| Gas-condensate reservoirs are thermodynamically complex and require sound reservoir engineering solutions. The main challenge is to recover more condensate because it has a higher market value than that of produced gas. Therefore, in this study we have revisited different gas injection practices to improve the liquid yield and to suggest an optimal gas injection scheme for a thick, low-permeability carbonate reservoir, which also has a compo- sitional variation with depth. The reservoir model is a 1500 feet thick carbonate reservoir with 5 md permeability in horizontal directions. The vertical permeability of the reservoir is 1 md. The reservoir fluid properties vary with depth, so the liquid-rich zone with 225 STB/MMscf condensate-gas ratio is located at the bottom of the reservoir. Constant volume depletion simulation studies indicated that the maximum liquid dropout of 47 percent occurred near the gas-oil contact. Therefore, a slight pressure reduction below the dewpoint pressure can immediately cause condensation of the liquid. To evaluate the effectiveness of gas cycling in gas-condensate reservoirs, a compositional model was built. We used compositional gradient calculations to generate the reservoir fluid model. This ensured our model matched the measured PVT data, which indicated a composition variation trend with depth. Since the reservoir fluid had non-hydrocarbon gases, such as carbon dioxide, nitrogen, and hydrogen sulfide, this made it possible to re-inject them into the reservoir and analyze their effect on the condensate recovery in comparison with the dry gas. Results showed that the 50 mol-% CH4/ 50 mol-% N2 gas injection initiated early in the reservoir production lead to higher recovery, 63 percent, which was 36 percent incremental recovery in comparison with the pressure depletion scenario, 27 percent. On the other hand, the lean gas, which in this study was comprised of 95 mol-% CH4 and 5 mol-% C2H 6 yielded almost the same cumulative condensate recovery, only 0.5 percent less than that of the 50 mol-% CH4/ 50 mol-% N2 gas mixture. These two alternative options were further studied in gas cycling operations conducted below the dewpoint pressure. The results of gas injection scenarios operated below the dewpoint pressure demonstrated that the lean gas yielded more condensate, 44.7 percent of condensate initially in-place, than the 50 mol-% CH4/ 50 mol-% N2 gas mixture, 43.7 percent, but this difference was only one percent. Again, the lean gas and the CH4-N2 gas mixture proved their technical feasibility in such reservoirs. An overall condensate recovery during the gas injection below the dewpoint pressure was significantly less than the gas injection initiated in the early stage of production. For instance, the lean gas recovered only 44.7 percent of condensate initially in-place when the gas injection started after 10 years of production. The cumulative recovery for the same lean gas reached 63 percent if the gas injection started in the beginning of reservoir development. Additional injection cases studied in this thesis were the pure gases, such as methane, nitrogen, and carbon dioxide, non-acid gases containing hydrocarbon gases and nitrogen only and sour gas injection with some amount of CO2 and H2S. The results of these gases or gas mixtures were close to the lean and CH4-N2 gases. This study further mentioned advantages and disadvantages of using particular gases in the thick, low permeable gas- condensate reservoir case, but to apply these results in the real field scale, thorough economic analysis must be performed in the future and can only provide a better choice for an injection gas. From the thesis results, we would, however, recommend injecting the 50 mol-% CH4/ 50 mol-% N2 gas mixture into the upper part of the reservoir, starting as early as possible to achieve the greatest condensate recovery. |
