The PID tuning basics
The basics of PID tuning is the workflow to determine parameters
A PID controller is a generic, feedback-based control technique considered as the workhorse in the automation world: 99% of all automation control loops are PIDs.
As a DCS engineer, it’s important to understand the workflow to determine PID parameters. Knowing in advance which steps to take and why to take them is necessary before jumping into the different PID tuning methods available. You’ll be able to get the most out of your PID tuning and ensure that you reach your desired plant performance. Explore te PID tuning basics.
What does PID tuning mean?
PID tuning refers to the workflow whereby PID parameters are determined based on the PID algorithm used, the open-loop process behavior, your control objectives and the engineering specifications of the desired closed-loop behavior.
Optimal tuning is finding the ideal set of P, I and D parameters efficiently. Within the large space of the search parameters, there is only one set that results in an optimal performance. Depending on the PID methods you use, the computed parameters can be closer or further away from this optimal set.
Process behavior is key! Optimal tuning can only be performed when you grab the correct process behavior. Most methods reduce the process to a very simple first-order behavior, although there are many processes that can not be described with these simple models. For example, steam boiler drums, superheaters, processes with long delays, (fed-)batch reactors and level controllers.
Finally, you shouldn’t forget the engineering specifications of the desired process behavior. When setpoint tracking is key, make that your criterion for optimality. When disturbance rejection is key, focus on that. When both are needed, exploit using feedforward.
Tune a controller to avoid overshoot if overshoot is not allowed, or tune a controller to operate in a very broad operating range if that is exactly what happens at the real plant.
In other words, PID tuning means that your control loop has a specific goal which you achieve by using the right P, I and D parameters. As a result, you’ll accomplish the best plant optimization.
Why PID tuning?
You now know that PID tuning is. PID tuning is necessary to have closed-loop control. When you want to, for example, control temperature, a PID controller needs to be tuned to keep the temperature at the setpoint value.
- The minimum requirement for tuning is that the controller can operate in a stable way in a closed-loop.
- When you go one step further PID tuning will stabilize the plant, by reducing oscillations. This will result in, for example, fewer alarms and fewer operation interventions.
- One step beyond brings you to the point that because of optimal tuning the plant efficiency can be boosted, the specific energy consumption can be reduced, more stability can be provided and a minimum amount of alarms and operator interventions.
The key points of PID tuning
PID tuning can result in many benefits. Paybacks of weeks, hours, or even minutes can be achieved when you take the time to adjust the PID parameters to the optimal value. Follow these key steps to tune your PID right the first time:
- Understand the behavior of the process you want to control, for instance by using (multiple) step testing.
- Use the right system ID methods.
- Check your DCS PID algorithm and change it if appropriate
- Make your own engineering wishlist in which you choose between optimal tracking and/or optimal disturbance rejection. Don’t forget the robustness and high-frequency gain.
- Compute optimal parameters, implement them and check the result on the real plant.
- Track and document the changes that you’ve made.