Edge Computing Technology Implementation for Enabling Autonomous Control at SRP Wells to Reduce Energy Consumption and Optimize Oil Production in D Oilfield of Pertamina Hulu Rokan

Abstract D Oilfield is the largest steam flood, consisting of 6,700 wells utilizing Sucker Rod Pumps (SRP) as the artificial lift method, operating continuously 24 hours a day. This continuous operation results in high energy consumption, reaching up to 1 million kW per day, and low pump efficiency of approximately 48.52% in fluid lifting activities. To address these challenges, autonomous control is implemented to reduce energy consumption while enhancing production efficiency. Edge computing technology enables the Autonomous Control Logic (ACL) feature, optimizing oil production and reducing energy consumption. Edge devices are integrated with existing legacy controllers at well sites to facilitate autonomous control of sucker rod pumps. These devices leverage a private cellular LTE network to ensure high-speed and secure communication. The primary parameter controlled autonomously is idle time, which is continuously analyzed and adjusted based on high-frequency data stored in the edge devices. The implementation of edge computing technology has demonstrated its effectiveness in enabling autonomous control by optimizing idle time settings across 30 SRP wells in D Oilfield. This has resulted in a significant reduction in pump running time to 3.43 hours per well per day, leading to a decrease in energy consumption of up to 33.65 kW per well per day. Production efficiency has improved to 4.21 barrels of fluid per day per kW (BFPD/kW), representing a 59.49% increase compared to previous conditions through enhanced pump effectiveness. The reduction in manual optimization efforts by approximately 3,000 minutes and the decrease in travel distance for optimization tasks by 15 kilometers have further improved operational efficiency and safety. Additionally, the operations team has experienced an increase in confidence, reaching 80%, in performing problem identification and analysis using edge computing data. This improvement has the potential to generate an estimated USD 2 million in annual value if deployed across 200 SRP wells in D Oilfield.