Pulp and paper producers looking at more sustainable production would be well-advised to evaluate if their fiber morphology measurements are sufficiently advanced and well suited for the detection and characterization of nano-cellulose fibers.
Online testing for better quality control
Paper, board and tissue manufacturers use pulp from a variety of sources, including long-fibered softwood pulps for strength, short-fibered pulps such as eucalyptus for opacity, bulk and softness, and recovered fibers from a variety of species and geographic areas. Those that control their own forestry operations and have integrated pulping operations can rely on a fairly consistent raw material, but need to measure the effect of pulping parameters on fiber quality. In the case of non-integrated operations, understanding the characteristics of the incoming fibers and having the ability to adjust the recipe is imperative to be able to meet customer product specifications.
For many decades, manufacturers have relied on lab tests such as freeness, shive content, size classification and various strength tests performed on handsheets to provide information on the properties of the fibers they are using. While informative, the data from these lab tests are neither timely enough to make process adjustments nor frequent enough to characterize the variability of fiber properties.
Now, with the advancement of online testing, automated sampling equipment and lab testing –combined with data historians, sophisticated control systems and emerging AI techniques – mill operators can benefit from far more detailed, timely information on fiber quality. These powerful online tools allow quality improvement and variability reduction while lowering manufacturing costs.
Big data & artificial intelligence
Most mills now have data historians that can store thousands of measurements from across the facility at time intervals of seconds to hours, for up to several years, and this data can be used to build powerful predictive models.
Modern fiber morphology analyzers can directly analyze thousands of fiber suspensions, reporting on the deviation of properties such as length, width, wall thickness, shape factor, kink index, fines content, shive content and coarseness. While such measurements help develop better insights into the characteristics of the pulp furnish, and cost little to perform, their real value only emerges when combined with other online and offline mill data to develop tools for better quality control.
Soft sensor and advanced control strategies
Soft sensors, or calculated online measurements, offer huge potential value in their use to control the refining process. A soft sensor specific to a mill’s process can be built using a combination of lab experiments and machine learning. New advanced process control techniques can then be applied, incorporating a
predicted paper strength variable to optimize the refining process.