443293 Big Data Analytics Applied to Real-World Control Systems - from Instruments to Advanced Controls

Wednesday, April 13, 2016: 4:36 PM
335A (Hilton Americas - Houston)
George Buckbee, Metso ExperTune, Waukesha, WI

Big Data Analytics Applied to Real-World Control Systems - From Instruments to Advanced Controls

A modern control system at a chemical plant or refinery can gather gigabytes of data, every day.  Savvy engineers have applied software tools and methodologies to extract immediate value.  This presentation shows how users are applying Big Data Analytics to improve the performance of all aspects of control systems, from instrumentation to advanced controls.

Manufacturing control systems have been a source for big data since the late 1980's.  Digital control systems started to capture data, and feed it to process data historians.  Initially, users were selective, and only captured the most important process measurements, such as temperatures, flows, and pressures.  As computer memory, storage, and network speeds increased capacity, users were quickly storing thousands of pieces of information about the process, at faster and faster sampling rates.  Today, in a modern refinery or a large chemical plant, it is common to collect tens of thousands of process measurements, valve positions, and control system statuses, at sampling rates as fast as once per second. It is even possible to store all of that data without loss of information in data compression.

Analyzing this flood of data to find the most important information has been a challenge to engineers for several decades.  This presentation shows new tools and techniques that have demonstrated success in finding, prioritizing, and resolving major issues in process plants.  Specific techniques and examples will be shown for:
* Identification of instrument failures
* Finding and quantifying mechanical issues with control valves
* Measuring the effectiveness of the control system
* Identifying control strategies that are no longer working properly
* Measuring the performance of Advanced Process Control

The author's analysis has shown that most plant control systems are operating far below their potential.  Furthermore, using Big Data Analytics, it is possible to drive significant, targeted improvements in a very short time.  Attendees will also learn how to estimate the potential economic benefit potential at their plant site.

Extended Abstract: File Uploaded