Leveraging mill-wide big data sets for process and quality improvement in paperboard production, TAPPI JOURNAL May 2016
Please Note: This document will be available in PDF format in the "My Electronic Documents" link on the home page once your order has been completed. Please make sure you have the latest version of Acrobat Reader. Click on the Acrobat Reader icon to check for the latest version, it’s FREE. To print a hardcopy of a PDF file correctly you must have a postscript printer. If you are not sure if your printer is a postscript printer please refer to your owner’s manual.
Purchase of electronic (downloadable) documents made at www.tappi.org by credit cards, followed by instant download CANNOT be cancelled. We do not offer refunds on electronic download documents.
The MWV mill in Covington, VA, USA, experienced a long term trend of increasing episodes of paper indents that resulted in significant quantities of internal rejects and production downtime. When traditional troubleshooting techniques failed to resolve the problem, big data analysis techniques were employed to help deter-mine root causes of this negative and increasingly frequent situation. Nearly 6000 operating variables were selected for a deep dive, multi-year analysis after reviewing mill-wide process logs and 60000+ PI tags (data points) collected from one of the major data historian systems at the MWV Covington mill. Nine billion data points were collected from November 2011 to August 2014. Strategies and methods were developed to format, clean, classify, and sort the various data sets to compensate for process lag time and to align timestamps, as well as to rank potential causes or indicators. GE Intelligent Platforms software was employed to develop decision trees for root cause analysis. Insights and possible correlations that were previously invisible or ignored were obtained across the mill, from pulp-ing, bleaching, and chemical recovery to the papermaking process. Several findings led the mill to revise selected process targets and to reconsider a step change in the drying process. These changes have exhibited significant impacts on the mill’s product quality, cost, and market performance. Mill-wide communications of the identified results helped transform the findings into executable actions. Several projects were initiated.
Papermaking, like other manufacturing, is facing an unprecedented explosion in data collection and availability. Big data analytics can develop predictive models and find critical insights that are essential to improve operational performance.