but for this paper it will be referred to as in-situ alcohol flushing or ethanol flooding. Alcohol

flushing has been demonstrated to be an effective method for enhancing the removal of DNAPLs

from sand in laboratory tests and from the subsurface in sandy aquifers [Augustijn et al., 1994;

Imhoff et al., 1995; Augustijn et al., 1997; Rao et al., 1997; Sillan et al., 1998; Jawitz at al.,

2000; Falta et al., 2000; Brooks et al., 2003].

 There is always a risk of mobilization when using alcohols as the cosolvent for in-situ

flushing [Lunn andKueper, 1999; Padgett andHayden, 1999]. Ethanol was chosen by LFR

[1998] from ternary phase diagrams because it exponentially increases PCE solubility but

provides a lower risk of PCE mobilization [Falta, 1998]. Shorter chain alcohols tend not to

lower the interfacial tension (IFT) associated with entrapped DNAPL, reducing the binding

capillary forces. Thus, the risk of IFT reduction and downward mobilization of DNAPL during

flushing is reduced when employing ethanol as the solubilization enhancing agent [Lunn and

Kueper, 1999].

1.7 Site Characterization

 The hydrogeology of the Sages site was characterized by soil cores, pumping tests, water

table levels, and an electromagnetic borehole flowmeter test. The results will be summarized

here but extensive studies at the site have been performed and published [LFR, 1997, 1998a,

1998b; Sillan, 1999; Jawitz et al., 2000; Mravik et al., 2003]. The water table is approximately 3

m bgs, and the natural hydraulic gradient is 0.0025. The media was characterized to be fine to

very fine sand down to 9 m bgs. In the lower zone, from 9 m to 10.7 m, very fine to silty sand

was observed down to the discontinuous clay layer at 10.7 m bgs. Hydraulic conductivities were

estimated at 6 m/day in the upper region and 3 m/day in the lower. Combined with the low