How well energy storage capacity is represented in the dataset?

[matteodefelice]

I am including here a question I have received from Alexander.

Can you give an estimation on how well energy storage capacity is
represented in the dataset? 45 TWh for Norway for instance seem to be a
bit too low, but maybe I am wrong.

This was my answer:

Regarding the storage (and in general to the whole dataset), we have tried to include "as much as possible" but, of course, it wasn't enough to match what we think it should be the 'reality', both for the installed power and reservoir capacity.

Norway, thanks to the NVE data that is the best you can found among the European countries, is quite accurate but however the total storage doesn't match the values in the ENTSO-E TP or the Nord Pool data, which is ~76 TWh. In general, for the reservoir capacity, we rescale all the values we have at national level to match the maximum storage capacity we have from ENTSO-E TP or other source (for example, TERNA for Italian hydro).

For the Norwegian power plants, NVE provides a variable named 'energiEkvivalent_kWh_m3' which is the energy that can be generated from each m3 of water. Then we multiply this with the volume of the storage in cubic meters to get the maximum storage capacity in GWh. Honestly I don't know why the values do not match but unfortunately right time I do not have the time to investigate those issues.

[hannaekfalth]

Dear Matteo,

Im interested in how energy storage capacities are estimated and just wanted to know if you have a link/sorce for the NVE database where they provide the 'energiEkvivalent_kWh_m3' and the volume of storage.

Thanks in advance!

[FebinKa]

Hi,

I have also looked into it recently and I'll let him correct me if it is wrong but I think he obtained the energy equivalent from the NVE Vannkraftdatabase and the volume from the column "OPPSTRVOL" in the table starting around p.240 in this report.

[matteodefelice]

We use essentially two files: Vannkraft.xlsx and Vannkraft_Magasin.shp, both downloaded from the NVE website. I can't find the URLs to download them by the way so one is here
Vannkraft.xlsx and a zip with the shapefile is here (the link will work only for 7 days).

The size of the storage is taken from the column energiEkvivalent_kWh_m3 of the Excel file and the volume is in the field volOppdemt of the shapefile.

[mtaylor72]

Can I just confirm that the MWh of storage quoted are 'usable' rather than the total in the reservoir? thanks Michael

[matteodefelice]

Yes, I confirm. In all the cases, the reported reservoir capacity (MWh and/or hm3) is the 'usable' part.

…

On Sat, Mar 13, 2021 at 11:11 AM mtaylor72 ***@***.***> wrote: Can I just confirm that the MWh of storage quoted are 'usable' rather than the total in the reservoir? thanks Michael — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#1 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABQQYASHTXNCUQ42ZJ7JK4TTDM24NANCNFSM4HFDRBIQ> .

-- Matteo

[mtaylor72]

Yes, I confirm. In all the cases, the reported reservoir capacity (MWh and/or hm3) is the 'usable' part.
…
On Sat, Mar 13, 2021 at 11:11 AM mtaylor72 @.***> wrote: Can I just confirm that the MWh of storage quoted are 'usable' rather than the total in the reservoir? thanks Michael — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#1 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABQQYASHTXNCUQ42ZJ7JK4TTDM24NANCNFSM4HFDRBIQ .
-- Matteo

Thanks for the quick repsonse!

[ebbekyhl]

Hi Matteo,
Allow me to follow up on this thread, although it is more than three years since you created it. I am looking at the storage energy capacities (i.e., the 'storage_capacity_MWh' attribute) for pumped-hydro storage in your dataset.

What is in particular of my interest is the plant named 'Villarino VLLRINO 1G - Almendra (Villarino) Hydroelectric Power Plant Spain'. According to your dataset, it has a storage energy capacity of 3.1 TWh and a power capacity of 851 MW. This corresponds to a discharge time (duration) of 3666 hours. I am not exactly familiar with how the Villarino plant operates, but it seems quite impressive. Is this true, is it a bug, or does it relate to the above disclaimer regarding how well energy storage capacity is represented?

I have furthermore compared your reported cumulative energy capacity of pumped-hydro storage (approximately 11TWh) with a study from 2015 by Geth et al. (https://doi.org/10.1016/j.rser.2015.07.145) who finds it to be one order of magnitude lower (approximately 1.5 TWh). Again, could the discrepancy relate to how well energy storage capacity is represented?

Best regards
Ebbe

[matteodefelice]

Dear @ebbekyhl, yes the Villarino plant is quite impressive: the size of its reservoir is also confirmed in this news https://www.lagacetadesalamanca.es/provincia/viaje-a-las-entranas-de-la-central-de-villarino-CH900206

The paper by Geth et al. is indeed a very good paper, but we think that our collected data might be more accurate because we tried to collect multiple sources for most of the biggest plants. For example, they don't report the storage size (E_stor) for many plants and this might be the reason for the discrepancy in total size.

[ebbekyhl]

Thanks for turning back to me and for clarifying, @matteodefelice.

[martavp]

Hi both,

Villarino is one of the hybrid (PHS + reservoir) hydro plants in Spain.
It is mostly used as a reservoir plant (i.e. it shows a seasonal pattern) as shown in the next image
https://www.embalses.net/pantano-71-almendra.html
(It has two official names 'Villarino' and 'Almendra')

I would suggest not considering it a PHS plant since its energy capacity (3TWh) is two orders of magnitude larger than the estimated energy capacity for PHS in Spain (around 72 GWh) and this can introduce significant errors.

[matteodefelice]

Thanks @martavp, this is indeed very interesting. It would be nice to be able to understand how much of this storage is "managed" as bi-directional. I tend to agree with you, probably we are drastically overestimating the PHS capacity in Spain...

[irm-codebase]

Are there plans to fix this?
I think other datasets (powerplantmatching) are correcting for the Spain issue...

Or perhaps this is an old problem that is already fixed?

[matteodefelice]

Hi everyone, May I ask you to open an issue on Github and propose a solution? I think it's worth keeping track of this kind of discussions. Would you suggest splitting in two? Or just classifying it as a reservoir? Let's continue there, what do you think?

…

On Wed, Aug 14, 2024 at 10:20 AM Ivan Ruiz Manuel ***@***.***> wrote: Are there plans to fix this? I think other datasets (powerplantmatching) are correcting for the Spain issue... Or perhaps this is an old problem that is already fixed? — Reply to this email directly, view it on GitHub <#1 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABQQYAWZJ75Q7KKZNV4DOZLZRMHLVAVCNFSM6AAAAABMP2FUOOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOBYGEZTENJQGU> . You are receiving this because you were mentioned.Message ID: ***@***.*** com>

-- Matteo

[OskarVagero]

I hope that it is OK to resume the discussion on the Norwegian storage capacity in the database in this issue.

I understand the processes as that the Norwegian storage capacity has been calculated by multiplying the energy equivalent [kWh/m3] with the "OPPSTRVOL" data from the previously mentioned report/shapefile. Comparing a few rows in the table with recent data indicates that this is only the sum of the upstream reservoirs and that it does not include the reservoir at the power plant itself. This makes sense given that OPPSTRVOL should mean "Upstream volume" and is likely the reason why the results are not correct.

Luckily, the Norwegian Resource and Energy Directorate (NVE) have published reservoir capacity data (both per hydro power plant and per reservoir), where the total storage capacity sums to 87.4 TWh. Direct download link: nmag_kraftverk.xlsx and the page describing and linking to the data.

I intend to do a pull request today or later this week updating the Norwegian data. I just need to do some processing. There is also some updates one could do on the classification between reservoir and run-off-river that I hope to get to.