Use of dynamic modelling and plant historian data for improved control design

Problems associated with process control equipment, such as valves, can be very subtle and difficult to detect. However, the impact that these problems can have on the performance of a process control system can be significant. This paper explores the use of dynamic, mechanistic modelling as a tool for detecting problems associated with process control equipment and as a technique for validating that corrective action which will solve these problems before such action is applied to the plant. The paper describes a case-study application at an INEOS ChlorVinyls site in Runcorn. In this case study, a dynamic model was able to identify the cause of a serious plant problem and determine the necessary re-configuration of the plant control systems, such that plant operations were improved. © 2009 Elsevier Ltd. All rights reserved.