Solar analyst with BloombergNEF, goose keeper. Author, "Solar Power Finance Without the Jargon". All opinions my own. She/her.
Book (edition 2) available through all major booksellers, or directly from my publisher on https://www.worldscientific.com/worldscibooks/10.1142/Q0437
@hweimer hydrogen storage is *much* more expensive than batteries though, which is why you want to keep it for periods of very high power prices. Because you are only really filling and emptying your hydrogen storage tanks once per year.
Whereas you want to cycle the batteries as often as possible - outside deep winter, a day that dawns with >0% charge is a waste of capex spent!
(At least that is what I would expect. I did find the constraints on H2 power output surprising).
@hweimer also the way that in the 2040s you flip to making your money in the winter with storage, and at this point you are basically setting your own power price.
A suggestion: have long-term storage and batteries discharge at different price points?
@hweimer I have had a good time playing with this, thanks - it well illustrates the challenges of the Energiewende.
(Also I spent a long time wondering why my stored hydrogen wasn't substantially discharging, when I didn't have enough hydrogen power plants to burn it).
And some of the emerging behaviours are interesting, eg the way in a high-renewables system you flip to making all your money at night with your storage, not with the PV in the day.
They're never going to get that "National Laboratory of the Rockies" name to stick though, even if they actually want to.
It's been eight years and people still won't call us BloombergNEF.
NREL, sorry, National Laboratory of the Rockies' 3Q 2025 global outlook on the solar industry contains some nice data points with good sourcing (and no paywall, unlike most of my work).
Eg, the US got 8.2% of its electricity from solar in the 12 months ending August 2025.
Page 7 is a good summary about what's going on with integrating renewables in high-penetration markets.
@rachelplusplus yes, it's a very famous graph, I included it in my book for example...
*New build* in 2023 was indeed about 444GW, then 2024 was about 592GW (higher! But definitely not well over 600GW). So the second chart on wikimedia is accurately citing my team's work, and is smart enough to timestamp the estimate as well - useful since estimates do fluctuate a little.
I don't recognise the Info is Beautiful 'data' at all, it is highly suspect.
@mjd yes, I have one coming up for clients and can probably send it round a few weeks after pub. But warning, it's not nearly so dramatic, and right now tells a different story.
(It is also only BNEF data, ie my team's data, because that is only fair. The IEA were cowards but I mostly see where they are coming from).
It's annoying because this is a real thing, we have underforecasted (and the IEA has done so worse, and tbh I don't know the Energy Institute) and there is indeed a lot of solar, but this chart is not correct!
Information is beautiful only if it is true!
This one, from "Information is Beautiful", (infobeautiful@vis.social), btw.
RE: https://vis.social/@infobeautiful/116109430224379547
Hi, it's me, from the "BloombergNEF" cited in that chart. While directionally accurate, 2024 was not over 600GW (2025 was about 655GW but this chart is older than that). There is uncertainty around solar build data but not that much.
So it's a bad construction using incorrect data and I wish it would stop spreading around.
@hweimer hmm I am trying to get to 100% renewables though. So you do need storage.
Interestingly in 2040 I have succeeded in making the power price flip between 6 and 0.1 cents by setting my charge / discharge price at 6, which makes sense. It's nice to know this successfully models breaking the power market! Have got to 55% renewables.
@hweimer hmm shouldn't interest payments be tax-deductible?
I have been pretty cautious and so failing to get my annual renewables production over 29% by 2033, but I have the feeling this model is failing to screw over me personally for my bad investment decisions the way the actual market probably would?
@hweimer yeah, also borrowing money doesn't seem to reduce my tax bill as I would expect?
It's quite hard to keep track of really key metrics like realized power price here - I think the average price per kWh is falling over time, which makes sense, but I'm not sure eg how batteries are affecting my effective power prices. Will leave it running for a bit and see what happens in 2035 or so.
RE: https://fediscience.org/@hweimer/116096351442711561
Other people should have a play with this, I am supposed to be doing much more boring work (have already discovered that you need to keep cash to pay your taxes for the end of the year, or borrow sharpish).
@c_ozwei hmm you're right, not sure where 10GW came from. Need to be more careful.
@hweimer hmm this is cool and interesting but before I run this for hours and hours I'd like to know how it is calibrating the pricing. Mostly the power pricing for a solar-only system seems pretty high compared to realized solar prices on the German market, though the price of solar at 1.2 euros a W is really high too.
.... North Macedonia, man. It's got 1.8 million people and under 3GW (edited) of total power generating capacity. And plans to build 3,014MW of PV projects, plus 2,207MW/4,961MWh of storage, afaict this year.
Maybe the solution for net zero is simply big cheap batteries after all.
(Two years ago I would have been sure this was fake / unserious / I was misreading).
@leonoverweel @dr2chase humans are bad at thinking about costs that are a mixture of capex and opex. And when you have something that is high-capex (like batteries) the capacity factor becomes absolutely critical.
Then when you have hours of the year where demand and supply are very different, the economic maths goes nonlinear. I would like to say "...and here at BNEF we absolutely understand and model that" but, um, I think we have successfully nailed some specific cases and that's it.
@leonoverweel @dr2chase I spend my life telling people to be a bit cautious about how to treat seasonality.
Batteries are really big! Bigger than I imagined! I was just walloped with news of really big (>3GW) solar and storage (mixed- co-located and standalone) plans in North Macedonia and the Philippines, being built this year. But still seasonality is going to mean that batteries cannibalize battery revenues sooner than imo some people realize.
@leonoverweel @dr2chase yeah, I saw that report and while I don't think it is bad or wrong per se, it's a little misleading for somewhere like Birmingham where demand will peak in the winter as well.
Like, the 'baseload' assumption is okay for a thought experiment (we used it in Saudi Arabia because it's like a data center) but the actual best pathway to net zero electricity in Birmingham is a lot more wind-heavy, and if you pushed solar + storage to 62% you would force out wind.