SEPA
United States
Environmental Protection
Agency
Robert S. Kerr Environmental
Research Laboratory
Ada. OK 74820
Research and Development EPA/600/M-89/019 August 1989
ENVIRONMENTAL
RESEARCH BRIEF
BIOPLUME Model for Contaminant Transport Affected
by Oxygen Limited Biodegradation
H.S. Rifaia, P.B. Bedient8 and J.T. Wilson"
Introduction
Many of the organic pollutants entering ground water are
potentially biodegradable in the subsurface. This potential has
been demonstrated in aquifers contaminated by wood-creosoting
process wastes (Borden et al., 1986) and gasoline (McKee et
al., 1972, Wilson et al., 1986). The persistence of many of
these organic compounds in the subsurface indicated that
some factors must be limiting biodegradation.
Current research at Rice University through the National Center
for Ground Water Research and the R.S. Kerr Environmental
Research Laboratory (RSKERL) of the U.S. EPA has been
aimed at identifying the major processes that limit biodegrada-
tion in aquifers and developing a mathematical model (BIOPLUME)
for simulating these processes.
Recent studies have shown the presence of quite active,
diverse microbial populations in the subsurface (Britton and
Gerba, 1984; Ghiorse and Balkwill, 1985). These organisms
have the ability to degrade a wide variety of organic contami-
nants (Wilson et al., 1983; Lee and Ward, 1985; Kuhn et al.,
1985; Gibson and Suflita, 1986; and Barker etal., 1987). Key
factors which seem to limit biodegradation is the lack of an
essential nutrient, typically nitrogen or phosphorous, or an
electron acceptor such as oxygen. Addition of the necessary
electron acceptor(s) stimulates the microorganisms and enhances
the restorative capacity of the contaminated aquifer.
In order to identify the rate limiting processes for biodegradation,
the equations describing microbial growth and decay and
transport of oxygen and contaminants were developed and
The authors are with the aNational Center for Ground Water Research,
Rice University, Houston, TX 77251, and °U.S. EPA's Roberts. Kerr
Environmental Research Laboratory, Ada, OK 74820.
solved in one and two dimensions. The main purpose of this
research was to develop the mathematical tools necessary to
describe and simulate the process of oxygen limited biodegrada-
tion of organics in ground water.
The United Creosoting Company, Inc. (UCC) site in Conroe,
Texas was evaluated in testing of the simulation models. Field
work at the site suggested that lack of oxygen was limiting the
microbial degradation of dissolved aromatic hydrocarbons pre-
sent in the shallow aquifer (Borden et al., 1986).
A numerical model, BIOPLUME, was developed to simulate
oxygen-limited biodegradation. BIOPLUME simulates advectipn,
dispersion and retardation processes as well as the reaction
between oxygen and the contaminants under steady, uniform
flow. The model allows the computation of two plumes sequen-
tially, one for the contaminants and one for oxygen. At every
time step, the two plumes are combined using superposition to
simulate the reaction between oxygen and the contaminants.
Other processes such as anaerobic degradation and diffusion
of oxygen from the unsaturated zone (reaeration) are simulated
as a first order decay in contaminant concentrations.
BIOPLUME was also applied to an aviation gasoline spill site at
Traverse City, Michigan (TCM). The model was used to
simulate the behavior of the degrading plume over a two year
period. Model predictions for the rates of mass loss closely
matched calculated rates from the field data.
Objectives
The overall objectives were:
1. Develop and test the equations for describing oxygen
transport, contaminant transport and microbial growth and
decay.
-------
2. Develop BIOPLUME based on an existing solute transport
model to include the major processes which limit biodegradation.
3. Evaluate BIOPLUME against selected analytical solutions
and against the UCC site data and the TCM site data.
4. Develop a PC version of the model (BIOPLUME II) and a
user's guide for the model.
