Open Climate Fix

US Based Integration of OpenClimateFix work

I know primary efforts are focused on productizing all the OpenClimateFix efforts in the UK, but I’m curious if it’d be worth it to start parallel efforts in the US (or elsewhere), namely on the logistical front.

If I understand correctly, there’s a much better idea of how any OpenClimateFix work would get integrated into the grid in the UK relative to the US. I could see arguments for both sides, focusing efforts in one area (the UK) before trying to tackle other areas or trying to tackle both areas (UK, US, maybe others) in parallel. Has there already been a discussion about this?

I’m somewhat wondering because if it’s a resource problem, I’ve got some time and can start trying to look into / chasing down what would be required on the US front if that’s desirable.


Dude, it would be amazing if you could start looking into what would be required to get our work implemented on the US grid. Thank you for offering!

In general, we’re after maximum climate impact, so it’s definitely important to scale up globally ASAP. Although, as you point out, there are arguments for not scaling too fast.

My understanding is that US grids (and many grids around the world) would appreciate better solar PV forecasts, and better maps of where the PV is installed. But I haven’t actually spoken to many folks in the US electricity industry. It would be amazing to get some concrete leads into the US :slight_smile:

1 Like

Just to add on to the discussion about the US grid, I would imagine the forecasts would have to be split by market/ISO (which are much more complicated than state boundaries)
with my understanding being that ERCOT and CAISO (Texas and California) are the most advanced operators with the most renewables who would be the most open to an idea like this. (Edit: Useful interactive tool I found that maps out different US entities that might be interested in the forecasts)

Finding an operator who is open to our forecast would also be useful just in terms of where to focus our pv search efforts, as the US is very large. I think the best impact would probably be California though, just in terms of amount of solar panels in deployment (2018 data):

(Wind for comparison too because I want to rep Texas):

Also I’ve read a lot of material talking about California’s trouble with the duck curve due to solar, so I bet having an accurate pv forecast would solve lots of problems. (Also I wanted to post this duck curve pic)

@jack yeah, definitely want to avoid not scaling too fast. I think given the current state, it’s worth reaching out to people and figuring out who in the US might be interested in better solar PV forecasts and better maps of where the PV is installed, as well as how much work it might take to deliver something useful to them (it should be fairly minimal effort to just have the conversations). Before doing anything more serious in terms of resources / effort with regards to US based productization, though, it’s probably worth more discussion.

@TylerBusby333 great points! Also, LOVE the duck curve pic :laughing:. I agree that we should tackle one place, and it sounds like either CA or TX would be a great starting point (or rather, the markets / ISOs that is largely comprised of those two states). It sounds like CA might be a bit more receptive, but at the same time if we’ve already limited it to two states maybe we can just start with those two, and then further narrow it down in the future depending upon the leads we find?

I’ve got some contacts that I can reach out to here in SF and see if they could point me in the right direction, and a couple contacts in Austin (although @TylerBusby333 I bet you have more), but other than that I was planning on just googling places and reaching out accordingly. Before doing that, though, any thoughts / ideas on better ways, or people that would be good starting points?

1 Like

@sallamander I can think of one person I could ask! I’ll report back if they have any leads. But definitely reach out to your Austin contacts as well.

Also I remembered that there’s apparently 30GW of utility scale solar “in the pipeline” in Texas, but it’s unclear how many years the pipeline spans, and how much of it will fall through, but it’s good to know that nowcasting will become more important in Texas as time goes on, even if it’s still relatively small scale compared to California.

I forget where, but I also saw that utility scale plants have to provide their own power forecasts, which might be another opportunity to improve accuracy with open source models (if this wasn’t already being considered).

Not too many leads so far, I did get an interesting report on existing forecast modeling in ERCOT though, it looks like they buy renewables forecasts from UL. Here’s the full report.

Some interesting quotes from the report:

Great discussion!

Here are a few quick thoughts about contacts in the US:

Apparently ISO-New England (RTO) have said publicly that they would love better forecasts. It’s one of their top efforts. Apparently Eugene Litvinov (ISO New England’s Chief Technologist) and crew working on getting better predictions.

NOAA are really interested in using ML to better exploit environmental data. e.g. see the “1st workshop on leveraging AI in the exploitation of satellite earth observations & numerical weather predictions”, April 23-25, MD.

California Energy Commission are interested in solar production forecasting

NextEra might be interesting - they have lots of PV, I believe.


EnerNOC & Voltus are two demand response aggregators in the US - they might appreciate better PV nowcasts.

PV dashboard companies like DNV GL.

Electricity price forecasters like Genscape.


1 Like

@TylerBusby333 thanks for that resource, that was informative. It looks like they have reasonably good forecasts built out for their wind farms, but not necessarily their solar farms (although I would imagine at least some of the components of the forecasts for each are shared). Sounds like a good lead to chase down!

@jack sounds like you’ve done a lot of googling and searching already, that’s good stuff, thanks for passing it along! I’ll start trucking down the list and see if it doesn’t also lead to other sources / places to reach out to.

1 Like