Brooke Hagen, OXY
With new artificial-lift and solids-management tools emerging every day, determining why some installations succeed while others fail is rarely as simple as it seems. In an effort to identify which future vibrational-energy tool installations would be successful, Oxy conducted an immersive look-back on its historical vibrational-energy tool installs.
The analysis covered 75 wells where Oxy had previously installed the vibrational-energy tool tool. Data collected on each well included downtime, production rates, fluid velocity, method-of-production (MOP) type, wellbore geometry, and foreign material type, among other variables. Wells were then classified as Good — successful installations where no chemical treatment, tubing cleanouts, or solids-related failures occurred after the install — or Bad — installations where chemical treatment was still required, tubing cleanouts occurred, or a solids-related failure took place.
Once data collection was complete, a new challenge emerged: there was no obvious answer as to why some wells succeeded while others did not. AI was deployed to analyze trends across the dataset and identify common contributing factors. The AI identified clear patterns and established operating ranges for several key variables, which were then used to build a candidate scorecard. This scorecard was validated against the historical population of vibrational-energy tool installs, and the AI was able to explain why some wells saw an oil-production uplift while others did not.
Building on these findings, the scorecard logic was deployed as a skill inside Oxy's internal AI assistant, giving engineers a fast, data-driven way to evaluate future well candidates for vibrational-energy tool installation.