&EPA
United States
Environmental Protection
Agency
National Health and Environmental
Effects Research Laboratory
Gulf Breeze, FL 32561
Research and Development
EPA/600/S-98/001 April 1998
ENVIRONMENTAL
RESEARCH BRIEF
Bioaugmentation with Burkholderia cepacia PR1301 for
In Situ Bioremediation of Trichloroethylene Contaminated Groundwater
Perry L. McCarty1, Gary D. Hopkins1, Junko Munakata-Marr1, V. Grace Matheson2,
Mark E. Dolan1, Louise B. Dion1, Malcolm Shields3, Larry J. Forney2, and James M. Tiedje2
A pilot field study was conducted at the Moffett Federal Airfield,
Mountain View, CA, to determine whether effective in situ
aerobic cometabolic biodegradation of trichloroethylene (TCE)
could be accomplished through bioaugmentation with a geneti-
cally modified strain of Burkholderia cepacia G4 (G4) together
with feeding of lactate to serve as an energy and growth
substrate for the organism. Strain G4 is highly effective at TCE
cometabolism but requires either phenol or toluene to induce
oxygenase enzyme activity. A strain of G4 was developed
through NTG mutagenesis that constitutively expresses the
oxygenase so that no inducer need be added. The modified
strain, B. cepacia PR130r (PR130I), can degrade TCE effec-
tively while growing on simple substrates such as lactate.
Strain-specific molecular probes were developed for monitoring
the presence and movement of PR1301 and were based upon
rep-PCR analysis.
Laboratory microcosm studies using Moffett aquifer material
indicated that the quantity of microorganisms that could be
injected was limited by oxygen availability. Within these limits,
addition of wild-type G4 grown on phenol or mutant strains
grown on lactate were effective initially at TCE cometabolism
when added daily to the columns. However, the strains with
constitutive oxygenase expression did not maintain TCE
degradative ability for long when lactate was used. Organism
1 Western Region Hazardous Substance Research Center, Stanford University,
Palo Alto, CA.
2 Center for Microbial Ecology, Michigan State University, East Lansing, Ml.
3 University of West Florida, Pensacola, FL.
presence in the microcosm effluents was found as long as
bioaugmentation continued, but not when it was discontinued.
Following discontinuation of bioaugmentation, microcosms fed
phenol improved in TCE cometabolism with time, reaching
over 90% removal, although at the end of the study, no G4 or
related mutant organisms could be found within the micro-
cosm. In contrast, a control microcosm fed phenol with each
exchange, but without bioaugmentation, removed about 60%
TCE initially, but with time, TCE removal efficiency decreased
to near zero.
Three field studies with bioaugmentation were conducted. In
each one, bioaugmentation with PR1301 along with lactate addi-
tion was initially reasonably effective at TCE removal. In addi-
tion, phenol was rapidly consumed to near detection limits
when added following a short period of bioaugmentation with
PR1301, and lactate addition, demonstrating that an initial phase
of bioaugmentation could be effective to establish a population
of phenol or toluene degraders if these substrates were to be
used at a site. However, in both the second and third field
trials, bioaugmentation with lactate feed alone in time became
ineffective at TCE removal, reaching near zero removal within
a few weeks. The inability of PR1301 to remain effective in
bioaugmented groundwater was also demonstrated in a brief
laboratory study using groundwater from the bioaugmented
well after the conclusion of the field study. This suggested
either predation of the introduced population or the inability of
PR130, to effectively compete for the added lactate, or perhaps
both, were the cause of the eventual failure of the system. In
order for bioaugmentation for TCE cometabolism to be suc-
cessful, methods for avoiding this competitive problem need to
be found.
Printed on Recycled Paper
-------
Introduction
Trichloroethylene (TCE) has been widely used as a solvent
over the past 50 years. Because of uninformed disposal prac-
tices, it has become a major groundwater contaminant. There
has been much interest in the potential of aerobic in situ
blotransformation processes for the destruction of TCE and
other chlorinated aliphatic hydrocarbons (CAHs) in groundwa-
ter since cometabolism of TCE was first demonstrated in soil
columns where natural gas and oxygen were added to stimu-
late the growth of native microorganisms. In cometabolism, an
enzyme (oxygenase), used by the microorganisms for initiating
primary substrate oxidation, fortuitously transforms many CAHs.
Field-scale evaluations of in situ biodegradation of CAHs have
been undertaken since 1985 at the Moffett Federal Airfield
(Moffett Field), Mountain View, CA. Initially, methane was used
as a primary substrate for aerobic cometabolism of several
CAHs. While the methane-consuming culture developed was
highly successful at transforming some CAHs, removal effi-
ciency was rather low with TCE. Therefore, other potential
inducers were sought. One of the most promising was phenol.
Phenol was then evaluated at Moffett Field over three seasons
using indigenous microorganisms only and was found to be
quite superior to methane for in situ TCE degradation, provid-
ing up to 90% removal at TCE concentrations of up to 1 mg/l.
