0,
0.10, and <0.05 mg/L, respectively, while methane and
iron II have increased to 3.1 and 5.2 mg/L, respectively.
Nitrate reduction, sulfate reduction, methanogenesis
from carbonate, oxygen respiration, and iron reduction
accepted 5.5, 1.7, 1.7, 0.3, and 0.5 mmol of electrons
per liter, respectively.
Biotransformation rates for BTEX were calculated after
first correcting for dilution of 2,3-dimethyipentane
(DMP). DMP was assumed to be refractory under the
anaerobic conditions present in this plume. First-order
bioattenuation rates were estimated by determining the
change in the natural logarithm of concentration at two
adjoining wells and dividing by the residence time of
water flowing between these wells. Results of this analy-
sis are shown in Table 1. The rates vary by about three-
fold. No apparent biodegradation of benzene occurred
under the anaerobic conditions present in this plume.
Microcosm Studies
Microcosm studies were conducted using aquifer mate-
rial collected at 0, 30, and 70 ft from the source to
Table 1. First-Order Decay Rates {d"') Estimated from Paid Data
30 Feat 70 Feet
Benzene
0.002 to 0.004
-0.004 to 0.002
Toluene
0.0S3 to 0.067
0.023 to 0.026
Ethylbenzene
0.003 to 0.010
0.003 to 0.011
p-Xylene
0.005 to 0.009
0.002 to 0.010
m-Xylene
0.005 to 0.014
0.004 to 0.008
o-Xylene
0.009 to 0.016
0.004 to 0.011
58
-------
evaluate the extent of biodegradation under ambient
conditions. The aquifer material was collected asepti-
cally and anaerobically using methods developed by
Robert S. Kerr Environmental Research Laboratory per-
sonnel. Ground water was collected anaerobically through
a 0.45-}im filter from monitoring wells adjacent to the core
locations.
The microcosms were prepared in 15-mL serum bottles
and constructed to simulate in situ conditions as doseiy
as possible. The experimental procedure followed was
essentially identical to the EPA protocol for estimation of
anaerobic microbiological transformation rate data (Fed.
Reg. Vol. 53, No. 115). BTEX was added to the micro-
cosms in concentrations similar to in situ conditions.
This addition yi/as necessary since some BTEX was lost
during ground water collection and preparation of the
microcosms in the laboratory. The microcosms were
constructed in an anaerobic glovebox with added so-
dium sulfide and resazurin to ensure reducing condi-
tions. Triplicate live and abiotic controls were sacrificed
at selected time points and analyzed to determine the
loss of BTEX and production of dissolved methane.
Figure 1 shows the variation in dissolved toluene and
methane in live and control microcosms from the 30 ft
location. Average concentrations are plotted as a solid
line. The dashed line is the first-order linear regression.
At day 60, a rapid increase in methane occurred in the
live microcosms. This increase greatly exceeded the
BTEX loss, indicating that other undefined substrates
were being biotransformed.
Toluene declined gradually until day 115 when a rapid
decline to less than 20 pig/L occurred by day 192. After
this rapid drop, toluene concentrations plateaued at be-
tween 5 and 15 |xg/L with no additional removal.
Biotransformation of the m+p-xylene, o-xylene, and
ethylbenzene was similar although it lagged the toluene
results somewhat The lag in ethylbenzene and xylene
removal coincided with a drop in toluene below 20 pig/L,
suggesting that biotransformation is sequential, with
toluene being the first to degrade. No apparent removal
of benzene occurred in any of the microcosms.
First-order removal rates for toluene, ethylbenzene, and
xylene isomers are shown in Table 2. Toluene biodegra-
dation was the most rapid, followed by ethylbenzene
and xylenes. The biotransformation rates shown are for
the entire data set and do not reflect the much higher
rates observed after the end of the lag period and before
a plateau is reached. First-order biodegradation rates
were higher in the 30 ft and 70 ft microcosms. Rates
may be higher because of the lower initial concentra-
tions present in these microcosms. In the 0 ft micro-
cosms, toluene is being removed but has not yet been
reduced to below 20 ng/L. In these microcosms, ethyl-
benzene and xylene biodegradation has been signifi-
cantly slower.
20.00
10.00
5
o
0.00
100"
10:
o
100
200
300
400
Time (days)
Figure 1. Biotransformation of toluene and production of ma th-
in* In laboratory microcosms from 30 ft location
(* a live, Q s abiotic).
Table 2. First-Order Biotransformation Rates from Sleeping
Bear Microcosms (d*1)
Compound 0 ft 30ft 70 ft Average
Toluene - Live
Toluene - Abiotic
Toluene - Net
0.0108
0.0021
0.0087
0.0111
0.0012
0.0099
0.0122
0.0016
0.0106
0.0098
Ethylbenzene - Live
Ethylbenzene - Abiotic
Ethylbenzene - Net
0.0037
0.0034
0.0003
0.0036
0.0026
0.0010
0.0103
0.0029
0.0074
0.0029
m&p-Xylene - Live
m&p-Xylene - Abiotic
m&p-Xylena - Net
0.0042
0.0039
0.0003
0.0040
0.0030
0.0010
0.0041
0.003S
0.0006
0.0006
o-Xylene - Live
o-Xylene - Abiotic
o-Xylene - Net
0.0036
0.0035
0.0001
0.0037
0.0026
0.0011
0.0033
0.0028
0.0005
0.0006
Comparison of Field and Laboratory
Results
Both the field monitoring and laboratory microcosm re-
sults indicate that benzene is not biotransformed under
anaerobic conditions at this site. Toluene, ethylbenzene,
and the xylene isomers were all transformed under an-
aerobic conditions in the field and laboratory micro-
59
-------
cosms. The match between field and laboratory rates is
not exact but is good considering the inherent differ-
ences in the techniques employed. Laboratory biotrans-
formation rates were typically a factor of 2 to 10 lower
than the field rates. Lower rates could be due to two
factors. The soil-to-water ratio in the microcosms is
about a factor of 3 lower than in the field. If most of the
microorganisms are associated with the aquifer mate-
rial, a lower microorganism-to-water ratio and lower
biotransformation rate would result. Also, since the data
from the lag and plateau periods were included in our
calculations, the estimated rates would be lower.
k.
60
-------
Field Treatment of BTEX In Vadose Soils Using Hydrofracturlng, Vacuum
Extraction, and Blotliters
Dolloff F. Bishop and Wendy Davis-Hoover
Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
Rakesh Govind
University of Cincinnati, Department of Chemical Engineering, Cincinnati, OH
Spills of fuels and leaking fuel tanks represent a major
source of vadose soil contamination. Such contamina-
tion, which includes aromatic hydrocarbons (benzene,
toluene, ethylbenzene, and xylene [BTEX]), leaches
through the vadose soil into ground water. Aromatic
hydrocarbons pose health risks when they leach into
ground water that is used as a drinking water supply.
EPA's Risk Reduction Engineering Laboratory (RREL),
in cooperation with the University of Cincinnati (UC), is
developing engineering systems to bioremediate fuel-
contaminated vadose soils. These systems include in
situ bioventing of soil and consolidated soils, a system
consisting of hydrofracturing and vacuum extraction to
transfer volatile organic compounds (VOCs) from the
soils to air, followed by air biofittration to mineralize the
extracted VOCs in the contaminated air.
Background
This presentation describes a planned field demonstra-
tion of the hydrofracturing, vacuum extraction, and air
biofiltration system. The hydrofracturing of consolidated
vadose soil is achieved when fluid is pumped down a
borehole until a critical pressure is reached that frac-
tures the soil. Sand-taden slurry then is pumped into the
fracture to create highly permeable pathways in the soil.
The resulting horizontal sand lenses, at vertical separa-
tions as close as 15 cm apart, produce improved overall
soil permeability for vacuum extraction of fuel hydrocar-
bons from the consolidated soil. The extracted air con-
taminated with the fuel VOCs then may be treated in air
biofilters to biodegrade the contaminant VOCs.
RREL and UC are developing improved biofilters to
control VOC emissions for RREL's Superfund research
program. The filter designs employ pellets or "straight-
passages" support media, as shown in Figure 1. The
VOC emissions in contaminated air passing through the
bfofilter are biodegraded by microorganisms attached
on the surface of the media. A nutrient solution is recy-
cled through the biofilter to support the microorganisms.
Research on improved biofilters reveals efficient re-
moval of biodegradable VOCs and a high degree of
mineralization.
Clean Air
Microorganisms
^ Immobifized on
Support Media
Nutrient
Solution
Air with VOCs
Fresh
Nutrients
Figure 1. Schematic of the biofilter system.
Field Demonstration
The site for the field demonstration has not yet been
selected but is likely to be an Air Force base in Ohio. At
the selected site, consolidated vadose soil contami-
nated with jet fuel VOCs will be hydrofractured by a
61
-------
RRELAJC team using the vacuum extraction equipment
installed to permit air extraction of the VOCs. Hydrofrac-
turing increases the overall soil permeability by nearly
two orders of magnitude. Since a single horizonal sand
lens, from fracturing with a radius of 7 m, typically pro-
vides approximately 0.1 m3/min of extracted air, three
horizontal sand lens (7 m radius) will be used to provide
0.28 m3/min of contaminated air for biofilter treatment
The field test will use two biofilters: one packed with
approximately 1 m (height) of porous ceramic pellets (6
mm avg. dia.), the other with approximately 1 m (height)
of porous ceramic straight-passages media (127 pas-
sages/cm2). Each biofilter will provide 2 min of empty
bed contact time within the media at an air flow of 0.14
m3/min. Steam and water will be used to raise the
temperature of the extracted air to approximately 25°C
and to prehumidify tie air.
The biofilters will be constructed at EPA's Test and
Evaluation (T&E) Facility in Cincinnati. The system will
include continuous gas chromatography for automatic
monitoring of appropriate air streams to characterize
biofilter performance. The tsfofilm on the support media
will be preacclimated to jet fuel hydrocarbons at the T&E
Facility. The skid-mounted biofilters with acclimated
biofilms will be transported to the site for connection to
the hydrofracture/vacuum extraction systems. The per-
formance of the integrated system will be characterized
for approximately 3 months.
62
-------
Section 4
Pilot-Scale Research
By studying bioremediation processes under actual site conditions on a small scale,
researchers can gather critical information on issues such as operation, control,
and management of residuals and emissions before moving to full-scale research.
Thus, pilot-scale research is a critical intermediate step in which the success of
laboratory experiments is further tested in an expanded but controlled setting.
Researchers reviewed the application of solvent washing with ethanol, followed by
treatment in an expanded-bed granular activated carbon (GAC) anaerobic biore-
actor, to remove pentachlorophenol from soil contaminated by wood preserving
operations.
In another study, researchers reported preliminary results on the use of trickle bed
biofilters with microbial supports to treat volatile organic compounds (VOCs) asso-
ciated with landfill ieachate stripping. With the emergence of biofiltration as a
practical and cost-effective means of removing low volumes of biodegradable
VOCs from large-volume air streams, researchers are studying application of this
technology for control of hazardous VOCs. Preliminary results were reported on
toluene removal. The trickle bed system also will be tested for degradation of
ethylbenzene, chlorobenzene, trichloroethylene, and methylene chloride.
The symposium's poster session included a presentation on the use of bioassays
for performing comparative assessments of complex mixtures before, during, and
after bioremediation efforts. Another poster presentation anticipated the use of
bioassays to assess the extent of bioremediation achieved in tests on optimal
composting conditions for the destruction of soil contaminants.
Bench- as well as pilot-scale treatability studies will be used by researchers
investigating the operation and performance of soil slurry bioreactors. In another
study, researchers used core samples to extrapolate the timeframe in which indige-
nous organisms might remediate a site contaminated with dioctyl phthalate and
xylene from the manufacture of vinyl wallcoverings.
63
-------
Treatment of PCP-Contaminated Soils by Washing with Ethanol/Water Followed
by Anaerobic Treatment
Amid P. Khodadoust, Julie A. Wagner, and Makram T. Suidan
University of Cincinnati, Cincinnati, OH
Steven I. Safferman
U.S. Environmental Protection Agency, Cincinnati, OH
Pentachlorophenol (PCP) has been used as a wood-
preserving compound since the 1930s. PCP-contami-
nated soils can be found at abandoned and existing
wood preserving sites, and at PCP manufacturing facili-
ties that are currently Resource Conservation and Re-
covery Act (RCRA) and Superfund sites (1,2,3). PCP is
considered a priority pollutant by the EPA, and its re-
moval from contaminated soils has been mandated
through the Comprehensive Environmental Response,
Compensation, and Liability .Act (CERCLA).
Among the technologies employing physical/chemical
processes for removal of pesticides from contaminated
soils, solvent washing of soil can be considered an
economical option. Factors that can influence the effec-
tiveness of solvent washing include the solubility of the
pesticide in the wash solvent, sorption capacity of the
soil for the pesticide, soil/solvent contact time, soil par-
ticle size distribution, the hydrophobic nature of the
pesticide (and the solvent pH), the pesticide application
matrix, soil moisture, the age of contamination, and
solvent cost and recovery.
Mueller et al. (4) found that ethanol was effective in
extracting polycyclic aromatic hydrocarbons (PAHs)
from wet contaminated soils. Traditionally, soil moisture
prevented the cleanup of the soil by using more active
solvents such as dichloromethane. Peters and Luthy (5)
demonstrated that water-miscible solvents such as
n-butylamine, acetone, and 2-propanol were capable of
removing PAHs from soils contaminated with coal tar.
The solvents considered for this study were acetone, for
above-ground soil washing, and ethanol, for both in situ
and above-ground soil washing. A PCP loading of 100
mg/kg soil was used as representative of PCP-contaml-
nated sites. The solvent washing procedure was used
for the removal of PCP from contaminated soils for
which both in situ and above-ground solvent washing of
soil were feasible. The in situ solvent washing (flushing)
of soil was simulated by continuously flushing solvent
through a packed bed of soil until PCP concentration in
the effluent did not decrease any further. The above-
ground (ex situ) soil washing was simulated by batch
tests (reverse isotherms) conducted on PCP-contami-
nated soil.
Expanded-bed granular activated carbon (GAC) an-
aerobic bioreactors (6) were used to treat the extracts
(spent solvents) from the soil solvent washing process.
The spent solvent from the soil washing tests was fed
to GAC bioreactors, where the PCP content of the wash
fluid was the biodegradable cometabolite and ethanol
served as the primary substrate.
For similar solvent throughputs, acetone was found to
be less effective than ethanol in flushing PCP from
contaminated soil. Thereafter the solvent washing tests
were conducted with ethanol. Solvent flushing was con-
ducted on 20 x 40, 60 x 80, and 100 x 140 U.S. mesh
soil fractions. The flushing solvent, 95 percent ethanol,
was applied at three different flow rates. Lower solvent
flow rates were more effective for the same solvent
throughput than higher flow rates in extracting PCP from
the soil, suggesting ttiat PCP desorption kinetics may
limit soil cleanup. The 20 x 40 U.S. mesh soil was
flushed with various ethanol/water mixtures, resulting in
higher PCP removal efficiency of the 50 percent and 75
percent ethanol solutions at lower solvent throughputs.
The data in Figure 1 show the results of the above-
ground soil washing test (reverse isotherm) on 100 x
140 U.S. mesh soil. The results indicate that during
batch extraction of soil, the 50 percent and the 75 per-
cent ethanol solutions achieved higher PCP removal
efficiencies than the other ethanol solutions. Similar re-
sults were obtained for the 20 x 40 U.S. mesh and the
clay fraction of the same soil, indicating that the superior
removal efficiency of the 50 percent and 75 percent
k.
64
-------
O Al Water
« 25% Ethanol
7 50% Ethanol
^ 75% Etnanol
q 95% Etrtanol
¦ 100% Ethanol
GO 80
Ce (mg/L)
100
120
Figure 1. Reverse Isotherm for 100 x 140 U.S. mesh soil.
I
ethanol solutions is not completely atfributable to the
solubility of PCP in ethanol, but rather partially to the
greater desorption of PCP from the soil in the presence
of water. In addition, the results from Figure 1 indicate
that the desorption of PCP was nearly independent of
the soil-mass-to-solvent ratio for the 50 percent and the
75 percent ethanol solutions. The results from Figure 2
indicate that the 50 percent and 75 percent etha-
nol/water mixtures obtained greater PCP recoveries
from 20 x 40 U.S. mesh soil at higher soil-mass-to-sol-
vent volume ratios. Similar results were obtained for the
100 x 140 U.S. mesh soil and the clay fraction.
SoilrSolvant Ratio (a/mL):
40 60 80
Ethanol In Solvent (%)
100
100
90
_80
^70
| 60
250
0.40
30
20
10
0
. #
Solent: 75% Ethanol
Solvent: Soil Ratio (g/mL): 0.01
o 12-hour test
¦ 21-day test
¦ i t i i i-V Z. i—i i i—i i
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 3 8 9 12 16 18 21 24
Tuna (days)
Figure 3. PCP removal from 20 x 40 U.S. mash soil with time.
conditions during acclimation of methanogenic cultures
and initial loading of PCP on the carbon. After stable
performance was attained, the PCP/ethanol feed rates
and the empty-bed contact times (EBCT) were altered
to study the chemical oxygen demand (COD) removal
and the reductive dechlorination of PCP. The effluent
COD and VFA (volatile fatty acids), and the effluent
concentrations of PCP and its biodegradation com-
pounds, were routinely monitored. After breakthrough of
PCP biodegradation products from the GAC occurred,
the high concentrations of para-chlorophenol (p-CP)
proved to be inhibitory to the methanogenic culture. A
practice of partial column GAC replacement with fresh
carbon followed the breakthrough period in order to
diminish p-CP inhibition. After a further breakthrough of
p-CP in the reactors, the PCP feed rate was lowered. A
gradual reduction of the carbon replacement rate com-
menced after attaining stable reactor operation. The
experimental data demonstrated an incomplete miner-
alization of perrtachiorophenol to carbon dioxide and
methane, and an equimolar conversion of PCP into
monochlorophenols. This indicates that the PCP feed
rate was the critical control parameter that effected re-
actor toxicity via the accumulation and inhibitory role of
p-CP (Figure 4).
Figure 2. PCP recovery from 20 x 40 U.S. mesh soil at high
soiksolvent ratios In batch extraction tests.
The effect of time on the removal efficiency of PCP from
20 x 40 U.S. mesh soil was studied (Figure 3). The
results from Figure 3 indicate that the maximum removal
of PCP from soil with 75 percent ethanol solution oc-
curred within a day, and that no additional removal was
obtained after longer periods of contact The results
from a shorter study revealed that maximum removal of
PCP occurred within 1 to 2 hours of the batch extraction
experiment
The extract (spent solvent) from the solvent washing
process was treated in expanded-bed GAC anaerobic
bioreactors. Two reactors were operated under similar
PCP (no btotof
# EI«i**PCP[Bau»)
a EfluftfltMCPl {aouil)
i°EJItu«*OCP»(*dua()
? SMuOT TCP*
10-«
0 100 200 300 400 500
Days
Figure 4. Effluent quality of chlorophertols In reactor A.
65
-------
References
1. U.S. EPA. 1986. Superfund Record of Decision (EPA
Region 6) United Creosoti'ng Co., Hilbig Rd., Conroe,
TX. EPA/ROD/R06-86/D14. U.S. Environmental Pro-
tection Agency.
2. U.S, EPA. 1989. Superfund Record of Decision (EPA
Region 6) United Creosoting Co., Conroe,
Montgomery County, TX. (2nd remedial action)
EPA/ROD/R06-89/053. U.S. Environmental Protec-
tion Agency.
3. U.S. EPA. 1987. Superfund Record of Decision (EPA
Region 4) Palmetto Wood Preserving, Dixiana, Lex-
ington County, SC. (1st remedial action)
EPA/ROD/R04-87/026. U.S. Environmental Protec-
tion Agency.
4. Mueller, J.G., et al. 1988. Preliminary study of treat-
ment of contaminated groundwater from the Tay-
lon/ille gasities site. HWRIC RR 077.
5. Peters, C.A., and R.G. Luthy. 1991. Coal tar disso-
lution in water-miscible solvents. Proc. 64th Water
Pollut Control Fed., Annual Conference, Toronto,
Canada.
6. Suidan, M.T. 1989. Treatment of biological inhibitory
waste with the expanded bed anaerobic carbon filter.
Proc. A&WMA/EPA Int. Symp. on Hazard. Waste
Treatment Biosystems for Pollut. Cont.
66
-------
Preliminary Evaluation of Attachment Media for Gas Phase Biofliters
Francis L. Smith, George A. SoriaJ, Paul J. Smith, Makram T. Suidan, and Pratim Biswas
Department of Civil and Environmental Engineering, University of Cincinnati, Cincinnati, OH
Richard C. Brenner
Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
The control and removal of volatile organic compounds
(VOCs) from contaminated air streams has become a
major air pollution concern since the enactment of the
1990 amendments to the Clean Air Act (1). For other
contaminated materials, such as water and soil, the use
of biological processes has become an accepted prac-
tice. More recently, biofiltration for treating contaminated
air has emerged as a practical technology for removal
of many VOCs. In comparison with other VOC control
technologies, such as carbon absorption and incinera-
tion, biofiltration can be more cost effective, particularly
for treatment of large volumes of air with low concentra-
tions of biodegradable VOCs (2). The low annualized
cost is because of the utilization of ambient temperature
microbial oxidation, rather than oxidation by thermal or
chemical means. Essentially, biofiltration uses ambient
temperature, self-regenerating, enzymatic catalysis.
Biofiltration consists of contacting a contaminated air
stream with a moist film of microbes attached to a
stationary synthetic or natural support material. The mi-
crobes then oxidize the VOCs to simpie end products,
such as carbon dioxide and water. By logical extension,
biofiltration as a hazardous VOC control technology has
been the subject of extensive research, and the process
design criteria have been identified (2-6).
This paper discusses the preliminary research per-
formed using trickle bed biofiiters with monolithic chan-
nelized (as well as palletized) microbial support media
for treatment of VOCs that are typical by-products of
landfill leachate stripping. Until now only toluene has
been tested, to characterize the trickle biofilter system.
In the future, this system will be tested for degradation
of ethylbenzene, chlorobenzene, trichloroethylene, and
methylene chloride. The objectives of the experiment,
relative to most biofilter research to date, are to investi-
gate the use of such biofiiters for treating chlorinated
compounds, while achieving high removal efficiencies at
high feed concentrations. Such compounds can be en-
countered in air streams from industrial processes as
well as environmental cleanup processes, such as soil
washing. The further research objective is to reduce to
practice biofiltration for the treatment of such VOC con-
taining air streams.
Experimental Apparatus
The trickle bed biofilter apparatus used in this study
consists of three independent, parallel trains, each con-
taining 4 ft of attachment media: Biofiiters A, B, and D.
Biofiiters A and B are filled with Coming Celcor chan-
nelized media, and D is filled with Manville Celite pellet-
ized media. The air supplied to each biofilter is highly
purified, for complete removal of water, carbon dioxide,
VOCs, and particulates. After purification, the air is split
off to each biofilter, humidified, injected with VOCs, and
fed to the biofiiters. The air is mass flow controlled, ant.
the VOCs are fed by syringe pumps. Each biofilter is
independently temperature controlled to 14°C.
Buffered nutrient solutions are fed to each biofilter. Each
biofilter independently receives a nutrient solution that
contains all the necessary macro- and micronutrients,
with a sodium bicarbonate buffer. Biofiiters A and B are
also equipped with effluent recirculation in order to pro-
vide even distribution of the biomass throughout the
Coming Celcor channelized medium. Biofilter B Is oper-
ated in a countercurrent mode (i.e., with the air flow fed
to the bottom). Biofiiters A and D are operated in a
cocurrent mode.
Results and Discussion
During startup each biofilter was fed 50 ppmv toluene at
a 12-minute residence time, and a nutrient solution feed.
of 20 LVday. Each biofilter was maintained at a constant
temperature of 14 ±2 aC and 7.7 ±0.2 pH range.
67
-------
Biofilter A The mass flow of toluene was increased
steadily up to 400 ppmv at a residence time of 12
minutes. The maximum percent removal of toluene that
could be obtained at 400 ppmv was only 80 percent. In
order to improve the performance, the effluent was re-
cycled to provide for even distribution of the biomass
throughout the support media. On day 127 after initial
startup, this recycle was started and the percent removal
increased to about 90 percent. The mass flow of toluene
was then increased to 500 ppmv at 12 minutes of resi-
dence time. When the percent removal of toluene was
stable at 99 percent, a residence time cycle test was
performed, with the residence time being varied from 12
to 1 minute and then back to 12 minutes. This was done
while holding constant the total mass of toluene fed per
day. Figure 1 shows the performance of the biofilter
during the residence time cycle test Note that the tolu-
ene removal stabilized at better than 96 percent up to 4
minutes of residence time. At 2 minutes of residence
time the removal efficiency dropped to about 90 percent,
and the performance dropped further to 65 percent at a
1-minute residence time. At a 1-minute residence time
the pressure drop was about 1 in. water.
600
400
200
50
40
30
20
10
0
o % Tokiww Removed
• Influtm Tolu*n« (OC)
• InAusnt Tokisna (Syiinge)
» EffluwK Toluan*
100
90
60
20
10
12
14
Figure 1. Biofilter A performance w.r.t. toluene removal during
a residence time cycle.
Biofilter B. The mass flow of toluene was increased
steadily up to 200 ppmv at a residence time of 12
minutes. The performance was particularly poor com-
pared to Biofilter A. The removal of toluene did not
exceed 70 percent and after day 87 the performance
started to drop. On day 113 the removal of toluene
dropped to about 57 percent, and at this point a decision
was made to Introduce effluent recycle to the biofilter.
On day 114 this recycle was started, and by day 122 the
removal of toluene was 85 percent At this point the
mass flow of toluene was increased steadily until it
reached 500 ppmv at a residence time of 12 minutes.
The biofilter was maintained at these conditions until a
stable effluent was obtained. At this point a residence
time cycle test was started and conducted in a manner
similar to that for Biofilter A. Figure 2 shows the perform-
ance of the biofilter during the residence time cycle test
i 500 i
% Tok*n» Aemovwd
tnttuant Totwn* (QC)
Inflow! Toluene (Syringe) |
° 200
"3 1°0 r
6 8 10
Retention Time, min
Figure 2. Biofilter B performance wxt toluene removal during
a residence time cycle.
Note that toluene removal during the cycle test stabi-
lized between 87 and 89 percent up to 6 minutes of
residence time, then 72 percent at 4 minutes, 45 percent
at 2 minutes, and 30 percent at 1 minute. At a 1-minute
residence time, the pressure drop was about 1.75 in.
water.
Biofilter D: The mass flow of toluene was increased
steadily up to 500 ppmv at a residence time of 12
minutes and a 99+ percent removal efficiency. At this
point a residence time cycle test was started and con-
ducted in a manner similar to that for Biofilter A. Figure 3
shows the performance of the biofilter during the resi-
dence time cycle test. Note that toluene removal during
the cycle test stabilized above 99 percent up to 2 min-
utes of residence time, then 95 percent at1 minute. The
pressure drop ranged from 4 in. water at 12 minutes to
22 in. water at a 1-minute residence time.
100
1 200
a * Tokan* Rtmwad
e Mwnl Tohwn* (<3C1
» MmM Toiucrw (fyrtngt)
j EffhMffl Tokfffw
80 c
2
5
"5
80
20
4 6 8 10
Ret&ntlon Time, min
14
Figure 3. Biofilter 0 performance w.r.t toluene removal during
a residence time cycle.
