Circular parameter assignment in Buildings.Fluid.HeatPumps.Carnot_TCon

[mwetter]

This issue is to correct the circular parameter assignment reported in lbl-srg/modelica-buildings#3226

[FWuellhorst]

From my side, there is another issue with regard to use_eta_Carnot_nominal:

If false, etaCarnot_nominal is used to calculate COP_nominal, but disabled in the annotation. This is misleading.
Two options:

1. Set COP_nominal to some fixed value, same as etaCarnot_nominal.
2. Enable etaCarnot_nominal at any time to calculate COP_nominal. However, this would make the whole point of COP_nominal somehow redundant.
   I would prefer (1), as you either have a certain COP at nominal conditions at hand or not.

However, I don't know if this could introduce a non-backwards compatibility, as the default values change.

[mwetter]

@FWuellhorst : Thanks for the suggestions. In the current implementation, I did not set COP_nominal to a fixed value as the value is highly depend on the temperatures. Rather, if use_eta_Carnot_nominal = false, the parameter COP_nominal is ignored (not used) because the model uses always the protected parameter etaCarnot_nominal_internal which is calculated as

  parameter Real etaCarnot_nominal_internal(unit="1")=
    if use_eta_Carnot_nominal
      then etaCarnot_nominal
      else COP_nominal/
           (TUseAct_nominal / (TCon_nominal + TAppCon_nominal - (TEva_nominal - TAppEva_nominal)))
    "Carnot effectiveness (=COP/COP_Carnot) used to compute COP";

Does this address your concern?

If so, I will see what default parameters can be removed in the examples to avoid duplicate bindings.

PS: I also tried to use the fixed attribute on eta_Carnot_nominal and COP_nominal but this then does not allow to provide a default value eta_Carnot_nominal = 0.3 in BaseClasses.Carnot which I think is useful to have.

[FWuellhorst]

@mwetter: A fixed value of COP_nominal is not possible, I agree. If use_eta_Carnot_nominal=false, I would expect the value is not used in the model. However, it is then used to calculate the COP_nominal.
I would go for this option, just leaving it empty:

If users then know the COP, they can insert it.

Do unspecified parameters raise warnings in the latest Dymola and OpenModelica? If so, leaving the default equation is indeed best. But I would then leave the etaCarnot_nominal enabled, as it influences the COP_nominal.

[mwetter]

The GUI above would require to set fixed=not etaCarnot_nominal and then provide an initial equation

  if use_eta_Carnot_nominal then
    COP_nominal = etaCarnot_nominal *TUseAct_nominal / (TCon_nominal + TAppCon_nominal - (TEva_nominal-TAppEva_nominal));
  end if;

However, with such an implementation, if a user already set a value for COP_nominal, then this would be an error. So, I think we need to keep the default assignment

parameter Real COP_nominal(unit="1") = etaCarnot_nominal*TUseAct_nominal/
    (TCon_nominal+TAppCon_nominal - (TEva_nominal-TAppEva_nominal))

and clarify this in the info section which I did in 6225c93. Is this clear or do you have another resolution?

[FWuellhorst]

Ok, then it's clear to me. Thanks for the documentation update.