A startup procedure for process industries using a multiple objective nonlinear program

Abstract

We develop a multiple objective nonlinear program to determine optimal set points for the startup period of a continuous mix manufacturing process. We model the objectives and constraints dictated by the problem formulation and solve the model. We discuss why and how process adjustments are made at the startup of operations. To advance the model, we develop a startup adjustment methodology. This methodology guides the process for startup conditions and real-time operation of the system. It is demonstrated on data from a breakfast cereal manufacturing process where the startup process under conventional approaches spans several days. We close the paper with discussions on process adjustments after steady state has been achieved and strategic issues in process regulation.