Doctoral Dissertations

Date of Award

5-1993

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Chemical Engineering

Major Professor

Charles F. Moore

Committee Members

Jim Downs, John Brados, Ken Kirby, John Pilpot

Abstract

In recent years, rising material and energy costs have forced process design toward greater utilization of heat integration and materials recycle. This trend reduces the equipment over-design and availability of surge capacity that have frequently been used to lessen the impact of disturbances and reduce the interaction between unit operations. With this change in design philosophy, control system design must move toward a more global view, recognizing that unit operations seldom function in an "isolated" environment.

The approach to control system design and analysis advocated here attempts to go beyond the traditional concepts. In doing so, the process is analyzed as a whole, instead of a sum of independent parts, and the traditional view of disturbance rejection is discarded in favor of a view that focuses on managing variation. Disturbance rejection is the ability of a control system to eliminate the impact of a disturbance on controlled variables. This type of action transforms the disturbance (variation) into movements in the manipulated variables. From a plant-wide perspective, a control system that demonstrates good local disturbance rejection may actually transform the variation to a more critical location in the plant. An approach to disturbance rejection that focuses on the management of variation can allow the plant to better meet the operating objectives.

This work applies a number of steady-state analysis tools to plant-wide control system design and analysis. Singular value decomposition (SVD), relative disturbance gain (RDG), and relative control gain (RCG) are applied to evaluate the variation transformation properties of various control strategies for a plant with material recycle. The relative importance of variables is incorporated in a manner such that the ability of a control strategy to best meet the overall operating objectives of the plant in the face of disturbances is evaluated. These tools are effectively used to screen our poor control strategies, to reduce the number of strategies that must be given further consideration to a tractable level.

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

Share

COinS