However, the use of phenol might pose regulatory problems
because of its known toxicity and taste and odor potential. A
possible alternative is the use of bioaugmentation with strains
of bacteria in which the oxygenase is constitutive, thus they
could not only degrade TCE, but also could do so without the
requirement for an inducing compound. Such bacteria may
grow on harmless water-soluble substrates and still maintain
their ability to transform TCE. However, such substrates may
be less selective and the introduced organisms may encounter
strong competition from indigenous organisms, resulting in little
TCE degradation. A possible approach to overcome this limita-
tion is to use daily bioaugmentation in an attempt to maintain a
competitive advantage. This was the approach evaluated in
this study. An additional interest is that stimulation of native
organisms with a specific substrate for in situ bioremediation of
TCE may enrich for a population of microorganisms that are
unable to cometabolize the target compound or that will de-
grade the target compound slowly. Bioaugmentation through
addition of bacterial cultures known to transform TCE rapidly
may enhance native biodegradation or even provide the sole
means of degradation in systems without indigenous TCE-
degrading organisms.
Tn5 mutagenesis was previously demonstrated to result in the
production of a constitutive TCE-degrading strain, but the in-
sertion of additional genetic information, particularly antibiotic
resistance, in this recombinant strain may subject its release to
strict regulatory and public approval. As an alternative, a non-
revertible regulatory mutant selected for spontaneous constitu-
tive TCE transformation through N-methyl-N'-nitro-N-
nitrosoguanidine (NTG) mutagenesis was produced in this study
and tested both in the laboratory and at Moffett field.
Objectives of Study
The objectives of this study were (1) to evaluate at laboratory
and field scale the potential for bioaugmentation with a bacte-
rial mutant containing a constitutive monooxygenase to en-
hance and improve in situ bioremediation of groundwater con-
taminated with TCE, (2) to determine the movement, fate, and
effectiveness of introduced microorganisms in an aquifer, (3) to
evaluate the value of introduced microorganisms and of
bioaugmentation for enhancing in situ biodegradation, and (4)
to evaluate the applicability of molecular tools in the monitor-
ing, operation, and control of in situ bioreclamation systems.
Overview of Study
The study was conducted by researchers at the Western Re-
gion Hazardous Substance Research Center (WRHSRC),
Stanford University; the Center for Microbial Ecology (CME),
Michigan State University; the University of West Florida; and
the U.S. Environmental Protection Agency, Gulf Breeze Envi-
ronmental Research Laboratory. The microorganism devel-
oped at the University of West Florida and used for
bioaugmentation was B. cepacia G4 PR1301 (PR1301), a non-
recombinant derivative of B. cepacia G4 (G4) that constitutively
expresses toluene ortho-monooxygenase (TOM) and is highly
effective at TCE cometabolism. Molecular probes for monitor-
ing the fate and effects of PR1 were developed at CME. The
laboratory microcosm studies and field studies were conducted
by WRHSRC, with molecular probe analysis of field samples to
determine the movement and fate of PR1301 conducted by
CME.
Methods
Development of PR1301 and Probes for Its
Detection
A non-recombinant strain (PR130])) capable of constitutive TCE
degradation was developed for this study. Bacterial cultures
used for PR1301, development were grown on two formulations
of media based on a basal salts minimal medium (BSM), BSM-
lactate (BSM, 20 mM lactate) and BSM-phenol-TTC (BSM,
0.025 mg/ml triphenyl tetrazolium chloride (TTC), 0.2 mg/ml
proteose peptone, 2 mM phenol). NTG mutagenesis and en-
richment for Tol-, Phe- mutants using a toluene vapor feeder
was performed as previously described. Tol- and Phe- mutants
were detected using their TTC dye-reduction assay with the
following changes: phenol (2 mM) was used as the primary
carbon source instead of toluene vapor, and 0.2 mg/ml pro-
teose peptone was added to the BSM-phenol-TTC purified
agar plates. Mutagenized cells were diluted to give approxi-
mately 150 colonies per 100 p.L plated. Assays for TOM through
the oxidation of trifluoromethyl phenol (TFMP) to the yellow
trifluoromethyl heptadienoic acid (TFHA) were performed as
previously described.
Rapid strain-specific nucleic acid probes for detecting B. cepacia
G4 (G4), including the above mutant PR1301 strain as well,
were developed for use in monitoring movement and survival
of PR130, and other G4 strains in laboratory and field micro-
cosms. The probes were made by cloning DNA fragments
amplified from genomic DNA of G4 using rep-PCR and primers
specific for repetitive extragenic palindromic (REP) sequences.
The specificity of the probes was determined by hybridization
against DNA fragments amplified from G4 and 80 genetically
distinct bacterial isolates from the Moffett field aquifer. Two out
of four probes tested were found to specifically hybridize to
DNA fragments of the expected size in the rep-PCR fingerprint
of G4, but not to the other strains tested. One of these probes,
a 650 bp fragment, produced a hybridization signal when as
few as 10 CPU of G4 were present in a mixture with 105 CPU
nontarget strains, indicating that the sensitivity of these probes
was comparable to those of other PCR-based detection meth-
ods. The probes were used to discriminate groundwater and
microcosm samples that contained G4 from those that did not.
-------
False positive results were obtained with a few samples, but
these were readily identified by using hybridization to the
second probe as a confirmatory step. The general applicability
of the method was demonstrated by constructing probes spe-
cific to three other environmental isolates.
