In the first of its kind, Process Vision has launched PhaseVu AI, a new AI-driven diagnostics solution that helps gas operators detect liquid carryover earlier, understand root causes, and move from reactive monitoring to predictive optimization.
Free liquids and mist escaping from separators and filters can contribute to measurement errors, compressor damage, process disruption, and off-specification gas. These conditions can be difficult to diagnose because conventional instruments generally measure individual process variables or analyze an extracted and filtered gas sample without directly showing the liquids present in the flowing pipeline.
PhaseVu AI, an intelligent diagnostics platform, is designed to improve separator and filter performance by combining real-time process data with direct visual evidence from inside gas pipelines.
It addresses one of the industry’s most persistent challenges: identifying and understanding liquid carryover and mist breakthrough before they create downstream safety, financial, and compliance issues. PhaseVu AI correlates live process variables—including flow, pressure, temperature, differential pressure, and level measurements—with LineVu visual data to create a digital twin of actual normal separation behavior, effectively building a process-specific digital twin from real-world data.
“Operators can have extensive process data and still lack direct evidence of what is moving through the pipeline,” said Paul Stockwell, Managing Director at Process Vision. “PhaseVu AI connects those two sources of information. It shows engineers what happened inside the line and identifies the operating conditions most likely to have driven it.”
Key Capabilities
PhaseVu AI delivers five core capabilities designed to improve operational awareness and decision-making:
- Event Diagnostics correlates observed liquid and mist events with changing process conditions to identify likely contributing variables.
- Dynamic Alarm Thresholds that replace fixed setpoints with adaptive thresholds based on actual operating conditions.
- Early Warning that identifies variables driving liquid carryover risk and supports proactive maintenance and process tuning.
- Visual Proof through time-stamped video evidence aligned with process data for reporting, root cause analysis, compliance verification, and dispute resolution.
- Continuous Learning that enables the model to adapt as operating conditions evolve.
Delivering Measurable Value
PhaseVu AI helps operators improve safety by minimizing the risk of compressor failure, reducing nuisance alarms, optimizing pigging and disposal costs, minimizing unplanned shutdowns, and strengthening evidence for contractual and regulatory gas quality compliance.
The platform’s model-based approach provides a significant step beyond traditional alarm management. Rather than relying solely on fixed thresholds, PhaseVu AI continuously compares actual visual conditions with predicted behavior, identifying deviations that can indicate developing problems before they become critical.
From Observation to Predictive Control
PhaseVu AI represents the next stage in process monitoring, transforming visual observations into a predictive operational tool. By integrating real-time visual evidence with live plant data, operators gain earlier warning of contamination risks, deeper understanding of process drivers, and greater confidence in operational decisions.
Process Vision is now offering short-term studies to establish baseline operating behavior and deploy smart alarms tailored to each facility’s unique operating envelope. PhaseVu AI can subsequently support permanent LineVu installations and continuing process-performance monitoring. For example, in one study with PhaseVu AI it was established that liquid level in the separator was one of the drivers of increased liquid carryover. Lowering the liquid level improved separator performance.
