The last thing a pellet mill needs is a big-ticket item such as a chip dryer or a drag chain to unexpectedly fail during the busy season, which can last from midsummer through early spring. Replacement parts may be backlogged and have to be shipped from overseas, meaning dollars are hemorrhaging from the plant in unfulfilled orders and idled capital equipment. “If we’re down for two days during the busy season, we’re losing hundreds of thousands of dollars,” says Mike Shearer, operations manager for Confluence Energy’s two Colorado-based pellet mills in Walden and Kremmling.

This is precisely why plants big and small plan a major outage at least once a year. It’s a time to take on those much-needed, larger repair items and maintenance practices that are so vital to getting the most bang for the operational buck at industrial pellet mills. More routine maintenance procedures are performed during weekly, biweekly or monthly shutdowns as well. Whether it’s a weekly maintenance outage or an annual capital shutdown, the plant needs to be idled the same way. The major difference, however, is that a capital shutdown takes months of planning. Ultimately though, both short- and long-term outages help ensure equipment longevity, plant safety and production efficiency, keeping margins as robust as possible.

Every week, Confluence Energy’s Walden and Kremmling facilities schedule routine maintenance from Friday through Sunday evening, Shearer says. As the weekly shutdown approaches, Shearer says it takes about five to six hours to empty the raw material bins. An early morning start produces empty bins by early afternoon. Shutting down the plant takes only about two hours. After all the product is dried, the operations team starts cooling down the burner and dryer systems, and once those have cooled, all the material from the hammer mills have been pelleted. The equipment is then locked out and the mechanic shift, consisting of four personnel, take over.

Charlie Daw, plant manager for Zilkha Biomass Energy’s 275,000-ton pellet mill in Selma, Alabama, says his plant conducts monthly outages for routine maintenance that last eight to 24 hours. “We’re trying to line it up where we don’t take the whole plant out, maybe we’ll just take one dryer out at a time for inspection,” he says.

Zilkha Biomass-Selma just began operations in April, so the team is still refining its maintenance schedule and procedures. For now, the entire plant goes down during its monthly maintenance and inspection outages. “We don’t like to go down with anything left in any hoppers,” Daw says. “There’s moisture and safety concerns. If we’re doing hot work, we don’t want wood chips present.” Zilkha Biomass, like Confluence Energy, shuts down the entire plant, starting in the front and working back to the pelletizers, before performing maintenance. 

At Indeck Ladysmith LLC’s 90,000-ton pellet mill in Ladysmith, Wisconsin, the safety, quality and logistics manager Darren Winchester says his team plans for a six- to eight-hour maintenance day once a week, but this may be biweekly during the busy season. He says while the plant is not manufacturing pellets during the outage, they can often package product while production is down for maintenance.

Shearer says routine weekly maintenance procedures range from changing gearbox oil and worn-out belts to replacing 2- or 3-horsepower motors. “If we think it’s going to fail, we replace it before it does,” he says. “We keep track of hours, so anything we see coming up, we take care of.” Confluence Energy initially didn’t track logged equipment hours precisely, Shearer says, making predictive maintenance nearly impossible. The team now keeps a solid list of critical wear items. “That way, you know what’s coming at you,” he says. Preventative vs Predictive The longer a facility accumulates a history of operations with accurate, dependable tracking of hours, the more it can rely on predictive versus preventative maintenance. Ray Budisavljevic took over as vice president of operations at Confluence Energy about six months ago. He says, once he came on board, he assigned unit numbers to all the equipment to track hours, and he developed a work-order system for maintenance, repair and replacement. “That’s all new here, we’re starting from scratch,” he says. “It’s important to know the life cycle of your parts.” For instance, Budisavljevic says the fan went out on the baghouse at one of Confluence Energy’s mills. “We knew we bought it two years ago and installed it two months after that, so we got 22 months out of the fan,” he says. “The manufacturer said that was about right, so we plugged that information into our system and scheduled it for replacement 20 months later.”

Confluence Energy utilizes software called Maintenance Pro to keep track of equipment hours and repairs. “The program issues work orders and scheduled maintenance orders, and also keeps financial records of what each piece costs us to maintain so we can make the right call as to when the equipment has reached its lifespan, before dumping more money into it,” Shearer says.

