Condition-based Monitoring Doesn't Work If You're Dabbling

John Schultz
Tags: maintenance and reliability, condition monitoring

The first time it happened was back in 1995. I was halfway through a presentation at a Chemical Manufacturers Association conference when a guy in the back of the room shouted, “What a load of bull!”

Not surprisingly, the room got quiet.

The gentleman was reacting to a slide which showed that the best companies in the chemical industry generate 55 percent of their workflow from condition-based monitoring (CBM) activities.

So, I stopped and said, “I appreciate your feedback. Would you care to share your experience?”

He explained that he was doing vibration analysis, infrared thermography, ultrasound, oil analysis, motor testing and non-destructive testing at his plant, but stated that CBM was only generating 5 percent of the workflow, not 55 percent. In his view, I was telling some tall tales to these people.

After thinking for a moment, I asked how many rotating machine trains he had in his plant. He said around 1,000.

“How many are on routine vibration monitoring?” I asked. He said 60.

That’s when it hit me. “So what I’m hearing is that you are applying the technologies to 6 percent of your asset base and they are generating 5 percent of your workflow. Does it make sense that if you applied the technologies to 60 percent of your asset base that they could generate 50 percent of your workflow?”

He just sat there, speechless. It was like he had never thought of it that way and had missed the benchmark coverage model discussions about what best practice plants do. Even though that was more than 10 years ago, I still see the same mistake today: People expect big results from CBM by just dabbling in it.

Another common example of dabbling is when companies apply only one or two technologies in their plants.

Let’s go back to the central concept behind condition monitoring.

Virtually every piece of equipment gives off early warning signals (heat, vibration or sound) before it ultimately fails. These warning signals, otherwise known as failure modes, can be detected with certain CBM technologies. But, here’s the catch. Industrial equipment has more than one failure mode. That means you often need multiple CBM technologies to detect them.

Consider the example of a chiller, where you could apply vibration, oil analysis, refrigerant analysis, on-line and off-line motor testing, and ultrasonic leak detection and still miss the early signs of a pending failure of the tube bundle. For that, you also need eddy current testing. In fact, depending on the material of construction and what it’s pumping, you could literally apply a dozen technologies to a chiller.

Therefore, the foundation of a successful CBM program is simple: Determine all the probable failure modes for your equipment and apply each of the technologies that will detect them.

Why do companies expect magical results from dabbling in CBM? Because doing more than that means they would have to make a commitment. That means risk. They would rather experiment first and then expand the program if they see the return on investment.

The problem with dabbling is that it simply doesn’t work. You won’t be able to detect the majority of problems that occur, so the equipment will run to failure anyway. As a result, everyone gets frustrated with the program. Eventually, the conclusion is “CBM doesn’t work here.”

The bottom line is, there is little to no payback from using one or two technologies – or applying CBM to a small amount of your assets – and hoping it will evolve into a successful program. Sure, by applying any technology to a targeted set of equipment, you can do “feel good” cost-avoidance calculations, but you will not impact your bottom line profitability with this approach.

The payback occurs when you integrate the full range of technologies across a high percentage of your asset base. That’s the only approach that lets you plan and schedule the majority of your work, get the crafts people out there with the right skills, get the right parts out there at the right time, make precision repairs and minimize the impact on your overall equipment effectiveness (OEE).

If you go the opposite direction, you’ll always have an evolutionary program. And, these types of programs never deliver best practice results. Best practice results must be engineered.

John Schultz earned the Certified Maintenance and Reliability Professional designation from the Society for Maintenance & Reliability Professionals and is a partner with Allied Reliability. Schultz and Allied Reliability are recognized leaders in applying condition-based maintenance to achieve the proactive maintenance model. Read more of his articles online at www.alliedreliability.com. For more information on this topic, e-mail info@alliedreliability.com or call 918-382-9400.