Firstly nuclear also requires storage otherwise you're paying $400-600/MWh on account of the idle capital. Secondly you don't need to lead with storage, you can build out up to 50% or so renewable without.
Thirdly if the costs of battery were to halve or quarter, such as by the conversion of lithium ion to sodium ion production which is already happening, then the reasoning evaporates entirely.
Nuclear doesn't need storage, since it's power output is not intermittent. The amount of wasted capacity varies with season, but minimum demand is usually 80% of peak demand. In the summer, this increases and the peak demand occurs during the day. This is a good place for solar: rooftop solar mitigating A/C is a great application of solar. But to fulfill base load, which is the vast majority of electricity demand, it is an ineffective choice.
(Assuming you mean MWh here and it's just a typo.). Where are you getting those numbers
Lazard's estimates that I linked have solar at $130-$230/MWh when charging directly from the grid, and $85-$160/MWh when charging from attached solar which shares the same inverters. (And for additional duration, just add more batteries and discharge them at less power)
So if you blend 50% stored electricity with 50% delivered from primary renewable generation at a fraction of the cost, it's way below nuclear's costs.
Finally, we don't even really know that those are nuclear's costs. We haven't finished Vogtle, so we are at scientific-wild-ass-guess stage of how much it will be when it finishes. Go back 10 years and the estimated cost of nuclear is almost reasonable, but every single year that we have had Vogtle delays and cost increases, the general estimates of nuclear's cost goes up. We don't know when it will stop going up, until we have a few more reactors built. And when you add in the cost of potential construction failure, like at South Carolina's Summer addition, started at the same time as Vogtle, nuclear goes up even more.
People always overestimate the potential of nuclear, for some reason, and reality bites them. In contrast, people always underestimate the potential of storage and renewables, and reality delights them. It's time to reset our expectations based on the copious amounts of data we have collected over the past decade.
Why not just measure in MWh? Because storage is useless without means to put that energy back in the grid. The electrical transformers to do this are not insignificant. So 100 MW / 100 MWh is more than a quarter of the cost of 100 MW / 400 MWh because of this overhead.
Current nuclear builds have indeed been expensive, the plants last a long time. As per your own source, it's still less than intermittent sources when you include the cost of storage. Furthermore, Lazard is not examining the transmission infrastructure needed o support low-density sources like solar and wind. So, yeah, despite the expense nuclear is still the most cost-effective option.
Furthermore, Lazard's estimates about nuclear are actually quite conservative. Nuclear plants were cheaper when built at scale. During the 1960s through the early 1980s, nuclear plants were often built for less than a quarter of what present builds [1]. Nuclear doesn't need technological improvement to get cheaper, it needs the economy of scale that previous plants enjoyed.
https://www.sciencedirect.com/science/article/pii/S030142151...
Solar is not $30/MW in any part of the world. The raw module cost is $200/MW (gross) presently and might reach $30 in an optimistic 2050 scenario.
Solar LCOE for recent projects is $15-50/MWh depending on climate and land/labour cost.
> And 4 hours of storage is pretty thin - most plans call for 12 hours of storage. And storage costs are only projected to increase: https://www.cnbc.com/2022/05/18/ev-battery-costs-set-to-spik...
> By comparison nuclear costs $131-$204/MW, so it's still cheaper after storage. The Lazard estimate also didn't include the transmission expansions necessary to support the distributed nature of renewable generation (explained further here: https://www.vox.com/videos/22685707/climate-change-clean-ene...)
There's no reason to compare a peaking storage facility to a net capacity buffer one.
For the use case of 8-12hr storage you can use the same inverter and frequency matching equipment as the solar panels use. If transformers are a cost factor then why would you transform the energy three times at 4-6x the needed power load? Put it in a battery on site and decrease the size of your solar farm's inverter. You need to add an efficient buck charger for the battery but that will cost much less than the money you saved from the smaller inverter. At current grid storage prices of $260/kWh you need about 7 years. Give it a year for price spikes related to EV adoption to settle and you only need 3 years -- sodium ion batteries are well on the path to industrialisation.
In the counterfactual world where you're offering guaranteed $130-200/MWh for 60 years and guaranteed loans and free insurance building such a system from solar and battery would be immensely profitable,
> Solar LCOE for recent projects is $15-50/MWh depending on climate and land/labour cost.
Do you mean MW here? Capacity costs is nowhere near $15 / MWh.
> For the use case of 8-12hr storage you can use the same inverter and frequency matching equipment as the solar panels use. If transformers are a cost factor then why would you transform the energy three times at 4-6x the needed power load?
