Rethink Overall Vibration Monitoring

Clinging to a single approach that made economic sense for your plant ‘back in the day’ could be an expensive strategy.

By Trent Phillips, CMRP, CRL, Novelis

Overall values are the most common measurements and calculations used in vibration analysis. What’s more, some reliability and maintenance programs rely solely on them. The goal is to remove monitored equipment from service once the overall vibration level exceeds a certain threshold. Although this approach would appear to be quite cost effective, in reality it frequently isn’t. In fact, overall vibration monitoring can become extremely costly for a facility.

The dilemma

High vibration levels can be caused by internal and external sources. They include, among others, imbalance, misalignment, belt defects, mechanical looseness, bearing-related issues, gear defects, and cavitation. Once identified, they should all be corrected. Keep in mind, however, that equipment often experiences multiple defects at once. For example, it’s possible for the amplitudes of certain frequencies to increase while the amplitudes of other frequencies decrease. The fact that these situations indicate a variety of specific conditions poses a problem for those relying solely on the overall-vibration approach.

An overall vibration value is centered on the frequency range being acquired and calculated based on a formula selected by the manufacturer of the vibration-monitoring device. Expressed in a mathematical representation of the energy exhibited by all defects combined, plus the vibration currently experienced in the machine, the overall vibration value cannot accurately differentiate among defects caused by various machine conditions.

The solutions

What should you do once an overall vibration level exceeds your target amplitude and the equipment is removed from service?

First, stop with the assumptions. They’re often made about the causes of high overall values, and work is completed based on them. Relying solely on overall values and only making assumptions about their cause can easily lead to incomplete information about the health of your equipment. This, in turn, can lead to misguided equipment repairs or detection of problems only in the painfully late stages of failure. As a result, extra resources and efforts are invested in determining the true source of elevated vibration levels—which translates as misspent time, unnecessary equipment maintenance, increased costs, and unwanted downtime.

There are several actions you can take to ensure your vibration program is effective, i.e., that it correctly identifies conditional changes in the equipment and sources of vibration.

• Make sure the most important equipment in the facility is monitored. Don’t arbitrarily assign monitoring intervals.
• Confirm that monitoring intervals allow enough time to identify, plan, schedule, and correct the identified findings before unwanted equipment failures occur.
• Verify that recommendations are implemented. Knowingly ignoring conditional changes in equipment health will result in downtime, extra cost, and lower capacity.

Be sure you understand the failure modes in each machine, based on principles of FMEA (failure-mode-effects analysis). Band alarms and analysis should be used to indicate changes in the condition of your equipment and, at the same time, identify their causes or sources. Specific bands can be easily created, measured, and trended around particular failure modes in equipment, including misalignment, imbalance, and bearings. This information leads to more accurate alerts of impending failure conditions than generic overall measurements—and, as an extra benefit, actually identifies the failing component.

Personnel considerations

Who should collect routine vibration data? This is an important issue given the fact that wasted time wastes dollars.

Operators and mechanics should be up to the task. Both can acquire comprehensive vibration measurements on equipment during the course of their normal work activities. They also can make sure machines are shut down if vibration levels exceed acceptable values and notify others regarding the need for corrective actions. This approach allows analysts to focus on collected data and determining root causes of defects.

Data considerations

What other valuable condition-monitoring data might be missing? Tracking process information such as temperatures, pressures, lubrication levels, and equipment speed is vital for achieving desired performance from plant equipment. It also represents one of the most overlooked opportunities within a reliability and maintenance program. 

Unfortunately, since most overall-vibration-measurement devices can’t log or process such information, many facilities are unable to apply proper analytics to it. A good vibration-data collector will be able to record and store these data, and routes can be created for personnel to guide them through its routine acquisition. The collected data can then be easily stored to meet the documentation requirements of your facility and trended to provide increased analysis capabilities that may otherwise go overlooked. Alarms can be automatically generated when certain measurements or observations are recorded.

Corrective action

How do you motivate others to take corrective actions? This is one of the biggest challenges in any condition-monitoring program.

The information that’s presented must be very concise and plainly show what action is required. It’s almost impossible to do this with overall vibration values. Although these values may hint that machinery conditions have changed, they won’t provide clear evidence of what has changed. As a result, precise conclusions can’t be formed.

In contrast, a comprehensive approach to vibration monitoring—with detailed collection and analysis of data—can provide a highly accurate indication of what’s wrong and what corrective action is required. Calling for more than simple overall data measurements, this type of approach is always the most effective method for identifying unwanted machinery conditions and determining specific component failures. MT

As global leader for reliability at Atlanta-based Novelis, Trent Phillips is responsible for training, coaching, auditing, and developing reliability programs. Contact: trent.phillips@novelis.adityabirla.com.

“Are You A Psychologist, A Condition-Monitoring Analyst, or Both?”

“KISSing Is Good for Reliability”

“Condition Monitoring: 10 Common Management Mistakes”