Comparison with University of Kansas Work

Researchers at the University of Kansas advocated dehydration as the dominant mechanism for re-establishing oil pathways through the gel.23-25 In recent work,25 they studied disproportionate permeability reduction provided by Cr(III)-acetate-HPAM gels in 4- to 5-darcy sandpacks. Using tracer studies with stilbene, they confirmed our finding of a high degree of connection for the oil phase.3,28 They also confirmed our finding that Sor in water-wet porous media was significantly higher after gel placement than before gel placement.1,3,7

Although the University of Kansas researchers argued in favor of a dehydration mechanism, confusion exists about their definition of “dehydration.” In addition to the definition that we accept (stated above), Ref. 25 contains at least two other definitions. A second definition (from Tables 4-6 in Ref. 25) is: [increase in oil saturation during oil injection after gel placement] relative to [gel saturation immediately after gel placement]. The third definition is that “dehydration” simply means oil injection. This ambiguity was common throughout Ref. 25.

In our view, a detailed analysis of their data29 indicates that ripping and gel displacement mechanisms dominate over the dehydration mechanism. A key revelation from analysis of the data in Ref. 25 was that a significant amount of polymer was produced with the effluent during oil injection. Typically more than half the polymer was recovered! This indicated that a gel destruction or removal mechanism other than dehydration was very important. Thus, removal of polymer from the core played an important role in establishing oil pathways, and removal of this polymer was not directly tied to dehydration. Instead, gel destruction or removal might be caused by (1) oil ripping the gel apart so it was basically a polymer solution or (2) displacing gel from the core as very small particles (too small to detect).

To quantify the relative importance of dehydration versus the destruction/displacement mechanism, consider experiment TN003 from Ref. 25. The “fraction of gel dehydrated” was listed as 68.5% and the fraction of polymer recovered was 45% (in Tables 4 and 5 of Ref. 25). Oil injection created open space in the gel equivalent to 59.7% of the original pore space (87.1-27.4 from Table 8 of Ref. 25). Recovery of 45% of the polymer suggests that 65.6% of the pore space that was opened by oil injection (i.e., 0.45x87.1/59.7) was actually caused by destruction or removal of the gel—not by dehydration. Performing these calculations for all cases in Ref. 25 shows that destruction/removal of gel was more important than dehydration.

A close comparison of Tables 5 and 6 in Ref. 25 indicates that raising the oil-flooding pressure gradient from 20 to 50 psi/ft caused no apparent additional dehydration. (In all cases, the average remaining polymer concentration (in the core) decreased when the pressure gradient was increased from 20 psi/ft to 50 psi/ft.) Consequently, only the gel destruction/removal mechanisms were significant.

In summary, despite claims in support of dehydration, the data in Ref. 25 supports ripping or gel displacement mechanisms over a dehydration mechanism. We note that the permeability of the sandpacks were from 4 to 5 darcys in the work at the University of Kansas. In contrast, our Berea core had a permeability of 0.47 darcys. One would expect ripping or gel displacement mechanisms to become more important as the permeability increases for water-wet porous media. Consequently, it is quite conceivable that the dehydration mechanism dominated in our 0.47-darcy Berea, while the ripping or gel displacement mechanisms dominated in the sandpacks investigated at the University of Kansas.