Modeling transmission lines in multi-regional models

[Antti-L]

No, there was no Pset_PN column in your table. The column AO was not in that table (there were blank columns between that and the ~TFM_INS table, so AO was not within the table).  

[Raulm]

I deleted all the region-specific columns, since I am using only the "All Regions" columns and it worked! I will now try the other parameters.

thanks again!

[saleh]

I have a quick question:
As you mentioned we can define any number of parallel trade links for the same commodity and between the same regions. Where should I define the second trade link? The first trade link is automatically generated in form of TB_ELC_REG1_REG2_01 using trade matrix in _trade_links_ file. How and where should I define the second trade link (which is new tech not base tech) between region 1 and 2 (i.e. TB_ELC_REG1_REG2_02)?

Thanks!

[Antti-L]

I am not sure what you mean by "base tech" and "new tech"??

I already tried to explain earlier in this thread that all TIMES trade processes can have both existing capacities and investments into new capacities. If you mean that you would like some of the trade link processes (defined in Scentrade_Trade_links) to appear in the "Base" scenario, and some others in a Subres scenario called "New", I don't think that is possible under VEDA-FE. All the trade processes defined in Scentrade_Trade_links will be defined in the Base scenario. But you can enable/disable any subsets of these processes in Scenario files, by using e.g. the START parameter. In that way you can have full control over the available trade link processes on a scenario basis.

Parallel trade links can be easily defined in Scentrade_Trade_links. Two methods are shown below:


The two tables shown in the figure above define 6+6=12 parallel and bi-directional links for ELC trade between the ROW and WEU regions of the DEMO model (24 TOP_IRE entries in total). By adding more rows you can have as many as you like...

[Antti-L]

Saleh: Could you confirm whether the fact that all the parallel trade processes are being "base techs" and not "new techs" causes some problems for you, or that this issue may be considered solved now?

[saleh]

Hi Antti,

I originally wanted to model transmission lines expansion. If I use only one trade process for each link then the model incurs some costs for keeping transmission capacity at the existing level (since technologies retire in their life time) and this happens even if each link has a big liftime. Therefore, I think the existing trade capacity should have no cost whereas the parallel link would have investment cost. I think this can be done by "base tech" and there is no need for "new tech", right?
Another question that I have is regarding bi-directional links, in electricity transmission lines I believe the capacity from region A to B is equal to the capacity from region B to region A and the total transmission capacity in each timeslice from either A to B or B to A should be less that transmission line capacity, however, I noticed that sometimes investments in transmission lines are not symmetrical in the model (i.e. the model invests more in one side of line), what does that mean and how should I model transmission lines?

Thanks!

[Antti-L]

Thanks for confirming that the "base techs" are OK for you, and no need for "new tech".

Yes, you can define costs for the parallel links as you like (e.g. no investment cost for links with non-zero existing capacities, and full investment costs for links with no existing capacities defined).

But I don't understand why you say that the model would incur some costs for keeping transmission capacity at the existing level. That should certainly not be the case:  If the existing level of capacity is X, and you define NCAP_PASTI=X with a lifetime of 200 years, you will not have any costs for keeping the transmission capacity at the existing level X, because the capacity is already there. Can you explain why there would be costs for keeping the capacity at X, even if the existing capacity remains at X?

I don't know why your model would invest on more new capacity on one side of a trade link, if the process parameters (lifetimes, availabilities) and the timeslice division are equal in both regions (but if I had access to you model I would be able to see why). Could you provide a reproducible example? Or could you at least show all the process parameters and primary groups for both of the two regions, as well as the non-symmetrical results for new capacity?  Is the difference notable, or just a small amount?

I usually define investments only on one side (not both sides), but it is up to the modeller to decide whether to have them on both sides or just on one side.

[saleh]

Thank you Antti, I did not know about NCAP_PASTI, it makes sense and I don't have that problem anymore. 
The difference in different regions is not notable, so I guess it should be fine (i.e. in the range of 10^-1).
One more question: when I use NCAP_PASTI, I don't have new investments to compensate for retirements, however, I still have COST_INV for all trade links in all years (not only for the trade links that have new investment). I don't understand why I should have these costs.

   
   

[Antti-L]

The costs are obviously the annualized investment costs for the past investment (defined by NCAP_PASTI).  
Many TIMES users DO want to have investment costs reported also for past investments, even though they are just a constant term in the objective function.

If you don't want to see those costs reported, although you want to define investment costs for new capacity (for the same process), there are several ways to eliminate the costs reported for the past vintage: you can either define the investment costs to be zero for the PASTYEAR vintage, or you could define a very short NCAP_ELIFE for the PASTYEAR vintage. An example is shown below, where I have defined NCAP_PASTI=1 for a past investment made in 2000 and a technical life of 100 years for that existing capacity, but an economic life of only one year for the existing capacity.  In this way, the TB_* processes would not have any investment costs reported for the past vintage (2000), but the costs for any new investments will be reported according to the economic life of 50 years.

[saleh]

Thank you Antti, that was very helpful. I am trying to model carbon cap in one region and I am not sure how I can do that. I wanted to assign the average carbon intensity of all regions except the region under the cap for the electricity import into the capped region and do iterations. I found an attribute that is presumably for defining emissions for import/export (IFLOSUM), however, I am not sure how to implement it in VEDA so that it only generates emissions for the imported electricity (not exported electricity) in the desired region in each timeslice.
The bigger problem is we might have both import and export in one timeslice, so we cannot calculate how much of imported electricity in that timeslice is actually consumed in the desired region. Therefore, implementing carbon cap gets very complicated. I'd appreciate your ideas.

[Antti-L]

By using IRE_FLOSUM(r,y,p,com1,ie,ts,com2,io) you can define a coefficient for the production or consumption (io=OUT/IN) of commodity com2, in proportion to the import/export (ie=IMP/EXP) of commodity com1 in region r, period y, timeslice ts. Under VEDA, the dimensions ie, io and com1 must be specified in the Other_Indexes column, while com2 goes to the standard commodity column. It works well in my models.

I hope that helps.

[saleh]

[Thanks. Please see my attached input.

   
 I get "Mismatch number of elements provided and expected in Other_Indexes column for attribute IRE_IFLOSUM" when I import to VEDA and in it seems VEDA cannot read my input properly.

   

 Would you please help me with that?

Also do you have any other idea about modeling carbon cap in one region? It seems this approach is not accurate since we may have both import and export in any specific timeslice so all of imported electricity might not be used in the region under the cap.

Thank you!

[Antti-L]

Try using ~ not , as the separator.

[saleh]

I want to define some other trade links (biomass) besides electricity trade links. However, when I define them in _Trade_links_ excel file, VEDA doesn't read them, but when I define them using VEDA wizard, they are shown in VEDA and the excel file that is generated by VEDA is exactly the same as how I defined them. Then I want to assign some attributes to this new links (efficiency) in a scenario file. However, when I sync that scenario file, VEDA does not recognize the trade links that are generated by itself anymore. Your help is appreciated.

[Antti-L]

I hope you are not addressing this question to me, are you?

I don't even know what the "VEDA wizard" is (but would be interested to know   ).
As I have told several times before on this Forum, I am not an expert on VEDA, only on TIMES.  
I would suggest a new thread about this VEDA problem, which seems largely unrelated to the original topic.