United States
Environmental Protection
Agency $,
Robert S. Kerr
Environmental Research Laboratory
Ada, OK 74820 '/ '
Research and Development
EPA/600/S2-90/017
July 1990
v°/EPA Project Summary
Innovative Processes for
Reclamation of Contaminated
Subsurface Environments
Larry W. Canter, Leale E. Streebin, M. Carlota Arquiaga, Francisco E. Carranza,
Dennis E. Miller, and Barbara H. Wilson
Research to better assess the capabil-
ities and limitations ol fixed-film bioreac-
tors for removing selected organic
contaminants from ground water or from
contaminated vapor streams produced
by air stripping of polluted ground water
and by soil venting operations is de-
scribed. Work was focused on volatile
chlorinated aliphatic hydrocarbons and
light aromatic constituents of distilled
petroleum products, two groups of
compounds which have been identified
in polluted ground water more frequently
and usually in higher concentrations than
other organic pollutants. Laboratory
scale fixed-film bioreactors containing
soil or diatomaceous earth materials were
employed to study the cometabolic
removal of trichloroethylene and related
compounds from aqueous and vapor
streams by biofilms sustained on primary
substrates consisting of gaseous ali-
phatic hydrocarbons. Additionally, the
removal of alkylbenzenes from air
streams by biofilms using these com-
pounds as primary substrates was exam-
ined in laboratory scale soil bioreactors.
The biodegradation processes involved
and the effects of bioreactor operating
parameters and systems configurations
on contaminant removal were evaluated.
Results obtained indicate a significant
potential for utilization of fixed-film
bioreactors in systems for above ground
treatment of contaminated ground water
and vadose zone gases.
This Project Summary was developed
by EPA's Robert S. Kerr Environmental
Research Laboratory, Ada, OK, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
Ground water is the primary source of
drinking water in the United States, with
estimates of approximately 80 percent of
all drinking water supplies obtained from
ground water sources. Historically,
ground water has been used for drinking
water without major treatment other than
removal of minerals and final disinfection.
During the past few years, however,
hundreds of synthetic chemicals have
been detected in a significant number of
drinking water supplies, thus raising
concerns about the potability and purity
of such waters.
Of the many organic chemicals that
have been detected in polluted ground
water, two groups of compounds have
been identified more frequently and
usually in higher concentration than any
others. These are volatile chlorinated
aliphatic hydrocarbons, such as trichio-
roethylene (TCE) and related substances,
and light aromatic constituents of distilled
petroleum products. Effective methodol-
ogies for removing these pollutants from
contaminated subsurface environments
and restoring the quality of polluted
ground waters are urgently needed.
-------
This report describes research con-
ducted to better define the feasibility of
treatment methodologies employing
fixed-film bioreactors for removal of
volatile chlorinated and light aromatic
hydrocarbons from aqueous and vapor
streams contaminated with these sub-
stances. Such methodologies would have
significant utility for above ground treat-
ment of polluted ground water and of
gaseous streams produced by air strip-
ping of ground water or by soil venting
operations. The work consisted of three
segments, conducted concurrently. In one
segment, treatment of trichloroethylene
(TCE) in solution in water by biofilms
sustained on a primary substrate of
methane or natural gas was studied, using
laboratory scale bioreactors containing
sand or beryl saddles as supporting
medium for the biofilm. Another segment
of the research effort examined vapor
phase treatment of TCE and 1,1,1-
trichloroethane (TCA) in laboratory scale
bioreactors containing biofilms sustained
by butane and supported on diatomace-
ous earth materials. The third segment
consisted of studies of the vapor phase
treatment of selected alkylbenzenes in
soil bioreactors by biofilms utilizing these
compounds as primary substrates.
Liquid Phase Treatment of
Trichloroethylene
The removal of TCE from contaminated
aqueous streams was studied in labor-
atory packed column bioreactors such as
that depicted schematically in Figure 1.
