Multivariate statistical monitoring of batch processes: an industrial casestudy of fermentation supervision

This article describes the development of Multivariate Statistical Process Control (MSPC) procedures for monitoring batch processes and demonstrates i ts application with respect to industrial tylosin biosynthesis. Currently, the main fermentation phase is monitored using univariate statistical proce ss control principles implemented within the G2 real-time expert system pac kage. This development addresses integrating various process stages into a monitoring system and observing interactions among individual variables thr ough the use of multivariate projection methods. The benefits of this appro ach will be discussed from an industrial perspective.