Investigation of Water Invasion Mechanisms and Dynamic Prediction of Horizontal Wells in Heterogeneous Bottom-Water Gas Reservoirs

Bottom-water gas reservoirs typically experience rapid water influx and a short water-free gas production period. Currently, variable-density perforation in horizontal wells serves as the primary water control technique for their development. However, under heterogeneous reservoir conditions, there remains a lack of robust technical support for the quantitative design of such perforations, accurate prediction of water cresting breakthrough points, and optimization of production strategies. To address this gap, physical simulation studies of water invasion investigating coupled flow in heterogeneous reservoirs and variable-density perforated wells in typical bottom-water gas reservoir horizontal wells in the East China Sea, a quantitative relationship between heterogeneous reservoir characteristics and variable-density perforation was established. This relationship enabled the optimization of segmented perforation density and the balancing of gas production profiles. Additionally, large-scale three-dimensional physical simulations further clarified the production performance and water control mechanism of variable-density perforation in horizontal wells of bottom-water gas reservoirs. Based on seepage mechanism analysis, combined with the method of images and superposition principle, a segmented water breakthrough prediction model for horizontal wells was developed that comprehensively accounts for reservoir heterogeneity, heel–toe effects, and trajectory fluctuations. This model predicts dynamic migration characteristics of nonuniform water invasion profiles, and the research results have been successfully applied to the HC gas field to verify the reliability of the methodology. Field application results demonstrate that: (1) the established design chart for variable-density perforation in heterogeneous reservoirs can effectively balance the bottom-water coning front profile, extend the water-free production period, and increase gas recovery by more than 3%; and (2) the water invasion prediction model for horizontal wells in heterogeneous bottom-water gas reservoirs can accurately identify multiple water-producing points along the wellbore, with a breakthrough time prediction error of only 6%. Notably, in the absence of dynamic monitoring data during reservoir development, the model can realize real-time tracking of gas–water front profiles and early warning of water influx solely based on production data. This enables timely optimization of production systems and the implementation of intervention measures, thereby significantly enhancing the development efficiency of bottom-water gas reservoirs.