Bioreactors were designed so all surfaces
in contact with feed water solutions
containing TCE were either glass or
Teflon. Columns (6 x 61 cm) were packed
with rewashed coarse sand (specific
surface area of 1599 or 673 ft2/ft3) for
most experiments, although beryl saddles
were used as the support for the biofilm
in one case. The systems were operated
in the unsaturated mode, with liquid flow
rates ranging between 1.9 and 8.4 ml/min
(equivalent to hydraulic loadings between
4.3 and 19.6 W/day/fP). Biofilms of
methanotrophic microorganisms capable
of cometabolizing TCE were established
in the reactors, utilizing enrichment
cultures prepared from Lincoln fine sand.
Influent water was amended with nitrogen
and phosphorus to enhance methane
utilization and biofilm development. After
appropriate acclimation, the systems
were operated over a range of hydraulic
and organic loadings, during which
bioreactor performance was assessed by
monitoring influent and effluent concen-
trations of methane and TCE, as well as
flows of water and air.
TCE removals in the bioreactor
columns packed with coarse sand gener-
ally ranged from 20 to 60 percent for most
of the experimental conditions examined
in this study. The specific percentage of
removal appeared to be a function of
several interrelated parameters, including
particularly the influent TCE and methane
concentrations, the packing media for the
columns, and water flow.
The percentage of TCE removed in the
bioreactors was found to increase signif-
icantly with increasing influent concen-
trations up to at least 500 //g/L of TCE,
as shown by Figure 2. However, when
TCE influent concentrations approached
1000 yug/L, removal was observed to be
drastically reduced in bioreactors con-
structed from either of the coarse sand
supporting media. Evaluation of methane
utilization in bioreactors receiving various
input concentrations of TCE revealed that
significant inhibition of methanotrophic
microorganisms comprising the biofilm
was likely to occur when TCE levels were
in the vicinity of 1000 yug/L or higher, as
indicated by Figure 3. Inhibition appeared
to be related to influent methane concen-
tration, with biofilms sustained on low
concentrations of methane appearing to
e e-
be most susceptible. Although complete
inhibition of methane utilization did not
occur until influent TCE concentrations
exceeded 1500 /ug/L, the ability of the
microbes to cometabolize TCE was
apparently lost completely when their
capability for metabolizing the primary
substrate was only partially compro-
mised.
Influent methane concentrations rang-
ing from less than 1 mg/L to more than
15 mg/ L in air delivered to the bioreactors
were observed to sustain biofilms capable
of removing TCE in the bioreactors. The
percentage of TCE removed increased
with increasing influent methane concen-
tration for influent TCE concentrations
below inhibitory levels, as shown in
Figure 4. However, the rates of increase
were relatively small, so the specific mass
of TCE removed per unit mass of methane
consumed was less at higher methane
concentrations. Nevertheless, signifi-
cantly higher removals of TCE were
attained in bioreactors receiving higher
influent levels of methane, provided other
operating parameters were the same. This
is illustrated in Figure 2, which presents
TCE removals at varying influent TCE
concentrations for systems operating
under identical conditions except for
influent methane concentrations, which
were either 3.4 or 13.4 mg/L in air.
In one set of experiments, natural gas
was substituted for pure methane as
primary substrate for the microbes com-
Glass Beads
• Coarse Sand;
TCE Solution Pump Tap Water
O Meth. Sampling Port
n TCE Sampling Port
o Flow Monitoring
Air
Methane
Peristaltic Pump
Water Effluent
Air Effluent
Figure 1. Schematic of fixed-film bioreactor.
-------
60
50
40
o 30
r
20
10
100
200
300
400
500
Infl. TCE Cone. (fjg/L)
Figure 2. Effect of influent TCE concentration on TCE biodegradation. (Water flow: 4.1 mL/min;
Air flow: 7.5 mL/min; Influent methane concentration: 3.4 mg/L; Media: Sand 3).
100
98
96
c
CO
-c
I
94
92
90
300
7200
600 900
Infl. TCE Cone. (jjg/L)
Figure 3. Inhibition of methanotrophs by TCE for an influent methane concentration of 3.2 mg/L.
1500
prising the biofilm. Although the data
obtained were not sufficient for definitive
conclusions, removals of TCE in bioreac-
tors operating on natural gas appeared
to be at least as good as removals attained
in those operating on pure methane.
Especially at influent TCE concentrations
in excess of 900 jug/L, removals in the
natural gas bioreactors seemed notice-
ably higher. This may reflect less suscep-
tibility of biofilms sustained on natural gas
to inhibition by high TCE concentrations.
