Much has been written about the potential failure to functional failure curve (P-F Curve). Yet, most condition-based or preventive maintenance (PM) tasks rarely spell out the potential failure conditions. Often, that determination is left to the skills of the operator or technician. Doing so is subjective at best, often leaving the organization to experience high rates of equipment failures – so much so that an executive jumped up during a two-day reliability training session and stated, “Jeff, the problem is that we are looking for the functional failure and not the potential failure.” Clearly, in that “ah-ha” moment, he got it. Let’s help you cross that bridge, too, if you’re not already there, to improve the identification of potential failures in your PM task writing.
From the Plant Floor is a new monthly column that explores reliability challenges faced by organizations and solutions to overcome them.
The interval from the potential failure to functional failure must be reasonably consistent to use the P-F curve. There must be enough time to do something about it as well. Ideally, this includes time to plan, get materials, and proactively schedule the restoration before reaching the functional failure point. Undoubtedly, using predictive technologies, i.e., vibration or oil analysis, allows for a higher level of diagnostics and trending, but not every asset or component lends itself to their use. In some cases, another approach may be more cost-effective too.
Assets are bought to perform a set of functions. A pump transfers something from one point to another. Assume the process requires 100 gallons of water per minute, and the pump is capable of 120 gallons per minute. A transfer rate below 100 GPM is a functional failure (F). The job of maintenance is to maintain the asset’s capacity, in this case, between 100 GPM (want) and 120 GPM (capability). When the flow drops from 120 GPM, the impeller is failing. While maximizing the useful life and managing risk, how much time do you need to plan the repair, get materials, and schedule the restoration? Assume that you decide the magic number is 104 GPM for the potential failure (P) point. The operator inspection program is updated with the threshold of 104 GPM, and on reaching it, the operator adds a work request electronically.
The challenge is to determine the asset’s numbers, i.e., capability and process requirements, and what indicates the potential failure using precision approaches compared to the subjective task steps in PMs such as “inspect the pump.” Your process may have a linear valve that clamps the flow at 100 GPM in a closed loop. If the linear valve proportionally opens to maintain that 100 GPM flow, the output to the valve can be used to determine the potential failure of the pump impeller. The valve may only be 30% open when the impeller is new. As the impeller wears, the valve opens to maintain a constant flow. Using the previous 104 GPM, maybe the valve is 60% open to achieve that flow when the impeller’s potential failure condition is reached. Again, a work request is initiated when the valve reaches the potential failure (P) point.
Many tools can identify potential failure conditions at the asset and the component levels. These tools include process instrumentation (flow, pressure, temperature), go/ no-go gauges, statistical process control (quality) charts, wear measurements, chain gauges, tensioner positions, belt spring tension gauges, sheave gauges, and other measuring tools. Use these tools at intervals (typically one-half the P-F interval) to find the potential failures.
To demonstrate using Figure 1, a cartoner uses an indexing chain carrying carton buckets that are filled with a product. Reasons for chain wear and elongation include improper lubrication and carton/ product jams. A new chain assembly exceeds $60,000, so stocking one is not an option for the company. As shown, the organization has set the potential failure (P) measurement of 72.25”. When the point is reached, a new chain is ordered, and a proactive window is scheduled for corrective maintenance once it arrives. Also, note that the process of finding the potential failure (P) is a condition-based inspection done at time intervals. Corrective maintenance does not occur until the inspection determines the need.
Approaches to determining the capabilities, process requirements (F), and the point of potential failure (P) include the piping and instrument diagrams (P&IDs), operations process information, SCADA systems, root cause analysis, and the skills of the operators and technicians. Educate the operators and technicians to provide feedback when they encounter PM task steps that lack precision. When feedback is given, ensure that the documents get timely updates to help guarantee a repeat performance. In the PM procedure, call out the likely failure modes and follow them with the task steps to identify the potential failure points. Publish a weekly metric showing the number of PM updates, and you will have a more robust PM program before you know it.
This story originally appeared in the June 2022 issue of Plant Services. Subscribe to Plant Services here.