Research & Innovation
PacWest 2021: First virtual conference a success
By Martin Fairbank, Ph.D.
Measurement and control throughout the process ruled PacWest 2021 virtual conference.
By Martin Fairbank, Ph.D.
PacWest, the annual Canadian pulp and paper conference of the western and Pacific Coast branches of PAPTAC, went virtual this year because of the COVID-19 pandemic. The sessions were split into five technical sessions containing 27 presentations spread out over three days (May 27, June 3 and June 8). PacWest, organized annually by professionals within the western Canadian industry, is known for its presentations on subjects of practical interest for pulp and paper mills, and this year was no exception. Many of this year’s presentations focused on ways to monitor and control mill processes, all the way from the chip pile to the effluent treatment plant.
An article in Pulp & Paper Canada’s winter 2021 issue featured the applications of online near-infrared (NIR) sensors, including chip moisture measurement. Guillaume Hans, a senior scientist at FPInnovations, where these sensors were originally developed in the 1980s, gave a presentation on recent work using the online chip measurement module to measure properties such as chip brightness and the content of extractives, lignin and total carbohydrates. He found that the sensor could readily distinguish between five different species based on the concentration of these components, and robust models could be built. This opens up the possibility of using the sensors to not only control feed rate based on dry weight, which has already been implemented in some mills, but to give further insight on chip management when managing a species mix.
Moving forward into the kraft pulping process, Ralph Lunn at Mercer Celgar gave a very interesting presentation on the use of equations developed many decades ago by John Tasman and co-workers at FPInnovations for Kappa factor (lignin content) and pulp yield. Both these parameters can be predicted from the H factor (dependent on temperature and time), liquor charge (effective alkali concentration EA) and liquor sulfidity S. The equations are quite complex, involving three wood-species-dependent constants for Kappa and four for yield. Furthermore, Tasman worked out the values of these constants for many species. Of the seven species used at the Celgar mill, only the yield constants were missing for hemlock. FPInnovations helped Celgar to establish these missing parameters. Lunn used the equations to determine how to maximize the yield at a target Kappa value. He worked out that by optimizing the EA and S levels, the pulp yield for his species mix could be increased by almost half a percentage point, which is worth a lot of money on an annual basis. PacWest presented Lunn with the H.R. McMillan trophy for the best presentation of the conference.
Many kraft pulp mills are now using online sensors in various parts of the mill, and several presentations were given on the use of these sensors for process control. Danny Zuver of Valmet presented the case for using the Valmet Recovery Analyzer (Alkali-R), which is a robotic analyzer that can perform up to 265 full ABC titrations per day, duplicating the lab method to determine total alkali, carbonate, hydroxide and sulfide. Used in combination with advanced process control, one mill was able to reduce the slaker TTA (total titratable alkalinity) variability by 72 percent and achieved a causticization efficiency of 82 percent.
While online titration duplicates the exact lab procedure for determining chemical strength, this can also be done using spectroscopic techniques. Jonathon Speed of Keit Spectrometers presented their FTIR device with no moving parts, suitable for harsh industrial environments, which has been purchased by a Swedish pulp and paper mill for online green liquor analysis. Alex Moline, area supervisor at the Domtar Dryden mill, talked about their use of FITNIR online near-infrared sensors for control of their recausticizing area using a combination of a soft sensor for TTA based on density, feed forward and feedback controls. Besides a reduction in variability and an improvement of two percent in causticizing efficiency, there was much less buildup of calcium carbonate scale on their top separator, as shown in the picture on the previous page.
Controlling the process
Devin Marchman from Spartan Controls gave a presentation on control of the Canfor A and B mill bleach plants in Prince George. The plants had many online measurements that generally worked well, but it was unclear how to use these measurements to help optimize their bleach usage. Marchman helped them develop a bleach plant control strategy that leveraged all the available measurements, provided better feedback to the operators, and was easy to maintain. It was a flexible design and resulted in bleach savings of $2.5 million per year from the two mills combined. Half of these savings were achieved by rebalancing and controlling the ClO2 dosage in the D0 and D1 stages, and one-third of them came from peroxide and caustic savings in the E stages.
An often-neglected area of the mill for process control is the evaporators, which concentrate the weak black liquor (~15 percent solids) to heavy black liquor (60 to 70 percent solids). Process control is difficult because of the long lag times when a process change is made. Val Parisien of Enero Solutions made a presentation on implementing evaporator control at the Mercer Celgar mill. The solution used model-based control that combined feed-forward, feedback and cascade control models, resulting in an overall 6.3 percent improvement in steam use in the evaporator plant.
Best student paper
The best student paper award went to Lee Rippon, a graduate student at UBC, for his paper on developing a machine learning model for detection and diagnosis of ring formation in lime kilns. This buildup of lime mud in the lime kiln is poorly understood and results in unscheduled downtime at many pulp mills to remove the buildup. Rippon’s approach was to try and correlate ringing events with temperature on the outside of the kiln (measured with thermal cameras), because the deposit acts as an insulating layer, resulting in a lower temperature at the location of the ring. In this first part of his research, he talked about how he developed a visual approach to data analysis and dealt with outliers in the data. He hopes to develop a supervised machine-learning approach and hopes to be able to present the final results at next year’s conference.
A maturity index
The final stage in any mill is effluent treatment. Although it is not as high-tech as other parts of the mill, it is important because it can shut down the whole mill if the process gets out of control. Understanding what to measure in order to keep things running smoothly is, therefore, an important topic. Kari Plamondon, lab supervisor at West Fraser’s Hinton mill, gave an informative talk on how she monitors the health of that mill’s aerated stabilization basin (ASB). The mill has developed a numerical “maturity index” by counting the types and frequency of different micro-organisms under a microscope; when the number gets higher, indicating older, less efficient organisms, corrective action is required.
Worth the risk
The PacWest organizing committee had to cancel the 2020 event. When they realized that an in-person event would still not be possible in 2021, they created PacWest 2021 in a virtual format to keep the interest in PacWest alive. There was a certain risk to this new format because it wasn’t certain how many would attend. In the end, some submissions had to be turned down due to a full schedule. Conference chair Kerry Morton was very pleased with the turnout and the quality of the presentations. The maximum number of people attending reached a maximum of 123 and was rarely below 80. PacWest 2022 will be held live in Jasper, Alta., from May 25 to 29.
Martin Fairbank has worked in the pulp and paper industry for over 35 years and is currently a consultant and technical writer.
This article is featured in the Fall 2021 issue of Pulp & Paper Canada.