Methods
Growth of microorganisms and removal of organics and oxygen
were simulated using a modification of the Monod function
(Borden and Bedient, 1986). These functions were combined
with the classic form of the advection-dispersion equation for a
solute undergoing linear instantaneous adsorption to obtain
Eqns. 1, 2, and 3 in Table 1.
Table 1. Two Dimensional Transport Equations with
Reaction Terms
<)H V(DVH-vH)
— = v(D v o - vO) - M • k • F .
V(DVM -vMJ
Ms.k.Y
— (1)
ko+o
(2)
O k .Y.OC
+ - bM (3)
Ko+0 Rm
Where,
D
v
O
H
0C
b
F
Y
dispersion tensor
ground water velocity vector
concentration of dissolved oxygen
concentration of contaminant
retardation factor for contaminant
concentration of microbes in solution
microbial retardation factor
Rm'Ms
contaminant half saturation constant
oxygen half saturation constant
first order decay of natural carbon
natural organic carbon concentration
microbial decay rate
ratio of oxygen to contaminant consumed
microbial yield coefficient
(g cells /g contaminant)
maximum contaminant utilization rate per unit
mass microorganisms
Instantaneous Reaction Assumption
H(t+1) = H(t)-0(t)/F; 0(t+1) = O where H(t) > O(t)/F (4)
0(t+1) = 0(t)-H(t)-F; H(t+1) = O where O(t) > H(t)-F (5)
where H(t), H(t+1), O(t) and O(t+1) are the concentrations
of contaminant and oxygen at time t and t+1 respectively.
One- and two-dimensional models were developed to study the
behavior of Eqns. 1, 2, and 3. One-dimensional simulations
indicated that in the region closest to the contaminant source,
biodegradation rates will be very high and result In nearly
complete removal of oxygen. In the body of the organic plume,
biodegradation will be limited by the rate of mass transfer of
oxygen into the plume. In the third region downstream of the
bulk organic plume, oxygen will be present in excess of the
oxygen demand and contaminants will be absent.
Sensitivity analyses performed with the 1-D model Indicated
that microbial kinetics had little or no effect on the contaminant
distribution. This suggested that the consumption of organics
and oxygen by the microorganisms could be approximated as
an instantaneous reaction (Eqns. 4 and 5, Table 1).
The two-dimensional simulations indicated that blodegrading
plumes are narrower than non-biodegrading plumes. This
characteristic has been confirmed at Traverse City, Michigan
(Twenter et al., 1985). The simulations also suggested that
reaeration could be a significant source of oxygen into a plume
(Wilson et al., 1986).
Model Development - BIOPLUME
BIOPLUME solves the governing equations (Eqns. 1 and 2,
Table 1) under the assumption of instantaneous reaction between
oxygen and the contaminants. The USGS solute transport
model (Konikow and Bredehoeft, 1978) was modified to allow
parallel computation of an organic plume and an oxygen plume.
The two plumes are combined at every timestep using
superposition.
The model is very versatile in that it allows the simulation of
retarded plumes. More important, however, is the capability of
simulating in situ restoration schemes such as injecting
oxygenated water into the aquifer. Sensitivity analyses on the
different model parameters indicated that the amount of mass
biodegraded is most sensitive to the hydraulic conductivity, the
coefficient of anaerobic decay and the coefficient of reaeration
(Rifai et al., 1988). The model had a weak sensitivity to
dispersion and the retardation factor. For contaminant plumes
which are naturally biodegrading, increasing the retardation
factor decreases the amount of mass biodegraded.
A PC version of the model (BIOPLUME II) has been developed
and is supported with a preprocessor to aid in defining the input
data. A postprocessor supports the preparation of output files
which can be used by most any plotting package for obtaining
graphical output. A user's guide (Rifai et al., 1987) which
outlines the modeling concepts and the use of the model as
well as illustrative sample problems is available.
