The Reality
On many projects, PID loop tuning is deferred. The project team or operations may prefer to wait until the station is fully in service, running under actual process conditions, before tuning the loops. This is understandable — tuning against real process behavior gives better results than tuning on simulated or unloaded conditions.
The Problem
If PID loops are left untuned (or at factory-default tuning parameters), they will not control well. Overshoot, oscillation, sluggish response, or instability are all possible. If this is not clearly communicated, operations may assume the control system is malfunctioning when it is simply not tuned.
What to Communicate
- Identify which loops are untuned. Provide a list of every PID loop with its current tuning parameters and note which ones are at default or placeholder values.
- Document it as a deficiency. In the project punch list or commissioning report, explicitly state that PID tuning is outstanding and expected control performance will not be achieved until tuning is complete.
- Set expectations with operations. Let them know that oscillation or poor control on specific loops is expected and will be resolved during tuning.
- Define when tuning will occur. Specify the conditions required — process must be running, flows must be at normal rates, etc.
Default Parameters to Set
Even before tuning, set reasonable starting parameters rather than leaving factory defaults:
- Proportional gain conservative enough to avoid instability (start low).
- Integral time long enough to prevent wind-up.
- Derivative at zero unless the process specifically benefits from it.
- Output limits clamped to a safe operating range.
Untuned PID loops are a known deficiency, not an unknown risk — as long as they are documented and communicated.