Application of Dynamic Modelling for Brownfield Process Simulation

High fidelity commercial process simulation packages can be used at different stages of a greenfield project, from initial plant design to operator training in the later stages. However, there are several factors that make the software less attractive for small and medium size brownfield projects, such as the licensing cost and the fact-finding manpower required for model design and simulator maintenance. In this paper a different approach is described. The knowledge gathered from operating personnel was used to implement higher level models based on plant data, operating heuristics, and experience rather than detailed component-level models. The computing platform used was a low-cost, well-established, control system vendor-independent software. Throughout this paper we demonstrate the success of implementing this model-based approach using two industrial examples. The first example is a simulation of the carbon monoxide (CO) distribution system at a metallurgical complex. The purpose of this simulator was to test new control strategies and PLC programs prior to implementation. CO being a toxic combustible gas, validation of the suggested control system changes was necessary for plant safety. Simulation of specific operating conditions helped alleviate concerns and improved plant-wide acceptability of the new operating philosophy. The second example is a boiler simulator at a petrochemical plant. The operator training simulator (OTS) was designed as a platform for experienced operators to practice handling rare or abnormal conditions. A customized medium-fidelity model was developed for this targeted application where the cost of a full scale high-fidelity model could not be justified.