Multivariate statistical process evaluation and monitoring for complex chemical processes

[摘要] ENGLISH ABSTRACT: In this study, the development of an innovative fully integrated process monitoringmethodology is presented for a complex chemical facility, originating atthe coal feed from different mines up to the processing of the coal to produceraw gas at the gasification plant. The methodology developed is real-time,visual, detect deviations from expected performance across the whole valuechain, and also provide for the integration and standardisation of data from anumber of different data sources and formats.Real time coal quality analyses from an XRF analyser are summarised andintegrated with various data sources from the Coal Supply Facility to provideinformation on the coal quality of each mine. In addition, simulation modelsare developed to generate information on the coal quality of each heap and thequality of the reclaimed coal sent to gasification.A real-time multivariate process monitoring approach for the Coal GasificationFacility is presented. This includes a novel approach utilising GeneralisedOrthogonal Procrustes Analysis to find the optimal units and time period toemploy as a reference set. Principal Component Analysis (PCA) and CanonicalVariate Analysis (CVA) theory and biplots are evaluated and extended forthe real-time monitoring of the plant.A new approach to process deviation monitoring on many variables is presentedbased on the confidence ( ) value at a specified T2-value. This methodologyis proposed as a general data driven performance index as it is objective,and very little prior knowledge of the system is required.A new multivariate gasifier performance index (GPI) is developed, whichintegrates subject matter knowledge with a data driven approach for real timeperformance monitoring. Various software modules are developed which wererequired for the implementation of the real time multivariate process monitoringmethodology, which is made operational and distributed to the clientson an interactive web interface. The methodology has been trademarked bySasol as the MSPEM™ Technology Package. Following the success of thedeveloped methodology, the MSPEM™ package has been rolled out to manymore business units within the Sasol Group.In conclusion, this study presents the development and implementationof the MSPEM™ application for a real-time, integrated and standardisedapproach to multivariate process monitoring of the Sasol Synfuels Coal ValueChain and Gasification Facility. In summary, the following novel developmentswere introduced:• The application of distance measures other than Euclidean measures areintroduced for space filling designs for computer experiments in mixturevariables.• An approach utilising Generalised Orthogonal Procrustes Analysis tospecify the optimal units and time period to employ as a reference set isdeveloped.• An approach to process deviation monitoring on many variables is presentedbased on the confidence ( ) value at a specified T2-value.• An integrated approach to a reactor performance index is developed andillustrated. • A comprehensive software infrastructure is developed and implemented

[发布机构] Stellenbosch University