EPA/600/R-93/054
vvEPA
l Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA/600/R-93/054
May 1993
Symposium on
Bioremediation of
Hazardous Wastes:
Research,
Development, and
Field Evaluations
Abstracts
The Fairmont Hotel
Dallas, TX
May 4-6, 1993
. PROJECTION
1445 ROSS AVENUE
DAIUS, I£XA3 7570?
1 1 ^
» ^
-------
EPA/600/R-93/054
May 1993
Symposium on Bioremediation of Hazardous Wastes:
Research, Development, and Field Evaluations
Abstracts
The Fairmont Hotel
Dallas, TX
May 4-6, 1993
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Printed on Recycled Paper
-------
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
commercial products does not constitute endorsement or recommendation for use.
-------
CONTENTS
Bio remediation Field Initiative 1
Retrospective Performance Evaluation on In Situ Bioremediation:
Site Characterization 3
John T. Wilson and Don H. Kampbell
Retrospective Performance Evaluation on In Situ Bioremediation:
Modeling and Risk Assessment 10
Tissa H. Illangasekare, David C. Szlag, and John T. Wilson
Evaluation of Full-Scale In Situ and Ex Situ Bioremediation of
Creosote Wastes in Soils and Ground Water 14
Ronald C. Sims, John E. Matthews, Scott G. Hiding,
Bert E. Bledsoe, Mary E. Randolph, and Daniel Pope
An Evaluation of Concurrent Bioventing of Jet Fuel and Several
Soil Warming Methods:
A Field Study at Eielson Air Force Base, Alaska 20
Gregory D. Sayles, Robert E. Hinchee, Catherine M. Vogel,
Richard C. Brenner, and Ross N. Miller
Documenting Bioventing of Jet Fuel to Great Depths:
A Field Study at Hill Air Force Base, Utah 27
Gregory D. Sayles, Robert E. Hinchee,
Richard C. Brenner, and Robert Elliott
Application of Wood-Degrading Fungi to Treat Contaminated Soils 34
John A. Closer, Richard T. Lamar, Mark W. Davis, and Diane M. Dietrich
Performance Evaluation 39
Field Evaluation of Phenol for Cometabolism of Chlorinated Solvents 41
Gary D. Hopkins, Lewis Semprini, and Perry L. McCarty
Innovative Bioremediation Strategies for Creosote:
Characterization and Use of Inocula 47
James G. Mueller, Suzanne E. Lantz, Jian-Er Lin, and P. Hap Pritchard
Decontamination of PCB-Contaminated Sediments Through the Use of
Bioremediation Technologies 51
John F. Quensen, III, Stephen A. Boyd, James M. Tiedje, and John E. Rogers
in
-------
CONTENTS (cont.)
Natural Anaerobic Bioremediation of TCE at the St. Joseph, Michigan,
Superfund Site 57
Peter K. Kitanidis, Lewis Semprini, Don H. Kampbell, and John T. Wilson
Comparison of Bioventing and Air Sparging for In Situ Bioremediation of Fuels 61
Don H. Kampbell, Christopher J. Griffin, and Frank A. Blaha
Determining the Genotoxicity of Bioremediation Products 68
Larry D. Claxton, S. Elizabeth George, Robert W. Chadwick,
Virginia S. Houk, L.E. Rudd, J.J. Perry, and John T. Wilson
Field Research 71
Field Demonstration of a Constitutive TCE-Degrading Bacterium for
the Bioremediation of TCE 73
Malcolm S. Shields, Michael Reagin, Robert Gerger, Rhonda Schaubhut,
Robert Campbell, Charles Somerville, and P. Hap Pritchard
Bioremediation of TCE: Monitoring the Fate and Effects of a Microorganism Used
in a Field Bioaugmentation Study 80
M.S. Shields, R. Snyder, M. Reagin, R. Gerger, R. Campbell,
C. Somerville, and P.H. Pritchard
Factors Determining the Effectiveness of Microbial Inoculation in Soils and Sediments:
Effectiveness of Encapsulation 86
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 90
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 100
Morton A. Barlaz, Michael B. Shafer, Robert C. Borden, and John T. Wilson
Pilot-Scale Research 107
Treatment of PCP-Contaminated Soils by Washing with Ethanol/Water Followed by
Anaerobic Treatment 109
Amid P. Khodadoust, Julie A. Wagner, Makram T. Siddan, and Steven I. Safferman
IV
-------
CONTENTS (cont.)
Page
Preliminary Evaluation of Attachment Media for Gas Phase Biofilters 115
Francis L. Smith, George A. Sorial, Paul J. Smith, Makram T. Suidan,
Pratim Biswas, and Richard C. Brenner
Process Research 121
Methanogenic Degradation Kinetics of Nitrogen and Sulfur Containing Heterocyclic
Aromatic Compounds in Aquifer-Derived Microcosms 123
E. Michael Godsy, Donald F. Goerlitz, and Dunja Grbic-Galic
Anaerobic Degradation of Halogenated and Nonhalogenated Phenolic Compounds .... 129
MM Hdggblom, M.D. Rivera, and L.Y. Young, and I.E. Rogers
Anaerobic Biodegradation of 5-Chlorovanillate as a Model Substrate for the
Bioremediation of Paper-Milling Waste 131
Barbara R. Sharak Genthner
Bioremediation of Soils and Sediments Contaminated with Aromatic Amines 137
Eric J. Weber, David L. Spidle, and Kevin A. Thorn
Anaerobic Biotransformation of Munitions Wastes 139
Deborah J. Roberts, Stephen Funk, Don L. Crawford, and Ronald L. Crawford
Cometabolic Biodegradation of 2,4-Dinitrotoluene Using Ethanol as a
Primary Substrate 145
Jiayang Cheng, Makram T. Suidan, and Albert D. Venosa
Effects of Metals on Anaerobic Treatment Processes 149
W. Jack Jones and In Chul Kong
Fate of Highly Chlorinated Dibenzo-/>-Dioxins and Dibenzofurans
in Anaerobic Soils and Sediments 155
Peter Adnaens, Dunja Grbic-Galic, and Gregoiy D. Sayles
Bioreactor Treatment of Nitrate Contamination in Ground Water:
Studies on the Sulfur-Mediated Biological Denitrification Process 161
Michael S. Davidson and Thomas Cormack
Chemical Interactions and pH Profiles in Microbial Biofilms 166
Joseph R. V. Flora, Makram T. Suidan, Pratim Biswas, and Gregory D. Sayles
-------
CONTENTS (cent.)
Characterization of Biofilter Microbial Populations 170
Alec Breen, Alan Rope, John C. Loper, and P.R. Sferra
Fundamental Studies in the Development of the Gas Phase Biofilter 174
Rakesh Govind, Vivek Utgikar, Yonggui Shan, Wang Zhao,
Madan Parvatiyar, Stephan Junginzer, and Dolloff F. Bishop
Sequential Anaerobic/Aerobic Treatment of Contaminated Soils and Sediments 184
Grace M. Lopez, Gregory D. Sayles, Karen Buhler, Dolloff F. Bishop, In S. Kim,
Guanrong You, Petra Klostermann, Margaret J. Kupferle, and Douglas S. Lipton
Poster Session 189
Approaches to the Development of Comparative Genotoxicity Risk Assessment
Methods for Evaluating Hazardous Waste Control Technologies 191
Larry D. Claxton
Modulation Effects of Biotechnology Microbial Agents Following Pulmonary
Exposure of Mice 193
5. Elizabeth George, Michael J. Kohan, John P. Creason, and Larry D. Claxton
Field Treatment of BTEX in Vadose Soils Using Hydrofracturing, Vacuum Extraction,
and Biofilters 194
Dolloff F. Bishop, Wendy Davis-Hoover, and Rakesh Govind
PCB Biodegradation during Aerobic Treatment of Sludge from
the French Limited NPL Site 197
/. W. Anderson, T. Smith, and J. T. Wilson
Development and Evaluation of Composting Techniques for Treatment of Soils
Contaminated with Hazardous Waste 204
John A. Glaset; Carl L. Potter, Edward D. Kennedy, Jeffrey J. McCoimack,
Joseph B. Fan-ell, and Michael Najar
Engineering Optimization of Slurry Bioreactors for Treating Hazardous Waste in
Soil and Sediments 206
John Closer, Paul McCauley, Majid Dosani, Jennifer Platt, Edward Opatken,
and Diana Roush
Biotreatability of a Vadose Zone Soil Contaminated with Dioctyl Phthalate 207
Don H. Kampbell, Dennis D. Fine, and Jerry W. Anderson
VI
-------
CONTENTS (cont.)
Page
Innovative Bioremediation Strategies for Creosote: Geographic Diversity of PAH
Degradation Capabilities at Wood-Treating Sites 208
James G. Mueller, Suzanne E. Lantz, Richard Devereaux, Deborah L. Santavy, and
P. Hap Pritchard
Design of an Expert System to Select an Appropriate Bioremediation Technique 214
Raymond C. Loehr and Greg E. Schmidt
A Data Visualization System for Bioremediation Analysis 217
Lewis A. Rossman, Kevin Savage, and John Franco
Poster Presentations Supported by
EPA's Hazardous Substance Research Center Program 219
Development of a Knowledge-Based Bioremediation Adviser 221
Shu-Chi Chang, Peter Adriaens, Iris D. Tommelein, and Timothy M. Vogel
Use of Composting Technologies to Treat a TNT-Contaminated Soil 224
James H. Johnson, Jr., Mohammed Moshin, Lily Wan, and Abdul Shafagatti
Effect of Pore-Scale Hydrodynamics on Bulk Reaction Rates 225
Bruce B. Dykaar and Peter K Kitanidis
In Situ Bioremediation Using a Recirculation Well 227
MM Lang, L. Semprini, and P. V. Roberts
Chlorinated Aliphatic Hydrocarbon Biodegradation by Methanotrophic Bacteria 229
Laurence H. Smith, Tomas Henrysson, and Perry L. McCarty
Anaerobic Biodegradation of BTEX Compounds at Seal Beach, California 231
Harold A. Ball, Martin Reinhard, and Eva O. Orwin
Biotransformation of Indole and Quinole under Denitrifying Conditions 232
J.N.P. Black, KM. Kauffman, D. Denney, and D. Grbic-Galic
Cometabolism of TCE by Nitrifying Bacteria 233
Michael Hyman, Roger Ely, Sterling Russell, Ken Williamson, and Daniel Arp
Assessing the Effect of Environmental Conditions on Chlorophenol Reductive
Dechlorination Pathways and Kinetics 235
Sandra Woods, Sheryl Stuart, David Nicholson, Teresa Lemmon, James Ingle,
and John Westall
vn
-------
CONTENTS (cont.)
Spatial Distribution of Nonaqueous Phase Liquid in Sand Cores Using X-ray
Computed Tomography 243
John L. Holmes, R. Lee Peyton, and Tissa H. Illangasekare
Scale-Up Implications of Respirometrically Determined Microbial Kinetic Parameters . . 245
P.J. Sturman, R.R. Sharp, J.B. DeBar, P.S. Stewart, A.B. Cunningham, and J.H. Wolfram
Dissipation of Polycyclic Aromatic Hydrocarbons in the Rhizosphere 246
M. Katherine Banks and A. Paul Schwab
Effect of Irreversible Sorption on Bioavailability 247
Amy T. Kan, Gongmin Fu, Mason B. Tomson, and Calvin H. Ward
The Effect of Population and Substrate Interactions on SBR Design Optimization 249
B. C. Baltzis, G.A. Lewandowski, S. Dikshitulu, and K. W. Wang
Vlll
-------
BIOREMEDIATION FIELD INITIATIVE
-------
RETROSPECTIVE PERFORMANCE EVALUATION ON IN SITU
BIOREMEDIATION: SITE CHARACTERIZATION
John T. Wilson and Don H. Kampbell
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
Ada, OK
INTRODUCTION
Bioremediation is difficult to assess in heterogeneous geological material. Often, oily phase
material is associated with fine textured material with low hydraulic conductivity. Remedial fluids
tend to pass around the fine textured material. Because the flux of nutrients and electron acceptor
through the fine textured material is small, there is little opportunity 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 mineral nutrients, and bioventing. Ground water flow under ambient conditions was
to the north or northeast. The flow of water during the remediation paralleled the natural 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 recharged 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.
-3-
-------
Hydrocarbon Release
RW-1
m
MW-31
61-B 61.A 61-J
61-C* • • A •
. V «MW-2A
EXISTING
SERVICE
BUILDING
N
NUTRIENT
RECHARGE
GALLERY
GROUNDWATER
RECHARGE
GALLERY
20 meters
Figure 1. Infrastructure at an in 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 recovery well (RW-1) and filtered
through activated carbon. The flow was split; part was amended with hydrogen
peroxide and mineral nutrients, and recharged in a nutrient recharge gallery. The
remainder was recharged in a ground water recharge gallery. The system was
designed to sweep hydrogen peroxide and nutrients under the service building. MW-
1, -2A, and -3 are monitoring wells. 61A through 61J are boreholes for cores.
-4-
-------
GROUND WATER REMEDIATION
Table 1 compares the reduction in concentration of benzene 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. Well MW-3 is a significant distance from the region with oily phase hydrocarbons;
it sampled the plume of contaminated ground water that moved away from the spill.
Prior to remediation, concentrations in wells MW-1 and MW-8 were equivalent. Well MW-1
was closest to the nutrient recharge gallery, and the aquifer surrounding MW-1 was completely
remediated; BTEX compounds were undetectable in ground water. In well MW-8, immediately
adjacent to the point of release, the concentration of benzene was reduced at least one order of
magnitude, and the concentration of benzene and BTEX compounds 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 contamination could
return once active remediation ceased. The site was cored along a transect downgradient of the
release. The transect extended laterally from clean material, 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 spill to clean material below. The cores were
extracted and analyzed for total petroleum hydrocarbons (TPH) and for the concentration of
individual BTEX compounds.
The relationship among the land surface, the water table, the region containing hydrocarbons,
and the bedrock are presented in Figure 2. Significant amounts of hydrocarbons remain within a
narrow interval approximately 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.
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 surrounds 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 oily phase hydrocarbon can be predicted by Raoult's law. The solution concentration in water
should be proportional to the mole fraction of the hydrocarbon in the oily phase.
-5-
-------
Table 1. Reduction in Concentration of Hydrocarbon Contaminants in Ground Water
Achieved by In Situ Bioremediation
Well
MW-1
MW-8
MW-2A
MW-3
RW-1
Benzene
Before
During
After
Total BTEX
Before
During
After
--Otg/liter)-
220
180
?
11
<1
<1
130
11
5
2
<1
16
0.8
2
<1
2,030
1,800
9
1,200
<1
164
331
1,200
820
2
<6
34
13
46
<1
-6-
-------
IX uoijeja8§CT3 [eojjjaA
Q -
f
a
o
o
C*
CQ
f,
•n
"§ £r
§ j*a
J3 w> §>
•£> e fT
fl'a*
u « .S
a g-o
§ o «
!> i- o
« i's.
81-8
111
•O T3 J3
ecu
- *o
w c -°
5.2 a
°l 8
_ ^ _
O. u -P.
s el1
sag
.2.2*
«-° -
811
IJl
ill
"^ CO ««
•S 1 a
- g o
c°c &
§o
-7-
-------
Table 2. Vertical Extent of Total BTEX Compounds and Total Petroleum Hydrocarbons at
Borehole D, the Most Contaminated Borehole in the Transect (Figure 1)
Elevation
(feet AMSL)
281.14 to
5280.31
5280.31 to
5279.97
5279.97 to
5279.56
5279.56 to
5279.14
5279.14 to
5278.97
5278.97 to
5278.64
5278.64 to
5278.22
5278.22 to
5277.14
TPH
BTEX
Benzene
--(mg/kg)-
<44
227
860
1176
294
273
<34
<24
<1
5.1
101
206
27
7.4
<1
<1
<0.2
<0.2
<0.2
4.3
0.68
0.26
<0.2
<0.2
Color and
Texture
Brown sand
Brown sand
Black sand
Black sand
Black sand
Black sand
Black sand
Brown to
yellow sand
-8-
-------
Table 3. Lateral Distribution of Total BTEX Compounds and Total Petroleum Hydrocarbons
along the Transect (Figure 1) at the Most Contaminated Depth Interval
Borehole
B
C
D
E
TPH
BTEX
Benzene
--(mg/kg)--
167
156
1,176
156
0.8
3.5
260
3.5
<0.2
<0.2
4.3
0.06
-9-
-------
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
and
John T. Wilson
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
Ada, OK
INTRODUCTION
Conventional methods for determining the extent of cleanup at a bioremediation site can
often be misleading. 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 is a conservative assumption and generally leads to
target levels of total petroleum hydrocarbons (TPH) concentrations on the order of 10 to 100 mg/kg
aquifer material. Many bioremediation schemes, however, may preferentially degrade 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 methodology to evaluate the possible risk, if any,
associated with benzene, toluene, ethylbenze, 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 questions: 1) Will BTEX reappear in ground water? 2) How long will it
take the plume to reappear? 3) What concentration level may be expected? The results of the case
study will also assist in providing a technical basis for monitoring schedules, locating compliance
wells, and constructing rational criteria for site closure.
-10-
-------
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 holding tank. These
hydrocarbons exist as light nonaqueous 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
transverse to the original NAPL plume in order to determine the remaining concentrations of TPH
and BTEX. Test 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 representative over the aquifer cross
section but may not be representative of the aquifer material containing entrapped LNAPL.
Laboratory parameter tests conducted on 28 soil core samples showed hydraulic conductivity varies
two orders of magnitude across the site 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 during coring reveal the presence
of layering in the aquifer.
MODEL SELECTION
The following observations at the site and in the laboratory 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 inherently
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; 4) Solute plumes are not vertically homogeneous and biological
activity will not be uniformly distributed vertically.
Bypassing of the LNAPL plume by nutrients and electron acceptor, attributable to a lowered
hydraulic conductivity of the central part, resulted in high TPH and BTEX levels in some cores.
Coupled with the 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 simulated with a three-dimensional random walk called RAND3D.
MODELING APPROACH
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
-11-
-------
deviation of the mean residual. 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 contaminant
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 assumptions are used in the solute transport modeling: 1) the concentration of BTEX
in the source zone remains constant, 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 was then
used to calculate the equilibrium concentration and, consequently, the mass flux. Estimated benzene
mass fluxes were converted 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 conservative, 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
develop within the source zone in areas with low BTEX and TPH concentrations, allowing water
flow to bypass the more highly contaminated areas. Laboratory determination of the hydraulic
conductivity in the samples containing 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 homogeneous. In general, the solute plume will consist of subplumes 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 equilibrium at 420 days.
The design of the compliance wells will have tremendous 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 existing monitoring wells at the site are screened over the top
5 ft of the aquifer. 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 dilution 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
-12-
-------
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 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. 5) Compliance well design should be carefully considered when
sampling a three-dimensional plume, as the well design can lead to significant 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
preferentially 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 channels created by naturally
occurring aquifer soil heterogeneities 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 adequate resolution to capture these
spill-site-scale heterogeneities. Even though the hydraulic conductivity values determined in the
laboratory on disturbed soil samples were used in this study, a more appropriate characterization
method would be well-designed bail tests (or slug tests), which capture the local layered
heterogeneities more accurately. These local hydraulic conductivity values allow us to obtain the
velocity field in the contaminant zone and to subsequently determine the contaminant mass flux.
Solute breakthrough curves determined by this method can then 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. Retrospective performance evaluation on in situ
bioremediation: site characterization. Proceedings of U.S. EPA Symposium on
Bioremediation of Hazardous Wastes: Research, Development, and Field Evaluations.
Dallas, TX, May 4-6, 1993.
2. Illangasekare, T.H., D.C. Szlag, J. Campbell, J. Ramsey, M. Al-Sherida, and D.D. Reible.
1991. Effect of heterogeneities and preferential flow on distribution and recovery of oily
wastes in aquifers. Proceeding of Conference on Hazardous Waste Research. Kansas 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
would also like to thank Lisa Weers, of the Colorado Department of Health, for her assistance.
-13-
-------
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
and
John E. Matthews, Scott G. Ruling, Bert E. Bledsoe, and Mary E. Randolph
U.S. Environmental Protection Agency
Ada, OK
and
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
managers that are overseeing or considering the use of bioremediation; and 3) develop a treatability
data base that will be available through EPA's Alternative Treatment Technology Information
Center (ATTIC). The performance evaluation project in Libby, Montana, mainly focuses on the first
objective.
The Champion International Superfund Site in Libby, Montana, was nominated by the
Robert S. Kerr Environmental Research Laboratory as a candidate site for performance evaluation.
Two forms of wood preservative were used at the site: creosote, containing polycyclic aromatic
hydrocarbons (PAHs) and pentachlorophenol (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 remediation activities; 2) identifying critical site characterization and treatment
parameters necessary to evaluate bioremediation performance for each of the bioremediation
treatment units; and 3) evaluating bioremediation performance based on the information obtained.
-14-
-------
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 locations of the three treatment processes at the site.
BIOLOGICAL TREATMENT PROCESSES
The upper aquifer above-ground treatment unit is provided 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
reoxygenation of the effluent prior to reinjection. The system was commissioned in February 1990.
An evaluation of the system components for equalization and for biotreatment is being
conducted. Equalization 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
nutrient amendment, influent pumping, bioreactor vessels, aeration, heating, and effluent pumping.
Figure 2 illustrates the components of the above-ground treatment system for extracted ground
water.
The LTUs have been used for bioremediation of contaminated soil from three primary
sources, including tank farm, butt dip, and waste pit areas. Contaminated 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 naphthalene, 3) 8 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. Components 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-permeability 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 infiltration, erosion, and direct contact.
Contaminated soil is applied and treated in lifts (approximately 9 in. thick) in the designated
LTU. Degradation rates, volume of soil to be treated, initial contaminant concentration, degradation
period, and LTU size determine the time required for remediation of a given lift. Based on an
estimated 45-day timeframe for remediation 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 concentration of approximately
100 mg/L of hydrogen peroxide. Injection flow rate is approximately 100 gpm into three injection
-15-
-------
Monitoring Wells
D Injection Wefts
Figure 1. Relative locations of the three treatment processes at the site, including the
bioreactor building for treatment of extracted ground water, LTUs, and monitoring
and injection wells for in situ treatment.
-16-
-------
Ground Water
from Extraction Wells
Steam:
Equalization Tank
Heat Exchanger
Heat
Exchanger
I Fluid
Nutrient
Supply Tank
To Rock Filter ana i
Infiltration System y
Figure 2. Components of the above-ground treatment system for bioremediation of extracted
ground water.
-17-
-------
clusters. Inorganic nutrients in the form of potassium 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 fig/L for benzene,
50 fug/L for arsenic, and a human health threat no greater than 10'5 for ground water concentrations
of other organic and/or inorganic compounds.
PERFORMANCE EVALUATION ACTIVITIES
Performance evaluation of the upper aquifer above-ground treatment system involves
evaluation of the bioreactor system. Treatment evaluation is focused on characterizing performance
with regard to system capability to remove PAHs and PCP from the ground water, optimizing
operation within the bioreactors, and investigating 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 reactor and from the operation of the pilot-scale reactor will be
combined with data provided by Champion International to provide an in-depth evaluation of
performance.
Performance of the soil bioremediation system in the LTUs involves evaluation of the
reduction in concentration of PAHs and PCP with time and with depth within the LTUs. The
primary purpose of the LTU soil 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 conditions 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 following equation:
V, = Vs/k + Va/k * n
where:
k = the number of samples
n = the number of analyses per sample
k*n = the total number of analyses
V, = the total variance
Va = the analytical variance
Vs = the sample variance
-18-
-------
In general, sampling efforts to minimize V, will result in the best precision. Analytical
procedures frequently achieve precision levels (Va/k*n) of 1 to 10 percent, while soil sampling
variation (Vs) may be greater than 35 percent. Sampling designs that will reduce the magnitude of
Vs should be employed where possible. Therefore, the sampling procedures used in this evaluation
were designed to minimize Vs and provide representative information about the transformation of
PAHs and PCP within the LTUs.
The LTUs were sampled in May, June, July, and September of 1991, and September of 1992.
Field-scale investigations concerning PAH and PCP concentrations were supported by laboratory
mass-balance investigations 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 associated 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 demand for injected hydrogen peroxide, and the
concentrations 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
potential 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 International to provide an integrated evaluation of
bioremediation 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.
-19-
-------
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,1 Robert E. Hinchee,2 Catherine M. Vogel,3 Richard C. Brenner,1
and Ross N. Miller4
*U.S. Environmental Protection Agency, Risk Reduction Engineering Laboratory,
Cincinnati, OH
2Battelle Laboratories, Columbus Division, Columbus, OH
3U.S. Air Force Armstrong Laboratories, Tyndall Air Force Base, FL
"U.S. Air Force Center for Environmental Excellence, Brooks Air Force Base, TX
INTRODUCTION
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 technologies, 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 relatively 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 Field Initiative, a cooperative program between EPA's Office
of Research and Development and Office of Solid Waste and Emergency Response, EPA's Risk
Reduction Engineering Laboratory 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 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.
METHODOLOGY
Site history, characterization, installation, and monitoring were summarized previously (2).
Figure 1 shows a plan view of the project. Briefly, four test plots have been established, all receiving
-20-
-------
£
o •» o
o
o
at
O
cw
0>
O
J*
as
it
;|j
:«"
«
•>-> K!
o iS
«s c
c§
So
g ^
o «
t>
OH
H
C/j
C >2
5 S
PH C
fc
-21-
-------
relatively uniform injection of air. The four test plots are being used to evaluate three soil warming
methods:
• Passive Warming. Enhanced solar warming in late spring, summer, and early fall
using clear plastic covering 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.
