Real-Time Digital Twin Integration for Optimizing Surface Network in a Giant Onshore Asset: KOC North Kuwait Case Study

Abstract A calibrated, up-to-date model provides valuable insights into network bottlenecks and enables optimization of chokes and artificial lift settings to enhance overall production efficiency. This paper presents the development of a digital twin for a giant onshore oil asset, featuring a surface network model that updates daily with real-time well and network data. The model captures new well connections, status changes, and re-routing, and automatically calibrates choke settings and surface network parameters. Surface network data is extracted daily from KOC's corporate databases and loaded into the KwIDF database using ETL (Extract, Transform, Load) processes, ensuring compatibility with the simulation environment. Real-time data from wells, manifolds, and facilities is sourced from KOC's live data systems. The foundational surface network structure—comprising pipeline lengths and diameters—is manually derived from GIS maps and stored in a central repository. An integrated tool automates model building and calibration through several dedicated workflows. The AIVGAC (Automatic IPR/VLP Generation and Conversion) workflow transforms calibrated well models into IPR/VLP curves. ANMB (Automatic Network Model Builder) updates the surface network and integrates real-time data. AIVWMV validates well model quality, while ACC (Automatic Choke Calibration) and ANMC (Automatic Network Model Calibration) fine-tune choke settings and flowline connections, respectively. The HR (Hourly Run) workflow executes the model hourly, generating virtual flow rates across the network. Finally, ACEO (Automatic Choke & ESP Optimization) recommends optimal choke and artificial lift parameters. This fully automated system ensures the digital twin remains current, accurate, and capable of supporting real-time operational decisions. Results are presented per gathering center (GC) on a dedicated GC Summary page, which displays model outputs alongside input data quality indicators. The Network Heat Map visualizes active pipeline connections and provides key parameters such as phase flow rates, pipeline specifications, and real-time measurements. Alarms and alerts are automatically generated for conditions like high pressure loss index, header switching (LP to HP and vice versa), low or no flow, and discrepancies between real-time and model data. The ACEO dashboard lists wells with their recommended optimal choke settings and ESP frequencies, enabling targeted operational adjustments. Additionally, the Network KPI page evaluates the quality of real-time and well test data used in the models, ensuring transparency and reliability in decision-making. Together, these interfaces provide a comprehensive, real-time view of network performance and optimization opportunities. This paper highlights the value of a fully automated digital twin that updates, calibrates, and delivers daily production optimization recommendations. By integrating real-time data and advanced workflows, the system enables continuous, data-driven decision-making to enhance operational efficiency and maximize asset performance.