How Do I Conduct a Successful Bump Test?

Bumping a Process and Tuning Your PID Controller: Best Practices

When tuning a PID controller, whether manually or with the help of software, bumping a process is a basic need. In any case, following industry best practises will save you a lot of time and money.

The bump test is crucial because it establishes the cause-and-effect relationship between the Process Variable (PV) and Controller Output in a control loop (CO). It's when you introduce a change in CO to watch how the PV reacts. Model parameters can then be generated and turned into tuning parameters for use with the loop's PID controller once that relationship has been established. While this may appear to be a simple task, there are various factors to consider:

• Keep It Quiet - Disturbances cause the process data to become jumbled. It will be nearly impossible to distinguish between what should be ascribed to your bump and what should be assigned to the disturbance when examining the findings of your bump test. Maintain a smooth and silent process as much as feasible.
• Mind Your DLO — As previously stated, the control loop spends the majority of its time at the Design Level of Operation (DLO). It's a range that corresponds to typical operation and should be taken into account while running your test. Testing outside of the DLO will only impede the PID's capacity to maintain effective and efficient control.
• Keep your CO sharp - Whether you do a simple step, a complete bump, a doublet, or even a PRBS, the shift in your CO should be noticeable. It may be difficult to understand how the change influences the PV's reaction if the CO is slowly ramped from one value to another. It's also vital that the adjustment eliminates any deadband or stiction that may have occurred during the procedure (i.e. valve or damper).
• Be Aware of Noise - There is some level of noise in almost every procedure. Make sure the procedure is outside of the noise band so you can see the PV's response clearly. A bump test that is at least 4-times the value of the related noise is a good rule of thumb.
• Allow the CO-PV relationship to show — The aim of the bump test is to understand the CO-PV relationship, therefore conduct your test properly. It is not always necessary to hold the PV steady at a new value, but a clearer response will result in a more accurate process model and better management.

These straightforward rules will help you avoid bump tests that yield unsatisfactory results and/or necessitate extra testing. It should go without saying, but when completing a bump test, always keep your own and your coworkers' safety in mind.

Consider attending a training course that focuses on understanding process dynamics and the PID controller, or contact us if you need more help on how to execute a test or how to tune your PIDs for best performance.