This paper introduces a compelling argument for development of and adherence to procedure based maintenance when implementing and executing a modern program to ensure maximum capacity of a plant and reliability of its equipment. The argument is based on a new analysis of four (4) statistically significant failure profile distribution studies over the period of the last 40 years, the latest of which was completed in 2001. While all of the studies involve failure profiles in mobile platforms (two for commercial aircraft, and one each for surface warships and nuclear powered attack submarines) the conclusions that can be drawn from them apply equally to fixed facilities, commercial transportation systems and utility infrastructures of all types. It also applies to categories of equipment such as motors. The author will estimate the odds of ensuring a decline in reliability by assuming what we used to think were “truisms” about failure profiles in equipment.  Several case studies are included to emphasize how these findings cross over to manufacturing, utility, and government equipment and systems.    

Introduction

In the field of maintenance the traditional approach has been to rely upon the intuitive knowledge and skill of the crafts-persons who conduct it. There is a great deal of pride of workmanship and, in all too many organizations, a great deal of psychic income in addition to significant overtime pay for successful emergency repairs to return equipment to operation after unplanned shutdowns. There is a mystique that accompanies all of this that many skilled crafts-person would like management to believe firmly. That is that there are too many variables in maintenance, making compliance with written procedures impossible and impractical; that the “way we’ve always done it” is the best and only way to conduct maintenance. This idea spills over into preventive maintenance, also. Crafts-persons believe that their own intuitive knowledge is preferable to a written procedure and/or a thoroughly defined checklist. Aside from these problems, most organizations have allocated no resources to creation and on-going support of procedures and checklists. Accordingly these organization are beating on the wrong way of conducting maintenance in order to assure reliability. This results in at least a lost opportunity for increased profits from existing assets and at worst a fatal management omission. Management is gambling with profits and losing big time with the approach that emphasizes “pride” of workmanship over an approach that has been proven to work.  

Lost in all of this is the concept of ensuring and sustaining reliability as both corrective and preventive maintenance is performed. Ideas about how things fail that we used to rely upon as a basis for preventive maintenance have been shown in the four failure profile studies over the past 40 years to apply to only a minor percentage of failures. In gambling terms, this means that odds are very long against a “win.” From this it can be shown that  time directed maintenance, in general, also should apply to only a minor portion of the failure modes which an organization must correct or mitigate.

This is because we seldom know the profile for failure and assume that most components exhibit a “wearout” characteristic, but assign a frequency for maintenance anyway, as if they did.  Further it can be shown that intrusive, time directed maintenance can be detrimental to reliability because humans are involved and they produce “infant failures.” Non-intrusive maintenance and monitoring tasks should be sought, instead. Indeed, because of the distribution of the failure profiles described in this paper, the only logical approach for the mitigating failures in the majority of equipment is through the use of non-intrusive tasks supported by the use of procedures to assure consistent results. 

As modern predictive maintenance tools and analysis methods have come into use, most of which are non-intrusive, the requirement for procedure-based maintenance becomes even more important. Analysis of data from modern tools such as vibration monitoring, lubricant and wear particle techniques, infra red observations, motor electrical condition monitoring and almost all other technologies depends for accuracy upon knowledge of the operating state of the equipment. Operating conditions and surrounding environmental parameters must be carefully established and recorded in order that thorough analysis can be performed. This can only be established by adherence to carefully written, detailed procedures and checklists.

Such procedures may be “imbedded” into equipment designed for data collection. However, procedures for connecting data must be carefully prepared and followed in order that there is complete agreement between imbedded and non-imbedded details.

More...Click here to read the rest of the story (950K PDF)