Industrial Vibration Monitoring for Predictive Maintenance

Program design before instrument selection

Most failed condition monitoring programs fail at the design stage, not the instrument stage. Hardware that detects an emerging fault is irrelevant if the alarm reaches no one, or if it reaches everyone and nobody knows whose decision it is. A successful program defines four things up front:

• Which assets are monitored, and why
• What baseline “healthy” looks like for each asset class
• What an alarm means, and who responds within what window
• How insights feed maintenance planning, not just alarm logs

Asset prioritization and route planning

Continuous monitoring is justified for assets where unplanned failure would impose unacceptable downtime, safety risk, or replacement cost. The rest belong on a periodic route, walked monthly, quarterly, or annually with a portable analyzer.

Prioritization is a one-time exercise that drives sensor count, system selection, and labor budget. Revisit it annually as plant configuration and product mix change.

Sensor selection and mounting for permanent monitoring

For permanent monitoring of rotating machinery, IEPE accelerometers in the 100 mV/g range mounted directly on the bearing housing are the workhorse. Premium MEMS units cover similar applications with the addition of low-frequency response useful for slow-speed rotors. For deliberate-vibration equipment (vibratory conveyors, screens), the Electro-Sensors VS-2 / VS-2XP series and VUM-800 are designed specifically to confirm normal vibration rather than detect anomalies.

Mounting matters more than sensor selection

• Stud mount achieves the catalog frequency response; magnetic and adhesive cut it sharply
• Mount as close to the bearing as the housing allows, distance from the source attenuates higher-frequency content first
• Avoid mounting on covers and panels that resonate independently from the structure
• Route cables to prevent fretting and triboelectric noise from movement
• Use locking connectors and strain relief in environments with continuous vibration

Thresholds, alarms, and actionable decisions

Threshold setting is where most condition monitoring programs lose credibility. Too tight and the system cries wolf until everyone ignores it. Too loose and the failure surfaces as an unplanned outage anyway.

1. Establish a baseline

   Capture vibration signatures of healthy equipment over enough operating cycles to characterize normal variation. Two weeks of data on a stable load is a typical minimum.

2. Anchor to a recognized standard

   ISO 10816 / ISO 20816 provides vibration severity classes for rotating machinery by power and mounting type. Use these as the starting point, then refine with site-specific baselines.

3. Use multi-level alarm logic

   A “warning” level triggers route inspection on the next shift; an “alert” level triggers planned intervention; a “trip” level triggers automatic shutdown. The thresholds and the response procedures must match.

4. Combine static and trend triggers

   A 30% increase in 1× RPM amplitude over 48 hours is more meaningful than the absolute value being below a static threshold.

5. Document the response

   Each alarm must have a documented owner, response window, and follow-up procedure. Without that, alarms become noise in a Slack channel.

System architectures: permanent vs. portable

Permanent online monitoring

Rack-mount systems such as the HBK / Beran 766 Permanent provide continuous data acquisition, alarm logic, and historian integration for critical assets. They suit power generation, large compressors, primary process equipment, and similar high-criticality applications. The 768 is optimized for smaller turbines in hydroelectric and gas pumping installations.

Portable analyzers

The HBK / Beran 766 Portable and the compact 767 Portable (under 12.5 lbs) cover walk-around routes, troubleshooting, and commissioning. Portable systems suit standard-tier assets where continuous monitoring isn’t justified and provide a second-opinion capability for permanently monitored equipment when an alarm needs further investigation.

Industrial machinery protection

For deliberate-vibration equipment and basic machinery protection, the Electro-Sensors VS-1 / VS-1XP for general protection and VS-2 / VS-2XP for vibratory conveyors provide simple, robust protection without the complexity of a full predictive program. Browse the condition monitoring catalog and industrial vibration sensors catalog.

Operational workflow that survives staff turnover

• Document the asset list, sensor map, and threshold logic in a controlled location, not in someone’s notebook
• Standardize the route, the analyzer configuration, and the inspection report template
• Train at least two people on every analyzer and software platform
• Calibrate sensors and analyzers on documented intervals; keep the certificates accessible to maintenance and audit
• Review program performance quarterly: did alarms predict failures, were false alarms reduced, did missed events have detectable signatures in retrospect?

FAQ