You have a quarterly or monthly route-based vibration monitoring program but you’re still experiencing failures. The problem is that route-based monitoring only provides an instantaneous reading. For many assets, issues can develop quickly and often without warning.
A prime example of this is a bearing cage defect. A cage defect can progress from first indication to complete bearing failure in a matter of a week, or even days - and depending on the timing, route-based monitoring may not catch it. One possibility is to switch to continuous vibration monitoring. Unfortunately, implementing a full continuous vibration monitoring program can be challenging, and for some assets may be overkill. What can you do?
The solution is to apply a hybrid approach that incorporates both continuous vibration monitoring and periodic route-based monitoring. Use the right tool for the right type of asset. Critical process bottlenecks and safety-related assets such as ventilation or cooling fans are ideal for continuous monitoring (see figure 1), while low-risk or low-speed assets are best suited to route-based monitoring. This hybrid approach is an efficient way of improving your maintenance program without breaking the bank and not letting the perfect be the enemy of the good.
Figure 1: Baseline vibration spectrum of bearing (left) taken by Dynapar OnSite™ Online Condition Monitoring System shows highest amplitude at the fundamental frequency of rotation (red dot). In spectrum taken after the emergence of a bearing defect (right), the amplitude of the fundamental frequency has reduced (red dot) even as 2X and 3X harmonics have increased; higher-order harmonics correlate to bearing looseness.
If there is one truth in engineering, it’s that there is no one perfect solution, just the best approach for a given problem. Establishing a hybrid monitoring program starts with understanding the strengths and weaknesses of the two approaches.
Route-based monitoring involves taking manual readings at intervals determined by the history of the asset or extrapolated from preventative maintenance schedules. It’s a good approach to monitor high numbers of points. On the downside, the monitoring tool represents a substantial investment. Taking accurate readings requires trained technicians, whether in-house or third-party.
Data is typically stored and managed locally, and the volume increases substantially over time. Finally, analysis and interpretation of the readings requires expertise that comes at a cost, whether that involves training a staff engineer, hiring a skilled technician, or working with a third-party reliability service.
Continuous online vibration analysis uses networked sensors installed on the equipment to gather data and forward it to a local sever or cloud for analysis and alerts. The frequency of readings makes continuous online vibration monitoring very effective at detecting fast-developing defects. Although the equipment traditionally has been expensive, a new generation of budget-friendly, cloud-based condition monitoring systems is making the technology practical for a wider range of assets.
As always, there are trade-offs. Even applied at discrete intervals, continuous online condition monitoring generates vast amounts of data. In most cases, that data is ported out to the cloud for storage, however. User-friendly, cloud-based software can be used to automatically run preconfigured analytics against the data, reducing the amount of expertise needed to convert sensor output into actionable information.
The ideal monitoring approach for a given asset successfully balances the cost of monitoring with the likelihood that the frequency will discover issues in time to prevent unscheduled downtime. For purposes of choosing a monitoring modality, consider the risk introduced by failure. This includes the risk of failure occurring; the duration of that failure; and the impact of that failure in terms of human safety, environmental safety, and overall production. Finally, consider the accessibility of the assets for manual readings and for repairs.
Once you’ve tabulated these factors, match the application to the monitoring approach that is best suited to the application.
Stable assets: Assets with mean times between failure (MTBFs) of a year or more can be tracked very effectively using route-based monitoring on a monthly or quarterly basis. Investing in a system to take data continuously on this type of asset is overkill. Continuous online monitoring is unlikely to significantly increase uptime compared to route-based techniques. What it will increase is the amount of “I’m fine” data that needs to be captured, analyzed, and stored.
Low-risk assets: If an asset has a direct standby unit, spares in inventory that can be quickly installed or is unlikely to impact overall productivity of the facility and/or operation, it can be considered risk low risk. Such assets are good candidates for route-based monitoring.
Low-speed assets: Route-based monitoring is very effective for low-speed rotating assets. Most continuous monitoring systems have significant low-frequency and high-frequency roll off in the vibration spectrum. This limits their effectiveness for assets operating at speeds below about 600 RPM.
For an application, look no further than the industrial ovens used for everything from toasting breakfast cereal to curing automotive paint. The motors powering the conveyors in these ovens might run at 25 RPM while carrying heavy loads, exposing them to significant stress. Route-based monitoring provides effective monitoring for not only vibration but for other process variables to give greater insight into the overall health of the asset.
Assets that fail frequently - despite route-based monitoring: Fast-developing defects require more frequent monitoring than a route-based schedule can typically deliver. If an asset is failing every six months or less, despite regular manual readings, consider continuous online vibration monitoring. It provides better protection and a more cost-efficient solution than a route-based approach.
Critical assets: Assets that act as bottlenecks to other operations should be monitored continuously. A conveyor that acts as the chokepoint for an entire processing line, for example, should be monitored to ensure continuous operation if the cost of failure is high, even if the cost of the asset is not, it should be considered critical.
As an example, the 100,000 square-foot production floor housing the processing line for a major food producer was served by a single rooftop blower. When it failed, production had to be shut down. A simple repair might cost $120,000. A catastrophic failure could run in excess of $1 million. Given the critical nature of the asset, it should have been monitored regularly. As a result of snow and ice and other impediments, it was only checked a few times a year.
