1,800 days of minute-by-minute solar generation as Open Data

@edentOpen Data solar · 6 comments · 250 words · Viewed ~527 times.

Back in 2014, I released a year's worth of solar generation data. I was enormously proud to recently discover that the data were cited in a couple of academic papers:

Colantuono, Giuseppe & Kor, Ah-Lian & Pattinson, Colin & Gorse, Christopher PV with multiple storage as function of geolocation (2018) page: 217-232. Elsevier BV. Solar Energy. https://doi.org/10.1016/j.solener.2018.03.020
Shukhobodskiy, Alexander Alexandrovich & Colantuono, Giuseppe RED WoLF: Combining a battery and thermal energy reservoirs as a hybrid storage system (2020) page: 115209. Elsevier BV. Applied Energy. https://doi.org/10.1016/j.apenergy.2020.115209

Someone privately asked if I had more data - so I'm happy to release the full dataset. About five-and-a-half years' of readings, from 2013-12-28 to 2019-06-23.

Download .ZIP file (7MB)

The format is of the CSV is simple. ISO 8601 date, Watts being generated, Cumulative power generation.

Some days and data are missing due to power cuts and Internet outages.

The location of the panels was Oxford, UK (51.725275,-1.232889). They were South-facing. More technical details can be supplied if necessary.

These data are licensed to you as CC BY-SA. If you use them in an academic paper, you are expected to publish as open access.

Enjoy!

6 thoughts on “1,800 days of minute-by-minute solar generation as Open Data”

2020-07-06 17:55

Alcides Fonseca said on twitter.com:

Is there any kind of prediction that would be useful for you? Like knowing if it will reach a given threshold in the next 24h.
Reply | Reply to original comment on twitter.com
2020-07-06 18:04

Tom Morris 🏳️‍🌈 said on twitter.com:

I'm coming round your place with a big umbrella and slowly blocking and unblocking the sun in a Morse code pattern to get scandalous messages encoded in the dataset.I hereby call this method—wait for it—sunblockchain.
Reply | Reply to original comment on twitter.com
2020-07-06 18:52

Paul Kelly says:

Thank you. This is very interesting data. I am probably being a bit dense, but could you clarify why the cumulative values usually start some way above zero at the beginning of the day?
Reply
1. 2020-07-07 09:03
  
  @edent says:
  
  It's a quirk of the Fronius system. I think it's because the fringes of early morning sunlight are enough to trickle in some power.
  Reply
2020-07-08 16:11

Andrew says:

Very interesting, thanks for sharing. What electricity supplier do you use? I've been thinking about getting solar panels and following a few calculations believe that I should be able to cover my electricity needs with panels + battery storage, but think I may not save much money due to the daily standing charge (approx. 20p per day). Many thanks for any help!
Reply
1. 2020-07-08 18:09
  
  @edent says:
  
  I'm with Bulb. If you join using http://www.bulb.me/terencee5219 we both get £50. They take part in the Solar Export Guarantee scheme - so you get paid for every kWh you export back to the grid.
  Reply

1,800 days of minute-by-minute solar generation as Open Data

6 thoughts on “1,800 days of minute-by-minute solar generation as Open Data”

Alcides Fonseca said on twitter.com:

Tom Morris 🏳️‍🌈 said on twitter.com:

Paul Kelly says:

@edent says:

Andrew says:

@edent says:

What links here from around this blog?

What are your reckons? Cancel reply

Share this post on…

6 thoughts on “1,800 days of minute-by-minute solar generation as Open Data”

Alcides Fonseca said on twitter.com:

Tom Morris 🏳️‍🌈 said on twitter.com:

Paul Kelly says:

@edent says:

Andrew says:

@edent says:

What links here from around this blog?

What are your reckons? Cancel reply