One of Greg McMillan‘s favorite sayings in his illustrious process automation career was something to the effect of, “If it weren’t for deadtime, I wouldn’t have a job.” Just do a search on “Greg McMillan” “deadtime”, and you’ll see how much of his expertise he’s shared on this subject over the years.
I mention this because his and Peter Morgan’s recent ControlGlobal.com article is titled Deadtime compensation opportunities and realities. The article explains the:
…benefits of implementing controller integral action by using filtered, positive feedback. … In this article, we discuss how adding deadtime to the reset feedback signal improves loop performance when there’s deadtime in the process or when the process response is characterized by an equivalent deadtime and lag.
…benefits of implementing controller integral action by using filtered, positive feedback.
…
In this article, we discuss how adding deadtime to the reset feedback signal improves loop performance when there’s deadtime in the process or when the process response is characterized by an equivalent deadtime and lag.
They highlight common misconceptions that include:
…the idea that deadtime compensation eliminates the effect of deadtime, disturbances are on the process output/process variable, the Smith Predictor is the solution, deadtime compensation is best reserved for deadtime dominant processes, and performance improvement requires turning on deadtime compensation.
The article highlights the issues and opportunities with deadtime compensation and Smith Predictors as they apply to different process dynamics–unmeasured load disturbances, lag-dominant processes, integrating processes, deadtime dominant processes, etc.
ISA standard form PID with positive feedback integral mode enabling external reset feedback (dynamic reset limit).
They cite studies by Greg Shinskey on deadtime compensation using external reset feedback (ERFB). The test results:
…show deadtime compensation via ERFB can improve performance for a wide spectrum of processes, and reveals the tuning adjustments needed and robustness observed. Surprisingly, the ability to handle the inevitable changes in process gain is increased.
Read the article for how the controller was configured for these tests, the process models used including self-regulating and integrating processes, stability margins, adding derivative action, and the authors’ conclusions.
Also, make sure to check out these recent posts where Greg makes three of his numerous books available for free as PDF files. These posts include:
The post If it Weren’t for Deadtime… appeared first on the Emerson Automation Experts blog.