
🛢️ Oil & Gas · Data Flow
How a gas compressor station predicts failures: vibration and process data, an ML model, and maintenance work orders before a trip happens.
Drawing diagram…
Data flow for compressor predictive maintenance: vibration monitoring (Bently Nevada), bearing temperatures and process data (suction/discharge pressure, flow) from the compressor station go through the control system to the historian. An anomaly detection model compares live data with a learned normal baseline and estimates time to failure, raising early warnings to the reliability engineer, who creates a planned work order in SAP PM; outcomes are fed back to tune the model.
flowchart LR VIB[Vibration monitoring] -->|Overall and spectra| DCS[Control System] TMP[Bearing temperatures] --> DCS PRC[Suction and discharge pressure, flow] --> DCS DCS --> HIS[(Historian)] HIS --> AD[Anomaly Model - normal baseline] AD -->|Deviation score, time to failure| RE[Reliability Engineer] RE -->|Planned work order| PM[SAP PM] PM -->|Findings and repairs| LBL[(Failure Records)] LBL -->|Tuning| AD
The operational network behind a pipeline SCADA system: field RTUs at pump and valve stations, the control centre, and a firewalled link to the business network.
The safety permit that controls hazardous work at a refinery or rig: request, risk assessment, isolation, gas testing, work and handback.
A digital oilfield platform: wellsite sensors and SCADA, a production data historian, and engineering applications for surveillance and optimisation.
How drilling and production data from wells is collected, quality-checked, stored and shared with engineering teams.
A simplified crude oil refinery: desalting, distillation, conversion units and blending into finished fuels.
What happens when a pipeline leak detection system spots a possible leak: confirmation, valve closure, field response and regulator notification.