Pulp & paper industry writes a digitalized future
The pulp and paper industry and its many manufacturers have experienced more ups and downs than a roller coaster in recent decades due to fluctuating demands for different types of paper and the evolving technologies for producing them.
“I’ve had five different owners over 30 years, but somehow kept the same phone number. Early in my career, life was good and we built new equipment. Then came the downturn, and we had to focus even more on reliability to minimize costs,” said Greg Drewiske, PE, manager of engineering and capital at Billerud Americas Corp., which produces coated papers in central Wisconsin. “Now, the mills that are still open have a lot of obsolete, proprietary equipment, and the need to incorporate more automation is even greater.”
Tom Shope, VP of digital transformation solutions at Endress+Hauser, added, “There’s fewer staff, so it helps to have more intelligence in devices. However, the question now is how to use all the data they’re generating? For example, how we can turn it into maintenance workorders and get them to skid builders or suppliers.”
Drewiske, Shope and other co-panelists explored how the pulp and paper industry can optimize and transform with an assist from automation and digitalization at Automation Fair 2022 this week in Chicago.
Right data, right hands
The panelists reported that more detailed, targeted and distributed data can greatly enable pulp and paper processes if it quickly reaches decision-makers who can make the best use of it.
“At the 8 a.m. meeting, operators and managers want to know why their process was lost last night, and this usually means pulling up Excel spreadsheets and cross-correlating them,” said Dan Timmers, solutions consultant for machine design, information and analytics at Rockwell Automation. “Now, more information is getting processed at the edge—such as signatures of flows on a day-to-day basis—and getting it to the cloud, which has the potential of giving us the equivalent of Google Maps for process control. This is especially important because the old guys that could walk by equipment and tell that something was wrong from the vibrations are retiring. Because this tribal knowledge is vanishing, we need controls that can tell us what the 4-20 mA signals can’t, and that transition is coming.”
Timmers cautioned that, even if information can be shown on a dashboard, it can’t be just blind data points, and requires components and software that are smart enough to immediately put it in useful context. “This capability is immensely useful. Some applications can predict motor failures two or three months ahead of time,” he said. “We need devices to fill in for experts that have retired, but those devices need to first establish performance baselines before they can look for anomalies.”
Education softens reality’s punch
To gain the benefits of added automation and more digitalized data, some wrinkles must inevitably be smoothed out. Drewiske reported that Billerud’s mill installed a single-loop distributed control system (DCS) about 10 years ago, which included HART-enabled devices that pointed out new details when they were connected.
“It turned out that 30% of our instruments had issues we didn’t know about before, such as valves sticking,” said Drewiske. “We also implemented control-loop, performance-monitoring software to identify tuning issues. In our drying section, we took current feedback, and relayed them to FactoryTalk Historian software.”
Shope added, “IT is taking over some of these areas, but the question remains: how do we get 30 years of plant-floor experience into just three years of staff training?” Because IT is getting more closely involved with production issues, Timmers advised users to designate a leader and cross-functional team to handle IT’s presence in the process space, such as deciding on a policy for configuring switches.
“We have to educate our IT guys,” said Drewiske. “Previously, if we couldn’t get email for 3-5 seconds, not many people would notice. However, if we lose control of our paper process for 3-5 seconds, a lot of people would notice.”
Tooting automation’s horn
Drewiske reported that increasing automation, digitalization and data analytics in papermaking is also important for meeting enterprise requirements and achieving corporate goals. “It’s pretty simple to meet basic production and product priorities,” he said. “However, it’s harder to also stay ahead of technical obsolescence, as well as address what a company wants to do as it goes from being family-owned to hedge fund-owned to internationally owned.”
Drewskie explained that plant-floor personnel tend to fix equipment and systems and not make a big deal about it, but this can result in other parts of the organization not being aware of OT’s importance or what it requires. “Equipment becomes obsolete, and parts can’t even be found on eBay, and then we have to negotiate. I always say the best time to negotiate is at 2 a.m. after 10 hours of downtime,” said Drewiske. “In any case, we have to make clear what our processes and equipment need to maintain availability and uptime, and hopefully invest in a phased approach.”
Unfortunately, another hurdle to securing investment in pulp and paper is that fact that operations issues and the cost of fixing them typically overshadow maintenance concerns and advocating for them, according to Drewiske. "We measure overall efficiency and target speed and quality, but when we’re looking at lost time, dealing with operating problems outweighs maintenance, and makes it hard to justify investing in it,” he explained. “If a company is looking at a multi-million-dollar project, they usually want a 30% return on investment, and it would take maintenance far too long to reach that ROI and justify that investment.”
Drewiske added that automation and digitalization take operations, maintenance and other data sources out of their formerly separate silos, making them easier to pull together. “We don’t look at these reports as much as we should, but we have a corporate reliability initiative to focus on developing systems that can integrate our data sources. I think integrating them is going to get easier in the future, once we get over the hump of bringing them all together.”
