The availability of historical and real-time data opens the door for the widespread application of multivariate models for process monitoring and optimization. However, model development, deployment and maintenance are time and resource-intensive, which often limits the number of process applications. At Arkema Inc. we developed custom applications in MATLAB based on the PI-SDK to manage all three stages in the model life cycle. The core of the development is a central model server that communicates with individual plant PI databases and continuously analyzes processes. We will present the technical aspects and challenges of the project as well as its financial impact.
Dr. Holger Amort is a Research Scientist in the Systems group of Arkema Inc. in King of Prussia. Holger earned his M.S. and Ph.D. in Physical Chemistry from the University of Hamburg/Germany and has worked in fuel cell R&D and polymer manufacturing with a specialization in process monitoring and optimization.