Technology Status Report: In Situ Flushing - CLU-IN

In Situ Flushing Project Summaries
GWRTAC Case Study Database
surface (bgs). At Hill AFB, OU1, the LNAPL has over 200 identified components. The major
constituents of the Hill OU1 LNAPL on a mass basis are decane, 1,3,5-trimethyl benzene and
undecane. Seasonal fluctuations in the water table smeared the LNAPL over the interval 5 m (16.4
feet) bgs to 8 m (26.2 feet) bgs. The sheet pile-enclosed cell measured 2.8 m x 4.6 m (9.2 x 15.1
feet) and penetrated 3.7 m (12.1 feet) into the clay confining unit. Leak testing of the sheet pile
cell confirmed that leakage of the sheet pile cell was insignificant. The water table within the test
cell was maintained at 4.9 m bgs. The test cell held fifteen multilevel samplers, three extraction
wells and four injection wells. The injection wells and extraction wells were screened from 4.9 m to
7.9 m (16.1 to 25.9 feet) bgs. At each of the fifteen multilevel sampler locations, from five to eight
individual samplers were bundled to yield a total of 96 multilevel sampling points. Peristaltic
pumps maintained a flow rate of approximately one pore volume per day.
The extent of contamination within the test cell was assessed by collecting soil samples during well
installation and analyzing for a selected group of target analytes present in the NAPL. The test cell
contamination was also characterized by conducting a 10-pore volume tracer test with a group of
partitioning and non-partitioning tracers. The test cell had an average NAPL saturation of about
0.06 prior to SPME flushing. Water samples were also collected and analyzed for the target
analytes, however, nearly all the contaminants were below detection limits. Finally, before the
SPME flood, contamination in the test cell was characterized using interfacial tracers. The method
estimates the contact area between NAPL and water in the test cell. The interfacial tracer method,
developed at the University of Florida, was tested for the first time in the SPME test cell at Hill
AFB. The result indicated that the NAPL-water contact area can vary more than one order of
magnitude within the test cell. This variability in NAPL-water contact area could have significant
impacts on in-situ flushing extraction efficiency. If there are mass transfer limitations that are
magnified with low NAPL-water contact area this will reduce flushing efficiency, or if the NAPL is
simply non-uniformly distributed, as evidenced by a low NAPL-water contact area, this can lead to
inefficiencies.
The SPME flood was conducted over a 18-day period by pumping nine pore volumes of the
precursor solution with some flow interruption periods. In the field test, a non-reactive tracer test
was conducted first to determine the hydrodynamic performance of the cell and then set the flow
rate for the rest of the flushing test. Next, a 0.16 pore volume pulse of partitioning tracers,
methanol (2700 mg/L) and 2, 2-dimethyl 3-pentanol (930 mg/L), flowed through the sheet pileenclosed
cell before the SPME test. The distribution and morphology of NAPL in the test cell was
also characterized with interfacial tracers that were first developed and tested during this project.
After the pre-flood tracer tests, 3.5 pore volumes of water were flowed through the test cell. Nine
pore volumes of the surfactant and alcohol mixture (3% Brij 97 (polyoxyethylene (10) oleyl ether)
and 2.5% n-pentanol) were perfused through the test cell. The viscosity of the surfactant + alcohol
mixture, measured in the field, was 1.66 centipoise. The surfactant/alcohol mixture was followed
by surfactant alone, a water flood, and a final partitioning tracer test.
To assess NAPL removal, the investigators collected soil cores before and after the addition of the
surfactant and alcohol mixture; they also conducted pre- and post-flush NAPL partitioning tracer
tests. Fluid samples were collected from the extraction wells (540 total) and multilevel samplers
(6500 total). Eighty-five pre-flush soil core samples and 40 post flush soil core samples were
collected from the test cell. Soil cores were sub-sampled and immediately extracted in the field.
Soil core and fluid samples were analyzed for target LNAPL components (p-xylene, n-decane, nundecane,
n-dodecane, n-tridecane, 1, 2 - dichlorobenzene, 1, 2, 4 - trimethylbenzene,
Ground-Water Remediation Technologies Analysis Center
Operated by Concurrent Technologies Corporation
Appendix - Page 38 of 164
Copyright GWRTAC 1998
Revision 1
Tuesday, November 17, 1998