Compared to solution gas drive, the waterflooding of light oil reservoirs often results in a multifold increase in oil recovery. This striking success has led to the current waterflooding operational strategy, generally mandated by regulatory bodies: completely replace voidage (VRR = 1), and maintain reservoir pressure at or above the bubble point. For viscous reservoirs, however, the incremental benefit of waterflooding over solution gas drive becomes smaller; for sufficiently heavy oils, waterflooding is futile whereas solution gas drive is surprisingly effective due to a decreased rate of gas bubble coalescence. This suggests that, for reservoirs with oils within some range of viscosities, a hybrid waterflooding / solution gas drive process may be optimum, with VRR < 1. Production data from a waterflooded viscous oil reservoir from the Milne Point Unit on Alaska's Norih Slope have been examined. The observed water oil ratios (WOR) do not typically conform to those predicted by numerical simulations using light oil physics and operational strategies. A performance curve of the WOR ratio with time has been developed and divided into four regimes. For the best performing wells, there exists a well developed regime with WORs that are constant (0.5 to 1) for extended periods of time, with a VRR < 1. A conceptual model has been developed to explain the empirical observations and is being tested. The model envisions the injected water initially forming a preferred communication path between the injector and producer, with the injected water mixing with the oil to form water-in-oil emulsions that are driven to the producer. Optimal conditions for emulsion formation require continual oil resaturation of the communication path, most easily achieved by activating solution gas to drive oil via a VRR < 1.
展开▼