Laboratory Microcosm Studies
In preparation for field studies, the effects of bioaugmentation
on the aerobic cometabolism of TCE in groundwater were
initially investigated in the laboratory using small-column aqui-
fer microcosms (17 ml total volume, 5 - 6.5 ml pore volume)
containing Moffett field aquifer material. In initial studies, non-
sterile non-bioaugmented microcosms fed phenol as a primary
substrate mimicked observed in situ behavior at the Moffett
field site, cometabolizing approximately 60 ja.g/1 TCE while fed
6.5 mg/l phenol. High density single bioaugmentation with G4
increased TCE removal in sterile aquifer material, while pro-
ducing mixed results in non-sterile material. Low density semi-
continuous bioaugmentation enhanced TCE transformation in
non-sterile microcosms. Phenol-fed microcosms augmented
with either G4 or PR1301 transformed twice as much TCE as the
non-augmented phenol-fed microcosm. In addition, should pri-
mary substrate addition be a regulatory concern, TCE degra-
dation was observed without primary substrate addition through
bioaugmentation using organisms expressing the TCE-trans-
forming enzyme.
In subsequent long-term studies, aquifer microcosms were
repeatedly bioaugmented and fed solutions containing 6.5 mg/
I phenol or 15 mg/l lactate and 250 ng/l TCE every two to three
days. The effectiveness of TCE cometabolism by an indig-
enous phenol-fed microbial population declined significantly
during a 280-day experiment. This behavior, possibly due to
the negative selective pressure of TCE cometabolism which
leads to the formation of toxic products, had not been observed
previously in shorter-term TCE transformation experiments.
The addition of G4 or PR1301 to microcosms along with phenol
or lactate initially allowed for substantial TCE degradation but
led to the eventual depletion of dissolved oxygen and a decline
in TCE transformation. After termination of bioaugmentation,
dissolved oxygen levels recovered in all microcosms, and those
microcosms that continued to receive phenol returned to or
surpassed previous TCE transformation levels, while unfed
and lactate-fed microcosms lost degradative activity. The intro-
duced organisms, however, did not appear to be responsible
for the recovered TCE degradation in the phenol-fed, formerly
bioaugmented microcosms. The source of activity in these
microcosms was not identified but is likely to have been effi-
cient TCE-transforming indigenous organisms selected by the
operating conditions within the microcosms. G4 and PR1 were
never found present in the effluents from the non-bioaugmented
columns but were repeatedly found in column effluents during
83 days of active bioaugmentation in bioaugmented columns.
However, within 10 days after bioaugmentation was stopped,
neither G4 nor PR1 were again detected. The greatly improved
TCE biodegradation that occurred in the previously
bioaugmented phenol-fed columns through the remainder of
operation to day 237 raised the question of whether the
bioaugmented organisms may be growing in the column. How-
ever, when they were dismantled and the column contents
were analyzed for G4, none was found present. We were
unable to confirm why the improved performance occurred in
the phenol-fed bioaugmented columns after bioaugmentation
was stopped. The lack of effective long-term TCE biodegrada-
tion in lactate-fed bioaugmented microcosms raised initial ques-
tions about the potential for successful field application.
Field Bioaugmentation Studies with PR1301
The field study of bioaugmentation for in situ bioremediation of
TCE was conducted at Moffett Field. This location was the site
of several previous studies of in situ biodegradation of CAHs.
In all cases, the previous studies made use of indigenous
microorganisms. The last previous study conducted at this site
was concerned with injection of phenol and toluene to stimu-
late indigenous microorganisms for cometabolism of TCE. It
was found that the dominant organisms utilized toluene ortho-
monooxygenase (TOM) for primary substrate and TCE oxida-
tion. Since the organism proposed to be used for
bioaugmentation, B. cepacia PR1?01 also produces TOM, a
new series of injection and monitoring wells was developed to
avoid the potential interference to this study from native micro-
organisms previously stimulated with phenol and toluene addi-
tion. However, the analytical system and all other features of
the test site were similar to those used in the past.
The in situ evaluation was performed using the same method-
ology as in our previous studies. Here, a series of stimulus-
response tests were performed under induced gradient condi-
tions of injection and extraction of groundwater. The stimulus
was the injection of groundwater that was blended with the
microorganism and chemicals of interest. The response was
the concentration history of the chemicals at the monitoring
locations. The profile of the field system is illustrated in Figure
1. This system consisted of an injection well (2SSEI) and an
extraction well (P2) located 9 m apart with three monitoring
wells in between. Both the injection and extraction wells con-
sist of standard (ca. 5-cm diam.) polyvinyl chloride pipe in-
stalled by using a hollow-stem auger. The wells contained 1.5-
m long slotted screens, installed 4.5-6.0 m below the ground
surface, and fully penetrated the aquifer zone that contained
sand and gravel. The sampling wells consisted of 3.18-cm
diam. stainless steel wire-wound sand points with 0.6 m screens.
The screen sections were located 4.7-5.3 m below the ground
surface, in the center of the sand-gravel layer of the aquifer.
These three wells were spaced 1, 2, and 3.5 m from the
injection well. In addition, two special monitoring wells 2SSE
(FP1) and 2SSE (FP2) fully penetrated the aquifer and were
located 0.5 m and 1.5 m, respectively, from the injection well.
