Grid and Power Flow Modeling in TIMES

[Antti-L]

As the users of VEDA-TIMES surely know, the modeling of electricity grids is basically quite easy in TIMES. But how about power flow modeling?

Implementing non-linear power flow equations into a TIMES model would perhaps not be very practical, but linearized power flow equations might be considered. I found a neat example of using linearized power flow equations for the modeling of power markets, and so I was tempted to take a look how the approach works in TIMES. The basic data for the example model is courtesy of Andrew Stiel (http://www.files.ethz.ch/cepe/AndrewStiel.pdf).

I now uploaded this small TIMES model here (uploads/30/GridDemo.rar), so that others can comment on the usefulness of the approach.

As mentioned above, the purpose of this example model is to demonstrate one possible way of modeling electricity grids with power flow equations in TIMES.  It has 21 DAYNITE level timeslices, and features the following aspects of grid modeling:

• A grid with 16 internal nodes, 21 internal transmission lines, and 3 interconnections to external nodes;
• A total of 20 electricity plants connected to different nodes in the grid;
• A node-specific demand connected to each of the 16 internal nodes;
• Modeling of grid losses for each of the transmission lines;
• Modeling of linear power flow equations for the whole grid.

In the example model, all the nodal demands have the same load curve, but that could be easily generalized. The model also has the small limitation that the internal transmission lines cannot have different availability factors in each direction, but that limitation could also be easily removed with small changes that will be supported by the next version of the TIMES code.

Defining the power flow equations manually is of course somewhat clumsy.  Therefore, should the power flow modeling approach be of general interest, one might also consider implementing into TIMES automatic generation of the equations, on the basis of user-supplied reactance data for each transmission line of the grid.

If you have any comments, please post them in this thread.

[Edit:] New VEDA2 version here:
  GridPowFloDemo.zip (Size: 172.83 KB / Downloads: 35)

[Pernille.S]

I think this is very interesting. In particular the effect of including the linearized power flow equations to the model results.

How will this work with aggregated nodes, for example a region in a country? I assume it will capture the power flow equation between the regions but not within a region.

Also, I think it would be ideal if the user could choose to include both AC and DC cables in the same model.

Pernille  

[wnijs]

I also think this is interesting. It would shift the classical general electricity price to nodal prices. It could be an important factor in the trade off between long distance transport of electricity and local production.

Wouter

[Antti-L]

For those interested, linear DC power flow equations have been available in TIMES for some time:

http://www.iea-etsap.org/docs/TIMES-ELC-Grid-Features.pdf

[F.Rehulka]

Hi,

I am interested in applying the power flow equations to a multi regions model, but in a first approach, without creating a grid associated to each region and thus only applying the power flow between regions. To do so, is it necessary to create a new subres with new commodities, process and attributes for the power flow equations, or is it possible to apply the equations to the already existing trade process by just using the PRC_REACT and GR_GENLEV attributes ? (the latest do not seem to work that's why I am asking...)

Thank you in advance.

[AKanudia]

Are you using the current version of VEDA FE?

[F.Rehulka]

Yes I am using VEDA FE 4.4.7.

[Antti-L]

You seem to have misunderstood the documentation. GR_GENLEV can only be used when you have created a separate add-on grid associated to each region. It is utilized for the allocation of the generation in each region to the nodes of the grid.  As you have not created such a grid, you cannot use GR_GENLEV, sorry. 

[F.Rehulka]

Thank you, no GR_GENLEV then. The PRC_REACT attribute should however activate the power flow equations for the existing trade process on which it is specified, shouldn't it ?

[Antti-L]

Indeed, it should.  No new grid needs to be specified for using the power flow equations, if your model already includes a grid, comprised of interconnected trade links between regions.

Are you saying that the equations are not generated at all, or that they just "do not seem to work"?

Maybe you could give some more information? Perhaps you could post a picture of your trade matrix definition? And maybe you could also upload the listing file (ZIP/RAR compressed) from the model run where you have included the PRC_REACT parameters for the electricity exchange links?

[F.Rehulka]

It is working...! I guess it was already working (I could indeed see the powerflo.vda being included in the .lst) but I could not see any changes in flows, maybe because of too little differences between the reactance values I tried. With large reactance differences, I can now see flows changing with and without the power flow equations.

Thank you !

[krakowski]

Hello,

Do you know from which version of VEDA_FE, the grid and power flow modeling in TIMES is working properly?

Thankyou in advance,

Vincent

[Antti-L]

New VEDA2 version of the Grid & PowerFlow demo is here:
GridPowFloDemo.zip