The optimization of chemical floods for a low-permeability, high-salinity, high-temperature carbonate formation presents several unique design challenges, which may be overlooked in short coreflood experiments. Low permeability formations require the optimization of polymer molecular weight distribution such that both mobility reduction and injectivity are maximized. The effects of high core-scale heterogeneity must be understood and mitigated, for example, by increasing core length. These subjects are explored in the context of an ongoing flood design project. Understanding and mitigation of pressure build-up issues -possibly related to high surfactant adsorption -is an ongoing challenge. Additionally, an attempt will be made to shed light on the discrepancy between available carbonate capillary desaturation curves for carbonates and the near-complete recovery reported in surfactant-polymer corefloods.
Effect of dispersivity on chemical floodsGash and Griffith (1980) suggest that, for rocks with low dispersivity, such as Berea, cores as short as 30 cm may yield high recovery in laboratory chemical floods. Clastic reservoirs often contain high-correlation length heterogeneities (e.g. stratification), which, when averaged across an entire section, may add a length scale-dependent megascopic component to dispersion measurements (Arya et al., 1988;Gelhar et al., 1992;Barth et al., 2001). Thus, Arya et al. (1988) differentiate between megascopic dispersivity, which controls volumetric sweep efficiency, and macroscopic dispersivity which determines real mixing in porous media and is under interest of miscible EOR. The ratio of mean correlation length of permeability field to mean distance travelled is proposed as a factor of nature of mixing: if this ratio is much less than one, dispersivity represents true mixing, and is proportional to the mean correlation length. In the opposite case, measured dispersivity results from megascopic heterogeneity as well as dispersion. The criteria α/L<<1 can thus be applied to determine whether the nature of dispersion measurements is truly mixing, or rather an artifact due to large scale heterogeneity. Too much of either can be detrimental to chemical flood performance, the former yielding dilution of the surfactant slug, the latter yielding poor sweep.