Into these wells were inserted a series of nylon bags contain-
ing glass beads that fully covered the depth of the groundwater
aquifer. These beads were used for organism colonization to
determine the presence of the injected strain PR13or The
formation groundwater is moderately saline, having a total
dissolved solids content of 1,500 mg/l, and was contaminated
by some chlorinated aliphatic hydrocarbons (CAHs), mainly
1,1,1-trichloroethane (TCA), but was devoid of chlorinated
ethenes the subject of this study. Thus, the target compound,
TCE, was added to injection water in a controlled manner.
Nitrate was present in the native groundwater (25 mg/l as
measured by ion chromatography) and served as a source of
nitrogen nutrient. Total phosphorus concentrations were low
(<0.1 mg/l, as measured by induced-coupled argon plasma
spectrometry), but near solubility limits of common phosphorus
minerals, which were probably the source of the needed phos-
phorus.
-------
Depth Below Ground Surface (m)
O) -&. M O
i i I I
>
Cl
>
i
Sa
1
Injection
Well f Sampling Wells
iy
I
\.
id and
f Gravel
Extra
-^ W
i ,
Aquifer
ction
ell
k
4 2SSE1 2SSE1 2SSE2 2SSE3 P2
C|' 2SSE(FP1) 2SSE(FP2)
1 1 1 1 1 1
2345
Distance from Injection Well (m)
Figure 1. Cross-sectional view of in situ bioaugmentation site at Moffett Field.
Chemical Introduction into the Aquifer
Chemicals were introduced into the injected water continuously
or in an automatic programmed manner. The extracted water
used for injection was treated before chemicals were added by
filtration through a nominal 1 UJTI filter, and UV disinfection.
Oxygen was introduced using a counter-current column for gas
transfer into the injection water. TCE, phenol, and lactate were
added to injection water by pumping water solutions containing
these chemicals. The injection water and sampled waters all
passed through stainless-steel tubing, which prevented pas-
sage of gasses and TCE through tubing walls. Thus, excellent
mass balances of all chemicals were maintained through the
system.
Field Analytical System
Water samples for analysis were obtained from the monitoring
well locations and from injected and extracted water by auto-
mated pumping to an automated data acquisition and control
system (DAC). This permitted the continuous measurement of
the principal chemical constituents, which were bromide tracer,
phenol, toluene, TCE, dissolved oxygen (DO), and pH. The
instruments operated by the DAC system are an ion chromato-
graph for the bromide tracer analysis, a reverse phase high
performance liquid chromatograph (HPLC) for phenol and tolu-
ene analyses, an anion exclusion HPLC with conductivity de-
tector for lactate, a gas chromatograph equipped with an elec-
tron capture (GC-ECD) and a Hall conductivity detector (GC-
Hall) for TCE analysis or a PID detector for toluene analysis, a
dissolved oxygen meter (Yellow Springs, OH), and a pH meter.
The lower concentration limits for the analyses were DO, 0.1
mg/l; bromide, 0.5 mg/1; TCE, 0.5 u.g/1; lactate, 0.5 mg/l; and
phenol, 1 u.g/1. All data were compiled automatically and stored
in a database on a personal computer at the test site.
Growth of Bioaugmentation Culture
A standard procedure was used for the growth of PR1301 used
in bioaugmentation for the first two field trials, and a modified
procedure was used for the third field trial. During the first two
field trials a culture of PR1301 was transferred from a broth tube
to 1 liter of media contained in a 2-liter flask. The media
consisted of 4 u.g/1 of sodium lactate in mineral media. These
were introduced into the flask through a foam plug, and the
mixture was aerated for 48 hr at room temperature in the
instrument-control room in the field. Aseptic conditions for the
microorganism were maintained throughout. The culture grown
in the flask was then transferred to 16 liters of similar media,
but containing 1 ug/l of sodium lactate in a carboy in which
pure oxygen was bubbled for a 24-hr period. Samples were
taken for suspended solids, lactate, and PRI analyses. The
contents of the carboy were then pumped daily into the ground-
water through the injection well at the normal injection flow rate
of 1.5 liter/min.
During the winter season of the second field study, all lactate in
the carboy was not always used, and suspended solids, repre-
senting PR1 growth, were somewhat lower than during the
normal season. For this reason, during the third field study,
conducted in the summer, organisms were grown under more
controlled conditions in the laboratory and then transported to
the field. The transportation time was less than 30 minutes.
Similar procedures were followed except that air was used for
mixing and growth of microorganisms in the carboys rather
than pure oxygen. In addition, two 16-liter carboys of microor-
ganisms were added per day rather than one in order to obtain
a more definitive study of the effectiveness of bioaugmentation.
Field Study Results
Three separate studies were conducted of bioaugmentation
with PR1301. It was originally planned to conduct one single
long-term study, but heavy rains occurred twice during the
study, resulting in excessive groundwater flow and hydraulic
head so that the study had to be interrupted. Nevertheless,
sufficient data were obtained from each separate study to allow
evaluation of the effectiveness of bioaugmentation. In addition,
-------
some aspects of the studies were repeated to provide more
conclusive results.
