In our earlier blogs we discussed the advantages of energy-based condition monitoring and how it can provide an effective and increasingly affordable way to improve reliability while decreasing maintenance and operations expenses. But what exactly is the relationship between condition monitoring, condition-based maintenance, and the long-heralded predictive maintenance?
Condition monitoring is the process of collecting measurable, quantifiable performance indicators from assets. Once analyzed, condition monitoring data trigger condition-based maintenance practices, where performance thresholds drive maintenance activities instead of traditional interval-based preventive maintenance schedules. While proactive and not time-based, condition-based maintenance is still a preventive maintenance paradigm.
Certainly, there are times where condition monitoring delivers predictive maintenance solutions; see, for example, the case study where Motors@Work’s condition monitoring alerts helped Des Moines identify and correct issues with a 1,250-horsepower finished water pump before its possibly catastrophic failure. However, the full potential of condition-based maintenance will only be realized as we begin to analyze the reams of condition-monitoring data we’re now collecting.
Today, we do not have sufficient mechanical and operational data yet to predict P-to-F time; yet, this duration affects the cost-effectiveness of preventive versus reactive (run-to-failure) maintenance. Because industry is just beginning to monitor operating data on a variety of assets, we trigger condition-based maintenance when an asset reaches a predetermined unacceptable operating level. While these thresholds originate from experts’ experiences and manufacturers recommendations, we don’t really know how these operating conditions affect P-to-F time, particularly when multiple complicating factors exist.
For example, in the late 1990s, after work by Dr. P. Pillay demonstrated the ubiquity of voltage unbalance and the effects of voltage variations (sags, swells) on motors in petro-chemical applications,[i] Austin Bonnett and Rob Boteler of Emerson Motors (now US Motors) decided to run motors to failure at varying voltage unbalances.[ii], [iii] From this work, we know that a motor operating with a 1% unbalance has half the life expectancy of the same motor operating at nominal, balanced voltages; a 2% unbalance yields one-quarter the life expectancy; 3%, one-eighth; and so on. Subsequent work shows that life expectancy changes based on whether the voltage unbalance occurs on the leading or lagging phase.[iv] But what about a motor that operates with varying levels and phases voltage unbalances? With sagged or swelled voltage and a voltage unbalance?
As we collect and analyze condition monitoring data from different assets in various applications, such as through Motors@Work’s partnership with IBM Watson, new insights about what’s normal vs. abnormal will emerge. We’ll unlock knowledge about what affects the P-to-F time of assorted assets in distinct application classes. And with that knowledge, we will unlock previously unthinkable levels of optimization.
To learn more on how Condition Monitoring can help maximize your motor life, download our white paper.
[i] P. Pillay, “Practical considerations in applying energy-efficient motors in the petro-chemical industry,” Proceedings of the IEEE Petroleum & Chemical Industry Conference (1995), Paper # PCIC-95-21.
[ii] A. Bonnet, “The impact that voltage and frequency variations have on AC induction motor performance and life in accordance with NEMA MG-1 standards,” IEEE (October 1998).
[iii] A. Bonnet & R. Boteler, “The impact that voltage variations have on AC induction motor performance,” Proceedings of the 2001 ACEEE Summer Study on Energy Efficiency in Industry (2001).
[iv] A. Von Jouanne & B. Banerjee, “Assessment of Voltage Unbalance,” IEEE Transactions on Power Delivery 16.4 (2001).