Rotating equipment failures cost industrial facilities millions in lost production every year. A single unplanned shutdown at a mining operation can cost hundreds of thousands of dollars per day. Pump failures at water treatment plants can disrupt service to entire communities. In power generation, an unexpected equipment failure can create consequences across the grid.

Condition monitoring detects developing equipment problems before catastrophic failures occur. It measures physical parameters – vibration, temperature, oil quality, and others – to identify bearing wear, misalignment, rotor imbalance, and other mechanical faults while machines continue running. The technology transforms maintenance from reactive firefighting into planned, proactive intervention.

This guide explains what condition monitoring is, the technologies it relies on, how it predicts failures through measurable physical changes, and how to structure a program that delivers real reliability improvements in an industrial facility.

What Condition Monitoring Actually Does

Condition monitoring uses sensors and analysis tools to track the health of rotating equipment over time. Rather than waiting for failure to signal a problem, monitoring programs measure physical parameters that change gradually as mechanical components deteriorate.

Think of it as the industrial equivalent of regular health checks. A general practitioner measures blood pressure, cholesterol, and other indicators to identify developing health problems before they become serious conditions. Condition monitoring does the same for pumps, motors, fans, compressors, and gearboxes – identifying signs of deterioration early enough that maintenance can be planned and executed without crisis.

The monitoring process measures specific parameters that change predictably as particular fault types develop. Vibration increases as bearings wear. Temperature rises as lubrication fails. Oil contamination grows as components degrade. Each fault type leaves a measurable signature in one or more monitored parameters.

The P-F Curve – How Failures Develop Predictably

Understanding why condition monitoring works requires understanding how failures develop. Components do not fail randomly – they deteriorate through predictable stages. The P-F curve is the standard way of illustrating this.

Point P on the curve represents the moment a potential failure first becomes detectable through monitoring. At this stage, the developing fault produces a measurable change in vibration, temperature, or oil composition – but the equipment is still functioning normally. Point F represents functional failure – when the equipment stops performing its intended function.

The interval between P and F is the condition monitoring window. This is the time available to detect the fault, assess its severity, order replacement parts, and plan the repair into a scheduled maintenance window. The longer this interval, the more flexibility maintenance teams have to respond efficiently.

Different monitoring technologies detect faults at different points along the P-F curve. Oil analysis often detects wear particle generation before any vibration signature is measurable. Vibration analysis typically identifies bearing defects six to twelve months before functional failure. Thermography and temperature monitoring generally provide shorter warning windows but are sensitive to fault types that vibration monitoring misses.

Deploying multiple technologies extends the detection window and improves the reliability of the diagnosis by cross-referencing findings from different measurement methods.

The Key Monitoring Technologies

Vibration Analysis

Vibration analysis forms the cornerstone of most condition monitoring equipment programs. Accelerometers mounted on bearing housings convert mechanical vibration into electrical signals. These signals are processed into frequency spectra that reveal which vibration frequencies are present and how strong they are.

Different fault types generate vibration at characteristic frequencies. Worn bearings produce vibration at frequencies calculated from bearing geometry. Misalignment generates elevated energy at twice running speed. Rotor imbalance appears at exactly running speed. FFT (Fast Fourier Transform) processing separates these overlapping signals so each fault’s contribution is visible and measurable.

Thermography and Temperature Monitoring

Thermography uses infrared cameras to detect temperature distribution across equipment surfaces. Hot spots in motor windings, bearing housings, and electrical cabinets indicate developing problems – often before vibration symptoms become measurable.

Bearing temperature monitoring provides a simpler but effective complement to vibration analysis. A bearing temperature that rises progressively above its normal baseline, while operating conditions remain constant, indicates developing wear or lubrication failure.

Oil Analysis

Oil analysis examines the lubricant from gearboxes and other lubricated machinery. Spectrographic analysis identifies metal particles suspended in the oil, revealing which specific components are generating wear debris. Viscosity measurement, acid number, and particle count data indicate lubricant condition and contamination level.

Oil analysis is particularly effective for gearboxes and large slow-speed bearings where direct vibration measurement is less sensitive. It often detects developing problems before any vibration signature is measurable.