This is why storage costs include both the output and the capacity. 100 MW / 100 MWh is different from 100 MW / 400 MWh. Why would someone use a 1:1 output to capacity ratio? So that renewable builders can advertise "100 MW of storage", without actually specifying the actual capacity. In fact some plants have a output to storage ratio less than 1. I've seen 200 MW / 100 MWh facility before.
> In the counterfactual world where you're offering guaranteed $130-200/MWh for 60 years and guaranteed loans and free insurance building such a system from solar and battery would be immensely profitable,
I think you're misunderstanding that figure. It's the levelized cost of energy: meaning every $200 MWh of energy stored and retrieved, you have to pay $200, or $0.20 per KWh. Existing energy costs are less than that. In the US it averaged $0.16/KWh. Storage alone would amount to more than what we currently pay for electricity, for half the energy we use. And remember you need to add generation, and transmission costs on top of that.
Solar. LCOE. Joules put on the grid and money received https://renewablesnow.com/news/portugal-reaches-record-low-p...
> I think you're misunderstanding that figure
No, I'm saying if you offer me a contract where you give me $200 and I give you 1MWh of electricity distributed over the day proportional to demand. You guarantee my loans and I have free insurance. I can spend $30-80 of that on the energy (doubling for seasonal demand), $80 on storage and pocket $40. Then in 10 years I'll replace the worn out battery and start pocketing $160/MWh. I will also be able to sell my other variable MWh for another $15.
LCOS is presently much higher than current electricity costs, but you're not proposing current costs, you're proposing the deal the nuclear industry gets.
Then there's the cost of storage: how much does it cost to store 1 MWh of electricity, and retrieve it later. The total cost of generating 100 MWh of solar energy, storing it, and releasing it into the grid later is the sum of both of these: 1 MWh of electricity is $30-40 from solar according to the Lazard doc. Then storing it and retrieving it is another $160 to $279. So it's a round trip cost of $30-40 to generate 1MWh of electricity, and $160-$280 to store and retrieve it, for a net cost of $190 to $320. But only half that energy is getting stored and retrieved, so the effective cost of producing solar energy round the clock is $120 and $180 per MWh. And, no costs of transmission when you move energy from generation to storage - the Lazard explicitly call this out as a gap in its calculations. This has substantial overlap with nuclear power, especially when you consider the absence of transmission costs in the former and the fact that nuclear plants are ultimately steam turbines that can step into existing fossil fuel transmission infrastructure.
The main concerns here is that the effects of market forces should the world start attempting to build battery storage en-masse is going to drive up the prices of batteries considerably. Lithium prices are up 400% over the past year: https://tradingeconomics.com/commodity/lithium Only about a quarter of the cost of a lithium ion battery is manufacturing, the rest is raw materials [1]. The global battery production amounts to 455 GWh per year [2], and the US uses 12 TWh of electricity per day. Unless we figure out a way to make mining a lot more efficient, the exponential growth of lithium ion batteries is going to run into supply issues. And this isn't counting the knock-on effect of making EVs more expensive, and delaying transition away from ICE vehicles. We need to reserve our supply of battery storage for EVs.
By comparison, you just need to build 3 nuclear plants for every one that currently exists in the US to reach 80% nuclear and 20$ hydroelectric storage, the latter of which already exists. You could probably use solar without storage to help with peak summer daytime demand. The intermittency doesn't really matter as much because when there's cloud cover, temperatures go down and there's less strain due to A/C.
1. https://www.visualcapitalist.com/breaking-down-the-cost-of-a...
2. https://www.spglobal.com/marketintelligence/en/news-insights...
You said a solar facility plus a storage facility has a cost of $120-180 (this is not neccessarily correct as in some places you might need to store 70% of energy).
Then said you could not combine them (eliminating half of the conversion loss and shrinking the transforming electronics) for less than $200.
Additionally storage isn't increasing in price. Lithium is. And LiFePO4 is already being replaced for this purpose with sodium with prussian blue and carbon (which also does not need copper bus bars). Manufacturing hits GW scale next year, and price parity will follow shortly thereafter.
You are right in that solar is largely used without storage. That is because solar + storage is only marginally cheaper than heavily subsidized and insured for free nuclear and thus is unaffordable. It is also because a mix of solar and wind is a vastly better strategy for reducing emissions with limited resources than anything else.
You also haven't included electrolysers in your calculus (which have just started doing what the solar and battery market did over the last year).
Steam engines are obsolete. Throwing away 70% of your energy with a huge complex machine which wears out quickly due to operating conditions just doesn't work.
By comparison nuclear costs $131-$204/MW, so it's still cheaper after storage. The Lazard estimate also didn't include the transmission expansions necessary to support the distributed nature of renewable generation (explained further here: https://www.vox.com/videos/22685707/climate-change-clean-ene...)