Analytical and numerical modelling of transient convective transfers on SAGD performance

The steam-assisted gravity drainage (SAGD) method has shown promise for efficiently recovering many heavy oil and oil sand reservoirs in Canada. In this study, a comprehensive review of thermal recovery literature enabled a comparison of different approaches to simulate transport processes in various steam injection recovery processes. Studies by Butler and coworkers on the SAGD subject served to explain the temperature distribution in the reservoir ahead of the steam zone, and how it influenced oil drainage rate.; A new analytical model was developed to simulate the expanding steam zone in the SAGD process. This new model was based mainly on the Stefan Phase-Change problem, and incorporated convective transfers taking place in the reservoir. These convective heat transfers included heat losses to the cap and base rock, and heat removed from the reservoir in the produced fluid stream. A rigorous mathematical solution was obtained for this new analytical SAGD model, which enabled the steam zone interface position and velocity to be determined at different times. Results from the new SAGD analytical model agreed mostly with results obtained from CMG's STARS™ thermal simulator, and results calculated from Butler's models. The agreement in the oil production rate between the new analytical model and STARS™ simulator was good for the middle-time period. Difference in the predicted oil production rates in the early-time period was due to the assumption that slope drainage is present from the start of the process by the proposed analytical model. Different external boundary conditions led to the difference in the predicted oil production rates for the new analytical model and the simulator in late-time period.; Results from the new analytical SAGD transient model agreed with corresponding results calculated from Butler's models. The steam zone interface velocity predicted by the new analytical model at different times agreed quite closely with those calculated by Butler's LinDrain (1994) model, with the agreement being closer in later-time periods. Oil production rate predicted by the new analytical model was less than those predicted by the TANDRAIN model, and Butler's 1985 model. The reason was believed to be due to the heat loss to cap rock and base rock, and due to heat removed due to fluid production from the reservoir; these losses are accounted for in the new analytical model.; After the new transient analytical model was validated, it was used to investigate the effects of flow potential and thermal diffusivity on the SAGD process. The results in these cases showed that the steam zone advanced further into the reservoir for low fluid production potential and high thermal diffusivity. In the first case, low fluid production potential resulted in less heat being removed from the reservoir, and hence, more heat accumulated inside the reservoir and contributed to the forward movement of the steam zone. In the second case, high thermal diffusivity value promoted the expansion of the steam zone into the reservoir. Effects of a water sand in communication with the oil reservoir on SAGD performance were also investigated, using CMG's STARS™ simulator. The presence of a bottom water layer was determined to have a lesser impact on recovery than the overlying water layer case. Oil recovery was determined to decrease with increasing water layer thickness. Increasing the area) coverage of the bottom water layer resulted in only a slightly reduced recovery as compared with the confined bottom water layer. However, increasing the areal coverage of the overlying water layer severely reduced the recovery efficiency of the process, as heat was diverted into the overlying water zone.