Field Bioaugmentation Evaluation
A summary of the experimental variables used during the three
separate field tests is given in Table 1. In all cases, 10 liters/
min of groundwater was extracted from the aquifer and 1.5
liter/min of air stripped, filtered, and disinfected groundwater,
augmented with phenol or lactate, and TCE was added to the
injection well. The primary substrates (lactate or phenol) were
pulse-injected three times a day as in previous studies. Here,
either lactate or phenol was added over a period of 15 to 30
min to provide time-averaged concentrations over the 8-hr
period as indicated in Table 1. DO, bromide, and TCE were
added continuously at the concentrations indicated.
Bioaugmentation through daily injection of 16 liters during the
first two field studies or 32 liters during the second was carried
out during the periods indicated.
The first study was conducted for 26 days or a little over 600
hr. The second study was conducted over 34 days, or about
800 hr. The third study was conducted for 40 days, or about
950 hr. During the first study, conditions were maintained
constant the entire period with lactate used throughout as the
primary substrate. During the second study, lactate was added
as the primary substrates for the initial 130 hr, the primary
substrate was then switched to phenol for up to 387 hr, and
then the primary substrate was witched back to lactate. In the
final study, lactate was used exclusively during bioaugmentation.
However, bioaugmentation was stopped after 530 hr. Phenol
was substituted for lactate as the primary substrate after 830
hr.
First Field Study
Both daily bioaugmentation and lactate addition at a time-
averaged concentration of 13 mg/l began at time zero. Figure 2
indicates the normalized bromide concentration measured at
the three monitoring and the one extraction wells. Based upon
time for 50% arrival, the time of movement of injected water
from the injection well to the first monitoring well 2SSE1 was
about 6 hr. Movement to the second (2SSE2) and third (2SSE3)
monitoring wells was about the same or about 18 hr. After
about 50% bromide arrival occurred, the bromide concentration
increased more rapidly at 2SSE2 than at 2SSE3, as expected*
The peculiar behavior of similar times for 50% arrival, but
divergent times for 100% arrival at the second and third moni-
toring wells had occurred previously in the first or south leg of
wells constructed at the Moffett Field site some 10 years
earlier. More normal arrival times were obtained in a southeast
leg that was constructed between the first study and this study.
This peculiar behavior apparently resulted from heterogeneities
occurring in the aquifer. The important feature, however, is that
bromide concentration approached 100% with time at all moni-
toring wells, a requirement to be able to adequately evaluate
removal efficiency for TCE. The concentration of bromide at
the extraction well was 12% to 13% of that in the injection
water, which is as it should be with an extraction rate 8 times
that of the injection rate. The bromide tracer studies indicate
the newly constructed leg was satisfactory for the
bioaugmentation study to proceed.
Figure 3 illustrates lactate concentration at the various monitor-
ing wells. Lactate was injected in pulses three times per day to
give the average time-averaged injection concentration of 13
mg/l. Some removal of lactate began immediately, as the
concentration at the second and third monitoring wells never
exceeded 3 mg/l, declining to non-detectable levels after about
two days of injection. Some lactate, too, was found at the first
monitoring well location for the first 130 hr of injection, and
then became non-detectable after that time. The lactate was
consumed readily by the microflora either added or already
existing in the aquifer, with most removal occurring within the
short distance between the injection well and the first monitor-
ing well.
The pH at the three monitoring wells was similar and varied
from about 7.1 at the beginning of bioaugmentation to about
6.9 - 7.0 at the end. The injection well gage pressure remained
the same at about 4.9 pounds per square inch throughout the
study until the rain began. This indicates that excessive
bioclogging did not occur throughout the bioaugmentation pe-
riod.
Figure 4 illustrates DO concentration in the injection well, the
monitoring wells, and the extraction well. Most DO demand
occurred between the injection and the first extraction well with
steady-state DO consumption after 400 hr equaling about 16
Table 1. Experimental Variables for the Three Field Studies of Bioaugmentation
Injection Concentration, mg/l
Field Hours of
Study Study
1 0
0
2 130
387
3 0
520
830
- 624
- 130
- 387
- 800
- 520
- 830
- 950
Bioaugmentation
(9/d)
5.0 + 0.9
3.5 ±1
3.5 + 1
3.5 ±1
10.5 ±1
0
0
.3
.3
.3
.3
Primary
Substrate
Lactate
Lactate
Phenol
Lactate
Lactate
Lactate
Phenol
Primary
Substrate
13
13
6
13
13
13
6
TCE
0.08
0.10
0.10
0.10
0.10
0.1.0
0.10
DO
32
32
32
32
32
32
32
Br
60
60
60
60
60
60
60
-------
o
2SSE1
2SSE2
2SSE3
Extract
0.0
100 200
300 400
Time, hr
500 600 700
Figure 2. Normalized bromide tracer concentration at the various monitoring wells following continuous injection of 60 mg/l beginning at time
zero during the first field study.
15'
10-
5-
20 40 60 80 100 120 140
Time, hrs
Figure 3. Lactate concentration versus time at the various monitoring wells following 13 mg/l continuous lactate addition beginning at time zero
during the first field study. Beyond 130 hours concentrations were all below the detection limit.
