The optimized process with data


Digitization, i.e., data analysis, is a useful and surely successful means of optimizing the operation of any production process, including the paper process. However, one needs to start from prerequisites such as having a good infrastructure, good software, good measurement technology and, last but not least, trained and experienced personnel. Digitization therefore requires collaboration, but the benefits it brings are significant, and through data analysis, it is possible to get a clearer view of what is happening in the production cycle.

The meaning of digitization is advanced data analysis. Being able to apply it in a production process results in its greater control and the ability to immediately identify a possible problem and, consequently, solve it. A digitized system collects, analyzes and cross-references millions of data with an accuracy and speed not otherwise possible. However, the human component, made up of the experience and knowledge essential to correctly interpret the indications provided by the numbers, also remains important.

Eugenio Ciucani, energy engineer at Holmen brought the example of what he achieved in his company by digitizing the paper production process.

The steps of a good project

In 2017, Ciucani told Miac 2023, a roadmap began at Holmen for the development of process digitization, based on the certainty that better data analysis can be helpful in increasing process reliability. First, a roadmap was established in which three phases of the digital systems implementation project were identified.

“The first step to get good data analytics is to have an equally good infrastructure,” says the engineer. “Nowadays, the machinery in a paper mill’s production process can provide a huge amount of data, but we need that data to be first of all of good quality and to be able to make use of good software and also a system with good storage so that it can be made homogeneous and accessible.” This means, for example, having a single platform from which to have access to different sources of information.

The second step-which, Ciucani points out, is a natural consequence of the first-is the ability to perform data selection. “We need, in effect, to have the right detection period and also the right resolution.”

Finally, the third step, involves the need to “communicate the information collected to all stakeholders” thus arriving at what he calls “a proper data visualization strategy.” All this without forgetting one last essential element, the experience of the operators and their skills, which it is important to maintain and cultivate.

These first three steps, the engineer explains, are a fundamental prerequisite for then proceeding with the digitization project, as they “already provide added value to the business. Once these first three steps are completed, you have a solid foundation on which to build the project.”

Example on the pump

The system thus structured-that is, with quality data, easily accessible from a variety of sources, and with a high-performance IT infrastructure-has been applied to Holmen Group’s production sites and, in some of them, has proven to be able to solve different problems that were compromising the smooth operation of production lines.

One such case involved the use of a pump. The monitored process data had shown, at one point, that certain properties had changed. “At first we didn’t realize what the cause was,” says Ciucani, “after a series of comparisons we realized that it stemmed from the water management system, specifically a pressure pump. 

Here we identified a first problem: the pressure sensor placed after the pump is part of the control system, and we observed that its frequency was very high. We knew that the system had some bearings and these were also showing a very high speed. Therefore, we decided to attach a new pressure transmitter to the system to thus have an additional point of data recording.” The new equipment showed variations on the pressure, with spikes of a few seconds, thanks to which the mill’s technicians were able to understand that the disturbance was coming from the pump. “In particular, we hypothesized that the problem was attributable to the presence of air bubbles in the flow.” And indeed, the application of an antifoam in the correct dosage eliminated the pressure spike. “Our hypothesis was correct, indeed there was an air entry in the flow, we just had to prove it, and by combining the process data analysis, we were able to identify the changes in the frequency and solve the problem.”

The weight case

The second case involved the prediction of variation in grammage. Grammages, Ciucani explained, can be measured in a variety of ways. One standard in the industry is the use of a camera system: every 30 seconds or so, a scanner monitors the surface of the sheet while, at the same time, the operator takes a sample of the product, making sporadic checks on the paper, which take place on average every 220 meters or so of paper produced, or every 45 minutes. Certainly, says the manager, this is not an optimal control method.

The first part of the intervention involved analyzing the status quo. “We used existing technology, scanners, and even manual checks performed by the operator. We noticed a significant variation in the operators’ performed checks, which could be due to a number of elements. So we used sensors applied to the process and measured flow, resistances and weight. But the interesting aspect was that, with the digital system, we could have sampling per second, so no longer every 220 meters or every 45 minutes. So the detection of possible variations became much more frequent, and we had a much more accurate view of what was happening on the line. And in fact, we noticed pulses that are a clear indication of a problem on the process.”

Thus, the new tool allows for higher quality and resolution data that give a more accurate view of what is happening in the process and, as a result, optimize it.

MD variations

Third and final case of application of digitizing systems involved the identification of MD variations on sensed data, both those in the machine direction (MD)-that is, along the fiber-and data in the cross direction (CD)-that is, against fiber. “In this case, we used cylinder-by-cylinder thermal mapping and analyzed the results in our system. This allowed us to notice a kind of pattern i.e., phenomena repeating. We focused on a 24-hour period and analyzed the main variations over that time frame.” The system allows measurement of weight, thickness and moisture, and could also calculate density. The sticking point, in this case, might have been density, with respect to which the graphical data indicated a recurrence.

“The first thing we had to do was to isolate the area where this pattern was occurring, and so we could see that the CD and MD were different. In particular,” says Ciucani, “we noticed that the pattern resembled another situation; we noticed that this variation in density had a meeting point with a variation in the level of an installed pump. So we did another check and got confirmation that that variation in the level of the pump was also responsible for the variation and error in the density.”

At this juncture, then, cross-referencing the CSF data showed that the product quality data also provided information on process quality.

Four centuries of history

Holmen has a history of more than four centuries behind it. It began in 1609 with a gun factory on the island of Kvarnholmen, a few kilometers from Stockholm, Sweden, the place from which it also took its name.

Over the centuries the company has changed and is now a full-fledged group present in thirteen countries around the world. Its headquarters, however, has remained Sweden, where most of its approximately 3,500 employees work, and it produces total sales as of 2023 of about SEK 22 billion, which is equivalent to more than 2 billion euros.

Over time, the Holmen Group has specialized in different market sectors, which has led it to structure itself with four business areas that are different but share the same philosophy, built around the eco-cycle of forests and the renewable products that can be made from them. The four business areas are:

– forests

– renewable energy

– wood products

– board and paper.

The paper area includes four mills-three located in Sweden and one in the United Kingdom-that produce 1.5 million tons of paper from virgin fibers for advertising, magazines, books, transport packaging, and the premium segment.

These are lightweight papers made from traceable raw material from sustainably managed forests.

The group continuously invests in the Paper business area, to which 4 percent of capital is allocated.

One area for paper and cardboard

At the end of January 2024, the Board of Directors of the Holmen Group decided to partially change the structure of its business areas. The decision, which is part of the strategy of continuous development of its activities that the group is conducting, has led to the amalgamation of two previously separate lines, those dedicated to activities related to, one, board and, the other, paper. The idea of creating a single business area, Holmen Board & Paper, is aimed at increasing its efficiency, competitiveness and development opportunities.

“By bringing together our expertise in the process industry into a single business area, we are increasing our capacity to develop our business in all areas, from consumer board to transport packaging, books and magazines,” Henrik Sjölund, president and CEO of Holmen, said in a press release.