A PMO Will Generate Resources

When it comes to plant operations, personnel regularly contend with scores of issues. Through this column, Klaus M. Blache attempts to help them address those concerns by answering common questions. This month’s Q&A focuses on one of the issues raised in last month’s discussion.

Q: How do I acquire the resources needed to support a predictive/condition-based maintenance effort?

A: Start by doing a PM Optimization (PMO). That means cleansing your maintenance tasks. A PMO process is a structured methodology that makes sure your technician/trade functions are adding value. It means throwing out, changing, and adding tasks.

Use an FMEA (Failure Mode and Effect Analysis) approach to understand the frequency, severity, and detectability of failure modes and potential impact on equipment function. Meet regularly with operations, maintenance, and engineering personnel to separate the actions, facts, and general information. Then, evaluate how well the action tasks will protect the machinery and equipment functions against potential failure modes.

One way to do this involves asking several sequential questions to enable the most desirable type of maintenance. This identifies/improves ineffective maintenance tasks. Ideally, you want all maintenance tasks to be cost effective. If you can’t eliminate the task, implement the highest level maintenance type possible, starting with condition-based functions to trigger maintenance events. If that’s not possible, then implement measured factual tasks, then subjective (observed without measurement), then time-based, and finally, reactive tasks.

Condition-based maintenance is always better, because it’s less expensive, non-intrusive, less labor intensive, and uses the full life of the component. Also, if the P-F (Potential-Failure) interval is short, then condition monitoring or eliminating the failure mode is most likely required. Use your current maintenance tasks, failure history, any additional details in the computerized maintenance management system, and employee experiences. Basically, do something.

You need to be doing at least 35% predictive maintenance (finding issues) to be at best practice and 25% for top quartile. This doesn’t include corrective work, i.e., fixing what you find. That’s counted as preventive. Figure 1 shows that the difference between where you should be (target) and the North America actual top quartile is not significant. I published the targets prior to 2009, based on an earlier study. The actual data is from a 2016 study based on 140 companies (representing about 3,000 facilities).

Make sure that the final PM tasks are clear, specific, and can be performed by technicians/trades/operators who have been trained and are following their standardized work. Overall, PMO helps make your PMs more efficient and effective. It enables resource optimization, while instilling best practices. When done right, it results in improved uptime and reliability (higher mean time between failure); increases wrench time; reduces variation, which can help quality; and makes resources available to enable the correct kind of maintenance.

On a bigger scale, this process helps you understand and make better decisions regarding asset risk. Every PMO effort in which I’ve been involved has resulted in a 30% to 50% reduction in tasks, freeing up valuable resources.

There are several ways to organize the PMO effort. The most complicated and lengthy process is an RCM (Reliability Centered Maintenance) approach. The most-used method is an FMEA organized approach. Many organize PMOs by similar asset groups. If you aren’t ready for either of those, you can still gain benefit by simply corralling the subject-matter experts and talking through the benefits of each PM. Ask what is done, why it is done, and whether or not it works.

Do you need to conduct a PMO? If you are not sure, ask yourself the following:

• Have you done the same PM six or more times without any needed corrective action?

• Is there high ambiguity in your task wording? If three technicians did them, would they all be done the same way?

• Where did your PM tasks come from? Have you ever reviewed them for effectiveness?

• PM tasks cause as many problems as they fix due to lack of accessibility, modular replacement, visual controls, and training. Human error is still the root cause of more than 50% of the issues.

• You can’t complete your predictive and condition-based maintenance tasks because you are too busy doing reactive repairs.

• Things are breaking before you get to do the PMs.

Also, if you haven’t done a PMO in the past five years, you need one.

The bottom line is, if you are looking for resources to do predictive and condition-based maintenance, start with a PMO. The only cost is short-term time. The benefit is long-term time you’ll give back to your employees. EP

Based in Knoxville, Klaus M. Blache is director of the Reliability & Maintainability Center at the Univ. of Tennessee, and a research professor in the College of Engineering. Contact him at kblache@utk.edu.