Timestep compatibility with other Module

[yvann]

Hello,

currently I'm working on the resolution of a MIP integrating a module used to estimat the reliability of the power grid. When I execute my model with the option Timestep without the reliability module and conversely, everything works fine. However, when I excecute with both, There are a compilation error:*** Abort"COMPILE ERRORS IN GAMS **", followed by several messages ** Did not find GAMS ... **

I guess it must have a problem of compatibility between the two. I wonder if anyone has ever had this kind of problem?

Thank You

[Antti-L]

It could be some compatibility problem.  Could you upload the run file and the listing file (*.LST) for the case with these errors (packed into a ZIP file), so that I could see what the problem is?

[yvann]

Thank you for your help.
You will find attached .zip folder containing the files  <case>.RUN  and <case>.LST

[Antti-L]

Ok, I think I see where the problem is.

BTW:  decided to delete the ZIP file from here, as you had used an option to include the full TIMES source code in the listing file.

I verified that the problem is not related to the TIMES code itself, but to your FIA extension, which is not correctly coded to be usable under the stepped mode.  And, as ETSAP does not provide support for the proper GAMS coding of user extensions, it is basically your responsibility.

But I can give some quick advice:  You should include the TIMESTEP controls directly in the equation headers (e.g. EQ_FIAOF(R,T,P,S)) and remove them from the code inside the equations.

I can also see that you are using a very old TIMES version (3.3.1). Note that TIMES support is basically always provided only for the latest versions.

[yvann]

Thank you for your help and advice.

[Antti-L]

A small follow-up:

As you included your full model in the listing file, I was able to test with the FIA extension myself under time-stepped mode. With just small corrections to the equ_ext.fia file, the model solved fine, both with TIMES v3.3.1 and the latest TIMES v3.8.6. Using OPTCR=0 the MIP solutions were also identical up to the second timestep (the third step took too much time with OPTCR=0 for me to wait for it to finish).

Therefore, I still don't see why you choose to use the old version.  The modified equ_ext files for both versions are here: uploads/28/equ_ext-fias.zip

[yvann]

Hello,
I have access to only two versions of TIMES, 3.0.2 and 3.3.1. this is why I work on 3.3.1. I tested the model with changes made it work very well .Thank you so much for your help.