A Reservoir Dynamic Prediction Model Based on the REROSIM Method: Digital Twin and Simulation Studies Driven by Digital Transformation

Abstract Reservoir management in mature waterflood fields demands accurate dynamic modeling and agile optimization under complex heterogeneity and ever-changing conditions. This paper presents a novel Reservoir Engineering Reservoir Simulation (REROSIM) approach– a data-driven interwell connectivity model augmented as a digital twin – to predict reservoir dynamics and optimize operations in the Changqing Oilfield of China. The REROSIM methodology represents the reservoir as a network of connection units linking injectors and producers, characterized by transmissibility (conductivity) and connected volume parameters instead of traditional gridblocks. History-matched with production/ injection data via automated machine-learning calibration, the model captures time-varying well connectivity and flow paths with high speed and reasonable accuracy. REROSIM integrates seamlessly into a digital twin framework: continuous field data feeds the model for real-time performance forecasting and scenario simulation, enabling proactive management decisions.We outline the theoretical underpinnings of the REROSIM/REROSIM model, including the conversion of 3D flow fields to a 1D network, Pollock streamline-inspired path tracing, and non-gradient optimization for large-scale control of injection and production schedules. A field case study in Changqing Oilfield demonstrates model calibration to historical waterflood performance and polymer flood data.Simulation experiments show close agreement with full-physics numerical simulation and significant advantages over traditional models like capacitance–resistance models (CRM) and streamline simulations in speed and adaptability. The REROSIM digital twin was applied to optimize injection allocation and infill well placement, yielding a predicted increase in sweep efficiency from 61% to 83% and a 10% improvement in injection utilization. The model also guided multi-level profile control (water shutoff) decisions by identifying dominant channeling pathways (high Lorenz coefficients) and selecting appropriate gel blocking strategies, which improved water-oil ratio and enhanced recovery.Through comprehensive validation, we compare REROSIM against conventional methods, highlighting its strength in capturing dynamic connectivity and enabling smart reservoir management. Finally, we discuss the implications for digital transformation in oilfields–emphasizing how physics-based data-driven models like REROSIM act as enablers for real-time optimization, uncertainty reduction, and closed-loop reservoir control. This work contributes a scalable, intelligent reservoir simulation paradigm blending engineering principles with advanced analytics to meet the demands of modern digital oilfields.