Optimizing ESP Efficiency and Reliability Using Non-Dominated Sorting Genetic Algorithm II (NSGA-II)

Abstract Electric Submersible Pumps (ESPs) are essential components of artificial lift systems in the oil and gas industry, valued for their ability to handle high production rates and operate across diverse reservoir conditions. Despite their advantages, ESPs are prone to frequent failures due to harsh downhole environments and suboptimal performance, leading to increased downtime, maintenance costs, and reduced operational efficiency. Issues such as excessive vibration, overheating, and unstable discharge pressures contribute to premature equipment failure. While existing optimization practices focus on real-time monitoring, predictive maintenance, and control system adjustments, they often address either efficiency or reliability independently, rarely combining both into a unified framework. To address this gap, this study proposes a multi-objective optimization framework using the Non-Dominated Sorting Genetic Algorithm II (NSGA-II), aimed at simultaneously improving ESP efficiency, reliability, and energy consumption. Efficiency is defined as the ratio of hydraulic output to power input, reliability is measured through mean time between failures (MTBF), and energy consumption is quantified as energy used per cubic meter of fluid pumped. A mathematical model incorporating operational parameters, such as current, pressure, temperature, and vibration, was developed to identify Pareto-optimal solutions that balance these competing objectives. The optimization results showed significant improvements: efficiency increased from 47.3% to 79.80% (a 68.72% gain), MTBF improved from 4.26 to 4.70 years (a 10.33% reduction in failure rates), and energy consumption decreased from 0.69 kWh/m³ to 0.48 kWh/m³ (a 30.43% reduction). Visualized through Pareto fronts and comparative analyses, the findings from this study demonstrate NSGA's ability to provide actionable insights for enhancing production and field operations.