Summary Steam-assisted gravity drainage (SAGD) is a commercially successful recovery process that produces heavy oils and bitumen. The method ensures both a stable displacement front of steam and economical rates by using gravity as the driving force with a pair of horizontal wells for injection/production. Although several ways of improving the performance have been discussed in the literature, the well configuration employed in the process has remained the same as originally proposed by Butler et al. (1981). A systematic attempt to improve the performance by using radically different well configurations has not been reported. This paper presents a study intended to examine the applicability of a new well configuration to SAGD processes in Athabasca and Cold Lake reservoirs in central and northern Alberta. The fully implicit thermal-reservoir simulator, CMG's STARS 2007, with fully coupled wellbores was used to account for frictional pressure drop and heat losses along the wellbore. 3D numerical simulation models were set up, and sensitivity analyses were conducted on injection pressure. After optimization of the injection pressure, an investigation of new well configurations was conducted using these models. The result of this work shows that the SAGD-process performance in Athabasca and Cold Lake reservoirs can be improved significantly by changing the well configuration.

Summary The solvent thermal resource innovation process (STRIP), a downhole steam-generation technology, has the capacity to show improved recovery factors with a significantly reduced environmental footprint compared with traditional thermal-enhanced-oil-recovery (TEOR) methods, most notably by delivering all the combustion heat to the pay zone. In this effort, a quarter-symmetry inverse-five-spot model and a multiphase, multicomponent reservoir-simulation framework were used to simulate the STRIP technology. Commercial simulators such as STARS - Thermal and Advanced Processes Reservoir Simulator [Computer Modelling Group Ltd. (CMG), Calgary, Alberta, Canada; CMG 2015b] often use the K-value approach to simulate TEOR. However, the method cannot simulate STRIP's carbon dioxide (CO2) and steam coinjection because the K-value method does not consider miscible gas injection. On the other hand, CMG's GEM - Compositional and Unconventional Simulator (CMG 2015a) includes the effects of miscible gases but does not provide comprehensive support for steam-injection processes, which are better handled by STARS. The novel simulation framework developed here leverages and combines the individual strengths of STARS (thermal features) and GEM (compositional features). In this framework, STARS simulated steam injection (but cannot directly simulate the effects of CO2) and was the governing model that synchronized temperature, pressure, and phase saturations for two parallel iterations of the GEM models (GEM-1 and GEM-2) at each timestep. Immiscible methane (CH4) was added to GEM models to maintain gas saturations equivalent to the STARS model. GEM-1 simulated hot-water and CH4 injection, but at increased rates to yield a pressure field and gas saturations equivalent to STARS. A final run of GEM-1 injected both CO2 and hot water and demonstrated the expected increase in oil production. Calibrated injection rates from GEM-1 were specified in GEM-2 to ensure equivalence of the pressure field. Next, the GEM-2 model also simulated hot-water and CH4 injection, but matched both water and oil productions along with oil saturations from the final GEM-1 run by altering relative permeabilities. Finally, the updated relative permeabilities were fed back to STARS, and iteration proceeded. Results from this framework were verified against a STARS steam-injection simulation. Finally, when considering coinjection of CO2, STRIP's superior performance was demonstrated through increased oil recovery and a lower steam/oil ratio (SOR).

No abstract