[amarait]

What replaces fossil fuels is some kind of breakthrough in batteries. At the moment its getting better every year were currently at less than $100 per KWh which is crazy but needs to be improved for allowing more off the grid energy consumption

[pjc50]

The lesson from solar is that it won't be "a breakthrough" but the gradual accumulation of a thousand different efforts at cost-shaving across the whole supply chain making batteries gradually but inexorably cheaper.

There won't be fanfare when fixed batteries start using sodium chemistry rather than lithium, for example, but that will start happening across the next few years.

[yndoendo]

Batteries are just one means to store renewable energy and mechanical storage is another. Re-designing the power grid to transfer peak to areas via HVDC is another, to spread into areas where the weather limits or constricts the renewables for that time of day or day itself.

"Taming the Sun" [0] goes into more details and talks about it better than I can.

People like to over simplifying complexity by reducing arguments to a single reasoning. It helps make everything seem more simple than it really is. It is a way to persuade people that lack understanding "all systems are complex". Even instructions on how to construct a peanut butter and jelly sandwich. How many years does it take of development before a child can actually preform that "simple" task?

[0] https://mitpress.mit.edu/9780262537070/taming-the-sun/

[goda90]

Better batteries are the road to replacing fossil fuels for transportation, but I feel like abundant nuclear energy is what we need to give a jump start to green steel, hydrogen, ammonia, etc, and electrifying bulk heating industrial processes.

[dalyons]

you may feel that, but its not going to happen. nukes are too expensive, and thus the energy they produce is too expensive.

[pydry]

We can either pray and wait for a technological breakthrough that makes storage tech way cheaper than gas or we can just use taxes and subsidies to make it happen now.

It's not so hard. Lavish subsidies were used to make nuclear power semi-sort-of-competitive even though it's way more expensive.

The same thing could have been done with solar and wind but apparently we thought the best course of action was just to wait until they became cheaper than coal without subsidies (& then Obama and Trump slammed solar with tariffs).

[fastball]

Or we can go full nuclear.

[dalyons]

why? much more expensive, much slower. This reflexive "nukes are the only way" meme amongst technical folk really has gotta die.

[fastball]

More expensive in what way? "Cost" is what everyone quotes about why nuclear isn't great, but isn't the whole idea behind shelving fossil fuels and switching to alternatives due to downsides that are secondary to cost?

To me, renewables (solar and wind namely) have many more downsides than nuclear. So if we are doing things not because of cost anyway, why not nuclear? What do you fundamentally care about?

The power density of wind and solar is abysmal. You need to cover huge amounts of land with your preferred solution (which doesn't work everywhere) to produce relatively meager amounts of power. You need to have grid-scale storage solutions which are currently not priced in to the costs being quoted. Even if you have that storage solution you need to be significantly over-capacity in terms of production so that storage can actually be filled during peak hours.

Meanwhile, nuclear: requires a fraction of land use (good for ecology), runs continuously (so doesn't need huge storage outlays), can run basically anywhere (reducing transmission costs).

The most important note is that "nuclear" is not entirely encapsulated by existing Gen III reactors. There are many more designs and ideas that are being developed as we speak, whether more interesting (read: safe/efficient) fuel mixes, modular/micro designs, and various other improvements.

"Cost" is a merely a reflection of how much human capital is required to make something happen. I'd much rather spend our human capital on technologies that have the potential to massively increase the energy available to humanity, rather than focusing on tech which we know has strict upper bounds on power output / scalability. Solar and wind is useful in certain areas, but the idea that they can provide the baseload for a decarbonized future is ridiculous to me, unless your starting point is "I don't think humanity needs to consume much more power".

[dalyons]

in $/kwh.

We are in fact doing things soley because of cost, and pretty much only because of cost, because capitalisim. Solar and wind are now cheaper than all alternatives in most situations, so they are rapidly becoming all thats being built. We are doing the cheapest thing, which just so happens to be great for carbon, luckily for us. If we get out of this climate mess it will almost be by accident, because we made solar cheap, not because we chose to do the right thing.

Honestly you need to look into numbers for some of your points, and you'll see the folly. Land usage, its a non issue. For eg, its estimated that if around 1/3rd of the land the US currently uses for corn ethanol was converted to solar, it would power the whole country. And thats existing used land without talking about the insane amount of empty spaces that exist. non issue.

