Environmental Engineering Reference
In-Depth Information
a year of the spill. Methyl chloroform concentrations in groundwater were up to 11,000
g/L; a
pump-and-treat system removed 41 million gallons of contaminated water until 1994, when the
system was shut down because cleanup goals had been attained at most monitoring wells. By 2003,
methyl chloroform concentrations had decreased to 2.3
μ
μ
g/L, with 5.1
μ
g/L 1,1-dichloroethylene
and 4.7
g/L 1,1-dichloroethane. The results of analyses for 1,4-dioxane at the site were nondetect
at less than 2
μ
g/L in all monitoring wells, based on one-time sampling in 2002, 18 years after the
initial release (Cal EPA, 2003).
Using data from monitoring wells placed within the 500-feet-long methyl chloroform plume at the
Santa Clara site, Wing (1997) plotted time-series concentration data from two groundwater extrac-
tion trenches as the log of the molar ratios of the sum of 1,1-dichloroethylene and 1,1-dichloroethane
to methyl chloroform against sample dates. Because no data for acetic acid was available, Wing
applied the approximate kinetic ratio of 3:1 (molecules acetic acid to 1,1-dichloroethylene produced
from methyl chloroform) and multiplied the molar concentrations of 1,1-dichloroethylene by four.
The specii c ratio plotted to interpolate the timeframe of release was (Wing, 1997)
μ
[
]
[
]
Ê
ˆ
4 1,1-dichloroethylene
+
1,1-dichloroethane
ln
.
Á
˜
[
]
methyl chloroform
+
1
Ë
¯
Wing's analysis corroborated the i nding by McNab and Narasimhan which showed that degrada-
tion effects overwhelm retardation effects. The slope of the linear regression line (correlation coef-
i cient
r
2
0.81) was 0.298 year
−1
, from which an overall half-life of methyl chloroform of 2.3 years
was obtained. To account for only the elimination reaction that transforms methyl chloroform to its
i eld-measured degradation product, 1,1-dichloroethylene, Wing removed 1,1-dichloroethane and
the factor of four that accounted for acetic acid to obtain a slope of 0.243 year
−1
. The resulting half-
life of methyl chloroform for the elimination reaction was determined to be 2.9 years (Wing, 1997).
The study of i eld data by Wing produced unexpectedly accurate results, providing encourage-
ment for applying this technique where an episodic release is suspected (e.g., a plume of unknown
origin that may have originated as a “midnight dumping” episode). However, caution is in order
despite the high degree of accuracy obtained by Wing. Sources of error and uncertainty include
hydrodynamic dispersion and molecular diffusion into i ne-grained sediments, which can be
expected to decline with distance from a continuous source (McNab and Dooher, 1998). The num-
ber, placement, and construction of monitoring wells may invalidate assumptions of a continuum
between data points where anisotropy creates preferential pathways (e.g., paleochannels) affecting
some monitoring wells and not others. Heterogeneity of hydraulic conductivity, redox conditions,
and temperature may all affect the ratios used to interpret the timeframe of release. Seasonal varia-
tions in hydraulic conditions such as water level l uctuations or seasonal recharge causing variation
in groundwater l ow direction may also confound the application of ratios to estimating the timing
of a release. To determine the extent to which these factors may be inl uencing the ratio data at a site,
contrasting the concentrations of methyl chloroform and its degradation products to the concentra-
tions of 1,4-dioxane where it has been established as a cocontaminant is recommended.
=
9.2.3 M
ODELING
1,4-D
IOXANE
M
IGRATION
TO
A
PPROXIMATE
THE
D
ATE
OF
A
R
ELEASE
Two modeling studies used the 1,4-dioxane data set from the Gloucester Landi ll (described in
Section 3.4.2) to rei ne the methods of solving the inverse problem to estimate the probable timing
of a 1,4-dioxane release to groundwater in a special waste unit. Laboratory wastes, presumably
including scintillation wastes, were discharged to a trench and periodically burned. Wastes contain-
ing 1,4-dioxane leached from the trench to groundwater and migrated about 1600 ft down-gradient.
The modeling studies sought to solve the linear inverse problem by using assumed values for con-
taminant transport parameters such as lateral and transverse dispersivity.
Search WWH ::
Custom Search