Editorial: Special Issue Contributed from CHEMECA 2006-Australasian Process Modelling and Control Selected Papers

Abstract

The papers in this issue of Chemical Product and Process Modelling are substantially those that arose from special sessions on``process simulation and control" (organised by Brent R. Young) and``mathematical modeling" (organised by Mark I. Nelson) at the 34th Australasian Chemical Engineering Conference (held between 17-20th September 2006, in Auckland, New Zealand). The papers in this special issue are iss2. The papers featured in this issue have been revised and extended from CHEMECA and rereviewed before publication here.All the papers in this issue use mathematics. However, this special issue only features a small number of the presentations at CHEMECA that use mathematics. Mathematics finds many practical applications within chemical engineering and consequently presentations involving mathematics were featured in many special sessions throughout CHEMECA. Some of these presentations will appear in special issues elsewhere. In particular, the papers from every session that were nominated for the John Brodie award are appearing in a special issue of the AsianPacific Journal of Chemical Engineering. Summary of papers ControlMany chemical engineering processes consist of multi-unit operations in which the output from one operation is an input into a subsequent operation. A variety of techniques are available to control such processes, some of which apply only to steady-state operation whilst others can be used to achieve dynamic control. The distinction between dynamic control and steady-state control is useful in practice because, for example, dynamic models are often unavailable during process design whilst steady-state process models are commonly available.The implementation of any control scheme requires the values of some of the process variables. The required values can often be measured on-line; however the measured values may differ from the `true' values as a consequence of random or systematic measurement errors. Plant data are frequently dynamic and thus a dynamic data reconciliation approach is required to estimate the `true' values of time dependent process variables. Effective processing of plant data to remove noise will lead to improvements in process control. Dynamic data reconciliation is the subject of the paper by Tellez-Schmill et al which uses a process simulator model to transform a constrained optimization problem into an unconstrained problem; the point being that the latter requires less computational effort than the former.Santoso et al (1) investigate the process operability of a high-purity distillation column for methanol-water separation using the framework of steadystate attainability. To assess the static operability of the process it is assumed that the input variables, i.e. the initial conditions of the process model, can be varied over some range in parameter space. This defines a hyper-volume (A) in parameter space. The set of initial conditions within (A) that evolve to the desired steady-state defines a second hyper-volume (B). A new metric, the Output Controllability Inde...