When transparency becomes a problem for a newspaper
By Pulp & Paper Canada
By Pulp & Paper Canada
Have you ever had the impression you’re reading two news stories at the same time as you read your favourite newspaper? Most likely, you are experiencing the effects of a paper transparency problem, w…
Have you ever had the impression you’re reading two news stories at the same time as you read your favourite newspaper? Most likely, you are experiencing the effects of a paper transparency problem, which refers to a situation where the reader can see the ink printed on the other side of the page from the side being read.
When newspaper transparency becomes a technical problem, it’s time to sound the alarm! Paper transparency issues are frustrating for readers, printers and producers of newsprint.
This problem arose in recent years at a Kruger plant in Trois-Rivires, Que. It was resolved only after a team of workers applied themselves diligently to the task. In May 2007, during the Congrs CFP held in Qubec, Christian Gaudet,* a process engineer at the Kruger Trois-Rivires plant, presented the steps that he and his team took to deal with the problem. This technical communication, entitled “Sheet Transparency on Paper Machine Number 7” won the 2007 Raimbault de Montigny Prize. In a recent interview, Gaudet talked about his team’s experience.
A measurable property
In November 2004, a customer who had been doing business with Kruger for several years complained about an intermittent problem of newsprint-related transparency. Readers were able to see the printing on the other side of the page as a background. In subsequent weeks, the problem was reported again. The customer was a printer with special presses that would permit only a single adjustment of ink per section. Following the first complaint, Kruger began its investigation.
The initial study found no correlation between the problem of transparency and the processes used at the plant. However, a subsequent study in the form of a printability test identified a first line of inquiry. It focused on a weight indicator in the context of a comparison between the sheet produced by a competitor and that produced on one of the two newsprint machines in the plant, PM7.
Kruger realized that the paper produced with this equipment was more transparent than the paper of a competitor. As for PM10, there was no problem of transparency.
The research helped zero in on a problem and identified an approach to solving it. “What we were trying to do,” said Gaudet, “was to transform the needs of the customer into measurable properties. Transparency was the issue for the customer. But we had to address it in terms of a property that could be quantified with our tools.”
That property was porosity. “We can’t address transparency,” explained the engineer. “But we can improve porosity, which has an impact on transparency. It could have been opacity, he went on to note. But in this case, the porosity factor was determinant. “This is what caused the greatest differences from the sheet produced by the competition.”
The porosity of a sheet of paper, noted Gaudet, is defined by how the fibres are entangled together. For example, in a pulp-making process, thick fibres produce a fibrous mat within which the fibres are spaced from one another. Sheet formed this way is very porous, as ink placed on the paper is absorbed deeply, settling into the free spaces. By contrast, a tightly woven fibrous mat holds more ink at the surface. Consequently, there is less risk of having a transparency problem on your hands.
The printability tests, conducted in February 2005, indicated that the sheet manufactured on PM7 had a much higher porosity than that produced by a competitor. Kruger employees carried out the test with the help of equipment designed to measure sheet air permeability.
Recognized throughout the industry, the PPS (Parker Print-Surf) porosity test allows the calculation of the volume of air passing through a sheet every minute. In this case, the test revealed that the porosity of the sheet from PM7 was 100 ml/min (PPS) higher than the competition’s.
Built in 1972, the Kruger paper machine was a “Papriformer” system. The company had two machines of this type in Quebec, the other operating in its Bromptonville plant in the Eastern Townships. With a width of 619 cm (244 in.), the machine installed at Trois- Rivires processes pulp composed of 80% thermo-mechanical pulp (TMP) and 20% de-inked pulp. For porosity, the pulp was a critical factor: the level of porosity can depend to a great extent on the amount of refining the pulp has undergone. For example, a very refined pulp would have a tight fibrous mat, which is likely to have a low porosity level. Under these conditions, the ink flows less easily and is less likely to penetrate the paper deeply.
The DMAIC method
At the beginning of April 2005, Kruger received three new complaints from its customer regarding transparency problems. Suddenly, the company lost the production contract for the first page of the newspaper, which had been printed up to this point on the presses of this printer.
In addition to the financial loss, the effect on Kruger’s reputation was the hardest blow. The company was not willing to let that rest. “That is not the Kruger philosophy,” observed Gaudet.
At the end of April, the papermaker established a committee of five employees responsible for solving the problem once and for all. The creation of the working group followed a tradition well-established at Kruger, noted Gaudet: “Once the problem is identified, a specific committee is formed.”
