Hi,

I am hoping for some more detailed advice for control loop settings in District Energy System models in Modelica. I have followed the guidance given with the PID controller documentation (https://simulationresearch.lbl.gov/mo...) but I need more help.

At the broadest level I need guidance on how to think about the Ti parameter for all the different elements of my DES model. I have many elements with control loops (timeseries loads, susbstation heat exchangers, pumps, plant devices like boilers and chillers) and all of them are controlled with PI controllers.

My timeseries loads change every T=600s so I thought maybe I should use that as the Ti term for all the parts of the system. But I have studied the other controllers and they definitely do not react every T=600s, its sometimes more often and sometimes less often.

Should I use the same Ti for all the controllers? If not how do I think about "tuning" a model with so many different control loops?

The control loop performance makes a huge difference in the feasibility of running my model. I have 150 buildings (timeseries loads) in my DES and if any of element of the system slows down because of controller error it significantly degrades the ability to even run the model. I want to run an 8760 hr simulation but right now I am looking at about a three day run time which is unacceptable for debugging purposes.

Thanks! Graham