Users Conference 2003
Learning to Balance - Baby Steps
Our plant is part of a large global chemical manufacturing company. Our performance is being compared to sister plants in Europe. This "competition" makes it necessary to provide accurate data to our ERP system and find ways to continually improve performance. A year ago it was impossible to accurately measure yields or determine where losses were occurring in most of our processes. Questionable measurements applied to strapping charts of unknown origin led to inaccuracy. Due to these problems among others, no matter where the loss occurred the production unit ate the loss. But even more importantly it was not possible to pinpoint the loss so that it could be reduced or eliminated. With the addition of instrumentation, PI automatic reporting, PI Totalizer, and PI ACE tank mass calculation much progress has been made. KPI’s that merge SAP information with process information from the PI System are being developed that would not have been possible in the past. This is just the beginning. The future includes plans for Sigmafine and possibly RLINK. This presentation provides our way of preparing for those applications.
Company
ATOFINA Chemicals, Inc.
Speaker
Dwight Stoffel
Dwight leads the Electrical, Instrumentation and Controls group at the corporation’s Calvert City, Kentucky plant. He has been employed by the plant for 25 years in Electrical/Instrumentation Project Engineering, Process Development and Operations roles. During that time, the plant has converted from complete pneumatic to complete DCS (distributed control system) control. The plant consists of both continuous and batch processes used in the manufacture of refrigerants, KynarÓ PVDF, and HCl. In the last six years he has concentrated on the installation and use of a plant process data historian. This project has provided plant data to all desktops at the plant, and some remote locations, providing easy access to the information necessary to improve plant operation and reduce operating costs. It has also provided data for the design of new projects. Statistical and AI software is being used to develop and perform run-time prediction of product and process values. Process history combined with current data has proven to be a powerful tool. He was previously employed by Reliance Electric Company as a Drives Control Engineer for four years. He received a BS in Electrical Engineering from Purdue University.
Back to Users Conference 2003 Presentations