A Thorough Coreflood Study of the Effect of Gas Viscosity on the Performance of Gas and WAG Injections under Near-Miscible Displacement Conditions in a Weakly Water-Wet Sandstone Rock

Abstract

Abstract A Water-alternating-gas (WAG) injection is a broadly practised technique in oil fields. Gas viscosity is a significant parameter that can affect the efficiency of gas and WAG injections. By conducting the current coreflood experiments at reservoir conditions, we aimed to investigate the effect of gas viscosity on gas and WAG injection performance in terms of oil recovery and differential pressure. Both WAG injection experiments were performed on the same Clashach sandstone core, under weakly water-wet and near miscible (gas/oil IFT = 0.04 mN.m-1) conditions, using two different hydrocarbon systems (C1-nC4 and C1-nC10). To eliminate the impact of the experimental artifact, a long and large core (2ft x 2 in) was employed. In addition, after each initial water injection, water was pumped through the core at multi-rates, for further investigation of the impact of capillary end effects on our experimental results. To facilitate the interpretation of the data and the comparison, the same injection strategy and methodology were followed in both coreflood experiments. In each injection scenario, four water slugs, starting with primary water flooding, were injected in an alternating manner with four gas cycles. The results of these WAG experiments showed that the cyclic oil recovery performance during different water and gas injection cycles increased as the number of WAG slugs increased. Investigating the effect of gas viscosity on the performance of oil recovery during gas and WAG injections revealed higher oil recovery performance during the tertiary (three-phase displacement) water injection cycles that were subsequent to the preliminary water flood periods, in WAG injection with C1-nC4 than that in C1-nC10. In contrast, the efficiency of oil recovery during the successive gas injection cycles (under three-phase conditions) was lower in C1-nC4 than that in C1-nC10. The ultimate oil recovery achieved by WAG injection under weakly water-wet and near miscible conditions reached 93 % and 94.5 % (IOIP %) in C1-nC4 and C1-nC10 respectively. On the other hand, the results showed also an extra oil quantity of 3.7 % (Sor%) recovered during the alternation of water and gas injections post-waterflood, by C1-nC10 compared with that in C1-nC4. Studying the impact of the gas viscosity on the injectivity showed a significant drop in the periodic gas injectivity, during different gas injection cycles in WAG injection for C1-nC10 compared with its values for C1-nC4. A comprehensive series of data sets, generated for two WAG injection experiments with different hydrocarbon fluids (C1-nC4 and C1-nC10) will be reported in this paper. WAG injection is a special case that involves complex multi-phase and multi-physics processes, which are well-known to be difficult to reliably predict by the current existing reservoir simulators. Therefore, representative and reliable experimental data are needed to improve our understanding of the complex underlying mechanisms of oil recovery by WAG injection and to develop improved models and methodologies for reliable predictions of the performance of WAG injection under reservoir conditions.