MRO Today - Showa Denko Carbon: Breaking the bottleneck

Breaking the bottleneck
Showa Denko Carbon took a big-picture look at its
maintenance headaches and improved the performance
of its worst offenders by almost 50 percent

by Clair D. Urbain

The old adage, planes don’t crash because of one system failure, they crash because of two or more failures, can easily be transferred to any industrial maintenance environment. Fortunately, plant “crashes” rarely result in the loss of life, but send ripples throughout the production process.

At Showa Denko Carbon, a leading manufacturer of large-diameter electrodes used in steel making, it wasn’t just crashes that were causing problems; it was the system slowdowns that made bottleneck processes even more of a bottleneck.

“When we look at our maintenance records, you can see a saw-tooth pattern form,” says Steve Carter, maintenance superintendent at Showa Denko Carbon Inc.’s Ridgeville, S.C., plant. “As you measure machine performance, it starts off high after a repair, then gradually slips until performance becomes unacceptable and the maintenance staff is called in to get equipment back in shape.”

Carter came to Showa Denko Carbon in 1995, charged with the responsibility of figuring out why maintenance was becoming more of a fire drill than a planned process. The company achieved ISO 9002 and QS 9000 registration in 1996 and had a computerized maintenance management system (CMMS) in place. But it was still seeing more unplanned repairs that caused production bottlenecks.

“We launched the Kaizen Maintenance Improvement Project (K-MIP) in 1997 to develop a process that would permanently solve chronic equipment problems we were facing,” he says.

Electrode production
Producing what amounts to a piece of graphite from 18 to 30 inches diameter and up to 10 feet long is a challenging process. It’s an extremely hot and heavy process that requires complex machinery working in a difficult environment.

First, coke and a coal tar binder are mixed in various proportions then extruded and packed in sand in saggers, then baked in kilns up to 1,500 degrees F. This carbonizes the coal pitch binder. This process also causes the rod to be porous, so it is moved into an autoclave, which under a vacuum, removes air from the pores. Then, hot petroleum pitch is forced into the pores.

Once the rod leaves the autoclave, it is cooled then final-baked in kilns at 800 degrees C.

From that point, they are moved to horizontal furnaces for graphitization where they are laid end to end and covered with metallurgical coke. An electrical current heats them to 5,500 F, which thermally converts the amorphous carbon to graphite. The final step is machining, where they are finished to a specific diameter and a threaded head and socket are milled into the ends of the rod.

“The weight and size of these electrodes and the production process creates an exceptionally harsh environment. Anything we can do to head off equipment problems helps us stay in production,” Carter says.

Getting a handle on the bottlenecks
To fully understand its maintenance needs, Carter and his colleagues at Showa Denko embarked upon a month-long analysis of existing processes and benchmarked them.

“We used an outside consultant to help us with this process,” says Carter.

Brad Peterson of Strategic Asset Management Inc., in Burlington, Conn., helped Showa Denko complete this analysis.

“We developed a year-long plan to do this project. We pulled information from all areas of the company including production, maintenance, project and process engineering, accounting, finance, sales, purchasing and warehouse departments. With Brad’s help, we focused on each asset’s performance metrics. We met every week, and carefully examined how we would measure the things we thought we should be tracking,” says Carter. “We looked back at four years of maintenance records and sorted our work orders into planned and unplanned activities. We found that even with what we thought was an effective CMMS, up to 40 percent of our maintenance jobs were still unplanned. We want to be at 10 percent.”

A management team identified common production equipment headaches and scored each piece of equipment on a scale of zero to 10 on the following criteria: Annual maintenance cost; breakdown frequency; criticality to the overall production process and safety/ environmental impact.

In all, 33 pieces of production equipment were analyzed. After sizing up the needs, the work team selected to address the coke vacuum material handling system used in the final graphitization process as its first task. On the plant floor, these two huge material handlers are called gulpers, and were a source of continual maintenance headaches and unplanned downtime.

