An Integrated Approach for Scheduling and Validation of Steady-State Production Well Tests in Unconventional Wells

Steady-state production well tests are a key component of day-to-day operations, providing critical data for regulatory reporting, volume allocation, and identification of well-work opportunities. This paper presents an automated workflow deployed for scheduling and validating steady-state production well tests across more than 2,300 ExxonMobil-operated wells in the Permian Basin. A Well-Test Scheduling (WTS) workflow was developed to provide daily recommendations to field personnel for prioritizing well tests. It leverages data such as well-to-separator mappings, recent well-test history, and ongoing well events such as frac hits or shut-in events. In facilities equipped with automated routing headers, recommendations are implemented via SCADA remote setpoint control, with adjustable test and purge durations. Upon test completion, an automated Well-Test Validation (WTV) program ensures test quality using high-frequency data and a rule-based validation system. This system assesses the quality of separation, infers the effective flow period, and verifies the measured oil, water, and gas rates. Results are communicated to field operators through a mobile app that is integrated with production accounting systems. The WTS process has improved well-test prioritization, ensuring at least one well test per month and increased testing frequency on a per-well basis. The well-test validation process has enhanced the accuracy of oil, water, and gas rate measurements, aiding daily surveillance by promptly detecting issues like water carryover, sand buildup in test separators, and water-in-oil emulsions. The automated process has reduced test turnover time from 4 days to less than 2 hours, significantly saving time for field operators and engineering teams while increasing test acceptance from 65% to 75%. Finally, the WTV process has established a comprehensive dataset linking flowing pressures with phase rates, laying the groundwork for machine learning applications in surveillance and optimization.