Periodically, in situ respirometry tests (3) are conducted to measure the in situ oxygen uptake
rates by the microorganisms. These tests allow estimation of the biodegradation rate as a function
of time and, therefore, as a function of ambient temperature and soil warming technique. Quarterly
comprehensive and monthly abbreviated in situ respiration tests are planned.
Final soil hydrocarbon analyses will be conducted in late 1993 and compared with the initial
soil analysis to document actual hydrocarbon loss due to bioventing.
RESULTS
Evaluation of Soil Warming Methods. Table 1 shows soil temperatures in each plot averaged
over the plot and over a 3-month period (a season). Each warming method calls for maintaining
temperatures higher than the unheated control, despite, for example, mean minimum ambient air
temperatures in January of about -20°C. The active and heat tape warming methods involve
maintaining summerlike temperatures during the winter.
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 temperature. The
trend of higher rates at higher temperatures is clear. To clarify this relationship, an Arrhenius plot
was constructed (Figure 3) plotting Iog10(rate) vs. inverse absolute temperature. If an Arrhenius
relationship exists, the data should be linear in this plot. The resulting linear fit gives:
Iog10[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
140-9280/T
-22-
-------
Table 1. Temperatures in Each Plot as a Function of Season
Temperature (°C)
Season
Autumn 1991
Winter 1992
Spring 1992
Summer 1992
Autumn 1992
Winter 1993
Active
10.8
10.1
17.4
18.1
17.0
15.9
Passive
5.7
0.5
0.7
15.6
7.1
1.3
Heat Tape
NA
NA
NA
NA
8.0
10.0
Control
5.7
-0.2
-0.5
9.6
3.8
0.2
Note: Temperatures are averaged over the plot and over the length of the season.
NA = Not available, prior to installation.
-23-
-------
O
(N
m
O
cn
*•* £
(4-1 O
o-a
c ^
o W)
•§s
c >•
<3 rt
eg S
"S 2
UM Qy
60 JS
q,) -4—»
"2 >-
O )/6Lu)
o
"5.
-24-
-------
0
CO
O
in
m
o
+
LJ
Q_
E
0)
CO
CO
a
l§
o
M JS
CO *-
t«
c —
o
-------
This function is in agreement with previous studies (4).
Soil Sampling. Soil samples for total petroleum hydrocarbon 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.
REFERENCES
1. Hoeppel, R.E., R.E. Hinchee, and M.F. Arthur. 1991. Bioventing soils contaminated with
petroleum hydrocarbons. J. Indust. Microbiol. 8:141.
2. Sayles, G.D., R.C. Brenner, R.E. Hinchee, C.M. Vogel, and R.N. Miller. 1992. Optimizing
bioventing in 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.
-26-
-------
DOCUMENTING BIOVENTING OF JET FUEL TO GREAT DEPTHS:
A FIELD STUDY AT HILL AIR FORCE BASE, UTAH
Gregory D. Sayles,1 Robert E. Hinchee,2 Richard C. Brenner,1 and Robert Elliott3
Environmental Protection Agency, Risk Reduction Engineering Laboratory, Cincinnati, OH
2Battelle Laboratories, Columbus Division, Columbus, OH
3Hill Air Force Base, UT
INTRODUCTION
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 technologies, 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 utilities cannot be disturbed, because bioventing equipment (air injection/withdrawal wells, air
blower, and soil gas monitoring wells) is relatively non-invasive.
Through the Bioremediation Field Initiative, a cooperative program between EPA's Office
of Research and Development and Office of Solid Waste and Emergency Response, EPA's Risk
Reduction Engineering Laboratory 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 unsaturated 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 understanding 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.
-27-
-------
METHODOLOGY AND RESULTS1
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 disturbing these
structures. A plan view of the installation is shown in Figure 1. The single air injection well,
continuously 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 ft3/min. Since October 1993, the injection rate has been about 45 ft3/min.
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 biodegradation 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 ft3/min. The oxygen levels at great distance are strongly influenced 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
conducted in September 1991 and 1992. The air flow rate was about 70 ft3/min for both
measurements. The lower 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 composition 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.
JSee previous report (2) for additional details.
-28-
-------
IW XASMP5
(0.2)
CW7
• SMP = Surface Monitoring Point
O CW = Soil Vapor Cluster Well
(1.5) = TPH in Ground Water (mg/L) (9/91)
A-A1 = Cross Section Trace
50 feet
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 gas monitoring wells; SMPs are
shallow soil gas monitoring points.
-29-
-------
o
m
CN
o
o
CM
o
in
O
O
. o
o
CM
in
in
(%)
c
o
•4-J
o
CD
E
o
0)
o
c
o
-M
£0
• ^—
Q
ft
.2 2
n> efl
£.1
£ S'S
"e ^
£ U i_
143 I 3
8 S'fi
2o 2
.,
0 .§* |»
a ^ °
a-a^
:3 o •"
O 5 <4_
co £ o
cs
30
|jos
-30-
-------
Table 1. Rates of Biodegradation, Averaged over the Depth Noted, at Three Wells
Rate (mg/kg/day)
Well
CW-1
CW-2
CW-3
Depths (ft)
20-90
60-90
10-90
September 1991
0.97
0.59
0.56
September 1992
0.30
0.36
0.32
-31-
-------
43
"? §
*-
03 O
3 «
^ I
3 «
Sag
1- C
H) C
II
U
C vi
li
•a B
a 1
"> ao
.S .E
c "S,
6
8 «
E -
as
«*»
OUOQ
-32-
-------
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.
REFERENCES
1. Hoeppel, R.E., R.E. Hinchee, and M.F. Arthur. 1991. Bioventing soils contaminated with
petroleum hydrocarbons. J. Indust. Microbiol. 8:141.
2. Sayles, G.D., R.C. Brenner, R.E. Hinchee, C.M. Vogel, and R.N. Miller. 1992. Optimizing
bioventing in deep vadose zones and moderate climates: Hill 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: Hinchee, R.E., and Olfenbuttel, R.F., eds., In
Situ Bioreclamation. pp. 541-545. Butterworth-Heinemann, Boston.
-33-
-------
APPLICATION OF WOOD-DEGRADING FUNGI TO TREAT CONTAMINATED SOILS
John A. Glaser
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH
and
Richard T. Lamar, Mark W. Davis, and Diane M. Dietrich
Forest Products Laboratory
U.S. Department of Agriculture
Madison, WI
INTRODUCTION
Past investigations of soil treatment systems using lignin-degrading fungi have largely been
confined to laboratory 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 abandoned wood-
treating site in Mississippi to evaluate fungal 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 Recovery 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 characteristics were suitable for conducting field investigations, 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 demonstration phase results was not complete at the
time of this writing and will be reported at a later date.
-34-
-------
METHODOLOGY
The treatability study was designed to evaluate the ability of three different fungal species
to degrade pentachlorophenol (PCP) in soil. The soil pile was sampled and analyzed for PCP and
creosote components (i.e., polycyclic aromatic hydrocarbons [PAHs]) prior to developing 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 conditions. Soil beds measuring 3 m by 3 m (10 ft by 10 ft) were
constructed of galvanized sheet metal. A leachate collection system was installed to direct the liquid
discharge 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. Further homogenization was accomplished by the mixing of different
portions of screened soil. The mixed soil was then 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 the 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 (RGB) without replication and a balanced incomplete
block (BIB) with treatments replicated four times. Six of the plots were allocated to the RGB design
and four to the BIB design. The BIB plots were subdivided into 1.5 m by 1.5 m (5 ft by 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 contamination between
treatments and controls. Soil moisture was monitored on a daily basis throughout the study and
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; immediately 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
Phanerochaete 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.
-35-
-------
Table 1. Feasibility Study Experimental Design and Pentachlorophenol Transformation
Results'
Treatment1"
Regime
1
2
3
4
Fungus/Control
P. chrysosporium
»
P. sordida
P. chrysosporiiiml
T. hirsuta
Inoculum
Loading
5%
10%
10%
5%
5%
PCP
Transformed
(Dry Wgt.)
15%
67%
89%
23%
PCP
Initial Cone.
(mg/kg)
576
1017
673
615
7
8
9
10
Inoculum Carrier
Control
No Treatment
Control
T. hirsuta
P. chrysosporium
Wood Chip Control
10%
10%
13%
14%
15%
55%
52%
10% (Day 0) 55%
2% (Day 14)
0%
687
737
334
333
360
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.
bWood chips were added to each plot at a 2.5 percent loading level with the exception of the No
Treatment Control, which received no amendment.
-36-
-------
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.
SUMMARY AND CONCLUSIONS
The superior performance of P. sordida in biotransforming PCP parallels previous field study
experience with P. chrysosporiwn 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 encountered 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 developing
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 chrysosporium: 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.
-37-
-------
Table 2.
Creosote Constituent Transformation Results*
% Decrease
Compound
Acenapthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo[a]anthracene
Chrysene
Initial Cone.
(mg/kg)
429
225
941
684
972
572
74
90
No Treatment
Plot
49
75
69
57
23
10
11
6
Carrier
Plot
68
57
49
48
42
22
13
14
P. sordida
Plot
95
95
90
85
72
52
24
33
'Initial concentrations were based on soil samples taken 1 day after treatment application. Each
value for initial concentration and percentage decrease is the mean of 24 observations.
-38-
-------
PERFORMANCE EVALUATION
-------
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
INTRODUCTION
Aerobic microorganisms grown on phenol or toluene can initiate the cometabolic oxidation
of chlorinated aliphatic hydrocarbons (CAHs) to stable nontoxic end products (1,2,3,4). Such
microorganisms have good potential 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 trichloroethylene (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 jtg/L
TCE (6). The objective of the current work is to determine how effective the phenol-utilizers were
at degrading TCE over higher concentrations ranging from 62.5 to 1,000 /*g/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 monitoring locations. Experiments were performed under the induced
gradient conditions of injection and extraction. In this study, TCE was injected along with phenol
and DO. This 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 /tg/L. The
TCE concentration was gradually raised by doubling the concentration after effective transformation
had been achieved at the lower concentration. After 1 week of operation, the injected TCE
concentration was raised to 125 jig/L, then to 250 /ig/L after another week, 500 ^g/L during the
subsequent 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 jtg/L. Bromide was added as a conservative tracer
as a basis of comparison for determining transformation extents.
-41-
-------
RESULTS
Figure 1 shows the normalized concentration breakthrough of TCE at the three observation
wells SSE1, SSE2, and SSE3, located 1,2.2, and 3.8 meters from the injection well. The normalized
concentration represents the observed concentration divided by the injection concentration 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 helpful 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 wells.
For TCE injection concentrations ranging from 62.5 /*g/L to 500 /ig/L, the extent of
biodegradation is similar despite the increases in injection concentration. Bromide tracer tests
conducted over this period show complete breakthrough of the conservative tracer, C/Co = 1.
Approximately 70 and 85 percent TCE removal was observed at wells SSE1 and SSE2, respectively.
Upon increasing the TCE injection concentration to 1,000 /ig/L, removals were lower.
While injecting 1,000 /tg/L TCE, the phenol concentration 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 increase. 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 frequently 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
/tg/L, but the quantifiable detection limit was about 25 /ig/L. The phenol concentrations at the SSE2
and SSE3 monitoring locations were generally too low to give a visible response, and thus are
presumed to be below 20 /ig/L, and probably below 10 /ig/L. Thus phenol removal was excellent and
probably greater than 99.9 percent in this system.
The TCE results are summarized in Table 1. The percentage removals listed are based on
average values at the end of the period following the changes in concentration. Removal up to 89
percent was observed at the SSE3 well, with 12.5 mg/L phenol added. Removals 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-order kinetics, TCE
transformation product toxicity is beginning to have a measurable effect, or sufficient reducing power
from phenol oxidation is not available.
Transformation yields are also 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 /ig/L
TCE. Lower yields are obtained at lower TCE concentrations or at higher phenol injection
concentrations.
-42-
-------
TCE Injection Concentration Gig/1)
u
^
u
w
u
H
s
N
HH
o
2
200
600
800
1000
TIME (HR)
Figure 1. Normalized breakthrough of TCE at observation wells during biostimulation at 12.5
mg/L phenol. TCE injection concentrations range from 68 to 1,000 /ig/L.
-43-
-------
U
w
U
8
N
O
z
Phenol Injection Concentration (mg/1)
800
900
1000
1100 1200
1300
TIME (HR)
Figure 2. TCE concentration response resulting from increased phenol addition.
-44-
-------
Table 1. Average Removal Efficiencies for TCE, at Various Well Locations, and
Transformation Yields
Percentage Removal
Phenol TCE Transformation
Added Added Yield
mg/l ng/l SSE1 SSE2 SSE3 g TCE/g phenol
12.5 62 60 78 89 0.0044
12.5 125 68 82 87 0.0087
12.5 250 70 82 88 0.018
12.5 500 68 84 88 0.035
12.5 1,000 78 85 90 0.062
19 1,000 75 82 85 0.045
25 1,000 78 85 90 0.036
-45-
-------
SUMMARY AND CONCLUSIONS
An indigenous phenol-utilizing population effectively degraded TCE up to 1,000 /ig/L. At
the highest phenol injection concentration, up to 90 percent of the 1,000 /xg/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 /tg/L. A
phenol concentration of 25 mg/L was required to obtain similar removal efficiency with a TCE
injection concentration of 1,000 /tg/L. These results suggest in situ biodegradation of TCE with
phenol to be quite promising.
Future studies at the site will explore a range of contaminants including vinyl chloride,
chloroform, and 1,1-dichloroethylene. The injection of a noncompetitive source of reducing power,
such as formate, and its effect on removal efficiency will be explored. Studies with toluene as a
growth substrate will also be performed for a comparison with phenol.
REFERENCES
1. Nelson, MJ.K., S.O. Montgomery, E.J. O'Neill, and P.H. Pritchard. 1986. Aerobic metab-
olism of trichloroethylene by a bacterial isolate. Appl. Environ. Microbiol. 52:383-384.
2. Nelson, MJ.K., S.O Montgomery, W.R. Mahaffey, and P.H. Pritchard. 1987. Biodegradation
of trichloroethylene and involvement of an aromatic biodegradative pathway. Appl. Environ.
Microbiol. 53:949-954.
3. Nelson, MJ.K., S.O Montgomery, and P.H Pritchard. 1988 Trichloroethylene metabolism by
microorganisms that degrade aromatic compounds. Appl. Environ. Microbiol. 54: 604-606.
4. Wackett, L.P., and D.T. Gibson. 1988. Degradation of trichloroethylene by toluene
dioxygenase in whole-cell studies with Pseudomonas putida Fl. Appl. Environ. Microbiol.
54:1703-1708.
5. Semprini, L., P.V. Roberts, G.D Hopkins, and P.L. McCarty. 1992. A field evaluation of
in situ biodegradation 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-utilizing microorganisms.
Submitted for publication.
-46-
-------
INNOVATIVE BIOREMEDIATION STRATEGIES FOR CREOSOTE:
CHARACTERIZATION AND USE OF INOCULA
James G. Mueller and Suzanne E. Lantz
SBP Technologies, Inc.
Gulf Breeze, FL
and
Jian-Er Lin
Technical Resources, Inc.
Gulf Breeze, FL
and
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 that utilize high-molecular-weight polycyclic
aromatic hydrocarbons (HMW PAHs), such as those found in creosote, coal tar, and crude oil, as
sole sources of carbon and energy for growth (1,2). In addition, microorganisms such as
Mycobacterium sp. strain PYR-1 (Dr. Carl Cerniglia, U.S. Food and Drug Administration [FDA]
National Center for Toxicological Research, Jefferson, Arkansas), which co-metabolize mixtures of
HMW PAHs and structurally related compounds, have been described (3,4). Concomitantly, our
knowledge of physicochemical factors limiting biological activity in the field (e.g., bioavailability) has
increased significantly, while innovative methods for assessing and monitoring field performance have
been developed.
Successful combination of these technological advances 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. Despite 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
-47-
-------
effectively employ inoculant microorganisms in the field. Thus, as the practical applicability of
bioremediation technologies are expanded to include more persistent chemicals, such as those found
in creosote (i.e., HMW PAHs), new inoculation procedures and implementation strategies need to
be developed.
Expected advantages of augmented or otherwise improved microbiology of contaminated
environments include 1) accelerated rates of biodegradation (e.g., shorter treatment time), 2)
furthered extent of contaminant removal ("cleaner" end product), 3) reduced probability of
undesirable "side-reactions" (e.g., production of more toxic intermediates by indigenous microbes),
and 4) ensured consistency and reliability of biodegradative performance in the field thus yielding
more cost-effective treatment technologies. The overall effectiveness 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, includingbioreactor operations, solid-phase (landfarming), 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 warranted. In the most
straightforward scenario, the inability of indigenous microflora to degrade or transform a persistent
chemical is known. In this case, the addition of relevant catabolic machinery, in conjunction with
requisite chemical resistances (e.g., heavy metals) or physical tolerances (e.g., low temperature), may
be accomplished via inoculation. A second scenario considers 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 facilitate a
more rapid increase in degrader biomass, hence potentially accelerating the biodegradation process.
Lastly, when indigenous microorganisms effect incomplete or partial catabolism, inoculants can be
used to provide alternative catabolic pathways. Of these possibilities, 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, microorganisms with requisite resistances and/or tolerances.
Few reports describe the use of microbial inoculants to positively influence the
biodegradation of targeted compounds in the field under solid-phase conditions, but most studies
have been conducted under laboratory conditions. Alternatively, the most successful use of
inoculants, 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 controlled 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 their presence. This may be because inoculation is often
performed in conjunction with other efforts to modify the chemical and/or physical environment,
hence the comparative data necessary to clearly elucidate the response as directly attributable to the
inoculant per se are not available. From more of a microbiological perspective, other potential
reasons for this lack of a discernible response include 1) inoculant microorganisms may not
successfully compete with indigenous microflora for essential micro-sites or ecological niches
essential for their survival, 2) inoculant bacteria are particularly prone to predation, and 3)
inoculants experience rapid die-off due to desiccation and other physiological shocks (e.g.,
temperature changes, osmotic changes).
-48-
-------
In an effort to overcome factors known to inhibit inoculation procedures for bioremediation
applications, a variety 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 provide an
effective means of inoculant storage and distribution.(Un et al., unpublished data). Furthermore,
newly developed encapsulated cell technologies provide an efficient means of applying high numbers
of viable, catabolically active inoculant cells (>1 x 1010 bacterial cells/g inoculant) to soil and
facilitating their timed release (Lin et al., unpublished data).
Encapsulation technologies currently are being used to improve the solid-phase
bioremediation of soils contaminated with organic wood preservatives (i.e., creosote 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 nutrients and electron acceptor; and 3) offer an ecological niche conducive to microbial
growth, proliferation, and catabolism. Results to date have identified effective encapsulation and
immobilization technologies for various PAH- and pesticide-degrading bacteria. Depending on the
desired end points, this strategy may offer a viable approach for remediating soils contaminated with
creosote and similar substances. Similarly, immobilized cells in liquid bioreactor systems (above-
ground or in situ bioreactors) are being tested for their ability to treat ground water impacted by
related compounds. In addition, research is being conducted to determine the effectiveness of
co-encapsulating microorganisms (e.g., HMW PAH-degraders) with nutrients, electron acceptors,
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 (IEG, Technologies, Inc. and
GfS, Germany). Other in situ designs have been developed in collaboration with Drs. John Cherry
(University of Waterloo-Canada) and Malcolm 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 Research 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 (contract no.
FTTA-003).
REFERENCES
1. Mueller J.G., PJ. Chapman, B.O. Blattmann, 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. Mueller, J.G., PJ. Chapman, B.O. Blattmann, and P.H. Pritchard. 1990. Isolation and
characterization of a fluoranthene-utilizing strain of Pseudomonas pauclmobilis. Appl.
Environ. Microbiol. 56:1079-1086.
-49-
-------
3. Grosser, R.J., D. Warshawsky, and J.R. Vestal. 1991. Indigenous and enhanced
mineralization of 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. Ind. 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. Patent No. 5,143,606.
7. Lin, J-E., J.G. Mueller, and P.H. Pritchard. 1992. Use of encapsulated microorganisms as
inoculants for bioremediation. Amer. Chem. Soc. Special Session on Biorernediation of Soils
and Sediments. September 21-23, 1992, Atlanta, GA.
8. Lin, J.-E., H.Y. Wang, and R.F. Hickey. 1991. Use of coimmobilized biological systems to
degrade toxic organic compounds. Biotechnol. Bioengineer. 38:
9. European Patent Office. 1989. Encapsulation Method. C.A. Baker, A.A. Brooks, R.Z.
Greenley, and J.M.S. Henis. Publication No. 0 320 483.
10. Stormo, K.E., and R.L. Crawford. 1992. Preparation of encapsulated microbial cells for
environmental applications. Appl. Environ. Microbiol. 58:727-730.
-50-
-------
DECONTAMINATION OF PCB-CONTAMINATED SEDIMENTS THROUGH THE USE OF
BIOREMEDIATION TECHNOLOGIES
John F. Quensen, III, Stephen A. Boyd, and James M. Tiedje
Michigan State University
East Lansing, MI
and
John E. Rogers
Environmental Research Laboratory
U.S. Environmental Protection Agency
Athens, GA
INTRODUCTION
It is now widely recognized that polychlorinated biphenyls (PCBs) can be reductively
dechlorinated in anaerobic environments such as sediments. This process was first suggested by PCB
congener distribution patterns observed in sediment core samples from the upper Hudson River (1).
Compared to both surficial sediments and the Aroclor 1242 originally input to these sediments,
chromatograms for deeper sediment samples showed a depletion of the more heavily chlorinated,
later eluting congeners and a corresponding relative increase in lesser chlorinated, early eluting
congeners. This transformation process has since been demonstrated to be the result of anaerobic
microbial activity (2), and has been observed in PCB-contaminated sediments from a number of
other sites (3,4).
The discovery of the anaerobic dechlorination of PCBs has stimulated interest in developing
a sequential anaerobic/aerobic biotreatment process for their destruction. While the 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, all products of the anaerobic 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 accumulate as a result of
anaerobic PCB dechlorination can serve to induce PCB-degrading enzymes in aerobic
microorganisms. Although more highly chlorinated congeners can be aerobically cometabolized, they
are not inducing substrates.
-51-
-------
To develop a sequential anaerobic/aerobic biotreatment process for PCBs, more about the
factors controlling the anaerobic dechlorination process needs to be understood. The extent of in
situ PCB dechlorination that has occurred varies among sites (Table 1). This is likely attributable
to both the congener specificity of the predominant PCB-dechlorinating microorganisms present and
certain environmental characteristics. Recognizing that environmental characteristics may have
changed since the time of in situ dechlorination, assays were conducted to determine the ability of
each sediment to currently support dechlorination. Sediments from each site were slurried with
reduced anaerobic mineral medium and inoculated with Hudson River microorganisms, and Aroclor
1242 (500 jug/g sediment) was added. The amount of chlorine removed from the 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 analytical 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 laboratory experiments
designed to better define the impact of these factors on the dechlorination process are summarized
below. Research to be undertaken under a cooperative agreement between Michigan State
University and EPA will address ways potentially limiting factors, 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 concentrations 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
microorganisms.
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 comparable 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 that 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 hydrocarbons content) might be important in inhibiting PCB dechlorination. Vacuum
pump oil was chosen for the above experiment because it does not contain potentially toxic or
inhibitory additives. Oils found in environmental samples also are being characterized and their
effect on PCB dechlorination is being determined.
-52-
-------
Table 1. Extent of Dechlorination and Selected Environmental Parameters for PCB-
Contaminated Sediments
Sediment
In situ
Dechlori-
nation
Assay
Dechlori-
nation
Total
PCBs
(ug/g)
Oil &
Grease
(ug/g)a
Total
Metals
(ug/g)b
Acushnet 17 5
Acushnet 19 26
Hudson H7 74
Lagoon A 37
Lagoon B 38
Raisin D 9
Raisin E (0-10 cm) 7
Raisin E (10-25 cm) 3
Saginaw 8R (10-20 cm) 7
Sheboygan R212 41
Sheboygan R28 63
Sheboygan R9 67
Silver L F3 32
0
0
1
0
0
1
1
.3
.5
.6
0
0
.8
0
0
0
.5
.1
.8
0
1,
1,
25,
14,
20
112
74
800
113
88
093
993
323
418
395
319
694
10,
11,
17,
38,
7,
3,
3,
13,
1,
62,
730
939
626
750
338
157
807
542
092
083
955
581
856
2
2
1
1
3
3
1
5
9
,796
,790
,143
,934
,352
,919
,152
542
,427
295
292
244
,135
a Oil and grease concentrations were determined gravimetrically and
corrected for PCB concentration.
b Total concentration of cadmium, chromium, copper, nickel, lead, and zinc.
Note: In situ dechlorination is defined as the percent decrease in meta and para chlorines that
has occurred naturally in the environment. Assay dechlorination is defined as the jtg
atoms of chlorine removed from PCBs in a 16-week laboratory incubation after addition
of Aroclor 1242 (500 jig/g sediment).
-53-
-------
Table 2. Extent of Aroclor 1242 Dechlorination
Avg m & p Cl's
Treatment Mean S.D.
0
5
10
20
40
0.85
1.10
1.16
1.26
1.30
0.11
0.06
0.02
0.04
0.03
1242 1.8
*mg oil / g sediment
Note: Dechlorination is as indicated by the average number of chlorines 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-contaminated Hudson River
sediments used in the dechlorination assay. No dechlorination from the ortho positions was
indicated. Addition of 40 mg oil/g sediment reduced the extent of dechlorination by half.
-54-
-------
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 /xg/g
sediment of copper, lead, zinc, or chromium) were added to pre-incubated anaerobic sediment
slurries. The slurries were then autoclaved and inoculated with Hudson River microorganisms, and
Aroclor 1242 was added. These experiments have been sampled over 8 weeks and preliminary
analysis suggests 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 conducted to ensure that the extraction process itself does not adversely affect
the ability of the sediments to support PCB dechlorination.
