Open christian-rli opened 5 years ago
Ok, so if I understand the documentation correctly and want to add the NREL 15 MW offshore wind turbine we talked about over at https://github.com/PyPSA/atlite/issues/54 to OEP, I have to post the .xlsx file here, right?
What happens next, do I need to do something else?
Ok, so if I understand the documentation correctly and want to add the NREL 15 MW offshore wind turbine we talked about over at PyPSA/atlite#54 to OEP, I have to post the .xlsx file here, right?
You understood correctly :)
What happens next, do I need to do something else?
You don't have to do anything else. I'll try to get the update done today and will post here when it's online. In any case it won't happen until some time in the afternoon.
Compare and merge turbines from atlite
model @christian-rli
Also consider this issue: https://github.com/PyPSA/atlite/issues/54
Your turbine has been added @euronion Thank you for letting me know of the other issues @Ludee . I'm going to have a look at them soon.
I found some turbines which are not in the wind_turbine_library on this website (https://www.wind-turbine-models.com/turbines/882-senvion-3.2m114) however their impressum states that the data is not public domain and one should ask the manufacturer the right to use their data. I was wondering whether asking manufacturer works and if there is a pitch one can use to not scare them that their data will be "stolen" if we publish them with an open license?
https://www.htw-berlin.de/hochschule/personen/person/?eid=10251 war vorher bei Senvion der ist extrem nett; Kollege von RLI-Gründungs-Chef Jochen geht es nur um Senvion? für Suzlon: https://www.linkedin.com/in/dagmar-gramatzki-11616917?originalSubdomain=de lieben Gruß von mir.
danke @beritRLI :)
We actually asked a lawyer on this matter. My notes of his reply read as follows:
To sum up a long session with explanations of different rights that interact with each other in hazy ways: It is very questionable, whether the manufacturer can claim a copyright on the technical description of their product. They are published measurements which neither constitutes a meaningful effort in terms of creating a database (necessary for "Datenbankherstellerrecht") nor does it seem to reach a sufficiently high enough level of originality (necessary for "Urheberrecht").
So technically, in my view you should be allowed to just take individual specification sheets and extract information as you please. Only if you should adapt on a collection of sheets that was done to get a larger overview "Datenbankherstellerrecht" applies. In fact it does apply to the wind_turbine_library.
All that aside, it is always nice to let the manufacturer know, however. As far as I know we don't have a template text to use for this.
@christian-rli I was looking for the 15 MW turbine @euronion posted in https://github.com/OpenEnergyPlatform/data-preprocessing/issues/28#issuecomment-593411988 last year for inclusion in https://openenergy-platform.org/dataedit/view/supply/wind_turbine_library but couldn't find it. Can you check whether it got lost somewhere?
You're right @fneum , the turbine is missing in the uploaded table. I do recall adding it, but it must have been overwritten in a later upload. Looking into an update I ran into bad documentation unfortunately, hence the late response.
It seems that this (last) update was done manually with a quick sql command. However these are no longer supported. These days only API uploads are allowed. That's a good decision, but in the case of the wind turbine library that leaves us with a bit of a challenge, since the chosen data types weren't supported with the API last time we tried.
@jh-RLI could you please write an upload script for the library based on a csv download of the current table? If you can, preserve column types. I'll give you the documentation we have available internally. Then we can think about a working system for updating and leave that info here.
The complete dataset is now uploaded here: https://openenergy-platform.org/dataedit/view/model_draft/wind_turbine_library
@christian-rli please move the table to supply schema @fneum the data from https://github.com/OpenEnergyPlatform/data-preprocessing/issues/28#issuecomment-593411988 is available at id:"140" in the table. Note that the table is currently in model_draft schema and will be moved to supply schema soon.
I created a jupyter notebook to upload the data:
@christian-rli did you have a chance to move this to the supply schema yet?
