Multivariable Control

Bringing the Lincoln Paper Mill Back from the Dead

Lincoln Paper Wasn't Even a Zombie Company. It Was Dead. Closed. Kaput. Then New Owners Brought in Multivariable Testing to Help Resurrect It Through a Top-to-Bottom Revamp of Procedures, Processes and Plant Culture

By Nancy Bartels

When the paper mill in Lincoln, Maine, closed in 2004, it wasn't a tragedy just for its owners and the workers there. Losing the mill was a serious blow to Lincoln, a town of about 5000 located in east central Maine. There had been a mill of some kind in Lincoln since the 1800s, and lumbering and paper-making were embedded in the area's history and culture.

For nearly three years, the mill doors remained closed, and it looked like the next stop would be dismantling the plant. Nobody wanted it. Then a group of investors, led by Keith Van Scotter, bought it up for about $33 million dollars.

But that was only the beginning of the project of getting Lincoln Paper and Tissue up and running again and, above all, profitable.

Lincoln is not a small facility physically, covering some 15 acres, although it does only about $150 million in sales, not all that much in the paper business. It makes 140,000 tons of paper and tissue a year and has six integrated pulp and energy production lines, two paper machines, a pulp dryer and three tissue machines. Its machinery runs the gamut from some of the oldest in the State of Maine to the newest—a tissue machine installed in 2006.  Its Paper Machine 4 is over 100 years old.

The business has a checkered history. CEO Van Scotter explains: "As a pulp and paper company, it was formed in 1882, and paper machines were added in the first half of the 20th century, but there has been manufacturing on the site since the 1850s. It's been through a couple of ownerships and a couple of bankruptcies—the first time in 1968. It was purchased and restarted then and tissue-making capacity was added in the 60s and 70s. Pulp mill improvements had been added in the 1950s and 60s, but due to a combination of factors, it went into Chapter 11 in 2000."

Unable to work out an arrangement to get it out of Chapter 11, by 2003, it had again been shut down and was about to be sold for scrap when Van Scotter and his associates negotiated a highly leveraged buyout and re-opened it as Lincoln Paper and Tissue.

Immediately Van Scotter went out to raise the $40 million to buy the modern Tissue Machine 8, but it would take more than that to get Lincoln back on its feet. "We had a tremendous need to drive the company forward without spending capital to get the kind of improvements we needed to stay in business. We were making a little bit of money then; we'd invested a lot of money, but we had to pay the bills. It was a survival thing."

Read Also: Thoroughly Modern Mills: Enabling Knowledge Workers for the Next Century

And it wasn't just Van Scotter's and his associates' money riding on the line. Lincoln now employs nearly 400 people, many of whom already had been through one factory close in a one-factory town, and Lincoln is the biggest employer in a 50-mile radius. Failure simply was not an option, and frugal from-the-ground-up improvement was the only way to go.

"The company had been in bankruptcy for three years. Only limited expenditures were made on maintenance and capital improvements," explains Van Scotter. "It functioned when we got it, but it needed a lot of work."

And Van Scotter had little time and less money to get his improvements in place. "We didn't have the capital a lot of companies did; we didn't have the people a lot of companies did. We had what we had."

Bringing in MVT

That's when Van Scotter called in his old friend Art Hammer of QualPro in Knoxville, Tenn. QualPro is a consulting service that offers a technique for process improvement called multivariable testing or MVT.

MVT takes the guesswork out of selecting business improvement strategies by testing a multitude of ideas at once and revealing the positive, negative or neutral impact of the ideas. Van Scotter breaks it down this way: "Twenty-five percent of the factors will help [improve a process]. Another 25% will make things worse, and 50% don't matter."

What MVT does is demonstrate which is which.

"The neat thing about MVT," says Van Scotter, "is that in classic process improvement, you test one factor at a time. MVT allows you to take ideas from everyone and test a lot of different factors at the same time, and through showing their statistical significance, you can tell what helps, what hurts and what doesn't make a difference. You likely won't find a solution running individual tests. The math is impossible. You'll miss out on all the interactions. And you just run out of time. It's physically impossible."

QualPro's MVT methodology focuses on harvesting the ideas and creativity of the entire workforce to develop innovative ideas and then statistically determines which of them have the strongest impact on improvement. At Lincoln Paper, the rule of thumb for suggestions is "can it be done quickly and cheaply." If it can, Van Scotter says, "Then we'll try it. The science and the statistics are hard, but the actually testing is really simple."

The idea is to take the results of brain-storming sessions that involved everyone—operators, engineers, managers, safety personnel, millwrights, electricians—and gather all the ideas they think will improve a process. Everything from changing the recipe for the process to the brand of coffee in the break room is under consideration. Nothing is off the table. These ideas are winnowed down to those that seem most practical. Then the testing of them begins.

The collaborative nature of the MTV process is one of the things that appeals to Van Scotter. "One of the neatest things about this is that MVT forces us to engage everybody, regardless of where they were in the organization, and we were working on real stuff. Nothing is worse than working on things that nobody cares about."

Van Scotter admits the process had its skeptics at first. "I had a couple of senior guys that were blockers. I had to let them go."

But this kind of change is easier to sell at a company such as Lincoln, where everyone feels his or her back is to the wall. "It's tougher to make this cultural change at a successful company. Complacency is hard to overcome. We didn't have to work that hard at Lincoln."

Test Results

For Lincoln Paper, the multivariable tests ranged from one shift to four months in length. Some were "emergency" tests to address an immediate problem. Some tackled longer-term process difficulties, whether an improvement of a particular machine's performance, development of a new product or sales and marketing strategy.

Sometimes, more than one series of tests is required. The first set of tests gives a broad outline of the solution and the second tests refine it.