FUTURE RESEARCH OBJECTIVES AND APPROACH
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 environmental 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 overcoming 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 chemically or biologically providing more reducing conditions. Subsequently, an
attempt will be made to overcome limitations attributable to metal toxicity by chelation, extraction,
precipitation, and bioleaching. Research aimed at overcoming limitations relating to the congener
specificity of the microorganisms involved currently is being investigated in a separate project funded
by General Electric. Together, these efforts represent a comprehensive evaluation of factors limiting
PCB dechlorination and approaches for overcoming these limitations.
REFERENCES
1. Brown, J.F., R.E. Wagner, D.L. Bedard, MJ. Brennan, J.C. Carnahan, RJ. May, and TJ.
Tofflemire. 1984. PCB transformations in upper Hudson sediments. Northeast. Environ.
Sci. 3:167-179.
2. Quensen, J.F., III, J.M. Tiedje, and S.A. Boyd. 1988. Reductive dechlorination of
polychlorinated biphenyls by anaerobic microorganisms from sediments. Science 242:752-754.
3. Brown, J.F, D.L. Bedard, MJ. Brennan, J.C. Carnahan, H. Feng, and R.E. Wagner. 1987.
Polychlorinated biphenyl dechlorination in aquatic sediments. Science 236:709-712.
4. Brown, J.F., R.E. Wagner, H. Feng, D.L. Bedard, MJ. Brennan, J.C. Carnahan, and RJ.
May. 1987. Environmental dechlorination of PCBs. Environ. Toxicol. Chem. 6:579-593.
-55-
-------
5. Quensen, J.F., III, S.A. Boyd, and J.M. Tiedje. 1990. Dechlorination of four commercial
polychlorinated biphenyl mixtures (Aroclors) by anaerobic microorganisms from sediments.
Appl. Environ. Microbiol. 56:2360-2369.
6. Bedard, D.L., R.E. Wagner, MJ. Brennam, M.L. Haberl, and J.F. Brown, Jr. 1987. Extensive
degradation of Aroclors and environmentally transformed polychlorinated biphenyls by
Alcaligenes 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 organic contaminants in soils. Environ. Sci. Technol. 24:142-144.
-56-
-------
NATURAL ANAEROBIC BIOREMEDIATION 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
and
Don H. Karapbell and John T. Wilson
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
Ada, OK
INTRODUCTION
In a sand aquifer near the town of St. Joseph, Michigan, trichloroethylene (TCE),
1,2-cis-dichloroethylene (c-DCE), 1,2-trans-dichloroethylene (t-DCE), and vinyl chloride (VC) are
present at ground water concentrations 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
concentrations. 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 contamination and of the natural mixing and
transformation processes. This presentation summarizes the methodology and the results of this
analysis.
The geologic formation of interest is an unconfined aquifer 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 the creek, which act as constant-head boundaries, at a nearly
-57-
-------
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 the
absence of oxygen (2) and in fact have been shown to do so under a variety of environmental
conditions. Common anaerobic electron acceptors and the associated microbial process, in the order
of their redox potential, are nitrate (denitrification); Mn(IV) (manganese reduction); Fe(III) (iron
reduction); sulfate (sulfate reduction); and carbon dioxide (methanogenesis). Many of the
chlorinated aliphatic compounds are highly oxidized and have redox potentials above those of
common electron acceptors (3). Generally, the less halogenated components are transformed 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 chemical characterization of the St. Joseph site that will be presented
supports previous laboratory and field findings. The characterization permits zones to be identified
where transformations are occurring, and permits flux estimates of the contaminants and
transformation products. 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
Laboratory (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 plume. Onsite gas
chromatography analysis was performed for TCE and its anaerobic transformation products. 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 succession of boreholes. Samples for volatile organic compounds (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 concentration and to estimate flux rates. Contouring and averaging 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 measurement and account for the shape of the concentration surface. Linear
geostatistical methods account for spatial 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 measurements to "hot spots." Also, they are highly skewed (i.e., asymmetrically
distributed).
-58-
-------
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 values), which make it
possible to evaluate the accuracy of estimated concentrations and masses. The approach accounts
for the skewness in the distribution of estimation errors by adding only one parameter to those used
in linear geostatistics (variograms or generalized covariances). 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 previous 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 concentration distributions obtained from the
two-dimensional contour analysis in the two transects that are roughly perpendicular to the ground
water flow. Complete transformation of TCE to ethene was associated with methanogenic
conditions in the aquifer. VC and ethene concentrations turned out to be correlated with the
highest methane concentrations. Zones of high TCE concentration 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
methanogenesis is not occurring, for example, because of an absence of appropriate bacteria,
electron donors, or redox conditions. 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 formation process(es). c-DCE was found at a slightly greater
depth than 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 concentration 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 methanogenic conditions. Active VC formation
and transformation to ethene occurred under methanogenic conditions that were observed at greater
depth.
Flux estimates were made based on the contour average concentrations and an estimated
ground water velocity 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 transects 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
-59-
-------
of four greater at transect 2 than at 1, suggesting more highly anaerobic conditions upgradient. The
differences between the two transects, however, might be due to 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 chlorinated 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
percent of the COD reduction was associated with reduction 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 anaerobic transformation
of TCE to nontoxic end products.
REFERENCES
1. McCarty, P.L., and J.T. Wilson. 1992. Natural anaerobic 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. Griddle, and P.L. McCarty. 1987. Transformations of halogenated
aliphatic compounds. Environ. Sci. Technol. 21:722-736.
4. Griddle, C.S., and P.L. McCarty. 1991. Electrolytic model system for reductive
dehalogenation in aqueous environments. Environ. Sci. Technol. 25:973-978.
-60-
-------
COMPARISON OF BIOVENTING AND AIR SPARGING
FOR IN SITU BIOREMEDIATION OF FUELS
Don H. Kampbell
U.S. Environmental Protection Agency
Ada, OK
and
Christopher J. Griffin
Solar Universal Technology, Inc.
Traverse City, MI
and
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, ethylbenzene, 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 can be in situ bioremediation processes that provide air flow to
vaporize and transport volatile organic pollutants upward to more amenable media for
mineralization. Microbial degradation processes also utilize the oxygen provided by the air. Injection
-61-
-------
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 surface 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 effectively 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 by 45 ft rectangular area overlying the plume of contamination.
An initial nutrient solution of 64 Ib nitrogen, 5 Ib potassium, and 13 Ib phosphorus was applied for
dispersion throughout the unsaturated subsurface. Aeration injection wells screened across the water
table were placed 10 ft apart in a 3 by 5 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-inch 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 subsurface 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 cartridge 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 uncapped one at a time, and a plug aliquot was removed and preserved for later
analysis of total petroleum hydrocarbons. 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 by 4 grid 10 ft apart. A minimum radius of influence of 15 ft was
determined by pressure changes at distances away from a sparge well (3). A blower rate of 10 cfm
was used.
RESULTS
Venting operations started during October 1990. Gasoline hydrocarbon vaipor in the vadose
zone initially increased to near 5,000 mg/L, then within 3 weeks decreased 80 percent (Figure 1). A
gradual decline continued 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 /xg/L, compared to 3.2 ft depth soil gas levels of 163 mg/L. Biomass
degradation activity from the turf rhizosphere was greater by a factor of 10 than at the same depth
of a barren soil control that was not acclimated to gasoline vapor.
-62-
-------
Vertical profile fuel carbons for October 1990 and October 1991 are shown in Figure 2. The
water table was at the same level of 15.3 ft for both time periods. A reduction of oily phase residue
in the vadose zone occurred 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 analyzed 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 October 1991 and September 1992. Heterogeneity obtained from vertical
profile sampling 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 reduction 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.
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 operations. Proceedings of Subsurface Restoration Conference, 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.
-63-
-------
Btovvnttog BJovcntlng
Startup R««tart«p
10
Start vp
Tme (Months she* 10/1/90)
lUptlcut*
Location
Figure 1. Soil gas hydrocarbons at 4-m depth for injection-only plot.
600-
598
m
594
o-9/90
• -10/91
GWTtble
1000 2000 3000
Fuel Carbon, mg/Kg Core Material
4000
Figure 2. Vertical profile oily phase residue in bioventing of north plot.
-64-
-------
Table 1. Ground Water Quality after Seven Months of Biosparging
Monitoring
Well
Control
Sparge Plot
Well Depth
-ft-
16
17.5
20.5
22
15
18
19.5
21
Benzene
Xylenes
Total Fuel
Carbon
Mg/L
9.9
228
70
57
1.9
>1
>1
>1
19
992
38
7.7
5.3
5.0
>1
>1
2880
4490
956
783
559
>6
>6
>6
-65-
-------
Table 2. Core Assay of Avgas Vapor during Drilling Operations
Control
Sparge plot
Initial
One Year Later
Replicate Average over 40 in. profile,
rag Avgas/ft
4070 (3 reps)
5980 (3 reps)
3920 (2 reps)
1220 (4 reps)
Reduction
%
80
-66-
-------
Table 3. Total Fuel Carbon in Vertical Profile
Sparge plot
Initial
One Year Later
Replicate average, mg/ft2 surface area
139,540 (4 reps)
85,230 (5 reps)
Reduction
%
39
-67-
-------
DETERMINING THE GENOTOXICITY OF BIOREMEDIATION PRODUCTS
Larry D. Claxton, S. Elizabeth George, Robert W. Chadwick, and Virginia S. Houk
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC
and
L.E. Rudd and J.J. Perry
North Carolina State University
Raleigh, NC
and
John T. Wilson
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
Ada, OK
INTRODUCTION
The presence of toxicants in the environment can be determined either by direct analytical
chemistry measurements of known toxic substances or by using bioassays. The wide variety of man-
made and naturally occurring toxicants in the environment, including carcinogens and mutagens, is
well documented. Carcinogens (substances that initiate the cancer process) and mutagens
(substances that cause heritable effects in cells 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).
Because environmental samples often contain mixtures of both identified and unidentified
genotoxicants, a need to evaluate environmental samples using bioassays exists. 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 monitoring is that the lack of hazard
-68-
-------
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 genotoxicity in the 1970s led to the use of these
assays for evaluating complex environmental mixtures. The genetic assay that has demonstrated the
greatest utility is the Salmonella typhimurium mutagenicity assay. Many genetic assay systems have
now been applied to understanding the toxicity of complex environmental situations. 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
understand the effect of bioremediation on genotoxicity.
LABORATORY-SCALE STUDIES EXAMINING BIOREMEDIATION PRODUCTS
In order to evaluate the potential for both the degradation and production of genotoxicants
during the bioremediation of crude oils, biodegradation products of three crude oils metabolized by
one of two species of fungi was monitored by analytical chemistry methods and by use of the 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
environmental substances of no, moderate, and high mutagenicity. 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 response after degradation. In all cases, "weathered" control samples (incubated
without the fungus), showed little to no change in mutagenicity.
This laboratory experiment illustrates the type of complexity inherent in understanding the
toxicity of bioremediated 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 nonmutagenic oil during fungal degradation have not been identified. Chemical
analysis would not be expected to provide an appropriate approximation of relative genotoxicity for
these samples. Bioassays, however, shown to screen for the destruction and creation of
genotoxicants, 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 bioremediation of the site, core samples will be taken so that
mutagenicity before and after remediation efforts can be compared.
-69-
-------
EVALUATING THE EFFECTS OF MULTIPLE POLLUTANTS
2,6-Dinitrotoluene (DNT) represents a class of compounds that can be found in hazardous
waste sites contaminated with munition chemicals. The organochlorine pentachlorophenol (PCP)
is a wood preservative found within many sites. The activation of DNT to genotoxic metabolites
involves enzymes in both the liver and the intestinal flora. PCP has both bactericidal activity and
can induce hepatic mixed function oxidase enzymatic activity in the liver. Studies were done
therefore to determine the effect of PCP on the production of mutagenic metabolites from DNT.
Results of the study indicated that PCP accelerates the biotransformation of DNT to genotoxic
metabolites and potentiates the formation 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 combinations of pollutants must be
closely monitored.
SUMMARY
Although bioassays have been used for many years to evaluate ecotoxicity in complex
environmental situations, 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 applied to bioremediation research to identify toxic
pollutants, 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 lexicological processes.
-70-
-------
FIELD RESEARCH
-------
FIELD DEMONSTRATION OF A CONSTITUTIVE TCE-DEGRADING BACTERIUM FOR
THE BIOREMEDIATION OF TCE
Malcolm S. Shields, Michael Reagin, Robert Gerger, Rhonda Schaubhut, and Robert Campbell
Center for Environmental Diagnostics and Bioremediation
Department of Biology, University of West Florida
Pensacola, FL
and
Charles Somerville
Technical Resources Inc.
Gulf Breeze, FL
and
P. Hap Pritchard
Gulf Breeze Environmental Research Laboratory
U.S. Environmental Protection Agency
Gulf Breeze, FL
INTRODUCTION
The degree to which trichloroethylene (TCE) has been 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 notoriety as a persistent point source pollutant, despite numerous reports of
both anaerobic and aerobic bacterial transformation capabilities. Aerobic bacteria are more rapid
TCE metabolizers, but only in a cooxidative fashion. TCE serves as a cooxidative substrate for
various 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).
-73-
-------
BACKGROUND
Our research has centered on the miocrobiology of P. cepacia G4, which expresses a unique
toluene ort/io-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 nonrecombimmt derivative of
G4, called G4 PR1 has been developed that constitutively expresses Tom, and consequently degrades
TCE without the need for a co-inducer (14). This paper deals with the characterization, alteration,
and application of this constitutive derivative.
RESULTS AND CONCLUSIONS
Bioreactor
A fixed film bioreactor was investigated for the exploitation 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 PR1 biofilm capable of TCE degradation: cellite pellets
(diatomaceous earth pellets manufactured by Manville Corp.), gravel, glass beads, activated charcoal,
sand, 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 bioreactor research.
All column designs tested thus far involved the metered application of nutrients (yeast
extract, peptone, and glucose 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-entrained
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 nutrient was input beyond initial colonization. PR1, present at _>_ 1.4 x 108bacteria/g
of the support material, completely removed the 80 /*M TCE (10 mg/L air) present in the gas phase
under the same conditions, over a 20-hour period.
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 ^M of air-entrained
TCE introduced at a rate of 4 mL/min over a 96-hour period.
A similar design of a 26-liter column reactor (Figure 1) allowed for nutrient (lactate) liquid
recycle and a continuous input of freshly grown cells from a lactate-fed chemostat during the
inoculation phase. Two separate tests of this reactor system at 0 percent refeed were conducted.
One resulted in a 92 percent removal of TCE at an average input concentration of 12.6 ^M (2
mg/L) in the air phase over a 72-hour test.
In another test, TCE was introduced at appreciably lower levels and with a significant
nutrient/liquid recycle. After a 48-hour inoculation period, air-entrained TCE (~ l^M [131 ^g/L])
was introduced at 100 mL/min total flow rate (c.a. 58 percent oxygen). Under these conditions,
approximately 90 percent of the TCE present in the gas phase was continuously removed during a
4-day test period (Figure 2).
-74-
-------
Spent Nutrtantt
0.36X fcg
CtyWOHCOOH
Figure 1. Schematic of 26-liter column reactor.
-75-
-------
Input
Output
LJU
o
50
Time (HRS)
100
Figure 2. Twenty-six liter biofilter containing crushed oyster shell and colonized with
Pseudomonas cepacia G4 PR1. Air-entrained TCE was delivered from a compressed
gas cylinder and mixed with oxygen to balance nutrient requirements.
-76-
-------
Tom expression in liquid culture or as a biofilra 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 nonmetabolizable trifluoroheptadienoic acid (TFHA) (a reaction
requiring both Tom and catechol-2,3-dioxygenase [C23O]). 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 Catabolic Range and Resistance to Toxic Intermediates
TCE in aquifers is frequently found in combination with other pollutants. Chloro-aromatics
like chlorobenzene and chlorophenol are of considerable concern due to their conversion via Tom
to toxic chlorocatechols, which are not metabolized by C23O. 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 pROlOl (a tetracycline derivative of pJP4, specifying the degradation of
2,4-D) was introduced into PR1. In so doing, the transconjugant PR1 (pROlOl) was created that
expresses enzymes for ortfzo-fission (i.e., catechol-l,2-dioxygenase [C12O]) of the resultant 3-
chlorocatechol, in addition to the nonproductive meta-cleavage mechanism of the C23O 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 inhibitory to cellular
metabolism, including its ability to degrade TCE. PR1 (pROlOl) is capable of growth on 2,4-D,
chlorobenzene, or 2-chlorophenol as the sole carbon sources and is also effective in degrading TCE
at approximately 20 /*m, following a 24-hour pre-exposure to 2-chlorophenol at 1 mM. PR1 is not
able to utilize these carbon sources, nor is it able to degrade TCE following exposure to these levels
of chlorobenzene or 2-chlorophenol. This brings the established substrate range of PR1 (pROlOl)
to include 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 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 parameters affected by the reactor design
and operation. Future considerations will include the organisms ability to metabolize TCE in an in
situ bioreactor.
The substrate range of Tom has been established to include TCE, vinyl chloride, els- and
frans-l,2-dichloroethylene, 1,1-dichloroethylene, benzene, toluene, phenol, ort/io-xylene, and ortho-,
rneta-, andpara-cresol. P. cepacia PR1 carrying pROlOl is capable of using these same substrates
in addition to 2,4-D, chlorobenzene, 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
-77-
-------
system remains to be demonstrated. The introduction of pROlOl into PR1 resulted in expansion
of the substrate range of this organism to include chlorobenzene 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. Knackmuss. 1990. Arch. Microbiol. 154:410-413.
3. Folsom, B.R., and PJ. Chapman. 1991. Appl. Environ. Microbiol. 57:1602-1608.
4. Fox, E.G., J.G. Borneman, L.P. Wackett, andJ.D. Lipscomb. 1990. Biochemistry 29:6419-
6427.
5. Harker, A.R., and Y. Kim. 1990. Appl. Environ. Microbiol. 56:1179-1181.
6. Henry, S.M., and D. Grbic-Galic. 1991. Appl. Environ. Microbiol. 57:236-244.
7. Kaphammer, B.J., JJ. 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, PJ. Chapman, and P.H. Pritchard. 1989. Abstr. K-68, p.
256. Abstr. 89th Annu. Meet. Am. Soc. Microbiol.
9. Nelson, MJ.K., S.O. Montgomery, and P.H. Pritchard. 1988. Appl. Environ. Microbiol.
54:604-606.
10. Nelson, MJ.K., S.O. Montgomery, EJ. O'Neill, and P.H. Pritchard. 1986. Appl. Environ.
Microbiol. 52:383-384.
11. Nelson, MJ.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, PJ. Chapman, and P.H. Pritchard. 1991.
Appl. Environ. Microbiol. 57:1935-1941.
13. Shields, M.S., S.O. Montgomery, PJ. Chapman, S.M. Cuskey, and P.H. Pritchard. 1989.
Appl. Environ. Microbiol. 55:1624-1629.
14. Shields, M.S., and M.R. Reagin. 1992. Appl. Environ. Microbiol.
15. Wackett, L.P., G.A. 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.
-78-
-------
17. Worsey, MJ., and P.A. Williams. 1975. J. Bacteriol. 124:7-13.
18. Zylstra, G.J., L.P. Wackett, and D.T. Gibson. 1989. Appl. Environ. Microbiol. 55:3162-
3166.
-79-
-------
BIOREMEDIATION OF TCE: MONITORING THE FATE AND EFFECTS OF A
MICROORGANISM USED IN A FIELD BIOAUGMENTATION STUD?
M.S. Shields, R. Snyder, M. Reagjn, R. Gerger, and R. Campbell
Center for Environmental Diagnostics and Bioremediation
Department of Biology, University of West Florida
Pensacola, FL
and
C. Somerville
Technical Resources Inc.
Gulf Breeze, FL
and
P.H. Pritchard
Environmental Research Laboratory
U.S. Environmental Protection Agency
Gulf Breeze, FL
INTRODUCTION
The development of a constitutive trichloroethylene (TCE)-degrading Pseudomonas cepacia
provides us with a unique opportunity to study several microbiological 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 effectiveness in biodegrading a target pollutant like TCE is an exploration
of the microbial behavior of such a laboratory construct under anticipated operational conditions.
These issues are more readily addressed in a contained bioreactor than in an. environmental
application. Problems associated with the use of laboratory bacteria in field releases include
optimizing the activity of the organism under environmental conditions and defining the risk
associated with the introduction of a non-native or genetically altered microorganism. The use of
a co-oxidative bacterial pathway does not permit direct selection for the organism because the
pollutant cannot be utilized 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
-80-
-------
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 temporal scales of their intended activity.
2. Response of indigenous populations of bacteria and bacteriovores to the added
biomass, nutrient enrichment, contaminant, and supplemental substrate metabolites.
PERSISTENCE AND ACTIVITY
P. cepacia G4 PR1 is a nonrecombinant derivative of P. cepacia G4 that constitutively
expresses toluene ortJto-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 tracking of that organism by
detection of those sequences in nucleic acid samples extracted from the environment. Our approach
centers around determination of the nucleotide sequence of three regions of PR1 DNA:
1. The Tom and catechol-2,3-dioxygenase (C23O) genes responsible for the oxidation
of several aromatic chemicals and TCE located on the large plasmid 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 coli to produce indigo from indole (a trait not demonstrable
in G4).
Since the Tom pathway is believed to be broadly distributed, 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 conjunction with Tom and C23O genes. The junction site of Tn 5
insertion in G4, however, should be unique over a 20-to-50 base pair sequence. Another "primer"
sequence can be obtained from the published sequence of Tn 5, and the two can be used in concert
as polymerase chain reaction (PCR) primers. It also would be possible to utilize internal Tn 5
sequences, or PR1 sequences, 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.
-81-
-------
EVIDENCE FOR TOM BEING PLASMID ENCODED
G4 and its derivatives contain two plasraids of approximately 50 and 112 kb (designated
pG4S and pG4L, respectively, for those found in strain G4; pPRlS and pPRIL for those in strain
PR1). Following 100 generations of nonselective growth, G4 PRl(pROlOl) 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 pROlOl and pPRlS, but not pPRIL.
Strain G4 (Km' and phenol inducible for TCE degradation [TCE+P]) was mated to a Kmr,
TCF derivative of PR1: PR1 (pROlOl, pPRlS, pPRlL'). A Km'transconjugant was isolated that
carried pG4L, pPRlS, and pROlOl. This strain was found to have regained the ability to degrade
phenol. TCE degradation, however, occured only after phenol induction: TCE+P,
The TCF strain PR1 (pROlOl, pPRlS, pPRIL') was rendered resistant to nalidixic acid
(Nal) and rifampicin (Rif) and mated to PR1 (pPRlS, pPRIL) (Nals, Rif). The 112-b pPRIL was
transferred from the Nals, Rif strain to the phenol; Nalr, Rif PR1 (pROlOl, pPRlS, pPRIL') strain
to create theNalr, Rif strain PR1 (pROlOl, pPRlS, pPRIL). PR1 (pROlOl, pPRlS, pPRIL) also
was able to degrade phenol and TCE. However, PR1 (pROlOl, pPRlS, pPRIL) degraded TCE
constitutively (TCE+C). This is persuasive evidence that Tom and C23O as well as their
transcriptional controllers are located on the large indigenous plasmid.
CLONING
The genes responsible for Tom and C23O have been cloned from PR1 into a 2.9-kb E. coli
cloning vector, pGEMSZ, as an 11-kbEco RI 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, C23O activity (encoded by the Tom pathway in G4) was detectable in this recombinant E.
coli. Various subclones of this Eco RI 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 RI fragment of pMS64 with total DNA isolated from G4
(pG4S, pG4L) and its derived strains—PR1 (pROlOl, pPRlS), PR1 (pROlOl, pPRlS,
pPRIL)-—took place only with the large, approximately 112-kb plasmid (i.e., pG4L or pPRIL) in
each case. We propose that pG4L will serve as the archetype for a new class of catabolic plasmid
known as TOM, which encodes an o/f/io-hydroxylation pathway for the degradation of benzene,
toluene, o-xylene, cresois, 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 G4 and TOM. Its existence now poses some provocative possibilities for environmental
remediation of several priority aromatic and aliphatic pollutants. This is not only from the point of
view of their constitutive degradation, but also is based on the observation that TOM appears to be
-82-
-------
01 2345678
,11111111
1
torn A
TomC230 E S EKBH HK K S B H
tomB
B pGEM4Z B B
B pGEM4Z B B
E pGEM4Z B
B pGEM4Z B
E pGEM4Z H H H
E pGEM4Z H
E pGEM4Z . S
E pGEM4Z K K K
•• - u»..l.. -- .. '
f pGEM4Z f *
E pGEM4Z
(Sm,al) pGEM4Z f
E PGEM4Z (Sm,al>
H Hmdlll
9 10 11 12 13 14kb sssp*
1 1 1 1 1 1 B BamHI
__— _. f EroRI
Hoal
B HN SI NB E
1 1 1 1 1 II DMS64
B
| pMS69
pMS70
B E
1 | pMS75
B B
1 | pMS76
1 (E pMS71
H E
1 1 DMS73
S E
1 1 DMS78
E
1 DMS79
E
| pMSRO
N N E
1 | | pMSSl
(HfT" PMS82
(Hpal) E
1 | pMS83
Figure 1. Eco RI fragment that encodes Tom activity.
-83-
-------
a broad host range, self-transmissible plasmid (a kanamycin-resistant derivative was transferred to
E. coli where it is currently stably maintained).
The cloning and expression of the genes for Tom and C23O (i.e., torn A and torn B) in £.
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 the 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 important, the fate of the genetic information and its capacity for phenotypic
expression.
