Page 365 - Enhanced Oil Recovery in Shale and Tight Reservoirs
P. 365
Forced imbibition 337
The literature on shale oil rocks indicates that they are most likely oil-
wet (Phillips et al., 2009; Wang et al., 2011). This oil-wet condition makes
it difficult for the aqueous phase to penetrate the matrix and displace the oil
out. Surfactants can alter the rock wettability from oil-wet to water-wet or
mixed wet (Sheng, 2012). Therefore, most of surfactant-related EOR
studies for shale oil reservoirs focus on wettability alternation and water
imbibition (e.g., Shuler et al., 2011; Wang et al., 2011; Ferno et al., 2012;
Xu and Fu, 2012; Morsy and Sheng, 2014b). Those studies used very thin
slices or small cores because the spontaneous imbibition process is very
slow (Sheng, 2013b). In practice, if the matrix is too large, the recovery
rate by spontaneous imbibition will be uneconomically slow, because the
imbibition rate is inversely proportional to a characteristic length, either
linearly, or squared (Mattax and Kyte, 1962; Cuiec et al., 1994; Kazemi
et al., 1992; Li and Horne, 2006; Ma et al., 1997; Babadagli, 2001). To solve
this problem, we need to implement forced imbibition to speed up the
imbibition process, like in the fracturing process. Apparently, refracturing
shale reservoirs improves recovery (Vincent, 2011). Huff-n-puff surfactant
injection may contribute to the process of refracturing and thus enhance
oil recovery. Interestingly, huff-n-puff is expected to work better than sur-
factant flooding in shale reservoirs. However, we did not see a huff-n-puff
field case report. Instead, a surfactant flooding in a Bakken formation was
studied.
A detailed characterization program consisting of logging, coring, pres-
sure testing, and fluid tracing was carried out to build a reservoir model to
evaluate the EOR potential of surfactant flooding in a Middle Bakken for-
mation. The reservoir model is also based on calibrated parameters from
history-matching experiments like surfactant adsorption parameters. The
permeability in the modeled area is 100 nD to 10 mD. The model consists
of a pair of horizontal wells with half-fractures from each well overlapping
60% but not connected. Only half of a fracture from each well is included in
the model. The base half-fracture length is 1200 ft, and fracture height is
300 ft. The model predicts that the produced oil in 12.5 years is more
than what can be produced from 12.5 years of primary production. A series
of economic sensitivity studies shows that the surfactant injection could have
not only EOR potential, but also economic potential. However,
detailed data in the model were not reported in the paper by Dawson
et al. (2015).

