When the grass root plant starts production, 70% to 80% of the PID loops are in default mode. There is nothing wrong with default settings. However, when time passes and alarms occur, the production manager will complain about spending too much time on controlling the plant. You need to check all the PID loops to make sure the plant will run smoothly. In this blog, you will read 5 of the most common mistakes made by DCS engineers in PID tuning. And more importantly, how you can avoid those mistakes.
Mistake 1: Not gaining process knowledge
When encountering a bad performing PID controller, the first thing you need to do is investigating and identifying the problem. This can be challenging when you don’t fully understand the process of the plant. To set the right KPIs for the control objectives of your PID loop, you need to have process knowledge. So, the first challenge is to gain this knowledge to be able to react accordingly. Even with a little more knowledge, you can significantly improve the stability of the plant.
Mistake 2: Only check the PID loops when there is an alarm
Everything is working fine at the moment, so you can sit back and relax until an alarm occurs or until there is a complaint from your operator. As you can imagine, that isn’t always a very good tactic to keep the plant running optimally.
Of course, it’s impossible to monitor the performance every morning if you have 1000 different PID loops. However, it can be wise to start with investigating the critical loops that affect other PID loops in your plant. You can identify the critical loops with a PID monitoring software or running an analysis in an historian or even in Excel by checking the history of the loop. This way your operator doesn’t have to put your PID loops in manual and you make sure the plant performance is optimal.
Mistake 3: Ignore instrument ranges (and units)
Instrument ranges are very important in PID tuning. If these instrument ranges have changed - for example, due to maintenance or the replacement of the instrument - the performance of the PID and therefore the plant behavior will be affected as well. An important follow-up task is to check these instrument ranges to avoid oscillations or even unstable behavior.
Mistake 4: Using the default PID equations
Only about 30% of all DCS engineers know that there are more than the three equations in a PID loop. The other 70% often don’t bother to change the default control equation of the loop. That’s unfortunate because alternative equations can allow a soft landing in your output when this is needed.
Almost every DCS has multiple equations and most of the act on the error. When configuring a PID loop, you can use the default algorithm which states that the Proportional, Integral, and Derivative act on the error.
However, there are also many other algorithms you can use, which might suit the loop better. For example, using equation B or C in e.g. Honeywell. When you use equation C, the Proportional and Derivative will act on the process value (PV), not on the error. So, if you change the setpoint of the PID loop, no output kick will happen. The output of the controller will move less aggressive to the new value. For some PID loops, having an aggressive output behavior doesn’t cause any problems but for some overshoot or undershoot is often dangerous.
Mistake 5: Individually tune interacting PID loops
Many loops in your plant aren’t independent and interact with each other. You can tune interacting loops by imposing the control objectives of disturbance rejection. Whenever an alarm occurs it’s often due to a setpoint change, but it can also be caused by a disturbance in another loop. When you have tuned an interacting loop individually and set it to automatic control the behavior of the loop will not be the same as when you tuned it individually. That’s because the interaction between the other loops is not taken into consideration during the tuning. If you are aware of the interactions, tuning will be more efficient.
What now?
Knowing the 5 common mistakes allows you to improve plant stability. It even can contribute to saving 10% of the plant’s energy. But how to start with these improvements tomorrow? Here are some best practices to optimize your plant performance with PID tuning :
- Identify all the critical loops in your plant, and tune them first.
- Always check the PID loops in the front end of your plant when there is an alarm
- Follow a systematic approach in tuning your PID.
To achieve optimal plant performance, it’s essential to prevent the PID loops to oscillate or have undesired overshoot. Therefore, it's essential to understand your PID tuning process.
Download the ebook: 'The Guide to PID tuning' here
Do you like to know what INCATools PID tuning software can offer you?
Request your demo here.