Ultrasonic Monitoring

Ultrasonic monitoring detects high-frequency sounds produced by bearing defects, steam trap failures, electrical arcing, and gas leakage. These sounds occur above the range of human hearing but are readily detected by specialised instruments.

Ultrasonic detection of bearing lubrication failure often provides earlier warning than vibration analysis. It is also the primary technology for detecting compressed air leaks and steam trap failures – fault types invisible to vibration monitoring.

Motor Current Signature Analysis

Motor current signature analysis examines the electrical current waveform drawn by a motor to detect both mechanical and electrical faults. Broken rotor bars, stator faults, and load-related problems all produce characteristic patterns in the current signal. This non-intrusive technique requires no sensor installation on the equipment itself.

Online vs Offline Monitoring Systems

Industrial facilities deploy condition monitoring in two main configurations, each suited to different operational requirements and equipment criticality levels.

Online condition monitoring systems use permanently installed sensors that continuously track equipment health. Data flows to centralised analysis software that automatically alerts maintenance teams when parameters exceed preset thresholds. For critical assets where unexpected failure causes immediate production loss, continuous surveillance is the appropriate approach. These systems capture transient events and intermittent fault signatures that periodic measurements would miss.

Offline monitoring relies on portable instruments and scheduled inspection routes. Technicians carry handheld vibration analysers through the facility on a predetermined path, collecting measurements from each machine on the route. This approach is cost-effective for monitoring large equipment populations where not every machine justifies permanent instrumentation.

Online condition monitoring is appropriate for the highest-consequence assets – equipment whose failure would immediately halt production or create safety risks. Portable vibration analysers cover the broader equipment population efficiently and economically.

Most facilities benefit from a tiered hybrid approach: online monitoring for critical assets, route-based offline monitoring for the standard equipment population, and minimal or no monitoring for low-value, easily replaced assets. This allocation concentrates monitoring resources where failure consequences justify the investment.

Common Failure Modes Detected Through Monitoring

Condition monitoring identifies specific fault types through recognisable measurement patterns. Understanding these patterns helps maintenance teams diagnose what is happening before a machine fails.

Rolling element bearing defects generate vibration at frequencies calculated from bearing geometry and shaft speed. Outer race defects produce frequencies typically three to five times shaft speed. Inner race defects appear at higher frequencies. Ball and cage defects each have their own characteristic frequency signatures. Trending these defect frequencies over time reveals progressive deterioration before catastrophic failure.

Gear tooth wear increases vibration at gear mesh frequency – the number of teeth multiplied by shaft speed. Sidebands around the gear mesh frequency at intervals of shaft speed indicate individual tooth problems. Monitoring gear mesh frequency trends reveals deterioration before tooth breakage occurs.

Shaft misalignment produces elevated vibration at twice running speed in radial directions, with significant axial vibration confirming the diagnosis. The relative level of axial to radial vibration helps distinguish angular misalignment from parallel offset.

Rotor imbalance generates strong vibration at exactly running speed. The amplitude increases with the square of speed – doubling shaft speed quadruples imbalance-related vibration. This speed-sensitivity characteristic distinguishes imbalance from other fault types that do not respond to speed in the same way.

Mechanical looseness generates multiple harmonics of running speed – peaks at 1X, 2X, 3X, and 4X running speed appearing together. Loose bearing fits, base plates, or foundations produce this pattern. The multi-peak pattern distinguishes looseness from single-frequency faults.

Key Parameters Measured in a Condition Monitoring Program

Effective programs track multiple parameters to build a complete picture of equipment health. Single-parameter monitoring misses important fault types and reduces diagnostic confidence.

Overall vibration velocity provides a general health indicator. ISO 20816 defines severity zones for different machine types and mounting configurations. Trending overall vibration in millimetres per second RMS over time reveals deteriorating conditions without requiring detailed spectral analysis of every data point.

Vibration frequency spectra provide the diagnostic depth to identify specific fault types. Regular spectral analysis on critical equipment allows fault frequencies to be tracked individually, providing early warning of specific component degradation.

Bearing temperatures should remain stable under consistent load conditions. A progressive rise of 10-15 degrees Celsius above the established baseline signals developing problems in the bearing or its lubrication.