Conclusion and Future Work
The removal efficiency for each residence time was
similar, for both increasing and decreasing residence
times. It appears, however, that when increasing the
residence time, which causes an increase in tfie VOC
concentration at constant mass loading, more time is
68
-------
required by the biofllter to achieve maximum efficiency.
Effluent recycle, to establish even distribution of
biomass throughout the media, was necessary to
achieve maximum efficiency with the channelized sup-
port media.
Biofiiter D showed the highest VOC removal efficien-
cies. The efficiencies ranged from 99+ percent to 95
percent for residence times of 12 to 1 minute. Biofiiter
A, also operated cocurrently, showed somewhat lower
removal efficiencies of from 99 percent to 65 percent for
times of 12 and 1 minute, respectively. Finally, Biofiiter
B, operated countercurrently, showed efficiencies less
than 90 percent of from 90 to 30 percent for times of 12
minutes and 1 minute, respectively.
The pressure drops for both A and B were quite low, with
a maximum of 1 in. and 1.75 in. of water, respectively.
This low pressure drop, even for the countercurrent
mode, Is very promising since it indicates that the major
operating cost for this biofiiter design (i.e., blower motor
power) will be lower than for most typical biofiiter de-
signs. The pressure drop for D was as high as 22 in.
water. Further investigation will evaluate techniques for
removing excess biomass from the pelletized support
media in order to achieve high removal efficiencies at
acceptably low pressure drops.
Continuing work will include investigating the effect of
the recycle flow rate on the performance of each biofiiter.
The effect of increasing the mass loading to 500 ppmv
at lower residence times also will be investigated. After
finishing the characterization of the biofilters with tolu-
ene, feeding the other VOCs will be initiated.
References
1. Lee, B. 1991. Highlights of the Clean Air Act Amend-
ments of 1990. J. Air Waste Manag. Assoc. 41 (1):16.
2. Ottengraf, S.P.P. 1986. Exhaust gas purification. In:
Rehn, H.J., Reed, G., eds. Biotechnology, Vol. 8,
VCH Verlagsgesellschaft, Weinham.
3. Hodge, D.S., V.F. Median, R.L Islander, and J.S.
Devinney. 1991. Treatment of hydrocarbon fuel va-
pors in biofilters. Environ. Technol. 12:655.
4. Leson, G., F. Tabatabal, and A.M. Winer. 1992. Con-
trol of hazardous and toxic air emissions by biofiltra-
tion. Presented at the 85th Annual Meeting and
Exhibition of the Air and Waste Management Asso-
ciation, Kansas City, Missouri, June 21-26.
5. Van Langenhove, J., A. Lootens, and N. Schamp.
1989. Inhibitory effects of S02 on biofiltration of al-
dehydes. Water, Air & Soil Pollution 47:81.
6. Van Lith, C„ S.L. David, and R. Marsh. 1990. Design
criteria for biofilters. Trans. Inst. Chem. Eng.
68B:127.
69
-------
Approaches to the Development of Comparative Genotoxicity Risk Assessment
Methods for Evaluating Hazardous Waste Control Technologies
Larry D. Claxton
Health Effects Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC
Environmental situations requiring remediation typically
are present as mixtures of multiple toxicants and ottier
pollutants. Most types of remediation efforts, including
biologicaJ remediation endeavors, produce additional
compounds that add to the complexity of evaluating the
potential health effects of a contaminated site during and
after remediation attempts. Likewise, when environ-
mental applications of microorganisms are used for re-
mediation purposes, the organisms and their products
must be evaluated for safety.
Given the presence of multiple pollutants and the poten-
tial production of other pollutants by remediation proc-
esses, most of the actual toxicants within a typical
remediation site are not identified. The Committee on
Environmental Epidemiology of the National Research
Council (1991) states, There is evidence that NCPs
[nonconventional pollutants] are a potentially important
source of hazardous exposure. Some preliminary toxi-
cologic studies suggest that NCPs have important bio-
logic properties, environmental persistence, and
mobility.... these unidentified substances represent risk
of unknown magnitude." These NCPs, therefore, limit
the ability to conduct risk characterizations of remedia-
tion sites when only analytical chemistry is used for
exposure assessment studies.
By incorporating biological tests into assessment stud-
ies, it is possible to improve the estimations of potential
human toxicant exposure before, during, and after re-
mediation efforts. When appropriately coupled with ana-
lytical chemistry, bioassays also can be used to identify
the major toxic pollutants. In addition, during the devel-
opment of remediation methods, bioassays can be used
for comparative assessments between differing techno-
logical approaches. Also, naturally occurring, mutant,
and genetically engineered microorganisms have great
potential for use in environmental remediation. Environ-
mental releases, however, may result in direct human
exposure to these organisms.
In the 1970s, using primarily Escherichia coli, re-
searchers allayed public and scientific concerns that
recombinant strains could survive and transfer genes for
toxins to normal human microflora to a degree that
would cause major health concerns. Because some
microorganisms, however, when present in large num-
bers can cause adverse health effects by other mecha-
nisms, it is important to explore the potential health
effects of environmentally released organisms (ERO).
Recent efforts have examined competition and survival
of EROs with normal human flora, their effect upon
xenobiotic metabolism (environmentally and in vivo),
and their effect on the competition and survival of patho-
gens. This poster presents the use of bioassays for
comparative assessments of complex mixtures and in-
troduces developing methods for the health assessment
of microorganisms that are targeted for environmental
release.
70
-------
Development and Evaluation of Composting Techniques for Treatment of Soils
Contaminated with Hazardous Waste
John A. Giaser, Carl L. Potter, and Joseph B. Farrell
Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
Edward D. Kennedy, Jeffrey J. McCormack, and Michaei Najar
International Technology Corporation, Cincinnati, OH
This research investigates the potential for composting
systems to remediate soils contaminated with organic
pollutants. Significant progress in optimizing conditions
and applying the power of biotechnology to large-scale
soil compost systems will require a working under-
standing of the processes and mechanisms involved.
Current commercial compost operations constitute
black-box systems where optimization is approached on
a hit-or-miss basis using scant process-related informa-
tion. This research is designed to characterize physical
and microbiological changes during composting, identify
the stage of the composting process that yields the
greatest destruction of contaminants, and determine
mass balances of pollutant biotransformation.
The closed-vessel bench-top compost reactor design for
this project was selected from many options to mimic a
static compost pile. Considerable project design effort
was invested in identification of materials-handling
equipment and optimization of materials-handling op-
erations. Since toxic soils will be handled during reactor
loading, dust management procedures were carefully
developed to eliminate spread of contaminants.
The composter consists of a modified 55-gal drum with
five thermal wells extending 9 in. into the reactor core.
Since heat production may be highly variable tfiroughout
the reactor, thermal wells are vertically spaced 5 in.
apart to allow temperature measurements at various
depths of the compost mixture. A screen located 4 in.
above the reactor bottom supports the compost pile, and
air forced through a 1-in. inlet port near the reactor
bottom flows upward through the mixture. Increased air
flow can be used to cool the reactor and help control
compost temperature during thermogenesis. Atop sam-
ple port enables collection of compost core samples for
analysis. A draw is provided on the bottom to permit
removal of excess liquid and collection of ieachate sam-
ples for chemical analysis. The reactor is tipped on its
side and rolled on a drum roller every 24 hours to
mix the compost and rupture anaerobic pockets. Based
on a prototype reactor investigation, a field of seven
stainless steel reactors wilt be constructed for a treata-
bility study using soil contaminated with creosote
components.
Investigation of optimal composting conditions for de-
struction of soil contaminants involves determination of
suitable soil types, moisture content, co-compost mate-
rials (e.g., wood chips, sawdust com cobs), and nutri-
tional adjustments (e.g., carbon/nitrogen ratio). The
study also seeks to identify metabolically active micro-
bial species and possible mechanisms of hazardous
chemical transformation. Since it is difficult to trace all
metabolites produced during the compost process, the
end product will be bioassayed to assess the extent of
detoxification by compost microorganisms. The finished
compost soil will be evaluated for toxicity by Ames Sal-
monella mutagenicity, earthworm toxicity, and seed ger-
mination tests.
71
-------
Engineering Optimization of Siurry Bioreactors for Treating Hazardous Wastes in
Soil and Sediments
John Glaser, Paul McCaufey, and Edward Opatken
Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
Majid Dosani, Jennifer Piatt, and Diane Roush
IT Corporation, Cincinnati, OH
The operation and performance of soil slurry bioreactors
are not well understood despite widespread use of these
reactors in the field. This project's purpose is to enhance
understanding of the factors controlling slurry bioreactor
operation and performance using fundamental mechani-
cal and treatability studies at bench and pilot scale. A
systematic plan to evaluate slurry bioreactors and the
required parameters for their optimal use is being devel-
oped by this research program. The information gained
from these studies should be useful in improving treat-
ment effectiveness In field applications.
Current research efforts are centered on developing and
selecting suitable bench- and pilot-scale bioreactor de-
signs and on preparing soil slurry. The bench-scale
bioreactor selected is an original design, conical in
shape with a round bottom. Three sample ports are built
on one vertical side. The entire unit is airtight with four
access ports configured on the lid to allow sealed entay
of the air inlet and outlet tube and impeller shaft. The
impeller is driven by a variable-speed motor. The fourth
port on the lid is larger than fie other ports and is used
for slurry charging. Soil slurry preparation includes
screening, presium'fication, soil particle size selection
(by hydrocyclone), and treatment evaluation in either the
bench- or pilot-scale bioreactors.
Bench-scale slurry reactors are being used to develop
a more complete understanding of the physical, chemi-
cal, and biological factors controlling optimal operation.
Bench- and pilot-scale reactors will yield information on
solids loading methods, optimal residence times, opti-
mal solids loading as a function of soil type, pollutant
mass balances, and effectiveness of nutrient and micro-
organism additions. As bench-scale bioreactor perform-
ance information becomes available, it will be used to
configure the larger pilot-scale bioreactors.
72
-------
Biotreatabiiity of a Vadose Zone Soil Contaminated with Dioctyl Phthalate
Don H. Kampbell
Robert S. Kerr Environmental Research Laboratory, U.S. Environmental Protection Agency, Ada, OK
Dennis D. Fine and Jerry W. Anderson
ManTech Environmental Technology, Inc., Robert S. Kerr Environmental Research Laboratory, Ada, OK i
Chemical spillage and waste disposal during'40 years
of manufacturing vinyl wallcovering had contaminated
portions of a field site with various organic chemicals,
predominantly xylene and dioctyl phthalate. Laboratory
treatability studies were conducted on core samples
obtained from the site. The objective was to determine
whether natural biodegradation processes would reme-
rfate the site. Nine core samples were collected at three
locations in 3-ft increments from site locations of low
contamination, moderate contamination, and high con-
tamination. Analysis of a nine-sample composite
showed that 95 percent of organic compound contami-
nation was dioctyl phthalate. All core samples contained
sufficient nitrogen and phosphorus to support viable soil
microbiological processes. Bacterial cell counts ranged
from 7 x 10s to 53 x 108 cells/g for all cores, but were
highest in moderately contaminated samples. The same
trend was apparent with dehydrogenase activity. Dioctyl
phthalate concentration means were 6, 101, and 333
mg/kg soil for low, moderate, and high contamination,
respectively.
Equal amounts of core material on a 32 g dry-weight
basis were .contained in capped, airtight 160-mL glass
bottle microcosms. Consumption of headspace oxygen
and carbon dioxide generation was measured during a
24-day, 22"C incubation period. Total mean oxygen con-
sumption was 10,77, and 66 percent of the 20.9 percent
oxygen in headspace air for low, moderate, and high
contamination, respectively. Headspace carbon dioxide
for the same microcosms was 1.6,12.3, and 13.1 per-
cent The rate of dioctyl phthalate biodegradation in
laboratory microcosm bottles was approximately
0.1 g/kg soil/day for the contaminated cores. Extrapola-
ting from laboratory data rates and assuming ideal field
conditions, the predicted total time to naturally cleanse
the site to less than 10 ppm phthalate would be 3 years
and 10 years, respectively, for the moderate and high
contaminated areas. Biodegradation of dioctyl phthalate
was enhanced by the presence of xylene in a separate
microcosm study.
73
-------
Section 5
Process Research
Process research involves isolating and identifying microorganisms that carry out
biodegradation processes and the environmental factors affecting these processes.
In this way, researchers establish the building blocks of new biosystems for treat-
ment of environmental pollutants in surface waters, sediments, soils, and subsur-
face materials. Painstaking experimentation and thorough evaluation are critical at
this level of research, since a firm scientific basis can facilitate the scaling up of a
promising bioremediation method or technology.
In one study, researchers used contaminated aquifer material from a wood preserv-
ing site in Florida to model the degradation kinetics of heterocyclic aromatic
compounds as carried out by indigenous microorganisms. This research concerned
bioremediation of creosote and pentachlorophenol, as did a study that achieved
dechlorination of phenolic compounds using an inoculum from the Hudson River.
Biodegradation of paper-milling waste was the context for research into tite use of
5-chlorovanillate as a substrate, which was chosen because it contains side groups
representative of those present on aromatic chlorinated compounds.
Research into aromatic amines, which are associated with the degradation of
munitions and other wastes, is focusing on whether indigenous enzymes can be
used to activate the capacity of soils and sediments to bind with this class of
contaminants. Two other studies, both of which are specific to munitions wastes,
have achieved an aqueous phase in the biotransformation of trinitrotoluene that
contains only small amounts of organic material and the biodegradation of dinitro-
toluene using an ethanol substrate in an anaerobic respirometer.
Preliminary data on the toxic effects of metals on naturally occurring microorgan-
isms suggest that metal type, aqueous metal concentration, and organic carbon
content may affect the transformation of chlorophenols in anoxic sediments. In
another study, tests on microcosms inoculated with aquifer material indicated a
dechlorination pattern for dibenzo-p-dioxins/dibenzofurans (PCDD/PCDF).
Process-scale studies were conducted on sulfur-based autotrophic denitrification
as a means of rendering potable water from the nitrate-contaminated ground water
basin underlying Orange County, California. Elsewhere, researchers used pH-de-
pendent Monod kinetics in their investigation of the production and movement of
gaseous products in methanogenic biofilms.
Biofiltration, considered a promising and economical biological treatment technol-
ogy, is the focus of two other studies. In one, gene probes, culturing techniques,
and DNA amplification fingerprinting is being used to identify and compare microbial
communities on various biofilters. In the other, experiments on the biodegradation
capacity of aerobic biofilters of different configurations were performed using
several air-entrained compounds.
75
-------
A study supporting the development of sequential anaerobic-aerobic landfarming
and composting technologies is under way. By monitoring methanogenic master
culture reactors operating at steady state, researchers expect to obtain fundamen-
tal information on reaction kinetics.
In the symposium's poster session, researchers presented data from laboratory
experiments on the degradation of polychlorinated biphenyls at a Superfund site
near Houston, Texas.
A second study examined whether the successful use of microbial isolates to
enhance bioremediation at wood preserving sites in the United States would be
effective for soils elsewhere in the world.
Tfi
-------
Methanogenic Degradation Kinetics of Nitrogen and Sulfur Containing
Heterocyclic Aromatic Compounds In Aquifer-Derived Microcosms
E. Michael Godsy and Donald F. Goerlltz
U.S. Geological Survey, Menio Park, CA
Dunja Grbid-Galid
Department of Civil Engineering, Stanford University, Stanford, CA
The fate of nitrogen and sulfur containing heterocyclic
aromatic compounds in subsurface environments is
controlled by various transport and transformation proc-
esses. The most important, but least understood, proc-
ess affecting ground water quality is biotransformation
of these compounds by indigenous microorganisms.
Heterocyclic compounds are frequently encountered in
the environment because they are major constituents of
both fossil and synthetic fuels and many pesticide mix-
tures (e.g., creosote).
This study presents the Monod-decay kinetics for the
conversions of benzothiophene, oxindole, 2(1H)-quino-
linone, and 1 (2H)-isoquinolinone to methane and car-
bon dioxide by a complex aquifer-derived methanogenic
microbial consortium. The stoichiometric oxidation of
indole, quinoline, and isoquinoline to the persistent com-
pounds oxindole, 2(1H)-quinolinone, and 1 (2H)-isoqui-
nolinone, respectively, requires an input of energy. A
thermodynamic explanation is presented to rationalize
the persistence of the oxidized intermediates.
Background
The source of contaminated aquifer material is located
adjacent to an abandoned wood-preserving plant within
the city limits of Pensacola, Florida (1). The wood-pre-
serving process consisted of steam pressure treatment
of pine poles with creosote and/or pentachlorophenol
(PCP). For more than 80 years, large but unknown
quantities of waste waters, consisting of extracted mois-
ture from the poles, cellular debris, creosote, PCP, and
diesel fuel from the treatment processes, were dis-
charged to unlined surface impoundments that were in
direct hydraulic contact with the sand-and-gravel aquifer.
To model the degradation kinetics, laboratory micro-
cosms were prepared in 4-L glass bottles. The micro-
cosms contained approximately 3 kg of aquifer material
anaerobically collected from the approximate centroid of
the active methanogenic zone (1). Compounds of inter-
est were added to a mineral salts solution (1) at concen-
trations similar to those present in the aquifer.
Subsamples for substrate utilization determinations
were removed from the microcosms at approximately
3-day intervals and analyzed by high pressure liquid
chromatography (HPLC) and verified when necessary
by gas chromatography (GC)/mass spectrometry (MS).
Concentrations of methane and carbon dioxide in the
microcosms were determined by GC (1). Total biomass
concentrations in the microcosms at the onset and at the
end of incubations were determined by total protein
analysis using the Coomassie brilliant blue staining
technique of Bradford (2) as described by Gdili (3).
Substrate depletion curves were fitted to the Monod sub-
strate utilization and growth with decay equations (4):
dS mmXaS
dt = Y{Ks+S)
mmX,S
dXa
dt Ks+S
where:
|im = maximum specific growth rate
Ks = half-saturation constant, mg L"1
V= yield coefficient, mg cells per mg substrate
utilized
S = substrate concentration at time t, mg L*1
Xa = active biomass at time t, mg L'1
kd = specific bacterial decay rate, day'1
The method of Marquardt (5) was used for the determi-
nation of parameter values and kj that best fit
the experimental substrate depletion data by minimizing
77
-------
the residual sum of squares using NLR (non-linear re-
gression) techniques.
Results and Discussion
The bacterial substrate utilization for all of the com-
pounds tested was modeled successfully using the
Monod-decay equations with the exception of the initial
oxidation of indole, quinoline, and isoquinoline. These
compounds were stoichiometrically oxidized with the
intermediates persisting for particularly long, variable,
and unreproducible times from the initial oxidation to
final mineralization to methane and carbon dioxide (up
to 184 days elapsed after the oxidation of isoquinoline
and the start of methanogenesis). The mass balances
on the degradation of benzothiophene and the oxidized
nitrogen intermediates were 86.4 percent of theoretical
methane and carbon dioxide production for benzothio-
phene, 87.6 percent for oxindoie, 91.7 percent for 2(1 H)-
quinolinone, and 88.5 percent for 1(2H)-isoquinolinone.
A stable oxidized intermediate similar to oxindoie or
2(1 H)-quinolinone that would be formed from the oxida-
tion of benzothiophene to 2-hydroxybenzothiophene
was not observed in this study or in a previous study (6).
The oxidation of the nitrogen heterocycles is endergonic
and must be coupled with some other reaction as shown
in Table 1; however, the nature of tfie couple is not clear
at this time. The reactions are endergonic even if the
reactions presented in Table 1 are coupled to methane,
carbon dioxide, and/or H2 production; however, the
overall conversion of the nitrogen sulfur organic (NSO)
compounds to methane and carbon dioxide provides a
negative Gibbs free energy (Table 2). The electrons
produced during the oxidation of parent compounds are
very likely being used as a source of reducing power,
perhaps as a source for the reduction of other aromatic
ring compounds present in complex mixtures such as
the contaminated ground water (e.g., the reduction of
phenol to cyclohexanone). In the laboratory micro-
cosms, where indole, quinoline, or isoquinoline are the
only major carbon and energy sources, the complete
oxidation of the parent compound may supply neces-
Table 1. Glbbs Free Energy Changes during Oxidation of
Nitrogen Compounds at 25°C and pH of 7
AGUjq, kJ mol'1
Indole to Oxindoie
168
CgH7N+ HfcO -~ C0HtNO+ 2 H* + 2 e~
Qulnollne to 2(1H)-Quinolinone
84
CjHyN + Hj,0 -» CgHrNO + 2 H* + 2 e"
Isoqulnolina to 1(2HHsoqulnollnone
56
CgHyN + HaO ->0^0+2^ + 26-
AGpaq calculated using values from Stull et al. (7) or estimated using
the method at Jobak in Reid et al. (8) for compounds not found in
the literature.
Table 2. Glbbs Free Energy Changes during Methanogenesis
of NSO Compounds at 25°C and pH of 7
AG^jq) kJ mol'1
Indole to CH4 and C02
C8H7N + 7H20 + h^-»3.5C02 + 4.5CH4+NHJ -205
Quinoline to CH 4 and CO?
C^N + a HjO + HT -> 4 COz + 5 ch4 + NHJ -245
Isoquinoline to CH, and C05
C9H7N + B HzO + H+ -~ 4 COj + 5 CH4 + NHJ -231
Benzothiophene to CH4 and C02
CgHgS + 7 HzO -~ 3.5 COz + 4-5 ch4 +
0.5 HzS + 0.5 HS" + 0.5 H* -185
sary reducing power required for the subsequent ring
reduction and ring deavage.
Monod-decay kinetic constants for the anaerobic degra-
dation of sulfur and oxidized nitrogen containing hetero-
cyclic compounds have not been reported in the
literature under methanogenic conditions as exist in the
aquifer at the study site (Table 3). The kinetic constants
for all of the compounds are quite similar and may
represent the same rate-limiting bacterial population;
however, no statistical test for this hypothesis exists
when the parameters are generated by NLR techniques
(9). Although acetate was not measured in this study, in
a previous study of the contaminated ground water at
the research site (1), acetate concentration increased
concomitantly with the degradation of individual pheno-
lic and nitrogen and sulfur heterocyclic compounds, sug-
gesting that acetate utilization was a rate-limiting step.
This limitation may explain why the values for n™ and Ks
are essentially the same for each of the compounds
tested. The values of the kinetic parameters agree with
previously published values for acetate utilization by
well-acciimated acetate-fed enrichment cultures of
methanogenic bacteria (10).
The apparent low biomass yield suggests that the mi-
crobial community from this ground water environment
has adapted to low nutrient conditions by using a major
portion of the available energy for maintaining cellular
integrity in a relatively hostile environment (11). It is
unlikely that other explanations (i.e., inefficiency at cap-
turing the free energy available or storing carbon as
intracellular storage products) would account for the low
Y values.
The biomass yields obtained by protein determination
for the compounds tested were compared to theoretical
values calculated on thermodynamic principles (12). Ex-
perimental and theoretical values of Y for the com-
pounds tested are given in Table 4. The theoretical
values are greater than the values obtained from the
78
-------
Table 3. Kinetic Constants Determined for Methanogenesis of the Heterocyclic Compounds Tested ± 95 percent Confidence
Intervals
m» (lay1)
Ks (mg L"1)
*„(day')
Y (mg mg"1)*
Benzothiophene
0.117 ±0.136
0.80 ± 5.06
0.000 ±0.129
0.025
Oxindole
0.160 ±0.002
1.10 ±2.10
0.000 ±0.012
0.029
2(1 H)-Quinolinona
0.089 ± 0.058
11.41 ± 0.06
0.000 ± 0.049
0.033
1 (2H)-lsoquinolinone
0.099 ± 0.342
5.00 ± 22.0
0.000 ± 0.027
0.035
a Values for V are determined from average of three protein determinations.
Table 4. Experimental and Theoretical Y Values for the
Compounds Tested In Microcosms
Compound
Experimental Y
(mg mg'V
Theoretical Y
(mg mg"1)
Benzothiophene
. 0.025
0.14
Oxindole
0.029
0.35
2(1 H)-QuinoUnone
0.033
0.21
1 (2H)-)soquinolinone
0.035
0.21
* Based on the average of three protein determinations each before
and after growth.
measured protein increase in the microcosms by ap-
proximately an order of magnitude.
The bacterial decay term in the biomass equation
is apparently not required to describe substrate utiliza-
tion and/or biomass increase in the batch growth micro-
cosms. The values determined by NLR are such that
\xm» kj, and as a result can be neglected. NLR
analyses of the compound disappearance data using
Monod equations without the decay term resulted in
essentially the same kinetic constants. Using the Monod
equations without decay should, however, alleviate the
problem of increased uncertainty associated with fitting
three parameters versus two.
References
1. Godsy, E.M., D.F. Goerlitz, and D. Grbid-Galid
1992. Methanogenic biodegradation of creosote
contaminants in natural and simulated ground-
water ecosystems. Ground Wat. 30: 232-242.
2. Bradford, M.M. 1976. A rapid and sensitive method
for the quantitation of microgram quantities of pro-
tein utilizing the principle of protein dye-binding.
Analyt Biochem. 72:248-254.
3. Galli, R. 1987. Biodegradation of dichloromethane
in waste water using a fluidized bed bioreactor.
Appl. Microbiol. Biotechnol. 27: 206-213.
4. Monod, J. 1949. The growth of bacterial cultures.
Ann. Rev, Microbiol. 3: 371-394.
5. Bard, Y. 1974. Nonlinear parameter estimation.
Academic Press, New York.
6. Godsy, E. M., and D. Grbid-Galid. 1989. Biodegra-
dation pathways for benzothiophene in methano-
genic microcosms, pp. 559-564. In: Mallard, G.E,
and S.E. Ragone, eds. U.S. Geological Survey
Toxic Substances Hydrology Program—Proceed-
ings of the Technical Meeting, Phoenix, Arizona,
September 26-30, 1988: U.S. Geological Survey
WRI Report 88-4220.
7. Stull, D.R., E.F. Westrum, Jr., and G.C. Sinke. 1969.
The chemical thermodynamics of organic com-
pounds. Marcel Dekker, New York.
8. Reid, R.C., J.M. Prausnrtz, and B.E. Poling. 1985.
The properties of gases and liquids. McGraw-Hill
Co., New York.
9. Bates, D.M., and D.G. Watts. 1988. Nonlinear re-
gression analysis and its applications. Wiley Inter-
science, New York.
10. Wang, Y.T., H.D. Gabbard, and P.C. Pai. 1992. In-
hibition of acetate methanogenesis by phenols. J.
Environ. Eng. 117: 487-500.
11. Battley, E.H. 1987. Energetics of microbial growth.
Wiley Interscience, New York.
12. McCarty, P.L 1971. Energetics and bacterial
growth, pp. 495-531. In: S.D. Faust and J.V. Hunter,
eds„ Organic compounds in aquatic environments,
Marcel Dekker, Inc., New York.
k.
79
-------
Anaerobic Degradation of Halogenated and
Nonhalogenated Phenolic Compounds
M.M. H§ggblom, M.D. Rivera, and L.Y. Young
Center for Agricultural Molecular Biology
Rutgers University, New Brunswick, NJ
J.E. Rogers
U.S. Environmental-Protection Agency, Athens, GA
Pentachlorophenol (PCP) and creosote have been used
widely as wood preservatives. Improper use and acci-
dental spills can cause serious contamination of soil and
water at wood-preserving sites. Approximately 10 per-
cent of creosote consists of phenolic compounds. Given
their water solubility, these phenolics as well as PCP can
leach from soil with subsequent transport to ground
water.
