Modern tools allow plants to predict and prevent costly failures.
by Mike Pemberton (Pumps & Systems) and William Livoti (WEG)
November 13, 2015

In today's industrial landscape, remote monitoring of rotating assets is critical for reliable operation. Specifically, vibration data from pump bearings and other monitored points allows users to analyze the data spectrum and identify problems, such as shaft misalignment and pump cavitation.

The effective use of vibration technology begins within the context of a well-managed predictive maintenance program. When not caused by improper installation, rotating machine failure is typically caused by component wear that can result from multiple sources, including a lack of proper maintenance, shaft misalignment, incorrect balance, damaging harmonic frequencies, lubricant type or misapplication.

End users, however, can mitigate these common problems by using vibration monitoring to quantitatively monitor and diagnose the root cause of failure. Condition monitoring is a powerful solution in the arsenal of effective predictive maintenance programs. While this technology and methodology is often well documented, understood and applied, today's economic climate and a growing shortage of skilled labor in North American facilities have led to a scarcity of experienced people to perform and manage these programs. As a result, many areas of preventive maintenance are neglected or suffer because of budget, manpower and training constraints.

Vibration monitoring is one part of a comprehensive predictive maintenance program, which typically includes oil sampling, performance testing, thermography and other types of well-established testing procedures. The long-term sustainability of manufacturing plants will continue to depend on this frontline defense to ensure process uptime and throughput into the foreseeable future.

One common way to capture vibration data is the use of portable handheld probes that incorporate accelerometers. These devices can also be installed on the equipment itself. The technology is easy to use, and the stainless-steel sensor housing protects the accelerometer, which operates across a wide frequency spectrum and temperature range with high accuracy and repeatability. Most plants rely on vendor-specific application programs to consolidate and analyze collected vibration data. In many cases, however, the actual vibration spectrum is not collected and only vibration levels can be established.

For those sites that collect data in an inconsistent manner, the lack of a well-structured vibration monitoring program can result in the following issues:

  • Failure to collect data because of time constraints, lack of manpower or proper equipment
  • Data is collected using inconsistent techniques or in different XYZ planes
  • Data may be misinterpreted or anomalies can be undetected
  • Failure to transfer captured data to the appropriate database
  • Failure to capture baseline data after refurbishment

To address these concerns, wireless remote monitoring has become the preferred method in many cases. Wireless technology eliminates the need for costly hardwiring for communication and also reduces sensor cable lengths. In addition, some new designs do not require sensor cables as they are independently powered by capacitor-like technology that generates micro levels of power based on the equipment vibration itself. Wireless technology has become robust and reliable and can be practically foolproof with backup systems. This equipment also provides repeatable results because the readings are taken at the exact same place on each piece of equipment. The primary benefit of capturing and analyzing vibration data is the ability to track the results and quality of equipment upgrades, rerates and redesigns.

For condition monitoring, 802.11 b/g (Wi-Fi) is the established wireless communication protocol because of its signal strength and ease of use. However, a few issues still must be addressed:

  • Demands on existing wireless devices and networks
  • Battery backup system, which also requires ongoing maintenance
  • High bandwidth required because of the large volume of data sent over the wireless link
  • Dynamic range, low noise levels, high-level processing capabilities and ability to capture data at the right time
  • Sensitivity (proper selection) and durability of the wireless sensors
  • Qualified people to analyze data; all equipment is analyzed by the same person using the same equipment
  • Single point of data entry into the predictive maintenance database

A well-designed vibration data collection and analysis program is one key part of any plant's predictive maintenance program. Monitoring provides maintenance with a report that clearly and concisely presents the results of vibration analysis.

The report provides relevant information to maintenance personnel that indicates the "as is" condition of your plant and recommends areas for process improvements. The effectiveness of the overall predictive maintenance program depends on a holistic approach, multiple skill sets and good communication.