If we want to mitigate the impact of climate change, we need to invest in decreasing Nuclear costs and building many more plants in the US.
Eg hydro, overprovisioning solar or wind, transmission to remove local weather variations, coupling wind and solar, demand flexibility.
Fervo just started its first full-scale new-gen geothermal plant, for instance; 24/7 firm power. You might like David Roberts interview with Tim Latimer about it: https://www.volts.wtf/p/enhanced-geothermal-power-is-finally...
I don’t know how far solar/wind can get us. But hydro sucks more than most people realize.
Adding PV means hydro can save much more of its water, just dispensing to “fill in the gaps”. This is already how the hydro in Norway and Sweden operates, you can see it daily if you look at the hourly power breakdowns by generation type.
But I agree, I hope we will get to a point where we can decommission the big dams..
That’s not quite how it works.
Consider the biggest hydro project in US, the Columbia river, with its 14 dams.
The system does fill up in the rainy season, but the crucial thing is that it’s not the dams that do fill up, but rather the whole watershed, meaning things like snowfall and ground water. This means that we have very limited amount of control over when we let the water through. We can’t just dam the river for an extended period: if we don’t use it for generating power, we must spill (waste) it. This means in practice that the dams are not batteries: to a large degree, it‘s use-it-or-lose-it.
To make these into batteries, we’d need to somehow refurbish the dams to tremendously increase the power generating capacity on each dam, so that instead of assumption of continuous flow (either through turbines or spillways), we make the flow more intermittent, so that we can make up for closed times by pushing more water through during open times.
This is tremendously difficult in practice: dams are simply not designed to allow for refurbishing with many more turbines or much more flow through them than they were originally designed for, the upstream reservoirs are not designed for quickly varying water levels, etc.
Point is, we can’t just “save the water”.
Yeah in reality there are two forms of hydropower: the classic reservoir dam, and 'run of the river' powerplants. You're absolutely right, even reservoir dams don't really fulfill the 'annual battery' idea, since they must maintain some minimum outflow for downstream consumers and can't shut it off entirely if it better suits the power generation goal.
The storage level at Lake Mead has an annual spread of ~100ft; the dam is currently at historic low capacity and could hold about twice its current level: https://www.usbr.gov/lc/region/g4000/lakemead_line.pdf
Reservervoir dams certainly can store water and dispense it when appropriate.
Like I said, this is happening every day already, you can see it in the hourly data on wind/hydro generation share, and you can see it in the annual storage capacity data for Norwegian hydro.
To put it in context: Columbia River basin produces 40% of US hydroelectricity. Out of 14 dams on Columbia River, 12 of them produce (each individually) more electricity than Hoover Dam, most by a factor of 3x or more. Very little of it can be turned into storage as of today, what does not get used must be spilled.
This is not to say that using dams for energy storage is a bad idea, it’s not. It just will not work well in practice with most of currently existing dams, at least without huge retrofits and/or screwing over downstream water consumers and upstream reservoir users.
There's a reason Russia blew up the Kakhovka hydro dam. The down-stream impacts of the flooding were more devastating than what they could reasonably accomplish with conventional weapons.
It killed <75 people.
They blew up the dam because it was also a bridge and removing it allowed them to redeploy troops to the east.
But the thing about dams is that they're over-engineered and massive. Even with high explosives it's difficult to demolish one. Hence why dam busting bombs were large (>4000 kg) and still required underwater detonation to boost their power [0].
It wasn't unreasonable for the Russians to expect Ukraine might be able to seize a bridgehead on the far side, rebuild an operable road over the dam, and then rush armor (and logistics) across it.
Much harder and more time consuming when there are no longer any load bearing remnants available.
[0] https://en.m.wikipedia.org/wiki/Operation_Chastise#The_attac...
It was, as much as it was unreasonable for russians - that's why they ran over to that side of river.
https://www.usbr.gov/lc/hooverdam/faqs/powerfaq.html
> Hoover Dam generates, on average, about 4 billion kilowatt-hours. And the lake covers an area of 1,495,806 acres
Rough estimate, 2GW worth of solar produces that much in a year. And requires I think 12000 acres of land. (Open to have phat fingered the calcs)
My take away is it's not even economic anymore and requires flooding 100 times more land then a solar plant.
edit: to add some context re the Hoover dam:
> Upon becoming Secretary of Commerce in 1921, Hoover proposed the construction of a dam on the Colorado River. In addition to flood control and irrigation, it would provide a dependable supply of water for Los Angeles and Southern California.
Ecosystem change =/= ecosystem destruction. The lake produced by a dam is a far more beneficial ecosystem to a much broader range of life than the river that preceded it
I like hydro but this is not at all what people I know working in this field believe to be true.
Hydro use is growing, especially in developing countries. However it’s a shrinking percentage of total energy generation. There is an absolute cap on theoretical hydro energy production, and it isn’t enough.
