Underground Gas Storage Digital Twin for Depleted Reservoirs Based on Mechanistic Models and Digital Technologies

Abstract Digital technology is helping improve safety, economics and efficiency while managing the whole-life cycle of underground gas storage (UGS), which is the most important method of natural gas regulation during peak consumption periods. Based on PetroChina's digital construction practice, this study presents an intelligent solution package for Huangcaoxia (HCX), located in east of Sichuan basin, China. The intelligent solution combines decades of expertise in mechanistic model development, rooted in theoretical principles of the petroleum industry with cutting-edge digital technologies. It seamlessly integrates elements such as reservoir geology, geomechanics, reservoir simulation, production operations, processing, and advanced machine learning techniques. Additionally, the solution also enables seamless cross-domain collaboration including technical with digital. Interactive dashboards enable users to monitor, manipulate, and predict the performance, ensuring intelligent management of UGS. When managing a UGS facility, activities can be classified into two categories, planning and operational. Planning usually refers to a relative long-term action, e.g., annual production plan while operational refers to behaviors that are currently undergoing, e.g., well choke opening adjustment operation, compressor conditions operational optimization etc. Both planning and operational activities play an important role in overall performance of UGS, the former facilitates accurate estimation for the upcoming period, while the latter ensures that the current status of the facility is safe and economical. The solution comprises of smart reservoir, smart well, smart network and smart process components, playing a crucial role in the abovementioned activity categories. Each component includes specialized features to help realize the ultimate potential of an intelligent UGS. Some features deal with the planning aspect, such as creating an annual production/injection strategy, judging the UGS production/injection scheme based on machine learning geomechanics simulation, providing early warnings for specific injection plan, performing operational scheduling to predict the following cycle's performance. Other features focus on the operational aspect, such as facility capability evaluation, choke hydrate risk identification that helps diagnose wellhead hydrates using real-time sensor data, transient production system that envisions how the fluid flows along the pipeline with real-time high frequency data being fed into the simulation engine, or dehydration optimization that enables optimization of the surface processing facilities. The intelligent UGS integrates operational experience, multidisciplinary insights, digital technology, as well as field onsite management needs into a single collaborative platform. The combination of the planning and operational features helps achieve intelligent long-term and real-time operational management, ensuring safer, more efficient, and optimized UGS operations.