To better understand the fate of mixed creosote and
PCP wastes in soil and ground water, interactions be-
tween the anaerobic dechlorination of PCP and the
biodegradation of phenol or methylphenols (ortho,
meta-, and para-cresol) were examined. Methanogenic
enrichment cultures were Rstablished and fed phenol or
one of the methylphenols at a concentration of 500 (xM
(see 1 and 2 for methodology). In addition, a parallel
series of cultures was established to which 50 |iM PCP
was added together with each of the phenolic com-
pounds. The source of inoculum was sediment from the
Hudson River, miie point 145, near Albany (3).
Our results indicate that PCP was reductively dechlori-
nated via 2,3,4,5-tetra- and 3,4,5-trichlorophenol to
3,5-dichlorophenol within 60 to 100 days, with 3,4-di-
chlorophenol observed as a minor product These two
dichlorophenols were not substantially dechlorinated
further within 160 days. The addition of propionate, phe-
nol, or cresoi did not stimulate dechlorination, and thus
did not appear to serve as electron donor for reductive
dechlorination.
In the absence of PCP, phenol, m-cresol, and p-cresol
(500 nM) were used within 30 days, while degradation
of o-cresol took more than 60 days. Methane was pro-
duced in excess of background controls, indicating that
the phenolics were mineralized. On the other hand, the
presence of 50 |iM of PCP completely inhibited degra-
dation of the phenolic compounds and methanogenesis.
Degradation of phenol and the cresols took place only
after PCP had been dechlorinated to 3,5-dichlorophe-
nol. Thus, in the presence of PCP, a 60- to 100-day lag
was observed before degradation of phenol and cresols
could proceed.
Toxicity of PCP and its dechlorination products was
substantiated with experiments in which PCP, 2,3,4,5-
tetrachlorophenol, and 3,4,5-trichlorophenol were
added at concentrations of 10 to 100 jiM to active phe-
nol-degrading cultures. PCP and the tetra- and trichlo-
rophenols at a concentration of 10 |iM and above
exhibited toxicity, while 10 to 25 jiM of 3,5-dichlorophe-
nol did not inhibit degradation of phenol. Therefore, high
concentrations of PCP might inhibit the anaerobic deg-
radation of creosote-derived phenolic compounds in
ground water and soil.
References
1. Haggblom, M.M., M.D. Rivera, I.D. Bossert, J. Ro-
gers, and LY. Young. 1990. Anaerobic degradation
of para-cresol under three reducing conditions. Mi-
crobial Ecology 20:141-150.
2. Higgblom, M.M., M.D. Rivera, and L.Y. Young. 1993.
Effects of auxiliary carbon sources and electron ac-
ceptors on methanogenic degradation of chlorinated
phenols. Environ. Toxicol. Chem. 12. (in press)
3. Haggblom, M.M., M.D. Rivera, and L.Y. Young. 1993.
Influence of alternative electron acceptors on the
anaerobic biodegradability of chlorinated phenols
and benzoic adds. Appl. Environ. Microbiol. 59. (in
press)
k.
60
-------
Anaerobic Biodegradation of 5-Chlorovanlllate as a Model Substrate for the
Bloremediation of Paper-Milling Waste
Barbara R. Sharak Genthner
Technical Resources, Inc., Gulf Breeze, FL
The anaerobic biodegradation of 5-chlorovanillate (5CV;
5-chloro-4-hydroxy-3-methoxybenzoic acid) was inves-
tigated. 5CV was selected as a model compound for
studying the biodegradation of paper-milling effluents
because it contains the methoxy, chloro, and carboxyi
side groups representative of those present on aromatic
chlorinated compounds released in paper-milling efflu-
ent. Using sediment from a river receiving discharge
from a paper-milling plant, an anaerobic enrichment
culture was developed that degraded 5CV. The major
pathway of 5GV degradation in this enrichment culture
was concluded to be stepwise demethoxylation to 5-
chloroprotocatechuate (5CP; 5-chloro-3,4-dihydroxy-
benzoic acid), decarboxylation to 3-chlorocatechol
(3CC; 3-chloro-1,2-dihydroxybenzene), and dechlorina-
tion to catechol, which was completely degraded (Fig-
ure 1). Dechlorination of XC was the rate-limiting step
of degradation (Figure 2).
COOH
SCMaravanBe vaiumpmiiicaMcMc >Cnk*t>eai«ct«* CMhI
Figure 1. Pathway for the complete degradation of 5-chlo-
rovanillie add in a primary freshwater sediment erv-
rlchment
Degradation of 5CV was investigated using a sequential
inoculation approach designed to avoid the formation of
3CC (Figure 3). In lieu of the 5CV enrichment described
above, the 5CV medium was initially inoculated with an
anaerobic bacterial co-culture that dechlorinates 3-chlo-
robenzoate (3CB). This co-culture had been previously
shown to sequentially dechlorinate and then de-
methoxylate 5CV to form protocatechuate, which was
not degraded further. Upon formation of proto-
catechuate, this culture was inoculated (indicated with
an arrow in Figure 3) with a vanillate-degrading bacterial
consortium that had been derived from the above-men-
500
•—# 5-chlorovanillate
O—~ 5-chloropratocatechuate
A—A 3-chlorocatechol
^^ catechol
t 250 ..
0
5
10
15
20
Time (days)
Figure 2. Anaerobic degradation of 5-chlorovanillate In a fresh-
water sediment enrichment
1000
9 9 S-chtorovanillatB
q q vaniWate
V—V protocatechuate
O—O catechol
g- 750
a.
5 500 '
250 ..
0
10
20
40
50
SO
Time (weeks)
Figure 3. Anaerobic degradation of 5-chlorovanillate using se-
quential inoculation.
tioned 5CV enrichment and is known to perform the
degradative steps downstream from dechlorination (i.e.,
sequential demethoxylation to protocatechuate, decar-
boxylation to catechol, and complete degradation of
catechol). As expected, protocatechuate was decar-
boxylated to catechol, which then was rapidly degraded.
81
-------
Under these conditions, 5CV was completely degraded
without formation of 3CC, as predicted (Figure 4). The
Co-cuttura 3162
a- x ="3°" =°2
OH OH OH
S-CWofowHAe Vantac ProtoeawAtMC Ctisctoi
Van4f Buiehmant
Figure 4. Pathway for th# complete degradation of 5-chlo
rovanillle acid using sequential inoculation.
time required for complete dechlorination increased,
however, from 4 to 17 weeks. Subsequent transforma-
tion of vanillate to protocatechuate took 29 weeks, in
contrast to less than 1 week in the original enrichment
in a sequential inoculation study currently under way,
the co-culture was adapted to degrading 5CV before
inoculation, which resulted in reducing the time required
for dechlorination and demethoxylation to 10 and 6
weeks, respectively. Inoculating the sequential experi-
ment with the vanillate consortium as soon as dechlori-
nation is complete might reduce complete degradation
of 5CV to less than 3 months—something that is cur-
rently under investigation.
To further investigate the anaerobic degradation of 5CV,
the 3CB co-culture and the vanillate consortium used in
the sequential inoculation experiment were inoculated
into several media to enrich for bacterial species per-
forming the individual transformation steps. The 3CB
co-culture was inoculated into 5CP and vanillate me-
dium to enrich for dechlorination and demethoxylation,
respectively. The vanillate consortium was maintained
in vanillate medium and inoculated into guaiacol (2-
methoxyphenol) medium to enrich for the demethoxyla-
tion, protocatechuate medium to enrich for the
decarboxylation, and catechol medium to enrich for the
catechol degradation. Transformation of target com-
pounds in these enrichment cultures was followed
closely using high pressure liquid chromatography
(HPLC) analyses. Immediately upon completing the
transformation of interest, the cultures were passed to
fresh medium. Complete degradation of vanillate takes
less than 1 week, while demethoxylation and decar-
boxylation take 3 and 4 days, respectively.
While studying the capacity of the co-culture to dechlori-
nate 3CB, a sulfate-reducing bacterium was isolated
and "identified as a new species, Desutfomicrobium es-
cambium. This bacterium is responsible for dechlorina-
tion of 3CB and currently is being tested for
dechlorination of 5CV and 5CP and demethoxylation of
vanillate. Isolation and identification of the bacteria able
to decarboxylate protocatechuate and degrade catechol
currently are under way.
Anaerobic degradation of chloroaromatic compounds
may prove useful in bioremediation of paper-milling ef-
fluents. An appropriate toxicity test currently is being
devised to assess the toxicity of various untreated pa-
per-milling effluents. Effluents found to be toxic will be
"bioremediated" by inoculating culture media that con-
tains increasing percentages of effluent with the bacte-
rial cultures developed in these studies. Cultures will be
monitored using HPLC analyses to determine biotrans-
formation and biodegradation of the aromatic com-
pounds in the effluents. Toxicity tests will be repeated on
filter-sterilized supernatant from these cultures to deter-
mine the effect of this treatment on toxicity.
-------
Bioremediation of Soils and Sediments Contaminated with Aromatic Amines
Eric J. Weber
Environmental Research Laboratory, U.S. Environmental Protection Agency, Athens, GA
David L. Spidle
Technology Applications, Inc., Athens, GA
Kevin A. Thorn
U.S. Geological Survey, Arvada, CO
Aromatic amines comprise an important class of envi-
ronmental contaminants. Concern over their environ-
mental fate arises from the toxic effects that certain
aromatic amines exhibit toward microbial populations
and reports that the compounds can be toxic and/or
carcinogenic to animals. Aromatic amines can enter the
environment from the degradation of textile dyes, muni-
tions, and numerous herbicides, including the pheny-
lureas, phenylcarbamates, and acyianilides. Because
these chemicals are synthesized from aromatic amines,
loss of aromatic amines to the environment also may
result from production processes or improper treatment
of industrial waste streams. The high probability that
aromatic amine contamination of soils, sediments, and
ground water aquifers will occur necessitates the
development of in situ bioremediation techniques for
treatment
This study has demonstrated by sequential extraction of
sediments treated with uC-aniline and 15N-NMR analy-
sis of dissolved organic matter treated with 1sN-aniline
that aromatic amines become covalently bound to the
organic fraction of soils and sediments through oxidative
and/or nucleophilic coupling reactions (1). It generally is
accepted that, once covalently bound, the bound resi-
due is less bioavaiiable and less mobile than the parent
compound. Thus, procedures that enhance tfie irre-
versible binding of aromatic amines to soil constituents
could potentially serve as a remediation technique (2).
The general approach to bioremediating soils and sedi-
ments contaminated with aromatic amines is to use the
naturaily occurring phenol oxidases in soil that are
thought to play a significant role in catalyzing the forma-
tion of bound residues (3). The binding capacity of sedi-
ments and soils for aromatic amines can be increased
by the activation of indigenous enzymes (e.g., peroxi-
dase) through the addition of hydrogen peroxide and
readily oxidizable substrates such as phenolic com-
pounds. Oxidation of the phenolic substrates by acti-
vated peroxidase results in the in situ generation of
oxygenated polymers, which provide additional binding
sites for aromatic amine pollutants (4).
Results and Discussion
Initial experiments have been conducted with pond sedi-
ments that were collected in the Athens, Georgia, area.
Concentration-dependent studies with aniline in sedi-
ment-water slurries (5 percent solids) demonstrated
that, at low aniline concentrations (1 x 10"6 M), approxi-
mately 90 percent of the aniline was irreversibly bound
in 48 hours; however, as the initial concentration of
aniline was increased, the fraction of aniline that was
bound became increasingly less, suggesting that there
was either a kinetic limitation to covalent binding or a
limit to the number of reactive binding sites. In an at-
tempt to increase the binding capacity of these systems,
the sediment was treated with peroxidase, hydrogen
peroxide, and ferulic acid, a readily oxidizable phenolic
compound. In a typical experiment, a sediment-water
slurry (5 percent solids) was first treated with aniline at
an initial concentration of 5.5 x 10"5 M. Within 4a hours,
only 15 percent of the aniline had bound to the sediment
Upon subsequent treatment of the sediment-water
slurry with activated peroxidase and ferulic acid, the
aqueous concentration of aniline decreased to below
detectable levels within several hours.
In other studies, 1SN-NMR is being used to make a
comparative study of phenoloxidases concerning: 1) the
types of covalent linkages formed in their presence; and
S3
-------
2) the degree to which they increase binding capacity.
Initial studies have focused on horseradish peroxidase
and tyrosinase. In both peroxidase- and tyrosinase-
catalyzed binding of 5SN-aniline to fulvic acid, INEPT and
ACOUSTIC 1SN-NMR spectra exhibited resonances that
indicated the formation of imine, anilide, anilino-qui-
none, and anilino-hydroquinone covalent linkages. Al-
though the distribution of covalent linkages was quite
similar for the two enzymes, peroxidase proved to be a
much more effective catalyst for binding; the binding
capacity of the fulvic acid was significantly greater for
the peroxidase-catalyzed reaction than for the reaction
catalyzed by tyrosinase.
References
1. Baughman, G.L, E.J. Weber, R.L. Adams, and M.S.
Brewer. 1992. Fate of colored smoke dyes. U.S.
Department of the Army, Frederick, MD.
2. Bollag, J.-M. 1992. Decontaminating soil with en-
zymes. Environ. Sci. Technol. 26:1876-1881.
3. Bollag, J.-M., and W.B. Bollag. 1990. A model for
enzymatic binding of pollutants in the soil. Intern. J.
Environ. Anal. Chem. 39:147-157
4. Berry, D.F., and S.A Boyd. 1985. Decontamination of
soil through enhanced formation of bound residues.
Environ. Sci. Technol. 19:1132-1133.
84
-------
Anaerobic Biotransformation of Munitions Wastes
Deborah J. Roberts
Department of Civil and Environmental Engineering, University of Houston, Houston, TX
Stephen Funk, Don L Crawford, and Ronald L Crawford
Center for Hazardous Waste Remediation Research, University of Idaho, Moscow, ID
2,4,6-Trinitrotoluene (TNT), hexahydro-1,3,5-hinitro-
1,3,5-triazine (RDX), and octahydro-1,3,5,7-tetranitro-
1,3,5,7-tetraazocine (HMX) are commonly found
together in soil where munitions loading and packing
operations have.contaminated large amounts of soil in
lagoons. These contaminants leach into the surrounding
soil and have been known to contaminate ground water
(10). Although this disposal approach is no longer prac-
ticed, large quantities of soils contaminated in this man-
ner still must be cleaned up. Incineration is the only
proven technology available for the remediation of soils
contaminated with explosives. Unfortunately, this tech-
nology is expensive for small locations, and only 40
percent less expensive for especially large sites, with
estimates approaching $800/ton (12).
The approach described in this study is based on the
similar application of anaerobic treatment procedures
that have been shown to be effective with the nitroaro-
matic compound dinoseb (2-seobutyl-4,6-dinitrophe-
nol) (6,7,12) to a soil contaminated with 12,OCX) mg/kg
TNT, 3,000 mg/kg RDX, and 300 mg/kg HMX as the
major contaminants.
Background
The literature on the biological degradation of TNT and
other hazardous energetic nitroaromatic compounds re-
viewed by Kaplan (8) shows that under both aerobic and
anaerobic culture conditions the initial step in the meta-
bolism of nitroaromatic compounds is typically a reduc-
tion of the nitro constituents to amino groups. This
reduction usually proceeds in steps with the para-nitro
group being ttie most susceptible to reduction. The re-
duction of the para-nitro group appears to be nonspecific
and will be performed by most cells and some cell
extracts as long as growth is exhibited or sufficiently
reduced conditions are used (unpublished data). The
next reduction usually occurs at one of the ortho groups,
producing diaminonitrotoluene isomers. The reduction
of the third nitro group occurs only under anaerobic
conditions (11).
Under aerobic conditions the production of unstable
hydroxylamino intermediates during the reductive proc-
ess can lead to the formation of azo or azoxy linkages
with other intermediates, followed by dimerization or
polymerization (2). Under anaerobic conditions, the re-
ductions occur more rapidly and the hydroxylamino in-
termediates do not accumulate or have as much
opportunity to form linkages. The production of diami-
nonitrotoluenes and aminodinitrotoluenes does not
completely reduce the toxicity of . the TNT molecule
(5,11). Amerkhanova and Naumova (1) found that the
amino derivatives of TNT were less toxic to some organ-
isms but more toxic to other organisms. Complete reme-
diation of a TNT-contaminated soil would require not only
conversion of TNT to its amino derivatives, but the further
conversion of these derivatives to nontoxic products.
Although research using Phanerochaete chrysosporium
has shown microbial mineralization of 14C-ring-labeled
TNT to 14C02 in aqueous culture (3,4), complete inhibi-
tion of growth was found and no degradation of TNT was
observed when 0.02 percent (weight/volume) of soil was
present in the growth medium of P, chrysosporium (13).
This paper reports on the application of a procedure for
the strictly anaerobic microbial treatment of nitroaro-
matic-contaminated soils to the remediation of muni-
tions-contaminated soils. The soil examined in this study
was heavily contaminated with munitions compounds
and is representative of the types of soils found at.
munitions loading, handling, and packing (LHAP) sites.
The procedure used for these soils has effectively
treated soils contaminated with the herbicide 2-seo-bu-
tyl-4,6-dinitrophenol (dinoseb) in the laboratory (7) and
in field studies (6,7,12). Anaerobic metabolism was
found to occur in two stages. The first stage is a reduc-
-------
tive stage in which TNT is reduced to its amino deriva-
tives as described by most researchers. The second
stage begins after the reduction of tie third nitro group.
For this, the report describes the optimization of the
reductive stage of TNT metabolism. This optimization
leads to the removal of the toxic amino intermediates in
a short period of time.
Results
Anaerobic conditions were established in soil cultures
by providing potato starch to indigenous aerobes, which
then used all the oxygen, creating anaerobic conditions.
The redox potential in the cultures usually dropped from
initial values of +250 mV to approximately +50 mV within
1 to 2 days of incubation. Although the cultures were
anoxic, the nitrotoluenes kept the redox potential above
0 mV. The redox values remain constant at this value
until after the 4-amino-2,6-dinitrotoluene is removed
from the culture supernatant, when they then drop to
approximately -200 mV during the remainder of the
incubations. When grown under optimal conditions, the
cultures regularly demonstrated the reduction of TNT to
4-amino-2,6-dinitrophenol (4A26DNT) and then to 2,4-
cfiamino-6-nitrotoluene (24DA6NT) as shown in Figure
1. The reductive stage of the remediation of munitions-
contaminated soil to the elimination of 24DA6NT was
accomplished by about 25 days of incubation in the
optimized cultures. RDX was removed to below detec-
tion limits in ttie cultures in as little as 24 days of incu-
bation with no identifiable intermediates detected,
although no radiolabeled tracer studies have been per-
formed. Visual observation of these cultures showed the
presence of a fungal mat on the surface of the cultures
after the completion of the reductive stage of TNT Me-
tabolism.
A summary of the optimal conditions as determined fo
date for the first stage of munitions-contaminated soil
remediation is presented in Table 1. Although TNT re-
moval rates were more rapid at pH 8, a separate experi-
ment showed that up to 60 percent of the label from
14C-TNT was incorporated into nonfilterable material
(Table 2), which may be dimerized or polymerized azoxy
compounds as described by Kaplan (9). In a separate
experiment performed at pH 6, the majority of the label
from 14C-TNT was recovered as the intermediates
4A26DNT and 24DA6NT after 24 days of incubation.
Further experiments to define the pH optima revealed
that biological activity was optimal between pH 6.5
and 7. A 50 mM ammonium phosphate buffer (pH 7.0)
completely inhibited TNT reduction. This inhibition also
was seen in phosphate-buffered cultures receiving 50
Table t. Optimal Conditions for the Anaerobic Remediation
of Munitions-Contaminated Soils
Condition Optimal
Range
Comments
PH
6.5
6.5-7.0
Polymerization at
alkaline pH
Temp. (°C)
35
25-37
Buffer
50 mM KHjPO*
N supply
25 mM NH4
15-30
50 mM inhibits TNT
degradation
OTNT • RDX
« intermediate A
t Intermediate O
4A26DNT q 240A8NT
a Intermediate B
a intermediate E
Note: TNT, RDX, and 4A2ADNT are measured by
concentration, the remaining are measured by
peak area.
V
10000
8000
6000 |
o
«
8
Z
4000 I
2000
10 15
Time (d)
20
25
Figure 1. Blodegradatlon of TNT and TNT biotransformation products under optimized anaerobic conditions.
flfi
-------
Table 2. Carbon Label Distribution In the Aqueous Phase' (% of T=0 (W-sd))
Retained Fractions Total
Culture Conditions on Filter 4A26DNT 240A6NT 1-3 Recovery
Anaerobic pH 8 56.8 (22) 5.8 3.2 39.7 105.5
Anaerobic pH 6 nd 54 38.8 5.7 95.5
Aerobic 71 (10.6) 15.3 5.2 7.4 98.9
' Includes polar intermediates and volatile organic acids
nd m not detected
mM ammonium chloride. TNT reduction rates were
found to rise with increasing amounts of ammonium up
to 25 mM ammonium chloride (1.33 g/L), which was
optimal.
Summary and Conclusions
The research described here has allowed the optimiza-
tion of the first phase of the remediation of munitions
compounds from munitions-contaminated soils. This
phase of remediation has been shown to be complete
within 24 days of incubation. The completion of the first
or reductive, phase of TNT degradation from these soils
results in an aqueous phase that contains very little
organic material and coincides with the removal of RDX
from the supernatant. The resultant aqueous phase from
the completed first phase cultures has been shown to
support the growtti of fungi present in the soil. It ap-
pears that the toxicity of the cultures is reduced by the
removal of TNT, the amino-intermediates, and RDX.
Toxicological studies with completed first stage super-
natant are planned.
References
1. Amerkhanova, N.N., and R.P. Naumova. 1979.
Comparative study of the acute toxicity of alpha-
trinitrotoluol and products of its biodegradation for
hydrobionts. Biol. Nauki. 2:26-28.
2. Carpenter, D.F., N.G. McCormick, J.H. Cornell, and
A.M. Kaplan. 1978. Microbial transformation of 14C-
labeled 2,4,6-trinitrotoluene in an activated-sludge
system. Appl. Environ. Microbiol. 35:949-954.
3. Fernando, T., and S.D. Aust. 1991. Biodegradation
of munition waste, TNT (2,4,6-trinitrotoluene), and
RDX (hexahydro-1,3,5-trinitro-1,3,5-triazine) by
Phanerochaete chrysosporium, p. 214. In: Ameri-
can Chemical Society ed., Emerging Technologies
in Hazardous Waste Management 2. American
Chemical Society.
4. Fernando, T., J.A. Bumpus, and S.D. Aust. 1990.
Biodegradation of TNT (2,4,6-trinitrotoluene) by
Phanerochaete chrysosporium. Appl. Environ. Mi-
crobiol. 56:1666-1671.
5. Funk, S.B., D.J. Roberts, and R.A. Korus. 1992.
Physical parameters affecting the anaerobic degra-
dation of TNT in munitions-contaminated soil, Ab-
stract Q142. American Society for Microbiology,
92nd General Meeting, New Orleans.
6. Kaake, R.H., D.J. Roberts, S.B. Funk, R.L. Craw-
ford, and D. L. Crawford. 19921 On-site anaerobic
biological treatment of nitroaromatic-contaminated
soils. 18th Annual RREL Research Symposium,
Cincinnati, OH.
7. Kaake, R.H., D.J. Roberts, T.O. Stevens, R.L
Crawford, and D.L. Crawford. 1992. Bioremediation
of soils contaminated with 2-sec-butyl-4,6-dini-
trophenol (dinoseb). Appl. Environ. Microbiol.
58:1683-1689.
8. Kaplan, D.L. 1990. Biofransformation pathways of
hazardous energetic organo-nitro compounds, p.
155. In: Kamely, D., A. Chakrabarty, and G.S.
Omenn. eds. Biotechnology and biodegradation.
Portfolio Publishing Company, TX.
9. Kaplan, D.L. 1992. Biological degradation of explo-
sives and chemical agents. Curr. Opin. Biotechnol.
3:253-260.
10. Kaplan, D.L., and A.M. Kaplan. 1982. 2,4,6-trinitro-
toluene-surfactant complexes: Decomposition, mu-
tagenicity, and soil leaching studies. Environ. Sci.
Technol. 16:566-571.
11. Roberts, D.J., S.B. Funk, and R.A. Korus. 1992.
Intermediary metabolism during anaerobic degra-
dation of TNT from munitions-contaminated soil,
Abstract Q136. American Society for Microbiology,
92nd General Meeting, New Orleans.
12. Roberts, D.J., R.H. Kaake, S.B. Funk, D.L Craw-
ford, and R.L. Crawford. 1993. Field-scale anaero-
bic bioremediation of dinoseb-contaminated soils.
In: M. Levin and M. Gealt, ed., Biotreatment of
Industrial and Hazardous Wastes. McGraw-Hill
Publishing Co., New York.
13. Spiker, J.K., D.L. Crawford, and R.L. Crawford.
1992. Degradation of 2,4,6-trinitrotoluene (TNT) in
explosives-contaminated soils by the white-rot fun-
gus Phanerochaete chrysosporium: Influence of
TNT concentration. Appl. Environ. Microbiol.
58:3199-3202.
-------
Cometabolic Biodegradation of 2,4-Dinitrotoluene Using Ethanoi
as a Primary Substrate
Jiayang Cheng and Makram T. Suidan
University of Cincinnati, Cincinnati, OH
Albert D. Venosa
U.S. Environmental Protection Agency, Cincinnati, OH
2,4-Dinitrotoluene (DNT) is a major by-product of the
manufacture of 2,4,6-trinitrotoluene (TNT) and is com-
monly found in ammunition wastewaters. Because of its
toxic nature and large-scale use, 2,4-DNT is listed as a
priority pollutant by EPA (1). Numerous studies have
investigated the biodegradation of 2,4-dinitrotoluene (2-
5) and these studies suggest that 2,4-DNT is inhibitory
to biological freatment processes.
In this study, the cometabolic biodegradation process of
2,4-DNT with ethanoi as the primary substrate under
anaerobic conditions was investigated using an anaero-
bic respirometer. The inoculum for the respirometer was
cultivated in a continuously stirred tank reactor (CSTR)
fed with 2,4-DNT and ethanoi. The pH and the tempera-
ture of the CSTR were kept at 72 and 35°C, respec-
tively. The detention time in frie reactor was 40 days and
the influent 2,4-DNT concentration was 92 mg/L. The
CSTR was operated until steady-state conditions were
achieved and 2,4-DNT was completely degraded. Bio-
logical methane potential (BMP) tests with 2,4-DNT and
ethanoi as substrates were conducted in the anaerobic
respirometer at the same pH and temperature as in the
CSTR. The effects of the initial 2,4-DNT concentration
on the anaerobic biodegradation of 2,4-DNT and on the
pathway of the anaerobic biotransformation of ethanoi
were studied.
The degradation of 2,4-dinitrotoluene with different initial
2,4-DNT concentrations is shown in Figure 1. Higher
initial concentrations of 2,4-DNT are more inhibitory.
The presence of 2,4-DNT also inhibits the acidogenesis
of ethanoi. The transformation of ethanoi can be ex-
pressed as,
CH^CH!2OH+ HiO= CH3COO- + H* +Hz
AGP = 2.3 kcal (1)
Figure 2 shows that propionate and methane were
formed in the reactors where ethanoi was fed as the sole
substrate or when low initial 2,4-DNT was present.