The new committee was composed of the superintendent for PM7, Claude Pelletier, Jean-Pierre Toupin, general foreman, Francis Lemonde, technical assistant, France Fernet, SSBB (a six-sigma expert), and the process engineer, Gaudet, who joined the team over the course of 2005.
Once the committee was formed, various procedures and tests took place over several months, wrapping up by the end of 2006. During this period, one tool proved especially useful: the problem solving method known as DMAIC (Define, Measure, Analyze, Improve, Control). This performance improvement model includes various stages that must be followed. It includes discussions, trials, cause analysis, validation tests, etc. With each step, rigour is the watchword. “You don’t move on to the next stage while the previous stage is still under way,” explained Gaudet. Kruger’s six-sigma team began first by defining the problem and setting goals. Reducing porosity rapidly became the target, the quantifiable parameter where progress was to be made. More specifically, Gaudet pointed out, the group wished “to reach an average of 320 ml/min of porosity (PPS) with a variation of 3 sigma, that is, 80 ml/min more or less (between 240 and 400 ml/min), 99.7% of the time. The ultimate objective was to obtain customer satisfaction and eliminate the problem completely.
With the subject defined and the goals well-established, the team set out to measure the extent of the problem. It was determined that over a one-year period, from January to December 2005, 46% of the tests carried out on PM7 revealed a sheet with a porosity above 400 ml/ min. This was well beyond the targeted average.
Then, the team undertook an analysis of the likely causes of the porosity. Several measures were taken to address this issue. The Ichikawa diagram, which became known as the “fishbone” diagram, proved very useful. In order to fill it out the committee members put ideas on the table, identifying the parameters which could have an effect on porosity.
The team also classified the variables according to six groups of importance, which included elements such as the configuration of the paper machine, thermomechanical pulp, de-inked pulp, proportion of mixed pulp, sheet properties, etc. Then the members selected 12 variables which had the great advantage of being easy to study, as Kruger had a substantial amount of data on them. The group targeted cationic demand, calendaring, retention, air draw, the wires, etc.
With these variables, the team was ready to launch a correlation study. On one particular day, very low cationic demand was observed. As soon as possible, in para
llel, they examined its impact on sheet porosity. Another day, the cationic demand was very high. The same exercise was executed and porosity data were examined. While the simple linear regression study was state-of-the-art, no significant factor was identified.
The Trois-Rivires team then decided to pursue another step. They selected other variables to be integrated, this time into a multiple regression model. This is a mathematical model with many mutually interactive parameters. This step was based on the reasoning that if a single variable did not have sufficient weight to account for the porosity problem, a comparison with other parameters might be more helpful. Such a study could reveal which factor held the most accountability in the variation of porosity.
The group then set out to collect values at different moments in the day of variables like the speed of the machine, the thickness of the pulp, the secondary mix pump, wetness, the draw between the sections, etc. In parallel, the degree of porosity at the same moment was established, and this information was inserted into the model. The statistics on these parameters varied from hour to hour, as did the data on the degree of porosity. While some factors had no impact on porosity, certain constants emerged. The most convincing results dealt with the draw of the sheet between the different sections of the press.
In terms of presses, PM7 has some very specific properties. “In the jargon of papermaking,” explained Gaudet, “it’s known as an ‘open draw’ machine. It’s necessary to really pull on the paper to draw it from one section to another.” By comparison, there is no need to draw the sheet being produced on PM10. When the sheet is drawn, explained Gaudet, its fibrous bed is dislocated slightly before it moves on its path. “If it is not drawn, the fibrous bed remains tightly woven,” he confirmed.
Armed with these results, the team members focused their attention on the draw between the dryer section and the third press. For this, they applied the five “whys:” Why, for example, is it necessary to increase the draw of the sheet at certain moments? Because the sheet is too wet. Why is it too wet? That depends on the consistency of the pulp, etc.
An efficient tool
In their quest for answers, the six-sigma team set up an experience design project. Experience design, explained Gaudet, replaces the need to run certain tests on the plant floor. “By mathematical calculations, we could see the effect of each variable on porosity. It’s a bit more complex, but it is a very efficient tool because it saves so much time.”
The members of the team created what is known as a fractional factorial design. They fed their tool with information regarding the five parameters (speed of the machine, the opening of the headbox nozzle, calendaring, percentage of de-inked pulp and wetness of the sheet).
For each of these elements, they inserted the range of possible data into the model, and played out all of the variations and scenarios. For example, regarding the headbox nozzle, the team entered the information on its different possibilities of opening. Then, the group introduced data concerning the flow of water or the consistency of the pulp.
