- Tactical Briefs
- Collaborative Manufacturing
- Control Panel Optimization
- Embedded systems & Trends
- Embedded Vision in Manufacturing
- Energy Efficiency
- Ethernet I/O Networking
- Factory Floor Network Deployment
- Factory Floor Network Reliability
- Fieldbus I/O
- Hands-on Guide to OEE
- HMI, From the Web to the Cloud
- Industrial PCs and the IoT
- Internet of Things
- Machine Safety
- Machine Safety Standards & Strategies
- Make a lasting connection
- Mechatronics @ Work: Insight & Technology Solutions
- Opening Up Your Gateway to Asia
- Real-time Operational Intelligence (RtOI)
- Robotics in U.S. Manufacturing
- Robots & Machines in Motion
- The Future of Industrial PCs
- The power of PackML
| November 29, 2012
New Columnist Greg McMillan on PID Control
A unified approach to advanced control can make utilizing the full power of PID less of an art and more of a documented, fast and straightforward procedure. This column, first in a series of “User Views,” will cover the original development of process control technologies, and the scope of the application opportunities.
After 33 years with Monsanto and Solutia, I can’t help thinking like a user. Even as a contractor, the user view pervades my motivations and goals. I would make a lousy sales and marketing person. Early aptitude tests confirmed the last thing I should do is try to run a company or be a manager. The tests said I should be a postal worker or scientist.
In a way, I live the life of a scientist by being in technology using field experience and a virtual plant as my lab to develop and test principles. What gets me excited is when a principle yields a common approach that can address a diverse spectrum of applications that seemingly require unique solutions, custom tuning, and different proportional/integral/derivative (PID) structures. Consultants may feel threatened by making what was special into the routine, until they realize that unifying principles create opportunities to move on to a higher level of control.
I think part of lack of recognition of the value of automation as profession is that PID solutions seem to require an expertise that is in the hands of a few. The PID structure has incredible flexibility, but this is a double-edged sword since there is no documented, methodical, general approach for the effective use of the PID features.
The model predictive control (MPC) is being used as a preferred solution for even PID applications because MPC software automatically identifies and implements process models. Tuning is often relegated to simply tweaking move suppression (how much do you allow each manipulated variable to change for each execution of the MPC). I found this out firsthand when I had designed a pH loop for a refinery scrubber. The control group manager said thank you, but two weeks after startup my PID solution was replaced with an MPC. When asked why, the manager said it was because they did not have any Shinskeys in the plant—a reference to early process control expert Greg Shinskey.
A unified approach using four simple solutions can make utilizing the full power of the PID less of an art and more of a documented, fast and straightforward procedure. This column, a series of “User Views,” will provide a perspective of the original development of the technologies for the solutions, and the scope of the application opportunities.
The first simple PID solution that we’ll look at for this column stops limit cycles, allows violation of the cascade rule, reduces oscillations across the split range point, facilitates optimization, improves equipment and environmental protection, and helps resolve the perpetual debate between academics and practitioners in terms of PID tuning and control loop objectives. Believe it or not, the solution involves technology that is over 90 years old.
By simply turning on a standard PID option created by the enabling technology, you prevent integral action from changing the PID output faster than the final control element, secondary loop, or a setpoint rate limit. If the secondary loop or final control element is not responding, PID reset action simply stops.
If a control valve has not moved due to backlash or stiction, or a variable frequency drive has not changed speed due to deadband or an I/O card resolution limit, the PID output does not continue to try to drive the output to make something happen. The suspension of the integral model prevents the ramping action that creates the limit cycle.
Furthermore, if a valve is truly stuck due to high temperatures, solids or coatings, the PID will not ramp off to its output limit. A fast readback of actual valve position and variable frequency drive (VFD) speed is used so the PID quickly knows whether the final control element is really responding.
Check in next month (Automation World January 2013) for a continuation of this discussion and other PID solutions.
>> Greg McMillan, [email protected], is a retired Senior Fellow from Monsanto-Solutia and an ISA Fellow. Greg received the ISA “Kermit Fischer Environmental” Award for pH control in 1991, and received the ISA Life Achievement Award in 2010. He is also the author of numerous books on process control. His most recent book 101 Tips for a Successful Automation Career was inspired by the ISA mentor program he started at ISA Automation Week 2011.
E-Book Special Report
IT Delivers on Automation’s Promise
Sign up to receive timely updates from the editors at Automation World and download this FREE Special Report on the transformative power of data in manufacturing. By integrating information and automation technologies, manufacturers are finally achieving major gains in productivity from their automated systems.