Such biofilms would be expected to be
composed of more diverse microbial
population than biofilms developed on
pure methane.
Changes in TCE removal by the bio-
reactors were also noted when water flow
rates were varied. For bioreactors packed
with the sand of higher specific surface
area and, hence, having longer unsatu-
rated hydraulic retention times, percen-
tages of TCE removed increased steadily
as water flow was increased in the range
of 3.8 to 6.2 mL/min (8.9 to 14.5 W/day/
ft2). This was believed to be the result of
reduced resistance to mass transfer at the
higher flow rates. Highest rates of TCE
removal in bioreactors packed with the
sand of lower specific surface area
seemed to be achieved when water flow
was in the neighborhood of 4 mL/min (9.3
ft3/day/ft2). This probably reflected the
negation of beneficial effects of improved
mass transfer at higher flow rates by
deleterious effects of decreased contact
time resulting from shorter unsaturated
hydraulic retention times in these columns.
Both of the coarse sands utilized as
packing media in these studies served
relatively effectively as supports for
biofilms capable of cometabilizing TCE,
although the operating parameters to
achieve optimum TCE removal were
different for each, as illustrated by the
observations concerning water flow rates
noted above. Bioreactors packed with the
sand of greater surface area were found
to be somewhat more susceptible to
clogging by excessive biomass when
influent concentrations of methane were
very high. In limited studies with bioreac-
tors packed with beryl saddles as biofilm
support media, much higher percentages
of low influent concentrations of TCE
were removed than in comparable sys-
tems packed with coarse sand. This may
have resulted from better distribution of
the microorganisms throughout the length
of the beryl saddles column, with con-
comitant longer contact of the biofilm with
the TCE-laden water. However, at high
TCE concentration (910 Aig/L), removal
was significantly less in the beryl saddle
system, possibly indicating increased
susceptibility of the biofilm to inhibition
by TCE.
Changes in the configuration of the
bioreactor systems were also investigated
in order to improve the effectiveness of
the systems for removing TCE from
influent water streams. This included
operating two columns in series and
operating single columns with multiport
injection of the primary substrate,
methane. Standard fixed-film bioreactor
columns (Figure 1), packed with coarse
sand and modified as appropriate, were
used.
-------
1
fi
Uj
60
50
40
30
20
10
10
15
Intl. Meth. Cone. (mg/L)
Figure 4. Effect of influent methane concentration on TCE biodegradation. (Water flow: 4.1 mU
mm; Air flow: 7.6 mL/min; Media: Sand 3).
In the systems constructed with two
bioreactor columns in series, effluent
liquid and vapor streams from the first
column were introduced directly onto the
top of the second column, together with
sufficient methane to achieve an influent
methane concentration in the second
column equal to that in the first column.
Since the effluent from the first column
usually contained essentially no methane,
the total mass of this compound intro-
duced to the system was double that
introduced in a single column system.
Multiport injection systems consisted of
single bioreactor columns modified to
achieve introduction of methane simul-
taneously at the top and at points one-
third and two-thirds down the length of
the column, as shown in Figure 5. The
objective was to obtain a more uniform
distribution of the biofilm throughout the
supporting media. An examination of
columns receiving methane only at the
influent end revealed most of the biofilm
development occurred within the first 8
to 15 cm of packing. The total mass of
methane introduced to a multiport injec-
tion column was the same as that intro-
duced to a single column receiving
methane only at the influent end.
As Table 1 shows, removals of TCE
achieved both in bioreactor systems
constructed with two columns in series
and with single columns with multiport
injection were significantly higher than
removals achieved by single column
systems in which methane was intro-
duced only at the influent end. The data
presented, representing two sets of
experiments using different influent
concentrations of TCE, indicate that
columns in series can provide 60 to 65
percent increases in TCE removal over
that attainable in single columns with
single point methane injection, while
single columns with multiport injection of
methane can be expected to remove 27
to 30 percent more TCE than single
column, single point injection systems.
Systems constructed with single bioreac-
tor columns with multiport methane
injection may be especially promising
because they appear to provide signifi-
cant increases in TCE removal essentially
without any requirement for additional
equipment or supplies.
