Veda2.0 Released!


Export and FLO_SUB
#1
Hi,

I'm having trouble locating a bug in a quite large TIAM-like model, so I try and post the topic here in case I missed something:

The problem is related the export of LNG to an external IMPEXP region. Since the model is a South American one, I have set TU_GASLNG_XXX_YYY processes that trade LNG inside the model, and TU_GASLNG_XXX_IMPEXP or TU_GASLNG_IMPEXP_YYY that trade it with the rest of the world. (In my case, "XXX" is mainly Trinidad and Tobago, and YYY would be e.g. Chile.)

The trade parameters are only an ACT_COST to represent transport expenses, and a FLO_SUB representing the revenue from selling the gas in the case of export towards IMPEXP. For import from IMPEXP, FLO_SUB is replaced by a FLO_COST parameter. Last, internal trade is modeled only by an ACT_COST (efficiency is supposed to be 1).

My problem is that no matter how high I set the FLO_SUB for XXX (Trinidad and Tobago), the region keeps exporting its LNG production to another model region YYY that will ultimately sell it. There is no additional transformation, it is only transferred to that region, then sold at a "lower price"... VEDA does not report any additional constraint on GASLNG or on the trade technologies, and I do not remember having set one.

Is there any hint of something I missed? Or maybe an alternative modeling technique?

Thanks and regards,

Sebastien
Reply
#2
It might be related to the default convention in TIMES, that the total regional variable cost component is assumed to be non-negative. You could check the variable cost components VAR and VARX of the objective (Reg_wobj). If the sum of them is zero for the region, then you are hitting the lower bound.

You can make the variable costs free by defining UC_RHS('OBJVAR','N')=<any value>.

But if it is not this issue, some more info would be needed...
Reply
#3
Thank you Antti, it was the right answer. I did not know of that non-negative restriction, so it did not make sense to me that the model would not try and lower its costs, even negative ones.

Could you please show me how you set the UC_RHS('OBJVAR','N')=<something> using VEDA please?

Thank you very much!
Reply
#4

I am glad the explanation to your problem was found.

It is quite easy to define the variable costs free in all regions, with a simple ~TFM_INS entry:

Reply
#5
Thank you very much Antti for you help!

Regards,

Sebastien
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)