Unlocking Business Value by Introducing Process Cameras

Objective/Scope This paper describes the initiative to unlock business value through technology deployment i.e. by introducing process cameras and AI models for real-time image processing. Gas compressors require dry gas to operate safely. If liquids or solids enter the compressor, they can cause severe damage, resulting in $ millions in repair costs, production loss, or even catastrophic safety incidents. Traditional sensors often declare gas as ‘dry’ based on calculated hydrocarbon dewpoint, but field experience with process cameras reveals this doesn't always reflect reality. Process Cameras ensure feed gas quality by detecting liquid contamination early, allowing proactive measures to prevent compressor damage and improve Reliability, Availability and Maintainability (RAM) in gas processing facilities. Using data generated by the process camera, correlating it with process parameters in a Machine Learning (ML) model, bad actors, bottlenecks and opportunities for improvement can be highlighted to improve compliance with an integrity management program in alignment with ASME B31.8S to assist operators in identifying and mitigating fluid-phase threats. Maintenance action, changes in procedure or implementation of a process improvement plan also benefit from evidence-based data from process cameras. It also enables comparison of feed quality between two processing trains and detects upstream performance deterioration by integrating camera technology with conventional instrumentation. The focus was given to a gas processing platform responsible for 60% of offshore production. The facility has two turbine-driven compressors with two stages each, scrubbers at multiple points, first and second stage coolers, separate gas dehydration systems with coalescer filters and absorbers, and a shared TEG unit. Historically, one compressor bundle fails frequently, leading to million-dollar losses. By installing process cameras, crews can act before contamination occurs.