Iptwins: visual analysis of injection-production correlations using digital twins

Abstract

During oil-gas production, appropriate water injection to different production layers can effectively maintain stratum pressure and implement sustainable extraction of petroleum resources. Studying the performance of oil displacement by water is largely significant for researching the distribution of remaining-oil and adjusting the oilfield development plan. Nevertheless, the multidimensional time-varying injection-production data and 3D spatial structures of underground injection-production networks pose special challenges for effective injection-production correlation analysis. Therefore, we propose a digital-twin-driven visualization to explore and simulate the dynamic patterns of injectors and producers. First, digital twins of underground injection-production network are constructed with static 3D geospatial scenes and dynamic injection-production data, providing users with intuitive visual exploration and flexible interaction. Then, we apply the multi-step time-varying Long Short-term Memory (LSTM) model for dynamic analysis and recommendation of injection development. Furthermore, abstract information visualizations are combined with the 3D virtual environment to support the real-time monitoring and dynamic simulation of injection-production process. Case studies based on real-world datasets and interviews with domain experts have demonstrated the effectiveness of our system for intelligent injection-production analysis.

Graphic abstract