Three-dimensional core reconstruction and performance evaluation of CO2 displacement in a tight oil reservoir

Abstract

In this study, a three-dimensional tight sandstone core was reconstructed digitally using serials thin sections cut from a core sample used in laboratory carbon dioxide (CO2) flooding experiments. Based on the digital core, multiphase flow simulations were carried out to evaluate the performance of three prior laboratory-tested injection strategies: CO2 continuous injection, water alternative gas (WAG: CO2) injection, and cyclic CO2 injection. The chosen image-based reconstruction led to a close porosity and permeability match with the sampled core in an economical way. The numerical evaluation closely replicated the laboratory-tested procedures, thus providing realistic fluid distributions before CO2 injection. To determine saturation functions for CO2 injection after water flooding, sensitivities of endpoint relative permeability and capillary pressure to CO2 continuous injection were quantified. Water relative permeability was identified as a key factor. The saturation functions were used in a history match between simulations and laboratory data where close agreements were observed for all three CO2 injection strategies. Then critical enhanced oil recovery (EOR) operational parameters (soaking time, slug ratio, cycle size, and test duration) were optimized to maximize CO2 displacement performance, which led to the following observations: (1) While extended soaking after CO2 continuous injection increased incremental oil recovery, such increase was limited due to concurrent soaking during low-rate CO2 injection. (2) In WAG injection, CO2 and brine slug sizes and ratios were optimized: increasing the cycle size performed better than adjusting the slug ratio for favorable mobility control and efficient CO2 EOR. (3) For cyclic CO2 injection, the total CO2 injection volume was the most important factor, followed by the number of cycles, to allow CO2 diffusion into tight core to maximize the swept pore volumes. The history-matched simulations suggest that both WAG and cyclic CO2 injection can achieve high displacement effectiveness for the tight reservoir, compared with that achieved with slight soaking after CO2 continuous injection. The injected brine slugs in the WAG injection reduced the CO2 needed at the expense of reduced storage capacity. Compared to WAG, cyclic CO2 injection, though realizing the highest oil recovery and gas storage, requires more CO2.