Bringing PID control performance to the highest level is a challenge, but keeping the performance high is a different game. In this blog, you’ll read reasons for degrading PID control performance and the various ways in which our customers take on the challenge of sustaining control performance monitoring. You’ll also read what to expect from the webinar “PID sustained control performance”: how to get to a high control performance monitoring and how to keep the performance up, including some practical cases.
PID controllers: how are PID loops often used in practice?
PID loops are normally used in one of these four settings:
- In automatic mode
- in “pseudo-auto” mode
- In manual mode
- In ‘constant tuning mode’
A PID controller in AUTO mode is how we all like it: it is in automatic (auto) mode, and doing its job. There are no oscillations or alarms caused by the PID loops. Great! When looking at data about the mode of the loop, it will be clear that the loop is in AUTO almost 100% of the time.
There are also cases where the PID loop is in auto mode, and it seems to be working well, but whenever there is a disturbance in the plant, the operators take the PID loop in manual, change the output of the controller and put the loop back in auto again. They “help” the PID loop. When looking at the data about the mode of the loop, the loop will show to be in AUTO close to 100% of the time, but it is actually not doing its job. The operators are actually taking control whenever the PID loop should have managed the situation.
Many PID loops are always kept in manual. The manual mode can be due to a lot of reasons. The most common reasons include bad tuning or a bad control structure (for instance choosing the wrong handle to control a process variable). Instrumentation issues (like sticky valves) can also prevent the loop to be put in AUTO.
‘Constant tuning mode’
This is a situation that is to be avoided at all times: the PID loop is being tuned almost daily. This is often due to varying operational windows, production levels, etc, which cause for instance valves to operate around 50% for a few days, then around 10% for the next days, and then around 90% for a few days, etc, which will cause the PID loop to oscillate due to the non-linear valve characteristics. Everytime an oscillation occurs, the control engineer is asked to retune the loop.
Good control performance monitoring: how to get there?
The main goal of control engineers is, of course, to aim for robust, high performance of the PID controllers in the plant. Tuning is often done at operator request, but that’s not a very structural approach. Quite often there is no real root cause analysis done. The control engineer only checks the loop the operator is complaining about, without considering the other loops that might affect the process parameter that’s being controlled.
Tuning is a multivariable process, so you have to take the other variables into account as well. We also notice that PID tuning is done when time permits.
We see this a lot, but it doesn’t bring a lot of value in the long run. A quick tune-up of a loop at operator request may help you today but won’t bring you many benefits in the future.
Organizations increasingly become aware that there is more value to be realized from the DCS in terms of basic control performance monitoring. What do you need to be able to improve basic control performance monitoring?
- Focussed effort to optimize the base layer controls (either in-house or outsourced, as long as it is a focussed effort)
- A review of not just the tuning, but also the structure of the loops (what is controlling what?)
- A good understanding of the process: knowing what you’re trying to control is crucial for choosing a successful control strategy
Benefits of good control performance monitoring
- Huge stability gains which can give you higher throughput and less energy consumption, for example.
- Higher percentage of time that PID loops are in AUTO (which will make your management happy as well).
- Happy operators because with more stability in the plant it’s easier for them to run the process
- A significant reduction in the number of alarms, helping your alarm KPI’s to improve drastically
On demamd webinar sustained PID controller performance: how to stay there?
he biggest challenge is keeping the PID control performance high. There are some best practices to improve your performance, such as knowing the limitations of PID control, always separating safety and control functions, and keeping it simple. Many more best practices are covered in the webinar.
Also, questions like “where to start”, and “what’s the role of control performance monitoring software” are addressed.
Do you want to learn from use cases and implement some of the lessons learnt in your own plant?
Watch the on demand webinar: ‘Sustained PID controller performance’ and learn how to keep your performance high based on some best practices from real-life examples.