These transformations are expressed as,
4 Hz + HCCTZ +hT = CHt + 3 HzO
AG0 = -32.Akcal (2)
3 Hz + CH^COO- + hT + HCOi =
CHzCHiCOCT + 2H20
AG° = -18.3/rca/ (3)
0.20
0.15
i
I
E 0.10
0.05
CVJ
0.00
" i —r~
¦¦ r • i T r 1
0 2 mg/L DNT -A
m 2 mgt DNT -fi
v 4 mg/L DNT -A
T 4 mg/L DNT -B
\ \
n 8 mg/L DNT -A
\ \
a 8 mg/L DNT -6
N. \ \
A 18 mg/L DNT -A
* 16 mg/L DNT -B
\ \\
« 32 mg/L DNT-A
Vh.\
V • 32 mg/L DNT -B
_i i _ i _ i
20 40
60 80
Tims (hre)
100
120
140
Figure 1. Anaerobic biodegradation of 2,4-DNT.
88
-------
0 100 200 300 400 500 600 700 800
Tima (hrs)
Figure 2. Ethanoi biotransformation with initial concentration of 2,4-DNT of 2 mg/L.
Ethanol -A
Ethanol-B
10
CH4 -A
CH4-B
5
0
Almost no propionate was formed, however, with high
initial 2,4-DNT concentrations (Figure 3).
Propionate was degraded to methane and carbon diox-
ide via acetate after all the previously formed acetate
was completely transformed to methane and carbon
dioxide.
CHsCHzCOO- + =
ch^coct + 3H2 + i-r + hcq*
ACP = 18.3/fca/ (4)
CHjPOCT + HzO =CHt + HCOi
ACP = -7.4/cca/ (5)
The BMP tests show that 2,4-dinitrotoluene with an
initial concentration as high as 32 mg/L was completely
degraded under anaerobic conditions. 2,4-DNT is self-
inhibitory, however, and inhibits the acidogenesis of
ethanol. The higher initial 2,4-DNT concentrations also
are inhibitory to the formation of propionate.
References
1. Keither, L.H., and W.A. Telliard. 1979. Priority pollut-
ants. I. A perspective view. Environ. Sci. Technol.
13:416-423.
2. Valii K., B.J. Brock, D.K. Joshi, and M.H. Gold. 1992.
Degradation of 2,4-dinitrotoiuene by the lignin-de-
grading fungus Phanerochaete chrysosporium. Appl.
Environ. Microbiol. 58:221-228.
3. Spanggord, R.J., J.C. Spain, S.F. Nishino, and K.E.
Mortelmans. 1991. Biodegradation of 2,4-dinitrotolu-
ene by a Pseudomonas sp. Appl. Environ. Microbiol.
57:3200-3205.
4. Lju, D., K. Thomson, and A.C. Anderson. 1984. Iden-
tification of nitroso compounds from biotransforma-
tion of 2,4-dinitrotoluene. Appl. Environ. Microbiol.
47:1295-1298.
5. McCormick, N.G., J.H. Cornell, and A.M. Kaplan.
1978. Identification of biotransformation products
froin 2,4-dinitrotoluene. Appl. Environ. Microbiol.
35:945-948.
6. Smith, D.P., and P.L. McCarty. 1989. Reduced prod-
uct formation following perturbation of ethanol- and
propionate-fed methanogenic CSTRs. Biotechnol-
ogy and Bioengineering 34:885-895.
--J. A I ..
o Ethanol -A
• Ethanol —B
CH4-A
CH4-B
Acetate-A
T Propionate -A
. Propionate -B
Time (hrs)
700 800
Figure 3. Ethanol biotransformation with Initial concentration of 2,4-DNT of 16 mg/L.
89
-------
Effects of Metals on Anaerobic Treatment Processes
W. Jack Jones
Environmental Research Laboratory, U.S. Environmental Protection Agency, Athens, GA
In Chul Kong
University of Georgia, Athens, GA
Recent interest in the use of bioremediation technology
for the cleanup of contaminated soils and sediments has
led to a greater understanding of the fate of a variety of
toxic organic compounds in natural environments. Most
of the research has focused on the biological transfor-
mation of a single organic contaminant in laboratory
microcosms using naturally occurring microbial inocula.
Yet, most polluted ecosystems and hazardous waste
sites often are contaminated with a mixture of organic
compounds as well as toxic inorganic wastes such as
metals. Although the transport and transformation of
organic compounds in natural environments such as
soils and sediments has been a topic of considerable
interest in recent years, little is known about the potential
toxicity of metals to naturally occurring microorganisms
and microbial processes of importance to the biotrans-
formation of organic compounds.
Soils and sediments are quite varied in composition but
generally consist of an array of mineral particles, organic
matter, microbial cells and debris, and inorganic solutes,
all of which are important in the formation of metal
complexes. Further, recent studies have indicated the
importance of microbial cell surfaces in the sorption and
immobilization of metals in natural ecosystems.
Results and Discussion
Ongoing studies at EPA's Environmental Research
Laboratory in Athens, Georgia, are investigating the ef-
fects of heavy metals on anaerobic microbial processes
on the biotransformation of organic compounds. As a
model system, the studies are focusing on the toxic
effects of metals on the reductive dechlorination of chlo-
rophenolic compounds in unadapted and chlorophenol-
adapted microbial communities from freshwater
sediments.
The onset, rate, and extent of biotransformation of sev-
eral mono-, di-, and trichlorophenols were examined
using unadapted and chlorophenol (CP)-adapted fresh-
water sediment slurries (pH 7.0) in the presence and
absence of added metal solutions (CuCI2, CdCI2,
K2Cr207). A summary of the experimental results is pre-
sented in Table 1. The times required for unadapted
control sediment slurries (no added metais) to dechlori-
nate the selected chlorophenols (10 mg/L) were 16 to
17 days for 2,4-DCP and 2,4,6-TCP; 21 days for 2,3-
DCP; 31 days for 2-CP; 37 days for 2,4,5-TCP; and 30
days for 3-CP. Addition of a 20 mg/L concentration of the
chloride salts of Cu(ll) and Cd(ll) had little or no effect
on the onset rate, or extent of dechlorination for most
chlorophenols tested. Addition of Cr(VI) at 20 mg/L,
however, increased the lag time before initiation of
dechlorination for most chlorophenols tested. Addition of
100 mg/L (or greater) of Cr(VI), as K2Cr207, caused total
inhibition of dechlorination of all CPs tested with the
exception of 2,4-DCP. Addition of 100 mg/L of Cr(VI)
increased the lag time of 2,4-DCP dechlorination from
7.5 days (control) to 45 days. Higher concentrations of
Cd(ll) (100 to 200 mg/L) also caused complete inhibition
of dechlorination for ail chlorophenols tested with the
exception of 2,4-DCP. The rate of dechlorination of 2,4-
DCP was reduced by 70 percent and 100 percent at 100
and 200 mg/L Cd(ll) concentrations, respectively. Addi-
tion of CuCla at 100 mg/L increased the lag time before
the onset of dechlorination for most chlorophenols
tested, but dechlorination of 2,4-DCP was still evident at
200 mg/L of Cu(ll). The distribution of the amended
metals between the aqueous and complexed (sediment)
phases of the experimental samples was determined by
inductively coupled plasma (ICP) spectrometry (Table
2). Representative data for experiments amended with
various metals and either 2,3- or 2,4-DCP are presented
in Table 2. In most experiments, the aqueous phase
concentrations of the added metals were tow, varying
from 0 to 2 mg/L for Cu(ll), 0 to 30 mg/L for Cd(ll), and
pn
-------
Table 1. Summary: Effects of Metal Amendment on the Onset and Rate of Dechlorination of Chlorophenols in Anoxic
Freshwater Sediments
Metal Amendment"
Cu(ll) [mg/L] Cr(VT) [mg/L] CD(II) [mg/L]
Subst. "Live"
Rate
Rate
20
100
200
20
100
200
20
100
200
2-CP
Rate*
0.64
0.65
0.55
N
0.55
N
N
0.56
N
N
Lag
12
10
24
15
10
3-CP
Rate
0.62
0.56
N
N
0.45
N
N
0.46
N
N
Lag (days)
59
58
65
58
2-3-DCP
Rate
0.63
0.74
0.26
N
0.60
N
N
0.60
N
N
Lag
10
7
37
10
7
2,4- DCP
Rata
0.83
0.76
0.63
0.35
0.75
0.18
VS
0.87
0.10
N
Lag
7.5
7
10
56
13
45
7
26
2,4,6-TCP
Rate
0.60
0.89
0.76
N
0.51
ND
VS
0.56
N
N
Lag
7
7
26
10
ND
14
1 Dechlorination rate Is expressed as mg CPio$« L'day"'.
6 Final concentration of metal amended as CuQz, CdClj, or
N, negligible loss; VS, vary slow; ND, not determined.
Table 2. Determination of Total and Aqueous Phase Metal Concentrations In Cherokee Pond Sediment Slurries (10% w/v)
Amended with Various Metal Solutions and Specific Chlorophenols
Metal Amendment
20 mg/L
100 mg/L
200 mg/L
Substrate
Amendment
Cu
Cd
Cr
¦ Cu Cd Cr
Experimental Results
Cu
Cd
Cr
Total" [mg/L]
2,4-DCP
2,3-DCP
25.1
26.9
16.0
20.2
19.7
ND
106.4
113.4
78.4
90.8
115.8
111.4
202.7
192.0
178.2
182.2
222.0
ND
Dissolved0
[mg/L]
2,4-DCP
2,3-CP
<0.03
<0.03
<0.03
<0.03
<0.03
<0.03
0.12
0.12
1.02
1.07
0.12
0.12
0.66
0.6
10.0
9.24
0.34
0.42
¦ Values are the mean of duplicate determinations.
b Total metal concentrations after digestion with concentrated nitric acids.
e Metal concentrations of aqueous supernatant of sediment slurry after centrifugation at 10,000 xg and filtering.
ND, not determined.
0 to 28 mg/L for Cr(VI) over the range of metals added
(20 to 200 mg/L).
The effects of metal salts [20 to 200 mg/L of Cu(ll),
Cd(ll)] on the reductive dechlorination of 3,4-DCP was
compared to two distinct freshwater sediments of similar
pH but of differing total organic carbon content Control
experiments (no added metals) for both sediment types
differed slightly with regard to ttie onset of dechlorina-
tion, but rates of dechlorination were similar. In general,
however, sediment cultures containing the higher or-
ganic carbon content (34 mg C/g) were more resistant
to the inhibitory effects of the added metal salts. The
onset and rate of dechlorination of 3,4-DCP were only
slightly affected in the higher organic carbon sediment
cultures amended with Cu and Cd salts at amended
metal concentrations up to 150 mg/L. Reductive
dechlorination in sediment cultures with a lower organic
carbon content (17 mg C/g) was completely inhibited
(total of 100 days incubation) in cultures amended with
a total metal (Cu or Cd) concentration of 50 mg/L
Further, longer lag times (40 days compared to 14 days)
were observed in those cultures with the lower organic
carbon content at a lower (20 mg/L) amended metal
concentration.
Results of recent experiments using 3,4-DCP-adapted
sediment cultures of low organic carbon content indi-
cated that dechlorination was more resistant to the in-
hibitory effects of the added metal salts than those
reported above for the unadapted sediment cultures.
Further, the effects of amended metals on the reductive
dechlorination of mono- and di-CPs also was examined
using mono- and di-CP-adapted sediment slurries. A
summary of ttie result, plotted as the percentage of
dechlorination activity (DAc) versus the amended metal
Q1
-------
concentration, is presented in Figures 1a-1c. The DAc
is the ratio of the times required for a 50 percent reduc-
tion (Tso) in the initial concentration of the amended CP
in metal-amended slurries and control slurries. Inhibition
of dechlorination was greatest in experiments amended
120
100
2-CP (Cu)
3-CP(Cu)
4-CP (Cu)
3,4-DCP (Cu)
80
40
20
0
0
100
150
200
2S0
300
Metal added [mg Cu/L]
120
100
2-CP (Cd)
3-CP (Cd)
4-CP (Cd)
3,4-DCP (Cd)
£
o
<
~
20
100
0
200
300
Metal added [mg Cd/L]
with Cd and Cu; reductive dechlorination of 3,4-DCP
was more resistant to the inhibitory effects of Cd com-
pared to dechlorination of mono-CPs. Interestingly,
amendment of Cr resulted in the least inhibition of re-
ductive dechlorination for all CPs tested. In fact, reduc-
tive dechlorination of 4-CP was completely resistant to
even the highest Cr concentration (200 mg/L) tested.
These preliminary data suggest that the metal type,
aqueous metal concentration, and organic carbon con-
tent may affect the transformation of chlorophenols in
anoxic sediments. Additional studies are in progress to
determine the inhibitory nature of metals on the reduc-
tive dechlorination of higher chlorinated chloroaromat-
ics, such as pentachlorophenoi.
120
100
2-CP (Cr)
3-CP (Cr)
4-CP (Cr)
3,4-DCP (Cr)
C. 60
40
0
100
200
300
Metal added [mg Cr/L]
Figure 1. Reductive dechlorination activity (DAc) of chlorophe-
nol-adaptad sediments in the presence of
(a) amended Cu, (b) amended Cd, and (c) amended
Cr. The % DAc Is the ratio of the Tso of the metal-
amended slurry and the control slurry. Data are pre-
sented for dechlorination of 2-CP, 3-CP, 4-CP, and
3,4-DCP.
92
-------
Fate of Highly Chlorinated Dibenzo-p-Dioxins and Dibenzofurans in Anaerobic
Soils and Sediments
Peter Adriaens
Department of Civil and Environmental Engineering, University of Michigan, Ann Arbor, Ml
Dunja Grbitf-Gaiid
Department of Civil Engineering, Stanford University, Stanford, CA
Gregory D. Sayles
Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
Chlorinated dibenzo-p
-------
1,2,3, 4, 7, 8 - HexaCDD
1,2, 3,4, 7, 8 - HeptaCDD
112 d
56 d ^ 280 d
¦ ¦ i
Active Au toe laved Chemical
224 d B
Actve
Autoclaved
Chemical
224 d
Active Autoclaved Chemical
Active
Autoclaved
Chemical
u
1,2,4, 6, 8 - PentaCDF
| o d 9 112 d
| 56 d 0 280 d
1, 2, 3, 4, 6, 7, 8 - HeptaCDF
od g 112 d A
56 d Eg 280 d
Active Autoclaved Chemical
Active Autoclaved Chemical
Od SI 56d B 224d B~
0.2 -
Active Autoclaved Chemical
Od «56d M 224d B
Active
Autoclaved Chemical
Treatment
Figure 1. Fate of selected PC DO and PCDF in Hudson River sediments (A) and in Pensacola aquifer material (B).
94
-------
Tabl* 1. Flrat-Ord«r Disappearance Rates of PCDD/PCDF in Hudson River Sediment and Aquifer Material Inoculated Microcosms
PCDD/F Congener
Inoculum
Active
Rate* (x 10-2 wfc-1)
Autoclaved
Nettorate (%)
HpCDD
HR
2.81 ± 0.07
2.3810.14
0.4310.14(15)
PS
3.44 ± 0.10
2.1910.02
1.25 ±0.10 (36)
HxCDD
HR
4.56 ± 0.30
3.1910.09
1.37 1 0.30 (30)
PS
2.4810.13
1.8510.16
0.6310.16 (25)
HxCDOi
HR
4.0010.16
4.1310.14
(-0.13 ±0.16)
PS
ND
ND
ND
HpCDF
HR
4.13 10.10
3.56 10.05
0.57 ±0.10 (14)
PS
4.88 ± 0.06
3.1310.01
1.75 ± 0.06 (36)
PentaCDF
HR"
2.69 ± 0.40
1.06 10.02
1.63 ± 0.40 (60)
PS
1
1.54 ±0.10
1.23 ±0.02
0.31 ± 0.10 (20)
•n-S-12
** Not ffi 95 percent confidence interval
component increased the rate of PCDD/PCDF disap-
pearance by 15 to 35 percent, dependent on the
PCDD/PCDF isomer spiked and the inoculum used.
Previously up to 10 percent of the HeptaCDD congener
was dechlorinated to an unidentified HexaCDD after 2
months (5). This paper presents the analysis of Hep-
taCDD-spiked microcosms inoculated with aquifer ma-
terial. Two hexachlorinated isomers have been
identified: 1,2,3,4,7,8- and 1,2,3,6,7,8-HxCDD, respec-
tively. The 6-chlorine from the lesser chlorinated ring
appeared to remove preferentially to the chlorine in the
4-position, as significantly more 1,2,3,4,7,8-HxCDD ac-
cumulated. The extensive removal of the peri-chlorines
of the HpCDD congener, resulting in an enrichment in
2,3,7,8-chlorinated congeners is similar to what was
found during photolysis (1).
Thus, a dechlorination pattern of 1,2,3,4,6,7,8-HpCDD
is proposed to be consistent with the discrepant degree
in accumulation of the two metabolites recovered (Fig-
a
1, z 3, 4, 7, a - HexaCDD 1, 2, 3, 6, 7, 8 - HexaCDD
Figure 2. Proposed dechlorination of 1,2,3,4,6,7,8-HeptaCDD In
methanogenic aquifer microcosms.
ure 2). A pentaCDF isomer formed from dechlorination
of 1,2,3,4,6,7,8-HpCDF has not been identified at this
time.
References
1. Tysklind, M., A,E. Carey, C. Rappe, and G.C. Miller.
1992. Photolysis of PCDF and PCDD on soil. Pro-
ceed. Dioxin 1992 Conference, Vol. 8, pp. 293-296,
Tampere, Finland.
Z Brown, J.F., Jr., D.L Bedard, M.J. Brennan, J.C.
Camahan, H. Feng, and R.E. Wagner. 1987. Poly-
chlorinated biphenyl dechlorination in aquatic sedi-
ments. Science 236:709-712.
3. Quensen, J.F. Ill, S.A. Boyd, and J.M. Tiedje. 1990.
Dechlorination of four commercial polychlorinated
biphenyl mixtures (Arodors) by anaerobic microor-
ganisms from sediments. Appi. Environ. Microbiol.
56:2360-2369.
4. Nies, L, and T.M. Vogel. 1990. Effect of organic
substrates on dechlorination of Aroclor 1242 in an-
aerobic sediments. Appl. Environ. Microbiol.
56:2612-2617.
5. Adriaens, P., and D. Grbic-Galic. 1991. Evidence for
reductive dehalogenation of highly chlorinated diox-
ins and dibenzofurans. Abstr. Dioxin 1991 Confer-
ence, Chapel Hill, NC.
95
-------
Bioreactor Treatment of Nitrate Contamination in Ground Water: Studies on the
Sulfur-Mediated Biological Denitrification Process
Michael S. Davidson and Thomas Cormack
Biotechnology Research Department, Orange County Water District, Fountain Valley, CA
Nitrate contamination of ground water resources has
become an increasingly serious problem in the United
States and Europe (1,2). Water containing more than 45
ppm nitrate (10 ppm nitrate-nitrogen) is considered unfit
for potable use. Consumption of water containing nitrate
in excess of this level appears to be associated with
adverse health effects, namely increased incidence of
methemoglobulinemia in infants and gastrointestinal
malignancies in adults (3).
The ground water basin underlying Orange County, Cali-
fornia, serves a population of some 2.3 million people.
Currently, more than 250,000 acre feet of the basin's
capacity is rendered unusable because of nitrate levels
exceeding the federally mandated limit of 45 ppm.
At least six technological approaches to mitigation of the
nitrate problem exist, including 1) recasing and/or redrill-
ing wells to access uncontaminated zones of the aquifer,
2) blending contaminated water with low-nitrate water to
achieve mandated standards, 3) demineralizing water
with reverse osmosis, 4) demineralizing water with elec-
trodialysis, 5) subjecting water to an ion-exchange proc-
ess, and 6) subjecting water to biological denitrification
(4). Orange County Water District has been exploring
the use of biological denitrification as a means of pro-
ducing potable-quality water from nitrate-contaminated
ground water reserves.
Background
The balance of existing research on biological denitrifi-
cation relates to heterotrophic processes in which an
organic compound, such as methanol, glucose, or glyc-
erol, is added to the high-nitrate water (5,6). Such or-
ganics serve as a carbon and energy source for
denitrification. While effective, the addition of organics
to the water is undesirable since residuals might render
the denitrified water toxic (i.e., if methanol is used). In
addition, organics can give rise to trihalomethanes
(THMs), if chlorine disinfection is used.
Autotrophic denitrification has been studied as a means
to avoid these shortcomings. This paper deals with the
use of elemental sulfur as an energy source for biologi-
cal .denitrification catalyzed by the autotrophic bacte-
rium, Thiobacillus denitrificans, which in the absence of
free oxygen reduces nitrate to dinitrogen gas by oxidiz-
ing sulfur to sulfate. All carbon required for biosynthesis
is derived from carbon dioxide/bicarbonates present in
the ground water. The denitrification process can be
conducted in fluidized bed bioreactors, fixed (static) bed
reactors, or in agitated (stirred) tank reactors.
Experimental Reactors/Methods
The denitrifying bacteria used in this study were ob-
tained from an anaerobic enrichment culture consisting
of basal mineral saits and sodium thiosulfate. Upon
development, the culture was transferred to flasks con-
taining basal salts and elemental sulfur. The culture
consists of a consortium of bacteria: Thiobacillus deni-
trificans plus a variety of heterotrophic denitrifying
bacteria.
The fluidized bed reactor was constructed from a 3.65 m
section of 5.08 cm (internal) diameter transparent poly-
vinyl chloride (PVC) pipe closed at the ends with PVC
unions. The reactor was mounted outdoors and fed witfi
nitrate-amended well water, which was conveyed to the
reactor using an adjustable peristaltic pump (Masterflex)
capable of flows of up to 13.2 LVmin. The reactor con-
tained sulfur particles with an initial size range of
-16/+30 mesh (U.S. Standard Screen Series). Depth of
the sulfur bed under static (nonfluidized) conditions was
1.83 m. The reactor was inoculated with the denitrifying
consortium.
The agitated (sulfur slurry) reactor system was con-
structed using a glass and stainless steel research fer-
menter (New Brunswick Scientific). The reactor (1.3-L
liquid capacity) system provides the capacity for tem-
perature control, variation of agitation rate, automatic pH
96
-------
control, and determination of dissolved oxygen tension.
The reactor was charged with precipitated sulfur (100
percent - 80 mesh). Initial reactor runs employed a sulfur
solids content of 1.5 percent (weight/volume). The reac-
tor was fed (via a peristaltic pump) a synthetic salts
solution formulated to simulate locally occurring high-ni-
trate ground water. The reactor system included a grav-
ity separator unit to recover unused sulfur from the
denitrified effluent. The separated sulfur was continu-
ously returned to the main reactor vessel by means of a
second peristaltic pump. The agitated slurry reactor was
inoculated with the denitrifying consortium used in the
fluidized bed reactor (see above).
During operation of both reactors, effluent samples were
obtained on an hourly basis by means of an automatic
sampling device (Isco). Samples were analyzed for ni-
trate, nitrite, sulfate, and pH. Studies also were con-
ducted on the effluent to determine total autotrophic and
heterotrophic bacterial counts.
Results/Discussion
Both reactor types were capable of maintaining pro
longed periods of stable operation (effluent containing
less than 0.3 ppm of either nifrate or nitrite from influent
streams containing from 45 to 100 ppm nitrate). The
autotrophic denitrification process resulted in a pH drop
from approximately 8.2 (influent) to 7.0 (effluent). It was
found that 1.64 mg of sulfate was added to the treated
water for every 1.0 mg of nitrate removed. Typical nitrate
removals for the fluidized bed reactor and the sulfur
slurry reactor are presented in Tables 1 and 2, respec-
tively. Row rate for the fluidized bed was 250 mL/min
(reactor temperature at 18"C) while the flow to the sulfur
slurry reactor was 5.0 mLVmin (reactor temperature at
30°C).
Table 1. Denltrlflcatfon by Fluid Bed Reactor
Influent Nitrate Effluent
Run Hour (ppm) Nitrate (ppm)
€600
46.7
0.7
6606
46.7
0.0
6612
46.7
0.0
6618
46.7
0.0
6624
47.0
0.0
6630
47.0
0.0
6636
47.0
0.0
6642
47.0
0.0
6646
42.5
0.0
6654
42.5
0.0
6660
42.5
0.0
6666
42.5
0.0
Table 2. Denitrification by Fluid Bed Reactor
Run Hour
Influent Nitrate
(ppm)
Effluent
Nitrate (ppm)
2303
572
0.0
2309
57.2
0.0
2315
57 i!
0.1
2321
57.2
0.1
2327
59.8
0.0
2333
59.8
0.0
2339
59.8
0.0
2345
59.8
0.0
2351
55.7
0.0
2357
55.7
0.0
2363
55.7
0.0
2369
55.7
0.0
Summary/Conclusions
The operational characteristics of two sulfur-based bio-
logical reactor systems were investigated. Maximum
hydraulic loading rates consistent with stable denitrifica-
tion were determined. Efficient operation of the process
in a field application might require supplementation of
the high-nitrate ground water reactor feed with trace (1
to 2 ppm) amounts of ammonium and/or phosphate to
supply bacterial biosynthetic requirements.
Both reactor types were successfully employed for the
sulfur-mediated biological denitrification process. The
fluidized bed reactor appears to offer several distinct
advantages including: 1) no complex solid-liquid sepa-
rating unit (clarifier).is needed; 2) no multiple staging is
required to prevent short- circuiting of water undergoing
denitrification; 3) no costly mechanical agitator drives
and reactor seals are required; 4) reactors can be con-
structed with large height-to-diameter ratios, thus mini-
mizing space requirements; and 5) no recycle pumping
of the sulfur substrate slurry is required.
References
1. Power, J.F., and J.S. Schepers. 1989. Nitrate con-
tamination of groundwater in North America. Agri.
Ecosys. Environ. 26:165-187.
2. Strebel, O., W.H.M. Duynisveld, and J. Bottcher.
1989. Nitrate pollution of groundwater in Western
Europe. Agri. Ecosys. Environ. 26:189-214.
3. National Academy of Sciences. 1981. The Health
Effects of Nitrate, Nitrite, and N-Nitroso Compounds.
National Academy Press, Washington, DC.
4. Aieta, M. 1988. Review of nitrate removal processes
for groundwater treatment. Table 3.4 Memorandum.
In: Preliminary facilities investigation for Chino Basin
.Q7
-------
Storage Program. Report prepared by James M.
Montgomery Consulting Engineers, Inc., Pasadena,
CA, for the Metropolitan Water District of Southern
California.
5. Gayle, B.P., G.D. Boardman, J.H. Sherrard, and R.E.
Benoit. 1909. Biological denitrification of water. J.
Environ. Engin. 115:930-943.
6. Rogalla, F., G. DeLarminat, J. Coutelle, and H. Go-
dart Experience with nitrate removal methods from
drinking water. Proc. Conference on Nitrate Contami-
nation: Exposure, Consequences, and Control. Nato
Advanced Department of Civil Engineering, Lincoln,
NE.
98
-------
Chemical Interactions and pH Profiles In Microbial Blofllms
Joseph R.V. Flora, Makram T. Suidan, and Pratim Biswas
University of Cincinnati, Cincinnati, OH
Gregory D. Sayles
U.S. Environmental Protection Agency, Cincinnati, OH
Substrate utilization kinetics within biofilms has been
modeled by coupling Fickian diffusive transport witfi
Monod-reaction kinetics (1,2). Substrate and product
concentration profiles are effected in biological systems
because of simultaneous diffusion and reaction. Com-
pounds used or produced during biological transforma-
tions can have an impact on the rate of substrate
utilization by affecting the local pH. For example, nitrifi-
cation and the utilization of halogenated organic com-
pounds produce acid equivalents and cause a decrease
in pH, while denitrification consumes acid equivalents
and causes an increase in pH. Production and utilization
of total carbon dioxide also alters the chemical equilib-
rium. The rates of substrate utilization and growth of
microorganisms are pH dependent and can vary signifi-
cantly from the bulk solution to the attachment wall of
biofilms because of the pH gradients influenced by dif-
fusional resistance to mass transport. An approach was
developed to analyze the effects of chemical interac-
tions and pH changes within the biofilm on the overall
rates of substrate removal in biofilms. This approach
incorporates ionic mass transport effects accurately and
accounts for the presence of background ions; charges
on the surface or microorganisms; and corrections for
the activity, electrophoresis, relaxation, and hydration of
ions. Models have been developed for autotrophic and
anaerobic biofilms. Results for an acetate-utilizing
methanogenic biofilm are described.
