Monitoring a PVC Batch Process with Multivariate Statistical Process Control Charts

Abstract

Multivariate statistical process control charts (MSPC charts) are developed for the industrial batch production process of poly(vinyl chloride) (PVC). With these MSPC charts different types of abnormal batch behavior were detected on-line. With batch contribution plots, the probable causes of these abnormalities were located. Examples are given for two different types of abnormalities, i.e., a bias in the batch loading and a control valve problem during polymerization. These examples show that MSPC charts in combination with batch contribution plots are a simple and powerful method for fault detection and identification of their cause in the operation and control of batch processes.