Multivariable Process Analysis

Multivariable process analysis becomes easier by delivering a visual presentation of all the variables on a single graph.

Complex processes and machines are controlled with hundreds of parameters, many of them interdependent. Despite product grade recipes, advanced algorithms and operating experience, it is not always possible to stay in control of the final product quality.

Intertek applies advanced software to create a visual representation of all the process parameters affecting your operation, there may be thousands of them, and places them on the same graph. With simple queries, the good operating zones are easily identifiable. An optimum process model can then be implemented on line, so that the optimum product is always produced.

It can be frustrating and expensive too when on some days your process runs smoothly and other days it does not. Intertek can help show you how to mitigate this situation and save your operation time and money.

Case Study:  Surface defect on speciality paper 

Objective:
To eliminate an edge defect so that the mill would no longer have to remove the edges prior to sending reels to the customer.

Problem:
The cause of the sudden onset of the defect, known as “cockling”, could not be found. The mill was required to trim 15% of the width off each edge and had been doing so for several months. There were periods when the defect would disappear but it would return some time later for no apparent reason.

Analysis:
There was no record of cockling severity, so an index of the defect had to be developed, and new process data had to be logged. The outcome was an assessment on a scale from 1 (worse) to 20 (best) to be made by the operator studying an end-of-reel sample. A process mapping exercise was then conducted identifying all the parameters that could possibly relate to the defect, and methods of logging them were determined if systems were not already in place. Data was recorded alongside the quality index for each reel of paper produced for a period of six weeks prior to the analysis.

Results:
The multi-variable analysis quickly revealed that there were four key variables showing two important attributes:
• They are easily controlled by the paper machine operators.
• Allowing for very few exceptions, when certain upper and lower limits are respected, a quality index of 16 or greater is achieved.

Benefits:
The mill applied these rules eradicating the defect and consequently the 30% broke. The exercise highlighted key process parameters that needed to be better controlled. The logging was continued and subsequent analysis enabled tuning of the ranges according to paper grade. 

Contact us to see how Intertek can help your organization with Multivariable Process Analysis.