AT123D vs. BIOSCREEN
This issue provides a comparison between the AT123D and BIOSCREEN analytical groundwater models.
Rather than simply providing a technical description of the models this review looks at variations
in model output for comparable input parameters.
Unlike numerical models, results in analytical models are calculated independently at each point.
This means that users do not need to design a network of grid nodes prior to modeling. In addition
calibration is significantly easier. This ease of use does not necessarily mean the results are any
less accurate. In fact the results can be just as valid for certain conditions. Obviously,
this is not to say that all analytical models are equal. As with numerical models there can be
significant difference as to which processes are simulated and how the processes are simulated.
Both of these models are available as public domain versions and have been incorporated in
integrated software packages.
AT123D
AT123D is a three-dimensional analytical groundwater transport and fate model. Processes simulated
include: advection, dispersion, diffusion, adsorption, and biological decay. AT123D was developed
in FORTRAN by G. T. Yeh (1981) at Oak Ridge National Laboratory. The version of AT123D in SEVIEW
was enhanced to simulate 1,000 years when linked to SESOIL and over 10,000 years when run on its own.
Contaminant releases can be simulated as instantaneous, continuous or varying loads. The ability
to utilize varying loads allows it to be linked to the SESOIL vadose zone model. Source load
configurations can be established as a point, line, plane, or volume release. AT123D has been used
by numerous regulatory agencies to establish baseline cleanup objectives.
BIOSCREEN
BIOSCREEN is a two-dimensional screening level groundwater transport and fate model. It was
developed in 1996 under contract from the US Air Force, by Ground Water Services, Inc. Unlike
most other models, BIOSCREEN is not a standalone program in that it requires Microsoft Excel to
run. Transport and fate processes simulated include: advection, dispersion, adsorption, and
biological decay.
BIOSCREEN is based on the Domenico model which assumes the source concentration remains constant
through time. The developers of BIOSCREEN determined this to be overly conservative and added a
declining source term. This declining source should not be confused with biodegradation. Instead,
the declining source concentration is designed to simulate the flushing of contamination from the
source via groundwater flow. This declining source should be used cautiously. Even the BIOSCREEN
documentation states:
"this is an experimental relationship,
and it should be applied with caution."
Source load configurations can be defined as a point, line, or plane. Unlike AT123D there is
no provision for a volume release scenario.
Model Parameters
Input parameters were based on the release of benzene in three aquifer types: sand, silt, and clay.
A volume of contaminated groundwater was established. This volume was used to define the release
coordinates in AT123D. As BIOSCREEN cannot simulate a volume release, the source was simulated as
a plane along the x and y axis. However, the volume of the source was used to calculate the total
mass used in the BIOSCREEN declining source term. BIOSCREEN modeling was performed using both an
infinite release and a declining source term. A concentration of 1.0 mg/L was established in both
models. Mass load to AT123D was calibrated to achieve an initial concentration of 1.0 mg/L in the
volume of contaminated groundwater. Additional BIOSCREEN runs were performed using the AT123D
mass load for the declining source term.
Contaminant Load
|
Soil Type
|
AT123D
|
BIOSCREEN
|
|
units
|
Kg
|
mg/L
|
Kg
|
|
Sand
|
0.00862
|
1.0
|
0.01981
|
|
Silt
|
0.00512
|
1.0
|
0.0226
|
|
Clay
|
0.177
|
1.0
|
0.0255
|
Contaminant Release Coordinates
|
|
X-Axis
|
Y-Axis
|
Z-Axis
|
|
|
Start
|
End
|
Start
|
End
|
Start
|
End
|
|
units
|
ft
|
ft
|
ft
|
ft
|
ft
|
ft
|
|
AT123D
|
-5.0
|
5.0
|
-10.0
|
0.0
|
0.0
|
-5.0
|
|
BIOSCREEN
|
-5.0
|
5.0
|
- -
|
- -
|
0.0
|
-5.0
|
Aquifer Parameters
Aquifer parameters were based on three soil types: sand, silt, and clay. Parameters for
these soil types were based on information provided in the SEVIEW 6.2 documentation. Additional
parameters were selected based on generic site conditions.
Aquifer Parameters
Soil
Type
|
Permeability
|
Gradient
|
Organic
Carbon
Content
|
Bulk
Density
|
Effective
Porosity
|
|
units
|
cm/sec
|
fraction
|
fraction
|
g/cm3
|
fraction
|
|
Sand
|
1.0E-3
|
0.001
|
0.001
|
1.40
|
0.30
|
|
Silt
|
1.0E-5
|
0.005
|
0.001
|
1.60
|
0.27
|
|
Clay
|
1.0E-7
|
0.030
|
0.010
|
1.80
|
0.20
|
Dispersivities
Soil
Type
|
Longitudinal
Dispersivity
|
Transverse
Dispersivity
|
Vertical
Dispersivity
|
|
units
|
ft
|
ft
|
ft
|
|
Sand
|
5.14
|
0.51
|
0.051
|
|
Silt
|
3.30
|
0.33
|
0.033
|
|
Clay
|
3.30
|
0.33
|
0.033
|
Chemical Parameters
Modeling was performed for a release of benzene. Values for the organic carbon adsorption
coefficient (Koc) and the water diffusion coefficient were obtained from the
SEVIEW 6.2 chemical database. The water diffusion coefficient is only used in the AT123D model.
