Do your assets burn out or fade away?

Nowlan & Heap discussed this very question in Reliability Centered Maintenance nearly 40 years ago. In John Moubray's book on RCM2, he described how the maintenance & reliability profession's understanding of failure profiles shaped the way maintenance was conducted from the industrial revolution to the present day. Consequently, numerous articles and websites on RCM and condition monitoring reference these figures frequently. If the past is any indication of the future, I suspect that as we enter the next age of maintenance and reliability, our understanding of these failure profiles will continue to evolve.

“At that time its initial reliability was poor, the conditional probability of failure was high, and this probability increased rapidly with age. However, the increase was linear and showed no identifiable wearout zone. Within a few months the reliability of this engine was substantially improved by design modifications directed at the dominant failure modes. The initial high failure rate brought the unmodified engines into the shop very frequently, which facilitated fairly rapid incorporation of the modified parts. Consequently the conditional probability of failure continued to drop, and ultimately the reliability of this engine showed no relationship to operating age.” (Nowlan & Heap, 1978)

This excerpt from Nowlan & Heap’s book highlights not only the evolution of a maintenance program during age exploration, it also highlights the evolution of equipment’s age-reliability characteristics when subjected to structured continuous improvement or defect elimination programs (which occurred over a 7 year period from 1964 to 1971).

In a more recent study, “U.S. Navy Analysis of Submarine Maintenance Data and the Development of Age and Reliability Profiles” by Timothy Allen (SUBMEPP Department of the Navy), SUBMEPP published their results and summarized some of the previous age-reliability profiles studies (reproduced below). The results of that report were in line with previous studies in the sense that the majority of assets or systems did not show increased failures with age.  Despite this common feature, it is evident that the individual reliability profiles have a wide amount of variation.

A recent survey posted on Weibull Analysis aimed at probing our collective understanding of the predominance of the different reliability profiles and this noted variation in the studies. The following shows the results from Question #3, related to the Wear-Out profile (Pattern B). Some of the answers offered figures presented in the widely discussed figures from Nowlan & Heap (i.e. UAL 1968); where they found 2% of the components matched a wear-out pattern; where they also found 11% of the components displayed a wear-out characteristic (bathtub, wear-out, and fatigue).

The correct answer can be determined by reviewing the results from the various studies in the last 30-40 years and reproduced in the table above. Consequently, it is not a fixed number at all and can vary quite a bit from industry to industry, plant to plant or decade to decade.

With these results in mind, where do you think your plant stacks up? How do you think these vary in your industry? Does your plant keep a Weibull database of your own assets to answer these types of questions?