Bug on Switzerland's real time nuclear production ?

[Oocedar]

After wondering often which swiss nuclear power was not running at full capacity, I realized that the data was always correcting itself at next iteration. When looking at the real-time production, the live nuclear load factor is always below reality (around 82%) which is systematically corrected at the next iteration when we look at the past hours (where it is indeed running almost at full, around 98%). The main problem is that it means that the displayed "Carbon intensity" for Switzerland is usually overestimated (and the color displayed could also be wrong) as there seems to always be 15-20% of nuclear missing. Thanks for your attention

[tonypls]

Thanks for spotting this. You're correct, the data is estimated then there is backfilling process of the data with the source data to correct any error of the estimation.

@pierresegonne What do you think? There could be an opportunity to improve estimation for nuclear as it's quite consistent.

[Oocedar]

In Switzerland the scheme of the 4 remaining nuclear reactors is usually pretty straightforward :

• they are supposed to operate at full capacity from October to April
• they take turns from May to September to reload fuel and do the maintenance work

So unless an incident pops up (which you would see 2h later when you receive the data), I guess you could base your estimate on the last load factor received from the data. I think it would be a notable improvement since the live value is so much more important than the "archived" ones. Best :)

[pierresegonne]

Interesting!

Unfortunately, the estimates we have are one size fit all currently, and we don't have a way to ingest that sort of ad-hoc knowledge to improve them.

By the way, we don't actually estimate the nuclear production in CH as explained here: https://github.com/electricitymaps/electricitymaps-contrib/wiki/CH-Estimation-Model

I guess it might be the source that gives us something suspicious

[Oocedar]

Thanks for this description on how the estimates work for CH ! But then if you don't estimate nuclear for Switzerland, it means that you get some sort of "pessimistic" data (always 15% below truth) that is corrected 2h later ?

I guess I can't help you, up to you to see what you can do, but as I use your site in my lectures on Energy, it would really help f the "live colour and production" was "trustable", it's difficult for me to explain why the plants don't run at maximum on the graph :) Best, C.

[pierresegonne]

But then if you don't estimate nuclear for Switzerland, it means that you get some sort of "pessimistic" data (always 15% below truth) that is corrected 2h later ?

I guess so. Looking a bit deeper at the parser, the production data we receive comes from ENTSO-E, which displays the same step:

You can look at it by yourself here

ENTSO-E typically receives this data from each country's TSOs, so I guess the problem must come from how the Swiss authority releases that data in the first place.

I suggest that you write to ENTSO-E to report the issue and they'll escalate to the original source. I'll also do it, but it's better to do it multiple times :)

[Oocedar]

Will do so, thanks a lot for your help and time and keep it up, we need the whole world to know and use your site !

Le mar. 25 oct. 2022, 10:57, Pierre Segonne @.***> a écrit :

But then if you don't estimate nuclear for Switzerland, it means that you get some sort of "pessimistic" data (always 15% below truth) that is corrected 2h later ?

I guess so. Looking a bit deeper at the parser, the production data we receive comes from ENTSO-E, which displays the same step:

[image: Screenshot 2022-10-25 at 10 55 10] https://user-images.githubusercontent.com/32778266/197729938-cb98c514-3327-4541-9d88-cffb90e7ba9c.png

You can look at it by yourself here https://transparency.entsoe.eu/generation/r2/actualGenerationPerProductionType/show?name=&defaultValue=false&viewType=TABLE&areaType=BZN&atch=false&datepicker-day-offset-select-dv-date-from_input=D&dateTime.dateTime=25.10.2022+00:00%7CCET%7CDAYTIMERANGE&dateTime.endDateTime=25.10.2022+00:00%7CCET%7CDAYTIMERANGE&area.values=CTY%7C10YCH-SWISSGRIDZ!BZN%7C10YCH-SWISSGRIDZ&productionType.values=B01&productionType.values=B02&productionType.values=B03&productionType.values=B04&productionType.values=B05&productionType.values=B06&productionType.values=B07&productionType.values=B08&productionType.values=B09&productionType.values=B10&productionType.values=B11&productionType.values=B12&productionType.values=B13&productionType.values=B14&productionType.values=B20&productionType.values=B15&productionType.values=B16&productionType.values=B17&productionType.values=B18&productionType.values=B19&dateTime.timezone=CET_CEST&dateTime.timezone_input=CET+(UTC+1)+/+CEST+(UTC+2)

ENTSO-E typically receives this data from each country's TSOs, so I guess the problem must come from how the Swiss authority releases that data in the first place.

I suggest that you write to ENTSO-E to report the issue and they'll escalate to the original source. I'll also do it, but it's better to do it multiple times :)

— Reply to this email directly, view it on GitHub https://github.com/electricitymaps/electricitymaps-contrib/issues/4657#issuecomment-1290218201, or unsubscribe https://github.com/notifications/unsubscribe-auth/A3SDI64TQS4AIX6JK3KJA63WE6OH3ANCNFSM6AAAAAARDAZZEE . You are receiving this because you authored the thread.Message ID: @.*** com>