10 ways to ruin your PdM program

While Preventive Maintenance only applies to at best 11% of your components, Predictive Maintenance (PdM) applies to a much larger percentage of your equipment, and it requires a much larger investment in equipment and training.

By Doug Plunkette, Allied Reliability Group August 13, 2014

Following up on the article I wrote on "The 10 Mistakes That Ruin Preventive Maintenance", here is something that is far worse. While Preventive Maintenance only applies to at best 11% of your components, Predictive Maintenance (PdM) applies to a much larger percentage of your equipment, and it requires a much larger investment in equipment and training. Just the same, I have been in the business long enough to see more than a handful of companies have to reinvent their Predictive Maintenance program as a result of one or more of the following mistakes.

In case you didn’t notice, I used the words "reinvent their Preventive Maintenance program" because it’s very difficult to just start the program over again and say someone made a huge mistake; it sounds much nicer to say we reinvented it. We typically call it something different the second time around, some folks might call it Condition Based Maintenance, Condition Based Monitoring, or even On-Condition Maintenance. However, the truth is that if you have to start over, a few people really messed up by doing one or more of the following on my list of things that can ruin your PdM program.

1. Not Completing the PdM Tasks on Time – This might be the most common way that companies ruin their PdM program, and those who do so usually start out doing well with their PdM program. They might even have a year or more of good data along with several significant finds with one or more of the technologies. Then the company they work for adds several new pieces of equipment or takes on more work and the next thing they know, they have a huge amount of emergency and demand maintenance and the workload begins to back up. Those who have been in the business a number of years know you can get away with skipping a task or two and hope like heck that you’re lucky, but once you begin missing defects or potential failures and a major asset crashes, all of the sudden your PdM program is not looking very effective. Let it happen more than a couple of times and someone will tell you it doesn’t work and isn’t needed.

You have to complete your PdM tasks on time all the time if you want the program to be effective. Taking the chance to skip these tasks is a bad gamble!

2. Not Correcting the Defects Detected by PdM – In a similar situation, when our PdM technicians do detect defects/potential failures, we need to act on them as soon as possible. It doesn’t do any of us any good to detect a fault, write a work order to have the fault corrected, and then allow the fault to occur. It makes the program look bad and when your program looks bad, people begin to question what it costs. Now, some maintenance folks will sit back and smile and say, "We told you it was going to fail and you didn’t give us the time to fix it." Well, last I checked, the job of planning and scheduling maintenance work belongs to those who work in maintenance. If the technologies detect a fault and the equipment needs to be repaired, we need to get it done.

3. Using 1 Technology to Detect All Failures
– I attended a class on vibration analysis once where the instructor told the group he could detect steam trap faults with vibration analysis. Now I have no doubt that one could actually do this, so the better question becomes why would you do this? When there is a technology that costs far less and you can very quickly determine if the trap is working or not working properly in less than a minute, why in the world would you use a technology that costs 5 times as much and takes 10 times longer to determine if the trap is working correctly? By the way, the instructor’s answer was, "Because I can."

My answer is slightly different. I believe in using the right tool for each job. It helps things get done in a more efficient and cost effective way. The strategy of using one PdM tool to do the job of a different tool that has been proven far better at detecting as specific fault is foolish and is a waste of time and money.

4. No One Understands the Reports – This might be the most common problem I see. While lots of companies are getting involved with Predictive Maintenance, the greatest percentage of companies I work with bid their PdM work out to a contract company who specializes in providing PdM services. They get started and the next thing they know, they have piles of paper or digital reports that show dozens of their assets in alarm. The maintenance supervisor and manager open the reports and they can’t make sense of what they are looking at. Worse than this, they now have a decision to make – do they take action on what they don’t understand?

If you’re going to have a PdM program, you need to have people trained to a level that they can read and understand the reports that come from your contractor.

5. Performing Your PdM Tasks at the Wrong Interval – This problem is typically one we see with new and inexperienced companies. They want to be able to use the technologies but have yet to put the business case together that shows how valid these tools are. How the investment in the tools and training will bear fruit in reduced equipment failures, reduced equipment downtime, and lower maintenance costs. We only get these results if we perform each technology at their prescribed intervals.

I worked with a company nearly 15 years ago who lost what I would call a mature PdM program because they allowed someone to just walk in and change the intervals on all of their PdM tasks. Harold the Hero claimed he had worked for a company who only did vibration analysis on critical assets once a quarter and once every 6 months on non-critical assets. Two years later, they had pumps, motors, and gearboxes failing without detection, and months later, they said goodbye to their entire PdM group. Five years later, Harold the Hero moved on and his old company had to reengineer their "new" condition based maintenance program.