-------
i*-****1*^^
100
200
300 400
Time, hrs
500
600
700
Figure 4. Dissolved oxygen concentration at the various monitoring locations during the first field study. Dissolved oxygen in native
groundwater was near zero.
mg/l. The oxygen demand from complete oxidation of the 13.5
mg/l of added lactate would be about 15 mg/l. Some oxygen
demand may have resulted from oxidation of natural organics
or inorganics occurring in.the aquifer, but most of the excess
probably was due to oxygen demand by the injected microor-
ganisms. As not all of the lactate in the carboy was used, there
would have been some oxygen demand from the remaining
lactate as well.
The normalized TCE concentrations at the three monitoring
wells and the extraction well are indicated in Figure 5. The
TCE injection concentration averaged about 70 u,g/l. It is some-
what difficult to interpret this information because of the very
strong sorption of TCE to aquifer material that occurs at this
test site. TCE needs to reach steady state removal before
adequate evaluation of removal effectiveness can be made.
Since most injected organisms would probably reside between
the injection well and the first monitoring well, and most of the
lactate was used here, the greatest portion of TCE removal
should occur within this zone as has been found from previous
studies. During the first 80 hr, TCE at 2SSE1 appeared to
approach a steady-state concentration of about 22% of the
injected concentration. However, after this time, TCE concen-
tration increased significantly and approached 80% of the in-
jected value between 280 and 450 hr. After that, removal
appeared to increase with the TCE concentration at 2SSE1
equaling about 60% of the injected concentration. These data
would indicate that bioaugmentation was moderately success-
ful. At about 600 hr a rainstorm occurred that ended the first
field study. Based upon these results, it appeared that
bioaugmentation at the given level with PR130], and feeding of
lactate resulted in successful removal of about 50% or more of
the injected TCE.
Monitoring was conducted for PR1391, through samples taken
daily from the three monitoring wells. Once bioaugmentation
began, PR1301 was detected in samples taken from the first
monitoring well during the first six days of bioaugmentation.
None was detected in water taken from the second or third
monitoring wells. However, after six days PR1301 was no longer
detected until almost the end of this field trial, and then was
found on only two other occasions. These two other occasions
occurred after 480 hr of operation at times when TCE removal
appeared to increase somewhat. No PR1301 was detected on
the glass beads removed at the end of this field trial.
Second Field Study
Once the field system had returned to normal following the end
of the rains, the second field trial was begun. As in the first
study, bioaugmentation was started at time zero with addition
of organisms once per day. The lactate was fed at the same
time-averaged concentration as previously. The objective of
the first five days was to see whether removals observed
during the first study would be seen again. Within the first two
days, TCE removal between the injection and the first monitor-
ing well appeared to level off at about 50% as indicated in
Figure 6. The average injected TCE concentration was about
80 jig/l.
After 125 hr the primary substrate was changed to a time
averaged 6 mg/l phenol to determine whether bioaugmentation
might be used to introduce a population of phenol degrading
organisms that would prevent phenol spread in an aquifer if it
were the primary substrate of choice for field implementation.
This concentration of phenol had been found from previous
studies to provide between 50% and 75% removal of TCE.
-------
2SSE1
2SSE2
2SSE3
Extract
0 100 200 300 400 500 600 700
0.0
Figure 5. Normalized TOE concentrations at the various monitoring locations during the first field study. Added TCE concentration was 70 ng/l.
2SSE1
2SSE2
2SSE3
Extract
0 100 200 300 400 500 600 700 800
0.0
Time, hrs
Figure 6. Normalized TCE concentrations at the various monitoring locations during the second field study. Added TCE concentration was 80
ng/l.
-------
Figure 6 indicates that as soon as phenol was added, stable
TCE removal of about 70% was obtained between the injection
and the first monitoring well. The stable removal over the next
250 hr demonstrated the benefit of phenol addition along with
bioaugmentation.
Phenol measurements indicated that during the first 24 hr,
about 0.5 mg/1 phenol was detected at the second and third
monitoring wells, but by the second day concentrations had
dropped below the detection limit of about 1 |ig/l. Some phenol
was found at the first monitoring well. This brief study indicates
that indeed bioaugmentation can be very effective for rapidly
establishing a phenol-using population to prevent the undes-
ired spread of phenol in the aquifer when first added. During
the previous study when phenol was first injected into this
aquifer, about two weeks were required for an indigenous
population to grow sufficiently for significant phenol degrada-
tion to occur. Then, several days were required for the phenol
concentration to drop below the detection limit. In this study,
phenol had never been injected into the aquifer at this location
and so the immediate response with phenol removal was
undoubtedly a result of the bioaugmentation. Thus, such
bioaugmentation would be very beneficial at the start of in situ
bioremediation where toxic compounds such as phenol or
toluene may be injected.
After 390 hr, a switch was made back from phenol to lactate to
determine whether the fairly good TCE removal obtained with
lactate bioaugmentation could be maintained. The response
shown in Figure 6 indicates TCE concentration then increased
at all monitoring locations and, within a few hundred hours,
approached the injection concentration. Therefore, these re-
sults indicate that bioaugmentation was no longer effective. DO
measurements indicated DO consumption between the injec-
tion point and the second monitoring points was about 24 mg/l,
which is significantly higher than the 16 mg/l obtained in the
first study. This probably resulted because all of the lactate in
the carboys was not consumed during this period so that a
greater amount of oxygen demanding material was introduced
into the aquifer. However, DO remained above 8 mg/l at all
monitoring wells throughout the study and so was sufficient for
TCE oxidation.
