Written By Tom Moriarty
Many organizations have preventive maintenance (PM) and predictive maintenance (PdM) programs that are not getting it done. They’re paying too much for what they put into their PM and PdM programs.
New plants or new systems always have deliverables in the specifications for the contractor to provide maintenance tasks. Original equipment manufacturers (OEM) almost always have a “one size fits all” maintenance program that covers the asset whether it is installed in Peru or above the Arctic Circle, regardless of redundancy or
duty cycles. OEMs almost never include PdM tasks (vibration analysis, infrared imaging, oil analysis, etc.) even when the
asset could greatly benefit from these proven technologies
As anyone who ever read an automobile owner’s manual can attest, an OEM maintenance plan often provides for protecting the asset through the warranty period. It also seems like they recommend overly intrusive maintenance so the asset owner has to pay for more parts and servicing over the life of the asset. Plants that have been in operation for a considerable period of time often have an additional problem. Over a number of years there have been upsets resulting in the operations manager or plant manager mandating more PM has to be done “to ensure this never happens again!” The problem is the new PM task probably hasn’t been developed through an appropriate process. In this case, the non-engineered PM task was likely done in haste, and also likely resulted in additional costs for labor, materials and production availability that over time dwarf the original event’s impact to the bottom line. This results in low-value or no-value tasks that, suck resources away from higher value
There are a number of choices such as classical reliability centered maintenance (RCM), various forms of streamlined RCM and a slew of other branded derivations of
RCM. But if I had limited funding, in an organization that had little or no appetite for new software purchases, my pick would be failure modes, effects and criticality analysis (FMECA), coupled with planned maintenance optimization (PMO). The first reason is that FMECA and PMO can be accomplished for plant assets using spreadsheets. Second, a good facilitator can be very efficient at leading a cross-functional team through a lot of assets in a reasonably short period of time. Third, nothing against RCM variants, but FMECA is the heart of any RCM methodology. I’m a big fan of learning to walk before you run, so I gravitate toward simple and effective. Fourth, PMO provides an opportunity to evaluate the costbenefit of the PM or PdM tasks — thus
weeding out the low-value or no-value tasks. Finally, FMECA and PMO provide an engineered approach on which future changes to maintenance tasks can be evaluated
or improved upon.
A good FMECA and PMO program will be an intuitive process that is easily taught to operators, craftsmen, supervisors, managers and engineers. It should allow rapid assignment of criticality based on weighted factors for safety, regulatory compliance, operational availability and corrective action costs. Use of a table of values for each criterion speeds the process. Probability is assigned to each failure effect, based on historical data or experience from operators and craftsmen who are closest to the assets.
When we have assigned criticality and probability these values can be combined into a risk prioritization number (RPN). Calculating the RPN allows sorting/prioritizing
to ensure the most important maintenance tasks are assigned the limited organizational resources, maximizing value. PMO provides a vehicle to evaluate the annualized labor, materials and contractor mcosts versus current tasks, or the “no task” expected outcomes. Again, there needs to be a net benefit for doing the task, or it must be required for safety or regulatory compliance; if not the task is not approved. Keep the process simple, eliminate low-value tasks and prioritize by safety, regulatory compliance, operational availability and maintenance costs.