Better Operating Targets and Alarms for Multi-mode and Multi-Grade Processes: Improve Quality and Reduce Waste

Most of the examples we use in our webinars are single-mode continuous processes. All processes, even those that appear as steady-state continuous, are more complex. There are separate operating modes for startup, shutdown, catalyst regeneration, reduced rate running, grade transitions, as well as alternate equipment configurations. For batch processes a product recipe defines the start and end-conditions of phases. These phases map onto modes in Geometric Process Control (GPC).  There are benefits in extending process knowledge by exploring, visualizing, and modeling the operating envelopes of modes (or phases) visually, individually as well as in aggregate, to discover variable and quality interactions.

GPC allows engineers to combine their process knowledge with data already captured in plant historians, LIMS, electronic batch records and KPI history to visualize, manipulate, and understand process operating envelopes spanning many modes and hundreds to thousands of process variables in a purely visual environment. It requires no explicit mathematical knowledge allowing users to query and discover relationships faster and creates real-time geometric models of each of the modes without requiring further user involvement other than configuring OPC connections to DCSs, PLCs and historians.

In this webinar we describe and demonstrate how our GPC technology naturally handles multi-mode and transient processes, including startup, grade-transition, and batch processes. Dividing process data into individual operating envelopes corresponding to each mode of operation is easily done, and the speed of analysis makes handling each mode individually efficient. We’ll also look at explicit time-dependence, whether it is based on time in transition between grades, catalyst use, batch time or phase time.

We’ll look at case studies demonstrating alarm rationalization, process operating limits, and dynamic operator feedback.