The key features in this work are the incorporation of
pH-dependent Monod kinetics and an attempt to ac-
count for the production and- movement of gaseous
products in methanogenic biofilms. Gaseous products
accumulate within the biofilm as bubbles that peri-
odically erupt to the surface, especially in systems under
high organic loadings (3). Since a portion of the carbon
dioxide that is produced does not remain in solution, the
extent of pH change within the biofilm will be affected.
In this study, the process of gas bubble formation and
eruption is modeled by assuming that the advection of
the gas generated from within the film to the bulk is
described by a one-dimensional gas flow rate per unit
area of the biofilm. Overall, the mass transfer of com-
pounds in the biofilm consists of the diffusion of dis-
solved compounds in the liquid phase (including the
biofilm) and the advection of gaseous products in the
gas phase.
Assuming that the biofilm is at steady-state with respect
to the consumption of substrate, a one-dimensional
mass balance on total acetate (CTAe), total inorganic
carbon (CT|C), and methane in the biofilm yields,
bXf QrtBxXf [HAc\
. Y ~
d^HAc d^Ac
dx
dx
K, + [HAd[.
'Ph
(1)
dJcct, cUhcq- dJco]'
dx
¦ +
dx
dx
d_
dx
(qgKnco£COA)
RT
QmaxXf [HAC]
KS + [HA$
pH
~dT +
d_
dx
f <7^ch4[CH4]
RT
qmMHAQ
K,+ [HAc)
pH
(2)
(3)
where:
J = the molar flux of the species,
x s the spatial coordinate perpendicular to the
biofiim surface.
qg = the gas volumetric flow rate per unit surface
area of biofilm.
Kf, = Henry's constant
R = the universal gas constant.
T = the absolute temperature,
b = the respiration coefficient.
Y = the cell yield per mass of acetate.
X( = the microorganism density in the biofilm.
qmax = the maximum specific growth rate at the
optimum pH.
99
-------
Kg = the Monod half-saturation constant.
fpH = the fractional decrease in the growth rate due
to deviations from the optimum pH.
The brackets [ ] refer to the molar concentration of the
species.
The flux is given by Fick's law for neutral species and
by the Nernst-Planck equation for ionic solutes (4). The
fractional decrease in growth rate because of deviations
from the optimum pH is represented by a Gaussian
curve (5),
(ph'i.caf
¦ tpH=e\ 0.5 J (4)
Equations 1 to 4 were solved in conjunction with ex-
pressions for acid-base and gas-liquid equilibria and
etectroneutraJity.
The profiles of (Ctac) and PH f°r a biofilm depth of
1,000 (im, bulk pH=7.0, bulk Ctac=10"2 M, and various
bulk partial pressures of carbon dioxide are shown in
Figure 1. The profile of Ctac predicted by using tradi-
tional biofilm models also is shown. Traditional biofilm
models typically assume a constant pH in the biofilm.
These models are expressed by a mass balance on total
acetate similar to equation 1 but use nek's law to de-
scribe the cfiffusion of both acetic acid and. the acetate
ion. The current model predicts that the pH increases in
the biofilm. This increase is due to the utilization of a
stronger acid (acetic acid) and the production of a
weaker acid (carbon dioxide). Depending on the partial
pressure of carbon dioxide (Pco2)> the pH at the wall of
the biofilm can be as high as 7.49. This deviation from
the optimum pH causes a drop in the rate of acetate
utilization and increases the level of Ctac within the
biofilm. The lower the Poo2. the lower the buffer intensity,
and the larger the increase in pH and Ctac in the fi'm-
Since fraditional models assume a constant pH, the
concentration of Ctac reaches very low levels in the biofilm.
— ¦
0 200 400 600 800 1000 0 200 400 600 800 1000
Distance into the Biofilm, jim
Figure 1. Profiles of total acetate (Ctac) and pH for various bulk
partial pressures of carbon dioxide, a biofilm depth
of 1,000 fim, bulk pH=7.0, bulk Ctac«10'2 M, and
Lcbl=0.
The impact of the bulk pH on the flux of substrate into
the biofilm is shown in Figure 2 for a bulk CTAc=10"2 M
for various bulk CT|C and a biofilm depth of 1,000 pin.
The flux predicted by using the traditional model also is
shown in tiie figure. Figure 2 shows that the traditional
model predicts a maximum flux at a bulk pH of 7.0. This
pH corresponds to the optimum pH of the microorgan-
isms specified using equation 4. In contrast, the current
model predicts an optimum pH less them 7.0 for acetate-
utilizing biofilms. Since the pH increases in acetate-util-
izing biofilms, operating at a lower pH causes a larger
fraction of the biofilm to exist closer to the optimum pH.
Decreasing the bulk CnC (or buffer intensity) causes a
greater increase in the pH within the biofilm and shifts
the .optimum bulk pH to lower values. Overall, the ap-
proach developed can be coupled to a reactor model,
and the operating conditions for a reactor (such as the
buffering capacity of the influent feed, tfie influent pH,
and the surface loading rates) can be specified to opti-
mize reactor performance.
0.2
—. Trttfliona Mow!
-- Thli 9lidy, Cnc ¦ NT'*-1
— TW« Sfctfy. Crc ¦ lO*431
Thto atitfy, Cnc -
0.1
UL
0.0
6
7
8
Bulk pH
Figure 2. The flux of substrate into the biofilm as a function of
bulk pH for a bulk Ctac-1 0'2 M for various bulk Ctic
and a biofilm depth of 1,000 pm.
References
1. Williamson, K„ and P.L. McCarty. 1976. A model of
substrate utilization by bacterial films. J. Water Pol-
lut. Control Fed. 48:9-24.
2. Suidan, M.T., and Y.-T. Wang. 1985. Unified analysis of
biofilm kinetics. J. Environ. Eng. (ASCE) 111:634-646.
3. McCarty, P.L., and F.E. Mosey. 1991. Modeling of
anaerobic digestion concepts (a discussion of con-
cepts). Water Sci. Technol. 24:17-33.
4. Cussler, E.L. 1984. Diffusion: mass transfer in fluid
systems. Cambridge University Press, Cambridge.
5. de Beer, D., J.W. Huisman, J.C. van den Heuvel, and
S.P.P. Ottengraf. 1992. The effect of pH profiles in
methanogenic aggregates on the kinetics of acetate
conversion. Water Res. 26:1329-1336.
100
-------
Characterization of Biofilter Microbial Populations
Alec Breen, Alan Rope, and John C. Loper
Department of Molecular Genetics, University of Cincinnati, Cincinnati, OH
P.R. Sferra
Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
Biofiltration, a promising, economical form of biological
treatment, is becoming a more common means of re-
moving environmental contaminants from waste gas (1).
The conversion of target compounds to innocuous end
products in the biofilter treatment process is mediated
by a heterogeneous bacterial population that develops
on the biofilter support material. Biofilter technology has
been advancing rapidly in Europe since the late 1970s,
but its application in the United States has been limited.
Originally used at facilities such as animal rendering
plants for odor abatement, biofilters are now being em-
ployed in the treatment of volatile organic compounds
(VOCs) such as toluene, benzene, and ethylbenzene as
well as more recalcfaant compounds such as methylene
chloride and trichloroethylene (TCE) (2).
In addition to sound engineering design, improved biofil-
ter efficiency ?s well as broadened VOC removal capa-
bilities are ultimately dependent on the catabolic
potential and activity of the biofilter microorganisms.
Biofilter microbial communities are largely uncharacter-
ized (3). Analysis of the structure and function of biofilter
communities should provide useful information for biofil-
ter performance optimization, detection and assessment
of biofilter perturbation, and comparison of biofilters
(e.g., between lab- and pilot-scale biofilters).
In this study, characterization of biofilter communities is
being conducted by using both molecular biological
techniques and standard microbiological protocols.
Gene probes specific for aromatic hydrocarbon oxida-
tion are being employed to determine which biochemical
pathways are present in the biofilter, and phylogenetic
probes are being used to determine what types of bac-
teria are predominant. In addition to these methods,
DNA amplification fingerprinting (DAF) is being evalu-
ated as a means of monitoring the microbial community
structure of the biofilter.
Background
Incorporation of some of the new DNA-based tech-
niques in the study of biofilter microbial ecology has the
potential to result in more precise and less time-consum-
ing analyses that will directly contribute to improved
biofilter performance. DAF technology has been suc-
cessfully employed in a number of studies and has
emerged as a powerful tool for rapid microbial identifi-
cation (4,5,6). DAF also has been useful in the mapping
of complex genomes (7). The utility of DAF in the analy-
sis of mixed cultures is being evaluated as part of this
study. Nucleic acids can be extracted directly from biofil-
ter material and subjected to polymerase chain reaction
(PCR) amplification in a relatively rapid timeframe. The
ultimate goal is to determine whether the fingerprints
generated can provide a means of comparing commu-
nity structure over time or among different biofilters.
Parallel studies using gene probes and standard cultur-
ing techniques are being conducted. The hope is ttiat
structural and functional assessments of biofilter com-
munities can be made by combining these methods.
Results
Initial experiments involved the generation of finger-
prints for previously characterized VOC catabolic strains
of Pseudomonas cepacia G4, P. putida F1, and Pseudc-
monas sp. JS150. These experiments served to evalu-
ate the efficacy of various primers in the PCR reaction
and provided fingerprints to compare with uncharacter-
ized biofilter isolates.
Preliminary mixed-culture experiments were conducted
using bench-scale culture microcosms inoculated with
biofilter material and maintained on various VOCs.
These cultures served as sources of biomass for the
evaluation of lysis and DNA extraction methods. All PCR
reactions were conducted using a single 8-base oligonu-
cleotide as a primer. Microcosms, one with toluene as a
101
-------
sole carbon source and the second with toluene, ethyl-
benzene, and p-xylene, were maintained for a period of
5 months. The resultant fingerprints showed consistent
profiles over a period of 3 consecutive days as well as
some common bands between the microcosms (Figure
1). The patterns generated show distinct bands, not an
indiscriminant smear as might have teen predicted
given the complexity of the sample.
Three operational VOC catabolizing biofliters, packed
with peat and maintained in different configurations,
were analyzed using DAF. These biofilters were main-
tained under different regimes: one with a pelletized
medium, one maintained on channelized medium in a
concurrent mode (gas stream and nutrient flow in the
same direction), and one maintained on channelized
medium in a countercurrent mode (gas stream and nu-
trient flow in opposing directions). The biofilters received
toluene as a sole carbon source for 6 months prior to
the initiation of sampling. Similar DAF profiles were
obtained for the biofilters maintained on channelized
t
Figure 1. DAF analysis of mixed microbial populations. Pat-
terns were generated from nudaic acid extracts from
shake flasks inoculated with biofilter material and
maintained on VOCs for 5 months. Lanes 1 and 7 are
123-base molecular weight standards. Lanes 2,3, and
4 are patterns from the toluene-maintained culture
over 3 consecutive days. Lanes 5 and 6 are patterns
generated at 48-hour intervals from the flask main-
tained on toluene, ethylbenzene, and p-xylene.
support media, while the pelletized biofilter yielded a
unique pattern. The individual organisms from these
biofilters appear to constitute a relatively diverse popu-
lation capable of growth on a number of VOCs including
benzene and ethylbenzene. Fungi appear to be only a
minor constituent of the population.
Summary and Conclusions
OAF has utility for generating rapid genetic profiles of
biofilter microbial communities. Multiple samples can be
easily processed, and culturing is not necessary. The
significance of these profiles as means of monitoring
biofilter performance is being evaluated by intensive
sampling over time. DNA fingerprinting technology has
potential as a rapid and economic tool for biofilter moni-
toring and for diagnostics. Ideally, a given DAF profile
for a particular biofilter will prove to be diagnostic for
particular modes of performance. Studies currently un-
der way will involve gene probing to provide a functional
component of the analysis.
References
1. Leson, G., and A.M. Winer. 1991. Biofiltration: An
innovative air pollution control technology for VOC
emissions. J. Air Waste Manag. Assoc. 41:1045-
1054.
2. Tong, G.E. 1991. Integration of biotechnology to
waste minimization programs. In: G.S. Sayler, R.
Fox, and J.W. Blackburn, eds., Environmental
. Biotechnology for Waste Treatment, pp. 127-136.
3. Lipski, A., S. Klatte, B. Bedinger, and K. Altendorf.
1992. Differentiation of gram-negative, nonfermenta-
tive bacteria isolated from biofilters on the basis of
fatty acid composition, quinone system and physi-
ological reaction profiles. Appl. Environ. Microbiol.
58:2053-2065.
4. Williams, J.G.K., A.R. Kubelik, K.J. Livak, J.A. Rafal-
ski, and S.V. Tingey. 1990. DNA polymorphisms am-
plified by arbitrary primers are useful as genetic
markers. Nud. Acids Res. 18:6531-6535.
5. Lehman, P.F., D. Lin, and B.A. Lasker. 1992. Geno-
typic identification and characterization of species
and strains within the genus Candida by using ran-
domly amplified polymorphic DNA. J. Clin. Microbiol.
30:3249-3254.
6. Harrison, S.P., L.R. Mytton, L Skot, M. Dye, and A.
Cresswell. 1992. Characterization of Rhizobium iso-
lates by amplification of DNA polymorphisms using
random primers. Can. J. Microbiol. 38:1009-1015.
7. Caetano-Anolles, G., B.J. Basam, and P.M.
Gresshoff. 1991. DNA amplification finger printing
using very short arbitrary oligonucleotide primers.
Bio/Tech. 9:553-557.
102
-------
Fundamental Studies In the Development of the Gas Phase Biofilter
Rakesh Govind, Vivek Utgikar, Yonggui Shan, Wang Zhao, Madan Parvatiyar, and Stephan Junginzer
Department of Chemical Engineering, University of Cincinnati, Cincinnati, OH
Dolloff F. Bishop
Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
The Superfund Amendments and Reauthorization Act
(SARA) emission summary for petroleum and chemical
manufacturing companies shows that the largest re-
leases in air are volatile organic compounds (VOCs).
Conventional technologies that have been used to con-
trol VOCs include: 1) adsorption onto porous materials,
such as activated carbon; 2) absorption into a liquid
stream followed by stripping of the liquid; 3) combustion
or incineration; and 4) pervaporation using membranes
to selectively remove the harmful volatiles. These tech-
nologies present several problems, involving disposal of
generated wastes, cost of equipment, and materials and
energy required for operation.
A biofilter bed consists of a bed packed with a solid
support with microorganisms in the form of a wet, bio-
logically active layer, referred to as biofilm, immobilized
on the surface of the support material. The solid support
material can be either fine particulate material, such as
soil, peat, compost, pellets of activated carbon, or ce-
ramic media, or structured monoliths with passages of
defined geometry. The microbial degradation of sub-
strate is assumed to take place in the biofilm.
A wide range of VOCs can be removed by microbial
degradation in biological filters. Air biofiltration is a well-
established technology in Europe (1).
As compared to other options, biofiltration is cheap,
reliable, and represents a more natural approach for
control of VOCs. Moreover, it can handle chemically
different compounds, and the compounds are com-
pletely degraded, unlike in some other technologies
where they are simply transferred from one phase to
otfier (2).
Background
Biological filtration of VOCs implies the removal of com-
pounds by contacting the contaminated air with micro-
organisms in a column or similar configuration. Three
basic types of biological filters are available: bioscrub-
bers, bio-trickling filters, and biofilters (3). A biofilter has
high specific surface area packing and water present as
liquid holdup in the column. Either gas entering the
biofilter has to be humidified or water has to be sprayed
intermittently into the biofilter to prevent the drying of the
column. Further, biofilters have to be washed with water
during operation to prevent the accumulation Of degra-
dation products.
Biological methods have been employed for purification
of waste gases since tfie early sixties. The concept of
using biodegradation to control hydrogen sulfide from
sewage works was first considered as early as 1923 (1).
Soil beds have been used for controlling odors from
sewage and wastewater plants. Bohn and Bohn (4)
have discussed the theory and potential applications of
soil beds for odor control. Such systems have recently
been used for control of volatile aliphatic compound
emissions (5).
Biofilters have been used extensively in Europe, espe-
cially in Germany and the Netherlands. Eitner (6) has
presented data indicating that a significant reduction of
hydrocarbon concentration is possible in approximately
1 week. Maximum removal rates were attained within 1
month of operation. He also has discussed the distribu-.
tion of microorganisms in the biofilter. Ottengraf and van
den Oever (7) have discussed the purification of waste
gas containing toluene, butanol, ethyl acetate, and butyl
acetate. The results show a zero-order dependence of
degradation rate on concentration.
Don and Feenstra (8) have presented data comparing
several alternative technologies for treatment of waste
hydrocarbon gaseous streams and showed that biofil-
tration is the most cost-effective treatment method.
103
-------
the soil bed biofiiter systems mentioned above are
inherently inefficient with respect to space utilization.
One biofiiter system, termed the BIKOVENT system,
consists of prefabricated concrete parts that form an
aeration plate to give uniform air distribution and drain-
age ducts. Developed by Drs. Hans Gethke and Detlef
Eitner of Aachen, Germany, the BIKOVENT has been
extensively used in Germany and Austria for odor con-
trol and for controlling VOCs in waste air streams. The
technology was recently introduced to the United States
by Biofiltration Inc., Gainesville, Florida.
A number of researchers have attempted to model
biofilm processes since the 1970s. Few of the available
models describe the situation of biofilm in three phase
gas-liquid-solid systems where the compound undergo-
ing degradation is present in the continuous gas phase.
Further, none of the models are useful when the kinetic
constants used in the rate expression are unknown.
This paper describes three experimental and theoretical
studies conducted on the biodegradation of toluene,
methylene chloride (MeCI2), trichloroethylene (TCE),
ethylbenzene, and chlorobenzene in air in aerobic
biofi Iters.
Measurement of Biofilm Kinetics
A packed bed microbiofilter was set up to conduct kinetic
experiments with immobilized biomass. The schematic
of the experimental system is shown in Figure 1. A
contaminant mixture in air was circulated through the
packed bed microbiofilter (which was operated as a
differential reactor) connected to the gas reservoir. Deg-
radation of the substrate was monitored based on the
substrate concentration in the gas reservoir. Mass bal-
ance equations were written for the system and the data
obtained on the microbiofilter were analyzed. Figure 2
shows a typical experimental run for the system.
Nutrient
Reservoir
Jacketed
Biofiiter
Gas
Reservoir
j Septum
l-OH
Water Bath
Teflon Bag
Biomass
on Pellets
500
400
300
5 200 -
100 -
Contaminant: Toluene
Time (hrs)
Figure 2.
Experimental results on biofilm kinetics for toluene
using the experimental apparatus shown in Figure 1.
Experimental Biofiiter Studies
The first study described the performance of a packed
bed activated carbon biofiiter (9). Complete degradation
of toluene, methylene chloride, and TCE was demon-
strated at the bench-scale at 2 minutes of gas retention
time. After about 180 days of operation, the biofiiter
flooded because of excessive biomass accumulation,
and the carbon pellets had to be cleaned periodically.
Figure 3 shows the removal efficiency of the biofiiter
as a function of time for the aerobic activated carbon
biofiiter.
The second study described the performance of a
bench-scale packed bed biofiiter with ceramic pellets
(Celite, Manville Corporation, CA). Complete degrada-
tion of all compounds except TCE was achieved in our
experiments. Figure 4 shows the removal efficiency of
the compounds in the packed bed Celite biofiiter.
The third study described the performance of a straight
passages biofiiter constructed from corrugated plates of
_ 100
a?
Totuona TCE M9CI2
Figure 1. Schematic of the experimental apparatus for deter-
mining biofilm kinetics.
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure 3. Experimental results on removal efficiency obtained
for the activated carbon packed bed aerobic biofiiter.
104
-------
100
90
# 80
•-UvClz —t—TCf:
kChlorolMfimn ¦
60
Tune (days)
120
Figure 4. Experimental results on removal efficiency obtained
for the Celite pellets packed bed aerobic biofilter.
Celite. The straight passages offer a means for exces-
sive biomass to leave the biofilter ted, which cioes not
occur in a packed bed system. Figure 5 shows the
removal efficiency of the Celite straight passage biofilter
for the compounds. The biomass loss from the straight
passage biofilter is related to the flow rate of the nutri-
ents, as shown in Figure 6.
60
Time (days)
Figure 5. Experimental results on removal efficiency obtained
for the Celite plates straight passages aerobic biofllter.
100
10
0.1
Oaa Flown a
~ 1600 mL/frtn
a 600 miymin
10
Two models for a biofilter system have been developed
that quantitate biofilm and plug flow regimes of biofilter
operation. These models have been applied to the ex-
perimental data from the above three studies. Insights
into biofilter design and operation gained from this
analysis will be presented subsequently. That presenta-
tion will include a preliminary design procedure for biofil-
ters and a comparison of costs with other technologies.
References
1. Leson, G., and A.M. Winer. 1991. Biofiltration: An
innovative air pollution control technology for VOCs
emissions. J. Air Waste Manage. Assoc. 41:1045
2. WPCF. 1990. Draft of report on VOC vapor phase
control technology assessment. Water Pollution
Control Federation.
3. Ottengraf, S.P.P., and R. Diks. 1990. Biological puri-
fication of waste gases. Chim. Oggi, 5:41-45.
4. Bohn, H.L., and R.K. Bohn. 1986. Soil bed scrubbing
of fugitive gas releases. J. Environ. Sci. Health.
A21:1236.
5. Kampbell, D.H. et al. 1987. Ftemoval of volatile ali-
phatic hydrocarbons in a soii bioreactor. J. Air. Potlut.
Contr. Assoc. 37:1236.
6. Eitner, D. 1984. Investigations of the use and ability
of compost filters for the biological waste gas purifi-
cation with special emphasis on the operation time
aspects (Ger.). GWA, Band 71, TWTH Aachen.
7. Ottengraf, S.P.P., and H.C. van den Oever. 1983.
Kinetics of organic compound removal from waste
gases with a biological filter. Biotech. Bioeng.
25:3089.
8. Don, J.A., and L Feenstra. 1984. Odor abatement
through biofiltration. Paper presented at symposium
Louvain-La-Neuve.
9. Govind, R., V. Utgikar, Y. Shan, and W. Zhao. 1992.
Studies on aerobic degradation of volatile organic
compounds (VOCs) in an activated carbon packed
bed biofilter. Paper submitted to ES&T for publication.
100
Nutrient Flowrate (L/m^hr))
Figure 6. Experimental data on biomass slough off from the
straining passages biofilter.
105
-------
Sequential Anaerobic/Aerobic Treatment of Contaminated Soils and Sediments
Grace M. Ldpez, Gregory D. Sayies, Karen Buhler, and Dolloff F. Bishop
Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
In S. Kim, Guanrong You, Petra Klostermann, and Margaret J. Kupferle
Department of Civil and Environmental Engineering, University of Cincinnati, Cincinnati, OH
Douglas S. Upton
Levine-Fricke Consulting Engineers, Emeryville, CA
Every year many new Superfund sites are added to the
National Priority List for extensive cleanup procedures.
Feasible and low-cost technologies must be developed
to expedite these procedures. One technology that
meets these requirements is bioremediation (1), which
uses naturally occurring microorganisms to convert haz-
ardous substances into less-toxic or nontoxic substances.
The biodegradation-rates of many highly chlorinated
compounds can.be accelerated by applying sequential
anaerobic/aerobic treatment (2). In general, tine bio-
chemical pathway providing the highest rate for the
initial steps of microbial destruction of highly chlorinated
organics is anaerobic reductive dechlorination. Once
partially dechlorinated, the resulting compounds typi-
cally degrade faster under aerobic, oxidizing conditions.
This scenario suggests that total degradation of the
contaminants might occur when both anaerobic and
aerobic treatments are sequentially applied.
The objective of this project is to conduct fundamental
and applied research to aid in the development of se-
quential anaerobic-aerobic landfarming and composting
technologies. These technologies wiil be used to biologi-
cally treat soils or sediments contaminated with highly
chlorinated aromatic compounds and other low-solubility
compounds that are susceptible to sequential treatment
Experimental Design
Methanogenic master culture reactors (MCRs) have
been established to be used as a source of acclimated
biomass for bench-scale biodegradation studies. The
seed for the MCRs was anaerobically digested sewage
sludge from a municipal wastewater treatment plant in
the Cincinnati area. Ethanol is used as a primary sub-
strate to enrich methanogenic microorganisms. These
reactors are kept under mesophilic conditions (35#C) in
a controlled-temperature chamber. The reactors are fed
semicontinuously with a mineral-salts medium and the
contaminants of interest (Table 1). The ethanol loading
rate is 1 g chemical oxygen demand/L/day. A computer-
controlled respirometer (CES AER-200) is used to re-
cord daily gas production and track the performance of
the MCRs.
Aerobic MCRs have been established by aerating a
portion of the anaerobic MCRs at 20#C. These reactors
also are fed semicontinuously using ethanol as the pri-
mary substrate. Anaerobic biodegradation products of
ttie compounds listed in Table 1 will be used as co-sub-
strates.
Table 1. Compounds of Interest
DDT
Hexach la ro benzene
Pentachloroptenol
Aroclor 1260 (PCBs)
Naphthalene (PAHs)
The rates and products of biodegradation will be meas-
ured in batch reactors under anaerobic and aerobic
conditions using aqueous slurries of soils and sedi-
ments. The soil or sediment will be spiked with contami-
nant and transferred to serum bottles. Inoculum,
co-substrate, and minimal salts media will be added,
and substrate and product concentrations will be meas-
ured over time.
Results
Anaerobic MCRs that are acclimated to each of the
compounds of interest are operating at steady-state.
The accumulated gas production data for the DDT and
106
-------
hexachlorobenzene MCRs for a 6-month period are pre-
sented in Figures 1 and 2, respectively. The concentra-
tions of the contaminants were increased until inhibition
was experienced in the reactors "as defined by loss of
gas production capability. From these data, it was con-
cluded that the optimal operating conditions for DDT and
hexachlorobenzene were at a concentration of 2 mg/L.
Conclusion
The MCRs are operating at steady-state, and the biode-
gradation studies are in progress. These experiments
will provide us with fundamental information on reaction
kinetics that can be used to evaluate the feasibility of
sequential anaerobic-aerobic treatment and to optimize
its performance under field conditions. Further studies
that will simulate field conditions more closely are
planned.
References
1. Geoffrey, S.H. 1992. Bioremediation: Myths vs. reali-
ties. U.S. Environmental Protection Agency.
2. Picro, A.M., D. Kafkewitz, C.-M. Kung, and G. Le-
wandowski. 1992. Dehalogenation and mineraliza-
tion of 2,4,6-frichlorophenoI by sequential activity of
anaerobic microbial populations. Biotechnology Let-
ters 14(2).
mg/L DDT
Gas Production
Date
Figure 1. Qaa production for anaerobic DOT raactor over a ft-month period.
¦ mofL HCB
Gas Production
BOO
filiiniii
T 3
CO
CJ
X
e
NNNNIVNNNNNINM
^ ^ ^ ^ ^ ^ O ^ O ^ ^ ^
cmnnnnnnnnncmnnn
«0 «0
oj fi p IS n n ,
C!5SCC!C!oS!:CJ
CO 01 O) 9 r O
sss
- ~
Data
Figure 2. Gas production for anaerobic hexaclorobenzena reactor over a 6-month period.
