Purpose. Condition monitoring should convert reliable measurements into maintenance decisions. This guide explains how to establish comparable baselines, interpret vibration trends, investigate oil-analysis changes and escalate abnormalities. All numeric examples are illustrative training data, not plant measurements or acceptance limits.
1. Define the equipment and failure mode
Begin with the asset ID, duty, criticality, rotating speed range, bearing and gearbox arrangement, lubrication method, and protection philosophy. A cement mill fan, elevator gearbox and process pump cannot be monitored with an identical diagnostic plan. List credible failure mechanisms such as imbalance, misalignment, bearing degradation, looseness, gear tooth damage, inadequate lubrication and contamination. Map each mechanism to a detectable symptom and a response owner.
| Asset / concern | Measured quantity | Useful context | Next validation |
|---|---|---|---|
| Process fan / rotor imbalance | Overall velocity, spectrum and phase where available | Running speed, damper position, flow, bearing point | Check build-up, rotor condition, looseness |
| Gearbox / gear or bearing wear | Vibration spectrum, oil particles, temperature | Load, speed, oil grade, sample point | Inspect lubrication, filtration and gear evidence |
| Motor / bearing distress | Acceleration/envelope, temperature, current | Speed, coupling condition, measurement direction | Verify repeatability and inspect lubrication |
| Hydraulic system / contamination | Particle count, moisture, differential pressure | Sample point, oil change history, system temperature | Resample and investigate ingress/filtration |
2. Collect measurements that can be compared
For each vibration route, use defined measurement points and directions, instrument settings, units and mounting methods. Record operating speed, machine load, process configuration and steady-state temperature. If a variable-speed machine is tested at a different operating point, do not simply compare the two overall amplitudes as if conditions were identical. Separate baselines by duty where necessary.
What is measured? Displacement (typically µm), velocity (typically mm/s RMS), and acceleration (typically m/s² or g) describe different aspects of motion; they are not interchangeable without signal frequency and measurement definitions. Overall velocity can support general machine-condition assessment, while acceleration and frequency-selective analysis can be more informative for some high-frequency faults. A single overall measurement cannot identify a fault with certainty.
ISO 20816 describes general approaches to machine-vibration evaluation, including vibration magnitude and its change. Applicable equipment-specific parts, manufacturer's guidance and instrument configuration must be selected before using any standard-based acceptance boundary. This page intentionally does not reproduce generic ISO severity thresholds as universal limits.
3. Worked vibration trend: detect change, not just an alarm crossing
Consider a hypothetical fan bearing with overall vibration velocity readings under comparable operating conditions. The readings below illustrate why a trend can deserve inspection even when no universal alarm threshold is provided.
| Week | Velocity (mm/s RMS) | Difference from baseline | Operator note |
|---|---|---|---|
| 1 (baseline) | 2.0 | 0% | Normal duty |
| 2 | 2.2 | +10% | Normal duty |
| 3 | 2.8 | +40% | Check repeatability |
| 4 | 3.6 | +80% | Investigate change promptly |
Worked calculation: Change from baseline = (3.6 − 2.0) ÷ 2.0 × 100 = 80%. This is a trend statistic, not a diagnosis or a safety limit. Confirm speed/load comparability, sensor placement and measurement settings before deciding whether to inspect alignment, looseness, impeller build-up, bearings or another mechanism. Use site alarm criteria and machine protection requirements independently.
4. Interactive vibration change calculator
Enter two comparable measurements in the same unit (for example mm/s RMS). The calculator shows absolute and percentage change; it does not classify machine safety or predict remaining useful life.
Interpretation requires consistent measuring points, sensor mounting, speed, load, units and instrument settings. A rising value is a reason to verify and investigate, not an automatic shutdown command.
5. Oil analysis: use a combination of results
An oil report should document the equipment, lubricant specification, sampling location, sample method and oil service hours. For a gearbox, review viscosity at the stated test temperature, water content, particle count, wear metals, oxidation and additive trends as appropriate. Compare with the correct fresh-oil reference, previous samples and laboratory methods.
| Indicator | Previous sample | Current sample | Investigative question |
|---|---|---|---|
| Water, ppm | 120 | 460 | Ingress, condensation, cooler leak or sampling problem? |
| Iron, ppm | 18 | 48 | Wear change, contamination or sampling variation? |
| Viscosity at 40°C, mm²/s | 214 | 198 | Grade consistency, dilution, shear, test uncertainty? |
| Particle cleanliness | Baseline recorded | Worse than baseline | Filter performance, ingress or maintenance contamination? |
These numbers are not generic acceptable/unacceptable thresholds. A sudden increase in water and wear indicators warrants confirming sample integrity, inspecting breathers and seals, reviewing filtration and cooler integrity if present, and arranging further laboratory or equipment checks according to risk. Never substitute one abnormal oil reading for engineering evidence of a specific failure.
6. Symptom-to-check troubleshooting table
| Observed pattern | Possible explanations (not diagnosis) | First verification |
|---|---|---|
| Strong speed-related vibration peak | Imbalance, eccentricity, resonance or process excitation | Confirm actual speed and frequency; inspect build-up and looseness |
| Elevated axial vibration | Misalignment, thrust loading or mechanical looseness | Verify bearing points, coupling condition, operating load |
| High-frequency bearing-related components | Bearing distress, lubrication deficiency or external noise | Repeat measurement, review lubrication, validate spectral evidence |
| Temperature increases with little vibration change | Lubrication, overload, cooling or sensor issue | Check load, oil system, cooling flow and temperature sensor |
| Moisture and particles increase in oil | Ingress, cooler fault, filtration or sampling contamination | Take confirmatory sample; inspect seals, breathers and cooler if fitted |
7. Investigation and escalation workflow
- Detect: Record abnormal measurement and time, location, units and duty.
- Validate: Recheck instrument, sensor mounting and repeatability under comparable conditions.
- Assess: Compare with site alerts, machine-specific limits, historical baseline and other indicators.
- Escalate: Notify the accountable engineer/operations supervisor with suspected mechanisms and uncertainty.
- Act: Follow approved risk-based inspections, operating restrictions, protection systems and work permits.
- Close the loop: Document root cause, corrective action, post-maintenance readings and changes to the monitoring plan.
8. Monitoring route readiness checklist
Tick the checks used on your route. The progress indicator is stored only on this page while it is open.
0 of 10 checks completed (0%).
9. References and further learning
- ISO 20816-1:2026 — Mechanical vibration: general measurement and evaluation guidelines (consult the full purchased standard and applicable equipment part for normative requirements).
- SKF — Introduction Guide to Vibration Monitoring, especially comparable operating conditions and trends.
Safety: This educational guide is not an engineering certification, universal vibration acceptance table, instruction to bypass protection or authorization to operate unsafe equipment. Qualified personnel must apply OEM instructions, relevant standards, site-specific risk assessments and statutory requirements.