RESPONSE OF NATIVE POPULATIONS
Monitoring the response of native bacteria and bacterivorous protists to the implementation
of this technology will involve a variety of parameters aimed at defining numerical and functional
responses of bacteria and protists. Total bacterial numbers will be determined by epifluorescence
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 incorporation of 3H 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 determined
by Winkler titration. Total dissolved organic carbon (Shimadzu TOC) and the carbon-to-nitrogen
ratio of paniculate 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 microbial organisms in the
aquifer material will also be tested by determining the ability of G4 to colonize biofilms developed
from aquifer material. This will be tracked with fluorescence techniques, using scanning confocal
laser microscopy.
Previous studies of contaminated ground water sites have indicated a significant numerical
response of bacterivorous protist populations as part of overall microbial response to carbon
enrichment from contaminants (7, 8). The ecological roles of these protists affect the dynamics 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 technique (9), as well as by direct fluorescent counts.
-84-
-------
REFERENCES
1. Shields, M.S., S.O. Montgomery, S.M. Cuskey, PJ. Chapman, and P.H. Pritchard. 1991.
Appl. Environ. Microbiol. 57:1935-1941.
2. Shields, M.S., S.O. Montgomery, PJ. Chapman, S.M. Cuskey, and P.H. Pritchard. 1989.
Appl. Environ. Microbiol. 55:1624-1629.
3. Shields, M.S., and M.R. Reagjn. 1992. Appl. Environ. Microbiol. 58:3977-3983.
4. Worsey, MJ., and P.A. Williams. 1975. J. Bacteriol. 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. Environ. 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 International Symposium on Microbiology of
the Deep Subsurface, pp. 3-39 to 3-45.
9. Sinclair, J.L., and W.C. Ghiorse. 1987. Appl. Environ. Microbiol. 53:1157-1163.
-85-
-------
FACTORS DETERMINING THE EFFECTIVENESS OF MICROBIAL INOCULATION
IN SOILS AND SEDIMENTS: EFFECTIVENESS OF ENCAPSULATION
Jian-Er Lin
Technical Resources, Inc.
Gulf Breeze, FL
and
James G. Mueller
SBP Technologies, Inc.
Gulf Breeze, FL
and
P. Hap Pritchard
Gulf Breeze Environmental Research Laboratory
U. S. Environmental Protection Agency
Gulf Breeze, FL
INTRODUCTION
Extensive studies have resulted in identification of many microorganisms with unique
degradative ability. The potential for inoculating contaminated soils and sediments with the
identified microorganisms to effect the degradation of pollutants has been proposed (1). If
inoculation is defined as the process of introducing microorganisms into a site in which the
inoculated microorganisms survive 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 microbial communities to degrade toxic wastes under environmental 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 will undoubtedly include inoculation
procedures to introduce unique and specialized metabolic capacities into a contaminated matrix.
Furthermore, inoculation also is important in enhancing the consistency and controllability of a
bioremediation process.
-86-
-------
Effectiveness of microbial inoculation in soils and sediments for bioremediation and pollution
control may be determined by the following factors: 1) concentration of active inoculants, 2)
interaction between added microorganisms and indigenous populations, 3) nutrient (including
electron acceptor) supplies to the target microorganism, 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 inoculation practice.
For this study, the use of cell encapsulation technologies was proposed to overcome some
of the existing difficulties associated with inoculation for bioremediation. To understand the effect
of encapsulated microbial inoculants on a biodegradation process, several encapsulation technologies
were established or evaluated. The use of these 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 developed for this study. They 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, vermiculite formulation, and polyurethane pellets, representing controlled-
release, solid-particle-supported, and polymer-immobilized inoculants, respectively. Four microbial
strains—strain CRE 7, Pseudomonas paucimobilis EPA 505, Mycobacteriutn sp. PYR-1, and
Alcaligenes eutrophus AEO 106 (pRO 101)—were encapsulated with these technologies. Various
additives, such as adsorbents, solid nutrients, and densification 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-state remediation (landfarming or composting), and a washing-water
bioreactor system that used the developed encapsulated microorganisms. Three kinds of
encapsulated inoculants were used in different treatment processes depending on the following
factors: 1) solubility of target compounds, 2) treatment purposes (remediation 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 microorganisms: 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 contaminated sites, and 3) degradation kinetics in conjunction
with a treatment system. The treated samples included phenanthrene, fluoranthene, creosote, and
2,4-dichlorophenoxyacetic acid-contaminated soils.
Experiments showed that the encapsulation technologies maintained a high viability of
inoculants for at lease 2 months, while the non-encapsulated inoculants lost their viability within 2
to 3 weeks. When the encapsulated inoculants were used for biodegradation in various test systems,
they resulted in the degradation of target compounds. Moreover, polymer immobilization of cells
could facilitate the reuse of inoculants 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-encapsulated
inoculants.
-87-
-------
So § "8 1
£2. "5 •- _2
te -TJ m~ ® "5
go
1
O
X
o
H
_o
*-t
J5
3
CX5
CX
S
c
w
T3
O
ex
_o
"cu
. o> •£ c -o c TS
E^ "B E '•$; « > u
= -ft 0 0-5 CD '-5
"-0-D £•£ Q. £
— .^ C C C ,.
O 0 J2 0 0 n ?"
w v) >— • u u *-£- ^1
1 1 (
^u~> -c-^ _
S'o. en D. E /-)
-g LU CD oo O)1-'
C """^ ^ D
CD "^ QJ k_ -y- ^3
^ •£ E ^E •§ < ~a
U § § § -a > J
'5 * Q_ § U 2 *
"*~ Q- ' *^~ "^ T" ^j
t/^ ^ ' C^ -*-^ ^^ -^^- ^^
CD
1-2 | "§ §
•^ 'w M- D D '•£
"S o S"i o i?
2* 8 J| ^
CD C m "E CD C
'E — *~ -2 E •-
O 3 ~D D CD ''t:
U U CD U •- —
S JI.IJ1
' '
c
•2 -o
-^ -K U
O Do
.w ~| ^
Q) n
| 'CD ^ ^f" «
"£ CD "n "^
£ ~a CD o .2 o
O C ^ U > p
V) ° O oo c C
-?» =5 i £ CN
!-§ ^^1 A
O o O "O 2 n
u oo u D E Ji
1 1 1
~o
_C oo
_y "5
O oo^
11
II
o
0 »
"
CD
"i
- D)
W
Q
4
u c
si
o o oo
V) V) >
CD
o
_D
D
CD
E
D
U
CD 0
•"= C
0)
-D
_0
V)
.^
i_
n
nX
'[/I
^
0
*
u
0
'CD
u
3
C
00
E
*- ° 2 o
o E § _ 2 E
9-^ :l ^ B CN
*
^ «
CD .-fc
>
-n — -t: ° CD 7= i-
ill-ill
-D aj =" J2 '*: oo 3
2 E -S S -2 .S o
CD S ^ -2 C - "
~i "2- S- 3 ,_
g a. Q. o t
Q- « « .
CD
C D-
. rs E
.£ -2 E J
3J2
II
I i'
O) O
c _«
o T u
c oo Jf
•- a £
OO J O
u "6 ~o
D oo C
cu _ 3
o ±E 2
la £ ro
O 0)
i^ -a
< 2
Q- D)
LU CD
"5 CU
-Q C
o cc
E -£
aJ ^
c
_o
n
u
0
c
CD
CD
V)
D
70
Q
'Q. "c
o
"~
ntarninant:
- repeatedly
- remove co
-o
c
O
VI
"E
_D
3
u
o
^c
CD
_N
15
0
E
E
oo
c
CD
-t—
3
^C_
(/)
CD
>
•^
-D
n
CD
D
a.
_CD
15
0
E
CD
-E
J
1
O
u
t/)
c
0)
T3
-D
C
D
-W*
^-
CD
_Q
O
00
-D
n
CD
E
croenviron
'E
cu
-D
'!
5.
X
~^
_
.^
"^
5
en
a
c
.0
CD
-£
t!
_CD
_D
_g
3
oo
conditions
X
'i_
"5
E
CD
E_
"o
r*i
i-
|>
"u
D
^00
'x.
C
JO
I
CD
C
O
O CD
n_ Q.
-------
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 contaminated systems need to be investigated further. Various factors
influencing the effectiveness of encapsulated inoculants in a contaminated site need to be identified.
REFERENCES
1. Pritchard, P.H. 1992. Use of inoculation in bioremediation. Curr. Opin. in Biotech. 3:232-
243.
2. Lin, J.-E., J.G. Mueller, and P.H. Pritchard. 1992. Use of encapsulated microorganisms as
inoculants for bioremediation. Abstract Book for the ACS Symposium 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 organic compounds. Biotechnol. Bioengineer. 38:273-279.
4. Lin, J.-E., and H.Y. Wang. 1991. Degradation of pentachlorophenol by non-immobilized,
immobilized, and co-immobilized Arthrobacter cells. J. Perm. Bioengineer. 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 fermentation on nutrient-supplemented vermiculite. Appl. Environ. Microbiol.
53:2138-2140.
7. O'Reilly, K.T., and R.L. Crawford. 1989. Degradation of pentachlorophenol by
polyurethane-immobilizedF/flvofeacte/Tum cells. Appl. Environ. Microbiol. 9:2113-2118.
8. Lin, J.-E. et al. 1993. Unpublished data.
9. Lin, J.-E., J.G. Mueller, and H.P. Pritchard. 1993. Use of encapsulated microorganisms to
prevent pollution by pesticides. HMCRI National R&D Symposium.
-89-
-------
COMBINING TREATABILITY STUDIES AND SITE CHARACTERIZATION
FOR RATIONAL DESIGN OF IN SITU BIOREMEDIATION USING
NITRATE AS AN ELECTRON ACCEPTOR
S.R. Hutchins, D.H. Karapbell, 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
and
B. Newell
Robert S. Kerr Environmental Research Laboratory (ManTech)
U.S. Environmental Protection Agency
Ada, OK
and
J.A. Johnson, V. Ravi, and J.K. Rumery
Robert S. Kerr Environmental Research Laboratory (Dynamac)
U.S. Environmental Protection Agency
Ada, OK
INTRODUCTION
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 remediation at field scale is not predicted in any straightforward way from the
laboratory study. Rational design compares the residence time and concentration of electron
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
-90-
-------
INJECTION WELL
WATER
TABLE
SAND
CLAY
SAND AND GRAVEL
BEDROCK
DEPTH
(FEET)
-o -
-5 -
-10-
-15-
-20-
-25-
-30-
-35-
-40-
TOTAL PETROLEUM
HYDROCARBON
(mg / kg)
0 30006000
f
Figure 1. Relationship between the lithology, water table, and vertical interval contaminated
with oily phase hydrocarbon at a pipeline spill undergoing in situ bioremediation.
-91-
-------
vertical extent of hydrocarbon contamination. 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 been selected for an evaluation of the relative contribution of the various
components of in situ bioremediation. Nitrate has been selected as the primary electron acceptor.
The demonstration also evaluates a potential synergism between nitrate and low concentrations of
oxygen. The water circulation system isolates the site into six cells of roughly equivalent size (Figure
2). Ground water without additional electron acceptors 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 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 60O 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 ammonium and phosphate salts
as described previously (2). No exogenous carbon source was added. Each microcosm was then
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 obtained from each microcosm and the volume was replaced 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 microcosm. 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 there was no significant effect of depth interval on the rate
of removal. Further, no correlation with either BTEX concentrations or concentration of TPH was
evident.
The average nitrate removal rates were 1.47 ± 1.09 mg/L/day at borehole 60N and 2.85 ±
1.87 mg/L/day at borehole 60O. Taking 2.85 mg/L/day as an estimate of the nitrate consumption rate,
-92-
-------
Key Description
TPH Concentration
(Kghn2)
1 - 3
3 - 5
5 - 7
7 - 9
9 - 11
> 11
200 feet
100 meters
Figure 2. Relationship between treatment cells, monitoring wells, boreholes, and water
circulation wells at a pipeline spill undergoing in situ bioremediation. Open
circles are pumped wells, closed circles are passive monitoring wells. Recharge
was distributed evenly to the cells. The infiltration wells are too numerous to
display.
-93-
-------
the circulated concentrations of 8 mg/L would be consumed within 67 hours at field scale.
Table 1. Rate and Extent of Nitrate Removal in Core Material from Locations 60N and 60O
Core
Sample
60N-C
60N-B
60N-A
60N-E
60N-D
60N-H*
60N-G
60N-F
60N-J
60N-I
60N-L
60N-K
60O-E
60O-D
60O-C
600-B
60O-H
60O-G
60O-F
60O-L
60O-K
60O-J
60O-I
60O-O
60O-N
60O-M*
60O-R
60O-Q
60O-P
60O-T
60O-S
60O-V
60O-U
Depth
(ft)
16.0 - 16.4
16.4 - 17.2
17.2 - 18.0
18.0 - 19.4
19.4 - 20.0
20.0 - 21.2
21.2 - 22.0
22.0 - 22.5
22.5 - 23.6
23.6 - 25.0
25.0 - 25.4
25.4 - 26.2
8.1 - 8.6
8.6 - 9.0
9.0 - 9.8
9.8 - 10.6
11.5 - 12.3
12.3 - 13.1
13.1 - 14.0
14.0 - 14.4
14.4 - 14.8
14.8 - 15.6
15.6 - 16.5
16.8 - 17.2
17.2 - 18.0
18.0 - 19.0
19.0 - 20.0
20.0 - 21.0
21.0 - 22.0
22.0 - 23.0
23.0 - 24.0
24.0 - 24.9
24.9 - 25.3
BTEX
(rag/kg)
1.6
1.4
2.2
1.4
11.3
32.7
249
264
331
3.1
0.6
0.4
0.1
0.1
0.1
0.1
0.3
0.2
1.4
1.1
2.5
4.1
8.8
17.9
9.1
31.9
45.7
127
132
13.5
0.5
0.5
0.2
TPH
(mg/kg)
40
112
256
95
567
695
3760
3620
4530
38
8
14
18
12
14
15
18
16
57
84
245
389
1060
179
334
1090
1420
3160
3110
219
9
9
7
NCyN Re-
moval Rate
(mg/L/day)
4.37 + 0.46
1.30 + 0.13
0.83 + 0.18
1.21+0.46
1.21 + 0.23
0.50 + 0.46
2.59 + 1.39
1.05 + 0.46
1.16 + 0.07
0.63 + 0.11
2.09 + 0.09
0.69 + 0.31
0.73 + 0.06
1.06 + 0.55
1.74 + 0.01
1.59 + 0.15
2.14 + 0.42
0.15 + 0.08
4.84 + 0.77
5.61+0.54
7.06 + 0.31
5.32 + 0.43
3.53 + 0.71
3.34 + 0.91
2.11 + 0.34
2.90 + 0.37
3.47 + 0.21
1.94 + 1.15
0.32 + 0.09
0.53 + 0.03
4.69 + 0.12
3.01+0.58
2.33 + 0.56
Nitrate-N
Removed
(mg/kg)
69.5
22.8
23.1
26.9
24.8
26.0
25.3
18.8
28.1
36.8
29.3
23.9
29.5
29.1
29.3
42.5
29.3
15.6
57.7
102.3
28.9
28.4
102.3
102.3
45.5
34.5
43.2
28.7
21.5
28.9
68.9
72.8
53.1
* Water table
-94-
-------
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 appropriately 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 would be expected if flow at the 30 to 35 ft level was primarily horizontal toward
the recovery well REC-2 and the bromide 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 2.2 ft/day in cluster 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 hypothesis
that flow is primarily vertical in wells 60-G2, 60-G3, 60-M2, and 60-M3.
TRACER TEST AND COMPUTER MODELING OF HORIZONTAL FLOW
Ground water was pumped from well REC-1 and circulated 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 semianalytical model
for the delineation of wellhead protection areas (U.S. EPA, Office of Ground Water Protection,
1991).
The model was run forward in time until it approached an equilibrium. It predicts two
interesting 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 uncontaminated 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 combination of pump-and-treat and in situ bioremediation.
-95-
-------
Park City Bromide Results 5/1-17/92
G Cluster Wells
40-r
100 150 200 250 300 350 400
150 200 250 300
Hours Into Study
350 400
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 clay layer.
-96-
-------
120 day capture zone
120 day capture zone
REC-2
Figure 4. Capture zones of circulated water and uncontaminated water predicted for ground
water circulation wells.
-97-
-------
The predictions of the models were evaluated by comparing the predicted concentration of
bromide tracer discharged from REC-2 to the actual concentrations. Over a 20-day tracer test, there
was considerable agreement between the prediction of the model and the concentrations of tracer
in the ground water from the discharge well (Figure 5).
CONCLUSIONS
If nitrate survives 67 hours in water moving 2.2 ft/day, it should move at least 6,1 vertical feet
after recharge. This should be adequate to carry it below the oil contaminated layer (cf. Figure 1),
and there should be no limitation on the supply of nitrate.
REFERENCES
1. Kennedy, L., and S.R. Hutchins. 1992. Applied geologic, 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: Bioremediation of
Hazardous Wastes. Office of Research and Development. U.S. Environmental Protection
Agency. EPA/600/R-92/126.
-98-
-------
25 -r
10
15
TIME (days)
20
25
30
Figure 5. Correspondence between recovery of bromide tracer in circulation well REC-2
and the predictions of the model.
-99-
-------
RATE AND EXTENT OF NATURAL ANAEROBIC BIOREMEDIATION
OF BTEX COMPOUNDS IN GROUND WATER PLUMES
Morton A. Barlaz, Michael B. Shafer, Robert C. Borden
North Carolina State University
Raleigh, NC
and
John T. Wilson
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
Ada, OK
INTRODUCTION
A combined field and laboratory study has been conducted 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 smelled 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 hydrogeologjc
characterization, a soil gas survey, 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 experiments were conducted to determine
the anaerobic biodegradation 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 due to 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
-100-
-------
BTEX, methane, nitrate, 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-
dimethylpentane (DMP). BMP 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 analysis are shown in Table 1. The rates vary by about
threefold. There was no apparent biodegradation of benzene under the anaerobic conditions present
in this plume.
MICROCOSM STUDIES
Microcosm studies were conducted using aquifer material collected 0, 30, and 70 ft from the
source to evaluate the extent of biodegradation under ambient conditions. The aquifer material was
collected aseptically and anaerobically using methods developed by Robert S. Kerr Environmental
Research Laboratory personnel. Ground water was collected anaerobically through a 0.45-^m 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 closely 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 microcosms in concentrations similar to in situ
conditions. This was 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 sodium sulfide and resazurin to ensure reducing conditions. 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 there was a rapid increase in methane 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 there was a rapid decline to less than 20
by day 192. After this rapid drop, toluene concentrations plateaued at between 5 and 15 jig/L with
no additional removal. Biotransformation of the m+p-xylene, o-xylene, andethylbenzene was similar
although it lagged the toluene results somewhat. The lag in ethylbenzene and xylene removal
coincided with a drop in toluene below 20 ^g/L, suggesting that biotransformation is sequential, with
toluene being the first to degrade. There was no apparent removal of benzene in any of the
microcosms.
First-order removal rates for toluene, ethylbenzene, and xylene isomers are shown in Table
-101-
-------
Table 1. First-Order Decay Rates (d'1) Estimated from Field Data
30 Feet 70 Feet
Benzene 0.002 to 0.004 -0.004 to 0.002
Toluene 0.053 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
-102-
-------
20.00
Time (days)
Figure 1. Biotransformation of toluene and production of methane in laboratory microcosms
from 30 ft location (stars - live, crosses - abiotic).
-103-
-------
2. Toluene biodegradation 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. This may be due to the lower initial
concentrations present in these microcosms. In the 0 ft microcosms, toluene is being removed but
has not yet been reduced to below 20 pg/L. In these microcosms, ethylbenzene and xylene
biodegradation has been significantly slower.
COMPARISON OF FIELD AND LABORATORY RESULTS
Both the field monitoring and laboratory microcosm results indicate that benzene is not
biotransformed under anaerobic conditions at this site. Toluene, ethylbenzene, and the xylene
isomers were all transformed under anaerobic conditions in the field and laboratory microcosms.
The match between field and laboratory rates is not exact but is good considering the inherent
differences in the techniques employed. Laboratory biotransformation rates were typically a factor
of 2 to 10 lower than the field rates. This 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 material, this would result in a lower microorganism-to-water ratio and
lower biotransformation rate. Also, since the data from the lag and plateau periods were included
in our calculations, the estimated rates would be lower.
-104-
-------
Table 2. First-Order Biotransformation Rates from Sleeping Bear Microcosms (d"1)
Compound 0 ft 30 ft 70 ft Average
Toluene-Live 0.0108 0.0111 0.0122
Toluene - Abiotic 0.0021 0.0012 0.0016
Toluene-Net 0.0087 0.0099 0.0106 0.0098
Ethylbenzene - Live 0.0037 0.0036 0.0103
Ethylbenzene - Abiotic 0.0034 0.0026 0.0029
Ethylbenzene - Net 0.0003 0.0010 0.0074 0.0029
m&p-Xylene - Live 0.0042 0.0040 0.0041
m&p-Xylene - Abiotic 0.0039 0.0030 0.0035
m&p-Xylene - Net 0.0003 0.0010 0.0006 0.0006
o-Xylene - Live 0.0036 0.0037 0.0033
o-Xylene - Abiotic 0.0035 0.0026 0.0028
o-Xylene - Net 0.0001 0.0011 0.0005 0.0006
-105-
-------
PILOT-SCALE RESEARCH
-------
TREATMENT OF TCP-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
and
Steven I. Safferraan
U.S. Environmental Protection Agency
Cincinnati, OH
Pentachlorophenol (PCP) has been used as a wood-preserving compound since the 1930s.
PCP-contaminated soils can be found at abandoned and existing wood preserving sites, and at PCP
manufacturing facilities that are currently Resource Conservation and Recovery Act (RCRA) and
Superfund sites (1,2,3). PCP is considered a priority pollutant by the EPA, and its removal 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 effectiveness 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 particle 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-contaminated
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
-109-
-------
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-contaminated soil.
Expanded-bed granular activated carbon (GAC) anaerobic 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 conducted 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 that 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 percent ethanol solutions achieved higher PCP removal efficiencies than the
other ethanol solutions. Similar results 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 ethanol solutions is not completely attributable 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 ethanol/water mixtures obtained greater PCP recoveries
from 20 x 40 U.S. mesh soil at higher soil-mass-to-solvent volume ratios. Similar results were
obtained for the 100 x 140 U.S. mesh soil and the clay fraction.
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 occurred 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 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 compounds, 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 commenced after attaining stable reactor
operation. The experimental data demonstrated an incomplete mineralization of pentachlorophenol
to carbon dioxide and methane, and an equimolar conversion of PCP into monochlorophenols. This
-110-
-------
00
B
I)
^*
a.
a.
i ' i
O DI Water
• 25% Ethanol
50% Ethanol
T 75% Ethanol
D 95% Ethanol
• 100% Ethanol
120
Figure 1. Reverse isotherm for 100 x 140 U.S. mesh soil.
o
a>
OS
a,
o
100
80
40
20
1
Soil:Solvent Ratio (g/mL):
V 0.2
• 0.75
O 1.5
1
20 40 60
Ethanol in Solvent
80
100
Figure 2. PCP recovery from 20 x 40 U.S. mesh soil at high soihsolvent ratios in batch
extraction tests.
-Ill-
-------
~
^^
"e
o
V
OS
ft.
O
0.
1UU
90
60
70
60
50
40
30
20
10
0
1 1 1 1 1 1 \ / / 1 1 1 1 1 1 1
- * • • * « * -
^b o • « '
3
_
~ Solvent: 75% Ethanol
Solvent:Soil Ratio (g/mL): 0.01 _
O 12-hour test
• 21 -day test
-
i i i i i i i // i i i i i i i
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 369 12 15 18 21 24
Time (days)
Figure 3. PCP removal from 20 x 40 U.S. mesh soil with time.
-112-
-------
indicates that the PCP feed rate was the critical control parameter that effected reactor toxicity via
the accumulation and inhibitory role of p-CP (Figure 4).
REFERENCES
1. U.S. EPA. 1986. Superfund Record of Decision (EPA Region 6) United Creosoting Co.,
Hilbig Rd., Conroe, TX. EPA/ROD/R06-86/D14. U.S. Environmental Protection 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 Protection Agency.
3. U.S. EPA. 1987. Superfund Record of Decision (EPA Region 4) Palmetto Wood
Preserving, Dixiana, Lexington County, SC. (1st remedial action) EPA/ROD/R04-87/026.
U.S. Environmental Protection Agency.
4. Mueller, J.G., et al. 1988. Preliminary study of treatment of contaminated groundwater from
the Taylorville gasities site. HWRIC RR 077.
5. Peters, C.A., and R.G. Luthy. 1991. Coal tar dissolution 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.
-113-
-------
o
6
6
c
o
IlJ
05
I-,
V
o
a
o
o
10
10'
10"
10
10~
10
- - Influent PCP
Effluent PCP (no biological activity)
• Effluent PCP (actual)
O Effluent WCPi (actual)
Effluent DCPt (actual)
V Effluent TCP> (actual)
100
zoo
300
Days
•»oo
soo
Figure 4. Effluent quality of chlorophenols in reactor A.
-114-
-------
PRELIMINARY EVALUATION OF ATTACHMENT MEDIA
FOR GAS PHASE BIOFILTERS
Francis L. Smith, George A. Serial, Paul J. Smith, Makram T. Suidan, and Pratim Biswas
Department of Civil and Environmental Engineering
University of Cincinnati
Cincinnati, OH
and
Richard C. Brenner
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH
INTRODUCTION
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 practice. 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 incineration, biofiltration can
be more cost effective, particularly for treatment of large volumes of air with low concentrations of
biodegradable VOCs (2). The low annualized cost is due to the utilization of ambient temperature
microbial oxidation, rather than oxidation by thermal or chemical means. Essentially, biofiltration
utilizes 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 microbes then oxidize the VOCs to simple 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 performed utilizing trickle bed biofilters with
monolithic channelized (as well as pelletized) 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 investigate the use of such biofilters
-115-
-------
for treating chlorinated compounds, while achieving high removal efficiencies at high feed
concentrations. Such compounds can be encountered 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 containing air streams.