107
-------
PCB Biodegradation During Aerobic Treatment of Sludge from the French
Limited NPL Site
J.W. Anderson and T. Smith
ManTech Environmental Technology, Inc., Robert S. Kerr Environmental Research Laboratoy, Ada, OK
J.T. Wilson
U.S. Environmental Protection Agency, Robert S. Kerr Environmental Research Laboratory, Ada, OK
French Limited is a Superfund site near Houston, TX.
The site is a sand pit that was used for disposal of
chemical waste, primarily pofycyclic aromatic hydrocar-
bons (PAHs), from API separator sludge, and polychlori-
nated biphenyls (PCBs), from discarded transformers.
At the French Limited Superfund Site, it had been
claimed that PCBs codisposed with API separator
sludge were degraded during aerobic biodegradation of
the petroleum hydrocarbons. The concentration of
PCBs in the sludge was reduced by a factor of 10, and
the sludge volume was reduced by a factor of 10.
Aerobic biodegradation of the lower chlorinated PCBs in
the presence of PAHs, in particular biphenyl, has been
demonstrated, but the degradation of higher chlorinated
species is more difficult (1,2,3). Therefore, the confirma-
tion of the observed PCB degradation in a controlled
laboratory experiment was desirable. A laboratory ex-
periment was proposed that would determine quantita-
tively the changes in the sludge under highly aerobic
conditions. The purpose of this experiment was to de-
termine if the vendor's claims of aerobic biodegradation
of the PCBs under highly aerobic conditions were valid.
Although the aerobic conditions were maintained for 1
year and significant degradation of the organics in the
aerobic reactor, measured as total petroleum hydrocar-
bons (Table 1), was observed, the total PCB concentra-
tion was not "significantly affected (Table 2). Only certain
of the lower chlorinated PCB congeners were signifi-
cantly degraded under aerobic conditions, specifically
BZ-7/9, BZ-6, and BZ-8/5 (Table 3). These congeners
were dichlorobiphenyls. Figures 1 and 2 illustrate that
the consumption of these PCB congeners occurred after
biological activity had decreased. No congener was
significantly degraded under anaerobic conditions
(Table 3).
Tabic t. Total Petroleum Hydrocarbon (TPH) Degradation
TPH (mg/L) Avg. SD (n * 3)
Start
18,919
5,944
Final
Anaerobic
18,566
864
Aerobic
6,939
369
Table 2. Polychlorinated Biphenyl (PCB) Depletion (mg/L. of
PCBs In Mixed Uquor; Average of Duplicate
Samples)
Aerobic Anaerobic
Initial 18.5 15.9
Final 14.5 14.B
References
1. Ahmed, M., and D.D. Focht. 1973. Can. J. Microbiol.
19 (47).
2. Baxter, R.A., R.E. Gilberg, R.A. Lidgett, J.H. Main-
prize, and H.A. Vodden. 1975. Sci. Total Environ. 4
(53).
3. Bedard, D.L., R.E. Wagner, M.L. Brennan, J.L
Haberl, and J.F. Brown Jr. 1987. App!. Environ. Mi-
crobiol. 53 (1094).
108
-------
Tab la 3. Changes in Concentration of PCB Congeners During Incubation under Aerobic aid Anaerobic Conditions
Aerobic Anaerobic Reduction (%)
Before
After
Before
After
Aerobic
Anaerobic
BZ-10/4
2.80
3.01
2.94
2.45
-7.6
16.9
BZ-7/9
0.30
0.00
0.38
0.39
100.0
-2.9
BZ-6
0.48
0.14
0.65
0.65
70.7
0.1
BZ-8/5
2.81
0.49
3.67
3.70
82.4
-0.8
BZ-31
2.56
2.41
2.54
2.57
5.8
-1.2
BZ-28
3.26
3.37
3.28
3.36
-3.3
-2.5
BZ-33/53
4.22
3.35
421
4.26
20.5
-12
BZ-52
4.68
520
4.48
4.60
-11.2
-2.6
8Z-44
2.67
2.87
2.57
, 2.63
-7.5
-2.6
Note: Concentrations expressed as percent of total PCBs.
Corr. Coat. =» 0.993 (rem Day 26 to Day 103
Oxygen Demand > 892 m^1/day
150
Days
200
250
300
Figure 1. Aerobic reactor oxygen consumption.
3
a.
9
a.
0.8
0.7
0.6
as
0.4
0.3
0.2
0.1
0
x s Anaerobic
+ = Aerobic
*
-~
50
100 150
Days
200
Figure 2. BZ-7/9 2,4- and 2,5-dlchlorobi phenyl.
250
109
-------
Innovative Bioremedlatlon Strategies for Creosote: Geographic Diversity of PAH
Degradation Capabilities at Wood-Treating Sites
James G. Mueller and Suzanne E. Lantz
SBP Technologies, Inc., Gulf Breeze, FL
Richard Devereux, Deborah L Santavy, and P. Hap Pritchard
Gulf Breeze Environmental Research Laboratory, U.S. Environmental Protection Agency, Gulf Breeze, FL
The use of specially selected microorganisms to en-
hance bioremediation efforts has proved effective in a
number of applications, especially when combined with
bioreactor systems. The successful use of such isolates
for the remediation of soil and water contaminated with
organic wood preservatives (e.g., creosote and pen-
tachlorophenol [PCP]) has resulted in the opportunity to
employ these technologies at similarly contaminated
sites throughout the world. Prior to worldwide dissemi-
nation of bioremediation strategies, however, concerns
regarding the introduction of foreign biota had to be
addressed Therefore, a research program was initiated
to ascertain: 1) whether microorganisms similar to those
used in previous bioremediation strategies could be
found in other soils; and 2) if so, whether tfie introduction
of these isolates would offer any advantage to the biore-
medation system.
Materials and Methods
To address these issues, soils contaminated with poly-
cyclic aromatic hydrocarbons (PAHs) were collected
from creosote-contaminated sites in Norway, Germany,
and the United States and screened for the presence of
bacteria capable of using phenanthrene (PHE) or
fluoranthene (FLA) as a sole source of carbon and
energy. Two soils from farmland in south-central Illinois,
with no known history of PAH exposure, were also sur-
veyed. Soil slurries (10 percent w/v) were prepared with
500 mg/L PHE or FLA in a sterile mineral salts medium
(1) and incubated in the dark with shaking (150 rpm) at
30°C for 14 days. Following 14 days of aerobic incuba-
tion and enrichment, cultures were diluted 1:5 (v/v) with
fresh mineral salts medium containing PAHs. This trans-
fer procedure was repeated two more times for a total
of four enrichments over a 10-week period
At selected time points, liquid samples were removed
from each vessel and screened for the presence of
bacteria capable of using PHE and FLA as primary
growth substrates. Once individual colonies were puri-
fied, they were transferred to carbon-free mineral salts
agar plates and complex.agar plates, then overlaid with
PHE or FLA. Plates were incubated for 14 days prior to
scoring for individual colonies exhibiting zones of clear-
ing of the PAH substrate. Colonies demonstrating PAH-
clearing abilities were purified and transferred to 125 mL
Erlenmeyer flasks containing 25 mL mineral salts broth
plus 500 mg/L PHE or FLA. Cultures were incubated for
5 to 7 days with shaking (150 rpm) at 30°C. Bacterial
growth at the expense of the PAH substrates was meas-
ured visually or by monitoring changes in absorbance at
OD^nm. As a control, growth in carbon-substrate free
mineral salts broth also was monitored.
Once the PAH-degrading ability of purified cultures was
validated, cultures were characterized by fatty acid pro-
file analysis (GC-FAME) and substrate utilization pat-
terns (Biolog Microplate System). The taxonomic
relationships among these strains was analyzed by
evaluating similarity measures from GC-FAME and sub-
strate utilization patterns with principal component
analysis (2). DNA/DNA hybridizations and 16S rRNA
sequence comparisons also were performed to deter-
mine the phyiogenetic relationships among the recov-
ered isolates according to published methods (3-6).
In parallel studies, microbial respirometric responses
(rate of liberation of carbon dioxide and the simultane-
ous consumption of oxygen) upon exposure to the fol-
lowing carbon sources were measured: 1) 500 mg/L
naphthalene (NAH), PHE, or FLA; 2) 500 mg/L readily
utilizable carbon (250 mg/L glucose + 250 mg/L glyc-
erol); or 3) 500 mg/L specification creosote No. 450
110
-------
(American Wood-Preserver's Association). Responses
were recorded at 8-hr intervals over an 8-day incubation
period (23°C, 100 rpm shaker speed) with a MicroOxy-
max respirometer (Columbus instruments, Columbus,
Ohio). These responses were compared to those ob-
served upon treatment with no supplemental carbon and
killed-cell controls (acidified to pH 2.0 with 1 N hydrogen
chloride plus 3.7 percent formaldehyde) to discern the
effect of nutrient amendment and aeration. At the end of
each incubation period, slurries from nutrient-amend-
ment-only, creosote-amended, and kii!ed-cell (control)
treatments were extracted and analyzed for the pres-
ence of creosote constituents as previously described
(7). These values were compared with those determined
at time zero with each soil.
Data Summary
Soils with previous PAH exposure possessed elevated
numbers of PAH degraders that corresponded with in-
creased respiratory activity and enhanced PAH biode-
gradation over uncontaminated soils. Following
enrichment with PHE or FLA in soil slurries, only soils
with known PAH contamination yielded bacteria that
grew at the sole expense of these PAH substrates. Thus,
all soils contaminated with PAHs harbored indigenous
bacteria competent for PAH-degradation, but soils with
no prior exposure to PAHs did not respond to the addi-
tion of these potential substrates.
The location of isolated strains and the corresponding
PAH substrate are summarized in Table 1. Physiological
Table 1. Bacteria Isolated from PAH-Contaminatsd Sites (or Their Ability to Degrade Phenanthrene or Fluoranthene
Soil of Origin
Isotate
Number
Culture
Number
Source/Reference
Enrichment
Substrate
1
EPA505
(10)
Fluoranthene
Pensacola, Florida
2
PJC 2282
G8ERL
Fluoranthene/pyrene
Pensacola, Florida
3
PJC 2283
G8ERL
Fluoranthene/pyrene
Live Oaks, Florida
4
PJC 2285
GSERL
Fluoranthene
Live Oaks site, Florida
S
PJC 2286
GBERL
Fluoranthene
Live Oaks site, Florida
6
PX 2287
GBERL
Fluoranthene
Pensacola, Florida
7
PX 2288
GBERL
Phenanthrerte
Pensacola, Florida
8
PX 2289
GBERL
Phenanttirene
Live Oaks, Florida
9
PX 2295
GBERL
Phenanthrene
Pensacola, Ron da
10
CRE7
(1)
Phenanthrene
Pensacola, Florida
11
CRE11 *
(1)
Phenanthrene
Pensacola, Florida
12
CRE12
W
Phenanthrene
Pensacola, Florida
13
AK Phen6
(11)
Phenanthrene
Prince William Sound, Alaska
14
N1F1
This Study
Fluoranthene
Rade, midt pa bygg, Norway
15
N2P5
This Study
Phenanthrene
Rade, ved kum, Norway
16
N2P6
This Study
Phenanthrene
Rade, ved kum, Norway
17
N3P2
This Study
Phenanthrene
LJIIestrom, karusell, Norway
18
N3P3
This Study
Phenanthrene
Lillestrom, karusell, Norway
19
N3F1
This Study
Fluoranthene
Lidestrom, karusell, Norway
20
N3F2
This Study
Fluoranthene
Lilestrom, karusell, Norway
21
N4F4
This Study
Fluoranthene
Lilestrom, provebronn, Norway
22
N5F4
This Study
Fluoranthene
Orammen, Norway
23
N6F4
This Study
Fluoranthene
Hommeh/rk, ved tonner, Norway
24
G1F1
This Study
Fluoranthene
Germany
25
G1F2
This Study
Fluoranthene
Germany
26
G1P1
This Study
Phenanthrene
Germany
27
G2P2
This Study
Phenanthrene
Germany
28
FDA PYR-1
(12)
Pyrene
Port Arkansas, Texas
29
SR3
(13)
Pentachlorophenol
Northwest Florida
111
-------
and biochemical analyses of isolated bacteria showed
that the PAH-degrading bacteria that recovered from
PAH-contaminated soil via the specified techniques ap-
peared to be related phylogenetically. For example, all
soils with a history of PAH exposure harbored bacteria
competent for PAH biodegradation, and principal com-
ponent analysis of GC-FAME data showed relatedness
between these strains (Figure 1). The strategy em-
ployed to enrich PAH-degrading bacteria thus appeared
to select similar types of microorganisms that are indige-
nous to contaminated soils at each site. Hence, the
export/import of these non-indigneous bacteria to aug-
ment bioremediation efforts would not seem to represent
the introduction of exotic biota.
2288,228»,
2295
N2P6, CRE7,
N2PS
30 -
c
N3F1
CRE12
20 -
u1P2
CRE11
a
Q.
2287
AK phwifi. 2288
G1P1.eiF1.SR3
-10
Q1F2 EPA50S
2285
-20
-40
-20
0
20
40
60
Principal Component 2
Bacterial similarity measurements using principal component
analysis from GC-FAME data (see Table 1 lor soil of origin
and PAH enrichment substrates).
Figure 1. GC-FAME principal component analyses.
From a PAH biodegradation perspective, the strategy of
relying exclusively on indigenous PAH degraders in
bioremediation efforts needs to be closely evaluated for
its ability to achieve site-specific cleanup standards in a
timely manner. If the stimulatory effect of controlled
nutriation, mixing, and aeration on the activity of the
indigenous microflora results in acceptable rates and
extents of biodegradation of targeted chemicals, tfien,
on a site-specific basis, it may be possible to rely solely
on the activity of such microorganisms to facilitate site
remediation (8,9). Alternatively, utilization of non-indige-
nous microbes in optimized bioremediation systems
could be advantageous for cost-efficient, effective biore-
mediation, while posing no discernible ecological risk.
Acknowledgments
We thank Myke'lle Hertsgaard and Barbara Artlet (Tech-
nical Resources, Inc., Gulf Breeze, Florida), Sheree
Enfinger (U.S. EPA, ERL, Gulf Breeze, Florida), and
Stephanie Willis (University of New Hampshire) for tech-
nical assistance; Brian Klubek (Southern Illinois Un'iver-
sity-Carbondale, Illinois) for soil analyses; Bruce
Hemming (Microbe Inotech Laboratories, St. Louis, Mis-
souri) for help in the interpretation of GC-FAME and
Biolog results; and Peter Chapman (U.S. EPA, ERL,
Gulf Breeze, Florida) and Carl Cerniglia (U.S. Food and
Drug Administration, NCTR, Jefferson, Arkansas) for
donating PAH-degrading strains for comparative analy-
ses. Soils from Norway and Germany were provided by
Jim Berg (Aquateam, Norway) and Wolfgang Fabig
(Umweltshutz Nord, Germany).
Financial support for these studies was provided by the
Norwegian State Railway (NSB), and the U.S. EPA (Gulf
Breeze). These studies were performed as part of a
Cooperative Research and Development Agreement
between the Gulf Breeze Environmental Research
Laboratory and SBP Technologies, Inc. (Atlanta, Geor-
gia) as defined under the Federal Technology Transfer
Act, 1986 (contract no. FTTA-003).
References
1. Mueller, J.G., P.J. Chapman, and P.H. Pritchard.
1989. Action of a fluoranthene-utilizing bacterial
community on polycyclic aromatic hydrocarbon
components of creosote. Appl. Environ. Microbiol.
55:3085-3090.
2. Jacobs, D. 1990. SAS/GRAPH software and nu-
merical taxonomy. In: Proceedings of the 15th An-
nual Users Group Conference. SAS Institute, Inc.,
Cary, NC. pp. 1413-1418.
3. Amann, R.I., C. Lin, R. Key, L. Montgomery, and
D.A. Stahl. 1992. Diversity among Fibrobacter iso-
lates: Towards a phylogenetic classification. Syst.
Appl. Microbiol. 15:23-31.
4. Devereux, R„ S-H. He, C.L. Doyle, S. Orkland, D.A.
Stahl, J. LeGall, and W.B. Whitman. 1990. Diversity
and origin of Desulfovibrio species: phylogenetic
definition of a family. J. Bacteriol. 172:3609-3619.
5. Lane, D.J., B. Pace, G.J. Olsen, D.A. Stahl, M.L.
Sogin, and N.R. Pace. 1985. Rapid determination
of 16S ribosomal RNA sequences for phylogenetic
analyses. Proc. Natl. Acad. Sci. USA 82:6955-6959.
6. Weisburg, W.G., S.M. Barnes, D.A. Pelleteir, and
D.J. Lane. 1991.16S ribosomal DNA amplification
for phylogenetic study. J. Bacteriol. 173:697-703.
112
-------
7. Mueller, J.G., S.E. Lantz, B.O. Blattmann, and P.J.
Chapman. 1991. Bench-scale evaluation of alterna-
tive biological treatment processes for the remedia-
tion of pentachlorophenol- and creosote-con-
taminated materials: Solid-phase bioremediation.
Environ. Sci. Technot. 25:1045-1055.
8. Berg, J.D., B. Nesgard, R. Gundersen, A. Lorent-
sen, and T.E. Bennett 1993. Washing and slurry
phase bio treatment of creosote-contaminated soil.
In: Proceedings of In Situ and Onsite Bioreclama-
tion Symposium, April 5-0, 1993, San Diego, CA.
(in press)
9. Berg, J.D., T.E. Bennett, B.S. Nesgard, and J.G.
Mueller. 1993. Treatment of creosote-contaminated
soil by soil washing and slurry-phase bioreactors.
In: Proceedings, International Symposium on Envi-
ronmental Contamination in Central and Eastern
Europe. October 12-16, 1992. Budapest "Hungary,
(in press)
10. Mueller, J.G., P.J. Chapman, B.O. Blattmann, and
P.H. Pritchard. 1990. Isolation and characterization
of a fluoranthene-utilizing strain of Pseudomonas
paudmobilis. Appl. Environ. Microbiol. 56:1079-
1086.
11. Mueller, J.G., S.M. Resnick, M.E. Shelton, and P.H.
Pritchard. 1992. Effect of inoculation on the biode-
gradation of weathered Prudhoe Bay crude oil. J.
Indust Microbiol. 10:95-105.
12. Heitkamp, M.A., and C.E. Cerniglia. 1988. Minerali-
zation of polycydic aromatic hydrocarbons by a
bacterium isolated from sediment below an oil field.
Appl. Environ. Microbiol. 54:1612-1614. t
13. Resnick, S.M., and P.J. Chapman. 1990. Isolation
and characterization of a pentachlorophenol-de-
grading, gram-negative bacterium. Abstr. Ann.
Meet. Am. Soc. Microbiol, p. 300.
113
-------
Section 6
Development of Computer-Based Assessment Systems
Potential applications of computer technology promise to assist researchers and
field personnel in making assessments related to the bioremediation of hazardous
waste sites. For instance, computer tools conceivably could collect and even
analyze information on the opportunities for successful site bioremediation, the
appropriateness of various methods under consideration, and the development of
cost-effective process designs. Moreover, a computer could perform such assess-
ment functions quickly and comprehensively.
Two poster presentations outlined database-oriented research. In one project, an
expert system is being developed for identifying the most appropriate method for
bioremediating petroleum-contaminated soils based on the parameters of the
particular site. The second project involves the development of software that can
be used by engineers and managers as a visualization tool when characterizing
contaminated sites.
115
-------
Design of an Expert System to Select an Appropriate
Bioremediation Technique
Raymond C. Loehr and Greg E. Schmidt
Environmental and Water Resources Engineering Program, College of Engineering,
University of Texas, Austin, TX
This project is developing a decision-making format that
can be used by individuals interested in determining the
appropriate bioremediation technologies to remediate
soils contaminated by spills of petroleum products. The
resulting information can be used for emergency re-
sponse and remediation decisions. The project will inte-
grate available information that the user can understand
and utilize in a short time. The output will provide guid-
ance to identify appropriate bioremediation action op-
tions for a specific situation. Thus, the project output will
meet the following objectives: 1) advance the under-
standing of applying bioremediation to specific prob-
lems; 2) enhance the use of bioremediation through
performance evaluations and technology transfer; 3)
evaluate the use of natural soil processes for treatment
of hazardous wastes; and 4) address informational im-
pediments tfiat constrain the use of new technologies.
The project will focus on in situ and ex situ bioremedia-
tion approaches that can be used for unsaturated soils.
The technologies considered in this project are:
In Situ Ex Situ
Btoventing Slurry reactor treatment
Air sparging Prepared bed treatment
Land treatment Aerated pile treatment
Passive remediation
As used in this project, petroleum spills are understood
to consist of spills of gasoline, diesel, JP-4 fuel, proc-
essed and crude oils, and other raw, partially processed,
and refined petroleum materials.
The anticipated users of the project output are expected
to be: 1) emergency response team personnel; 2) onsite
coordinators; 3) regional and state project managers;
4) industry personnel; 5) consulting engineers; and
6) educational organizations interested in training stu-
dents and others in need of the information.
Bioremediation
Bioremediation is the controlled application of the natu-
rally occurring process of biodegradation. The process
has been recognized as a relatively inexpensive and
efficient method for removing organic chemicals from
contaminated soils.
Despite the application of biological processes to many
waste treatment situations, the wider use of bioremedia-
tion for contaminated soils continues to be limited by a
general lack of understanding of the fundamentals that
are involved and the processes that can be used. To
date, each application has been, in many respects, an
isolated occurrence that is unable to benefit from past
experience and unable to provide guidance for future
use. In recent years, however, both the relevant funda-
mentals for use with contaminated soils and the techni-
cal processes that can be applied in bioremediation
have become more widely understood. Such knowledge
will be used in this project.
Bioremediation processes have been used for soils that
have contained: 1) petroleum hydrocarbons such as
gasoline, diesel fuel, crude oil, and creosote; 2) pesti-
cides and their derivatives; 3) chlorinated solvents such
as methylene chloride; and 4) hydrocarbons such as
pentachlorophenol, naphthalene, and anthracene. An
increasing number of sites are being remediated and
dosed using bioprocesses, magnifying the need for an
expert system.
Programming Language and Format
The expert system will be designed for a personal com-
puter format. The system will be available for use by the
audience discussed above.
The computer language used in the development of the
expert system had to meet three criteria. First, it had to
be in use throughout the country and not require a
116
-------
substantial financial investment or be difficult to set up.
This requirement would ensure a wide user base. Sec-
ond, the language had to be easy to upgrade and edit.
This would allow for updating of the program to stay up
to date with the changing field of bioremediation. Third,
the language had to have superior graphics capabilities.
This would be needed to effectively communicate the
various bioremediation techniques and illustrations that
would accompany text. The language that directly meets
all three criteria is HyperCard, a language developed for
information communication that is standard on all
Macintosh systems.
The format of ttie program will contain three main
branches:
Branch 1 Explanation of techniques
Branch 2 Description of illustrating methods
Branch 3 Analysis of the user's situation
The first branch will allow the user to become familiar
with the various bioremediation techniques. It will in-
clude a written explanation of each process and will list
the various soil or contamination criteria under which
each can be effectively applied. This branch will provide
a user who was exposed only recently to the bioreme-
diation of hydrocarbons, for example, with a solid back-
ground; or will provide a comprehensive review of a
technique that a person has not used in a while.
The second branch of the program will provide the user
with case studies that illustrate decision processes.
These studies will allow the user to match information
with techniques. Then the expert system will either af-
firm the user's choice and provide supporting evidence,
or correct the choice and provide the reasoning. This
•branch will allow the user to get a feel for the decision-
making approach one should follow.
The final branch of the program will allow the user to
input information pertaining to a specific situation. The
expert system then will output the "best" choice or
choices for this case and the logic behind the decision.
Note that the program is oniy an aid to the user. No
program can adequately replace the logic of a user
intimately involved with the particular situation. The pro-
gram will only show the correct thought process and
output the best possibilities. The user then should take
these possibilities and thoroughly consider them, taking
into account parameters that the computer could not.
Summary
An expert system is being developed to acquaint system
users with the various bioremediation techniques appli-
cable in the unsaturated zone and to indicate the condi-
tions for which particular techniques are appropriate.
The program will contain three main branches. The first
will describe the bioremediation processes. The second
will allow the user to view the decision process ttirough
case studies, and the final branch will allow the user to
input specific information so that the program can output
suggestions on available techniques.
117
-------
A Data Visualization System for Bioremediation Analysis
Lewis A. Rossman
Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
Kevin Savage
» Center Hill Research Center, University of Cincinnati, Cincinnati, OH
John Franco
Department of Computer Science and Engineering, University of Cincinnati, Cincinnati, OH
Site characterization data play an essential role in de-
termining the feasibility, design, and operation of a biore-
mediation system. These data represent a series of
point measurements taken from wells, borings, and test
pits of a highly heterogeneous, three-dimensional sub-
surface environment The site engineer is faced with the
task of interpreting these test-point measurements from
a variety of perspectives. For example, the feasibility of
a bioremediation strategy may depend on identifying a
sufficient volume of contaminated media located within
an acceptable soil stratum that is free of biological toxi-
cants. Site characterization data should be displayed in
a meaningful way so that they can be properly inter-
preted and used correctly when making remedial action
decisions. The goal of this project is to develop a pow-
erful yet simple-to-use data visualization tool for site
engineers and managers that assists them in evaluating
the feasibility of bioremediation at contaminated waste
sites.
Approach
Our data visualization software, called ICASE (Inte-
grated Computer Assisted Site Evaluation), takes meas-
urements made at irregularly spaced locations at a site
and uses a stratified interpolation process to estimate
values within a uniform three-dimensional grid under-
neath the site. This procedure is carried out for any
parameters of interest, which can include both soil prop-
erties and chemical species. It then is possible to display
the geographical distribution of single parameters or
pairs of parameters on a map of the site in a variety of
formats that include:
• Shaded contour diagrams on a plan view of the site
within any depth layer.
• Two-dimensional cross-sectional views across the
site (in the form of single slices, fence diagrams, or
an animated sequence of slices).
• Three-dimensional shaded volume views (similar to
medical imaging) within the subsurface, with the abil-
ity to slice open the volume along any plane to get a
cross-sectional view.
In addition to the data visualization methods, ICASE is
being equipped with a bioremediation knowledge base.
This knowledge base is a collection of rules and condi-
tions (perhaps even a scoring system) that will infer the
feasibility of a particular form of bioremediation technol-
ogy based on conditions measured at the site. These
conditions include soil and aquifer geotechnical proper-
ties along with contaminant properties and nutrient lev-
els. At the user's command, ICASE will let the
knowledge base reason over the site's characteristics
and graphically display areas where a particular type of
bioremediation technology may (or may not) be techni-
cally feasible.
Current Status
The first year of the project has focused on developing
map display and data visualization features. Base map
display, ad hoc data queries, contouring, and two-di-
mensional cross-sectioning have all been completed.
The three-dimensional visualization methods currently
are being enhanced. The major effort in the second year
will be the development of the bioremediation knowl-
edge base, its incorporation into ICASE, and the testing
of the completed system on a suite of site remediation
case studies.
118
-------
Section 7
Hazardous Waste Research Center Program
A number of presenters at the symposium's poster session provided information on
bioremediation projects being carried out under the auspices of EFA's Hazardous
Substance Research Centers (HSRCs). EPA established the HSRC program in
response to provisions in the 1986 amendments to the Comprehensive Environ-
mental Response, Compensation, and Liability Act (CERCLA). These provisions
authorized EPA to establish HSRCs with a mission to study all aspects of the
"manufacture, use, transportation, disposal, and management of hazardous sub-
stances" and made the Agency responsible for the "publication and dissemination
of the results of such research." The program is managed by the director of EPA's
Office of Exploratory Research (OER) in the Office of Research and Development.
