Roche, Francis William; (1998) On-line fault detection in fermentation development facilities. Doctoral thesis (Ph.D.), University College London (United Kingdom).

Text On-line_fault_detection_in_fer.pdf Download (7MB)

Abstract

This thesis reports the results of an investigation into the means which can be used to automate the supervision of a fermenter typical of the type used in antibiotic fermentation process development. The work uses detailed data sets obtained from experimental fermentations as a basis to evaluate possible on line supervision techniques. The basis of this work is the use of multiple data sources to determine whether a fermentation is proceeding as planned. These data can be tested for internal consistency, and compared with data from previous fermentations to generate an automatic method of detecting faults. The use of an on line Kalman Filter as an observer to supply information to generate analytical redundancies was investigated. As part of this work, an analysis is presented of the accuracy and precision of information available from a typical fed batch fermentation. The report concludes with a recommendation as to which methodologies are most appropriate for monitoring large numbers of small scale development fermentations, and a review of potential areas for further work, including AI and Multivariate Statistical analysis is included.

Download activity - last month

Export as CSVJSONXML

Download activity - last 12 months

Export as CSVJSONXML

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item