EXPERIMENTAL APPARATUS
The trickle bed biofilter apparatus used in this study consists of three independent, parallel
trains, each containing 4 ft of attachment media: Biofilters A, B, and D. Biofilters A and B are
filled with Corning Celcor channelized media, and D is filled with Manville Celite pelletized 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 biofilters. The air is mass flow controlled, and 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. Biofilters A and B are also equipped with effluent recirculation in order to
provide even distribution of the biomass throughout the Corning Celcor channelized medium.
Biofilter B is operated in a countercurrent mode (i.e., with the air flow fed to the bottom). Biofilters
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 L/day. Each biofilter was maintained at a constant temperature of 14
±2 °C and 7.7 ±0.2 pH range.
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 recycled 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 residence 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 minutes 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 toluene 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.
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 compared 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
-116-
-------
600
400
>
6
P,
ex
o 200
fi
(U
o
C
o
o
C
0)
"o
K
Figure 1. Biofilter A performance w.r.t. toluene removal during a residence time cycle.
-117-
-------
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 performance of the
biofilter during the residence time cycle test. Note that toluene removal during the cycle test
stabilized 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 conducted in a manner similar to that for Biofilter A. Figure 3 shows the
performance of the biofilter during the residence time cycle test. Note that toluene removal during
the cycle test stabilized above 99 percent up to 2 minutes of residence time, then 95 percent at 1
minute. The pressure drop ranged from 4 in. water at 12 minutes to 22 in. water at a 1-minute
residence time.
CONCLUSIONS 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 the VOC concentration at constant mass loading, more time is required by the
biofilter to achieve maximum efficiency. Effluent recycle, to establish even distribution of biomass
throughout the media, was necessary to achieve maximum efficiency with the channelized support
media.
Biofilter D showed the highest VOC removal efficiencies. The efficiencies ranged from 99+
percent to 95 percent for residence times of 12 to 1 minute. Biofilter 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, Biofilter B, operated countercurrently, showed efficiencies less than
90 percent of from 90 percent 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 biofilter design (i.e., blower motor power) will
be lower than for most typical biofilter designs. 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 biofilter. The effect of increasing the mass loading to 500 ppmv at lower
residence times will also be investigated. After finishing the characterization of the biofilters with
toluene, feeding the other VOCs will be initiated.
-118-
-------
600
50° F-
g
ft
ft
400 f-
d 300
o
6
01
K
0)
a
a>
40 5
o
E-
60
20
6 B
Retention Time, min
10
12
14
Figure 2. Biofilter B performance w.r.t. toluene removal during a residence time cycle.
600
., 400
B
ex
a,
C 200
o
a
0)
o
C
o
O
-------
REFERENCES
1. Lee, B. 1991. Highlights of the Clean Air Act Amendments of 1990. J. Air Waste Manag.
Assoc. 41(1):16.
2. Ottengraf, S.P.P. 1986. Exhaust gas purification. In: Rehn, HJ., 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 vapors in biofilters. Environ. Technol. 12:655.
4. Leson, G., F. Tabatabal, and A.M. Winer. 1992. Control of hazardous and toxic air
emissions by biofiltration. Presented at the 85th Annual Meeting and Exhibition of the Air
and Waste Management Association, Kansas City, Missouri, June 21-26.
5. Van Langenhove, J., A. Lootens, and N. Schamp. 1989. Inhibitory effects of SO2 on
biofiltration of aldehydes. 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.
-120-
-------
PROCESS RESEARCH
-------
METHANOGENIC DEGRADATION KINETICS OF NITROGEN AND SULFUR CONTAINING
HETEROCYCLIC AROMATIC COMPOUNDS IN AQUIFER-DERIVED MICROCOSMS
E. Michael Godsy and Donald F. Goerlitz
U.S. Geological Survey
Menlo Park, CA
and
Dunja Grbic'-Galic'
Department of Civil Engineering
Stanford University
Stanford, CA
INTRODUCTION
The fate of nitrogen and sulfur containing heterocyclic aromatic compounds in subsurface
environments is controlled by various transport and transformation processes. The most important,
but least understood, process 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 mixtures (e.g., creosote).
This study presents the Monod-decay kinetics for the conversions of benzothiophene,
oxindole, 2(lH)-quinolinone, and l(2H)-isoquinolinone to methane and carbon dioxide by a complex
aquifer-derived methanogenic microbial consortium. The stoichiometric oxidation of indole,
quinoline, and isoquinoline to the persistent compounds oxindole, 2(lH)-quinolinone, and 1(2H)-
isoquinolinone, 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-preserving 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
moisture from the poles, cellular debris, creosote, PCP, and diesel fuel from the treatment processes,
-123-
-------
were discharged to unlined surface impoundments that were in direct hydraulic contact with the
sand-and-gravel aquifer.
To model the degradation kinetics, laboratory microcosms were prepared in 4-L glass bottles.
The microcosms contained approximately 3 kg of aquifer material anaerobically collected from the
approximate centroid of the active methanogenic zone (1). Compounds of interest were added to
a mineral salts solution (1) at concentrations 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 Galli (3).
Substrate depletion curves were fitted to the Monod substrate utilization and growth with
decay equations (4):
_dS m
dt Y(KS + S)
where:
/im = maximum specific growth rate, day1
Ks = half-saturation constant, mg L"1
Y = yield coefficient, mg cells per mg substrate utilized
5 = 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 determination of parameter values (/im, K,,
and kd) that best fit the experimental substrate depletion data by minimizing the residual sum of
squares using NLR (non-linear regression) techniques.
RESULTS AND DISCUSSION
The bacterial substrate utilization for all of the compounds 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 benzothiophene, 87.6 percent for oxindole, 91.7 percent for
2(lH)-quinolinone, and 88.5 percent for l(2H)-isoquinolinone. A stable oxidized intermediate
similar to oxindole or 2(lH)-quinolinone that would be formed from the oxidation of
benzothiophene to 2-hydroxybenzothiophene was not observed in this study or in a previous study
(6).
-124-
-------
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 the 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). It can
be reasoned that 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 microcosms, where indole, quinoline, or
isoquinoline are the only major carbon and energy sources, the complete oxidation of the parent
compound may supply necessary reducing power required for the subsequent ring reduction and ring
cleavage.
Monod-decay kinetic constants for the anaerobic degradation of sulfur and oxidized nitrogen
containing heterocyclic 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,
there is no statistical test for this hypothesis 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 phenolic and nitrogen and sulfur heterocyclic compounds, suggesting that
acetate utilization was a rate-limiting step. This limitation may explain why the values for nm and
Ks are essentially the same for each of the compounds tested. The value of the kinetic parameters
agree with previously published values for acetate utilization by well-acclimated acetate-fed
enrichment cultures of methanogenic bacteria (10).
The apparent low biomass yield suggests that the microbial community from this ground
water environment has adapted to low nutrient conditions by utilizing 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 capturing 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). Experimental and
theoretical values of Y for the compounds tested are given in Table 4. The theoretical values are
greater than the values obtained from the measured protein increase in the microcosms by
approximately an order of magnitude.
The bacterial decay term (-kjX,) in the biomass equation is apparently not required to
describe substrate utilization and/or biomass increase in the batch growth microcosms. The values
determined by NLR are such that /im» kd, and as a result kd 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.
-125-
-------
Table 1. Gibbs Free Energy Changes During Oxidation of Nitrogen Compounds at 25*C
and pH of 7
Indole to Oxindole
C8H7N + H2O -» CgH7NO + 2 H+ + 2 e" 168
Quinoline to 2(lH)-Quinolinone
O>H7N + H2O -» C9H7NO + 2 H* + 2 e" 84
Isoquinoline to l(2H)-Isoquinolinone
CoH7N + H2O -» CQH7NO + 2 H+ + 2 e' 56
calculated using values from Stull et al. (7) or estimated using the method of Jobak in
Reid et al. (8) for compounds not found in the literature.
Table 2. Gibbs Free Energy Changes During Methanogenesis of NSO Compounds at 25°C
and pH of 7
Indole to CH4 and CO2
C8H7N + 7 H2O + H+ -> 3.5 CO2 + 4.5 CH4 + NH* ' 205
Quinoline to CH4 and CO2
CgH7N + 8 H2O + H+ -* 4 CO2 + 5 CH4 + NH^ - 245
Isoquinoline to C7fy and CO2
8 H2O + H+ -> 4 CO2 + 5 CH4 + NH -231
Benzothiophene to CH4 and CO2
+ 7 H20 -> 3.5 CO2 + 4.5 CH4 + 0.5 H2S + 0.5 HS" + 0.5 H+ - 185
calculated using values from Stull et al. (7) or estimated using the method of Jobak in
Reid et al. (8) for compounds not found in the literature.
Table 3. Kinetic Constants Determined for Methanogenesis of the Heterocyclic
Compounds Tested ± 95% Confidence Intervals
Benzothiophene
Oxindole
2(lH)-Quinolinone
1 (2H)-Isoquinolinone
\im (day1)
0.11710.136
0.16010.002
0.08910.058
0.09910.342
Ks (mg L'1)
0.80 1 5.06
1.1012.10
11.4110.06
5.00122.0
kd (day1) ]
0.00010.129
0.00010.012
0.00010.049
0.00010.027
rdngmg-1)'
0.025
0.029
0.033
0.035
Values for Y are determined from average of three protein determinations.
-126-
-------
Table 4. Experimental and Theoretical Y Values for the Compounds Tested in Microcosms
Compound Experimental Y (mg mg~ ' )"
Benzothiophene
Oxindole
2(lH)-Quinolinone
l(2H)-Isoquinolinone
0.025
0.029
0.033
0.035
Theoretical Y (mg rag'1)
0.14
0.35
0.21
0.21
* Based on the average of three protein determinations each before and after growth.
-127-
-------
REFERENCES
1. Godsy, E.M., D.F. Goerlitz, and D. Grbic'-Galic'. 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 protein 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. Grbic'-Galic'. 1989. Biodegradation pathways for benzothiophene in
methanogenic microcosms, pp. 559-564. In: Mallard, G.E, and S.E. Ragone eds., U.S.
Geological Survey Toxic Substances Hydrology Program—Proceedings 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 compounds. Marcel Dekker, New York.
8. Reid, R.C., J.M. Prausnitz, 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 regression analysis and its applications. Wiley
Interscience, New York.
10. Wang, Y.T., H.D. Gabbard, and P.C. Pai. 1992. Inhibition 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.
-128-
-------
ANAEROBIC DEGRADATION OF HALOGENATED AND
NONHALOGENATED PHENOLIC COMPOUNDS
M.M. Haggblom, M.D. Rivera, and L.Y. Young
Center for Agricultural Molecular Biology
Rutgers University
New Brunswick, NJ
and
I.E. Rogers
U.S. Environmental Protection Agency
Athens, GA
Pentachlorophenol (PCP) and creosote have been widely used as wood preservatives.
Improper use and accidental spills can cause serious contamination of soil and water at wood-
preserving sites. Approximately 10 percent 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 between the anaerobic dechlorination of PCP and the biodegradation of phenol or
methylphenols (ortho-, meta-, and/?ara-cresol) were examined. Methanogenic enrichment cultures
were established and fed phenol or one of the methylphenols at a concentration of 500 /*M (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 compounds. The source of inoculum was
sediment from the Hudson River, mile point 145, near Albany (3).
Our results indicate that PCP was reductively dechlorinated via 2,3,4,5-tetra- and 3,4,5-
trichlorophenol to 3,5-dichlorophenol within 60 to 100 days, with 3,4-dichlorophenol observed as a
minor product. These two dichlorophenols were not substantially dechlorinated further within 160
days. The addition of propionate, phenol, or cresol did not stimulate dechlorination, and thus does
not appear to serve as an electron donor for reductive dechlorination.
In the absence of PCP, phenol, m-cresol and/?-cresol (500 /iM) were utilized within 30 days,
while degradation of o-cresol took over 60 days. Methane was produced in excess of background
controls, indicating that the phenolics were mineralized. On the other hand, the presence of 50 /*M
of PCP completely inhibited degradation of the phenolic compounds and methanogenesis.
-129-
-------
Degradation of phenol and the cresols took place only after PCP had been dechlorinated to 3,5-
dichlorophenol. 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
/xM to active phenol-degrading cultures. PCP and the tetra- and trichlorophenols at a concentration
of 10 /iM and above exhibited toxicity, while 10 to 25 /*M of 3,5-dichlorophenol did not inhibit
degradation of phenol. Therefore, high concentrations of PCP might inhibit the anaerobic
degradation of creosote-derived phenolic compounds in ground water and soil.
REFERENCES
1. Haggblom, M.M., M.D. Rivera, I.D. Bossert, J. Rogers, and L.Y. Young. 1990. Anaerobic
degradation ofpara-cresol under three reducing conditions. Microbial Ecology 20:141-150.
2. Haggblom, M.M., M.D. Rivera, and L.Y. Young. 1993. Effects of auxiliary carbon sources
and electron acceptors 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 andbenzoic acids. Appl.
Environ. Microbiol. 59. (In press)
-130-
-------
ANAEROBIC BIODEGRADATION OF 5-CHLOROVANILLATE AS A
MODEL SUBSTRATE FOR THE BIOREMEDIATION 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 investigated. 5CV was selected as a model compound for studying the
biodegradation of paper-milling effluents because it contains the methoxy, chloro, and carboxyl side
groups representative of those present on aromatic chlorinated compounds released in paper-milling
effluent. 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 5CV degradation in
this enrichment culture was concluded to be stepwise demethoxylation to 5-chloroprotocatechuate
(5CP; 5-chloro-3,4-dihydroxybenzoic acid), decarboxylation to 3-chlorocatechol (3CC; 3-chloro-l,2-
dihydroxybenzene), and dechlorination to catechol, which was completely degraded (Figure 1).
Dechlorination of 3CC was the rate-limiting step of degradation (Figure 2).
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-
chlorobenzoate (3CB). This co-culture had been previously shown to sequentially dechlorinate and
then demethoxylate 5CV to form protocatechuate, which was not degraded further. Upon formation
of protocatechuate, this culture was inoculated (indicated with an arrow in Figure 3) with a vanillate-
degrading bacterial consortium that had been derived from the above-mentioned 5CV enrichment
and is known to perform the degradative steps downstream from dechlorination (i.e., sequential
demethoxylation to protocatechuate, decarboxylation to catechol, and complete degradation of
catechol). As expected, protocatechuate was decarboxylated to catechol, which was then rapidly
degraded. Under these conditions, 5CV was completely degraded without formation of 3CC, as
predicted (Figure 4). The time required for complete dechlorination increased, however, from 4 to
17 weeks. Subsequent transformation 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 experiment with the vanillate consortium as soon as dechlorination is complete might
reduce complete degradation of 5CV to less than 3 months—something that is currently under
investigation.
-131-
-------
o
4-
I
-
S
e
=9
<
a
O
£
T3
'§
O
I
•g
.O
*4-»
•S
T3
-------
.C
u
I
V)
§
•3
•3
«
ab
•8
o
1
-------
13
1
I
s M
<•••» *n
I
c
I
I
§
•8
o
UOI^DJ^Uo
-134-
-------
CM
CD
o>
o
o
O
0)
LLI
O
*••
=
I
I
c
•«
-a
I
J?
M
i
o
I
.
*
bfi
•8
-------
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 performing the individual transformation steps. The 3CB co-
culture was inoculated into 5CP and vanillate medium 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 demethoxylation,
protocatechuate medium to enrich for the decarboxylation, and catechol medium to enrich for the
catechol degradation. Transformation of target compounds 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 decarboxylation take 3
and 4 days, respectively.
While studying the capacity of the co-culture to dechlorinate 3CB, a sulfate-reducing
bacterium was isolated and identified as a new species, Desulfomicrobium escambium. This
bacterium is responsible for dechlorination of 3CB and is currently 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 effluents. An appropriate toxicity test currently is being devised to assess the
toxicity of various untreated paper-milling effluents. Effluents found to be toxic will be
"bioremediated" by inoculating culture media that contains increasing percentages of effluent with
the bacterial cultures developed in these studies. Cultures will be monitored using HPLC analyses
to determine biotransformation and biodegradation of the aromatic compounds in the effluents.
Toxicity tests will be repeated on filter-sterilized supernatant from these cultures to determine the
effect of this treatment on toxicity.
-136-
-------
BIOREMEDIATION OF SOILS AND SEDIMENTS
CONTAMINATED WITH AROMATIC AMINES
Eric J. Weber
Environmental Research Laboratory
U.S. Environmental Protection Agency
Athens, GA
and
David L. Spidle
Technology Applications, Inc.
Athens, GA
and
Kevin A. Thorn
U.S. Geological Survey
Arvada, CO
INTRODUCTION
Aromatic amines comprise an important class of environmental contaminants. Concern over
their environmental 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, munitions, and
numerous herbicides, including the phenylureas, phenylcarbamates, and acylanilides. Because these
chemicals are synthesized from aromatic amines, loss of aromatic amines to the environment may
also 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 14C-aniline
and 15N-NMR analysis of dissolved organic matter treated with 15N-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
residue is less bioavailable and less mobile than the parent compound. Thus, procedures that
enhance the irreversible binding of aromatic amines to soil constituents could potentially serve as
a remediation technique (2).
-137-
-------
The general approach to bioremediating soils and sediments contaminated with aromatic
amines is to utilize the naturally occurring phenol oxidases in soil that are thought to play a
significant role in catalyzing the formation of bound residues (3). It is thought that the binding
capacity of sediments and soils for aromatic amines can be increased by the activation of indigenous
enzymes (e.g., peroxidase) through the addition of hydrogen peroxide and readily oxidizable
substrates such as phenolic compounds. Oxidation of the phenolic substrates by activated 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 sediments that were collected in the
Athens, Georgia, area. Concentration-dependent studies with aniline in sediment-water slurries (5
percent solids) demonstrated that, at low aniline concentrations (1 x 10"* M), approximately 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 attempt 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 48 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 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 15N-aniline to fulvic acid, INEPT and
ACOUSTIC 15N-NMR spectra exhibited resonances that indicated the formation of imine, anilide,
anilino-quinone, and anilino-hydroquinone covalent linkages. Although 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 enzymes. 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.
-138-
-------
ANAEROBIC BIOTRANSFORMATION OF MUNITIONS WASTES
Deborah J. Roberts
Department of Civil and Environmental Engineering
University of Houston
Houston, TX
and
Stephen Funk, Don L. Crawford, and Ronald L. Crawford
Center for Hazardous Waste Remediation Research
University of Idaho
Moscow, ID
INTRODUCTION
2,4,6-Trinitrotoluene (TNT), hexahydro-l,3,5-trinitro-l,3,5-triazine (RDX), and
octahydro-l,3,5,7-tetranitro-l,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 practiced, large quantities of soils
contaminated in this manner still must be cleaned up. Incineration is the only proven technology
available for the remediation of soils contaminated with explosives. Unfortunately, this technology
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 nitroaromatic compound dinoseb
(2-sec-butyl-4,6-dinitrophenol) (6,7,12) to a soil contaminated with 12,000 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 reviewed by Kaplan (8) shows that under both aerobic and anaerobic
culture conditions the initial step in the metabolism of nitroaromatic compounds is typically a
reduction of the nitro constituents to amino groups. This reduction usually proceeds in steps with
-139-
-------
the para-nitro group being the most susceptible to reduction. The reduction 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 process can lead to the formation of azo or azoxy linkages with other intermediates,
followed by dimerization or polymerization (2). Under anaerobic conditions, the reductions occur
more rapidly and the hydroxylamino intermediates do not accumulate or have as much opportunity
to form linkages. The production of diaminonitrotoluenes 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 organisms but more toxic to other
organisms. Complete remediation 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 14CO2 in aqueous culture (3,4), complete inhibition 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 nitroaromatic-contaminated soils to the remediation of munitions-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-sec-butyl-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 reductive stage in which TNT is
reduced to its amino derivatives as described by most researchers. The second stage begins after the
reduction of the 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 utilized 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-diamino-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 detection limits in the cultures in as little as 24 days of incubation with no identifiable
intermediates detected, although no radiolabeled tracer studies have been performed. Visual
-140-
-------
80 n
o TNT; • RDX; v 4A26DNT; D 24DA6NT;
o intermediate A; A intermediate B;
* intermediate D; * intermediate E.
Note: TNT, RDX, and 4A26DNT are measured by
concentration, the remaining are measured by
peak area.
10000
8000
- 6000
• 4000
2000
Figure 1.
Biodegradation of TNT and TNT biotransformation products under optimized
anaerobic conditions.
-141-
-------
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 metabolism.
A summary of the optimal conditions as determined to date for the first stage of
munitions-contaminated soil remediation is presented in Table 1. Although TNT removal rates were
more rapid at pH 8, a separate experiment 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 (5, 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 also was seen in phosphate-buffered cultures receiving 50 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 optimization 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 growth of
fungi present in the soil. It appears that the toxicity of the cultures is reduced by the removal of
TNT, the amino-intermediates, and RDX. Toxicological studies with completed first stage
supernatant 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-l,3,5-trinitro-l,3,5-triazine) by Phanerochaete
chrysosporium, p. 214. In: American 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. Microbiol.
56:1666-1671.
-142-
-------
Table 1.
Optimal Conditions for the Anaerobic Remediation of Munitions-Contaminated
Soils
Condition
ph
Temp. (°C)
Buffer
N supply
Optimal
6.5
35
50 mM KH2PO4
25 mM NH4
Range
6.5-7.0
25-37
15-30
Comments
Polymerization
50 mM inhibits
at alkaline pH
TNT degradation
Table 2. Carbon Label Distribution in the Aqueous Phase* (% of T=0 (+/-sd))
Culture Conditions
Anaerobic pH 8
Anaerobic pH 6
Aerobic
Retained on 4A26DNT 24DA6NT Fractions Total
Filter^ 1-3 Recovery
56.8 (2.2)
nd
71 (10.6)
5.8
54
15.3
3.2
38.8
5.2
39.7
5.7
7.4
105.5
98.5
98.9
""Includes polar intermediates and volatile organic acids
nd = not detected
-143-
-------
5. Funk, S.B., DJ. Roberts, and R.A. Korus. 1992. Physical parameters affecting the
anaerobic degradation of TNT in munitions-contaminated soil, Abstract Q142. American
Society for Microbiology, 92nd General Meeting, New Orleans.
6. Kaake, R.H., DJ. Roberts, S.B. Funk, R.L. Crawford, and D. L. Crawford. 1992. On-site
anaerobic biological treatment of nitroaromatic-contaminated soils. 18th Annual RREL
Research Symposium, Cincinnati, OH.
7. Kaake, R.H., DJ. Roberts, T.O. Stevens, R.L. Crawford, and D.L. Crawford. 1992.
Bioremediation of soils contaminated with 2-,sec-butyl-4,6-dinitrophenol (dinoseb). Appl.
Environ. Microbiol. 58:1683-1689.
8. Kaplan, D.L. 1990. Biotransformation 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 explosives and chemical agents. Curr. Opin.
Biotechnol. 3:253-260.
10. Kaplan, D.L., and A.M. Kaplan. 1982. 2,4,6-Trinitrotoluene-surfactant complexes:
Decomposition, mutagenicity, and soil leaching studies. Environ. Sci. Technol. 16:566-571.
11. Roberts, DJ., S.B. Funk, and R.A. Korus. 1992. Intermediary metabolism during
anaerobic degradation of TNT from munitions-contaminated soil, Abstract Q136.
American Society for Microbiology, 92nd General Meeting, New Orleans.
12. Roberts, DJ., R.H. Kaake, S.B. Funk, D.L. Crawford, and R.L. Crawford. 1993.
Field-scale anaerobic 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 fungus
Phanerochaete chrysosporium: Influence of TNT concentration. Appl. Environ. Microbiol.
58:3199-3202.
-144-
-------
COMETABOLIC BIODEGRADATION OF 2,4-DINITROTOLUENE
USING ETHANOL AS A PRIMARY SUBSTRATE
Jiayang Cheng and Makram T. Suidan
University of Cincinnati
Cincinnati, OH
and
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 commonly 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 treatment processes.
In this study, the cometabolic biodegradation process of 2,4-DNT with ethanol as the primary
substrate under anaerobic conditions was investigated using an anaerobic respirometer. The
inoculum for the respirometer was cultivated in a continuously stirred tank reactor (CSTR) fed with
2,4-DNT and ethanol. The pH and the temperature of the CSTR were kept at 7.2 and 35°C,
respectively. The detention time in the 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. Biological methane potential (BMP) tests with 2,4-DNT and ethanol 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 ethanol 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.
-145-
-------
The presence of 2,4-DNT also inhibits the acidogenesis of ethanol. The transformation of
ethanol can be expressed as,
CH3CH2OH+H20=CH3COO-+H++H2 AG°=2.3few/ (D
Figure 2 shows that propionate and methane were formed in the reactors where ethanol was fed as
the sole substrate or when low initial 2,4-DNT was present. These transformations are expressed
as,
4H2+HCO;+H+=CH4+3H2O
(3)
Almost no propionate was formed, however, with high initial 2,4-DNT concentrations (Figure
3).
Propionate was degraded to methane and carbon dioxide via acetate after all the previously
formed acetate was completely transformed to methane and carbon dioxide.
CH3CH2COO-+3H2O=CH3COO-+3H2+H++HCO3 AG°=18.3fexz/ (4)
CH3COO~+H2O=CH4+HCO;
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 pollutants. I. A perspective view. Environ.
Sci. Technol. 13:416-423.
