Doctrines and Realities in Reservoir Engineering

Abstract

Abstract Doctrines often develop with minimal empirical support. This observation is clear in reservoir engineering for heavy and viscous oils. Our objective is to develop a deeper appreciation of the empirical realities to permit improvement of depletion plans and thereby enable projects that are deemed uneconomic by the application of the misleading doctrines. Accordingly, this paper reviews the evidence for and against three doctrines in current use to develop depletion plans: (i) optimal recovery is obtained using a voidage replacement ratio (VRR) of 1, (ii) the Buckley-Leverett formulation applies uniformly to heavier oils, and (iii) viscous fingering dominates unstable multiphase flows. Of primary importance is the doctrine that optimal waterflood response is obtained for a VRR of 1 both on an instantaneous and cumulative basis. A VRR of 1 is mandated in some regulatory jurisdictions and accepted unconditionally by many large oil companies. In the last decade, however, empirical, laboratory, and simulation investigations have demonstrated that, particularly for heavier oils, periods of VRR < 1 significantly increase the oil recovery and reduce water production. Optimal voidage management increases oil recovery by an amount similar to a variety of EOR techniques at minimal cost. A second doctrine is that the multiphase flows within the reservoir occur via the slipping of phases past each other, as formalized by the Buckley-Leverett formulation. Again, empirical and laboratory evidence demonstrates that in many cases two-phase flow occurs by emulsification of one phase, e.g., injected water disperses as small micron size droplets within a continuous oil phase. Such emulsion flows yield results that contradict current doctrines; e.g., decline curves are not monotonic but have steps associated with the type of emulsion. Likewise, the efficacy of polymer flooding is independent of polymer concentration after it exceeds the minimum value needed to support emulsification of water into the oil. These novel predictions have considerable commercial value. Finally, viscous fingering is often attributed to the instability of flow when the displacing phase (water) has a viscosity smaller than that of the displaced phase (oil). Though elegant mathematical formulations explain such unstable flows under idealized laboratory conditions, the reality is that slow pressure diffusion within heavy oils is the originating factor in channelized reservoir flows because commercial injection rates cause altered porosity distributions that initiate the preferential path flow. The appreciation of the empirical realities permits improvement of commercial depletion planning and enables a greater number of projects. These concepts are applicable in almost all heavy-oil reservoirs; they may also be applicable for lighter oils that possess oil chemistry that is typical of heavier oils, e.g., high acid content.