EPA has established five research consortia, with each serving two adjacent federal
regions. These include:
Northeast Hazardous Substance Research Center—Region-Pair 1 and 2, which
includes the New England states, New York, New Jersey, and the territories of
Puerto Rico and the U.S. Virgin Islands. The lead institution is the New Jersey
Institute of Technology, and the center's director is Dr. Richard Magee. Other
consortium partners include the Massachusetts Institute of Technology, Tufts Uni-
versity, Rutgers University, Stevens Institute of Technology, Princeton University,
and the University of Medicine and Dentistry of New Jersey.
A poster presentation from this center concerned research into the effect on
sequencing batch reactor design of kinetic interactions in microorganism popula-
tions and among pollutants.
Great Lakes and Mid-Atlantic Hazardous Substance Research Center— Re-
gion-Pair 3 and 5, which comprises the Great Lakes states and the mid-Atlantic
states of Virginia, West Virginia, Maryland, Pennsylvania, and Delaware. This
three-university consortium is headed by Dr. Walter Weber of the University of
Michigan; Michigan State University and Harvard University are partner institutions.
This center is developing an expert system to help field practitioners assess
bioremediation approaches for a variety of contamination sites. Other researchers
are investigating the use of ozone, alone and in combination with hydrogen perox-
ide, to mineralize 2,4,6-trinitrotoluene. For this study, chemical ozone was chosen
for pretreatment of contaminated soil because it is the most benign and field-appli-
cable oxidant.
South/Southwest Hazardous Substance Research Center— Region-Pair 4 and
6, which is made up of Gulf Coast and southern states. Louisiana State University
heads this center, in partnership with the Georgia Institute of Technology and
Rice University. The center's director is Dr. Louis Thibodeaux of Louisiana State
University.
119
-------
This center sponsored research into the reversibility of naphthalene adsorp-
tion/desorption on soil.
Great Plains and Rocky Mountain Hazardous Substance Research Center—
Region-Pair 7 and 8, which includes the states on the eastern side of the Great
Basin along with the Great Plains states. This large consortium is run by Dr. Larry
Erickson of Kansas State University. The other six participating institutions are
Montana State University and the Universities of Iowa, Missouri, Montana, Ne-
braska, and Utah.
Research carried out by this center demonstrated the use of computed tomography
for studying small-scale, three-dimensional, nonaqueous phase liquid movement
and entrapment in porous media. Another study used a curve-fitting model with a
single Monod kinetic expression as a means of estimating rates of in situ biodegra-
dation at contamination sites. A third project evaluated the contribution of plants
and associated microflora to the biodegradation of recalcitrant organic compounds
in soil.
Western Region Hazardous Substance Research Confer—Region-Pair 9 and
10, which includes the West Coast states along with Alaska, Arizona, Hawaii, and
Idaho. Stanford University and Oregon' State University make up this consortium.
Dr. Perry McCarty of Stanford University is the center's director.
A pair of researchers affiliated with this center presented information derived from
modeling the microbiological transformation of pollutants. This study focuses on the
need for scaled-up knowledge about in situ biorestoration from the pore to the field
level to facilitate measurement, prediction, and assessment in ground water and
soil remediation.
Another poster illustrated the use of a vertical recirculation well to promote
cometabolic in situ transformation of vinyl chloride using methanotrophic bacteria.
A focus of the presentation was the evaluation of different oxygen and methane
delivery systems. Other researchers presented an evaluation of trichloroethylene
(TCE) destruction carried out by methanotrophic bacteria in a two-reactor treatment
system.
Researchers based at Stanford conducted teste on benzene, toluene, ethytben-
zene, and xylene (BTEX) contaminated soil using In situ bioreactors and laboratory
microcosms. Data from these experiments, which were conducted using soil from
Seal Beach, California, indicate that a combination of the electron acceptors nitrate
and sulfate may enhance anaerobic biodegradation of BTEX.
Another study involved the use of biologically active sediment from an oil refinery
waste pond as inoculum in nitrate-containing enrichment cultures; of the eight
aromatic substrates present in the laboratory cultures, indole and quinoline under-
went a transformation.
A study of aerobic TCE-degrading bacteria is using Nitrosomonas europaea as a
model system to examine the physiological consequences of TCE toxicity on the
cometabolic process. In other research, preliminary studies examined the effects
of certain environmental conditions on the rate of anaerobic biotransformation of
chlorophenols.
120
-------
Development of a Knowledge-Based Bioremediation Adviser
Shu-Chi Chang, Peter Adriaens, Iris D. Tommelein, and Timothy M. Vogel
Great Lakes and Mid-Atlantic Hazardous Substance Research Center, University of Michigan, Ann Arbor, Ml
Bioremediation is an increasingly promising technology
for cleaning up hazardous waste sites, but because of
a lack of data and knowledge to make informed deci-
sions, owners and consultants often do not think of it as
a viable treatment alternative. Yet bioremediation is es-
tablishing itself with a credible track record as a cost-ef-
fective technology for certain types of contaminated
sites. A limitation, however, is that bioremediation proc-
esses are not universally applicable. Thus, it is important
to determine whether bioremediation can be applied on
any specific site, before resorting to other treatments.
This study aims at gathering and classifying disparate
pieces of knowledge on bioremediation processes in
such a way that they can be used in an interactive
knowledge-based advisory system, called the Bioreme-
diation Adviser. Such an advisory System would be use-
ful to field practitioners in determining the applicability of
bioremediation and enable novices to gain insight in the
factors affecting the technical feasibility and success of
bioremediation.
The Bioremediation Advisor will help bridge the gaps
among practitioners' collection of traditional data, theo-
retical knowledge about bioremediation treatment proc-
esses and characteristics of pollutantdegrading
microbes, and biological characteristics of the existing
environment (Figure 1). Ultimately, this information will
be used in conjunction with a ruie-based system to guide
field data collection and to match field data with biore-
mediation technologies to suggest the feasibility of
bioremediation as a treatment for a given contamination
and recommend possible bioremediation technologies.
Background
Whether bioremediation is feasible for cleaning a con-
taminated site depends on many factors including the
characteristics of the spill and the site conditions, and
knowledge of the degradability of the contaminant
These factors, in turn, depend on knowledge of the
contaminants (e.g., physical and chemical properties),
and of the microorganisms present (e.g., enzyme-pro-
ducing capabilities) in the contaminated environment
The literature contains considerable information on
laboratory experiments on the degradation of pollutants,
albeit studies with pure microbial cultures or undefined
mixed cultures. Although such information will be con-
tained within the Bioremediation Advisor, some of it will
be precluded from application in the field. Alternatively,
some laboratory studies have provided direct evidence
for the potential of either indigenous or added microbes
in bioremediation schemes.
Figure 1. Proposed expert system scope.
Results and Conclusions
To date, a literature survey has been performed, includ-
ing a review of papers and textbooks, pertaining to
the biological degradation as well as the chemical
properties of more than 70 organic contaminants. The
characteristic chemical properties and observed biode-
gradation of 20 relevant contaminants of concern, rang-
ing from halogenated aromatic and aliphatic compounds
to pesticides, have been classified in HyperCard stacks.
Environmentally relevant chemical properties include re-
dox potential, aqueous solubility, organic partition coef-
ficient, and molecular structure. When values were not
available in the literature, they were estimated based on
established methods. The compiled biodegradation pa-
rameters include the nature of the degradation mecha-
121
-------
nism (i.e., oxidation or reduction, and mineralization or
transformation), and the environmental redox potential
under which degradation may occur.
Since both the redox of the environment and that of the
compound determine to a great extent what type of (and
whether) biodegradation will occur, correlations are cur-
rently being developed using these parameters for a
range of organic pollutants. The pollutants themselves
have been grouped according to a system reflecting
their relative redox potential, taking into account the
level of carbon atom substitution (scaled from 1 to 4),
the number of halogen atoms and hetero-atoms, and
presence or absence of aromaticity.
Even though data collection from the literature consti-
tutes the bulk of the work reported here, the HyperCard
format is continuously being developed to suit the
eventual implementation in the Bioremediation Adviser's
architecture.
122
-------
Use of Composting Technologies To Treat a TNT-Contaminated Soil
James H. Johnson, Jr., Mohammed Moshin, Lily Wan, and Abdul Shafagati
Great Lakes and Mid-Atlantic Hazardous Substance Research Center, Howard University, Washington, DC
Trinitrotoluene (TNT), an aromatic hydrocarbon known
to be carcinogenic and mutagenic, has been extensively
used by the military as an explosive for decades. As a
result of its uses and disposal techniques, the U.S.
government now owns approximately 1 million tons of
soils contaminated with TNT and other explosive com-
pounds.
Previous laboratory and field research has shown that
TNT can be microbially transformed. The transformation
products are the result of the reduction of the nitro
groups and include diaminonitrotoluene and aminodini-
trotoluene isomers. Many of these compounds appear
to be toxic. The result is that this method of transforma-
tion of TNT may not reduce risks to known populations.
The aim of the current project is the mineralization of
TNT. The approach is to pretreat a TNT-contaminated
soil using a chemical oxidant to yield trinitrobenzoic acid
(TNBA) followed by microbial transformation of TNBA
using composting technology. Many methods have been
developed to oxidize the methyl (-CH3) group(s) of a
benzene ring to a carboxyl (-COOH) group. Many, how-
ever, require a very harsh environment (i.e.t high-tem-
perature, high-pressure, or severe acidic or basic
conditions). The most benign and field-applicable oxi-
dant and the one chosen for this project is ozone (with
a radical initiator). Initial efforts using an aqueous sys-
tem have shown that as the number of nitro groups
increased (i.e., methylbenzene to TNT), the oxidizing
efficiency of ozone decreased (80 percent to less than
50 percent). The use of ozone in combination with hy-
drogen peroxide or ultraviolet light is currently under
investigation.
The disappearance of TNBA was investigated under
thermophilic (55°C) conditions. Ninety-eight percent of
the TNBA was reduced in a period of 20 days in a soil
matrix contaminated at 10 mg/kg. The high pressure
liquid chromatography indicated the presence of two
new peaks, suggesting the appearance of metabolites.
The identification of the metabolites and 14C-TNBA stud-
ies are under way.
123
-------
Effect of Pore-Scale Hydrodynamics on Bulk Reaction Rates
Bruce B. Dykaar and Peter K. Kitanidis
Western Region Hazardous Substance Research Center, Stanford University, Stanford, CA
The objective of this research is to determine the effec-
tive first-order volumetric reaction rate of a biologically
reactive solute. The effective reaction rate is the rate that
would be measured in a column experiment, and that
would subsequently be used in a transport model com-
mensurate wifh the scale of the Darcian velocity. The
fundamental question addressed is how does the effec-
tive reaction rate depend on the cumulative effects of
pore-scale geometry and mechanisms. A model is de-
veloped that includes geometrical structural features
such as pore shapes and surface roughness, spatially
variable transport mechanisms such as diffusion and
convective mass transfer, and spatially variable reaction
rates within a biofilm or microcolony.
The microbiological processes by which subsurface mi-
croorganisms transform pollutants are controlled by the
detailed pore-scale physical mechanisms that affect
the organisms' functioning. The understanding of the
mechanisms and processes controlling in situ bio-
restoration are best at the pore scale. For the purposes
of ground water and soil remediation, however, it is
desirable to make measurements, predictions, and as-
sessments at a much larger scale. )t is not feasible to
bridge this gap in scales by simply having a computa-
tional domain that spans botti. First the data are not
enough to specify the spatial structure with such detail;
second, current computational resources are inade-
quate to handle such large problems, even if the data
did exist. A method for upscaling from the pore to field
scale is required.
This study starts with a detailed description of the
mechanisms and processes thought to be significant in
the biodegradation of a reactive contaminant. The de-
tailed model includes: 1) the pore water velocity field
through a channel with variable surface roughness and
tortuosity; 2) spatially variable diffusion in the pore
water, biofilm, and solid matrix; and 3) spatially variable
reaction rates within the biofilm. The approach adopted
in this work is to average spatially the equations govern-
ing the pore-scale transport of the reactive contaminant
over an appropriately large volume. The upscaling meth-
odology is based on a moment method approach. The
upscaling procedure yields the parameters that describe
the transport of the less-detailed, spatially averaged sol-
ute concentration, and in particular the effective first-or-
der volumetric reaction rate.
As a first step, a simple geometric model of a porous
medium was used to examine the effects of the sub-
grain-size structural features and processes on the
measurable laboratory-scale reactivity of a biologically
reactive solute. In addition to providing some rough
insight into the problem at hand, it will provide a means
for verifying the subsequent, more complicated models.
The geometry of the problem consists of two parallel
plates a distance 2a apart. The surfaces of the parallel
plates are assumed to be coated with a shallow uniform
biofilm of a reactivity that is approximated with a first-or-
der surface reaction rate coefficient X. At the column
scale, the measurable quantity of interest is the global
scale solute concentration (C*), which is defined as the
point concentration averaged over the distance between
the plates. An analytical solution for the temporal behav-
. tor of C* is developed for the case of a solute undergoing
diffusion Dm and a first-order surface reaction L For a
very wide range of parameter values, including those
values associated with enhanced subsurface biological
activity, it is found that 1) for times satisfying a2/Dm
-------
In Situ Bioremediation Using a Recirculation Well
M.M. Lang, L. Semprini, and P.V. Roberts
Western Region Hazardous Substance Research Center, Stanford University, Stanford, CA
Perry L. McCarty (presenter)
Stanford University, Stanford, CA
A promising microbial system for in situ bioremecfiation
is aerobic cometabolic degradation by methanotrophic
bacteria. Methanotrophic bacteria produce the enzyme
methane monooxygenase (MMO), which can degrade
many hazardous volatile organic compounds. During
cometabolic degradation, the target contaminant is for-
tuitously degraded by the MMO that is produced in order
for the microorganisms to use the supplied electron
donor (methane). Studies at the Moffett Naval Air Station
(3,4,5,6) have demonstrated successful removal of-sev-
eral volatile organics, including trichloroethylene (TCE)
and vinyl chloride (VC), by promoting the growth of
methanotrophic bacteria in the subsurface.
Several investigators (7,8,9) have examined the flow
characteristics of vertical recirculation wells. This poster
investigates the use of a vertical recirculation well to
promote cometabolic transformation of VC using
methanotrophic bacteria. A vertical recirculation well con-
tains both an injection and extraction screen with a pump
between the two screens to force recirculation. Ground
water enters the well through the upper screen and is
pumped in the downward direction. Oxygen and meth-
ane are added to the water below the pump, and the
enriched water is injected through the lower screen. The
screens have a 10 m separation. The target contami-
nants are recirculated, which increases their contact time
with the biologically active zone, allowing a greater extent
of contaminant removal. In previous in situ bioremedia-
tion applications (3), the target contaminant passed
through the biologically active zone only once, and any
remaining contaminant required further treatment above
ground. The simulations presented here evaluate differ-
ent oxygen and methane delivery schemes for a recircu-
lation well operated to promote in situ bioremediation.
References
1. Little, C.D., A.V. Palumbo, S.E. Herbes, M.E. Lind-
strom, R.L Tyndall, and P.J. Gilmer. 1988. Trichlore-
thylene biodegradation by a methane-oxidizing
bacterium. Appl. Environ. Microbiol. 54:951-956.
2. Fox, B.G., J.G. Bourneman, L.P. Wackett, and J.D.
Lipscomb. 1990. Haloalkene oxidation by the soluble
methane monooxygenase from Methytosinus
trichosporium OB3b: Mechanistic and environmental
implications. Biochemistry 29:6419-6427.
3. Roberts, P.V., L Semprini, G.D. Hopkins, D. Grbid-
GalkS, P.L. McCarty, and M. Reinhard. 1989. In-situ
aquifer restoration of chlorinated aliphatics by
methanotrophic bacteria. Dept. of Civil Engineering,
Stanford University. Technical Report No. 310.
4. Semprini, L, and P.L. McCarty. 1991. Comparison
between model simulations and field results for in-
situ biorestoration of chlorinated aliphatics: Part 1.
Biostimulation of methanotrophic bacteria. Ground
Water 29 (3):365-374.
5. Semprini, L„ P.V. Roberts, G.D. Hopkins, and P.L.
McCarty. 1990. Afield evaluation of in-situ biodegra-
dation of chlorinated ethenes: Part 2. Results of
biostimulation and biotransformation experiments.
Ground Water 28 (5):715-727.
6. Semprini, L., G.D. Hopkins, P.V. Roberts, D. Grbid-
Galid, and P.L. McCarty. 1991. Afield evaluation of
in-situ biodegradation of chlorinated ethenes: Part 3,
Studies of competitive inhibition. Ground Water 29
(2)239-250.
7. Herrling, B„ J. Stamm, and W. Buermann. 1991.
Hydraulic circulation system for in-situ bioreclama-
tfon and/or in-situ remediation of strippable contami-
nation. Proceedings of In Situ and Onsite
Bioremediation, Int. Symposium. San Diego, CA.
8. Philip, R.D., and G.R. Walter. 1992. Prediction of flow
and hydraulic head fields for vertical circulation wells.
Ground Water 30 (5):765-773.
9. MacDonald, T.R., and P.K. Kitanidis. Modeling the
free surface of an unconfined aquifer near a recircu-
lation well. Ground Water, (in press)
125
-------
Chlorinated Aliphatic Hydrocarbon Biodegradation by Methanotrophic Bacteria
Laurence H. Smith, Tomas Henrysson, and Perry L. McCarty
Western Region Hazardous Substance Research Center, Department of Civil Engineering,
Stanford University, Stanford, CA
Methanotrophic bacteria, which oxidize methane for en-
ergy, have been found capable of oxidizing chlorinated
aliphatic hydrocarbons (CAHs) such as trichlo*
roethylene (TCE) by cometabolism. A two-reactor treat-
ment system for TCE destruction by methanotrophic
bacteria is being evaluated. The system consists of a
growth reactor (completely mixed) that continuously pro-
duces the active culture for addition to a TCE-contami-
nated waste stream in a separate transformation reactor
(plug flow). This reactor configuration eliminates TCE
transformation product toxicity during cell growth, bene-
fits from the controlled conditions in the growth reactor,
and gives enhanced TCE removal through the absence
of competitive inhibition and more favorable plug flow
kinetics in the transformation reactor.
Methane/TCE Interactions. The TCE transformation rate
by a methanotrophic mixed culture in batch studies was
found to be enhanced by the presence of a high con-
centration of methane (5 to 6 mg/L) at high TCE concen-
tration (5 to 9 mg/L). This result was contrary to the
expectation that the presence of methane would cause
a reduction in the TCE transformation rate because of
competitive inhibition. At a lower TCE concentration (0.9
mg/L), the TCE transformation rate was found to be
lower in the presence of methane, indicating that com-
petitive inhibition does occur. It appears that at the
higher TCE concentration, energy depletion and cell
inactivation attributable to TCE transformation product
toxicity controlled the TCE transformation rates, while
competitive inhibition controlled at the lower TCE con-
centration. These results indicate that methane addition
to the transformation reactor in the proposed treatment
system may be beneficial in some cases.
Mixed-Culture Growth Conditions. Two aspects of
growth reactor performance were studied, reactor solids
retention time (SRT) and nutrient nitrogen concentration
(nitrate nitrogen, mg N/L). Three laboratory growth re-
actors were operated at various combinations of 2-day
and 8-day SRT and 82 mg N/L and 165 mg N/L influent
concentrations. The results showed that similar TCE
transformation capacities (Tc, g TCE/g cells) could be
obtained at both SRTs and that the higher growth yield
of the 2-day SRT led to a higher TCE transformation
yield (Ty, g TCE/g methane). Nitrogen starvation did not
affect the Tc at the 2-day SRT, but led to higher growth
yields and higher Ty.
Poly-beta-hydroxybutyrate (PHB). The Tc of resting
(non-fed) methanotrophic cells has been suggested to
be a function of the cells' PHB content, which increases
under nutrient deficiency. PHB serves as an internal
source of reducing power when the energy substrate
(methane) is absent. When the cultures of two labora-
tory growth reactors were compared, Tc increased with
PHB content
The activity of the soluble methane monooxygenase
(MMOs, the enzyme that performs TCE oxidation) was
determined by measuring naphthalene oxidation rate.
The naphthalene oxidation rate was directly proportional
to PHB content in samples taken from the 8-day SRT
growth reactor, which had PHB contents of 2 to 8 per-
cent When the samples were amended with 20 mM
formate, a readily available electron donor, the specific
naphthalene oxidation rate was constant and higher
than in unamended samples. These results indicate that
in resting cells MMOa activity is limited by the availability
of a suitable source of reducing power.
126
-------
Anaerobic Biodegradation of BTEX Compounds at Seal Beach, California
Harold A. Ball, Martin Reinhard, and Eva O. Orwin
Western Region Hazardous Substance Research Center, Stanford University, Stanford, CA
Perry L. McCarty (presenter)
Stanford University, Stanford, CA
Orange County Water District built a facility at the Seal
Beach site in California where a significant gasoline spill
resulted in contamination of the ground water aquifer
with hydrocarbons including benzene, toluene, ethyl-
benzene, and m-, p-, and o-xylene (BTEX). The facility
was designed for the operation of in situ bioreactors
wherein strategies for bioremediation of the contami-
nated soil were tested. The bioreactors consisted of
aquifer sediment-filled stainless steel cylindrical vessels
with near complete capability to control and monitor both
hydrodynamic flow and influent/effluent composition.
The operation of two anaerobic/anoxic bioreactors fo-
cused, on developing the native microbial populations
under natural and induced denitrifying conditions. The
ground water at the site had a high naturally occurring
background concentration of sulfate (85 mg/L). The in-
fluents to the "natural background" control and denitrify-
ing reactors were contaminated ground water from the
site and contaminated ground water amended with ni-
trate (13 mg/L as nitrogen), respectively. Although some
biological removal of toluene and m,p-xylene was ob-
served in the nitrate-amended bioreactor, toluene and
m,p-xylene also were removed in the unamended con-
trol reactor, though to a lesser degree.
In the laboratory, batch bottle microcosms with sediment
and ground water from Seal Beach then were used to
verify the field results and evaluate other conditions
under which in situ biotransformation could be en-
hanced at the Seal Beach field site. Both toluene and
m,p-xylene were removed in the unamended micro-
cosms, just as observed in the field. Corresponding loss
of sulfate in the sample bottle suggested that the ob-
served aromatic removal was probably due to sulfate-
reducing bacteria. Addition of nitrate to the samples
stimulated denitrification, which resulted in faster disap-
pearance qf toluene, complete loss of ethylbenzene, but
less complete removal of m,p-xylene. Addition of a non-
indigenous hydrocarbon-degrading consortium to the
microcosms resulted in some enhancement of o-xylene
transformation. In several nitrate-amended microcosms
in which nitrate was not replenished after it was
completely utilized, sulfate reduction commenced.
Thereafter, both nitrate and sulfate were used in the micro-
cosms. Biotransformatior of benzene in several micro-
cosms was tied to active sulfate reduction. In those
microcosms that had both active denitrification and sul-
fate reduction, complete loss of the full range of BTEX
aromatic compounds occurred. The results of this ex-
periment indicate that utilization of a combination of the
electron acceptors nitrate and sulfate may provide bene-
fits in restoration of BTEX-contaminated sites.
127
-------
Biotransformation of indole and Quinoiine Under Denitrifying Conditions
J.N.P. Black, R.M. Kauffman, D. Denney, and D. Grbi<5-Gali<5
Western Region Hazardous Substances Research Center, Department of Civil Engineering, Stanford University,
Stanford, CA
Perry L. McCarty (presenter) ,
Stanford University, Stanford, CA
The anaerobic conditions prevalent in many contami-
nated ground water plumes provide an impetus to inves-
tigate the anaerobic biodegradation of petroleum
hydrocarbons. In this study, sediment from an oil refinery
waste pond that had been found to be biologically active
against single-ring homocyclic aromatic compounds un-
der denitrifying conditions was used as inoculum in
nitrate-containing enrichment cultures. The cultures
were prepared according to a protocol of rigorous oxy-
gen exclusion; they contained 2 mm of nitrate and 100
mm of an aromatic substrate, individually, in a mineral
medium; and they were incubated within an anaerobic
chamber. The aromatic substrates were indene, naph-
thalene, indole, quinoiine, furan, benzofuran, thiophene,
and benzothiophene. Two live and two autoclaved con-
trols were prepared per substrate.
Within 20 days, cultures exhibited transformation of in-
dole and quinoiine as determined by high pressure liquid
chromatography (HPLC). The other aromatic substrates
have not been transformed after 7 months of incubation.
At 9 days, the concentration of indole had decreased
and an as yet unidentified more polar transformation
product had been produced. By 37 days, both indole and
the transformation product were no longer detectable.
Quinoiine was decreased in its cultures by 20 days and
a transformation product that HPLC retention times in-
dicated was 2-quinolinoi had appeared. By 39 days,
quinoiine was undetectable and 2-quinolinol had accu-
mulated. By 54 days, 2-quinolinol was undetectable. Ion
chromatography revealed that nitrate was consumed in
all live cultures and that nitrate consumption occurred
simultaneously with indole and quinoiine transformation.
Gas partitioner analysis of headspace samples failed to
detect methane or hydrogen sulfide. Examination of the
quinoline-fed culture with phase-contrast microscopy re-
vealed a microbial community of diverse morphology.
Neither nitrate nor any of the aromatic substrates were
consumed in autoclaved controls.
Sediment-free subcultures retained the ability to com-
pletely transform indole and quinoiine. These subcul-
tures were dominated by motile, Gram-negative rods
growing singly and in dense and macroscopically visible
floes. Subcultures exposed to hydrogen consumed ni-
trate without degrading the nitrogen heterocycles. The
addition of 0.3 percent mercury chloride prevented con-
sumption of any substrates.
128
-------
Cometahollsm of TCE by Nitrifying Bacteria
Michael Hyman, Sterling Russell, and Daniel Arp
Western Region Hazardous Substance Research Center, Laboratory for Nitrogen Fixation Research,
Oregon State University, Corvallls, OR
Roger Ely and Ken Williamson
Department of Civil Engineering, Oregon State University, Corvallis, OR
The soil nitrifying bacterium Nitrosomonas europaea is
a lithoautotroph that obtains all of its energy for growth
from the oxidation of ammonia (NH3) to nitrite (NO?-).
Ammonia oxidation is catalyzed by ammonia monooxy-
genase (AMO), which converts NH3 to hydroxyiamine
(NHzOH). The oxidation of NH2OH to NOr provides
both the reductant required for AMO and the sole source
of electrons for adenosine triphosphate (ATP) synthesis.
AMO also can oxidize alternate substrates including
alkanes, alkenes, aromatics, and many Ct to C3 halo*
genated aliphatics, including trichloroethylene (TCE).
The products of these oxidations are not assimilated and
the oxidation process requires simultaneous NH3 oxida-
tion as a source of reductant. These oxidation reactions
are therefore regarded as cometabolic.
TCE cometaboiism by autotrophic nitrifying bacteria has
several attractive features. First, AMO is a constitutive
enzyme and requires no induction. Second, NH3 is
inexpensive and very water soluble. Third, nitrifying bac-
teria are distributed widely in soils. In situ processes
could rely on indigenous nitrifier populations and avoid
the use of genetically engineered and potentially less-
competitive species.
