Determination of a Gas Field Production Pipeline Network Through Rolling Forecast Virtual Flow Metering Based on a Surface and Subsurface Integrated Model
摘要
This study presents a novel integrated model combining surface and subsurface data with multiphase flow calculations to achieve rolling forecast virtual measurements for gas field production pipeline networks. Utilizing model splitting, first-order difference weighted moving average, Sequential Quadratic Programming (SQP) algorithm, and Kalman filter methods; we optimized flow data for enhanced accuracy and stability. The model addresses scenarios with both known and unknown reservoir Inflow Performance Relationship (IPR) curves. For unknown IPR curves, the model adapts by generating new curves using historical data or employing rolling calculations. Case studies from a gas field in China demonstrate the model’s robust stability, broad applicability, and precision, fulfilling high engineering standards and significantly improving real-time flow safety simulations and production processes. Future work will integrate gas-oil ratio and water content variations to refine phase separation measurements, enhancing the model’s versatility in diverse production scenarios.