-------
Application of BIOPLUME
United Creosoting Site, Conroe, Texas
Organic contaminants present at the Conroe, Texas field site
are predominantly polycyclic aromatic hydrocarbons from the
wood creosoting process (Borden et al., 1986). A chloride
plume Is also present in the shallow aquifer and is moving at 15
ft./yr. A detailed description of the site's geology, hydrology and
subsurface microbiology can be found in Bedient et al., 1984,
Lee et al., 1984, and Wilson et al., 1985.
A three well Injection-production test was performed at the site
to estimate the effective retardation factors and to evaluate the
blotransformation of the hydrocarbons present in the aquifer
(Borden and Bedient, 1987). Results indicate that degradation
Is the major process limiting contaminant transport at the site.
A significant loss of contaminant mass was observed at the two
production wells.
BIOPLUME was calibrated to the observed chloride distribution
and was used to simulate hydrocarbon and oxygen transport.
Figure 1 shows contours of equal hydrocarbon concentration
forthree cases. Casel is the best estimate of conditions at the
site. Case 2 assumes no reaeration and Case 3 assumes no
biodegradation. It is clear that reaeration appears to be the
major source of oxygen to the plume at the site and that the
observed distribution could not be simulated with a transport
code that ignores biodegradation.
Traverse City Field Site, Michigan
The basic contaminant at the Traverse City Field site is aviation
gasoline. The water-table aquifer is about 50 ft. deep and
overlies 100 ft. of impermeable glacial clay. The water table
varies from 12 to 18 ft. below land surface, with ground water
velocities approximately 5 ft./day (Twenter et al., 1985.) The
plume is about 200-400 ft. wide (Figure 2). A pumping field was
installed in April, 1985 to halt the migration of BTX contaminants
(Benzene, Toluene, and Xylene) across the property boundary.
The research effort at the site included field sampling and
modeling of the contaminant plume with BIOPLUME. Data
collected by Rice University and RSKERL indicated that strong
anaerobic biological activity was occurring in the immediate
spill area (dissolved oxygen was absent). Four distinct regions
of contamination were identified: 1) The body of the plume:
high concentrations of BTX and trace concentrations of oxygen;
2) Zone of anaerobic treatment: low to moderate concentrations
of BTX, significant concentrations of methane at depths of 17
to 35ft., and trace concentrations of oxygen; 3) Zone of aerobic
treatment: low to moderate concentrations of oxygen (0.5-3
mg); and 4) Pristine zone: high concentrations of oxygen
(8 mg/l) and trace concentrations of BTX.
In addition, data was collected on a regular basis from a number
of wells at the site. The total BTX from about 25 wells were
averaged overthree month intervals beginning with the second
quarter of 1985. The data were also averaged vertically.
a)
Figure 1. Simulated hydrocarbon plumes for three cases: (a) case 1, best estimate simulation of the plume including biodegradation due to
horizontal mixing and vertical exchange with unsaturated zone; (b) case 2, simulated plume assuming no reaeration; and (c) case
3, simulated plume with no biodegradation. Contours are lines of equal hydrocarbon concentration in mg/L
-------
Concentration contours were developed from the averaged
data and are shown In Figures 2,3 and 4. As can be seen, the
plume is changing in time and exhibiting a significant loss of
mass.
The contour figures were used to calculate the mass of BTX
remaining in the system at the end of each quarter (Figure 5).
The mass captured by the pumping field is also shown on
Figure 5, and it is obvious that degradation accounts for most
of the mass loss at the site. Dispersion and volatilization would
not account forthe total mass loss observed. The rate of mass
loss calculated from Figure 5 is about 1.25 percent per day.
BIOPLUME was calibrated to the observed data prior to the
installation of the pumping field. The data in Figure 6 show the
results of the simulation along the centerline of the plume. The
model predictions matched the observed concentrations
except in the vicinity of well M31. This is to be expected since
anaerobic degradation was identified in that zone of the site,
but was not included in the model.
The calibrated model was then used to simulate the field data
from April 1985 to the end of 1986. The predicted rate of
degradation was about 1.0 percent per day in comparison to the
observed rate of 1.25 percent per day.