An important question arose as to why the bioaugmentation
was not effective at the end of the second field study. One
hypothesis was that the outside temperatures had decreased
as this was the winter season, and conditions for growth were
not as optimal as in the fall study. As a result, the quantity of
microorganisms injected per day dropped from 5 grams to 3.5
grams. The aquifer temperature itself remained constant at
about 18°C. Additionally, sporadic rain occurred after about
300 hr and this was associated with an increase in injection
pressure. While this might be a result of bioaugmentation, our
results from the previous study suggest that it was more due to
the rainfall causing an increase in the static hydraulic pressure
in this confined aquifer. Heavy rainfall occurring around 800 hr
terminated the second field study.
As before, samples were obtained from the various monitoring
wells for analysis for the presence of PR1301. PR130,> was
detected in the first monitoring well every day for the first four
days, and then it was not detected except for two times during
phenol injection at around 300 hr. None was found during the
last period when lactate was added after 390 hr. PR1301 glass
beads obtained at the end of the lactate feed and at the end of
the phenol feed.
Third Field Study
A third field study was begun the following summer when there
was no concern that rainfall might occur to adversely affect the
results. In an attempt to better confirm the previous TCE
removal associated with bioaugmentation and lactate feed,
cultures were grown in the laboratory under more ideal condi-
tions, and the amount of PR1301 for injection was doubled.
Here, two 16-liter carboys containing a total of about 10 g dry
weight of PR1,01 was used per day. Laboratory measurements
of TCE degradation rates by the culture were obtained daily to
ensure that the injected population was efficient in TCE bio-
degradation.
TCE removal during the third study is indicated in the normal-
ized TCE concentration graph shown in Figure 7. During the
first 124 hr, TCE removal looked very good, mimicking the
removal obtained initially in the first two studies, and seemed
to reach a steady state at about 80% to 90%. Then after 130
hr, the TCE concentration began to increase in a manner
suggesting TCE removal had stopped, and that only sorption
was affecting TCE removal. The upward TCE trend continued
through 520 hr when bioaugmentation was stopped. It contin-
ued to increase while only lactate was added until 830 hr,
when the primary substrate was switched to 6 mg/l phenol.
TCE removal then began, with the TCE concentration at all
locations continuing on a downward trend reaching about 50%
removal when the third field study was stopped after 950 hr of
operation.
DO concentration measurements indicated DO consumption
increased throughout the study, and by 500 hr equaled about
30 mg/l, leaving only about 7 mg/l at 2SSE1 and 2.5 mg/L at
2SSE2. This increase in oxygen consumption with time, greater
than that from the previous studies, can be attributed primarily
to oxidation of the biomass that had been added to the system.
When bioaugmentation stopped, DO concentration began to
rise immediately. In spite of the large DO demand, sufficient
DO was always present in the aquifer and so was not the
cause of the poor TCE cometabolism observed during
bioaugmentation and lactate addition.
Conclusion
Laboratory Study of PR1301 Survival
The relatively good TCE removal with bioaugmentation and
lactate addition found in the first field study and during the first
few days of the second and third studies, followed by poor TCE
removal during later periods of the last two studies, suggested
that a population of predators to PR1301 may have developed in
the field. This was also suggested by the fact that arrival of
PR1301 at the monitoring wells occurred during the first few
days of bioaugmentation, but rarely after that. In order to obtain
some confirmation of the possibility of a predator population
causing this problem, samples of groundwater and aquifer
organisms were obtained at the end of the field study by rapid
groundwater withdrawal and its collection from the injection
well where the population was expected to be the highest. This
resulted in a groundwater sample containing about 2,000 mg/l
total suspended solids. Microscopic observation of the ex-
tracted water with about 2,000 mg/l total suspended solids,
indicated a high population existed of motile bacteria that were
not representative of PR1301. A few small ciliated protozoa
were also observed.
-------
1.20
1.00-
0.80-
0.60-
0.40-
0.20-
0.00
200
400
600
800
1000
Time (Hours)
Figure 7. Normalized TCE concentrations at the various monitoring locations during the third field study. Added TCE concentration was 100 ng/l.
Bioaugmentation was conducted for the first 520 hr only, 13 mg/l lactate was added for the first 830 hr, and 6 mg/l phenol was added
during the last 120 hr.
Two separate studies were conducted to compare the ability of
PRIao, cultures to persist in the above groundwater. In the first,
several individual bottles, each containing 25 ml of PR130,
grown as usual in the 16-liter carboys (yielding 320 mg/l total
suspended solids) together with 75 ml of the above groundwa-
ter, were Incubated at room temperature. Controls with buff-
ered nutrient water and organisms and other buffer controls
without organisms were prepared similarly. All samples were
mixed at room temperature. Periodically, individual bottles were
supplemented with 3.9 mg/l TCE and tine rate of consumption
over time was monitored. In the second study, 800 ml of the
above groundwater was mixed in a 2-liter flask with 800 ml of
culture grown as usual in 16-liter carboys. A 2-liter flask control
contained 800 ml of buffered nutrient water and 800 ml of
culture, but no groundwater. Both flasks were mixed at room
temperature. Daily, 75 ml was withdrawn from each flask for
analysis of TCE decomposition rate and was replaced with 75
ml of buffer water containing 16 mg/l sodium lactate, similar to
the concentration added in the field study. For withdrawn
samples, 25 ml was mixed with 75 ml of buffer water, 2.6 mg/l
TCE was added, and the rate and extent of TCE utilization with
time was measured. A control containing 100 ml of buffer water
and 2.6 mg/l of TCE was also evaluated for possible physical
loss of TCE.
