Optimization of Injection-Production Volumes in Underground Gas Storage Based on Improved Non-Dominated Sorting Genetic Algorithm II

As critical infrastructure for seasonal natural gas peak-shaving, the operation of underground gas storage (UGS) must consider multiple factors including risk, economics, efficiency, and technology. Traditional UGS operation schemes are heavily dependent on subjective experience and lack intelligent methods to fully leverage historical data. This shortcoming leads to higher risks and increased compressor energy consumption. Taking S UGS as an example, the sensitivity factors of injection-production capacity are analyzed based on geological development and multi-cycle injection-production operation data. With injection-production rates as a decision variable and while considering safety and economic factors, objective functions and constraints are defined from the formation, wellbore, and surface. The proposed injection and production cycles are both 15 days, and the total injection and production volumes are 1200 × 104 m3 and 800 × 104 m3. An optimization model was constructed using the INSGA-Ⅱ and TOPSIS to determine the optimal gas injection-production volume allocation scheme. Compared with the initial scheme, the optimal injection-production volume allocation scheme reduces compressor energy consumption by 49.19% and 49.80% and formation pressure standard deviation by 78.88% and 77.21%, respectively. This effectively lowers injection-production energy consumption while improving safety, thereby ensuring the long-term safe and efficient operation of UGS.