2. Valli K, BJ. Brock, D.K. Joshi, and M.H. Gold. 1992. Degradation of 2,4-dinitrotoluene
by the lignin-degrading fungus Phanerochaete chrysosponum. 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-dinitrotoluene by a Pseudomonas sp. Appl. Environ. Microbiol. 57:3200-3205.
-146-
-------
0.20
mg/L
2 mg/L
4 mg/L
4 mg/L
8 mg/L
8 mg/L
16 mg/L
16 mg/L
32 mg/L
32 mg/L
DNT -A
DNT -B
DNT -A
DNT -B
DNT -A
DNT -B
DNT -A
DNT -B
DNT -A
DNT -B
0.00
100
120
140
Time. Hrs
Figure 1. Anaerobic biodegradation of 2,4-dinitrotoluene.
o
£
E
o
o
I I
o Acetate -A
• Propionate -A
Acetate -B
Propionate -B
Ethanol -A
Ethanol -B
CH4 -B
0
200 300
600
700
800
400 500
Time, Hrs
Figure 2. Ethanol biotransformation with initial concentration of 2,4-DNT of 2 mg/L.
o Ethanol -A
• Ethanol -B
9 / CH4 -A
)/J --- CH4 -B
v Acetate —A
* Propionate -A
D Acetate -B
• Propionate -B
100 200 300 400 500 600 700 800
Time, Hrs
10
JB
"o
a
E
B.
0
Figure 3. Ethanol biotransformation with initial concentration of 2,4-DNT of 16 mg/L.
-147-
-------
4. Liu, D., K. Thomson, and A.C. Anderson. 1984. Identification of nitroso compounds from
biotransformation 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 from 2,4-dinitrotoluene. Appl. Environ. Microbiol. 35:945-948.
6. Smith, D.P., and P.L. McCarty. 1989. Reduced product formation following perturbation
of ethanol- and propionate-fed methanogenic CSTRs. Biotechnology and Bioengineering
34:885-895.
-148-
-------
EFFECTS OF METALS ON ANAEROBIC TREATMENT PROCESSES
W. Jack Jones
Environmental Research Laboratory
U.S. Environmental Protection Agency
Athens, GA
and
In Chul Kong
University of Georgia
Athens, GA
INTRODUCTION
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 transformation 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
biotransformation 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 effects of heavy metals on anaerobic microbial processes on the biotransformation
-149-
-------
of organic compounds. As a model system, the studies are focusing on the toxic effects of metals on
the reductive dechlorination of chlorophenolic compounds in unadapted and chloro phenol-adapted
microbial communities from freshwater sediments.
The onset, rate, and extent of biotransformation of several mono-, di-, and trichlorophenols
were examined using unadapted and chlorophenol(CP)-adapted freshwater sediment slurries (pH
7.0) in the presence and absence of added metal solutions (CuCl2, CdClj, K^CrjO,). A summary of
the experimental results is presented in Table 1. The times required for unadapted control sediment
slurries (no added metals) to dechlorinate 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(II) and Cd(II) 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 KjC^Q^ 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(II) (100 to 200 mg/L) also caused complete inhibition of dechlorination
for all 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(II) concentrations, respectively.
Addition of CuCl2 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(II). 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 low, varying from 0 to 2 mg/L for Cu(II), 0 to 30 mg/L for Cd(II), and 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(II), Cd(II)] 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 the onset of dechlorination, but rates of dechlorination were similar. In
general, however, sediment cultures containing the higher organic 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 indicated that dechlorination was more resistant to the inhibitory 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 the result, plotted as the percentage of
dechlorination activity (DAc) versus the amended metal concentration, is presented in Figures la-lc.
The DAc is the ratio of the times required for a 50 percent reduction (Tso) in the initial
concentration of the amended CP in metal-amended slurries and control slurries. Inhibition of
-150-
-------
Table 1. Summary: Effects of Metal Amendment on the Onset and Rate of
Dechlorination of Chlorophenols in Anoxic Freshwater Sediments
Subst.
Rate
2-CP
3-CP
2-3-
DCP
2,4-
DCP
2,4,6-
TCP
Rate*
Lag
Rate
Lag
(days)
Rate
Lag
Rate
Lag
Rate
Lag
"Live"
Rate
0.64
12
0.62
59
0.63
10
0.83
7.5
0.60
7
Metal Amendment11
Cu(II) [mg/L]
20
0.65
10
0.56
58
0.74
7
0.76
7
0.89
7
100
0.55
24
N
0.26
37
0.63
10
0.76
26
200
N
N
N
OJ5
56
N
Cr(VI)
20
0.55
15
0.45
65
0.60
10
0.75
13
0.51
10
100
N
N
N
[mg/L]
200
N
N
N
0.18 VS
45
ND
ND
VS
CD(II) [mg/L]
20
0.56
10
0.46
58
0.60
7
0.87
7
0.56
14
100
N
N
N
200
N
N
N
0.10 N
26
N
N
"Dechlorination rate is expressed as mg CPloss L'May"1.
bFinal concentration of metal amended as CuCl2, CdClj, or
N, negligible loss; VS, very slow; ND, not determined.
-151-
-------
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
Totalb
[mg/L]
Dissolved0
[mg/L]
Sub-
strate
Amend-
ment
2,4-DCP
2,3-DCP
2,4-DCP
2,3-CP
Metal Amendment
20 mg/L
Cu
Cd
Cr
100 mg/L
Cu
Cd
Cr
200 mg/L
Cu
Cd
Cr
Experimental Results
25.1
26.9
<0.03
<0.03
16.0
20.2
<0.03
<0.03
19.7
ND
<0.03
<0.03
106.4
113.4
0.12
0.12
78.4
90.8
1.02
1.07
115.8
111.4
0.12
0.12
202.7
192.0
0.66
0.6
178.2
182.2
10.0
9.24
222.0
ND
0.34
0.42
"Values are the mean of duplicate determinations.
"Total metal concentrations after digestion with concentrated nitric acids.
•Metal concentrations of aqueous supernatant of sediment slurry after centrifugation at 10,000 xg and
filtering.
ND, not determined.
-152- -
-------
I
4!
Q
(a)
2-CP (Cu)
3-CP (Cu)
4-CP (Cu)
3,4-DCP (Cu)
50 100 150 200 250
Metal added [mg Cu/L]
(b)
2-CP (Cd)
3-CP (Cd)
4-CP (Cd)
3,4-DCP (Cd)
300
100 200
Metal added [mg Cd/L]
300
100 200
Metal added [mg Cr/L]
(c)
2-CP (Cr)
3-CP (Cr)
4-CP (Cr)
3,4-DCP (Cr)
300
Figure 1. Reductive dechlorination activity (DAc) of chlorophenol-adapted sediments in the
presence of (a) amended Cu, (b) amended Cd, and (c) amended Cr. The % DAc
is the ratio of the T50 of the metal-amended slurry and the control slurry. Data are
presented for dechlorination of 2-CP, 3-CP, 4-CP, and 3,4-DCP.
-153-
-------
dechlorination was greatest in experiments amended with Cd and Cu; reductive dechlorination of
3,4-DCPwas more resistant to the inhibitory effects of Cd compared to dechlorination of mono-CPs.
Interestingly, amendment of Cr resulted in the least inhibition of reductive dechlorination for all CPs
tested. In fact, reductive 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 content may affect the transformation of chlorophenols in anoxic sediments. Additional
studies are in progress to determine the inhibitory nature of metals on the reductive dechlorination
of higher chlorinated chloroaromatics, such as pentachlorophenol.
-154-
-------
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, MI
and
Dunja Grbic'-Galic'
Department of Civil Engineering
Stanford University
Stanford, CA
and
Gregory D. Sayles
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH
INTRODUCTION
Chlorinated dibenzo-p-dioxins (PCDD) and dibenzofurans (PCDF) have been generated as
unwanted by-products in many industrial and most incineration processes. Although their
widespread distribution in different environmental compartments has been recognized, little is known
about their fate in the ultimate environmental sinks, soils, and sediments. In particular, the
susceptibility of PCDD and PCDF to biological oxidation and reduction reactions has only recently
received attention, albeit limited.
Thus, the fate of selected highly chlorinated dibenzo-p-dioxins and dibenzofurans has been
studied in PCB (polychlorinated biphenyl)-contaminated Hudson River sediments and creosote-
contaminated aquifer samples. All samples were incubated anaerobically (i.e., in methanogenic
conditions) and spiked with 144 ±14 jig/kg of the following PCDD and PCDF congeners:
1,2,3,4,7,8-HxCDD; 1,2,4,6,8,9/1,2,4,6,7,9-HxCDD; 1,2,3,4,6,7,8-HpCDD; 1,2,4,6,8-PentaCDF; and
1,2,3,4,6,7,8-HpCDF.
-155-
-------
BACKGROUND
Dehalogenation of highly chlorinated PCDD/PCDF on soil and paniculate material has
previously been demonstrated during physical-chemical processes, such as photolysis. During
dehalogenation from the octa- to the tetra-chlorinated isomers, although the lateral chlorines (2,3,
7, and 8 positions) were selectively retained in the case of dioxins, others were preferentially
removed where dibenzofurans are concerned (1).
Reductive dehalogenation of PCBs with a high degree of chlorine substitution has been
demonstrated to occur in previously contaminated Hudson River sediments and in methanogenic
microcosms containing pristine or adapted sediments spiked with either Aroclor mixtures or
individual PCB congeners (2-4). The PCB-molecule is preferentially dechlorinated at the meta- or
/wra-position. Because of the structural similarity between dioxins or furans and PCBs, the potential
for susceptibility of the PCDD/PCDF to reductive dehalogenation processes was investigated.
RESULTS AND DISCUSSION
The fate of the spiked PCDD/PCDF (except for l,2,4,6,8,9/l,2,4,6,7,9-HxCDD)in anaerobic
sediments and soils (expressed as C/Co) is shown in Figure 1. All time samples for the active
microcosms represent an average of three replicates, while those for chemical (no inoculum) and
autoclaved controls represent two replicates. In all cases, the losses over time in the chemical
controls are negligible compared to the other treatments. The disappearance of PCDD/PCDF in
the autoclaved controls, however, is in most cases nearly as extensive as that in the active
microcosms, indicating the importance of sorption processes during the incubation period. The total
decrease of PCDD/PCDF appears to be slightly less in the low organic carbon (0.02 percent) aquifer
samples, when compared to the 7 to 8 percent organic carbon Hudson River sediment samples.
The first-order rate constants for disappearance—calculated from log (C/C0) plots—are given
in Table 1. Although it is difficult at this stage to differentiate between sorption alone and sorption
combined with biodegradation, the net rates calculated indicate that the biological component
increased the rate of PCDD/PCDF disappearance 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 heptaCDD-spiked microcosms
inoculated with aquifer material. Two hexachlorinated isomers have been identified: 1,2,3,4,7,8- and
1,2,3,6,7,8-HxCDD, respectively. 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 accumulated.
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 (Figure 2). A pentaCDF isomer
formed from dechlorination of 1,2,3,4,6,7,8-HpCDF has not been identified at this time.
-156-
-------
1,2,3,4,7,8-HexaCDD
1,2,3,4,6,7,8-HeptaCDD
o
o
o
Aetlv* Autoclaved Chemical
Active Autoelaved Chemical
1,2,4,6,8-PentaCDF
Active Autoclaved Chemical
Active Autoclaved Chemical
Active Autoclaved Chemical
Active Autoclaved Chemical
1,2,3,4,6,7,8-HeptaCDF
Active Autoclaved Chemical
Active Autoclaved Chemical
Treatment
Figure 1. Fate of selected PCDD and PCDF in Hudson River sediments (A) and in Pensacola
aquifer material (B).
-157-
-------
Table 1. First-Order Disappearance Rates of PCDD/PCDF in Hudson River Sediment and
Aquifer Material Inoculated Microcosms
PCDD/F
Congener
HpCDD
HxCDD
HxCDDi
HpCDF
PentaCDF
Inoculum
HR
PS
HR
PS
HR
PS
HR
PS
HR**
PS
Active
2.81 + 0.07
3.44 ± 0.10
4.56 ± 0.30
2.48 ± 0.13
4.00 + 0.16
ND
4.13 + 0.10
4.88 + 0.06
2.69 + 0.40
1.54 + 0.10
Rate* (x 10-2 wk-1)
Autoclaved
2.38 ± 0.14
2.19 ± 0.02
3.19 + 0.09
1.85 + 0.16
4.13 -I- 0.14
ND
3.56 + 0.05
3.13 ± 0.01
1.06 + 0.02
1.23 + 0.02
Nettorate (%)
0.43 ± 0.14 (15)
1.25 ± 0.10 (36)
1.37 ± 0.30 (30)
0.63 ± 0.16 (25)
(-0.13 -1- 0.16)
ND
0.57 + 0.10 (14)
1.75 ± 0.06 (36)
1.63 + 0.40 (60)
0.31 ± 0.10 (20)
* n = 6-12
** Not in 95 percent confidence interval.
-158-
-------
Cl
Cl
Cl
o
o
Cl
Cl
Major
ci
ci
Minor
ci
ci
CI
Cl
o
o
Cl
Cl
Cl
Cl
o
o
Cl
Cl
Cl
Cl
1,2,3,4,7,8-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.
-159-
-------
REFERENCES
1. Tysklind, M., A.E. Carey, C. Rappe, and G.C. Miller. 1992. Photolysis of PCDF and PCDD
on soil. Proceed. Diorin 1992 Conference, Vol. 8, pp. 293-296, Tampere, Finland.
2. Brown, J.F., Jr., D.L. Bedard, MJ. Brennan, J.C. Carnahan, H. Feng, and R.E. Wagner.
1987. Polychlorinated biphenyl dechlorination in aquatic sediments. Science 236:709-712.
3. Quensen, J.F. Ill, S.A. Boyd, and J.M. Tiedje. 1990. Dechlorination of four commercial
polychlorinated biphenyl mixtures (Aroclors) by anaerobic microorganisms from sediments.
Appl. Environ. Microbiol. 56:2360-2369.
4. Nies, L., and T.M. Vogel. 1990. Effect of organic substrates on dechlorination of Aroclor
1242 in anaerobic sediments. Appl. Environ. Microbiol. 56:2612-2617.
5. Adriaens, P., and D. Grbic'-Galic'. 1991. Evidence for reductive dehalogeriation of highly
chlorinated dioxins and dibenzofurans. Abstr. Dioxin 1991 Conference, Chapel Hill, NC.
-160-
-------
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
INTRODUCTION
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, California, serves a population of some
2.3 million people. Currently, over 250,000 acre feet of the basin's capacity is rendered unusable due
to 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 redrilling 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 electrodialysis, 5) subjecting water to an ion-
exchange process, and 6) subjecting water to biological denitrification (4). Orange County Water
District has been exploring the use of biological denitrification as a means of producing potable-
quality water from nitrate-contaminated ground water reserves.
BACKGROUND
The balance of existing research on biological denitrification relates to heterotrophic
processes in which an organic compound, such as methanol, glucose, or glycerol, is added to the
high-nitrate water (5,6). Such organics 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.
-161-
-------
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 biological denitrification
catalyzed by the autotrophic bacterium, Thiobacillus denitrificans, which in the absence of free oxygen
reduces nitrate to dinitrogen gas by oxidizing 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 obtained from an anaerobic enrichment
culture consisting of basal mineral salts and sodium thiosulfate. Upon development, the culture was
transferred to flasks containing basal salts and elemental sulfur. The culture consists of a consortium
of bacteria: Thiobacillus denitrificans 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 polyvinyl chloride (PVC) pipe closed at the ends with PVC unions. The
reactor was mounted outdoors and fed with nitrate-amended well water, which was conveyed to the
reactor using an adjustable peristaltic pump (Masterflex) capable of flows of up to 13.2 L/min. The
reactor contained 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 constructed using a glass and stainless steel
research fermenter (New Brunswick Scientific). The reactor (1.3-L liquid capacity) system provides
the capacity for temperature control, variation of agitation rate, automatic pH 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 reactor was fed (via a peristaltic pump) a synthetic salts solution formulated
to simulate locally occurring high-nitrate ground water. The reactor system included a gravity
separator unit to recover unutilized sulfur from the denitrified effluent. The separated sulfur was
continuously 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 nitrate, nitrite, sulfate,
and pH. Studies were also conducted on the effluent to determine total autotrophic and
heterotrophic bacterial counts.
RESULTS/DISCUSSION
Both reactor types were capable of maintaining prolonged periods of stable operation
(effluent containing less than 0.3 ppm of either nitrate 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
-162-
-------
reactor and the sulfur slurry reactor are presented in Tables 1 and 2, respectively. Flow 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 mL/min (reactor temperature at 30°C).
SUMMARY/CONCLUSIONS
The operational characteristics of two sulfur-based biological reactor systems were
investigated. Maximum hydraulic loading rates consistent with stable denitrification 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 separating 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 constructed with large height-to-
diameter ratios, thus minimizing 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 contamination 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 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. 1989. Biological denitrification
of water. J. Environ. Engin. 115:930-943.
6. Rogalla, F., G. DeLarrninat, J. Coutelle, and H. Godart. Experience with nitrate removal
methods from drinking water. Proc. Conference on Nitrate Contamination: Exposure,
Consequences, and Control. Nato Advanced Department of Civil Engineering, Lincoln, NE.
-163-
-------
Table 1. Denitrification by Fluid Bed Reactor
RUN HOUR INFLUENT NITRATE (ppnri EFFLUENT NITRATE (pprn)
6600 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
6648 42.5 0.0
6654 42.5 0.0
6660 42.5 0.0
6666 42.5 0.0
-164-
-------
Table 2. Denitrification by Fluid Bed Reactor
RUN HOUR INFLUENT NITRATE (vvm\ EFFLUENT NITRATE fpprn)
2303 57.2 0.0
2309 57.2 0.0
2315 57.2 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
-165-
-------
CHEMICAL INTERACTIONS AND pH PROFILES IN MICROBIAL BIOFILMS
Joseph R.V. Flora, Makram T. Suidan, and Pratim Biswas
University of Cincinnati
Cincinnati, OH
and
Gregory D. Sayles
U.S. Environmental Protection Agency
Cincinnati, OH
Substrate utilization kinetics within biofilms has been modeled by coupling Fickian diffusive
transport with Monod-reaction kinetics (1,2). Substrate and product concentration profiles are
effected in biological systems because of simultaneous diffusion and reaction. Compounds utilized
or produced during biological transformations can have an impact on the rate of substrate utilization
by affecting the local pH. For example, nitrification and the utilization of halogenated organic
compounds 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 equilibrium. The rates of substrate utilization and growth of microorganisms are
pH dependent and can vary significantly from the bulk solution to the attachment wall of biofilms
because of the pH gradients influenced by diffusional resistance to mass transport. An approach was
developed to analyze the effects of chemical interactions 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 account for the production and movement of gaseous products in methanogenic
biofilms. Gaseous products accumulate within the biofilm as bubbles that periodically 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 compounds in
the biofilm consists of the diffusion of dissolved compounds in the liquid phase (including the
biofilm) and the advection of gaseous products in the gas phase.
-166-
-------
Assuming that the biofilm is at steady-state with respect to the consumption of substrate, a
one-dimensional mass balance on total acetate (Cj^), total inorganic carbon (Cnc), and methane
in the biofilm yields,
[H4c]
dx dx dx dx
pB
f (3)
dx dx{ RT } Ks + [flXc] '**
where:
J = the molar flux of the species
x = the spatial coordinate perpendicular to the biofilm surface
qg = the gas volumetric flow rate per unit surface area of biofilm
KJ, = Henry's constant
R = the universal gas constant
T = the absolute temperature
b = the respiration coefficient
Y = the cell yield per mass of acetate
Xf = the microorganism density in the biofilm
qmax = the maximum specific growth rate at the optimum pH
K. = 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 Pick's law for neutral species and by the Nernst-Planck equation for ionic solutes
(4). The fractional decrease in growth rate due to deviations from the optimum pH is represented
by a Gaussian curve (5),
(4)
Equations 1 to 4 were solved in conjunction with expressions for acid-base and gas-liquid equilibria
and electroneutrality.
The profiles of (CTAc) and pH for a biofilm depth of 1,000 jim, bulk pH=7.0, bulk CrAc=10-2
M, and various bulk partial pressures of carbon dioxide are shown in Figure 1. The profile of Cj.Ac
predicted by using traditional 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 Pick's law to describe the diffusion 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).
-167-
-------
Depending on the partial pressure of carbon dioxide (Pcoz)*tne 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 C^ within the biofilm. The lower the PCQV the lower the
buffer intensity, and the larger the increase in pH and C^. in the film. Since traditional models
assume a constant pH, the concentration of C^ reaches very low levels in the biofilm.
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 CnC and a biofilm depth of 1,000 j*m. The flux predicted by
using the traditional model also is shown in the 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
microorganisms specified using equation 4. In contrast, the current model predicts an optimum pH
less than 7.0 for acetate-utilizing biofilms. Since the pH increases in acetate-utilizing 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 approach developed can be coupled
to a reactor model, and the operating conditions for a reactor (such as the buffering capacity of the
influent feed, the influent pH, and the surface loading rates) can be specified to optimize reactor
performance.
REFERENCES
1. Williamson, K., and P.L. McCarty. 1976. A model of substrate utilization by bacterial films.
J. Water Pollut. 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 concepts). 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.
-168-
-------
10
-2
s
d
-3
o 10
iH
J
a
8 io-4
10
-5
this study.
- - thl» study,
7.5
7.4
7.3
SB
PH
7.2
7.1
7.0
"/ y this study, F^ =10
i ; this study, ?,_„•= 10~a
:/ • - - this study, ?„,"= 10
0 200 400 600 800 1000 0 200 400 600 800 1000
Distance into the Biofilm,
Figure 1. Profiles of total acetate (C^) and pH for various bulk partial pressures of carbon
dioxide, a biofilm depth of 1,000 pm, bulk pH=7.0, bulk Cj-A^lO'2 M, and LCBL=0.
cd
0.2
g
O
\
53
V
O
0
a
).1
0.0
traditional model
thi, study, Cnc=1
tills study. Cnc=10_a
this study. ^=10
:a
bulk pH
Figure 2.
The flux of substrate into the biofilm as a function of bulk pH for a bulk CTAc=10"2
M for various bulk C^, and a biofilm depth of 1,000 /im.
-169-
-------
CHARACTERIZATION OF BIOF1LTER MICROBIAL POPULATIONS
Alec Breen, Alan Rope, and John C. Loper
Department of Molecular Genetics
University of Cincinnati
Cincinnati, OH
and
P.R. Sferra
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH
INTRODUCTION
Biofiltration, a promising, economical form of biological treatment, is becoming a more
common means of removing 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 employed in the treatment of volatile organic compounds
(VOCs) such as toluene, benzene, and ethylbenzene as well as more recalcitrant compounds such
as methylene chloride and trichloroethylene (TCE) (2).
In addition to sound engineering design, improved biofilter efficiency as well as broadened
VOC removal capabilities are ultimately dependent on the catabolic potential and activity of the
biofilter microorganisms. Biofilter microbial communities are largely uncharacterized (3). Analysis
of the structure and function of biofilter communities should provide useful information for biofilter
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 oxidation are being employed to determine which biochemical pathways are
present in the biofilter, and phylogenetic probes are being used to determine what types of bacteria
are predominant. In addition to these methods, DNA amplification fingerprinting (DAF) is being
evaluated as a means of monitoring the microbial community structure of the biofilter.
-170-
-------
BACKGROUND
Incorporation of some of the new DNA- based techniques in the study of biofilter microbial
ecology has the potential to result in more precise and less time-consuming analyses that will directly
contribute to improved biofilter performance. DAF technology has been successfully employed in
a number of studies and has emerged as a powerful tool for rapid microbial identification (4,5,6).
DAF also has been useful in the mapping of complex genomes (7). The utility of DAF in the
analysis of mixed cultures is being evaluated as part of this study. Nucleic acids can be extracted
directly from biofilter 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 community structure over time or among different biofilters.
Parallel studies using gene probes and standard culturing techniques are being conducted. The hope
is that structural and functional assessments of biofilter communities can be made by combining
these methods.
RESULTS
Initial experiments involved the generation of fingerprints for previously characterized VOC
catabolic strains of Pseudomonas cepacia G4, P. putida Fl, and Pseudomonas sp. JS150. These
experiments served to evaluate the efficacy of various primers in the PCR reaction and provided
fingerprints to compare with uncharacterized 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 oligonucleotide as a primer. Microcosms, one with toluene as a sole
carbon source and the second with toluene, ethylbenzene, and^-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 been predicted given the complexity
of the sample.
Three operational VOC catabolizing biofilters, packed with peat and maintained in different
configurations, were analyzed using DAF. These biofilters were maintained 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 nutrient 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 support media, while the
pelletized biofilter yielded a unique pattern. The individual organisms from these biofilters appear
to constitute a relatively diverse population 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
DAF 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
-171-
-------
Figure 1. DAF analysis of mixed microbial populations. Patterns were generated from nucleic
acid extracts from shake flasks inoculated with biofilter material and maintained on
VOCs for five months. Lanes 1 and 7 are 123-base molecular weight standards.
Lanes 2, 3, and 4 are patterns from the toluene-maintained culture over three
consecutive days. Lanes 5 and 6 are patterns generated at 48-hour intervals from the
flask maintained on toluene, ethylbenzene, and/?-xylene.
-172-
-------
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
monitoring and for diagnostics. Ideally, a given DAF profile for a particular biofilter will prove to
be diagnostic for particular modes of performance. Studies currently under way will involve gene
probing to provide a functional component of the analysis.
REFERENCES
1. Leson, G., and A.M. Winer. 1991. Bio filtration: 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,
nonfermentative bacteria isolated from biofilters on the basis of fatty acid composition,
quinone system and physiological reaction profiles. Appl. Environ. Microbiol. 58:2053-2065.
4. Williams, J.G.K., A.R. Kubelik, K.J. Livak, J.A. Rafalski, and S.V. Tingey. 1990. DNA
polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucl. Acids Res.
