In many heavy oil reservoirs in Alberta and Saskatchewan, a high water saturation zone occurs below or above the oil zone. Such water zones can take a number of forms, and in special circumstances can even be beneficial. However, in most cases, they cause channelling in steam injection operations, leading to low oil-steam ratios. This paper considers cyclic steam stimulation in particular, in the presence of high water saturation zones. The results are based upon simulation studies for Lloydminsrer and Cold Lake reservoirs.
It is shown mat cyclic steam stimulation response depends strongly on: water-oil zone thickness and permeability ratios, vertical permeability, oil viscosity, steam injection rate and steam injection interval. Partial penetration is one of the options that can help. For a given set of circumstances, operation may still be viable in the presence of a communicating water zone. Top water is likely to be more undesirable than bottom water.
Communicating water zones are a frequent occurrence in the heavy oil formations of Alberta, Saskatchewan, California, and other areas. Most of the time, such zones occur below the oil zone. In some cases, in Athabasca, Cold Lake, and California, they lie above the oil zone. A water zone can take a number of forms: it can be a transition zone, it can be a low oil saturation zone, or it can be a 100% water zone. or even an aquifer. Frequently, the mineralogy of the high water saturation zone differs considerably from that of the oil zone. Local geology and well log interpretations are important for assessing me effect of communicating water zones on cyclic steam stimulation response In particular, me formation description derived from well logs is of great value in cyclic steaming formations with contiguous water zones. This paper presents selected results of a simulation study to examine the effect of communicating high water saturation zones adjacent to an oil-bearing formation.
The reservoirs considered in this study were a typical Lloydminster formation (rypified by the Aberfeldy reservour) and a typical Cold Lake formation. Relevant darn. for the simulations is given in Table 1; Figs. 1 and 2 give the oil-water and oil-gas relative permeability curves used in the simulations. The oil viscosity-temperature curves are shown in Figs. 3 and 4, for the Aberfeldy and Cold Lake formations, respectively.
A three-phase, three-dimensional fully implicit Steam-water-oil simulator was developed for [his work. It [feats oil as a single non-volatile component. Thegrid used consisted of 5x5x2 or 5x5x3 blocks, representing a quadrant for single well stimulations discussed here. A typical complete run (three cycles, each consisting of 30 days of injections days of soak, and 90 days of production) took 200 to 700 seconds (CPU) of Amdahl 8080 time. The material and heat balance errors were well below 0.1 %.
A factor that can influence simulation results is the well production control in the present work, a productivity index of 10 was employed with a minimum bottomhole pressure limit.