Just done @fneum. Sorry for the wait. It's updated here now: https://openenergy-platform.org/dataedit/view/model_draft/wind_turbine_library
I tried moving the table from model_draft
to supply
directly, which failed, because the old table was still there. Then before moving that one to model_draft
I had to delete the upload by @jh-RLI , move the original table to model_draft, then delete it there, upload the data again (Thank you for providing the upload script!) to finally move it to the supply schema. BTW I added a turbine_id (152) while I was at it
All of this could have been much easier, had there been something like a renaming function. I couldn't find a documentation on anything like that, so I believe there is none. Or am I missing something @jh-RLI ? If no, I'd create an issue for that :)
Thanks!
@christian-rli By rename function you mean something like "rename table"? At least I am not aware of such a function. I would have used the backend DB access or written a script (which probably wouldn't work because the table was already in the restricted schema). It seems that we lack such functionality and all related functions to edit data that has already been moved to a restricted schema on the oep.
@christian-rli there is a turbine with false curve data in the turbine library. The same wind speed appears more than once in the curve data which leads to errors in the windpowerlib. Could you have a look at it? It's the one with turbine_type = IEA-15-240-RWT.
Thanks for letting me know @birgits . I'll look into it.
There are indeed quite a few duplicate wind speeds in the mentioned turbine type. Interestingly, there are matching power curve values, so maybe the decimal precision of the wind speeds just wasn't enough to account for this. Anyway, I removed all duplicate wind speeds, always keeping an entry with a value in the middle. Many odd numbered wind speeds remain, but no duplicates: supply__wind_turbine_library.csv
Unfortunately I ran out of time when uploading the table and will be gone for several weeks now. @jh-RLI or @henhuy could you help out with the upload, please? @jh-RLI has provided an upload script further up in this issue. Just make a copy of the original table, move it to model_draft, delete, upload the updated table supplied by me above and move it into the supply schema here. I have given both of you admin rights to the table. Also to you @birgits .
I can upload it next week.
I finally made some time for it tomorrow :) Sorry was quite busy last week.
I have updated the data. Please check as I don't know what exact values should have changed. @birgits
Thank you @jh-RLI! The buggy wind turbine is now correct, but something else must have gone wrong in the course of the upload, because for some wind turbines, there is now a bracket [ too much in the list of wind speeds. See e.g. turbine_type = "E-82/2300". These brackets do not occur in the csv file @christian-rli posted above, so I guess the problem comes from the upload?
This is wired. I read the csv with pandas, maybe there was a bug in there or I accidentally added the "]". I will try again :)
Hm for some reason when writing the data to the table (in these 3 cases where the falsy [ is included in the wind speeds column) it just starts over with writing the array. I do not have time to investigate this at the moment. Currently I dont think this bug is in the OEP.
Fixed "id" 9, 114 & 136. Let me know if you find more :)
Thank you! IDs 97, 82 and 54 are also broken. I will also include a check in the windpowerlib so that the download won't fail in case of faulty turbine data.
They are also fixed now.
@jh-RLI now it seems like the turbine library is missing completely.. Could you please check?
Oops, I forgot to move it back to the supply topic. Now it's back again.
Perfect, thank you! All turbine data looks good now :)
I have received an email pointing out another error in the library:
ENO126/4800 turbine has a small bug where several entries are in one column and others are empty.
I will fix this as soon as I have more time :)
The wind_turbine_library is a collection of wind turbine data, containing among other things power curves. Data review and metadata creation were discussed in #26 . Metadata are maintained here: v1.4 (and v1.3, which is maintained until the OEP-transition to v1.4 is complete). This issue will remain open for curation purposes. Feel free to comment annotations or point out errors. Feedback will be implemented as soon as possible.
Low threshold contributions can just be a comment in this issue.
If you'd like to add several turbines, it would help us if you could upload your changes to the table model_draft.wind_turbine_library_contributions, which has the same structure as the final library. This way, we can update the wind_turbine_library with much less effort. There is a python-script to help you with this. It allows you to upload a csv-file of the following format. Alternatively you may upload the file with your contributions here as an xlsx file. (Saving a csv-file in libreoffice as .xlsx works fine. GitHub unfortunately doesn't allow csv uploads)
Please cite as: "Wind Turbine Library" © Reiner Lemoine Institut | ODbL-1.0