Construction of Virtual Model and Development Optimization Application for Low-Permeability Reservoirs in Ordos Basin

Due to the complex geological conditions and strong reservoir heterogeneity of low-permeability to ultra-low-permeability reservoirs in the Ordos Basin, traditional static models have problems such as lagging parameter updates and low efficiency in development scheme optimization. Based on digital twin technology, this study integrates multi-source monitoring data to establish a virtual model adapted to the low-permeability reservoirs in this basin. A dynamic calibration mechanism for geological parameters, namely the “data preprocessing, sensitivity analysis, machine learning inversion, iterative update” workflow, is proposed to realize real-time matching of key reservoir parameters. Combined with the practice in the Yanchang Formation block of Wuqi Oilfield, an intelligent optimization algorithm is integrated into the virtual model to complete multi-objective optimization of development schemes. The research shows that the digital twin-driven virtual reservoir model can break through the bottlenecks of traditional technologies, improve the accuracy of geological cognition of low-permeability reservoirs and the scientificity of development decisions, and provide technical support for the efficient utilization of oil and gas resources in the Ordos Basin.