The BIOPLUME II model is presently being used to design and
operate an on-going biorestoration experiment at the Traverse
City field site. A nutrient mix containing phosphate, nitrogen
and an oxygen source are Injected Into a portion of the
contaminated aquifer next to the Hangar building (Figure 2).
The objective is to stimulate the microorganisms and enhance
the bioremediation activity occurring at the site. The BIO-
PLUME II model was used to determine the required number of
injection wells, the injection flow rates and the required time for
cleanup under different scenarios of oxygen delivery. Preliminary
data indicate that oxygen and inorganic nutrients are being
consumed and overall contaminant concentrations are decreasing.
The results of the restoration experiment will be published when
the test is concluded.
Conclusions
The following conclusions can be made from the previous
analysis:
1. Biodegradation processes have significant effects on
contaminant plumes. Naturally occurring biodegradation causes
a narrow plume and loss of mass In the body of the plume.
2. BIOPLUME simulates oxygen limited biodegradation and
incorporated reaeration and anaerobic processes as a first
order decay process.
3. BIOPLUME was calculated at the Conroe, Texas and
Traverse City, Michigan sites where biodegradation was observed.
Quarter 2,1985
Quarter 4, 1985
0M28
Well Location
& Number
BDL
Below Detection
Level
Hanger/Admin.
Building
150
Figure 2. Contaminant plume at Traverse City, quarters 2 & 4,1985.
-------
Quarter 1,
• M28
BDL
Well Location Below Detection Hanger/Admin.
& Number Level Building
meters 150
Figure 3. Contaminant plume at Traverse City, quarters 1 & 2,1986.
Quarter 3, 1986
Quarter 4, 1986
(M28
BDL
Well Location Below Detection Hanger/Admin.
& Number Level Building
meters 1 SO
Figure 4. Contaminant plume at Traverse City, quarters 3 & 4,1986.
-------
1200
1000 -
*-s 800
600 -
(0
5 40°
1-
200-
BTX Mass in Plume
— Mass Loss due to Interdiction Field
2/85 3/85 4/85 1/86 2/86 3/86 4/86
Time (Quarters)
Figure 5. Contaminant mass loss at Traverse City.
M28 M26 TP4 TP3 M30 M31 IN2 M2
IN4 M4
Figures. Calibration results at Traverse City. Contaminant
concentrations along plume centerline.
The model provided a better match than a standard transport
code.
4. Retardation can be handled in the new version, BIOPLUME
II, so that biodegradation effects on retarded plumes can be
simulated and analyzed.
5. BIOPLUME has the capability of simulating lusilu. restora-
tion by modeling the injection of oxygenated water, which
appears to be a viable alternative to the classic pump/treat
schemes.
Literature Cited
Barker, J.F., Patrick, G.C., and Major, D. (1987) Natural Attenuation
of Aromatic Hydrocarbons in a Shallow Sand Aquifer, Ground
Water Monitoring Review. 1, 64-71.
Bedient, P.B.. Rodgers, A.C., Bouvette, T.C., Tomson, M.B.,
and Wang, T.H. (1984) Ground-Water Quality at a Creosote
Waste Site, Ground Water. 22. 318-329.
Borden, R.C., and Bedient, P.B. (1986) Transport of Dissolved
Hydrocarbons Influence by Reaeration and Oxygen Limited
Biodegradation: 1. Theoretical Development. Water Resources
,22,1973-1982.
Observed Data
BIOPLUM Ell Model
Borden, R.C.. Bedient, P.B., Lee, M.D., Ward, C.H.. and
Wilson, J.T. (1986) Transport of Dissolved Hydrocarbons
Influenced by Reaeration and Oxygen Limited Biodegradation:
2. Field Application, Water Resources Research. 22, 1983-
1990.
Borden, R.C. and Bedient, P.B. (1987) In Situ Measurement of
Adsorption and Biotransformation at a Hazardous Waste Site,
Water Resources Bulletin. 23, 629-636.
