Thesis for the Degree of Licentiate of Technology


Year 1995
Anne Kytömäki

Statistical Process Control Methods and Applications in Process Control

ABSTRACT-The scope of this work was to examine the applications of the statistical methods in process industry, in process analysing, control and modelling, and to implement Statistical Process Control (SPC) methods to support the control of the polypropylene process. The methods were applied in the polypropylene production in Borealis Polymers Oy, and in the formaldehyde and resin productions in Neste Resins.

In the theoretical part of the thesis, quality improvement methods, the control of the process variation, the methods of the statistical process control and the methods of the design of the process experiments were examined.

In the experimental part, statistical methods were applied in analysing and modelling of the processes of the polypropylene, formaldehyde and resins. In the polypropylene production, the control chart method was taken into use. The first object for the method was melt flow rate, and the second object was additives. The purpose was to discover the abnormal variations, find the reasons for these variations, and to react to those improving the quality of the product. The methods of the chemometrics were used to examine the important process parameters which effect to the melt flow rate of the product, and to model the melt flow rate. In the resin production, the variations of the product characteristics were examined, and in the formaldehyde production, the methanol content of the formaldehyde was modelled with chemometric methods.

It was foud out, that statistical methods are good tools to examine the distribution and the capability of the process, and to observe the improvements. Graphics and process capability indexes are good ways to describe the state and the capability of the process. The education, motivation and commitment of the plant personnel and the management were observed to be very important. The chemometric modelling results were as good as traditional black-box modelling results.

In future, statistical process control and automatic process control should be used so, that SPC would be utilized for controlling the long term variation and APC for controlling the short term variation. Also process experiments would gave new process knowledge.

This info last modified 23 Sep 2017 - contact webmaster