Electricitymap-contrib: ENTSO-E Pumped Hydro consumption missing

[Fred-3000]

It seems that consumption values for pumped hydro (storage) are missing for countries based on ENTSO-E data.
They are reported on their website, so I guess something is wrong in the parser.

[willbeaufoy]

Which countries are you looking at? I checked now on the frontend and the UK, France and Germany are all reporting pumped storage values.

[Fred-3000]

Belgium, Germany, Portugal, and more...
The positive values (when producing electricity) are shown, but the negative values (when storing electricity) are not.

[willbeaufoy]

Ah yes I see what you mean. So for example hydro storage for Austria is reported as -.0.0 at 03:00 UTC this morning:

However ENTSOE reports that at this time hydro storage consumption was -422MW (and there was no hydro storage production at this time).

[willbeaufoy]

I've found the cause, it's because of this commit from a month ago that ignores consumption values when calculating production. If you revert the code that was changed then we get the pumped storage data again. @FelixDQ @PaulCornelissen (authors of the commit) was this effect intended in your commit?

[PaulCornelissen]

Nope, that was certainly not the intended effect. You can read the history here: #2676 and #2638

In short: In NL there's sometimes significant self consumption at power plants (e.g. CHP plants at large consumers). Sometimes it's more than the production for an entire generation type (mostly for unknown). That would cause the production for unknown to be negative and the parser would fail.

I suppose that we should add an extra condition to differentiate between consumption for hydro(storage) and other types of generation.

[willbeaufoy]

Nope, that was certainly not the intended effect. You can read the history here: #2676 and #2638

In short: In NL there's sometimes significant self consumption at power plants (e.g. CHP plants at large consumers). Sometimes it's more than the production for an entire generation type (mostly for unknown). That would cause the production for unknown to be negative and the parser would fail.

I suppose that we should add an extra condition to differentiate between consumption for hydro(storage) and other types of generation.

Ok I will see if I can fix it.