Vapor Phase Treatment of
Trichloroethylene and
1,1,1 -Trichloroethane
The cometabolic removal of TCE and
TCA from contaminated air streams was
studied in bioreactors using two sizes of
Manville's Celite diatomaceous earth
products (R630 and R635) as the solid
support for microbial growth and butane
as the primary substrate. The bioreactors
were constructed using borosilicate glass
columns 60 cm long by 5 cm in diameter.
The headspaces of the columns were
plumbed to receive vapors of TCE and
TCA, a mixture of butane in air, and a
nutrient media solution. To inoculate the
columns, a slurry of Rollin muck soil
previously adapted to n-butane was
circulated until substantial removals of
butane were observed at approximately
2.5 weeks.
Addition of chemical began on day 18
with influent concentrations of 90 //g
TCE/L air, 200 //g TCA/L air, and 8 mg
butane/L air. Influent and effluent con-
centrations of butane, TCE, and TCA were
monitored daily. Respective removals of
TCE, TCA, and butane in the bioreactor
with the R635 solid support were 82
percent, 93 percent, and 65 percent on
day 19. Removals on day 19 for the
bioreactor utilizing the R630 support were
74 percent for TCE, 76 percent for TCA,
and 59 percent for butane. On day 28,
the influent concentrations were
increased to 770 fig TCE/L air and 990
fig TCA/L air. At these influent concen-
trations, maximum removals for TCE and
TCA ranged from 16 to 25 percent with
both solid supports.
Density estimates of bacteria and
protozoa in the recirculated fluids were
determined by acridine-orange direct
counts. The recirculated water contained
a bacterial count of 1.8 x 1010 /mL The
water also contained approximately 6400
protozoa/mL
No products of biotransformation were
determined in this study. The probable
intermediates of oxidation of low molec-
ular weight alkanes such as methane,
propane, or butane are readily degradable
alcohols and ketones. Recent work has
shown 2,2,2-trichloroethanol is the chlor-
inated intermediate of TCA oxidation. The
oxidation of chlorinated ethylenes most
likely yields an epoxide with rapid con-
version to biodegradable hydrolysis
products.
Vapor Phase Treatment of
Alkylbenzenes
The objective of this study was to
examine various soils for their ability to
remove aromatic hydrocarbon vapors
from waste air streams. Three soils of
differing textures were tested for their
ability to remove vapors of benzene,
toluene, ethylbenzene, and o-xylene. The
soils examined were Rubicon sand from
Traverse City, Michigan; Durant loam from
-------
TCE Solution
o Meth. Sampling Port
D TCE Sampling Port
o Flow Monitoring
•&B-
Air Effluent
Figure 5. Schematic of bioreactor with intermediate methane sampling ports.
Table 1. Removal of TCE by Bioreactors Incorporating Single Column, Two Columns in Series,
and Single Column with Multiport Injection
Influent Influent Air Methane Methane TCE
Bioreactor TCE Cone. Water Flow Flow Cone. Removal Removal
Configuration mg/L mL/min mL/min mg/L % %
Single
Column
Two Columns
in Series
Single Column
with Multiport
Injection
Control
Column
Single
Column
Two Columns
in Series
Single Column
with Multiport
Injection
420
420
420
420
267
267
267
6.0
6.0
5.9
6.0
5.9
5.9
5.9
6.4
6.5
6.6
6.1
6.8
7.0
6.7
13.5
13.3
13.8
—
13.5
13.6
14.2
99.8
99.8
99.6
—
99.8
99.8
99.7
33.6
54.0
44.1
0.0
26.8
44.2
34.1
Control Column
267
6.0
6.2
0.0
Ada, Oklahoma; and Dougherty sand from
Stratford, Oklahoma. The variables of
hydrocarbon loading rate, soil type, and
soil moisture were examined to determine
their influence on removal efficiencies.
The soils were originally acclimated in soil
columns and then used for batch micro-
cosm studies. Reaction constants were
developed from the batch experiments.
A 7.6 cm internal diameter beaded
process pipe was used to construct the
soil columns. The column length varied
soil columns. The column length varied
from 86.4 cm to 96.5 cm. The columns
were filled with soil, then an air stream
was introduced into the bottom of the
column to establish flow up through the
bioreactor. Hydrocarbon removal was
determined by comparison of the inlet and
outlet concentration. Flow rates were
measured prior to each sampling period.