Despite these features, N. europaea, like all other aero-
bic TCE-degrading bacteria, suffers a toxicity from TCE
oxidation resulting from the formation of a reactive TCE-
epoxide intermediate. Much of this TCE toxicity involves
inactivatfon of AMO. This research uses N. europaea as
a model system to study the physiological conse-
quences of TCE toxicity on the cometabolic process.
The objective is to determine the factors that allow for
the maximal sustainable rate of TCE degradation.
Physiological studies are being used to provide operat-
ing parameters for concurrent studies using reactors
capable of sustainable TCE degradation.
In short-term (10 min) experiments correlating TCE deg-
radation (measured as Cr ion release) and the extent of
inactivation of several enzyme activities, N. europaea
has been found to oxidize 60 nmol of TCE/mg protein
before complete inactivation of AMO. This value ranges
from 30 to 100 nmol TCE/mg protein, depending on the
specific activity of batch-grown ceils. TCE toxicity is
irreversible without protein synthesis and affects botfi
AMO and the enzymes and/or proteins associated witti
NH2OH oxidation. Since sustainable TCE oxidation will
require the ability of cells to constantly resynthesize
these inactivated proteins, these data suggest tfiat the
maximal sustainable rate of TCE degradation will occur
at considerably lower concentrations of TCE than those
that support the maximal initial rate of degradation. This
is supported by the observation that the growth of N.
europaea in the presence of TCE can occur only at very
tow TCE concentrations <5 (nmol/mL) where the extent
of enzyme inactivation is sufficiently low that net growth
can occur despite the need to resynthesize inactivated
components. To quantify a maximal tolerable level of
TCE inactivation, the kinetics of recovery of AMO activity
in cells after partial and complete inactivation of AMO by
TCE have been examined. These kinetics have been
compared with cells in which AMO has been specifically
inactivated by light These results indicate that a level of
TCE inactivation of greater than 20 percent adversely
affects the rate of recovery of AMO activity.
Reactor-based studies investigating TCE-degradation
over longer periods (>1 hr) have been initiated and have
confirmed that sustainable TCE degradation can be
maintained but only at low TCE concentrations (<250
ppb). An interesting consideration in these studies is
whether cells respond to partial TCE inactivation by the
de novo syndesis of AMO (and other proteins) and what
contribution newly synthesized enzymes make to the
kinetics of TCE degradation. Initial studies following the
129
-------
kinetics of de novo protein synthesis using 14COz have
been conducted, and these studies are continuing.
A third approach to the question of TCE toxicity involves
studies of TCE-degrading, alkene-oxidizing bacteria.
These species contain epoxide-metabolizing systems
and may be able to circumvent TCE-mediated toxicity
by consuming the epoxide intermediate. Initial studies
indicate that a substantial increase in tolerance to TCE
toxicity compared to N. europaea is not realized in this
system. Some evidence indicates that TCE-epoxide
may in fact inhibit or inactivate the epoxide metabolism
of these bacteria.
130
-------
Assessing the Effect of Environmental Conditions on Chlorophenol Reductive
Dechlorination Pathways and Kinetics
Sandra Woods, Sheryl Stuart and David Nicholson
Western Region Hazardous Substance Research Center, Department of Civii Engineering,
Oregon State University, Corvallis, OR
Teresa Lemmon, James Ingle, and John Westall
Western Region Hazardous Substance Research Center, Department of Chemistry, Oregon State University,
Corvallis, OR
A knowledge of biotransformation pathways and kinetics
is essential to assess risks at contaminated sites, imple-
ment biological treatment processes, or design effective
bioremediation strategies. In addition to the microbial
consortium, environmental conditions such as pH, oxi-
dation/reduction potential, or the presence of toxicants
may alter biodegradation pathways and kinetics from
those observed.in the laboratory.
Pentachlorophenol (PCP) was selected for study be-
cause it is toxic to a wide variety of organisms (1), it is
widely distiibuted in the environment, and the anaerobic
biotransformation pathways for chlorophenols have
been well, studied. PCP has been used extensively as a
wood preservative and pesticide. Hundreds of sites in
the United States are contaminated with PCP as a result
of wood-treating activities. Many of these sites are on
the National Priority List for cleanup under the Super-
fund Program.
The goal of this study is to better understand the effects
of certain environmental conditions on tfie rate of an-
aerobic biotransformations of chlorophenols. Prelimi-
nary studies were conducted to characterize the
pentachlorophenol degradation pathway for an accli-
mated culture. This consortium will be used for all ex-
periments, and the kinetics of these biotransformation
reactions will be measured.
A computer-interfaced reactor system has been devel-
oped to allow measurement of biodegradation rate con-
stants under constant conditions of biomass, pH,
sulfate, sulfide, and acetate concentrations. The reactor
is being used to develop progress curves under varying,
but constant, environmental conditions.
Results
Pentachlorophenol Biotransformation Pathway by the
Acclimated Consortium. Pentachlorophenol biotransfor-
mation pathways were determined for a methanogenic
consortia fed 5,300 mg/L acetate, 3.4 m (0.9 mg/L)
pentachlorophenol, and nutrients for 9 months. Within
1.2 days, 4.38 |iM PCP was biotransformed at 99.7
percent efficiency (Figure 1). Degradation of PCP was
accompanied by the production of all three tetrachloro-
phenols, as well as 3,4,5-, 2,4,5-, and 2,3,5-trichloro-
phenol. The "sum" identified in Figure 1 represents the
sum of the concentrations of PCP and all tetra- and
tri-chlorophenols identified in the reactor. After 0.6 days,
the mass balance appears to fall off due to the produc-
tion of dichlorophenols, which were not included in the
"sum."
6.0
Sum
5.0
2
S.
4.0
PCP
3.0
-o-
2, 3, 5; 6 - TeCP
2.0
-2,3,4,6 - TeCP . 2,3, 4, 5 - TeCP
0
0.2
0.4
0.6
0.8
1.0
1.2
Time (Days)
Figure 1. Progress curve for PCP reductive dechlorination by
a PCP-acclimated methanoganic consortium.
131
-------
2,3,4,5-Tetrachlorophenol (2,3,4,5-TeCP) reached a
maximum concentration of 0.37 after 0.5 days and
was not detected in the reactor after 1 day. 2,3,5,6-TeCP
accumulated rapidly in the reactor to a maximum con-
centration of 3.1 or approximately 70 percent of the
initial PCP concentration. The accumulation of 2,3,5,6-
TeCP ceased after 0.88 days when PCP was removed
from the reactor. 2,3,4,6-TeCP, the meta dechlorination
product, was observed at very low levels (less than 0.07
jiM) during the experiment
The experiment demonstrates that the consortium ac-
quired the ability to dechlorinate PCP at all three chlo-
rine positions after exposure to PCP for 6 months.
Although dechlorination at the ortho position is preferred
by unacciimated consortia (1,2,3,4)", removal of para
and meta chlorines from PCP to produce 2,3,5,6-TeCP
and 2,3,4,6-TeCP is observed following acclimation. The
development of a split degradation pathway has gener-
ally not been observed in previous laboratory experi-
ments, although it has been documented in rice paddy
soils exposed to PCP (5,6). Additionally, each of the
tetrachlorophenols observed in this study also were re-
ported in studies performed examining PCP biodegra-
dation by sludges individually acclimated to either 2-CP,
3-CP, or 4-CP (1).
Similar progress curves were developed for the three
tetrachlorophenols and their resulting metabolic prod-
ucts. The overall pathway appears in Figure 2.
'CI
CI
I#
(o)«.c
Figure 2. Summary of the observed reductive dechlorination
pathway for PCP and Its metabolites by a PCP-
acclimated methanogerric consortium.
Development of a Reactor System to Control Environ-
mental Conditions. A computer-interfaced reactor sys-
tem has been constructed and is now being refined. It
provides control of pH, sulfide concentrations, acetate
concentrations, the apparent oxidation/reduction poten-
tial, and the biomass concentration (there is little growth
compared to the initial biomass in the reactor). Acetate
and sulfate are added with computer-controlled dispens-
ers to change or maintain the concentrations of these
species.
The reactor system provides constant monitoring of the
following parameters using an analog-to-digital conver-
sion board:
• pH with an Orion Ross glass electrode
• Apparent EH at a platinum electrode
• Potential at a silver/sulfide ion selective electrode
The same Orion double junction Ag/AgCI electrode is
used as the reference electrode for all the above indica-
tor electrodes to eliminate problems that can arise when
two or more reference electrodes are present in the
same solution. The sulfide electrode is calibrated by
multiple standard additions of Na^ at the end of data
collection. Since no on-line sensor for acetate or sulfate
exists, samples are periodically withdrawn from the re-
actor to determine their concentrations by ion chroma-
tography.
The Eh, sulfide electrode, and pH measurements for an
experimental run are shown in Figures 3 and 4. The
apparent oxidation/reduction potential mirrors the sul-
fide concentration and remained at a fairly constant level
of -530 mV for a period of nearly 60 hours. During this
period, the reactor's ability to maintain constant acetate
concentrations was evaluated by the programmed addi-
tion of acetic acid. Slug additions of acetic acid or so-
dium acetate were made three times to bring the acetate
concentration to a desired level. The pH changed after
slug additions of acetate acid or sodium acetate. Once
adjusted, the pH was maintained at a fairly constant
value by automated addition of acetic acid to compen-
sate for acetic acid consumption.
-0.10
-0.15
•0.20
t-0.33
^ -0.40
" -0.45
-0.50
-0.55
•0.60
0 1 000 2000 3000 4000 5000 6000
_ . _ _ . Time (Mini
• Apparent Redox
- Sulfide
Figure 3. Apparent Eh and sulfide concentrations.
. JContmuQut Acaac Add AdaesnL
132
-------
a.o
7.8
7.6
7.4
g.g j j ;— -—
IMS sua 0) Acne AM)
0.4 1 ; t1 ¦ ^""11 . .
6.0 i—; i ! : i
0 1QOO 2000 3000 4000 5000 6000
lima (MJn)
Figure 4. pH measurements.
Summary and Conclusions
Based on the results of preliminary experiments, the
complete biotransformation pathway observed for the
reductive dechlorination of PCP by this acclimated con-
sortium is shown in Figure 2. PCP was dechlorinated at
three positions to produce 2,3,4,5-TeCP, 2,3,4,6-TeCP,
and 2,3,5,6-TeCP. 2,3,4,5-TeCP was dechlorinated at
the ortho position to form 3,4,5-trichlorophenol (3,4,5-
TCP), which then gave 3,5-dichlorophenol (3,5-DCP)
and lesser concentrations of 3,4-DCP as persistent
products. 2,3,4,6-TeCP produced both 2,4,6-TCP and
2,4,5-TCP. 2,4,6-TCP was dechlorinated sequentially at
the ortho positions to produce 2,4-DCP and 4-CP; and
2,4,5-TCP was dechlorinated to produce 3,4-DCP and
2,4-DCP. Sequential ortho dechlorination of 2,3,5,6-
TeCP yielded 2,3,5-TCP and 3,5-DCP.
Ortho dechlorination was observed most frequently.
Every possible metabolite due to ortho dechlorination
was observed except for the production of 2,3,4-TCP
from 2,3,4,6-TeCP. Therefore, eight of the nine possible
ortho dechlorination products were observed. In con-
trast, only two para dechlorination'products (of seven
possible products) and four meta dechlorination prod-
ucts (of nine possible products) were observed. Of the
possible polychtorinated phenolic congeners, only five
are not included in the pattiway: 2,3-DCP, 2,5-DCP,
2,6-DCP, 2,3,4-TCP, and 2,3,6-TCP. These five unde-
tected di- and trichlorophenol congeners possess at
least one ortho chlorine, and would not be expected to
be observed because of the consortium's preferential
removal of ortho chlorines.
A reactor system has been developed to allow measure-
ment of reductive dechlorination kinetics while maintain-
ing relatively constant biomass concentration, pH,
sulfide concentration, and apparent oxidation/reduction
potential. Once the control of acetate and sulfate con-
centrations have been refined, reductive dechlorination
kinetics will be measured for varying, but constant, en-
vironmental conditions.
Acknowledgements
Funding for this study was provided by the Office of
Research and Development, U.S. Environmental Pro-
tection Agency, under agreement R-815738-01 through
the Western Region Hazardous Substance Research
Center and by the Presidential Young Investigator Award
Program of the National Science Foundation (ECE 84-
51991). The content of this paper does not necessarily
represent the views of these agencies.
References
1. Mikesell, M.D., and S.A. Boyd. 1986. Complete re-
ductive dechlorination and mineralization of penta-
chlorophenol by anaerobic microorganisms. Appl.
Environ. Microbiol. 52:861-865.
2. Boyd, S.A., and D.R. Shelton. 1984. Anaerobic
biodegradation of chlorophenols in fresh and accli-
mated sludge. Appl. Environ. Microbiol. 47:272-277.
3. Mikesell, M.D., and S.A. Boyd. 1985. Reductive
dechlorination of the pesticides 2,4-D, 2,4,5-T, and
pentachlorophenol in anaerobic sludges. J. Environ.
Qual. 14:337-340.
4. Woods, S.L, J.F. Ferguson, and M.M. Benjamin.
1989. Characterization of chlorophenol and chlo-
romethoxybenzene biodegradation during anaerobic
treatment. Environ. Sci. and Tech. 23:62-68.
5. Ide, A., Y. Niki, F. Sakamoto, I. Watanabe, and H.
Watanabe. 1972. Decomposition of pentachlorophe-
nol in paddy soil. Agric. Biol. Chem. 36:1937-1944.
6. Kuwatsuka, S., and M. Igarashi. 1975. Degradation
of PCP in soils. II. The relationship between the
degradation of PCP and the properties of soils, and
the identification of the degradation products of PCP.
Soil Sci. Plant Nutr. 21:405-414.
j/yjSkiflOlSoaium
AddSKjy rt9ottuwi*c
-------
Spatial Distribution of Nonaqueous Phase Liquid in Sand Cores Using X-ray
Computed Tomography
John L. Holmes and R. Lee Peyton
Great Plains and Rocky Mountain Hazardous Substance Research Center, University of Missouri-Columbia,
Columbia, MO
Tissa H. Illangasekare
University of Colorado at Boulder, Boulder, CO
The small-scale spatial distribution of entrapped
nonaqueous phase liquid in porous media is important
in controlling dissolution kinetics, bioremediation per-
formance, and the effectiveness of pump-and-treat
schemes. This distribution is difficult to measure, how-
ever, which limits experimental understanding of the
effect of entrapment processes on remediation. This
study investigated the use of x-ray computed tomogra-
phy (CT) to measure nondestructive^ the spatial distri-
bution of trichloroettiane (TCA) concentration inside
constructed sand cores.
The 76-mm-diameter and 52-mm-long plexiglass cores
were uniformly packed to a bulk density of 1.57 g/cm3
with sand having a dso of 0.49 mm, producing a mean
porosity of 0.41. The cores were slowly saturated with
water under vacuum pressure. Then a core was placed
in a Siemens Somatom DRH CT unit such that the
longitudinal axis of the core was horizontal and normal
to the CT scan plane. The core was scanned prior to the
introduction of TCA using a sweep of 21 adjacent scans
with a slice thickness of 2 mm and pixel size of 0.32 x
0.32 mm. Then an injection of 1.0 mL of TCA was made
at tiie center of the core using a needle and syringe
placed through a septum in the core wall. Immediately
after injection, an additional sweep of 21 adjacent scans
was made.
Each scan produced a 256 x 256 matrix of x-ray attenu-
ation coefficients (n). A theoretical equation was devel-
oped to relate p. to mass TCA in each pixel. The equation
requires an estimate of the effective energy (E) of pho-
tons passing inside the core. The mean E was computed
for this particular core geometry and elemental compo-
sition of sand using the known photon energy spectrum
emitted by the X-ray tube. The equation was applied to
compute the mass in each pixel, and images of mass
distribution were produced. For these mass calcula-
tions, p. values were averaged over groups of 6 x 6
pixels, producing computed masses for volume ele-
ments of 2 x 2 x 2 mrn. Using a mean E inside the core
of 75 kev, the total computed mass was equal to or
greater than 95 percent of the injected mass. Sub-
sequent sweeps of scans measured the change in mass
in each pixel and movement of the plume with time. This
study demonstrates the use of CT for studying issues
related to small-scale, three-dimensional, nonaqueous
phase liquid movement and entrapment in porous media.
134
-------
Scale-Up Implications of Respirometrically Determined
Microbial Kinetic Parameters
PJ. Sturman, R.R. Sharp, J.B. DeBar, P.S. Stewart, A.B. Cunningham, and J.H. Wolfram
Great Plains and Rocky Mountain Hazardous Substance Research Center, Center for Interfacial Microbial.
Process Engineering, Montana State University, Bozeman, MT
Recent attention to bioremediation of contaminated sur-
face water and aquifer systems has necessitated the
development of methods to estimate in situ rates of
biodegradation. Specifically, biodegradation kinetic pa-
rameters must be determined with sufficient accuracy to
justify their use in predictive models for field degradation
rates. Respirometry has been used for many years to
determine the oxygen demand and stoichiometry of the
degradation of organics in wastewater. More recently,
researchers have adapted the standard biochemical
oxygen demand (BOO) respirometry apparatus to deter-
mine the kinetics of microbial degradation within the
batch BOD vessel. Metfiods used to date have relied on
a thorough knowledge of initial and final conditions
within the respirometer for accurate determinations of
kinetic parameters. The method used in this research
employs a curve-fitting model that uses a single Monod
kinetic expression with cell decay to approximate the
accumulated oxygen demand curve over time. The
curve-fitting routine uses initial substrate concentration
and the actual oxygen demand curve as inputs, and
iteratively calculates a best fit solution for 11^* K,,
biomass yieid, initial biomass concentration, and cell
decay. Mean values and 95 percent confidence intervals
(CIs) were determined for these kinetic parameters. The
kinetic parameters then were used in a bioprocess
model to estimate the extent of biotransformation activ-
ity at a field site. To assess the effects of kinetic parame-
ter variation on the size of the area biotransformed, the
values of two important parameters (n^ and K,) were
varied through their CIs. Results indicate that the area
biotransformed in a 4,800-day model run varies only
slightiy from the least favorable (low high Ks) to
the most favorable (high Umax, low Ks) microbial kinetic
condition.
135
-------
Dissipation of Polycyclic Aromatic Hydrocarbons in the Rhizosphere
M. Katherine Banks
Great Plains and Rocky Mountain Hazardous Substance Research Center, Department of Civil Engineering,
Kansas State University, Manhattan, KS
A. Paul Schwab
Great Plains and Rocky Mountain Hazardous Substance Research Center, Department of Agronomy,
Kansas State University, Manhattan, KS
Vegetation may play an important role in the biodegra-
dation of toxic organic chemicals in soil. The beneficial
effects of vegetation may include uptake, metabolism,
or volatilization of hazardous organics by the plants,
and/or increased biodegradation by the rhizosphere mi-
croflora. The objective of this research was to evaluate
the contribution of plants and associated microflora to
the biodegradation of recalcitrant organic compounds in
soil. The degradation of selected polycyclic aromatic
hydrocarbons (PAH) in the rhizosphere of four plant
species was investigated in a greenhouse experiment.
Soil was contaminated with 100 mg/kg of PAH and
planted'with one of the following: alfalfa (Meticago sa-
tiva), fescue (Festuca arundinacea), big bluestem (An-
dropogon gerardii), or sudan grass {Sorgham vulgare
Sudanese). Plant roots, plant shoots, and soils were
analyzed for PAH content after 4, 8, 16, and 24 weeks.
Contaminants were detected in some cf the plant sam-
ples, and PAH concentrations were smaller in rhi-
zosphere than in unvegetated soil. Target compounds
were not detected in soil leachates. Microbial numbers
were greater in the contaminated rhizosphere of all plant
species when compared to unvegetated soil. These re-
sults indicate that the interaction between plants and
rhizosphere microflora enhances remediation of soils
contaminated with hazardous organics.
136
-------
Effect of Irreversible Sorption on Bioavailability
Amy T. Kan, Gongmin Fu, Mason B. Tomson, and Calvin H. Ward
South/Southwest Hazardous Substance Research Center, Rice University, Houston, TX
Biodegradation of organic pollutants in the environment
depends largely on the bioavailability of the adsorbed
pollutants. Steinberg et al. (1) has shown that ethylene
dibromide (EDB) from a contaminated soil sample is not
biodegradable by indigenous microbes after 38 days,
whereas added (14C) EDB is rapidly mineralized.
Similarly, the sediment bound polycydic aromatic hydro-
carbons (PAHs) from Tamar Estuary, United Kingdom,
retained remarkable compositional uniformity to that of
parent compounds (2). The bound PAH is not available
for leaching, microbial breakdown, or photodegradation.
The present experiment was conducted to determine
the reversibility of naphthalene adsorption/desorption on
soil. Naphthalene readily adsorbed onto the soil and
reached equilibrium within 1 day. The adsorption iso-
therm yielded an organic carbon-based partition coeffi-
cient (Koc) of 793 cmyg, which was comparable to either
the literature or estimated values (3).
Desorption experiments were conducted by replacing
successive fractions of the aqueous phase with clean
water. The supernatant fluid was separated from the soil
by centrifugation at approximately 1,500 g. The hyster-
esis phenomenon was observed for naphthalene ad-
sorption/ desorption (i.e., the adsorbed naphthalene
was not reacfily desorbed from soil). Only about 11 to 23
percent desorbed in the first five successive desorption
steps, whereas 97 percent of the adsorbed naphthalene
should desorb if assuming reversible desorption. The
desorption appears to be controlled by two rate-limiting
mass transfer steps. When first-order kinetic reactions
were assumed to approximate the desorption process,
the reaction half-life was 6 days for the fast desorption
and 451 days for the slow desorption. The rate con-
stants were one to three orders of magnitude slower
than the literature or estimated values (4,5).
The results of this study show that naphthalene or other
structurally similar compounds can be adsorbed to the
soii/sediment and rendered less available for extraction,
chemical, and biological transformation.
References
1. Steinberg, S.M., J.J. Pignatello, and B.L Sawhney.
1987. Persistence of 1,2-dibromoethane in soils: En-
trapment in intraparticle micropores. Environ. Sci.
Techno'!. 21:1201-1208.
2. Readman, J.W., and R.F.C. Mantoura. 1987. A re-
cord of polycydic aromatic hydrocarbon (PAH) pol-
lution obtained from accreting sediments of the
Tamar Estuary, U.K.: Evidence for non-equilibrium
behaviour of PAH. Sci. Total Environ. 66:73-94.
3. Karickhoff, S.M., D.S. Brown, and T.A. Scott. 1979.
Sorption of hydrophobic pollutants on natural sedi-
ments. Water Res. 13: 241-248.
4. Wu, S.-C., and P.M. Gschwend. 1988. Numerical
modeling of sorption kinetics of organic compounds
to soil and sediment partides. 24:1373-1383.
5. Brusseau, M.L., and P.S.C. Rao. 1989. Sorption
nonideality during organic contaminant transport in
porous media. Crit Rev. Environ. Control. 19:33-99.
137
-------
The Effect of Population and Substrate Interactions on SBR Design Optimization
B.C. Baltzis, G.A. Lewandowski, S. Dikshitulu, and K.W. Wang
Northeast Hazardous Substance Research Center, New Jersey Institute of Technology, Newark, NJ
Sequencing batch reactors (SBRs) operate in a cyclic
mode, with each cycle usually consisting of five distinct
periods: fill, react, settle, draw, and idle. Such reactors
are known to offer a number of advantages over con-
ventional activated sludge systems, including greater
flexibility in meeting changes in feed conditions, effec-
tive control of the process and the quality of the dis-
charge, and higher volumetric efficiency. Also, a
separate clarifier is not required. The optimal design of
SBRs for treatment of hazardous and toxic chemicals
requires development of process models for making
predictive and scale-up calculations.
In this project, 5-L SBRs are being used to investigate
two categories of issues: the first deals with the effect of
population interactions (i.e., competition) on the per-
formance of the unit,- the second involves identifying the
interactions among pollutants at the kinetic level and
tt>en determining the effects on SBR design.
Phenol was used as the model compound in studying
the effect of microbial competition. Two species were
employed: Pseudomonas putida (ATCC 17514) and
Pseudomonas resinovorans (ATCC 14235). Substrate
concentrations were monitored using high pressure liq-
uid chromatography (HPLC) measurements. Total
blomass concentration measurements were based on
optical density, while biomass concentrations of individ-
ual species were monitored through colony counts on a
nutrient agar and a citrate-containing agar on which only
P. resinovorans could grow. A relatively slow fill phase
with aeration was found to permit biodegradation to start
and to lead to better results. Depending on the phenol
concentration in the untreated waste, the residence
time, the fraction of the cycle time allocated to the fill
phase, and the ratio of minimum to maximum volume of
the reactor contents in a cycle, the biomass composition
was found to change substantially with time. Unless the
conditions are properly selected, a mixed culture cannot
be maintained in the long run. A detailed model has been
derived and used in extensive numerical studies. These
studies determined the conditions leading to stable op-
eration of the unit. The model predicts and the experi-
ments verify that under certain conditions of operation
multiple types of reactor behavior are possible, as de-
termined by conditions at start-up of the unit.
Phenol and 4-chiorophenol (4CP) are being used as
model compounds for studying the effects of substrate
interactions. Two cultures have been tested. One was
not capabie of degrading 4CP unless phenol was pre-
sent; this cometabolic process led to complete minerali-
zation of 4CP. The second culture was capable of
completely mineralizing botti phenol and 4CP, but each
compound affects the degradation of the other. These
cross-inhibitory kinetics have been mathematically de-
scribed, and the SBR model predicts the best values for
the operating parameters. Experiments are under way
to verify the model's predictions.
138
-------
TECHNICAL REPORT DATA
(Pleat read instructiont on the reverie before completing)
_
1. report no.
EPA/600/R-94/160
2.
3. RECIPII
4. TITLE AND SUBTITLE
BIOREMEDIATION OF HAZARDOUS WASTES - RESEARCH,
DEVELOPMENT AND FIELD EVALUATIONS - 1993
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Fran Kremer
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency, National Risk
Management Research Laboratory, Cincinnati, OH 45268
10. PROGRAM ELEMENT NO.
11. CQNTHACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
13. TYPE OF REPORT AND PERIOO COVERED
14. SPONSORING AGENCY CODE
IS. SUPPLEMENTARY NOTES
16. ABSTRACT
The proceedings of the 1993 Symposium on Bioremediation of Hazardous Wastes,
hosted by the Office of Research and Development (ORD) of the EPA in Dallas, Texas .
The symposium was the sixth annual meeting for the presentation of research conductec
by EPA's Biosystems Technology Development Program (BTDP) and by affiliated Hazardou:
Substance Research Centers (HSRCs). The document contains abstracts of recent
research projects, ranging in scope from molecular biology in the laboratory to
cleanup evaluations in the field. 32 papers and numerous posters presented at the
symposium are organized into seven program areas: Bioremediation Field Initiative,
Performance Evaluation, Field Research, Pilot-Scale Research, Process Research,
Development of Computer-Based Assessment Systems, and Hazardous Substance Research
Centers. The proceedings also contain a brief synopsis of introductory remarks mad<
by opening speakers.
17 KEY WORDS AND OOCUMENT ANALYSIS
a. DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI FieldfGroup
bioremediation, biological
treatment, hazardous wastes
ORD, R&D
18, DISTRIBUTION STATEMENT
P.ELEASF TO PIIRI TC —
19, SECURITY CLASS /This Report)
IIMfl ACCTCTCn
21. NO. OF PAGES
20. SECU R7Tirirjraxs;
UNCLASSIFIED
22. PRICE
EPA Form J220-1 (R»y. 4-77) p««viou» toiriONiioeiOLiTE
------- |