Techniques Comparison
| Technique | Detects | Warning lead time | Equipment cost | Skill required |
|---|---|---|---|---|
| Vibration analysis | Bearing defects, imbalance, misalignment, looseness, gear wear | Weeks to months | £2,000–20,000 | Medium–High |
| Infrared thermography | Electrical hot spots, overloaded connections, refractory faults, bearing overheating | Days to weeks | £1,500–15,000 | Low–Medium |
| Oil analysis | Lubricant degradation, wear particles, contamination (water, fuel) | Weeks to months | £50–200/sample (lab) or £3,000+ portable | Low (sampling); Medium (interpretation) |
| Ultrasound | Early bearing defects, air leaks, steam traps, electrical discharge (partial discharge) | Days to weeks for bearings | £500–5,000 | Low–Medium |
| Motor current analysis (MCSA) | Rotor bar cracks, eccentricity, bearing defects, broken coupling | Weeks to months | £3,000–8,000 | High |
| Process parameter monitoring | Performance degradation (efficiency, flow, pressure drop) | Immediate to weeks | Low (existing instruments) | Low |
Vibration Analysis
The most comprehensive technique for rotating machinery. Detects mechanical faults from imbalance and misalignment (appearing immediately as elevated 1× and 2× running speed) to early bearing defects (appearing as HFD/envelope spectrum changes months before failure). Most applicable to: pumps, compressors, fans, gearboxes, motors, turbines.
See our vibration analysis guide for full coverage.
Infrared Thermography
An IR camera displays temperature differences as colour maps. Electrical connection faults, loose connections, unbalanced loads, and failing components generate heat before they fail. Annual thermographic survey of LV/MV electrical panels, busbars, and motor control centres is considered best practice in most facilities.
Applications: electrical switchgear and panels, motor housings, bearing housings (trend temperature), refractory walls, heat exchangers, flat roofs, insulation gaps.
See our thermography guide for full detail on temperature rise limits and severity classification.
Oil Analysis
Lubrication oil analysis detects three things: (1) lubricant condition (viscosity, TAN/TBN, oxidation, additive depletion); (2) contamination (water, fuel, coolant, particulates); (3) wear particle analysis (material and quantity reveal which component is wearing). Sampling frequency: typically monthly or every 500–1,000 hours for critical gearboxes, compressors, and hydraulic systems.
See our oil analysis guide for interpretation and action limits.
Ultrasound
Ultrasound detectors (40 kHz) detect high-frequency sound from: (1) bearing defects in early stages — micro-impacts before they cause vibration visible in conventional analysis; (2) compressed air, gas, and steam leaks; (3) steam trap condition (passing steam in a failed-open trap). Valuable, affordable, and requires less skill than vibration analysis. Lead time for bearing detection is typically earlier than vibration analysis by weeks to months.
Building a CM Programme
Steps to implement condition monitoring:
- Asset register: List all machines to be monitored with criticality rating (A/B/C)
- Failure mode analysis: For each critical asset, identify the most likely failure modes and the CM technique that detects each mode earliest
- Technique selection: Select 2–3 techniques per asset class — e.g. pumps: vibration + oil analysis; motors: vibration + thermography
- Route planning: Create measurement routes by area; plan collection intervals by asset criticality
- Baseline collection: Collect 3–5 measurements on each machine known to be in good condition — set statistical alarm levels
- Analysis and reporting: Analyse data at regular intervals; generate exception reports for anomalies
- Work order integration: Connect CM findings to CMMS — raise predictive work orders, track completion, verify that maintenance action resolved the anomaly