Grids need reliability. Battery tech is not close to providing long enough durations to fulfill load with just renewables. I would much rather have nukes than natural gas filling that role.
This is roughly 1980. We are now in 2025. Battery tech has been increasing massively and besides that we now have HVDC tech which can transport very large amounts of energy from end of the country to another without significant losses. The whole 'baseline power' argument is getting really long in the tooth, it is mostly a matter of dogma at this point. It's not 'nukes' or 'natural gas', it's 'what is the best possible mix for this current moment in time.
The problem with nuclear power is that it is so expensive to be on standby that you need to buy their output even when you don't need it. So energy market pricing tends to be dominated by the least effective sources rather than by the most effective sources. If nuclear plants were left to fend for themselves they'd be out of business in a year. More so if you consider the cost of decommissioning.
Well, the race is definitely on. But another couple of years of reduced costs for solar and wind deployments and it may well be that nuclear projects underway will end up being cancelled before construction is complete.
One really nice thing about nuclear is that the fuel is highly portable. Small reactors next to datacenters take away a lot of complexity; transport, grid connectivity, etc. Plus they're already being built in industrial-ish areas.
Very long duration storage in my opinion is going to be thermal.
Standard Thermal's approach seems very simple and promises to deliver 365/24/7 heat (ultimately sourced from PV) at 600 C for cost competitive with Henry Hub natural gas. It's difficult to see how nuclear competes with that.
There is a very interesting tech using superconducting loops to store power. I think that that will be the 'battery of the future' but it is going to take a while before that sort of thing is safe enough (and cheap enough) for things like vehicles. But for stationary short term 'ride through' situations it has already been deployed, and also for stabilization of the grid in the presence of fast fluctuating loads or generators.
Superconducting storage is inherently short term, as the capex per unit of energy storage capacity is rather high. I doubt it's competitive with batteries even for diurnal storage. It might have niche uses, for example smoothing demand (on time scale of hours, not days) from intermittent high power users like electric arc furnaces.
Batteries work fine right now. Much better than nuclear plants that will take another decade to build. There's probably a few dozens twh of batteries being deployed in between. They retain their charge quite long too.
It's more an economical tradeoff than a technical one. But mostly, people doing off grid setups manage fine in some parts of the world. Winter months are challenging at higher latitudes. But there are also wind turbines and cables as alternatives. Cables allow you to import/export power from further south or further west or east. It's one of those things that might be a lot easier than building a lot of new nuclear power plants.
The issue with gas (aside from emissions) is that gas turbines are a bit hard to get. The lead time for getting new ones is pretty terrible and companies making them aren't exactly eager to make risky investments in extra production capacity for turbines that may or may not be needed in a 5-10 years. And once you manage to get them, you have to supply them with gas; which is expensive. Gas prices fluctuate a lot.
If you look at this through an economic lens, the next ten years will see:
- Hundreds of GW of solar deployed. Every year. Still growing. Probably quite a few TW of capacity added.
- A bit less wind turbines but still quite a lot. The US seems to have some irrational anti wind mill sentiment currently. But that might change. Huge potential there for upgrading old ones too. And offshore wind obviously.
- > 3 twh of battery produced for use in transport, grid, and domestic storage.
- Some new gas plants coming online here and there. But not a lot. Gas usage will grow but renewables will out grow it.
- Probably growing number of GW of cables and long distance cables. Possibly even cross Atlantic.
- Lost of coal plants being decommissioned (just too expensive). Offsetting most of the gw added by gas.
- A sprinkling of nuclear plants coming online. Mostly in Asia. If it's not approved yet, there's no way in hell many will come online before 2036 in the US/EU. A few GW per year.
The big picture here is that if you need a lot of power, you want low cost and feasibility. Gas is feasible. If you can get the turbines. But it's not cheap. If you want cheap, you mainly need to decide on the number of gw of panels you can afford to buy. And where to put them. Or wind. But the point is it's very feasible and quick to execute once you decide on a budget. Battery is not cheap (but still dropping in price year over year). And you may not actually need infinite buffers. You can fall back to gas when needed but as little as you can get away with (because $$).
Any AI/data centers still in business in ten years will probably be competing on energy cost. That's likely not going to be gas or nuclear mostly.
reply