MARCON 2021: Case study highlights

The University of Tennessee Reliability and Maintainability Center hosted the 2021 Maintenance and Reliability Conference from March 10-12. The conference was virtual this year, instead of on location in Knoxville, and offered informative opening sessions and a deep lineup of presentation in different tracks: Reliability Culture; Maintenance Strategy; Work Management; Digital Transformation; Reliability Strategy; and R&M Implementation. Plant Services attended many sessions from home, and this article will focus on three specific case studies that use data and analysis to guide maintenance practices.

Pinpoint the chronic losses

Andy Pitts of Owens Corning, presented a case study about the operation of autonomous guided vehicles (AGVs) at one of its plants. Much of the presentation was about how to identify where faults or chronic losses are occurring, and he outlined the plant’s detailed process and focused analysis.

In general, Pitts spoke of a plant maintenance/operator working partnership, as that of a doctor/patient relationship. “Maintenance should help operators learn more about the equipment and interpret what they see when they’re out there working,” Pitts said. “If you have issues with your blood pressure, you don’t go to the doctor every day to check it. They help you get the equipment and use it.”

Maintenance and operations work together, Pitts said, to define as a facility: what are our losses? In terms of maintenance, it’s easy to define losses as breakdowns. “We really try to challenge ourselves to think of the true losses, maybe not breakdowns, but using mechanics’ time more efficiently or speed losses, but really trying to be disciplined in quantifying what your losses are, so you know that you’re working on things that will pay off with a big positive impact,” Pitts said.

“We spend a lot of time thinking about what are our chronic losses,” Pitts said. A chronic loss can often just be part of the “background noise.”

“They occur frequently, and they can be easily restored but not resolved, so it keeps coming up and keeps coming up,” Pitts said. These chronic, long-term losses can also be difficult to quantify.

Owens Corning uses focused improvement analysis tools to accomplish three things: results; training or knowledge transfer; and standardization. It’s easy to define the symptoms of a problem, Pitts said, but not the true cause. And once you do the work, the long-term change comes from documenting and replicating solutions.

“You’ve got infinite problems you can go work on. How do you decide which one is the right one?” Pitts asked his attendees. He recommends a “data driven approach,” going to whatever metrics you have at the plant. From the AGV’s six months of CMMS data, considering unit cost as the KPI to influence, the plant identified one AGV that was the biggest cost to the maintenance department in time, labor, and parts.

Once you’ve identified the problem, Pitts said it’s important to carefully select your team to work on the problem. “Make sure you have representation throughout the plant. If you have a maintenance problem, get operators involved and people from different parts of the plant,” Pitts said.

Owens Corning’s biggest problem with AGV #8 was faults in the navigation scanner positioning, which stopped the AGV, sometimes causing traffic jams. “Maintenance is needed to correct the stop,” Pitts said. The AGV’s maintain position by scanners and reflectors. If it sees less than three of those, it gets a fault, he said.

After identifying a chronic issue, assembling the team, and investigating the problem, it’s time for a problem statement, which “keeps you on track,” Pitts said. The problem statement for Owens Corning case study was as follows: “AGV #8 stops while performing work at random times, at multiple locations along the AGV path. This stop occurs on all shifts and is a result of not finding a minimum number of reflectors, with a maintenance loss of $12,500 per year.”

From there, the team decided on 90% as the target improvement rate, and set up a timeline for achieving that goal. Once the team identifies how to solve the problem through the use of analysis tools and brainstorming, they finally get to implementing improvements. “You’ve done a lot of work to really narrow your focus, so you’ve got a simple action, that’s going to have a big impact,” Pitts said.

Vibration analysis for lubrication issues

Sonoco Products Company also used detailed investigation to identify an issue with press bearings on a paper machine at one of its plants. “If you’re not familiar with the paper machine, this area is very wet and humid. There are a lot of factors that can contribute to a failure,” said Baxter Stephenson of Sonoco.

