Reliability Centered Maintenance (RCM) is method for determining the most appropriate failure and consequence management strategies for your physical assets in your current operating context. The first four questions in the RCM method, as defined in standard, SAE JA-1011, “Evaluation Criteria for Reliability Centered Maintenance (RCM) Processes,” utilize the time proven engineering method, Failure Modes and Effects Analysis (FMEA).
Using FMEA, RCM digs down to the level of failure modes and their causes, often identifying failure mechanisms that few know little about. Indeed Engineering knowledge is beneficial when performing RCM analyses, but if you don’t have engineers, you can still do it.
The knowledge needed to perform RCM effectively covers the asset, its design and its construction, how it is used and what it is expected to deliver for your business. The application of the asset, once chosen, is largely in the hands of operators who may or may not have technical backgrounds. They need to understand the process in which the asset is used, its role in that process, and how to get the asset to perform that role. The asset itself and how it works (on the inside) may, in some cases, be relatively unknown to the operators.
Most of us reading this article will know how to drive a car safely. We need to know some basic functionality and how to use it correctly, but we don’t need to know how it works. For example, we need to know how to accelerate and stop, steer, signal and park. Yet we don’t need to know how the brakes work, how to accelerator works, how the steering works or how the signals work. We may do some very basic maintenance like cleaning, checking tire wear, checking fluid levels, but many of us leave the rest of the work to our mechanics, trusting them to take care of those systems for us and to tell us if anything needs extra attention.
Plant and mobile equipment operators don’t really need to know how the machinery works, but they do need to know how to use it to get the systems they operate to perform. That implies a needed degree of knowledge of how their operational systems are used to produce their outputs. Maintainers know how the machinery and systems work in more detail, but may not always know how they are used in a system to produce an output. For example, your maintainers in a refinery may know how the pumps work, but not how the fractionation process, which utilizes numerous pumps actually works.
RCM requires both sets of knowledge. The first question in RCM is about functions – something that only the operators can truly answer. When we look at failure modes and what to do about them, we need the maintainer’s knowledge. Describing effects of failure modes requires both. Sometimes the failure cannot be prevented nor predicted. We then get out of the realm of the maintainer and may get into the realm of the designer. The timing of proactive tasks can be set to minimize or eliminate operational disruption, but we need operational knowledge to set that timing properly.
The main bodies of knowledge and experience need to come from operators and maintainers. Even for newly designed systems with little or no operating history, operators and maintainers of similar systems will be invaluable in the analysis. Engineers, are very helpful with their operational and asset specific knowledge, their ability to help with the math when determining task frequencies and their ability to dig into details that may elude even the most diligent of maintainers. Indeed, they do add value.
But do you need engineers for RCM? The answer is “no but”. The main contributors are operators and maintainers. Engineers add value but without them, the analyses can still be very thorough and rigorous. For certain, it is far better than not doing RCM at all.
Should you trust your RCM work to engineers alone? No. Unless your engineer has both operational and maintenance experience with your systems and equipment they will come up short on the practical insights needed in doing RCM properly. Occasionally RCM will be left in the hands of engineers to perform. In my opinion, that is a mistake, particularly if using more junior engineers. Their lack of practical experience will show up as either superficial analysis (not enough depth due to a lack of deep experience) or excessively detailed analysis (same reason coupled with an attempt to compensate by digging too deep). In my experience, any RCM analysis performed by only one person (regardless of background) will likely be flawed.
Can you use multiple engineers? Yes, and make sure they have complimentary knowledge. For example a maintenance engineer coupled with a process engineer could do a very good analysis, but they would still benefit from the hands on experience of maintainers and operators. You can’t get around it – that field experience is the most valuable contributor in an RCM analysis.
Can you have too many engineers involved? Yes. I’ve seen this happen where engineers that had similar expertise were used. Perhaps personalities played a role here, but they had a bit of an intellectual rivalry. In their bids to out-do each other, they came up with increasingly rare, yet plausible failure modes, almost all of which resulted in Run-to-failure decisions. In that case we did have experienced maintainers and operators in the analysis team, they could see where it was headed long before we finished identifying all the exotic failure modes, but we were powerless to stop them. They were both respected engineers who were considered experts and no one dared to second guess them. The analysis took a few days longer than necessary to complete, was academically brilliant, but little of value was added. I’m always careful with my RCM team selection making sure we have a balance of knowledge and expertise at the table and little to no duplication or overlap.