Using an Optimal Digital Twin to Integrate Geoengineering Solutions and ML-Assisted Diagnostics in an Underground Gas Storage System

This paper presents a digital twin developed for an underground gas storage (UGS) system in China's gas-rich Sichuan basin to manage the country's evolving energy landscape. Integrating real-time data, simulation models, and machine learning (ML), the system manages automatic data acquisition, dynamic model updates, and real-time management. Interactive dashboards enable users to monitor operations, ensure flow assurance, predict future dynamics, and provide diagnostic and safety functions. The digital twin sets a new standard for intelligent and safe UGS operations and helps operators to achieve decarbonization. The UGS digital twin seamlessly integrates geoengineering, ML-assisted diagnostics, and online workflows across petroleum engineering and geomechanics domains. High-frequency data from digital oilfield instrumentation and user inputs via dashboards feed into structured query language (SQL) and open platform communications (OPC) based databases. Automated workflows use these data and multidisciplinary models for UGS management. Key features include: On-demand reservoir geomechanical coupled simulation, to predict geomechanical risks for production and injection cycles respectively and advise the maximum drawdown and injection pressures.Daily reservoir-to-surface network simulation, for operational planning and scheduling to proactively respond to changes in operational requirements.Daily wellbore and network simulation, to identify flow assurance issues, e.g. sanding, liquid holdup, erosion, and hydrate formation, and evaluate pipeline transportation efficiency to monitor system status, and identify operational risks.Real-time flow twin simulation, to analyze pipelines for liquid inventory, pigging, and gas leak detection and ensure the system operates at lower risk and in a safer mode.Real-time ML-assisted detection, to diagnose choke malfunctioning and wellhead hydrates for maintaining optimal efficiency and safety. This groundbreaking solution sets a new benchmark for UGS in China by integrating multidisciplinary insights into a single collaborative platform. The digital twin automates field development strategies, shifting from experience-based to model-based approaches. Predictive techniques for pipeline hydrate formation help operators anticipate and mitigate issues, reducing operational risks, dehydration costs, and downtime. Consequently, stakeholders can respond quickly to market demands, thereby enhancing overall organizational effectiveness. Valued at approximately 2.5 million USD, this project supports 50% of peak natural gas consumption in cities like Beijing and has inspired three similar ventures in the region. It aims to potentially reduce CO2 emissions by 5 million tons annually, ensuring safer, more efficient, and optimized UGS operations. The UGS digital twin leverages advanced technology, including geoengineering and ML, to revolutionize daily UGS operations. By facilitating multidisciplinary collaboration within a unified environment, it enables a smarter, more intelligent system, ensuring that UGS operations function more safely, efficiently, and optimally. Furthermore, with a strong focus on sustainability, this solution positively impacts the economy, environment, and corporate social responsibility, establishing itself as a leader in digital transformation within the clean energy sector.