Hacker News Reader

greenforecaston Ask HN: What have you created that deserves a second chance on HN?

You can halve[0] the carbon footprint of your electricity use (eg EV charging, heating/cooling) simply by using at the right time of day. GreenForecast.au uses simple-but-effective ML to predict the next 7 days of Greenness with good temporal stability, allowing you to plan loads for you or your business.

It also predicts wholesale power prices across that time.

I would be pleased to provide API access.

[0] Depending on state, time of year, etc. Scroll down for details in Q&A. Where I live, yesterday's worst time was 85% fossil fuels and best was 40%.

greenforecaston Ask HN: What are you working on this year?

In the current stage of our transition to renewables, the "greenness" of our electricity varies wildly throughout the day. I'm building a ML system to predict the next 7 days of "greenness" (and electricity price) in the hope that folks can reduce their carbon footprint, for example by:

- choosing a 'green' time of day to charge an EV

- 'overdrive' heating/cooling during times of high greenness, reducing usage during

- dream: a large scale user could take greenness into account when scheduling

- dream: the nightly weather report includes a green forecast

Feedback on idea/execution greatly appreciated!

http://greenforecast.au

greenforecaston Show HN: Predicting the Greenest time to use electricity (AU)

Sorry, not quite sure what you mean? Do you mean displaying the total power output (aka total demand) in MW rather than as a percentage? (In case that's correct, it's an interesting idea, I've been going back and forth on it, how would that info be useful?)

greenforecaston Show HN: Predicting the Greenest time to use electricity (AU)

Thanks to you! Much appreciated. I hope it can be useful to some people!

greenforecaston Show HN: Predicting the Greenest time to use electricity (AU)

Thanks! There's one big energy market operator in Australia, they publish heaps of data but it's a pain to process [0][1].

I'm no expert in US energy... Presumably each of your (many) markets will have different available data, but more-or-less equivalent. Probably do-able, but will need to be re-done several times over!

[0] https://nemweb.com.au/Reports/Current/ [1] https://aemo.com.au/energy-systems/electricity/national-elec...