Advanced Analytics to Discover the Root Cause of Multi-Variate Control Issues

Abstract for 2017 AIChE Spring Meeting – San Antonio, TX

Presenter: Steve Obermann

Advanced Analytics to Discover the Root Cause of Multi-variate Control Issues

Process plants are complex and interactive dynamic environments. With process side streams, recycle, and heat integration, it can be difficult and time-consuming to hunt down the root cause of process upsets. Furthermore, reduced plant staffing levels leave very few individuals with the combination of technical depth and wide-angle process view required to solve such problems.

This paper describes modern tools and techniques for solving large and complex multi-variate control issues. A Big Data approach, using real-time data from process control systems, is shown to reduce the time necessary to find the root cause of issues. Specific tools to be discussed include oscillation detection and analysis, interaction mapping, and pattern recognition. This paper will show how such tools and techniques can help plants with limited staffing to quickly resolve issues and return to stable, optimal operation.