For storage, solar+24hr storage is now cheaper than new gas, and dropping fast (https://ember-energy.org/latest-insights/solar-electricity-e...).

Yes there are new nuke designs that are cool, but they're at least 10-20 years away from deployment at scale, by which time renewables and storage will be much cheaper still, and the transition will be mostly over. Im not anti at all, they're just too late, too slow and too expensive.

I think you need to catch up on developments in the last few years, and re-evaluate what seems ridiculous to you, a lot has changed very quickly. Cheap energy abundance via renewables is now a very likely outcome.

[fastball]

The fundamental disagreement here is (in my opinion) on what needs to be encouraged. If what you say is true and renewables are cheaper anyway, then also as you say capitalism should in out and that's what we will use in the near term anyway. So shouldn't we be investing more now on the things that you think are 10-20 years out, in order to accelerate them?

Becuase I'm interested in the future. The math with wind and solar checks out if all you care about is current energy needs. But we've already achieved most of the efficiency we can with at least PV. Even in a hypothetical future where you have some sort of quantum PV panels using MEG, your best possible hope is only 3x current efficiencies. But again, I'm more interested in our long-term future. Nuclear (fission and fusion) have much more unbounded potential than wind and solar.

Back to cost, the numbers in the article you link are cherry-picked. They rely on deploying solar to "the sunniest regions in the world" to get that performance. Most of the world is not the sunniest, unfortunately. Beyond that, the corn fields and insane amounts of empty space you mention are generally not co-located with areas of high power usage, making transmission another factor (which is doubly a factor since PV is such low-voltage that you require significant transformer infrastructure in order to step things up for transmission). So I strongly disagree, land usage is absolutely still an issue. There are also externalities caused by covering huge swathes of land with PV panels.

And it would need to be huge swathes of land, because in case it wasn't clear I would like to see humanity have huge amounts of power at our disposal – significantly more than we are using today. My back of the napkin map is that it would take 50,000 km2 of solar to accommodate current US energy needs. But I'd like to 100x our energy supply. That would require 5m km2, which is half the entire land area of the USA.

And honestly, I'm still skeptical of the price difference. PV needs lots of things (transformers, transmission, storage, disposal, land use, etc) that are frequently not priced in. Meanwhile the numbers quoted for nuclear fission reactors are frequently absolutely all in, including the cost of decommissioning the reactor at some indeterminate point in the future and pre-allocating funds for disposal.

tl;dr – your right that solar/wind is already quite cost effective and moving rapidly on its own pace just fine. So if anything needs collective support to me it is nuclear which has potential for the future that solar/wind just lacks.

[pydry]

could do but im not sure what there is to be gained from unnecessarily spending trillions more to decarbonize.

[triceratops]

Get private insurance to fully cover nuclear and I'm onboard.

[heisgone]

This might be the only way I could have any trust in Nuclear. I heard recently from a journalist that the Fukushima plant paid Yakuza owned newspapers to avoid negative press well before the incident. The technology is great, humans are not.

[conductr]

“We” are only in control of “us”. The rest of the world will keep burning fossil fuel

[triceratops]

> The rest of the world will keep burning fossil fuel

As the article spends so many paragraphs to explain to us, the rest of the world is increasingly not burning fossil fuel for their new energy needs. Most of the fuel it burns is for the energy it already uses. And solar is starting to take a bite out of that too.

[conductr]

That's great but if we had accelerated our transition, it wouldn't have lead to them accelerating their transition. The things are not linked or correlated.

[triceratops]

Solar has gotten cheaper and cheaper largely because as we build more panels and batteries, we find efficiencies. If we had accelerated our transition then the increased demand would have driven even more efficiencies and it would be even cheaper it is now.

If it were cheaper, developing countries would buy even more of it, accelerating their own transition.

[carra]

We are also going to need a breakthrough in how batteries are produced and disposed of. Otherwise the environmental impact of the many millions of batteries themselves may prove unsustainable too.

[tonfa]

That would be nice indeed, but shouldn't prevent us from prioritizing reducing CO2 emissions first.

[cogman10]

Maybe better disposal practices. Regulations standardizing batteries would also do a lot of good.