In the first field study, the TCE degradation rate in groundwater
cultures was about 20% higher than cultures in buffer water,
perhaps because the groundwater contained some active
PR1301. However, by day four, the rate with the groundwater
culture was only about 88% of that in buffer water, and by day
ten the groundwater culture removed no TCE while the buffer
water culture was still removing TCE at about 10% of the initial
rate. These data suggest that the groundwater was detrimental
to the activity of PR1301. The second study was perhaps more
definitive as lactate was added in order to maintain some
activity among the cultures similar to what was done in the
field. Here again, the TCE utilization rate on day zero was
about 8% higher with the groundwater culture, but within three
days, the groundwater culture was no longer consuming TCE
but the buffer water culture maintained a TCE degradative
ability throughout the last nine days of the study at a rate of
about 23% of that obtained on day zero. This again indicates
that groundwater conditions were detrimental to PR1301, activ-
ity. The results are not conclusive as to whether predation
caused the activity loss or whether PR1301 simply lost out in the
competition for lactate and for this reason could not maintain a
high level of activity for long. In any event, the laboratory study
confirmed that PR1301 could not maintain activity well in the
groundwater environment. Three field studies with
bioaugmentation were conducted. In each one, bioaugmentation
with PR1301, along with lactate addition was initially reasonably
effective at TCE removal. In addition, phenol was rapidly con-
sumed to near detection limits when added following a short
period of bioaugmentation with PR1301 and lactate addition.
These results demonstrated that an initial phase of
bioaugmentation could be effective to establish a population of
10
-------
phenol or toluene degraders if these substrates were to be
used at a site. However, in both the second and third field
trials, bioaugmentation in time with lactate feed alone became
ineffective at TCE removal, reaching near zero removal within
a few weeks. The inability of PRI^,, to remain effective in
bioaugmented groundwater was also demonstrated in a brief
laboratory study using groundwater from the bioaugmented
well after the conclusion of the field study. This suggested
either predation of the introduced population or the inability of
PR1301 to effectively compete for the added lactate, or perhaps
both, were the cause of the eventual failure of the system. In
order for bioaugmentation for TCE cometabolism to be suc-
cessful, methods for avoiding this competitive problem need to
be found.
Acknowledgments
This study was supported through Cooperative Agreement
CR822029 from the Gulf Ecology Division, U.S. Environmental
Protection Agency. Special thanks are extended to Dr. Laurence
H. Smith and Dr. Alfred M. Spormann for their assistance in
culture growth during the third phase of the field study.
Disclaimer
The U.S. Environmental Protection Agency through its Office of
Research and Development partially funded and collaborated
in the research described here under Cooperative Agreement
No. CR819630 to North Carolina State University. It has been
subjected to the Agency's peer and administrative review and
has been approved for publication as an EPA document. Men-
tion of trade names or commercial products does not constitute
endorsement or recommendation for use.
Quality Assurance Statement
This project was conducted under an approved Quality Assur-
ance Program Plan. The procedures specified in the plan were
used without exception. Information on the plan and documen-
tation of the quality assurance activities and results are avail-
able from the principal investigator.
Research Products Cited
Miller, R. V. 1996. Genetic Stability of Genetically Engineered
Microorganisms in the Aquatic Environment. In: T. E. Ford
(ed.), Aquatic Microbiology - An Ecological Approach —
1997. Blackwell Scientific Publications, Boston.
Kidamba, S. P., M. G. Booth, T. A. Kokjohn, and R. V. Miller.
1996. RecA-dependence of the Response of Pseudonionas
aeroginosa to UVA and UVB Irradiation. Microbiology
142:10331040.
Ripp, S., and R. V. Miller. 1995. Effects of Suspended Particu-
lates on the Frequency of Transduction Among Pseudomo-
nas aeruginosa in a Freshwater Environment. Appl. Environ.
Micmbiol. 61:1214-1219.
Replicon, J., Frankfater, A., and R. V. Miller. 1995, A Continu-
ous Culture Model to Examine Factors that Affect Trans-
duction Among Pseudonionas aeruginosa Strains in Fresh-
water Environments. Appl. Environ. Microbiol. 61:3359-
3366.
Kidambi, S. P., S. Ripp, and R. V. Miller. 1994. Evidence for
PhageMediated Gene Transfer Among Pseudomonas
aeruginosa Strains on the Phylloplane. Appl. Environ.
Microbiol. 60:496-500.
Kidambi, S. P., S. Ripp, and R. V. Miller. 1994. Evidence of
Phage-Mediated Gene Transfer Among Pseudomonas
aeruginosa Strains on the Phylloplane. Appl. Environ.
Microbiol. 60:496-500.
11
-------
Q
co
O
0.
in
§
I
to
.2
8 2
CD C/J
O) CD
< CC
OS
» C
O
------- |