- 8:30 AM

Smart Magement Strategies for Smart Manufacturing Plants

Jeremiah O'Brien, Process Data Control Corp, 1803-A W. Park Row Dr., Arlington, TX 76013 and Jimmy L. Humphrey, J. L. Humphrey & Associates, 3605 Needles Dr., Austin, TX 78746.

This paper will review management, innovation evaluation, and implementation aspects of Smart Plants as elucidated by a "Smart Distillation" survey performed by Dr. Jimmy Humphrey in the fall of 2007. It asked the following questions: Suppose all key distillation columns within a company were equipped so that accurate real time material and heat balances could be determined around each column, and further suppose that data may be shared locally and globally with the enterprise in real time -- what advantages do you see, and what are the challenges?

The main "advantages" that respondents to this survey identified were: (a) Continuously operate columns at tighter product specifications to reduce off-spec products and increase energy efficiency; (b) Monitor, diagnose, and troubleshoot distillation columns (also heat exchangers, furnaces, reactors, etc.) via broadband using centrally located, highly skilled technical staffs; (c) Constantly compare performance of similar columns (globally) to identify and improve those that are underperforming; (d) Perform process diagnostics on individual columns and compare actual with theoretical performance; and (e) Coordinate and schedule production from information that is available globally to maximize overall profits.

The "key challenges" that respondents to the survey identified were: (a) Is there such a thing as an accurate material and heat balance ? Are compositions and flow rates based on inaccurate measurements? Can compositions be measured in real time? Distillation does not always operate as a steady state process; (b) More emphasis needs to be placed on applying wired and wireless sensors as well as soft sensors. Getting accurate real time measurements, and applying accurate predictive models are critical to taking proactive actions to prevent plant problems before they occur; (c) How do we handle the volumes of data that will be generated? These data may cause confusion and delays without considerable supporting software and graphics. Can we develop intelligent data analyses to quickly process the amount of data generated and pinpoint the problems?; (d) How can benefits outweigh the costs for small plants?; and (e) Will we have enough people with the expertise to intelligently mine the data?

Both the advantages and challenges will be assessed in this paper using an evaluation methodology consistent with Smart Plant design and operating concepts.