Lubricant condition – tracked through oil analysis – degrades through oxidation, contamination, and additive depletion. Regular sampling detects adverse trends before lubricant condition affects equipment reliability.

Implementing a Condition Monitoring Program

Structured implementation produces better results than beginning measurements without preparation.

Equipment criticality assessment determines where to focus initial efforts. Rank equipment by failure consequences across safety, production, and cost dimensions. Direct the first phase of program implementation at the top 20% of critical assets. This concentrates resources where monitoring delivers the most value and generates early successes that justify program expansion.

Baseline measurements establish what normal looks like for each machine. Collect measurements when equipment is in known good condition – after installation or following a major overhaul. These baselines become the reference for all subsequent trending and fault diagnosis.

Measurement point standardisation ensures data collected at different times by different technicians can be meaningfully compared. Mark measurement locations permanently on bearing housings. Document sensor type, orientation, and mounting method for each point. Inconsistent measurement locations introduce variability that masks genuine condition changes.

Alarm threshold setting defines the boundaries between normal, concerning, and alarming parameter values. ISO 10816 and ISO 20816 provide severity guidelines as a starting point. Site-specific thresholds based on individual machine baselines often prove more sensitive and appropriate than generic guidelines.

Aquip provides implementation support for condition monitoring programs, from criticality assessment and baseline data collection through to alarm configuration and diagnostic interpretation services.

Integrating Monitoring with Maintenance Planning

Condition monitoring data delivers its full value when it drives maintenance planning decisions. The connection between measurement data and maintenance action determines whether a monitoring program is a reliability tool or simply a data collection exercise.

Remaining useful life estimation allows maintenance to be planned based on actual equipment condition. If bearing trending data indicates continued deterioration at the current rate, maintenance planners can estimate when intervention will be needed and schedule accordingly. This is fundamentally different from replacing components on a fixed time interval regardless of their actual condition.

Work order prioritisation improves significantly when based on objective measurement data. Equipment showing rapid deterioration moves up the maintenance schedule. Machines in demonstrably good condition can wait. This risk-based prioritisation directs labour resources and spare parts budgets where they deliver the most value.

Parts inventory optimisation reduces capital tied up in spare parts. Rather than stocking bearings for every machine just in case, condition monitoring identifies which specific assets will need attention in coming weeks or months. Targeted inventory management cuts carrying costs while ensuring critical spares are available when needed.

Condition monitoring services provide expert diagnostic analysis and reporting that connects measurement data to specific maintenance recommendations – the translation step that turns data into actionable decisions.

Future Developments in Condition Monitoring

The predictive maintenance landscape is evolving rapidly. Several developments are reshaping how industrial facilities monitor equipment health.

Wireless sensor networks eliminate the installation cost associated with cable runs to each measurement point. Battery-powered sensors transmit data via mesh networks or cellular connections, making comprehensive monitoring economically viable for equipment that was previously too costly to instrument permanently.

Machine learning algorithms trained on large fault databases can automatically identify complex vibration patterns without requiring analysts to manually calculate and interpret fault frequencies. These systems improve their recognition accuracy as they process more data.

Digital twin integration combines condition monitoring data with detailed equipment models. Virtual replicas simulate how current operating conditions affect component wear rates. Operators can evaluate how different operating strategies would affect equipment longevity.

Edge computing processes sensor data locally before transmission, reducing bandwidth requirements and enabling real-time alarming even when network connectivity is intermittent. Critical alerts reach maintenance teams within seconds rather than after cloud processing delays.

Aquip monitors these developments and incorporates proven technologies into the condition monitoring solutions it supplies, ensuring facilities benefit from advances in sensor technology and analytical capability as they become commercially mature.

Conclusion

Condition monitoring transforms maintenance from reactive crisis response into proactive equipment management. It detects developing problems months before failures occur, giving maintenance teams the lead time to plan and execute repairs without production disruption.

Explore condition monitoring products to find the right technology mix for your equipment population. Review training services to build the diagnostic capability your team needs to turn monitoring data into reliable maintenance decisions.

To discuss how a condition monitoring program can be structured for your facility, contact us and a specialist will help design an approach matched to your equipment and operational requirements.