Upon completion of the studies, the
columns were taken apart and sections
of soil were stored in separate sealed
containers for later use in batch experi-
ments.
Soil microcosms were prepared from
the acclimated soil using 50 and 160 mi-
serum bottles. The amount of soil added
to a bottle was dependent upon the
moisture content and estimated porosity
of the soil. A headspace to open pore
space ratio of 60:1 was used so that hours
in a microcosm would correspond to
minutes in a soil bed. Each microcosm
was injected with hydrocarbon vapors
and sampled at predetermined time
intervals.
The hydrocarbon loading rate was
varied by changing the inlet hydrocarbon
concentration and by changing the flow
rate of the inlet vapor stream. Inlet
concentration did not significantly affect
the percentage of removal when the
concentration was adjusted from a lower
to higher concentration, 6 //g/mL to 26
Aig/mL of total hydrocarbons. Changing
the inlet concentration from a higher to
lower concentration, 50/ug/mL to 12 /ug/
mL total hydrocarbons, resulted in greatly
increased removals of benzene (31
percent removal increased to 96 percent
removal), and to a lesser extent, increased
removals of toluene (removals increased
from 23 percent to 31 percent). The
increased removal activity was short-
lived and diminished as the biomass to
hydrocarbon ratio readjusted to the
reduced loading rate. Reducing the inlet
flow rate produced greater removals until
the removal process became transport
limited. At this point, further reduction of
the inlet flow rate resulted in little or no
increase in hydrocarbon removal.
Although all soils examined displayed
an ability to remove a portion of each
vapor stream component, soil types had
a large influence on removals. Rubicon
sand produced the most efficient remov-
als for the four contaminants tested.
Dougherty sand removed some benzene
but substantially smaller amounts of the
other components. The Durant loam soil
column removed almost equal percen-
tages of each of the four components and
displayed the most consistent first-order
-------
removal characteristics. Benzene remov-
als in microcosms constructed from
Durant loam soil are shown in Figure 6.
Soil moisture content had a great
influence on the hydrocarbon removal
rate in the microcosm experiments, with
the highest removal rate constants found
on the higher moisture content (Figure 7).
Closer examination of soil moisture levels
is needed to determine the points of low
moisture and high moisture inhibition. In
addition, aromatic hydrocarbon removal
in the presence of typical gasoline
alkanes should be tested. Factors which
700
80--
O)
I 60
3
1
may affect transport limitations, including
grain size distribution, organic carbon
content, and moisture content, need to be
examined.
Conclusions and
Recommendations
The results obtained in this research
indicate that fixed-film bioreactors have
significant potential utility for removal of
volatile chlorinated and light aromatic
hydrocarbons from polluted ground water
20--
H 1 1 1 1 f-—h
-*-\ 1 h
30 40
Time (hours)
0 10 20
Moisture Content 19.6%
O —Control
* —Sample
Figure 6. Benzene removal in Durant loam microcosm.
50
60
-HM
70
700
80 f-
D>
| 60 +
CO
-------
individual compounds appears to be properties and microbial ecosystems of laden with aromatic hydrocarbons will
dependent on the type of soil used In different soils. The development of opti- require further studies to better define
preparation of the bioreactor, probably mized methodologies employing soil characteristics of soils most appropriate
reflecting differences both in physical bioreactors for treating vapor streams for bioreactor construction.
-------
Larry W. Canter, Leale E. Streebin, M. Carlota Arquiaga, Francisco E. Carranza,
Dennis E. Miller, and Barbara H. Wilson are with University of Oklahoma,
Norman, OK 73019.
William J. Dunlap is the EPA Protect Officer (see below).
The complete report, entitled "Innovative Processes for Reclamation of
Contaminated Subsurface Environments," (Order No. PS 90-199 5141 AS;
Cost: $17.00, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
Ada, OK 74820
United States Center for Environmental Research
Environmental Protection Information
Agency Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S2-90/017
2601
U.S. EPA / REGION 5
OFFICE OF GRQUNQWATER
230 SOUTH DEARBORN STREET
CHICAGO It- 60604
------- |