Chemical Properties
|
Chemical
|
Organic
Carbon
Adsorption
Coefficient
(Koc)
|
Water
Diffusion
Coefficient
|
|
units
|
(μg/g)/(μg/ml)
|
cm2/sec
|
|
Benzene
|
58.9
|
9.8E-6
|
Results
Results are based on the maximum predicted groundwater concentration at a point located 10
meters (32.8 feet) downgradient of the release. Maximum predicted concentrations for AT123D
were determined using the SEVIEW Point of Compliance reports. Determination of the maximum
predicted concentrations in BIOSCREEN was not as easy. Multiple model runs were performed until
the maximum concentration was observed. A summary of the results are presented below.
According to AT123D it would take benzene 10 years to reach the maximum concentration at a
point 10 meters downgradient in sand. BIOSCREEN predicted that it would take 62 years under steady
state conditions and about 14 years using the declining source term.
Sand Aquifer
Maximum Concentration 10 Meters Downgradient
|
Model
|
Maximum
Concentration
|
Time to
Maximum
Concentration
|
Ratio to
AT123D
Results
|
|
units
|
mg/L
|
years
|
fraction
|
|
AT123D
|
0.120
|
10
|
1.0
|
BIOSCREEN
(infinite source)
|
0.609
|
62
(steady state)
|
5.75
|
BIOSCREEN
(declining load)
|
0.319
|
13.92
|
2.65
|
Based on AT123D it would take 127 years to reach the maximum groundwater concentration at a
point located 10 meters downgradient in the silt aquifer. However, according to BIOSCREEN it
would take 850 years at the same location under steady state conditions and about 300 years
using the declining source term.
Silt Aquifer
Maximum Concentration 10 Meters Downgradient
|
Model
|
Maximum
Concentration
|
Time to
Maximum
Concentration
|
Ratio to
AT123D
Results
|
|
units
|
mg/L
|
years
|
fraction
|
|
AT123D
|
0.00785
|
127.25
|
1.0
|
BIOSCREEN
(infinite source)
|
0.715
|
850
(steady state)
|
91.1
|
BIOSCREEN
(declining load)
|
0.412
|
298
|
52.5
|
For benzene in clay AT123D predicts that it would take 820 years to reach the maximum groundwater
concentration at a point located 10 meters downgradient. According to BIOSCREEN benzene would
not have even reached that location in 820 years. In fact, BIOSCREEN predicts that it would take
12,000 years to reach the maximum groundwater concentration 10 meters down gradient for steady
state conditions and about 4,500 years using the declining source term.
Clay Aquifer
Maximum Concentration 10 Meters Downgradient
|
Model
|
Maximum
Concentration
|
Time to
Maximum
Concentration
|
Ratio to
AT123D
Results
|
|
units
|
mg/L
|
years
|
fraction
|
|
AT123D
|
0.0164
|
820
|
1.0
|
BIOSCREEN
(infinite source)
|
0.715
|
12,000
(steady state)
|
43.6
|
BIOSCREEN
(declining load)
|
0.443
|
4,525
|
27.0
|
Further agreement between the models in the sand aquifer was achieved buy using the AT123D
load as the BIOSCREEN source term mass. Although, the BIOSCREEN concentrations were still
significantly higher contaminant travel times to downgradient locations were almost identical.
However, this was not true of the silt and clay aquifers.
Discussion
Results between the models are most similar for the sandy aquifer. This is most likely because
the flow rates are high enough that diffusion has limited influence on overall contaminant
transport. This is especially true of contaminant mobility when the AT123D load was used to
calibrate the BIOSCREEN declining source term in the sand aquifer.
BIOSCREEN also predicted higher concentrations than AT123D for the silt and clay aquifers, and it
severely underestimated contaminant mobility. The limited contaminant mobility may be due to the lack
of the water diffusion process in BIOSCREEN. However, it could also be due to variations in the
methods used to simulate other transport and fate processes. This raises concerns for those who
believe that it is not necessary to model such long periods
of time. This is based on the assumption that the contamination will never reach that location at
levels of concern.
Conclusions
AT123D consistently predicted lower groundwater concentrations for all aquifer types. This would
result in higher risk-based cleanup objectives and a reduction in remedial activities. As AT123D
is a more complex model it also provides for greater confidence in the results. Not only does
AT123D provide lower results, it can be used over a larger range of contaminant release scenarios.
Based on this review the following conclusions were made:
- Both models are fairly easy to setup and run.
- The SEVIEW Point of Compliance reports made evaluation of the AT123D results much easier.
- Contaminant mobility is significantly lower in BIOSCREEN.
- Predicted contaminant concentrations are significantly higher in BIOSCREEN.
- The experimental declining source term in BIOSCREEN should not be used.
- Limited contaminant mobility in BIOSCREEN raises serious concerns; at the very least
it should be run until the maximum concentration is obtained.
Given these results it may appear that AT123D may not be conservative enough to meet regulatory
requirements, however, this is not the case. In fact numerous regulatory agencies have used
AT123D to develop their baseline cleanup objectives. What this review indicates is that BIOSCREEN
is even more conservative than AT123D.
|
|
|
