later in the flood than the other depths. In most MLS locations, the time of peak PCE

concentration was achieved prior to the arrival time of peak ethanol with the exception of the

top-most sampling location where there was an observed delay. At MLS-1, peak PCE attainment

time was slightly delayed compared to the peak ethanol appearance with only one exception, the

lowest depth. The highest removal concentrations were detected in MLS-1, MLS-3, and MLS-4

by almost one order of magnitude.

 Cumulative distribution functions (CDFs) were generated from the MLS flood results.

Figure 2-9 presents the CDFs for the maxima of both the percent ethanol and the percent of PCE

solubility achieved. This was the same process as the scaling performed on aqueous PCE

concentrations in the source zone in sections 2.5, 2.6, and 2.10 except reported as a percent.

Beginning with the PCE results, 60% of the MLSs achieved 10% or less of maximum solubility

and 90% attained less than 50% of maximum solubility. While only 20% of the MLSs received

less than 85% ethanol concentration. These results indicate that high concentrations of ethanol

reached most MLS zones, but few achieved high maximum solubility percentages. Although it

appears that remedial fluid interrogated the source zone thoroughly, PCE solubilization may be

limited by additional factors besides merely the presence of cosolvent. The actual flow path of

the remedial fluid in the subsurface is not known. Although high ethanol may reach a MLS, it

did not have to contact PCE. As shown in Figure 2-5, higher PCE saturations were observed in

regions of lower fluid velocities. This is a result of lower media permeability, resultant PCE

pooling, and the dissolution removal of PCE in higher flow zones. Thus, the main limitation of

solubility enhancement was demonstrated, the inability to contact and remove all DNAPL.

However, this pilot test verified the ability of ethanol to enhance and deplete PCE in well flushed

regions. This will be further explored below.