MRO Magazine

Starting to See Results

By Cliff Williams   

Human Resources Machinery and Equipment Maintenance

This article is part of an ongoing series. The introduction appeared in Machinery & Equipment MRO's February 2005 issue and the series has run in every issue since. Previous instalments are archiv...

This article is part of an ongoing series. The introduction appeared in Machinery & Equipment MRO’s February 2005 issue and the series has run in every issue since. Previous instalments are archived online at In this issue, we pick up where we left off in our April 2007 issue, as maintenance manager Bob Edwards and his team at the Plentya Paper Company agree to proceed on making changes to their preventive maintenance methods.

Monday morning rolled around and I was pleased that everyone seemed eager to hear John, our vibration consultant, and Carol, the maintenance planner, give a talk about predictive maintenance.

“I can’t wait to hear what John has to say; he talked about some really interesting stuff last week,” said Ivan, the millwright who was now very firmly planted on the bandwagon.

“Well, maybe, we can learn something new,” commented Stan, another millwright and perhaps the biggest unbeliever.


John started the meeting with a slide that asked a question: “Why do we do maintenance?”

Several replies came forth. “So the machines will run.” “Because they’re broken.” “So we can make paper and the company can make a profit.”

“Good answers,” said John. “But I’d like to suggest that the goal of maintenance is to get the longest life cycle possible for the equipment without adversely affecting the mill’s performance.

“Irrespective of which strategy you use, you just need to be sure that you are using the best one for each application. You could run all equipment until it fails completely and get the longest life that way, but it will usually cost more to repair because of secondary damage and take longer to repair — and so cause more downtime, really adversely affecting the mill.

“The trick is to know when to use which strategy. Carol has talked with you about breakdown and preventive maintenance and when you can use those methods. Today, we’re going to talk about predictive maintenance and when to use it. I guess the first thing we need to establish is: What exactly is predictive maintenance?”

“Vibration analysis!” shouted Ivan.

“That’s a tool we use.”

“Ultrasonics!” It was Ted (another millwright) this time, remembering what John had done with the problem on the steam system at the dryer section.

“Again, a tool.”

“Let’s think about the car examples we used in our meeting on breakdown and preventive maintenance,” suggested Carol. “When do you change the tires on the car?”

“When they’re flat!” “When they’re bald!” “When the tread gets low!”

“OK, OK!” interrupted Carol. “If you wait until they get flat or bald then you are taking risks that could end up costing you more — time and money — and even worse they could cause an accident.

“If you change them when the tread is below the recommended levels, then you can arrange for an appointment at the garage so you don’t waste time, you can get an oil change done at the same time, you can make sure it is convenient for you and you can be confident that they’re not going to cause an accident. So you can see which is the preferred choice.”

“Let’s look at what we are actually doing,” said John.

“We’re taking the condition of the tire into account when deciding to change it. Through tests the tire companies have determined that the recommended minimum tread is the point at which a failure of the tire is imminent, that it has got to the longest life cycle possible without adversely affecting the performance of the car.

“As Carol said, imagine the damage and injury that could occur if the tire failed when you were driving at 100 km/hour. Your mill is no different from the car; you want to do the maintenance when it suits you, not when you’re halfway through an important order.

“You want to be able to plan to do other work while the mill is down and you want to ensure that safety isn’t compromised. Imagine the damage if one of the big rolls suddenly seized.

“Predictive maintenance is doing maintenance based on some quantifiable and usually scientific measurement that indicates the condition of the equipment. It’s sometimes known as condition-based maintenance. As Ivan and Ted said, this measurement could be vibration or ultrasonic detection, but there are literally hundreds of other tools that can be used — it can even be something as simple as thickness that can be measured with a steel rule.

“The ultimate purpose of predictive maintenance is to allow maintenance departments to make the most informed decision possible. There have been times with some of my other customers when they have seen alarms from their vibration monitoring and decided to keep on running due to customer pressure, with the knowledge that the equipment is failing.

