- Conference Article
- 10.2118/225055-ms
Maximizing Asset Value by Utilizing a Digital Twin Integrated with Real-Time Data for a Complex Surface Network with Over 1400 Producers: A Case Study of KOC North Kuwait
- May 12, 2025
- Walid Mohamed + 4 more +4
Increasing water production in mature oil fields is a critical issue faced by many fields worldwide. Subsurface uncertainties and poor water injection strategies have led to a significant and unintended increase in water-cut. This paper discusses how a digital twin of the well and surface network model was utilized to significantly reduce water production while maintaining or increasing oil production. Custom-built workflows, employing advanced algorithms, were created to identify optimum criteria, such as adjusting artificial lift parameters, wellhead chokes, or closing low oil-producing wells. The algorithm performed several thousand iterations, considering water handling costs and current oil prices while optimizing production. A dedicated Automated Water Production Optimization (AWPO) workflow was developed to minimize water production. This workflow is part of a larger system that integrates well modeling, surface network modeling, and the calibration of chokes and surface pipeline networks. Custom-built workflows ensure the quality of well models and real-time data. The input to the AWPO workflow is a calibrated surface network model. A simulated annealing algorithm is utilized to optimize water production, running several thousand iterations to adjust artificial lift parameters, choke settings, and even close wells while maintaining current oil production rates or slightly reducing oil production, but only when the ratio of water handling cost to current crude oil price is significantly high. This ensures that any reduction in oil production is offset by substantial savings in water handling costs. The workflow honors all operational constraints, such as pumps operating within their designated envelopes, motor loads within set tolerable limits, and flowing bottomhole pressure above bubble point pressure. Current, optimized, and total costs of optimization are calculated at the back end. Many existing methodologies have mainly concentrated on targeting the subsurface domain or completion technology to reduce water production. The approach used in this paper, which involves utilizing custom-built automated workflows and algorithms with a digital twin of the well and surface network model, is a pioneering idea. In a giant onshore asset with over half a million barrels of crude oil production and over 2,000 wells, several wells will be closed at any given time due to different operational reasons. The current philosophy within the asset is to prioritize the re-opening of wells based solely on the last oil rate from the well test. This approach is not holistic as it doesn't consider the impact on the surface network model. This paper discusses the use of digital twin technology, which utilizes integrated well and surface network models along with dedicated workflows and algorithms to identify and rank wells that will have a net positive impact on the overall production in the asset. A custom-built Automated Closed Wells Re-evaluation (ACWR) workflow was created to identify and rank the closed wells with the maximum potential to significantly increase oil production. The ACWR workflow is part of a larger, integrated workflow that automates the process of well model validation, creation and updating of the surface network model, and calibration of wellhead chokes and surface pipeline connections. The workflows capture and integrate real-time, daily readings and connection data to create a true digital twin of the asset. Dedicated workflows ensure the quality of the well models and real-time data. The calibrated model is fed into the ACWR workflow. The workflow utilizes particle swarm optimization to determine the optimal number of closed wells to re-open, maximizing oil production. The algorithm runs several thousand iterations while honoring operational constraints such as water-cut, reservoir pressure, motor load, and pump upthrust and downthrust conditions. The recommendations and rankings are displayed in a dedicated user interface (UI). This paper demonstrates the importance of an integrated and automated workflow in identifying and ranking the re-opening of closed wells. It also recommends the optimum artificial lift and choke parameters for opening the wells. The methodology demonstrated in this paper is pioneering, as most of the industry doesn't utilize digital twin technology in prioritizing the re-opening of closed wells, which results in several random misleading recommendations.
Read more