This may sound strange to some of you, but that new CBM program looks a heck of a lot like their old PdM program.

6. "Forget the Training, Just Use the Technology"
– This is kind of like the old comedy routine where someone says, "You know your PdM program is in trouble when…"

Well, you know your PdM program is in trouble when your boss says go ahead and buy that new ultrasound tool, and when you let him know it is $17K for the tool and the training, he replies, "How much for just the tool?"

Trust me, the technologies require both training and experience to work, you really do need both. If you can’t afford the training, don’t buy the tool.

7. Using PdM to Advance Your Run-to-Failure Program – Let’s just say I have worked with a few companies over the past 20 years who have this "worst practice" down to a science. With the understanding that the term "worst practice" is the opposite of best practice, worst practice companies want to use their PdM technologies to get every last second possible out of a failing asset. Now, I understand that there are times where one might want to attempt to finish a run or an order, but using the technologies to temporarily delay the inevitable is not what the technologies were designed or invented for. When your technicians inform you that they have detected a potential failure, you should immediately look for the next window to change the component as soon as possible. I was at a customer site several years ago where when I walked into the building one morning, you could hear a low speed piece of rotating equipment just squealing as soon as the door opened. On the way to the maintenance shop, I walked by one of their PdM technicians; he was carrying his vibration analysis equipment. I stopped him and asked him where he was going, and he smiled and replied, "I’m going to take readings on that bearing you can hear squealing."

Now I don’t claim to be a Level 3 Vibration Analyst, but something tells me that the bearing he was going to check was in the process of failing and needed to be replaced. And, less than two hours later, it seized and the drive shaft fractured as well. If you’re going to invest in the tools, you need to believe in them as well.

8. Using the "Inexpensive Solution" Instead of the Right Solution – Like nearly everything else in the world, you have the option to buy the real deal or a cheap solution. Eight years ago, I worked with a client on the West Coast who had their operators taking single point readings with IR pens with the belief that he had an infrared inspection program. While taking temperature readings of different equipment on a daily basis might seem like a good idea, unless someone is recording and trending the data, a single point reading is not the best option. The shame is that with just a few thousand dollars more, they could have a camera that would have given them far more. The same can be said for every technology on the market. Do yourself and your company a favor, research the options and purchase equipment that has a proven track record.

9. Taking on Too Much at the Start
– Starting a PdM program is exciting. I have to say it feels good to finally move into the next generation of maintenance where we actually have the tools available to be proactive in detecting potential failures and correcting them before the equipment actually chews itself into thousands of little pieces. This excitement can also lead to biting off far more than you can chew. If you want your new PdM program to work, please take my advice and start this with your critical assets only. Get the most critical assets under control with the top 5 technologies before you even consider looking at the balance of your plant. Break your plant equipment up into blocks starting with the most critical and then moving on as you get each list into your CMMS, on your routes, and under control.

10. The Boss Knows Nothing About Maintenance – This last problem came from a comment on "The 10 Things That Can Ruin a Preventive Maintenance Program". It’s quite common today to have an engineer move into the role of maintenance manager or supervisor, and while they might be fantastic engineers, when they transition into these roles they know little or nothing about maintenance and can make some very rash decisions. While there are lots of good books on how to manage a maintenance and reliability program, there are very few courses available that can bring someone up to speed in a short time. The reality here is that those of us who have careers in maintenance are no different from the rest of society where blame is an acceptable excuse for failure. I look at this a bit differently – in the world where we are viewed as a burden, we must use data and facts to support the decisions we make. If your company has made the transition into Predictive Maintenance, there should be a business case to support the program, making it bulletproof to changes at any position. I would even say that those who have found success in this business did so by showing the data and results of each change they made to move in the direction of building a proactive maintenance culture.

In closing, I would like to add that in my travels, I have seen six companies who have solid Predictive Maintenance programs in place. While I looked for common ground, there were only two things they all had in common:

– Each of the six companies actively used data to show how their PdM program was improving equipment reliability and saving money.

– Each of the six companies followed, and stuck to, the industry standards for how often they should perform their PdM tasks and had a 90% completion rate or better.

It’s no secret that PdM works, but you have to be committed to the process.

Doug Plucknette is a Principal, World-Wide RCM Discipline Leader at Allied Reliability Group. This content originally appeared here. Edited by Anisa Samarxhiu, Digital Project Manager, CFE Media, asamarxhiu@cfemedia.com