There are several different philosophies on maintenance, and Budisavljevic says which one is followed depends on how much control is desired. “To me, preventative maintenance has always been oil changes, getting your greasing done and ongoing maintenance to get the expected life out of a part or machine,” he says. “Predictive maintenance is life cycle, being ready to replace the part before it goes out on you. A lot of places run to failure, and they have the part waiting and deal with the unintended consequences. Others don’t run to failure, so they’ll run it to 95 percent and replace it regardless if it’s needed because they don’t want unexpected downtime.”

Knowing equipment costs per hour is essential. Does it cost more to schedule replacement of a worn or near-end-of-life part before it fails, or is it more expensive to squeeze every ounce of life out of a part and change it out when it does finally go? When considering operation of high-energy-consuming equipment such as dryers, for instance, replacing degraded items before they fail is a major cost-saver. Thompson Dryers says when seals and airlock blades are worn or ducting is damaged, air leaks into the drying system. Increased air flow into the system will increase the energy used to evaporate a pound of water. Regular maintenance of these components prevents air leaks, which can cost hundreds of thousands of dollars a year in most operations. The company says a reliable drying system will operate 8,400 hours per year or more, but proper maintenance is critical. The dryer drum, trunnion wheels, conveyors and motors need to be properly aligned; bearings and the drum’s tracks must be properly lubricated; worn parts must be replaced; the drum tracks, trunnion wheels and seal rings should be resurfaced regularly to ensure proper sealing and reduce stress loading and shock vibration.

Confluence Energy’s new tracking system is paying off with maintenance on its Bliss-brand pellet mills. “We know now that the center shafts go out at about 10,000 hours,” Shearer says. “We replaced all of them in the mills at the Kremmling facility. In Walden, we have four pellet mills and we’ve already replaced two of the center shafts, and we have two left to do in the next three months.” He says shaft replacement is about a three-day project, so it’s one of the larger repairs performed during its weekly outages from Friday through Sunday. 

Winchester says predictive and preventive maintenances tie in with each other. The whole purpose of the weekly shutdowns at Indeck Ladysmith is preventative maintenance: checking belts; greasing; inspecting the drying systems; identifying wear, performance or durability issues with the conveyor chain, auger or auger trough; and checking the wipers on the airlock system. Through such measures, the team develops a history and is better able to predict when, for example, to change the airlock wipers over or flip them—whether that’s every six months or six weeks. “So that to me is when we talk about predicting it,” he says. “Through our regular preventative practices, we know that every three months we need to change out the bucket elevator belt. We schedule it in addition to keeping an eye on it.” 

With only three months of operational history at Zilkha Biomass-Selma, Daw says his team relies mostly on preventative maintenance at this point. “That’s what the OEMs recommend—every seven or 30 days, changing oil, lube, doing what they recommend,” he says. “We’re hoping to quickly get into predictive maintenance as we continue to learn the equipment and how to operate it. Preventative is time-based, and predictive is more condition-based. As we learn what ‘good’ looks like, we’ll move more to predictive maintenance.” Daw says he intends to leverage infrared cameras more to gauge, for example, bearing temperature. “If we see that one day it’s at 100 degrees, and the next it’s at 101, and then goes to 105, that’s a step change, a predictive indicator that it’s going to fail,” he says. “Now we’re doing more preventative maintenance based on OEM recommendations, but we’re hoping to move toward predictive.”

Monthly, Zilkha Biomass-Selma performs some corrective repairs and inspections. “There’s concerns about fouling in the dryer—the induced draft fan—so we open it and make sure we don’t have material fouling the blades,” Daw says. “Another example is, on our hammer mills we inspect for tolerance and the hammer internals to make sure we don’t have a large wear pattern, and that they’re wearing evenly, so we inspect that regularly to make sure we don’t need to make repairs or replace parts.” He says operator inspections are also conducted before outages to identify key indicators for what has to be done once the plant is down.

Daw says unexpected finds during maintenance outages have come to be expected. “It’s newer equipment, we don’t have a history with it yet, so we’re always encountering issues that are not expected,” he says. “A bearing was supposed to wear out in six months, but it wore out in three. It’s mostly minor things, but sometimes it can be major.”

