Optimization of process operation strategies by combining process models with plant operating data

Competition, cost pressure, and market fluctuations lead to a persistent increase in complexity and degree of integration of chemical plants. The utilization of accurate process models facilitate plant efficiency optimization in real-time and improve process transparency for the plant personnel. To reduce development time and costs, existing models from the process development phase of the chemical plant can be used. Next to the general availability of a plant model and plant operating data, a systematic strategy is helpful to successfully implement a validated process model into the process control system of a chemical plant. This contribution presents a strategy to develop a technology platform with a validated process model containing the following steps: analysis of the economic potential, software selection, process analysis, steady-state detection, parameter identification, and process optimization. The presented strategy is applied to an industrial plant of BASF SE in Ludwigshafen, Germany. The application of this platform exposes great economic potential. Firstly, a significant cost reduction can be achieved by reusing existing models during the development phase. Secondly, with the help of the technology platform soft-sensors are created, bottlenecks identified, and an optimization of process operating strategies is undertaken.