Benzene Leaching Assessment in Alaska
GURDARSHAN S. BRAR
Hetrick and Travis 1988). SESOIL was incorpo-
INTRODUCTION
rated into a system called PCGEMS (Graphical
The United States Army Engineer District, Alas-
Exposure Modeling System for the PC), a com-
ka, Environmental Technical Engineering Office,
plete information management tool developed
asked CRREL for technical assistance in evaluating
for EPA-OTS and designed to help users perform
the SESOIL model for aiding in assessing benzene
exposure assessments (General Science Corpora-
pollutant leaching in Alaska. The major tasks out-
tion 1987, 1989). General Science Corporation
lined in the request were 1) to work an example
(1990) further made improvements to PCGEMS
problem for a diesel-contaminated site, 2) do ana-
and incorporated it into a new system called
lytical checks and 3) make manual SESOIL an-
RISKPRO (Hetrick et al. 1993).
alytical calculations for one cycle.
Drewett et al. (1993) used the SESOIL model at
Objective
Fort Greely to assess the presence and concentra-
The objective of this report was to evaluate the
tions of petroleum hydrocarbons in the soil and
sensitivity of the SESOIL model for assessing ben-
zene leaching in the Alaskan environment.
sensitivity analysis was made for benzene con-
tamination in Building 110 at Fort Greely by using
input data files of Drewett et al. (1993).
LITERATURE REVIEW
Donigian and Rao (1986a) comprehensively re-
SESOIL history
SESOIL stands for seasonal soil compartment
viewed the model and concluded that the chemi-
model. It is a one-dimensional vertical transport
cal transport part of the code needed improve-
model for the unsaturated soil zones and is de-
ments. The model did not handle the large
signed to predict solute distribution in the soil
variations in the hydraulic properties of the soil
profile and watershed seasonally. In 1981, Arthur
profile. Furthermore, they reported that SESOIL
D. Little, Inc. (ADL) developed the program for the
overpredicted the mass of aldicarb movement to
Environmental Protection Agency's (EPA) Office
of Water and Office of Toxic Substances (OTS). The
erty (1993) found that SESOIL does not simulate
original code included three soil columns. In 1984,
ADL updated the model and included a fourth soil
the aqueous phase, but assumes that all contami-
layer as well as soil erosion algorithms (Bonazoun-
nants are dissolved.
tas and Wagner 1984). In 1985, the EPA's Environ-
Ladwig et al. (1993) tested the SESOIL model
mental Research Laboratory in Athens, Georgia,
for benzene, ethylbenzene, toluene, xylene and
thoroughly evaluated SESOIL (Watson and Brown
1,2-dichloroethane under Wisconsin conditions
1985), found numerous deficiencies in the model
and discovered that the model was most sensitive
and recommended improvements. Furthermore,
to soil type, biodegradation rate, residual concen-
the model was extensively tested at the Oak Ridge
tration, residual layer thickness and organic car-
National Laboratory (ORNL) and modified to en-
bon content of the soil. The model was least sensi-
hance its capabilities (Hetrick et al. 1986, 1989;
tive to climate, soil disconnectedness index, pH