Britton, G., and Gerba, C.P. (Eds). (1984) Groundwater Pollution
Microbiology. John Wiley and Sons, New York, NY, 377.
Ghiorse, W.C., and Balkwill, D.L (1985) Microbial Characterization
nf RuhsurfafiB Fnvironmants. GroundWaterQualitv. Ward. C.
H Gieger, W., and McCarty, P.L. Eds., John Wiley and Sons,
New York, NY, 387.
Gibson, S.A. and Suflita, J.M. (1986) Extrapolation of
Biodegradation Results to Groundwater Aquifers: Reductive
Dehalogenation of Aromatic Compounds, Applied and
Environmental Microbiology. 4, 681-688.
Konikow, L.F., and Bredeheoft, J.D. (1978) Computer Model of
Two-Dimensional Solute Transport and Dispersion in Ground
Water, Automated Data Processing and Computations.
Techniques of Water Resources Investigations of the U.S.
Geological Survey, Washington, DC., 90.
Kuhn, E.P., Colberg, P.J., Schnoor, J.L., Wanner, O., Zehnder,
A.J.B., and Schwarzenbach, R.P. (1985) Microbial Transformation
of Substituted Benzenes During Infiltration of River Water to
Ground Water: Laboratory Column Studies, Fnvironmental
Science & Technology. 19, 961-968.
Lee, M.D., and Ward, C.H. (1985) Microbial Ecology of a
Hazardous Waste Disposal Site: Enhancement of Biodegradation,
-------
Proceedings. Second Intl. Conference on Ground WaterQuality
Research, OSU Printing Services, Stillwater, OK, 25-27.
Lee, M.D., Wilson, J.T., and Ward.C.H. (1984) Microbial
Degradation of Selected Aromatics at a Hazardous Waste Site,
Developments in Industrial Microbiology. 25, 557-566.
McKee, J.E., Laverty, F.B., and Hertel, R.M. (1972) Gasoline in
Ground Water, Journal Water Pollution Control Federation. 44,
293-302.
Rifal, H.S., Bedient, P.B., Borden, R.C., and Haasbeek, J.F.
(1987) BIOPLUME II - Computer Model of Two-Dimensional
Transport under the Influence of Oxygen Limited Biodegradation
In Ground Water. User's Manual, Version 1.0, Rice University,
Houston, TX.
Rifai, H.S., Bedient, P.B., Wilson, J.T., Miller, K.M., and Armstrong,
J.M. (1988) Biodegradation Modeling at an Aviation Fuel Spill
Site, ASCE Journal of Environmental Engineering. 5,114.
Twenter, F.R., Cummings, T.R., and Grannemann, N.G. (1985)
Ground-Water Contamination in East Bay Township, Michigan,
Techniques of Water Resources Investigations of the U.S.
Geological Survey, Report 85-4064, Washington, DC.
Wilson, B.H., Bledsoe, B.E., Kampbell, D.H., Wilson, J.T.,
Armstrong, J.M., and Sammons, J.H. (1986) Biological Fate of
Hydrocarbons at an Aviation Gasoline Spill Site, Proceedings.
NWWA/API Conference on Petroleum Hydrocarbons and Organic
Chemicals in Ground Water - Prevention, Detection, and
Restoration, Houston, TX, 78-89.
Wilson, J.T., McNabb, J.F., Cochran, J.W., Wang, T.H., Tomson,
M.B., and Bedient, P.B. (1985) Influence of Microbial Adaptation
on the Fate of Organic Pollutants in Ground Water, Environmental.
Toxicology & Chemistry. 4, 721.
Wilson, J.T., McNabb, J.F., Balkwill, D.L.. and Ghiorse, W.C.
(1983) Enumeration and Characterization of Bacteria Indigenous
to a Shallow Water-Table Aquifer, Ground Water. 21, 134.
------- |