+- VedaSupport Forum (https://forum.kanors-emr.org)
+-- Forum: VEDA 2.0 (https://forum.kanors-emr.org/forumdisplay.php?fid=19)
+--- Forum: Questions and Observations (https://forum.kanors-emr.org/forumdisplay.php?fid=20)
+--- Thread: Dummy import for slack variables (/showthread.php?tid=1098)
Dummy import for slack variables - Mahmoud - 25-10-2021
Dear all, I am facing a problem deactivating dummy imports flows into the slack variables of a UC. Here's a screenshot of the .DD file that contains this information. Is it normal that these slacks are still there in the ouputs by 2050? Thanks a lot for your help Mahmoud
RE: Dummy import for slack variables - Antti-L - 25-10-2021
> Is it normal that these slacks are still there in the outputs by 2050? Hmm... the dummy import flows are not slack variables. But the UC constraints do always have slack variables, VAR_UCxxx. And yes, it is normal that these slacks appear in the results, either with non-zero levels or marginals. If you mean that the dummy flows are still there in the outputs by 2050, I am not sure what you mean by "in the outputs". But if you mean that the flows have positive values in the results, even though you have bounded them to zero, that would certainly not be normal. Defining FLO_BND(r,t,'IMPNRGZ',com,'ANNUAL','UP') = EPS would generate equations forcing the IMPNRGZ flows of com to zero.
RE: Dummy import for slack variables - Mahmoud - 25-10-2021
I meant that I have defined FLO_BND(r, t, 'IMPNRGZ', 'com','ANNUAL', 'UP') = 0. That is my understanding of my screenshot from the DD file. And Yes, even with this equation, the slack variables have postive values.
RE: Dummy import for slack variables - Antti-L - 25-10-2021
As I said, it is normal that the UC slack variables have positive values. But concerning the dummy import flows, they will inevitably be zero if you have indeed bounded them to zero, and the solver finds a feasible solution. Should you insist that this is not the case, it would be interesting to see it. Could you then please provide a reproducible case?
Or, could you also post the full DD file (of the scenario you showed an extract), and the listing file, to see if there is something unusual?
RE: Dummy import for slack variables - Mahmoud - 26-10-2021
Thank you for your answer Antti. Here's the DD file of the UC we are adding here. Thank you in advance
RE: Dummy import for slack variables - Antti-L - 26-10-2021
Your scenario uses $OFFEPS, and so you should know that any zero values defined for parameters will then only remove any earlier data for the parameters. As I am sure you know (because you are deliberately using $OFFEPS), this means that your "zero" FLO_BND parameters will have no effect: they are ignored. If you want to define zero FLO_BNDs that have an effect, you must define the EPS values for them: FLO_BND(r,t,'IMPNRGZ',com,'ANNUAL','UP') = EPS; That can easily be accomplished either by using $ONEPS, or the interpolation option 2: FLO_BND(r,'0','IMPNRGZ',com,'ANNUAL','UP') = 2; To conclude: $ONEPS is the default under VEDA, but you have decided to use $OFFEPS instead, and as a result your FLO_BNDs will actually have no effect.