Two major variables emerged from the experience design project: the speed of the paper machine and the opening of the headbox nozzle. With this project, noted Gaudet, “we were getting more concrete; dealing with the variables that really had an impact on porosity.” Thus, by reducing the speed of the paper machine to 90 ml/min and by opening the headbox nozzle to the minimum, the team obtained a porosity of 346 ml/min. This index of porosity was reduced by 110 ml/min compared to a situation where these two elements were adjusted to the maximum.
There remained a problem however. The factor of speed could only serve as an indicator. Reducing the speed of PM7 was out of the question, for financial reasons.
Despite all of the tests, the team’s goals had still not been reached at this stage. Therefore, the team turned to another parameter they knew was crucial: thermo-mechanical pulp (TMP). They felt it should be possible to establish a correlation with porosity. The team noted that the CSF drainability index of this pulp had been increased in the summer of 2004. This roughly coincided with the first complaint about transparency. They wanted establish this link beyond any doubt, and began by moving forward with various simple linear regression tests.
The team fed PM7 with drainability set at 105 ml, then back to 90 ml. They repeated the same scenario several times, and the situation became very clear: in step with the sequences, porosity increased, then fell, before increasing again, etc. “This factor alone, the variation of the drainability of the TMP, explained 40% of the variation in porosity,” affirmed Gaudet. “It was major. There is no other factor that can explain more variation.”
The Kruger team defined the extent of the problem, prior to its analysis. Now they could examine solutions.
Regarding the draw of the sheet, the operations people had a very good idea, reported Gaudet. They installed a laser pointer on the paper machine, a device that could be bought at any hardware store. The red ray, projected at the same spot, serves as a visual indicator to workers. Before entering a new section, the sheet must be aligned with the laser. This ensures the optimal draw. If it is not aligned, the workers can proceed with adjustments. In short, people on the shop floor no longer have to wonder whether or not to adjust the draw. While the installation of the laser pointer brought positive results, it did not reduce the porosity of the sheet. However, it did decrease the variation, and brought more uniformity. “Instead of being subject to major variations, we were more in control,” Gaudet pointed out.
The second solution proposed was to open the headbox nozzle to the minimum. The tests for this had been convincing. It had the effect of reducing porosity by 30 ml/min.
The third and final solution concerned the drainability of the thermomechanical pulp. Recognizing that PM7 could not operate with a high CSF index, this was adjusted in terms of the porosity of the sheet. If, for example, the porosity attained a peak of 400 ml/min, the CSF would be reduced. Experience proved that drainability must be reduced approximately 40% of the time. This adjustment is performed on PM7 (on PM10 there is no impact). The result: by reducing the CSF from 105 ml to 90 ml, an average reduction of porosity of 67 ml/min is obtained. “This is the major impact on porosity,” confirmed Gaudet. However, there is a cost. To produce a pulp that has water flow in it less easily, from wood chips, then fibres, requires much more energy than the standard process. The operation adds $250,000 per year to the electricity bill. “But we had no choice,” said Gaudet. There are a number of energy optimization projects being studied at Kruger in the de-inking area and in energy recovery, for example.
Controlling the processes
In October 2006, the Kruger six-sigma team was able to declare the paper transparency issue mission accomplished. When the team began its work, PM7 had an average sheet porosity of 400 ml/min; one and a half years later it was between 245 and 375 ml/min. Furthermore, it was accompanied by a much lower variation. Before the tests, the sheet produced reached 400 ml/min porosity, 46% of the time. After October 2006, the 400 ml/ min cap was never exceeded.
The goal was reached through the combined action of these three solutions. “These three factors together provide less variation in the porosity, and a lower level of porosity,” Gaudet said. Reducing the CSF index of the thermo-mechanical pulp was the key factor, he added.
In closing, a word about process controls: for Gaudet, this is an important aspect that cannot be neglected. Once the problem is identified and solutions found, it is vital to ensure the p
roblem does not reoccur. For this reason, the headbox nozzle was locked, and cannot be modified. As for the draw of the sheet, the operations people are proud of the laser pointer indicator solution. They visually verify it constantly. Finally, regarding the drainability of the pulp, the operating procedure is clear: when the plant is producing paper for a customer, workers adjust the CSF. “It is a solid procedure with the personnel,” noted Gaudet.
“The use of the DMAIC problem-solving method allowed us to achieve the porosity target, to reduce its variation and to eliminate the problem of transparency with the customer,” concluded Gaudet. After 2006, the problem never came up again to the great satisfaction of the customer.
*Note: Since the time of the interview, Christian Gaudet has left the Kruger plant in Trois-Rivires.