18:6531-6535.
5. Lehman, P.F., D. Lin, and B.A. Lasker. 1992. Genotypic identification and characterization
of species and strains within the genus Candida by using randomly 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 isolates 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.
-173-
-------
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
and
Dolloff F. Bishop
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH
INTRODUCTION
The Superfund Amendments and Reauthorization Act (SARA) emission summary for
petroleum and chemical manufacturing companies shows that the largest releases in air are volatile
organic compounds (VOCs). Conventional technologies that have been used to control 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 technologies 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, biologically active layer, referred to as biofilm, immobilized on the surface of the support
material. The solid support material can be either fine paniculate material, such as soil, peat,
compost, pellets of activated carbon, or ceramic media, or structured monoliths with passages of
defined geometry. The microbial degradation of substrate 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 completely degraded, unlike in some other technologies where they are simply
transferred from one phase to other (2).
-174-
-------
BACKGROUND
Biological filtration of VOCs implies the removal of compounds by contacting the
contaminated air with microorganisms in a column or similar configuration. Three basic types of
biological filters are available: bioscrubbers, 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 degradation products.
Biological methods have been employed for purification of waste gases since the 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 Bonn (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, especially 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 distribution 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 biofiltration is the most cost-
effective treatment method.
The soil bed biofilter systems mentioned above are inherently inefficient with respect to space
utilization. One biofilter system, termed the BIKOVENT system, consists of prefabricated concrete
parts that form an aeration plate to give uniform air distribution and drainage ducts. Developed by
Drs. Hans Gethke and Detlef Eitner of Aachen, Germany, the BIKOVENT has been extensively
used in Germany and Austria for odor control 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 undergoing 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 (MeCl2), trichloroethylene (TCE), ethylbenzene, and
chlorobenzene in air in aerobic biofilters.
-175-
-------
MEASUREMENT OF BIOFILM KINETICS
A packed bed raicrobiofilter 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. Degradation of the substrate was monitored based on the
substrate concentration in the gas reservoir. Mass balance equations were written for the system and
the data obtained on the microbiofilter was analyzed. Figure 2 shows a typical experimental run for
the system.
EXPERIMENTAL BIOFILTER STUDIES
The first study described the performance of a packed bed activated carbon biofilter (9).
Complete degradation of toluene, methylene chloride, and TCE was demonstrated at the bench-scale
at 2 minutes of gas retention time. After about 180 days of operation, the biofilter flooded due to
excessive biomass accumulation and the carbon pellets had to be cleaned periodically. Figure 3 shows
the removal efficiency of the biofilter as a function of time for the aerobic activated carbon biofilter.
The second study described the performance of a bench-scale packed bed biofilter with
ceramic pellets (Celite, Manville Corporation, CA). Complete degradation of all compounds except
TCE was achieved in our experiments. Figure 4 shows the removal efficiency of the compounds in
the packed bed Celite biofilter.
The third study described the performance of a straight passages biofilter constructed from
corrugated plates of Celite. The straight passages offer a means for excessive biomass to leave the
biofilter bed, which does 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 nutrients, as shown in Figure 6.
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 experimental data from the
above three studies. Insights into biofilter design and operation gained from this analysis will be
presented subsequently. That presentation will include a preliminary design procedure for biofilters
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 purification of waste gases. Chim. Oggi. 5:41-
45.
4. Bonn, H.L., and R.K. Bonn. 1986. Soil bed scrubbing of fugitive gas releases. J. Environ. Sci.
Health. A21:1236.
-176-
-------
NUTRIENT
RESERVOIR
JACKETED
BIOFILTER
BIOMASS
ON PELLETS
TEFLON BAG
Figure 1. Schematic of the experimental apparatus for determining biofilm kinetics.
-177-
-------
500
CONTAMINANT: TOLUENE
O run 1
H run 2
A run 3
A run 4
2 3
TIME (hrs)
Figure 2. Experimental results on biofilm kinetics for toluene using the experimental apparatus
shown in Figure 1.
-178-
-------
120
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 biofilter.
-179-
-------
60
TIME (DAYS)
120
Figure 4. Experimental results on removal efficiency obtained for the Celite pellets packed bed
aerobic biofilter.
-180-
-------
&
BJ
0
>••*
UH
fc
w
100
90-
80-
70-
60-
50-
40-
30-
20-
101
0
0
-X- MeCI2
3- Chlorobenzene
TCE -*- Toluene
Elhylbenzene
60
TIME (DAYS)
120
Figure 5. Experimental results on removal efficiency obtained for the Celite plates straight
passages aerobic biofilter.
-181-
-------
100^
•=3
cs
B
Q
0.1
GAS FLOWRATE
C! 1600 rri/irin A 600 ml/ton
©
10
JOO
NUTRIENT FLOWRATE (L/m2.hr)
Figure 6. Experimental data on biomass slough off from the straining passages biofilter.
-182-
-------
5. Kampbell, D.H. et al. 1987. Removal of volatile aliphatic hydrocarbons in a soil bioreactor.
J. Air. Pollut. Contr. Assoc. 37:1236.
6. Eitner, D. 1984. Investigations of the use and ability of compost filters for the biological
waste gas purification 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.
-183-
-------
SEQUENTIAL ANAEROBIC/AEROBIC TREATMENT OF
CONTAMINATED SOILS AND SEDIMENTS
Grace M. Ldpez, Gregory D. Sayles, Karen Biihler,
and Dolloff F. Bishop
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH
and
In S. Kim, Guanrong You, Petra Klostermann,
and Margaret J. Kupferle
Department of Civil and Environmental Engineering
University of Cincinnati
Cincinnati, OH
and
Douglas S. Lipton
Levine-Fricke Consulting Engineers
Emeryville, CA
INTRODUCTION
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 hazardous 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, the biochemical 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 typically 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.
-184-
-------
The objective of this project is to conduct fundamental and applied research to aid in the
development of sequential anaerobic-aerobic landfarming and composting technologies. These
technologies will be used to biologically 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 primary substrate 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 record 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 are also fed semicontinuously using ethanol as the primary substrate. Anaerobic
biodegradation products of the compounds listed in Table 1 will be used as co-substrates.
The rates and products of biodegradation will be measured in batch reactors under anaerobic
and aerobic conditions using aqueous slurries of soils and sediments. The soil or sediment will be
spiked with contaminant and transferred to serum bottles. Inoculum, co-substrate, and minimal salts
media will be added, and substrate and product concentrations will be measured over time.
Table 1. Compounds of Interest
DDT
HEXACHLOROBENZENE
PENTACHLOROPHENOL
AROCLOR 1260 (PCBs)
NAPHTHALENE (PAHs)
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 hexachlorobenzene MCRs for
a 6-month period are presented in Figures 1 and 2, respectively. The concentrations 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 concluded that the optimal operating conditions
for DDT and hexachlorobenzene were at a concentration of 2 mg/L.
-185-
-------
Accumulated Gas Production
DDT Reactor
CM CM CM CM CM CM
CM CM
>
CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCM
O> O> O> O> O> O> O) O> O> O> O* 1* 1* "* ^^ °^
o> o»
sll-^^^-^^S
Date
Figure 1. Gas production for anaerobic DDT reactor over a 6-month period.
O
O
-186-
-------
Accumulated Gas Production
CB Reactor
• mg/l HCB
- Gas Production
CMCMCMCMCMCMCMCMCMCNCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCM
O>^>O>O>O>O>O>O)O>O>O>O>O>O>O>O>O)O)O>O>
CM
r-
CO OO
O> O> O> T- i- O C
T- T- T—
Date
Figure 2. Gas production for anaerobic hexaclorobenzene reactor over a 6-month period.
-187-
-------
CONCLUSION
The MCRs are operating at steady-state, and the biodegradation 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. realities. U.S. Environmental Protection
Agency.
2. Picro, A.M., D. Kafkewitz, C.-M. Kung, and G. Lewandowski. 1992. Dehalogenation and
mineralization of 2,4,6-trichlorophenol by sequential activity of anaerobic microbial
populations. Biotechnology Letters 14(2).
-188-
-------
POSTER SESSION
-------
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 other pollutants. Most types of remediation efforts, including biological 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
environmental applications of microorganisms are used for remediation purposes, the organisms and
their products must be evaluated for safety.
Given the presence of multiple pollutants and the potential production of other pollutants
by remediation processes, 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 NCP's [nonconventional pollutants] are a potentially important
source of hazardous exposure. Some preliminary toxicologic studies suggest that NCP's have
important biologic properties, environmental persistence, and mobility. . . . these unidentified
substances represent risk of unknown magnitude." These NCP's, therefore, limit the ability to
conduct risk characterizations of remediation sites when only analytical chemistry is used for
exposure assessment studies.
By incorporating biological tests into assessment studies, it is possible to improve the
estimations of potential human toxicant exposure before, during, and after remediation efforts.
When appropriately coupled with analytical chemistry, bioassays also can be used to identity the
major toxic pollutants. In addition, during the development of remediation methods, bioassays can
be used for comparative assessments between differing technological approaches. Also, naturally
occurring, mutant, and genetically engineered microorganisms have great potential for use in
environmental remediation. Environmental releases, however, may result in direct human exposure
to these organisms.
In the 1970s, using primarily Escherichia colt, researchers 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,
-191-
-------
however, when present in large numbers can cause adverse health effects by other mechanisms, 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 pathogens. This poster presents the use of bioassays for comparative assessments
of complex mixtures and introduces developing methods for the health assessment of microorganisms
that are targeted for environmental release.
-192-
-------
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-I mice were challenged intranasally
with environmental 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. Environmental Pseudotnonas maltophilia strain BC6, and
P. aeruginosa strains BC16, BC18, and AC869, as well as clinical P. aeruginosa isolates DG1 and
PAO1, were detectable in the intestinal tract 14 days following treatment. In CD-I mice, P. cepacia
strain AC1100 impacted the small intestinal anaerobic count, lactobacilli, and obligately anaerobic
Gram-negative rods. Both strains AC869 and AC1100 modified the cecal obligately anaerobic
Gram-negative rods and lactobacilli 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 PAO1 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.
aeruginosa, 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 pathogens, or translocation of bacteria to other organs. This abstract does
not reflect EPA policy.
-193-
-------
FIELD TREATMENT OF BTEX IN VADOSE SOILS
USING HYDROFRACTURING, VACUUM EXTRACTION,
AND BIOFILTERS
Dolloff F. Bishop and Wendy Davis-Hoover
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH
and
Rakesh Govind
University of Cincinnati
Department of Chemical Engineering
Cincinnati, OH
INTRODUCTION
Spills of fuels and leaking fuel tanks represent a major source of vadose soil contamination.
Such contamination, which includes aromatic hydrocarbons (benzene, toluene, ethylbenzene, and the
xylenes [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 biofiltration to mineralize the extracted VOCs in the contaminated air.
BACKGROUND
This presentation describes a planned field demonstration 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 fractures the soil.
Sand-laden slurry is then pumped into the fracture to create highly permeable pathways in the soil.
The resulting horizontal sand lenses, at vertical separations as close as 15 cms apart, produce
improved overall soil permeability for vacuum extraction of fuel hydrocarbons from the consolidated
-194-
-------
soil. The extracted air contaminated with the fuel VOCs may then be treated in air biofilters to
biodegrade the contaminant VOCs.
RREL and UC are developing improved biofilters to control VOC emissions for RREUs
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 biofilter are
biodegraded by microorganisms attached on the surface of the media. A nutrient solution is recycled
through the biofilter to support the microorganisms. Research on improved biofilters reveals
efficient removal of biodegradable VOCs and a high degree of mineralization.
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 contaminated with jet fuel VOCS will
be hydrofractured by a RREL/UC team using the vacuum extraction equipment installed to permit
air extraction of the VOCs. Hydrofracturing 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 provides 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 meter (height) of
porous ceramic pellets (6 mm avg. dia.), the other with approximately 1 m (height) of porous
ceramic straight-passages media (127 passages/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 the 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 biofilm 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 performance of the integrated system will be characterized for
approximately 3 months.
-195-
-------
CLEAN AIR
AIR WITH VOCs
MICROORGANISMS
IMMOBILIZED ON
SUPPORT MEDIA
NUTRIENT
SOLUTION
FRESH
NUTRIENTS
Figure 1. Schematic of the biofilter system.
-196-
-------
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
and
J.T. Wilson
U.S. Environmental Protection Agency
Robert S. Kerr Environmental Research Laboratory
Ada, OK
INTRODUCTION
French Limited is a Superfund site near Houston, TX. The site is a sand pit that was used
for disposal of chemical waste, primarily polycyclic aromatic hydrocarbons (PAHs), from API
separator sludge, and polychlorinated biphenyls (PCBs), from discarded transformers. A technology
vendor at the French Limited Superfund Site had 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 confirmation of the observed PCB degradation in a controlled laboratory
experiment was desirable. A laboratory experiment was proposed that would determine
quantitatively the changes in the sludge under highly aerobic conditions. The purpose of this
experiment was to determine 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 hydrocarbons (Table 1), was
observed, the total PCB concentration was not significantly affected (Table 2). Only certain of the
lower chlorinated PCB congeners were significantly degraded under aerobic conditions, specifically
BZ-7/9, BZ-6, and BZ-8/5 (Table 3). These congeners were dichlorobiphenyls. Figures 1 and 2
-197-
-------
Table 1. Total Petroleum Hydrocarbon (TPH) Degradation
TPH (mg/L) AVG. SD (n = 3)
Start 18,919 5,944
Final
Anaerobic 18,586 864
Aerobic 6,939 369
-198-
-------
Table 2. Polychlorinated Biphenyl Depletion (rag/L of PCBs in Mixed Liquor; Average of
Duplicate Samples)
Aerobic Anaerobic
Initial 18.5 15.9
Final 14.5 14.8
Depletion (%) -21.9 -6.8
-199-
-------
Table 3. Changes in Concentration of Polychlorinated Biphenyl Congeners during Incubation
under Aerobic and Anaerobic Conditions
Aerobic
Before After
Anaerobic
Before After
Reduction (%)
Aerobic Anaerobic
BZ-10/4
BZ-7/9
BZ-6
BZ-8/5
BZ-31
BZ-28
BZ-33/53
BZ-52
BZ-44
2.80
0.30
0.48
2.81
2.56
3.26
4.22
4.68
2.67
3.01
0.00
0.14
0.49
2.41
3.37
3.35
5.20
2.87
2.94
0.38
0.65
3.67
2.54
3.28
4.21
4.48
2.57
2.45
0.39
0.65
3.70
2.57
3.36
4.26
4.60
2.63
-7.6
100.0
70.7
82.4
5.8
-3.3
20.5
-11.2
-7.5
16.9
-2.9
0.1
-0.8
-1.2
-2.5
-1.2
-2.6
-2.6
Note: Concentrations expressed as percent of total PCBs.
-200-
-------
100
90
BO
70
60
c
(V
D>
X
30
20
10
Figure 1
Aerobic Reactor Oxygen Consumption
Corr . Goer. - 0.833 from Day Z6 to Day 103
Oxygen Demand = 892 mg/ I/day
SO
150
Days
200
250
Figure 1. Aerobic reactor oxygen consumption.
-201-
-------
o.e
0.7
O O.B
0.
(0 0.5
•H
o
(-
0.4
c
-------
illustrate that the consumption of these PCB congeners occurred after biological activity had
decreased. No congener was significantly degraded under anaerobic conditions (Table 3).
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. Mainprize, 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. Appl.
Environ. Microbiol. 53 (1094).
-203-
-------
DEVELOPMENT AND EVALUATION OF COMPOSTING TECHNIQUES FOR
TREATMENT OF SOILS CONTAMINATED WITH HAZARDOUS WASTE
John A. Glaser,1 Carl L. Potter,1 Edward D. Kennedy,2 Jeffrey J. McCormack,2
Joseph B. Farrell,1 and Michael Najar2
Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
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 understanding of
the processes andmechanisms involved. Current commercial compost operations constitute black-box
systems where optimization is approached on a hit-or-miss basis using scant process-related
information. 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 operations.
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 throughout 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.
A top sample 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 leachate samples 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 will be constructed for a treatability study using soil contaminated with
creosote components.
Investigation of optimal composting conditions for destruction of soil contaminants involves
determination of suitable soil types, moisture content, co-compost materials (e.g., wood chips,
-204-
-------
sawdust, corn cobs), and nutritional adjustments (e.g., carbon/nitrogen ratio). The study also seeks
to identify metabolically active microbial 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 Salmonella mutagenicity, earth
worm toxicity, and seed germination tests.
-205-
-------
ENGINEERING OPTIMIZATION OF SLURRY BIOREACTORS
FOR TREATING HAZARDOUS WASTES IN SOIL AND SEDIMENTS
John Glaser,1 Paul McCauley,1 Majid Dosani,2
Jennifer Platt,2 Edward Opatken,1 and Diane Roush2
'Risk Reduction Engineering Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH
2IT 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 mechanical
and treatability studies at bench and pilot scale. A systematic plan to evaluate slurry bioreactors and
the required parameters for their optima! use is being developed by this research program. The
information gained from these studies should be useful in improving treatment effectiveness in field
applications.
Current research efforts are centered on developing and selecting suitable bench- and pilot-
scale bioreactor designs 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 entry 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 the other ports and is used for slurry charging. Soil slurry
preparation includes screening, preslurrification, 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, chemical, and biological factors controlling optimal operation. Bench- and pilot-scale
reactors will yield information on solids loading methods, optimal residence times, optimal solids
loading as a function of soil type, pollutant mass balances, and effectiveness of nutrient and
microorganism additions. As bench-scale bioreactor performance information becomes available,
it will be used to configure the larger pilot-scale bioreactors.
-206-
-------
BIOTREATABILITY OF A VADOSE ZONE SOIL CONTAMINATED
WITH DIOCTYL PHTHALATE
Don H. Karapbell
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
Ada, OK
and
Dennis D. Fine and Jerry W. Anderson
ManTech Environmental Technology, Inc.
Robert S. Kerr Environmental Research Laboratory
Ada, OK
Chemical spillage and waste disposal during 40 years of manufacturing vinyl wallcovering had
contaminated portions of a field site with various organic chemicals, predominately 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 remediate the
site. Nine core samples were collected at three locations in 3-ft increments from site locations of
low contamination, moderate contamination, and high contamination. Analysis of a nine-sample
composite showed that 95 percent of organic compound contamination was dioctyl phthalate. All
core samples contained sufficient nitrogen and phosphorus to support viable soil microbiological
processes. Bacterial cell counts ranged from 7 x 108 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, air-tight
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 consumption 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 percent. The rate of dioctyl phthalate biodegradation in laboratory microcosm bottles was
approximately 0.1 g/kg soil/day for the contaminated cores. Extrapolating 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.
-207-
-------
INNOVATIVE BIOREMEDIATION 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
and
Richard Devereux, Deborah L. Santavy, and P. Hap Pritchard
Gulf Breeze Environmental Research Laboratory
U.S. Environmental Protection Agency
Gulf Breeze, FL
INTRODUCTION
The use of specially selected microorganisms to enhance 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 pentachlorophenol [PCP]) has resulted in the opportunity to employ
these technologies at similarly contaminated sites throughout the world. Prior to worldwide
dissemination 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 the introduction of these isolates would offer any advantage to the
bioremediation system.
MATERIALS AND METHODS
To address these issues, soils contaminated with polycyclic 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 utilizing 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 surveyed. 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 incubation and enrichment,
-208-
-------
cultures were diluted 1:5 (v/v) with fresh mineral salts medium containing PAHs. This transfer
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 purified, 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 clearing 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 measured
visually or by monitoring changes in absorbance at OD550nm. As a control, growth in
carbon-substrate free mineral salts broth was also monitored.
Once the PAH-degrading ability of purified cultures was validated, cultures were
characterized by fatty acid profile analysis (GC-FAME) and substrate utilization patterns (Biolog
Microplate System). The taxonomic relationships among these strains was analyzed by evaluating
similarity measures from GC-FAME and substrate utilization patterns with principal component
analysis (2). DNA/DNA hybridizations and 16S rRNA sequence comparisons also were performed
to determine the phylogenetic relationships among the recovered isolates according to published
methods (3-6).
In parallel studies, microbial respirometric responses (rate of liberation of carbon dioxide and
the simultaneous consumption of oxygen) upon exposure to the following 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 glycerol); or 3) 500 mg/L specification creosote No. 450 (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 MicroOxymax™ respirometer (Columbus Instruments,
Columbus, Ohio). These responses were compared to those observed 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-amendment-only, creosote-amended, and killed-cell
(control) treatments were extracted and analyzed for the presence 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 increased respiratory activity and enhanced PAH biodegradation 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 addition of these potential substrates.
The location of isolated strains and the corresponding PAH substrate are summarized in
Table 1. Physiological and biochemical analyses of isolated bacteria showed that the PAH-degrading
bacteria that recovered from PAH-contaminated soil via the specified techniques appeared to be
related phylogenetically. For example, all soils with a history of PAH exposure harbored bacteria
-209-
-------
o
I
u
g
u
•o
C3
D
Q
o
(U
IS
£
|
4->
6
O
T3
(U
"I
o
0)
c
0
E
JC
o
LLJ
O
^ O
O £
o £
^ ®
j «*—
O 0)
CO DC
.>
Q. _l _l
O CD
C C
CD CD
v_ k.
a--5-
^5 ^5 CD
c c c
CD CD CD
ca ca CA (0 (o co
CA
CA
CO § 8 ra O § 8
O ca ca O co co co
to CDCD>CDCDCDCD'wCacacO:= = = = =CoCDCDCl}CDorJ
i ft f\ J'Q_ n DL n n ff rr' rr" ' • "~ ' ' '~ **^ "^ rn fn fn fn ** —
CMCMCMCMCMCMCMCMCMCS
-210-
-------
competent for PAH biodegradation, and principal component analysis of GC-FAME data showed
relatedness between these strains (Figure 1). The strategy employed to enrich PAH-degrading
bacteria thus appeared to select similar types of microorganisms that are indigenous to contaminated
soils at each site. Hence, the export/import of these non-indigneous bacteria to augment
bioremediation efforts would not seem to represent the introduction of exotic biota.
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, then, 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-indigenous microbes in optimized bioremediation systems could be advantageous
for cost-efficient, effective bioremediation, while posing no discernible ecological risk.
ACKNOWLEDGMENTS
We thank Mykelle Hertsgaard and Barbara Artlet (Technical Resources, Inc., Gulf Breeze,
Honda), Sheree Enfinger (U.S. EPA, ERL, Gulf Breeze, Florida), and Stephanie Willis (University
of New Hampshire) for technical assistance; Brian Klubek (Southern Illinois University-Carbondale,
Illinois) for soil analyses; Bruce Hemming (Microbe Inotech Laboratories, St. Louis, Missouri) 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 analyses. 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, Georgia) as defined under the Federal Technology Transfer Act, 1986
(contract no. FTTA-003).
REFERENCES
1. Mueller, J.G., PJ. 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 numerical taxonomy. In: Proceedings of the
15th Annual Users Group Conference. SAS Institute, Inc., Gary, NC. pp. 1413-1418.
3. Amann, R.I., C. Lin, R. Key, L. Montgomery, and D.A. Stahl. 1992. Diversity among
Fibrobacter isolates: Towards a phylogenetic classification. Syst. Appl. Microbiol. 15:23-31.
-211-
-------
c
0>
c
o
Q.
E
o
u
o
c
FDA PYR-!
2282, 2283
-40
-20
Principal Component 2
Bacterial similarity measurements using principal component analysis from
GC-FAME data (see Table 1 for soil of origin and PAH enrichment substrates).
Figure 1. GC-FAME principal component analyses.
-212-
-------
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. Pelletein and DJ. Lane. 1991. 16S ribosomal DNA
amplification for phylogenetic study. J. Bacteriol. 173:697-703.
7. Mueller, J.G., S.E. Lantz, B.O. Blattmann, and P J. Chapman. 1991. Bench-scale evaluation
of alternative biological treatment processes for the remediation of pentachlorophenol- and
creosote-contaminated materials: Solid-phase bioremediation. Environ. Sci. Technol.
25:1045-1055.
8. Berg, J.D., B. Nesgard, R. Gundersen, A. Lorentsen, and T.E. Bennett. 1993. Washing and
slurry phase biotreatment of creosote-contaminated soil. In: Proceedings of In Situ and
Onsite Bioreclamation Symposium, April 5-8, 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 Environmental 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 paucimobilis. 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 biodegradation of weathered Prudhoe Bay crude oil. J. Indust. Microbiol. 10:95-105.
12. Heitkamp, M.A., and C.E. Cerniglia. 1988. Mineralization of polycyclic aromatic
hydrocarbons by a bacterium isolated from sediment below an oil field. Appl. Environ.
Microbiol. 54:1612-1614.
13. Resnick, S.M., and P.J. Chapman. 1990. Isolation and characterization of a
pentachlorophenol-degrading, gram-negative bacterium. Abstr. Ann. Meet. Am. Soc.
Microbiol. p. 300.
-213-
-------
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
INTRODUCTION
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 response and
remediation decisions. The project will integrate available information that the user can understand
and utilize in a short time. The output will provide guidance to identify appropriate bioremediation
action options for a specific situation. Thus, the project output will meet the following objectives:
a) advance the understanding of applying bioremediation to specific problems, b) enhance the use
of bioremediation through performance evaluations and technology transfer, c) evaluate the use of
natural soil processes for treatment of hazardous wastes, and d) address informational impediments
that constrain the use of new technologies.
The project will focus on in situ and ex situ bioremediation approaches that can be used for
unsaturated soils. The technologies considered in this project are:
In Situ Ex Situ
Bioventing 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, processed and crude oils, and other raw, partially processed, and refined petroleum
materials.
