Thoughts from a mine maintenance engineer

A mine maintenance engineer writes:

My topic of interest is on how to evaluate the risk, failure effects and cost implications on components and work in general, in order to achieve the best cost per hour scenario and availability of equipment. For an example, if we look at a component. It has high hours!! but is still in good condition with no signs of trouble ahead.  What would be the issues if we decided to continue to run this component?  If it does fail, what would our costs be? Would they be the same as the change-it-now scenario. What about the effects of failure. Could there be a fatality etc. involved? Or would there be no real failure effect to worry about?

Say we have a small failure (a leaking duo-cone seal on the final drive) on a component with hours near the component’s scheduled preventive overhaul time. However, we are confident that many more hours are achievable with this component.  Should we just change the duo-cone seal or change the whole component that we know has all the internals in good condition and really could still continue on for many more hours before it requires change out?  Our people tend to have the mind set to change it, but they may not be thinking about the big picture.

This is the issue I would like you to examine. People have the mind set to just change components out, without weighing risk, cost implications, and cost savings that are very possible to achieve. The thinking process based on the 7 RCM questions, it appears to me, could assist them in looking for best case scenarios by considering risk.

The idea of merging RCM thinking with day-to-day practice springs from inquisitive engineering minds. They ask, reasonably, whether overhaul intervals are cast in stone?  Specifically, would a partial repair be a “better” decision than an overhaul in a given situation?

Decision stems from policy. A policy is a generalized rule that an organization uses to govern day-to-day decisions. Without policies maintenance administration would bog down by having to rehash past experience every single time the question came up. The technician or supervisor who decided to pull the final drive, given its advanced age, was acting upon a policy. Therefore, it is incumbent on us to ask the deeper question, “Can we find a better policy?” Can we set a rebuild trigger or rule that will better cover the general case?

The current rebuild age based policy will have been set during an “initial” RCM or some other type of analysis. Or, it may have been prescribed by the OEM to be consistent with its own warranty or service policy. But can a single number, say an overhaul interval of say 10000 hours, adequately reflect the true failure behavior of a specific equipment unit in the fleet? That’s the basic problem discussed in the following slide presentation.

We don’t want our CBM p0licy to recommend an equipment for overhaul prematurely because that would be costly and would reduce the unit’s availability. On the other hand we would not want to wait too long before overhauling because that would result in too many units failing in service, thus increasing unreliability. How do we select the best moment to intervene? In other words how can we set an overhaul policy that will extract the highest level of profit from the asset over the long run? How can we find an “optimal” rebuild policy that strikes the best balance between risk and prevention cost by considering not only age but also relevant condition data? The article “Confidence in predictive maintenance” describes the Living RCM (LRCM) methodology that will enable the maintenance engineer to arrive at a general, yet optimized, policy. The article “CBM Optimization” describes the optimization methodology. And the “Achieving reliability from data” course expands upon the theory and practice of LRCM.

© 2016, Murray Wiseman. All rights reserved.