Carter assembled a team of colleagues who either worked with the gulpers or who worked on them. Now known plant-wide as Equipment Health Team No. 1, members include Mike Pantone, maintenance engineer; Joey Westbury, facilitator in the graphitization area; N.T. Vasuki, senior process engineer; Randy Goethe, maintenance technician; Mike Jones, area project engineer; Keith Beckman, equipment operator and Bob Whitten, vice president of marketing. Whitten’s duty: provide management support and be the outsider who can ask the basic questions that can unearth valuable information.

Digging out the data
The gulpers are two huge vacuum and hopper systems suspended on overhead cranes. Working in high ambient temperatures, the units vacuum coke up from the storage area then transport it and lay it over the electrodes in the furnaces prior to the final graphitization process. They also vacuum the remaining coke out of the ovens once the process is completed. The initial research showed the units incurred $225,000 for maintenance costs; more than half of those costs were emergency repairs.

“In our first meetings, no one really knew how the gulpers actually worked. We all worked on them, but no one ever really looked at the equipment as one complete asset,” says Pantone.

To carefully assess all causes of downtime and reduced efficiency, the team scoured maintenance records for recurring problems. Although the CMMS had been in place since the plant opened in 1983, gathering information proved challenging, says Pantone.

“We pulled work orders on both units and started classifying the jobs as planned vs. unplanned and segregating emergency (breakdown) work orders into equipment components. We found that our descriptions over time were not very clear and we often had to dig back into parts purchased for the job to identify the component that needed work. It was a time-consuming process,” he says.

“We found that we had 396 breakdowns in 1996 that accounted for 56 percent of all work requests. That averaged more than one breakdown a day. It’s hard to make sales goals with those numbers,” says Westbury.

After referencing the original equipment operation manuals, the team then “swarmed” the machine. They spent a day with it, talking with operators and observing how it worked. They timed its operation to set a performance standard and in the process, identified several quick fixes that improved operating performance by 50 percent.

Once the data was collected and the team had a much better idea of how the gulpers work, they pulled together a root cause failure analysis. This logic tree process identifies problems and breaks them into distinct causes that can be evaluated separately. The team identified eight distinct problem areas on the gulpers; each team member was assigned overall responsibility of a specific problem.

Some of the greatest discoveries were found in the CMMS records.

“We found that our preventive maintenance (PM) work orders didn’t mean a whole lot,” says Pantone. “When we pulled them all together, we found that some maintenance tasks were being completed twice as often as they should have while others were missed entirely. We are now rewriting all scheduled preventive maintenance work orders so they are more specific and complete. When we get the PMs rewritten, we want to see the breakdown rate drop to next to nothing."

The team identified several engineering changes that will further improve gulper performance.

“We have prioritized these and now it’s up to engineering to make these changes so we can implement them,” Vasuki says. “We have improved the gulpers’ performance from 12 minutes per cycle to as low as four minutes. We now set the standard operating cycle to be six minutes, and it is timed and recorded every shift.”

The final step goes on and on
The most critical component of the team’s success relies on continually monitoring gulper performance.

“If a gulper doesn’t perform to the level we have set, the operator calls us immediately so we can identify the problem and repair it,” says Carter. “It’s become part of our culture, where in the past, the performance had to fall so low that it caused a bottleneck.”

President Steve Ulmer concurs. “At first, I thought this was a one-year project. Instead, it’s a never-ending process. You never achieve perfection, just improve. We have identified key problem areas that cascade throughout the plant. It starts when a colleague can’t do his or her job because some piece of equipment malfunctions, which causes rescheduling, which causes production problems, which in turn affects our customers because they may not be able to get the product they need.”

The initial K-MIP team’s work revealed that actual wrench time by maintenance colleagues was low.

“Instead of working on equipment, they spent much of their time chasing parts," says Ulmer. " Our system did not allow us to kit the parts they need to do a job. As we rebuild our maintenance process, we will have a support structure that gives them the parts and direction to do their job well. We are finding that there are so many unseen costs from lack of equipment reliability that any money you spend to improve reliability will save you money over the long run."

(For more information on Showa Denko Carbon, view
"Team focus: This isn't another 'flavor of the month'" and
"How to build a reliability-focused re-engineering team".)

This article appeared in the April/May 1998 issue of MRO Today magazine. Copyright, 1998.