Proactive Artificial Lift Replacement in Northern Oman: A Field-Scale Case Study

Abstract

Abstract The widespread use of artificial lift systems in mature fields increasingly requires a performance-driven approach due to aging equipment, insufficient reservoir support, and recurring equipment failures. Conventional artificial lift management is often reactive, relying on alarms or fixed run-life thresholds to trigger intervention, which can lead to unplanned downtime, production losses, and increased operational risk. Although previous research demonstrated potential of artificial intelligence and advanced analytics for early detection of artificial lift system failures, in nearly all cases, these applications have remained at the pilot stage or focused on a single artificial lift system. Very few studies demonstrate field-scale integration across both beam pump (BP) and electrical submersible pump (ESP) portfolios, leaving a gap that this study directly addresses. This paper presents a field-wide proactive artificial lift replacement methodology which is employed to a mature field in northern Oman. The process is founded on conventional engineering diagnostics combined with a systematic workflow of activities to establish the basis for proactive artificial lift replacement. More than 1,000 producing wells were screened, comprising approximately 60% BP and 40% ESP. Short-listed wells underwent verification testing, intake pressure trend analysis, producer–injector interference assessment, water-injection performance review, and dynamometer card interpretation. For ESP wells, expected post-replacement production was forecasted using PROSPER software. For BP wells, expected production was estimated using a theoretical equation incorporating pump and operating parameters. The integrated workflow identified four high-confidence replacement candidates. The candidates comprised two BP wells impacted by pump wear, valve leakage, and insufficient injection support and two ESP wells, one identified based on repetitive electrical trips and the other based on extended run life. Forecasted incremental oil gains ranged from 26 to 83 bbl/d per well, representing a combined forecast production opportunity of 220 bbl/d across the four candidates. One BP well was subsequently implemented, with post-replacement testing confirming an incremental oil gain of 38 bbl/d, closely matching predicted range of 38–39 bbl/d. As a secondary layer of proactive surveillance, an AI-based dashboard was developed to monitor multivariate operational data, including run-life progression, trip behavior, motor loading, intake pressure, and production stability, as part of the proactive replacement process. The replaced well was consistently identified as high risk by the AI tool, consistent with the engineering evaluation. The results demonstrate that combining proactive intervention with AI-assisted monitoring can support the transition from reactive artificial lift management toward a more proactive condition-based decision-making approach. Developed workflow provides an application tool for selecting replacement candidates, forecasting incremental production, supporting replacement timing decisions, and enabling continuous monitoring of artificial lift performance in mature fields.