And the results are often surprising and counterintuitive. For example, according to Van Scotter, through MVT, his staff learned that one of the most hurtful factors in winder performance was lubricating the chucks. "Who would have thought that?" he asked. "I know I didn't. I don't think anyone else did either. You're supposed to lubricate stuff, right? Well, we found out not.  Once we figured out there were several things we could do, we ran a refining experiment and changed the way we worked the machine. On a second test, you work with those help/hurt factors to discover a better way to run your process."

Read Also: What's Green and Safe and Makes Lots of Pulp?

Art Hammer, who himself was the on-site man from QualPro at Lincoln, says, "One of the most interesting things about MVT in general is the mindset that when you're having a problem with quality or production, you slow the machine down. We have found with MVT that not one time did slowing the machine down help. Sometimes speeding the machine, along with other things, will improve the system. Speeding the machine up is always cheaper. That's counterintuitive, but it's true."

MVT is also a good antidote to the "we've always done it that way" mindset. Van Scotter observes, "How many people are running a business where you say, ‘This is what we gotta do. It adds cost, but we know we gotta do it'? That's a placebo. A lot of procedures are based on theory and opinions. MVT gives you a robust fact set and provides discipline."

Both Simple and Hard

On one hand, MVT is hard. The mathematics behind it is based on the Plackett-Burman Matrix, a statistical formula based on complex equations. "The idea is to test a number of variables," says Van Scotter, "Three, seven, 11, 19—any number. Take the factors and set them based on what the matrix tells you to do. If you're working with 15 factors, you must run 16 experiments. But once you get all the data, you apply the math to figure out which factors work. It's simple math. Add, subtract, divide."

But even before you get to the testing phase, Van Scotter warns, you have to make sure what you're trying to improve can be measured properly. "You cannot take this for granted," he says. "Say you want to ‘improve production.' Is that gross? Net? Over time? You have to figure that out before you actually do the experiment. [The thing is] the quality of measurement systems in most companies is really bad. You need to check that measurement system first. Most of the time, the data you're working with is wrong. The first thing you have to do is fix the measurement system."

Big Machines, Big Problems

Tissue Machine No. 8 is the newest equipment at Lincoln. Bought in 2006, it was the first major investment Van Scotter and his associates made in the plant when they started it up. But, says Van Scotter, "The damn burners didn't work right. Two years down the road, and we're on our sixth version of the burners. We shut down for two days, had the experts in, the vendors, the engineers. We got ready to start up, and we had black tissue—soot all over the tissue. We're burning oil to help make the hot air to dry the tissue. It was a total disaster."

He continues, "The vendor's suggestion was to rip the stuff back out and put back what we just took out. That's another 48 hours down, and this is a Friday afternoon. That's not going to cut it. Our technicians and operators said, ‘Let's try to do something to fix the problem.' So in 10 hours on a Friday night, we run some MVTs and discovered three factors that reduced the problem so we could get to a saleable product."

MVT came to the rescue again with the No. 4 paper machine. Problems with it had cost 10 days of lost production in six weeks. The Lincoln Team ran MVTs on 100 factors to understand the situation. In the end, it was a mechanical problem, but as a result of the testing, Lincoln learned some valuable, money-saving lessons.

Lincoln had taken softwood craft out of the recipe because it was driving up costs. Popular plant wisdom was that removing it was causing the problem. But MVT showed that adding or taking out the softwood craft neither helped nor hurt machine performance. And adding it also added $50.00 a ton to the cost.

Feeding the Bottom Line

These anecdotes underlie the fundamental strength of the MVT approach. It is infinitely adaptable. Once a team is trained to use it, it can be deployed to solve problems from batch production issues to new product development. It doesn't require calling in the consultants every time an issue arises.

And the benefits are real and measurable. In Lincoln's case, MVTs have lead to such items as a $250,000-per-year reduction in oil costs for running an inline kiln, achieved by running a couple of 16-run MTVs. A seven-factor MVT on making 92-brightness paper reduced the cost of getting to 92 brightness by half—all done in two days of testing. On the No. 8 tissue machine, $300,000 a year was saved on strengthening chemicals by virtue of MVT. The same tests were run on the No. 6 tissue machine and saved the company another $150,000 a year.

Admittedly Lincoln Tissue and Paper may be something of a special case. The biggest champion of MVT at Lincoln was the CEO, which never hurts a project. There was a highly motivated workforce that understood that throwing money at problems was not an option and knew all too well the consequences of failure. By starting with small projects and working up, buy-in has come along. Now 120 people in the plant are trained in MVT and are believers in the system. The workforce understands that "we have always done it that way" is not an acceptable approach to process improvement and that testing improvements via MVT is not an add-on but just another part of their jobs.

But Lincoln's success is undeniable. It's still in business, which after its travails of the last decade is something of a miracle in itself. "When we restarted on 2004, not many people held out much likelihood of us surviving," admits Van Scotter. "There were only two types of people in Maine and in the industry—those that were skeptical and supportive and those that were skeptical and not supportive. But there was nobody that wasn't skeptical."

Now, Lincoln is paying its bills—thanks in part to MVT. "This is the difference between paying all the bills and not paying them. We started with nothing, and we've still got it. We were leveraged at the beginning. We've paid our debt down, and it's because of the tools we've had with MVT. It's not the only difference-maker, but it's been a huge one," says Van Scotter.

Van Scotter closed our interview with a story he says epitomizes Lincoln's position. Two Maine men are ice fishing and encounter a very large, aggressive bear. They start to run away, and one man says, "You can't run away from a bear." The other replies, "I don't have to beat the bear. I just have to run faster than you."

Lincoln Tissue and Paper is running faster than its competitors thanks to MVT.