Environmental Fate and Transport of Solvent-Stabilizer Compounds - Environmental Investigation and Remediation - page 170

DUKE FOREST LANDFILL SOIL SORPTION
STUDY (BALL AND BARTLETT, 1992)
A screening-level soil sorption study of 1,4-dioxane transport in native soils was conducted at
Duke University. A landi ll near the university had received wastes containing 1,4-dioxane,
and monitoring wells showed concentrations of several thousand micrograms per liter.
Researchers performed soil sorption tests on sieved and homogenized cores of aquifer sol-
ids retrieved from below the landi ll site to verify that 1,4-dioxane would not bind strongly
to soil material. Soil samples sterilized with sodium azide were mixed with water and 1,4-
dioxane. For 4 days, the slurry was rotated to maintain active mixing; then it was centrifuged
to separate the aqueous-phase solution from the soil for analysis. Because of 1,4-dioxane
losses in sample handling (13%) and low analytical recovery on some analyses (e.g., 44%),
the distribution coefi cient K d could not be reliably estimated. Nevertheless, the study clearly
demonstrated that the potential for 1,4-dioxane to sorb to soil is low. The maximum estimated
retardation factor was 4.2, that is, 1,4-dioxane would not move slower than one-quarter the
rate of groundwater l ow. A related study summarized previous investigations of the Duke
Forest landi ll site and noted that a consultant report determined an average retardation factor
of 1.2 (Liu et al., 2000).
organic compound to the length of the chloride plume, which presumably originated at the same loca-
tion and time and which is assumed to have a retardation factor of 1.0 (Patterson et al., 1985). A com-
parison of retardation factors determined with the plume-length approach is given in Table 3.22 .
Table 3.22 shows that estimates of retardation factors calculated with the Schwarzenbach and
Westall equation can be as little as one-sixth the i eld-measured values. The Schwarzenbach and
Westall equation does not adequately account for adsorption to mineral surfaces; it was derived
from data for hydrophobic polar compounds, which do not sufi ciently represent the behavior
of hydrophilic compounds such as 1,4-dioxane. Table 3.22 also demonstrates that migration of
1,4-dioxane is only minimally retarded. It therefore migrates substantially faster than most other
organic contaminants.
3.4.3 M ODELING 1,4-D IOXANE T RANSPORT
Modeling tools can provide useful insights into the relative rates of migration of contaminants
under assigned subsurface conditions such as groundwater l ow rate, organic matter content of
soils and aquifer solids, retardation factors, biodegradation rate constants, abiotic rate constants,
source mass strength, release duration, and other factors. There are many analytical approaches
and many models available to simulate contaminant transport. The subject of modeling contami-
nant transport is beyond the scope of this topic; however, a brief exercise using a screening-level
model is proi led here to provide a sense of the relative calculated rates of migration of 1,4-dioxane,
other stabilizer compounds, and methyl chloroform. The screening-level model reviewed here is
BIOCHLOR v. 2.2, a spreadsheet model based on modii cations to the Domenico analytical solute
transport equation, which can simulate one-dimensional advection, three-dimensional dispersion,
linear adsorption, and biotransformation via reductive dechlorination for chlorinated solvents
(Aziz et al., 2000, 2002). BIOCHLOR is used primarily to evaluate the efi cacy of employing
monitored natural attenuation as a remedial solution at chlorinated solvent release sites; the model
is not intended to accurately predict migration of contaminants. BIOCHLOR includes the assump-
tion that biodegradation occurs only in the aqueous phase; therefore, the model ignores degrada-
tion of contaminants sorbed to soil and aquifer solids. When users of BIOCHLOR set longitudinal
dispersivity to high values, BIOCHLOR is prone to errors in predicting migration when applied to