On the use of models with delay in PI(D) autotuning

Many model-based tuning methods for PI(D) controllers, especially - but not exclusively - in the process control domain, employ models with delay. Doing so has advantages, basically in terms of tuning rules´ simplicity, but also potential drawbacks, especially from the standpoint of performance (and sometimes robustness). This manuscript aims at initiating and fostering a critical discussion on the use of models with delay, and shows that in several relevant situations such models can be detrimental to the achieved tuning quality, at least unless some precautions are taken in the way they are parametrised. An example procedure is proposed that realises said precautions, showing the obtained advantages. The scope is here limited to First Order Plus Dead Time (FOPDT) models for brevity, although extensions are possible.