Electricitymap-contrib: Singapore has solar data

[jarek]

https://www.emcsg.com/ has generation mix for Singapore. It's overwhelmingly gas, but it is real time... Raw data comes from https://www.emcsg.com/ChartServer/blue/ticker (updated half-hourly).

The data has demand and "system loss" in MW, and mix in percent. Should generation be calculated based on demand plus system loss, then?

The data also contains prices for the following categories: USEP, LCP, Primary Reserve,
Secondary Reserve, Contingency Reserve, and Regulation. Is one of those of interest for the price parser?

[jarek]

Found this for the explanation of the prices https://www.emcsg.com/marketdata/guidetoprices

[systemcatch]

@jarek nice find on the data. I searched around and found this report from the government which may be useful in cross checking some figures. System loss looks quite low but Singapore is very compact so it makes sense I guess.

[corradio]

Indeed generation should be based on demand + loss.
For the prices I think that's a tricky issue. In general we have to rethink how we want to show prices, because currently they are very hard to compare.
Let's leave prices out for now, or simply take the spot price.

[jarek]

OK, I'll write a parser for the generation and the USEP now then.

[corradio]

Awesome!

On Thu, Jul 27, 2017 at 8:33 PM, jarek [e-mail] wrote:

OK, I'll write a parser for the generation and the USEP now then.

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
https://github.com/tmrowco/electricitymap/issues/653#issuecomment-318448661,
or mute the thread
https://github.com/notifications/unsubscribe-auth/ABlEKO9UkUIuIcyo-N10KgC85hr2eY3qks5sSNgVgaJpZM4OkHwm
.

[jarek]

Now done :tada:

[alixunderplatz]

@corradio @jarek
I found a link to Singapore's PV production.
https://www.ema.gov.sg/solarmap.aspx

Unfortunately, it's a png-image only.
https://www.ema.gov.sg/cmsmedia/irradiance/plot.png?ref=636512085214679676

Installed inverter capacity is ~109 MW, so only 1% of natural gas capacity / <2% of current daily peak demand.
Installed PV module capacity is ~141 MW according to:
https://www.ema.gov.sg/cmsmedia/Publications_and_Statistics/Statistics/31RSU.pdf

It's a quite insignificant share at the moment, but is there any way to use OCR or something similar for reading the data?

---

Edit: I've just seen PV was also described in the documentation in the Singapore parser, also mentioning its small impact on the mix. So the source for real time data is the only thing adding new info ;)

[systemcatch]

@alixunderplatz OCR would be fun to try.

Running the following commands on your image.

convert sing_pv.png -threshold 30% sing_bw.gif
tesseract sing_bw.gif output
Tesseract Open Source OCR Engine v3.03 with Leptonica

Gives me this result in a .txt file.

We're in business!

[alixunderplatz]

@systemcatch That simple, wow :) 👍 ❤️

As you can see, there is an output of 0.33 MW now, although it's night in Singapore. I don't think it comes from light pollution but it is kind of an "offset" value from their radiation data.
Would make sense to implement a threshold like 1 MWac or sth.

Edit: right now its 0 MWac, so forget the stuff about the threshold :)

[alixunderplatz]

@systemcatch I noticed the image's URL is changing - it is refreshing every minute according to the site https://www.ema.gov.sg/solarmap.aspx

NEXT EDIT: Numbers behind the image URL are irrelevant.

https://www.ema.gov.sg/cmsmedia/irradiance/plot.png

[brunolajoie]

haha that's brilliant! 🔍

[jarek]

Maybe we can email them and ask if same info is available in text format somewhere

[systemcatch]

Built a function that gets this data. Only question is whether it should be added to SG.py or in parsers/lib like AU_solar.

[systemcatch]

All done here, feel free to reopen if anything new comes up.