To provide advance warning of developing defects, the company’s third-party reliability service installed a continuous online vibration monitor. The unit began taking data on October 4. Around the middle of the month, the RMS vibration trend data plot for one of the bearings began to rise, indicating a bearing defect (see figure 2). After consultation, a decision was made to continue operations while monitoring the asset, in order to schedule the repair in a way that minimized the impact on production.
On November 3, the defect worsened abruptly. The system sent alerts. The team reviewed the data, monitoring system closely until the end of the shift. At that point, the issue was audible. The machine was shut down and repaired overnight. The continuous monitoring made it possible to extend production while avoiding the expense of catastrophic failure.
Figure 2: Data from the Dynapar OnSite Condition Monitoring System shows a dramatic jump in the RMS vibration trend on October 24 as the defect first formed, followed by a sharp increase November 3 when the bearing went critical.
Difficult to access assets: Continuous online vibration monitoring provides a solution for assets that are hard to access for manual readings. In a recent example, the winder roll in on a paper machine was experiencing frequent bearing failures as a result of constant speed variation. Route-based monitoring could not be used because the roll could not be accessed while the machine was running, due to safety concerns. The continuous speed variation also meant that simple vibration monitoring was not sufficient.
The company installed a continuous online vibration monitoring system with cloud-based data storage and analytics. The web-based analytics converted time waveform data into RMS waveform and vibration spectra. Using speed readings from the encoder, the automated analytics scaled the vibration data appropriately.
On January 24, the vibration levels triggered an alert. With the continuous monitoring system, the mill was able to keep the line running until February 5, when they replaced the bearing (see figure 3).
Figure 3: RMS trend of winder roll bearing vibration shows emergence of defect (inset) on January 24 (red arrow). Vibration levels returned to baseline after repair on February 5.
Reaching an asset such as a cooling tower fan or an elevator motor may require extraordinary safety precautions (and a comfort with heights). And some assets are simply a pain in the neck to monitor, perhaps because of tight spaces, because the enclosure heats up to triple digits in the summer, or because accessing the asset requires removal of protective guards.
Safety-related assets: Any assets whose failure presents the risk of immediate injury or death should be tracked by continuous online vibration monitoring. This includes assets whose failure can cause immediate harm to nearby operators and those whose failure could create a dangerous situation, for personnel and the facility as a whole.
A wood products manufacturer was suffering a high rate of failure of dust collectors used to capture sawdust in its manufacturing facility. Dust capture is safety critical – airborne dispersal of the sawdust creates a Class 1/Division I explosive environment. When the dust collectors go down, so does production. The operation needed to find an early warning system to help preempt failure.
The application is both safety critical and production critical, so it is best served by continuous online vibration monitoring. The granularity of the data not only identified issues and preempted failure, it also helped to diagnose the root cause. The monitoring system was configured to capture vibration data once an hour. Viewed over the course of a month, the RMS vibration trend showed a dip.
When the reliability tech viewed the time waveform, he noticed that the spikes caused by the impact of the bearing rollers against the races temporarily dropped for 24 to 48 hours after lubrication, before increasing again. To test the theory that the issue was insufficient lubrication, they increased the amount of lubricant dispensed every month. Failures immediately dropped and performance improved.
Troubled assets: When route-based monitoring detects a developing defect, maintenance has a choice between making the repair immediately or delaying the repair while continuing to monitor the issue. Immediate repair avoids the cost of replacing an asset that still has viable lifetime but it involves a calculated risk that the asset won’t fail between readings. Route-based monitoring become steadily less cost-efficient when frequencies increase to weekly, or even daily. Continuous online monitoring not only takes readings at intervals short enough to detect rapid changes in condition but can be configured to automatically send alerts when vibration exceeds certain thresholds.
Variable-speed assets: Certain classes of assets constantly accelerate and decelerate, which can cause premature bearing failure. The wind/unwind rolls in web processing, for example, frequently fail as a result of this operating mode.
Modern facilities can have hundreds, if not thousands of potential points of failure. No single condition monitoring approach can be effective in all cases. Route-based monitoring is inadequate to track fast-developing defects. Even the latest generation of affordable continuous online vibration monitoring systems becomes expensive for high numbers of points. The best approach is a hybrid strategy.
Audit the facility and determine which assets are good candidates for continuous online systems and which are better served by route-based monitoring. By building a hybrid monitoring program, you can streamline maintenance, increase uptime, ensure safety, and maximize productivity and profitability across the operation.
About the Author
Derek Lammel, Reliability Specialist, Dynapar Corporation. Derek Lammel, a Category III Vibration Analyst, has a wide-ranging experience in multiple pillars of predictive maintenance utilizing vibration analysis, thermography, oil analysis, and ultrasound technologies. Derek spent 9 years in the United States Navy working abroad in the aviation maintenance with Patrol Wing Reconnaissance Squadron 45. Derek joined Dynapar in 2019 and is the subject matter expert in vibration analysis for Dynapar’s product and services. Prior to joining Dynapar, Derek worked for SKF Reliability Systems as a field reliability engineer in various industries such as pulp and paper, steel producing, food and beverage, and mining developing predictive strategies to increase uptime for customers.