Objective The operation of underground gas storage is a systematic project. Its major concerns involve economic benefits, technical risks, work efficiency, and many other aspects. The gas injection-production optimization of gas storage involves many factors and is required in the entire process of the gas storage operation. However, previous studies mainly focus on single-factor optimization and few efforts were made to constrained multi-objective optimization.

Methods Based on the multi-objective optimization algorithm NSGA-Ⅱ model, a multi-objective optimization function was established, and multiple stratum-wellbore-surface constraint variables were set. In addition, based on actual production data from 11 wells in Block 2 of Wen 23 Gas Storage, various constraints such as sand flow, coning flow, erosion flow, liquid carrying flow, surface equipment limited flow, and export pressure were set to carry out multi-objective optimization simulation. The multi-objective optimization prediction model was used to simulate and optimize the conventional gas production plan and the emergency peak shaving plan. Two hundred times of iterations were performed to calculate the optimal flow range. The data of singlewell gas production under different constraints were compared.

Results In six conventional gas production plans, the stratum production ratios varied greatly, and various constraints directly affected the production control of gas recovery wells of the gas storage. The singlewell gas production at the average stratum pressure of 34 MPa, when compared with that at 36 MPa, changed in the range of -16.7% to 22.5%. Under emergency peak shaving conditions, when the surface constraints are removed, the results calculated by the model had better convergence.

Conclusion The multi-objective optimization prediction model for gas production of gas storage aims to provide a decisionmaking basis for the multi-objective optimization of multi-cycle gas storage and to ensure the efficient operation of gas storage. However, this optimization model mainly considers technical factors instead of economic cost analysis, so further studies are suggested to, based on the actual use costs, consider more factors in their pratice of optimizing the gas production of gas storage.