Paul Nomikos

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Chemical Engineering


Professor John F. MacGregor


Multivariate statistical procedures for the analysis and monitoring of batch and semi-batch processes are developed. The only information needed to exploit the procedures is a historical database of past successful batches. Projection methods based on principal component analysis and partial least squares are utilized to compress the information contained in the multivariate trajectory data and final product qualities, by projecting them onto low dimensional spaces. When additional information about the initial conditions and set-up of the batch process is available, multi-block approaches can be used to integrate the additional data into the proposed schemes.

The proposed methodology facilitates the analysis of operational and quality control problems in past batches, and allows for the development of simple multivariate statistical process control charts for on-line monitoring of the progress of new batches. Control limits for the proposed charts are developed using information from the historical reference distribution of past successful batches. The approach is capable of detecting subtle changes in the batch operation, and provides procedures for diagnosing assignable causes for the occurrence of observable upsets. The method's potential in analyzing past batches and tracking the progress of new batch runs, is illustrated through a simulation example and data collected from industrial polymerization reactors.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."