Multi-regional model parameters

[fg]

Dear TIMES-Supporters,

in my model I'd like to distinguish clearly between the imports from external regions in the primary sector (e.g. crude oil imports) and the internal trades between the several regions of my model named TB_* and TU_* for bi- and unilateral trades.
The problem however is, that both of these processes belong to the process set IRE.

Parametrizing the costs and bounds of internal exchanges seems to be possible via the Pset_PN field in the trade parameters file:

~TFM_INS
TimeSlice LimType Attribute Year Other_Index Attrib_Cond Val_Cond AllRegions Pset_Set Pset_PN Pset_PD Pset_CI Pset_CO Cset_Set Cset_CN Cset_CD
DAYNITE ACT_COST 2010 Imp 0,001 IRE TB_*, TU_* ELC
DAYNITE ACT_COST 2010 Exp 0,0001 IRE TB_*, TU_* ELC
DAYNITE ACT_BND 2010 Imp~Exp 10000 IRE TB_*, TU_* ELC

However, when I'm in the Backend looking into the results and choosing just IRE processes, also these from external regions are shown, but I just wanna examine the internal exchanges.

My workaround would be to set the imports from the external regions belonging to the MIN set and just calling them IMPOIL1 or so...

I wonder if there's a more elegant way to do it.

Thanks for you help!
fg

[Antti-L]

However, when I'm in the Backend looking into the results and choosing just IRE processes, also these from external regions are shown, but I just wanna examine the internal exchanges.

Yes, in TIMES, any processes that have their topology defined by TOP_IRE are IRE processes.  Consider a model with trade links between two internal regions REG1 and REG2.  Now, if you run the model with only REG1 as an internal region, thereby leaving REG2 as an external region, the role of these processes remains the same: they still represent trade between REG1 and REG2, even though REG2 is now an external region.

It is quite common to run multi-regional models with different subsets of internal regions, where trade links can be in some cases between internal regions and in some cases between an internal and an external region.  On the other hand, the MIN identifier is commonly used for mining processes. I don't see why you would want to associate the imports from external regions with mining.  But you can indeed define new sets of processes in VEDA-BE in a quite flexible way.

[fg]

It is quite common to run multi-regional models with different subsets of internal regions, where trade links can be in some cases between internal regions and in some cases between an internal and an external region.

Dear Antti-L, thank you for giving advice. However, I did not know that changing the role of a region is common practice... What is the deeper sense in it?
Let's say you wanna model the United States. You would define the US-states as internal regions and everything else whats being imported from overseas would be modeled as coming from outside. Or you wanna model Europe: One would define the European countries as internal regions and all the imports of gas (mostly Russia) and/or oil (MENA region) are traded with RestOfWorld. Why would it make sense to change those regions to internal or change any of the internal regions to external??

On the other hand, the MIN identifier is commonly used for mining processes. I don't see why you would want to associate the imports from external regions with mining.

Exactly. For my understanding it would make no real difference for the model. Mining and/or importing are just processes which work as sources of primary energy defined with additional costs for the model. Usually a state would use as much of its own resources through mining until importing becomes cheaper. So why differentiating between those primary processes?
Am I thinking too practical on these issues?

But you can indeed define new sets of processes in VEDA-BE in a quite flexible way.

How would I do that? Defining new sets of processes which are shown in VEDA-BE?

[Antti-L]

Why would it make sense to change those regions to internal or change any of the internal regions to external??

If you have a large model containing European, North American, African and Asian regions, the model would take a long time to solve, especially if you have a large number of timeslices. Therefore, for a study focusing on Europe, maybe using more timeslices, it can be quite useful to run the model only with the European regions internal. And when you study is focused on South Africa, you might not want to include e.g. North Korea or Mongolia in the model, and so I would leave them external. The full model would be best suited for global studies on e.g. climate change policies.

So why differentiating between those primary processes?

Well, if you want to call electricity imports "mining of electricity", then fine, but how about electricity exports? I think calling them trade processes simply makes more sense. In addition, note that mining is a domestic activity, which contributes to the GDP, employment and gross income, while the import expenditures go to an alien foreign vendor. GDP can be calculated as the sum of consumption, investment, government spending and net exports, and imports thus have a negative contribution. So, when doing economic analysis (or analysing security of supply), I think it makes a notable difference.

But you can indeed define new sets of processes in VEDA-BE in a quite flexible way.

How would I do that? Defining new sets of processes which are shown in VEDA-BE?

Yes. Goto Sets → New set(s), and define, for example, a new subset of IRE processes containing only those that do not have TU_* or TB_* in their process names.

[fg]

Thank you very much for your overall professional response.
The internal/external modelling approach now makes full sense to me.

Moreover, defining new sets in the Back-End is both really useful and easy - it will use it from now on!