Why did the company initially suspect there was an issue? Bearings were only lasting half of their expected life. “We dispatched our reliability team to go and do some investigating,” Stephenson said. Their main form of investigation was via vibration analysis. They found excessive vibration around the large rolls that hold the paper. “Based on what we were seeing, it was believed to be due to an inadequate amount of lubrication,” Stephenson said.

Vibration measurements were also taken of the rolls in storage, which were stored fairly close to the machine. “There was a lot of excessive ambient vibration going to those rolls,” Stephenson said. Additionally, the bearing housing of the rolls was on top of wooden timbers, which were accelerating and amplifying the vibration from the machine.

To confirm that a lack of lubrication was the cause of the excessive vibration, the team did more analysis, before and after lubrication application. Once grease was added to the bearings, they measured between 90 to 95% overall reduction in friction and impact. “Lubrication was definitely a driving factor in the failure,” Stephenson said, and analysis showed that the bearing operating that way was at risk for a catastrophic failure of the cage itself.

They investigated further to measure how much lubrication was delivered from a pneumatic grease gun, because they suspected pressure in the gun was causing some push back.

Implementing a new reliability program

Yevgen Kolvov of Teva, a pharmaceutical manufacturer, also presented a case study that focused on equipment lubrication with vibration analysis and ultrasound monitoring. The company introduced reliability engineering a year ago to cut down on maintenance costs and increase productivity.

Current practices for oil changes or lubricating bearings and other mechanical components, gearboxes or hydraulic units, was done annually or biannually. “We didn’t have any condition monitoring set up, so decisions were made historically based on the recommendation from the manufacturer,” Kolvov said. “The practice was very poor and we wanted to improve on that and start something more robust, and something that will give us more data to understand what’s going on with our systems.”

The reliability team wanted to understand what was the right level of lubrication in the bearing; how to know when it’s the right time to change the oil; what happens if the period between oil changes was too short or too long; what if lubrication was not really the issue? Their analysis would also consider the condition of the lubricant.

In the pharmaceutical industry, every lubricant is highly regulated and approved to be on-site before it can go into use. Some of the challenges for the maintenance department included: use of the right lubricant for the right application; safe and clean storage of the lubricant; and filtration of the lubricant upon receipt from the manufacturer.

The team identified best practices to improve lubrication storage and usage, starting with setting up a lubrication room, and a color-coded tagging procedure for different oils. Over the years, the company had accumulated nearly 40 approved lubricants by bringing in new equipment. An assessment of all the lubrication applications shortlisted six products that could cover the entire facility.

A clean chain of custody is extremely important in their industry, and Kolvov explained the process: oil gets delivered from the supplier, and filtration is required to remove any particle or water contamination; oil is sampled prior to transfer into storage; the storage tanks are also equipped with filtration units, so the oil gets another filtration during transfer; another sample is taken (pre- and post-transfer) and sent to the lab for analysis; periodic filtration is repeated while the oil is in storage to maintain the oil at the highest quality; and when it’s needed in the field, the oil goes from transfer containers to field devices using a special fitting that prevents contamination and oil spillage.

Once the oil gets applied to the asset, an oil condition monitoring program tracks lubrication levels. “We’re able to store the data from the previous analysis and that helps us spot patterns and trends and predict failures,” Kolvov said. An online platform helps make sense of the results.

Teva has also instituted an automatic lubrication program with the help of a monitoring system from UE Systems. “It is essentially a solution that allows us to have a maintenance free unit and installed on the bearings or whatever we selected for this. They are capable of operating standalone,” Kolvov said. An ultrasonic monitor connected to the bearing monitors its condition. “We can control the level of lubricant inside the auto-lubricator, and lubricate from anywhere in the world,” Kolvov said. After a period of monitoring, AI technology determines the baseline decibel sound level, and sets a threshold. “As soon as the sound level hits the lubrication level, we get an automatic notification. You can start lubricating by pushing a button,” Kolvov said.

This story originally appeared in the May 2021 issue of Plant Services. Subscribe to Plant Services here.