I am experienceing an issue with the Com_OFF attribute, which does not seem to behave as expected.
I tested it in a demo model by setting Com_OFF = 1 for the commodity AGRELC between 2004 and 2060, along with G_OFFTHD set to 1.
However, instead of being fully deactivated, AGRELC appears to be produced out of nowhere at no cost.
Has anyone encountered similar issues with the Com_OFF attribute ? Any insights would be greatly appreciated.
Good question.
What the model generator has thus far been always doing with COM_OFF, is to remove the corresponding milestone years from the sets RTC(r,t,c) and RTCS_VARC(r,t,c,s), such that the commodity balance equations will not be generated for those years. But that is indeed not sufficient for removing all the process flows of that commodity.
The documentation nonetheless says that COM_OFF would be "Specifying that the commodity c in region r is not available between the years y1 and y2 ". I guess the "not available" can well be considered to imply that the flows should be turned off as well. If that's what would be considered agreeable, I could implement that in the code accordingly.
Anyone having an opinion about it, please post it here.
07-03-2025, 09:35 PM (This post was last modified: 07-03-2025, 09:51 PM by vincedh.)
Hi Antti,
Thank you for your answer, I understand. In fact I wanted to use Com_OFF because I had very high commodity marginals during some periods although the processes that could produce these commodities where not available yet (START was defined later on). I guessed that cutting these commodities off is those periods would alleviate a bit the optimization. But if I also cut all the inbound flows with PRC_FOFF, I guess this would do the job.
For this use of Com_OFF, I would indeed appreciate that it cut all process flows of the commodity in question (following what is indicated in the documentation as you mentioned). But I am not sure if other people would like it only not to generate the commodity balance equation.
