Unstable Waterflood Performance Diagnostic Methods

Abstract

AbstractThis paper discusses a review and adaptation of some classic waterflood performance analytical methods, such as X-plot, comprehensive Y-plot (cY-plot), and WOR vs cumulative oil (Np) for the case of unstable immiscible displacement (viscous-oil fingering effect). These methods were reviewed based on fractional flow analysis (FFA) for unstable immiscible waterflood.These classic techniques account for the solution of the one-dimension frontal advance Buckley-Leverett theory (1942), assuming stable flow. In addition, the traditional semilog linear relationship between oil-water relative permeability ratio and water saturationis assumed (constant parameters A and B). Those assumptions tend toover predict ultimate oil recovery for the case of viscous-oil waterfloods because flow functions do not capture the viscous fingering effect.This work proposes to redefine aforementioned classic waterflood performance analytical methods with novel oil and water relative permeability expressions derived from the effective-fingering model(EFM) presented by Luo et al. (2016), which accounts for viscous fingering effects. In addition, an accurate exponential expression of kro/krw ratio as function of water saturation and an exact solution for a water saturation-dependent parameter B (named Bj) are proposed. New approaches of classic analytical methods were derived, and both laboratory and field cases were tested at the light of new equations.Adaptation of classic equations (stable) to solutions that account for unstable flow results in more reliable diagnostic-plot techniques for the case of viscous-oil, allowing to correct predictions of oil and water production in the case of heavy-oil waterflooding Additionally, new equations resulted in unified solutions that can be applied for both stable and unstable waterflood and help to improve reliability when estimating ultimate oil recovery, volumetric sweep efficiency, and various reservoir parameters. In the presence of viscous fingering, the water breakthrough and oil recovery from new X, cY, and WOR functions are viscous-finger number dependent (Nvf). The bigger the Nvf the lower the oil recovery, the earlier the water breakthrough, and the narrower the water saturation ranges.In its entirety, these novel waterflood performance analytical methods incorporate viscous fingering features in the traditional flow functions, encouraging the ability to predict ultimate oil recovery for both unstable and stable waterflooding cases and for chemical flooding (i.e., polymer with future adaptation) in heavy-oil reservoirs and facilitating the optimization of heavy-oil enhanced oil recovery (EOR) projects. These results might provide a basis to adapt other classic waterflood performance analytical methods.