Abstract
As governments around the world prepare for a transition period to a decarbonised energy and economic future, petroleum is needed to smoothen that transition. Based on the analysis of the International Energy Agency's 2020 projections, around 770 billion barrels of oil are required to meet demand from now to 2040. However, according to British Petroleum's Statistical Review of World Energy 2020, as at the end of 2019, the global total reserves of recoverable conventional and unconventional oils is approximately 1734 billion barrels. Out of that, the conventional easy-to-produce light oil accounts for only 30% (i.e. accounts for only 520.2 billion barrels). Therefore, the remaining 249.8 billion barrels of oil needed to satisfy demand up to 2040 must come from unconventional oils, namely heavy oils and bitumen. However, these unconventional resources are very difficult to produce and the current production methods have very high environmental footprints. Consequently, in accordance with climate crisis mitigations, the vast reserves of the virtually unexploited heavy oils and bitumen must be developed using advanced and greener extraction technologies, such as the yet-to-be-fully-understood THAI process which provides partial upgrading of heavy oil/bitumen via in situ combustion. Using validated numerical models which are developed using the CMG's reservoir thermal simulator, the STARS, which is also used in this study, field scale reservoir simulations of the THAI process were performed with the wells arranged in staggered line drive (SLD) and direct line drive (DLD). Over the 834 days of operating time, the cumulative oil recovery in SLD is 32% of oil originally in place (OOIP) which is equivalent to 26,100 m(3) whilst that in DLD is 27% OOIP. This shows that more oil (i.e. an additional 5% OOIP) was cumulatively recovered in SLD compared to in DLD model. It is found that smaller reservoir volume was swept by the combustion front in DLD and thus making the heat-affected reservoir volume smaller than that in SLD model. Furthermore, in DLD, due to the nearness of the injector well to the toe of the horizontal producer (HP) well, oxygen production began much earlier, compared to in the SLD. It is also found that the temperature of the mobile oil zone is higher in the SLD model compared to that in the DLD model. This implies that higher quality oil is produced when the wells are configured in the SLD pattern. Therefore, this first-of-a-kind work has shown that SLD arrangement is far more efficient, safer, and produces higher quality oil than DLD pattern and that actual process engineering designs should use SLD wells configuration.