“They have usually set the alarms to allow for this if necessary and then they increase the monitoring so they can shut down if they absolutely have to. It’s all about making informed decisions and not just running on a wing and a prayer.”

“I don’t have to tell you guys that everything at the mill has to generate a positive cash flow or we don’t do it,” said Carol. “So imagine how happy everyone would be if we told them that they could keep on running the paper machine instead of shutting down for 12 hours to change a roll — and how happy Bob, our maintenance manager, would be if we told him we were not going to change the $40,000 bearings every year like we do now.”

“Hey, hang on Carol, let’s be real about this,” said Stan, still not ready to jump on board. “Every other mill is doing just that — that’s why we do it that way. It obviously makes sense or we wouldn’t do it — right?”

“Well not really,” replied Carol. “Remember Bob’s comments to Pete last meeting? People started the strategy of changing equipment on a time basis because they really believed things failed that way. Today we know that’s not true, that it is the stress that is applied to equipment that determines the rate at which it fails.

“Imagine if we run the paper machine without the oil lubrication; we wouldn’t expect the bearings to last a year. That’s an obvious stress. Also think about how dirty the machine gets sometimes. We could contaminate the oil a little, then the motors have to work harder to drive the machine. Sometimes the operators increase the pressure on the rolls to get rid of the water. All of these have an impact on how the bearings fail.

“Ivan, how many times have you complained that the pumps in Stock Prep are being started up with the valves closed or the seal water turned off? You could have two identical pumps, but if one was started up that way and the other started up properly, then they’re not going to fail at the same rate.”

“What Carol is describing is true,” said John. “There are other industries that have really taken predictive maintenance to a higher level and what they’ve found is that maintenance costs and downtime have both been reduced by 50%, so it does work.

“That makes sense — you’re not changing parts for the sake of it, you’re not shutting down to change them unnecessarily, and perhaps the biggest benefit is that predictive maintenance looks at the equipment while it’s running, when it’s under load — and that’s when it usually fails. How many times have you seen something fail when it’s shut down?

“It’s pretty difficult to figure out if a bearing is failing when it’s shut down, since it behaves differently under load. And when you’ve finished inspecting it, how can you be sure you haven’t contaminated or damaged it,” asked John.

“What we do in predictive maintenance is look for a measurement that indicates that the equipment is starting to fail and set a point at which we need to know about it — an alarm or trigger point that allows us to do the maintenance in an organized way. Sometimes on critical equipment that means shutting it down right away, but we’ll talk about that later.

“Then we must decide how often to monitor it, based on recommendations, experi
ence, history or statistical analysis, although there are expert systems on the market that can do all this for you.”

“Bob has agreement from our plant manager, Joe Davis, that we can carry on with predictive maintenance, so you’ll see John here more often,” Carol added. “But vibration and ultrasonics won’t be the only tools we’re going to use.

“We’ll get a quote for setting up oil analysis for the paper machine and look for someone to carry out an infrared survey. Some of the guys have worked with John on the paper machine and Ivan is going to set up a route for doing vibration monitoring on the pumps in stock prep, but for the rest, we’re going to need to get everyone involved.”

“Whoa! Where are we going to get the money?” asked Stan.

“Yes, just hold on with that enthusiasm, Carol,” I said. “You have to remember that Joe wants a business case from me first.”

“That’s right,” said John. “Even those companies that are advanced in predictive maintenance do cost justifications to prove it will benefit the whole company and not just maintenance. If it doesn’t make economic sense, they don’t continue.”

“Thanks, Carol and John, that was great,” I said as I wrapped up the meeting. “As Carol said, we’re going to need everyone’s help in making this succeed.” Next, I asked John if he could stay around to talk about the frequency of inspections on the paper machine and the pulper gearbox.

Cliff Williams is the maintenance manager at Wrigley Canada in Toronto, Ont., and a consultant with TMS Total Maintenance Solutions of Markham, Ont. He can be reached at

Key Points

* There is no cookie-cutter approach

* It must benefit the bottom line

* Involve everyone


Stories continue below

Print this page