But really, we simply need a lot of virgin batteries regardless because we don't have enough. Recycling and disposal will only really take off once the market is mostly saturated (which we don't appear to be anywhere near).

I'd also point out that modern LiPo batteries are 90% recyclable with no special techniques needed. That's because by weight, the batteries are mostly iron and nickel. Recycling them is really as simple as just melting them down. It only gets tricky if you want to collect the lithium, silicon, and other trace materials (and there are already recycling plants that are handling that).

[triceratops]

Funny how no one seems to consider the environmental impact of digging up fossil fuels when they discuss solar.

It's similar to how you can identify Real Bird Lovers. They stay silent when they see pictures of oil-covered birds after an Exxon Valdez or a Deepwater Horizon. Show them a windmill and boy do they get passionate about bird safety and welfare.

[agumonkey]

there's also a lot of wastes, with different urban planning and build code a lot of cooling and heating would be avoided

[cogman10]

I think this is the most frustrating part. You wouldn't even need to massively change out neighborhoods to get a huge benefit. If a developer built with district and heating planned, you could have an entire neighborhood heated and cooled without needed AC and heaters in every building. You could still have single family homes that are massively more efficient per unit.

The problem is that has to be planned almost from the beginning. Which shouldn't be a huge deal. My neighborhood had a water tower built at the same time the neighborhood was built. There's no reason district heating couldn't have occupied the lot right next to it.

[1970-01-01]

Does not necessarily need to be a battery. Flywheels, heat, and even synthesis of fuels are also solutions to the problem.

[cogman10]

If you look at the energy density, cost/kwh, and lifetime maintenance of most of those, you'll find that batteries handily win. Further, batteries have room for innovation and growth in all those categories.

We won't, for example, make a more cost-efficient flywheel or heat storage. They are effectively as efficient as they'll ever be.

IMO, it necessarily has to be batteries. The other alternatives are nowhere near as good.

[qiqitori]

On the other hand, you need to buy a new set of batteries every 15 or so years. The other things you mentioned generally don't need regular replacement, and when they do, it's not the whole setup.

[cogman10]

You don't need to replace the whole setup, just the batteries. All the power electronics and interconnects are already there and will last as long as they last.

You also don't technically need new batteries almost ever. Batteries (typically) don't really die, they just lose capacity. After a 15 year runtime instead of storing 10mWh they now store 7mWh. That's still 7mWh. After another 15 years it'll be down to around 5mWh.

[EvanAnderson]

At grid scale I'd imagine that pumped hydroelectric storage would beat batteries for TCO, but there are significant geographic constraints.

[cogman10]

I mean, it is a literal pipe dream :)

Batteries can be deployed nearly instantly. My power company is planning on building a new battery plant next year, it announced it the year prior.

I know a pumped hydro plan that has literally been in the works for the last 20 years and shows no sign of actually being started (still being planned).

[bena]

I think the overall point is that we will never get there.

Renewables will never be cheap enough to fully replace fossil fuels, batteries will never be good enough.

No matter what, as long as the cost of extracting and burning fossil fuels is less than the result of what gets produced by the consumption, someone will be doing it.

It’s why crypto will never solve the energy issue. Why AI/GPT/LLM won’t either. Especially when the cost of that output is pegged to the cost of generating the above.

[toomuchtodo]

The price of LFP batteries in China is $35/kWh for cells, $52/kWh for fully integrated systems [1]. Roughly 1TW of solar is being deployed globally every year. We’re already there, it’s just a matter of pushing the pedal down harder. China destroys half a million barrels a day of global oil demand every year they build EVs at current output levels, which are still increasing.

Fossil fuels are already dead, it’s just time horizon. How fast we want to go is a function of how much fiat we want to shovel into PV solar and battery manufacturing.

[1] https://reneweconomy.com.au/watershed-moment-big-battery-sto...

[hnaccount_rng]

But that isn’t the case _today_. Unless you have existing facilities (which we do have and which gives fossils momentum) it’s strictly cheaper to build new renewables! The problem for renewables is that the p99.9 price is much higher than the p95 price. But finance will be solving that part.

[toomuchtodo]

Citation:

https://pv-magazine-usa.com/2025/07/01/solar-cost-of-electri...

[dalyons]

you're just wrong. they already are cheap enough, its already happening. PV+battery is now cheaper than new gas & coal. Very soon to come, existing gas.