-214-
-------
The anticipated users of the project output are expected to be a) emergency response team
personnel, b) onsite coordinators, c) regional and state project managers, d) industry personnel, e)
consulting engineers, and f) educational organizations interested in training students and others in
need of the information.
BIOREMEDIATION
Bioremediation is the controlled application of the naturally 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 bioremediation 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 fundamentals for use with contaminated soils and the technical 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 a) petroleum
hydrocarbons such as gasoline, diesel fuel, crude oil, and creosote; b) pesticides and their derivatives,
c) chlorinated solvents such as methylene chloride, and d) hydrocarbons such as pentachlorophenol,
naphthalene, and anthracene. An increasing number of sites are being remediated and closed using
bio-processes, magnifying the need for an expert system.
PROGRAMMING LANGUAGE AND FORMAT
The expert system will be designed for a personal computer 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 substantial financial
investment or be difficult to set up. This requirement would ensure a wide user base. Second, 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 the program will contain three main branches:
Branch 1 Explanation of techniques
Branch 2 Description of illustrating methods, with examples
Branch 3 Analysis of the user's situation
-215-
-------
The first branch will allow the user to become familiar with the various bioremediation
techniques. It will include a written explanation of each process and will list the various soil or
contamination criteria under which each can be effectively applied. This will provide a user who was
only recently exposed to the bioremediation of hydrocarbons, for example, with a solid background;
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 affirm 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 will then output the "best" choice or choices for this case and
the logic behind the decision.
Note that the program is only an aid to the user. No program can adequately replace the
logic of a user intimately involved with the particular situation. The program will only show the
correct thought process and output the best possibilities. The user should then take these
possibilities and thoroughly consider them, taking into account parameters the computer could not.
SUMMARY
An expert system is being developed to acquaint system users with the various bioremediation
techniques applicable in the unsaturated zone and to indicate the conditions 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 through
case studies, and the final branch will allow the user to input specific information so that the
program can output/suggestions on available techniques.
-216-
-------
A DATA VISUALIZATION SYSTEM FOR BIOREMEDIATION ANALYSIS
Lewis A. Rossraan
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH
and
Kevin Savage
Center Hill Research Center
University of Cincinnati
Cincinnati, OH
and
John Franco
Department of Computer Science and Engineering
University of Cincinnati
Cincinnati, OH
INTRODUCTION
Site characterization data play an essential role in determining the feasibility, design, and
operation of a bioremediation system. These data represent a series of point measurements taken
from wells, borings, and test pits of a highly heterogeneous, three-dimensional subsurface
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 toxicants. Site characterization data should be displayed in a meaningful
way so that they can be properly interpreted and used correctly when making remedial action
decisions. The goal of this project is to develop a powerful 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.
-217-
-------
APPROACH
Our data visualization software, called ICASE (Integrated Computer Assisted Site
Evaluation), takes measurements made at irregularly spaced locations at a site and uses a stratified
interpolation process to estimate values within a uniform three-dimensional grid underneath the site.
This procedure is carried out for any parameters of interest, which can include both soil properties
and chemical species. It is then 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 ability 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 is a collection of rules and conditions (perhaps even a scoring
system) that will infer the feasibility of a particular form of bioremediation technology based on
conditions measured at the site. These include soil and aquifer geotechnical properties along with
contaminant properties and nutrient levels. 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 technically 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-dimensional cross-sectioning
have all been completed. The three-dimensional visualization methods are currently being enhanced.
The major effort in the second year will be the development of the bioremediation knowledge base,
its incorporation into ICASE, and the testing of the completed system on a suite of site remediation
case studies.
-218-
-------
POSTER PRESENTATIONS
SUPPORTED BY
EPA'S HAZARDOUS SUBSTANCE RESEARCH CENTER PROGRAM
The following research is being carried out under the auspices of the EPA's Hazardous
Substance Research Center (HSRC) program. EPA established this program in response to
provisions in the 1986 amendments to the Comprehensive Environmental Response, Compensation,
and Liability Act. These provisions authorized EPA to establish HSRCs with a mission to study all
aspects of the "manufacture, use, transportation, disposal, and management of hazardous substances"
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 (ORD).
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 University, Rutgers University, Stevens Institute of Technology, Princeton University, and the
University of Medicine and Dentistry of New Jersey.
Great Lakes and Mid-Atlantic Hazardous Substance Research Center—Region-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.
-219-
-------
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 Georgia Institute of Technology and Rice University. The center's director is Dr. Louis
Thibodeaux of Louisiana State University.
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, Nebraska, and Utah.
Western Region Hazardous Substance Research Center—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.
-220-
-------
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, MI
INTRODUCTION
Bioremediation is an increasingly promising technology for cleaning up hazardous waste sites,
but because of a lack of data and knowledge to make informed decisions, owners and consultants
often do not think of it as a viable treatment alternative. Yet bioremediation is establishing itself
with a credible track record as a cost-effective technology for certain types of contaminated sites.
A limitation, however, is that bioremediation processes are not universally applicable. It is thus
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 Bioremediation Adviser. Such an advisory system would be useful 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.
BACKGROUND
Whether bioremediation is feasible for cleaning a contaminated 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-producing 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.
While such information will be contained 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.
-221-
-------
The Bioremediation Advisor will help bridge the gaps among practitioners' collection of
traditional data, theoretical knowledge about bioremediation treatment processes and characteristics
of pollutant-degrading microbes, and biological characteristics of the existing environment (Figure
1). Ultimately, this information will be used in conjunction with a rule-based system to guide field
data collection and to match field data with bioremediation technologies to suggest the feasibility
of bioremediation as a treatment for a given contamination and recommend possible bioremediation
technologies.
RESULTS AND CONCLUSIONS
To date, a literature survey has been performed, including a review of papers and textbooks,
pertaining to the biological degradation as well as the chemical properties of over 70 organic
contaminants. The characteristic chemical properties and observed biodegradation of 20 relevant
contaminants of concern, ranging from halogenated aromatic and aliphatic compounds to pesticides,
have been classified in HyperCard stacks. Environmentally relevant chemical properties include
redox potential, aqueous solubility, organic partition coefficient, and molecular structure. When
values were not available in the literature, they were estimated based on established methods. The
compiled biodegradation parameters include the nature of the degradation mechanism (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 currently 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 constitutes 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.
-222-
-------
e.g., attachment. pH,
temperature, inorganic
Pollutant-degrading
Microbe
Characteristics
In situ Physical/Chemical
Characteristics
e.g., u.u., nitrate,
toxicants, predators
Engineered
New "Site"
Characteristics
In situ Biological
Characteristics
Figure 1. Proposed expert system scope.
-223-
-------
USE OF COMPOSTING TECHNOLOGIES
TO TREAT A TNT-CONTAMINATED SOIL
James H. Johnson, Jr., Mohammed Moshin, Lily Wan, and Abdul Shafagatti
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 compounds.
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 aminodinitrotoluene isomers. Many of these compounds appear to be
toxic. The result is that method of transformation 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, however, require a very harsh environment (i.e., high-temperature, high-pressure, or severe
acidic or basic conditions). The most benign and field-applicable oxidant and the one chosen for
this project is ozone (with a radical initiator). Initial efforts using an aqueous system 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
hydrogen 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
studies are under way.
-224-
-------
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 effective 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 commensurate with
the scale of the Darcian velocity. The fundamental question addressed is how does the effective
reaction rate depend on the cumulative effects of pore-scale geometry and mechanisms. A model
is developed 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 microorganisms 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 biorestoration are best at the
pore scale. For the purposes of ground water and soil remediation, however, it is desirable to make
measurements, predictions, and assessments at a much larger scale. It is not feasible to bridge this
gap in scales by simply having a computational domain that spans both. First, there is not enough
data to specify the spatial structure with such detail; second, current computational resources are
inadequate 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 detailed 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
governing the pore-scale transport of the reactive contaminant over an appropriately large volume.
The upscaling methodology is based on a moment method approach. The upscaling procedure yields
the parameters that describe the transport of the less-detailed, spatially averaged solute
concentration, and in particular the effective first-order volumetric reaction rate. •
As a first step, a simple geometric model of a porous medium was used to examine the
effects of the subgrain-size structural features and processes on the measurable laboratory-scale
-225-
-------
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-order 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 behavior of C* is developed for the case of a solute undergoing diffusion Dm and a first-
order surface reaction X. 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
Presenter: Perry L. McCarty, Stanford University
INTRODUCTION
A promising microbial system for in situ bioremediation 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 fortuitously degraded by the MMO that is produced in order
for the microorganisms to utilize the supplied electron donor (methane). Studies at the Moffett
Naval Air Station (3,4,5,6) have demonstrated successful removal of several 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 contains 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 methane 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 contaminants are recirculated,
which increases their contact time with the biologically active zone, allowing a greater extent of
contaminant removal. In previous in situ bioremediation 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 different oxygen and
methane delivery schemes for a recirculation well operated to promote in situ bioremediation.
REFERENCES
1. Little, C.D., A.V. Palumbo, S.E. Herbes, M.E. Lindstrom, R.L. Tyndall, and P.J. Gilmer.
1988. Trichlorethylene biodegradation by a methane-oxidizing bacterium. Appl. Environ.
Microbiol. 54:951-956.
-227-
-------
2. Fox, E.G., J.G. Bourneman, L.P. Wackett, and J.D. Lipscomb. 1990. Haloalkene oxidation
by the soluble methane monooxygenase from Methylosinus trichosporium OB3b:
Mechanistic and environmental implications. Biochemistry 29:6419-6427.
3. Roberts, P.V., L. Semprini, G.D. Hopkins, D. Grbid-Galid, 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. A field evaluation of
in-situ biodegradation 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. Grbi6-Gali<5, and P.L. McCarty. 1991. A field
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
bioreclamation and/or in-situ remediation of strippable contamination. 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 recirculation well. Ground Water, (in press)
-228-
-------
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 energy, have been found capable of
oxidizing chlorinated aliphatic hydrocarbons (CAHs) such as trichloroethylene (TCE) by
cometabolism. A two-reactor treatment system for TCE destruction by methanotrophic bacteria is
being evaluated. The system consists of a growth reactor (completely mixed) that continuously
produces the active culture for addition to a TCE-contaminated waste stream in a separate
transformation reactor (plug flow). This reactor configuration eliminates TCE transformation
product toxicity during cell growth, benefits 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 concentration of methane (5 to
6 mg/L) at high TCE concentration (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 due to competitive
inhibition. At a lower TCE concentration (0.9 mg/L), the TCE transformation rate was fot J to be
lower in the presence of methane, indicating that competitive 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 concentration. 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 reactors 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).
-229-
-------
Nitrogen starvation did not affect the Tcat the 2-day SRT, but led to higher growth yields and higher
Tr
Pofy-beta-hydraxybutyrate (PHB). The Tc of resi.ng (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 laboratory growth reactors were compared, Tc increased with PHB content.
The activity of the soluble methane monooxygenase (MMO,, 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 percent. When the samples were amended with 20 mM formate,
a readily available electron donor, the specific naphthalene oxidation rate was constant and higher
tiian in unamended samples. These results indicate that in resting cells MMO, activity is limited by
the availability of a suitable source of reducing power.
-230-
-------
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
Presenter: Perry L. McCarty, Stanford University
Orange County Water District built a facility at the Seal Beach site in California where there
was a significant gasoline spill resulting in contamination of the ground water aquifer with
hydrocarbons including benzene, toluene, ethylbenzene, and m-, p-, and o-xylene (BTEX). The
facility was designed for the operation of in situ bioreactors wherein strategies for bioremediation
of the contaminated 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 focused 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 influents to the "natural background" control and denitrifying reactors were
contaminated ground water from the site and contaminated ground water amended with nitrate (13
mg/L as nitrogen), respectively. Although some biological removal of toluene and m^-xylene was
observed in the nitrate-amended bioreactor, toluene and m,p-xylene also were removed in the
unamended control 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 enhanced at the Seal Beach field site. Both toluene and m,/?-xylene were
removed in the unamended microcosms, just as observed in the field. Corresponding loss of sulfate
in the sample bottle suggested that the observed aromatic removal was probably due to
sulfate-reducing bacteria. Addition of nitrate to the samples stimulated denitrificaytion, which
resulted in faster disappearance of toluene, complete loss of ethylbenzene, but less complete removal
of m,p-xylene. Addition of a nonindigenous 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 utilized in the microcosms. Biotransformation of benzene
in several microcosms was tied to active sulfate reduction. In those microcosms that had both active
denitrification and sulfate reduction, there was complete loss of the full range of BTEX aromatic
compounds. The results of this experiment indicate that utilization of a combination of the electron
acceptors nitrate and sulfate may provide benefits in restoration of BTEX-contaminated sites.
-231-
-------
BIOTRANSFORMATION OF INDOLE AND QUINOLINE
UNDER DENITRIFYING CONDITIONS
J.N.P. Black, R.M. Kauffman, D. Denney, and D. Grbic-Galic
Western Region Hazardous Substances Research Center
Department of Civil Engineering
Stanford University
Stanford, CA
Presenter: Perry L. McCarty, Stanford University
The anaerobic conditions prevalent in many contaminated ground water plumes provide an
impetus to investigate 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 under denitrifying conditions was used as inoculum in nitrate-
containing enrichment cultures. The cultures were prepared according to a protocol of rigorous
oxygen 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, naphthalene, indole, quinoline, furan, benzofuran, thiophene, and
benzothiophene. Two live and two autoclaved controls were prepared per substrate.
Within 20 days, cultures exhibited transformation of indole and quinoline 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. Quinoline was decreased in its
cultures by 20 days and a transformation product that HPLC retention times indicated was 2-
quinolinol had appeared. By 39 days, quinoline was undetectable and 2-quinolinol had accumulated.
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 quinoline
transformation. Gas partitioner analysis of headspace samples failed to detect methane or hydrogen
sulfide. Examination of the quinoline-fed culture with phase-contrast microscopy revealed 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 completely transform indole and quinoline.
These subcultures were dominated by motile, Gram-negative rods growing singly and in dense and
macroscopically visible floes. Subcultures exposed to hydrogen consumed nitrate without degrading
the nitrogen heterocycles. The addition of 0.3 percent mercury chloride prevented consumption of
any substrates.
-232-
-------
COMETABOLISM OF TCE BY NITRIFYING BACTERIA
Michael Hyman,1 Roger Ely,2 Sterling Russell,1 Ken Williamson,2 and Daniel Arp1
Western Region Hazardous Substance Research Center
^boratory for Nitrogen Fixation Research
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 (NO2"). Ammonia oxidation
is catalyzed by ammonia monooxygenase (AMO), which converts NH3 to hydroxylamine (NH2OH).
The oxidation of NH2OH to NO2~ provides both the reductant required for AMO and the sole
source of electrons for adenosine triphosphate (ATP) synthesis. AMO can also oxidize alternate
substrates including alkanes, alkenes, aromatics, and many Cl to C3 halogenated aliphatics, including
trichloroethylene (TCE). The products of these oxidations are not assimilated and the oxidation
process requires simultaneous NH3 oxidation as a source of reductant. These oxidation reactions are
therefore regarded as cometabolic.
TCE cometabolism 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 bacteria are widely distributed 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 aerobic TCE-degading bacteria, suffers a
toxicity from TCE oxidation resulting from the formation of a reactive TCE-epoxide intermediate.
Much of this TCE toxicity involves inactivation of AMO. This research uses N. europaea as a model
system to study the physiological consequences 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 operating parameters for concurrent studies using
reactors capable of sustainable TCE degradation.
In short-term (10 min) experiments correlating TCE degradation (measured as Cl" 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 cells. TCE toxicity
-233-
-------
is irreversible without protein synthesis and affects both AMO and the enzymes and/or proteins
associated with NH2OH oxidation. Since sustainable TCE oxidation will require the ability of cells
to constantly resynthesize these inactivated proteins, these data suggest that 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 only occur at very low 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 synthesis of AMO (and other proteins) and what
contribution newly synthesized enzymes make to the kinetics of TCE degradation. Initial studies
following the kinetics of de novo protein synthesis using 14CO2 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. There is some evidence that TCE-epoxide may in fact inhibit or inactivate the epoxide
metabolism of these bacteria.
-234-
-------
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 Civil Engineering
Oregon State University
Corvallis, OR
and
Teresa Lemmon, James Ingle, and John Westall
Western Region Hazardous Substance Research Center
Department of Chemistry
Oregon State University
Corvallis, OR
INTRODUCTION
A knowledge of biotransformation pathways and kinetics is essential to assess risks at
contaminated sites, implement biological treatment processes, or design effective bioremediation
strategies. In addition to the microbial consortium, environmental conditions such as pH,
oxidation/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 because it is toxic to a wide variety of
organisms (1), it is widely distributed 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 Superfund Program.
The goal of this study is to better understand the effects of certain environmental conditions
on the rate of anaerobic biotransformations of chlorophenols. Preliminary studies were conducted
to characterize the pentachlorophenol degradation pathway for an acclimated culture. This
consortium will be used for all experiments, and the kinetics of these biotransformation reactions
will be measured.
-235-
-------
A computer-interfaced reactor system has been developed to allow measurement of
biodegradation rate constants 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. Pentachloro-
phenol biotransformation 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 fjM
PCP was biotransformed at 99.7 percent efficiency (Figure 1). Degradation of PCP was
accompanied by the production of all three tetrachlorophenols, as well as 3,4,5-, 2,4,5-, and 2,3,5-tri-
chlorophenol. 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 production of dichlorophenols, which were not included >'n the "sum."
2,3,4,5-Tetrachlorophenol (2,3,4,5-TeCP) reached a maximum concentration of 0.37 /im 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 concentration of 3.1 fjM 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 /*M) during the experiment.
The experiment demonstrates that the consortium acquired the ability to dechlorinate PCP
at all three chlorine positions after exposure to PCP for six months. Although dechlorination at the
ortho position is preferred by unacclimated 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 generally not been observed in previous laboratory
experiments, although it has been documented in rice paddy soils exposed to PCP (5,6).
Additionally, each of the tetrachlorophenols observed in this study were also reported in studies
performed examining PCP biodegradation 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 products. The overall pathway appears in Figure 2.
Development of a Reactor System to Control Environmental Conditions. A computer-interfaced
reactor system has been constructed and is now being refined. It provides control of pH, sulfide
concentrations, acetate concentrations, the apparent oxidation/reduction potential, and the biomass
concentration (there is little growth compared to the initial biomass in the reactor). Acetate and
sulfate are added with computer-controlled dispensers to change or maintain the concentrations of
these species.
The reactor system provides constant monitoring of the following parameters using an
analog-to-digital conversion board:
• pH with an Orion Ross glass electrode
-236-
-------
6.0
5.0
-~ 4.0
•z.
o
o:
o
o
3.0
2.0
1.0
0
0
Figure 1.
Sum
0.4 0.6 0.8
TIME (days)
1.0
1.2
Progress curve for pentachlprophenol reductive dechlorination by a pentachloro-
phenol-acclimated methanogenic consortium.
-237-
-------
OH
Figure 2. Summary of the observed reductive dechlorination pathway for pentachlorophenol
and its metabolites by a pentachlorophenol-acclimated methanogenic consortium.
-238-
-------
• Apparent EH at a platinum electrode
• Potential at a silver/sulfide ion selective electrode
The same Orion double junction Ag/AgCl electrode is used as the reference electrode for all the
above indicator 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 NajS at the end of data collection. Since there is no on-line sensor for acetate or
sulfate, samples are periodically withdrawn from the reactor to determine their concentrations by
ion chromatography.
The £„, sulfide electrode, and pH measurements for an experimental run are shown in
Figures 3 and 4. The apparent oxidation/reduction potential mirrors the sulfide 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
addition of acetic acid. Slug additions of acetic acid or sodium 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 compensate for acetic acid consumption.
SUMMARY AND CONCLUSIONS
Based on the results of preliminary experiments, the complete biotransformation pathway
observed for the reductive dechlorination of PCP by this acclimated consortium 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 contrast, only two para
dechlorination products (of seven possible products) and four meta dechlorination products (of nine
possible products) were observed. Of the possible polychlorinated phenolic congeners, only five are
not included in the pathway: 2,3-DCP, 2,5-DCP, 2,6-DCP, 2,3,4-TCP, and 2,3,6-TCP. It is important
to note that these five undetected di- and tri- chlorophenol 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 measurement of reductive dechlorination
kinetics while maintaining relatively constant biomass concentration, pH, sulfide concentration, and
apparent oxidation/reduction potential. Once the control of acetate and sulfate concentrations have
been refined, reductive dechlorination kinetics will be measured for varying, but constant,
environmental conditions.
-239-
-------
-0.10
-0.60
1000
2000
3000
Time (min)
4000
5000
6000
Apparent Redox *= Sullide
Figure 3. Apparent £„ and sulfide concentrations.
-240-
-------
6.0
1000
2000
3000
Time (min)
4000
5000
6000
Figure 4. pH measurements.
-241-
-------
ACKNOWLEDGEMENTS
Funding for this study was provided by the Office of Research and Development, U.S.
Environmental Protection 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 reductive dechlorination and mineralization
of pentachlorophenol 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 acclimated 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 chloromethoxybenzene 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
pentachlorophenol 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.
-242-
-------
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
and
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 performance, and the effectiveness
of pump-and-treat schemes. This distribution is difficult to measure, however, which limits
experimental understanding of the effect of entrapment processes on remediation. This study
investigated the use of x-ray computed tomography (CT) to measure nondestructively the spatial
distribution of trichloroethane (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/cmj with sand having a d50 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 CTunit 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 the 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 attenuation coefficients (/*). A theoretical
equation was developed to relate /* to mass TCA in each pixel. The equation requires an estimate
of the effective energy (E) of photons passing inside the core. The mean E was computed for this
particular core geometry and elemental composition 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 calculations, /x values were averaged
over groups of 6 x 6 pixels, producing computed masses for volume elements of 2 x 2 x 2 mm. Using
a mean E inside the core of 75 kev, the total computed mass was equal to or greater than 95 percent
-243-
-------
of the injected mass. Subsequent 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.
-244-
-------
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
Center for Interfacial Microbial Process Engineering
Montana State University
Bozeman, MT
with support from:
Great Plains and Rocky Mountain Hazardous Substance Research Center
Recent attention to bioremediation of contaminated surface water and aquifer systems has
necessitated the development of methods to estimate in situ rates of biodegradation. Specifically,
biodegradation kinetic parameters 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 (BOD) respirometry
apparatus to determine the kinetics of microbial degradation within the batch BOD vessel. Methods
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 /*„„, K,, biomass yield, initial biomass concentration, and cell decay. Mean
values and 95 percent confidence intervals (CIs) were determined for these kinetic parameters. The
kinetic parameters were then used in a bioprocess model to estimate the extent of biotransformation
activity at a field site. To assess the effects of kinetic parameter variation on the size of the area
biotransformed, the values of two important parameters (fimax and KJ were varied through their CIs.
Results indicate that the area biotransformed in a 4,800-day model run varies only slightly from the
least favorable (low /*„,„, high KJ to the most favorable (high pmax, low KJ microbial kinetic
condition.
-245-
-------
DISSIPATION OF POLYCYCLIC AROMATIC HYDROCARBONS
IN THE RHIZOSPHERE
M. Katherine Banks1 and A. Paul Schwab2
Great Plains and Rocky Mountain Hazardous Substance Research Center
Department of Civil Engineering
Department of Agronomy
Kansas State University
Manhattan, KS
Vegetation may play an important role in the biodegradation 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 microflora.
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 saliva), fescue (Festuca arundinacea), big bluestem (Andropogon
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 of the
plant samples, and PAH concentrations were smaller in rhizosphere 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 results indicate that the
interaction between plants and rhizosphere microflora enhances remediation of soils contaminated
with hazardous organics.
-246-
-------
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 polycyclic aromatic
hydrocarbons (PAHs) from Tamar Estuary, U.K., 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 isotherm yielded an organic carbon-based partition coefficient (K^.)
of 793 cnij/g, 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 hysteresis phenomenon was observed for naphthalene adsorption/
desorption (i.e., the adsorbed naphthalene was not readily 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 constants 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 soil/sediment and rendered less available for extraction, chemical, and biological
transformation.
-247-
-------
REFERENCES
1. Steinberg, S.M., J J. Pignatello, and B.L. Sawhney. 1987. Persistence of 1,2-dibromoethane
in soils: Entrapment in intraparticle micropores. Environ. Sci. Technol. 21:1201-1208.
2. Readman, J.W., and R.F.C. Mantoura. 1987. A record of polycyclic aromatic hydrocarbon
(PAH) pollution 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 sediments. 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 se^' ment particles. 24:1373-1383.
5. Brusseau, M.L., and P.S.C. Kao. 1989. Sorption nonideality during organic contaminant
transport in porous media. Grit. Rev. Environ. Control. 19:33-99.
-248-
-------
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 conventional activated sludge systems, including greater flexibility in meeting
changes in feed conditions, effective control of the process and the quality of the discharge, 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 performance of the unit;
the second involves identifying the interactions among pollutants at the kinetic level and then
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
(ATCC14235). Substrate concentrations were monitored using high pressure liquid chromatography
(HPLC) measurements. Total biomass concentration measurements were based on optical density,
while biomass concentrations of individual 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 operation of the unit. The model predicts and the
experiments verify that under certain conditions of operation multiple types of reactor behavior are
possible, as determined by conditions at start-up of the unit.
-249-
-------
Phenol and 4-chlorophenol (4CP) are being used as model compounds for studying the
effects of substrate interactions. Two cultures have been tested. One was not capable of degrading
4CP unless phenol was present; this cometabolic process led to complete mineralization of 4CP. The
second culture was capable of completely mineralizing both phenol and 4CP, but each compound
affects the degradation of the other. These cross-inhibitory kinetics have been mathematically
described, and the SBR model predicts the best values for the operating parameters. Experiments
are under way to verify the model's predictions.
• U.S GOVERNMENT PRINTING OFRCE 1993.750- ooz/ '60169
-250-
------- |