Electricity from no where

[yasmine94]

Hi everyone ,

I am currently working on modeling the power sector transition in The Maghreb region, considering four distinct scenarios. While the results for REG 1 (Tunisia) and REG 2 (Morocco) are relevant and align with our expectations, I am facing inconsistencies with the results for REG 3 (Algeria).

In the case of Algeria, the dummies table appears to be empty, raising concerns about the reliability of the data. Additionally, I noticed that the electricity demand is significantly higher than what the model is currently producing. For example, the projected electricity demand by 2050 is estimated to be 500 PJ, but the model is only yielding 4 PJ. This discrepancy suggests that the electricity supply is seemingly originating from a mysterious source, as there is no recorded production or importation.


I am seeking insights and guidance from the community to identify potential mistakes or issues in the modeling process. Has anyone encountered similar challenges in power sector modeling or have suggestions on how to address this discrepancy?

Thank you in advance for your valuable assistance!

[Antti-L]

My ten cent suggestion: Check your the QA_Check.log file in the work folder, and see if it contains warnings, e.g.

 *** RPC in TOP not found in any ACTFLO/FLO_SHAR/FLO_FUNC/FLO_SUM

Such warnings may point you to modeling errors, e.g. related to the input flows into the demand processes.

[yasmine94]

My ten cent suggestion: Check your the QA_Check.log file in the work folder, and see if it contains warnings, e.g.

 *** RPC in TOP not found in any ACTFLO/FLO_SHAR/FLO_FUNC/FLO_SUM

Such warnings may point you to modeling errors, e.g. related to the input flows into the demand processes.

Thank you Antti for your answer, i checked the QA_check.log and I found this: 
*** GAMS warning detected; Input data have been filtered via GDX

[Antti-L]

Ok, then you have domain violations, which may be related to your problems. Check the warnings in the listing file (*.LST), and fix those domain violations in your model...

[yasmine94]

Ok, then you have domain violations, which may be related to your problems. Check the warnings in the listing file (*.LST), and fix those domain violations in your model...

Dear Antti,
thank you for your answer, I checked the .LST file of the scenario I am running, but I was not able to identify the problem.
I attached the .LST file requested.

[Antti-L]

You have: 
**** 0 ERROR(S)  3924 WARNING(S)
3924 domain violations! It is pretty serious as such.  Most of them seem to be about timeslices, and some of them about missing IMP/EXP indicators. I think you should definitely correct all of them.

But yes, you are right that these errors in your model might not be related to the modeling error of the electricity chain for Algeria. To identify that error, it should be quite easy for you, simply by going through the energy chain, from the demand (which you said to be around 500 PJ in 2050) backwards to the supply. Obviously, there must be some process satisfying that demand, otherwise the model would be infeasible. Perhaps that might be the ELCDEMAND01 process? Then just check the input flows of that process and see if they are still correct, and then the processes outputting those commodities, and so on. I am sure you would easily find the modeling error point by just doing that small effort.