One example Daw gives is the headaches caused by the drag chain. “Once we opened the conveyor, there was quite a bit of wear, more than what we anticipated,” he says. “It was mainly due to the design of the drag chain—it’s a high-incline drag chain, about to the point where physics says this is impossible. We ended up having to rebuild the chain and replace it, and we’re working with a vendor now to figure out a better situation. That was one of the ones where we said, ‘Wow, this is more than we planned for.’ We planned for a 24-hour shutdown, but it went 36 hours. We replaced an entire section of the chain.”  Annual Outages The biggest differences between routine weekly or monthly and annual shutdowns, beyond sheer duration, are the comprehensive planning involved, and size and cost of projects undertaken. Annual outages are also the times many pellet mills plan capital improvement projects. Winchester calls these outages capital shutdowns, and he was in the midst of one at press time. Indeck Ladysmith’s once-a-year outage typically lasts one week, but depending on the scope of work it could go shorter or longer. “If the airlock needs replacement, or the housing is getting worn through, it might take a couple-three days to replace, so the magnitude of what we’re doing during our capital shutdowns ties into what we’re doing weekly.” For instance, a drag chain might need replacement with a cost of $12,000. The lead time for delivery after ordering may be months, and the mill might need a crane onsite to move it into position. Therefore, planning for an annual outage starts three to four months in advance.

A capital shutdown could involve a facility expansion, but Winchester says much of what occurs is efficiency improvements to reduce down time. “There are a lot of ways to incrementally improve production efficiency, without adding a mill,” he says.

Daw says Zilkha Biomass-Selma is planning for a four-day outage late this fall in November or December. “As soon as we see critical issues, we start planning and get it on our monthly or bimonthly maintenance schedule,” he says. “And we’re planning jobs now for November/December. As soon as we find out, we start planning. Parts are a long lead time, many are coming from foreign companies. Most of our products are made in Europe so we have to rely on EU markets. That adds four weeks because of shipping and customs, so the planning starts quickly and depends on the length of the job.”

Some key items for inspection during the upcoming annual outage at Selma include the plant’s environmental equipment—wet electrostatic precipitators, scrubbers, the regenerative thermal oxidizer—along with the compressed air and steam systems.

At Confluence Energy’s facilities, Shearer says they’re down for major repairs three to four weeks a year in the spring during the off-season. “We replace wear items like rolls and dies, and change the teeth in our Laidig raw material handling system,” he says. “We start planning three to four months in advance. It’s also the time when we do capital improvements. Last year, one capital improvement we performed at the Kremmling facility was putting in a new MEC Co. dryer. It was a half-a-million-dollar project that took four months to plan. We were down for three weeks and as soon as we decided to shut down, we had subcontractors on site and ready to go.” Production Efficiency Consider it ironic, but plants have to go down to be effective when they’re up. The most important qualifier to this is being in control of when the plant is down—avoiding having to shut down versus planning a shutdown. “Efficiency ties into downtime, so reducing downtime directly impacts production efficiency,” Winchester says. Other process efficiency improvements come through debottlenecking and finding opportunities to improve throughput, he adds. “Maybe it’s a limitation of your conveyor—it can only move so much material—or your product recipe and raw material species and your respective dies, or critical elements like moisture content, knowing what it is and controlling that. You need process measurement devices like moisture meters that are reliable and accurate, so you can dial it in and minimize variation and, importantly, stay on target. That has a huge impact in how efficient your process is.”

Winchester says he can’t emphasize enough how vital planned downtime is to business. “It’s true that when you’re sitting idle, you’re not making money. You’re losing money,” he says. “But, it’s critical for the plant to perform. Fundamentally, equipment runs out. It’s a critical element of business that’s not to be overlooked. There’s a lot of work staying on top of that.”

Planned outages are extremely important to production efficiency, Daw says. “Mainly because in startup, we don’t know what we don’t know yet. In order for our production efficiency to stay high, we need planned shutdowns to repair properly and quickly. Those outages are very critical to our production.” Author: Ron Kotrba Senior Editor, Pellet Mill Magazine 218-745-8347 [email protected]