Ask Greg McMillan - What role do you see dynamic simulation playing in the achieving the best automation system dynamics for flow, pressure, and surge control?

Greg's Response:

In flow, pressure and surge control loops, the greatest sources of deadtime is in the automation system. Often for flow control, the speed of the response is not an issue except in suction flow control for surge control and when the primary process loop is fast requiring a fast secondary loop to avoid violation of the cascade rule. For steam headers, it was realized the pressure controllers needed to directly throttle fast precise linear control valves. For some gas and particularly liquid pressure control it was recognized that the controllers must directly manipulate variable frequency drives (VFDs) because control valves could not be made fast enough.

The 5Rs (resolution, repeatability, response time, rangeability, and resolution) are the key to better performance. For flow measurements, mass flow meters offer order of magnitude better 5Rs as discussed in the Control Talk Column “Knowing the best is the best”. Transmitter damping settings must be minimized. Note that the beneficial effects of faster and more sensitive measurements may cause trend recordings to appear to have more oscillations because noise and excursions are less filtered. For control valves, achieving the best 5Rs is detailed in the Control Feature article “How to specify valves and positioners that do not compromise control”. The pervasive rule for over 50 years that boosters instead of positioners should be used on fast loop is potentially dangerous resulting in butterfly valves slamming shut due to positive feedback from a booster’s high output port sensitivity, Boosters with a slightly open bypass valve should be added to positioner output and the positioner tuned for high gain proportional-only control. The controller should have an execution rate that is 0.1 seconds or faster. There are some special fast digital controllers with an execution rate of 0.05 seconds. There are some analog control holdouts I discovered for polymer and furnace pressure controllers manipulating VFDs. Note that VFDs may not have good 5Rs, which will be the subject of my next post.

For much more knowledge see “Pressure Control: Without Deadtime, I Might Be Out of a Job” and “Compressor Surge Control: Traveling in the Fast Lane” in my book Process Control Case Histories and Chapters on Effect of Controller, Measurement, and Valve and VFD Dynamics in my book Tuning and Control Loop Performance Fourth Edition.