A Real-Time Intelligent Inversion Method for Formation Pressure While Drilling Based on UKF and SA Algorithm

Rapid inversion of formation pressure after a gas kick and adjusting the bottom hole pressure to a safe range are crucial for suppressing continued gas invasion. However, conventional inversion methods that rely on shut-in pressure build-up curves lack timeliness, while existing approaches are often plagued by the non-uniqueness of solutions and inadequate robustness, which consequently restricts their reliability and precision in field applications. To address these issues, this paper proposes a real-time intelligent inversion method for formation pressure while drilling based on the Unscented Kalman Filter (UKF) and Simulated Annealing (SA) algorithm. This method utilizes UKF to integrate surface measurement data with transient multiphase flow equations, enabling real-time estimation of the gas influx rate under different bottom hole pressures and reservoir lengths. Subsequently, a formation pressure inversion model is constructed using reservoir seepage flow equations. Finally, SA algorithm is employed for dynamic self-adaptive global optimization to simultaneously invert formation pressure and permeability. To validate the model’s robustness, random errors were introduced into the outlet flow rate, standpipe pressure, and casing pressure. The results demonstrate that the formation pressure inversion errors under these conditions were 1.83%, 1.75%, and 1.77% respectively. Compared with the 1.68% error observed under noise-free conditions, these represent only marginal increases of 0.15%, 0.07%, and 0.09% respectively. This finding confirms the model’s strong robustness against measurement noise. The accuracy of the model was validated using three case wells from different blocks in the South China Sea. Comparative analysis revealed that the errors between the inverted formation pressure results and actual field data from all three wells were below 2.17%, confirming the model’s considerable accuracy. This method not only provides a novel approach for formation pressure inversion, but also offers a theoretical foundation and reliable assurance for achieving safe, efficient, and precise control of wellbore pressure.