Hydro is low carbon and renewable. But it isn’t green, and it’s not enough.
Speaking of hydro, I do think pumped hydro-storage ought to be looked at a lot more for energy storage (esp. versus giant lithium-ion battery banks), especially as we transition to inconsistent renewable sources like solar and wind. I'd assume that creating new, isolated bodies of water wouldn't incur as much ecological damage as blocking off existing rivers or greatly increasing our mining of rare earth minerals)
Wind caps out at 42% with Denmark, solar at 15% in Australia. Many countries have nearly 100% of their electricity coming from hydroelectricity. Besides nuclear power, hydroelectricity is one of the few non-intermittent sources of renewable energy - sure, rainfall does technically make hydro intermittent in a sense, but it's not going to change output on a dime when the sun goes down or the wind stops blowing.
Well, you also have geothermal power, but that's even more geographically constrained than hydro.
I never said hydro wasn’t renewable. I said it wasn’t clean. Most people think of hydro as renewable and low carbon and clean. Two are true, one is false. Hydro is dirty and, imho, we should phase it out the same way we should phase coal out.
Nuclear isn’t renewable by most definitions. But it is low carbon and clean.
Nuclear was good 40 years ago. Solar is getting good. Wind I don’t know much about. Hydro is an ok bootstrap but is much dirtier and more problematic than most people realize.
https://web.stanford.edu/group/efmh/jacobson/Articles/I/Comb...
Here's a review article on 100% RE studies:
https://ieeexplore.ieee.org/document/9837910
"The main conclusion of most of these studies is that 100% renewables is feasible worldwide at low cost."
The reality is even assuming that we can not overcome shortages in solar and wind by overprovisioning and storage (and studies say otherwise), it does not make any sense to build nuclear instead of solar/wind as long as we are still running coal. We get much bigger CO2 reduction bang for our buck with solar and wind. Building nuclear would therefore effectively increase our CO2 over alternatives. This is especially true as nuclear plants have a relatively long ROI (in terms of CO2.
https://www.energy.gov/ne/articles/doe-report-finds-hundreds...
If you do them regularly, then sure they do. You already have trained workers, an intact supply chain and contact with the regulators.
Nuclear would be entirely unsuited for this task. Nuclear provides baseload, it doesn't fill in gaps. If you try to run the reactor intermittently to counterbalance an intermittent source the cost of its output increases massively.
For EU this is no longer the case. The difference in market price between low and high can be above 100x. In theory a power plant could earn as much in 4 days as an other plant earn in a year worth of power generation. This is why all those nations started to bailout the power bills of businesses and citizens last winter. A single month for some people costed more than a years worth of power. For companies with contract obligations, paying what ever the market demanded was the lesser evil.
The US is not in the same situation, but the energy grid there is still a market based one. There is also other technologies that could in theory compete in such volatile market.
Europe has enough salt formations to store many petawatt hours of hydrogen, far far more than would be needed.
Producers of green hydrogen are currently more interested in delivering green steel, which pays much better than hydrogen-burning turbines. The general idea is that this will in the future reduce prices down to energy grid levels, and a researcher here in Sweden working on such project estimated prices to drop to those levels around ~2060-2080.
This could happen much earlier if prices continue to increase as they do, but who knows. It would make for a good A/B testing to produce both and see which one was the cheaper option, and if the green hydrogen power plant fail they can always just produce more hydrogen for steel production.
Note, however, that as long as your grid is still burning natural gas for power, it doesn't make much sense to burn green hydrogen on it too. Eventually the natural gas will get very expensive (CO2 charges if nothing else), but for now price spikes are short term because LNG can be brought in (this is what Europe has done after Russia shut off the gas.)
Green hydrogen is going to have to be something that is produced, because the world uses about 100 million tonnes of hydrogen a year. Ammonia is an essential commodity chemical made from hydrogen. Some 6% of current world natural gas consumption goes to making hydrogen. These markets can and will be served by green hydrogen even before hydrogen is used for grid generation, and this will serve as dispatchable demand to help smooth renewable intermittency even without hydrogen being burned for power generation. And once there are large stockpiles of green hydrogen, it will be a small step to divert some of it for backup generation.
https://www.deseret.com/2022/5/9/23046910/green-hydrogen-ene...
The project will take that excess solar and wind capacity and through a process called alkaline electrolysis it will separate oxygen and hydrogen from water through 220 megawatt electrolyzers, producing up to 110 tons of hydrogen a day.
It is called green hydrogen because it is derived from renewable power sources.
...
The other “wow” factor of the project is the salt cavern storage reservoirs.
“Those salt caverns will be the largest single storage site for hydrogen, globally,” Ducker said.
He pointed out that the battery storage capacity across the United States sits at two gigawatt hours via lithium ion batteries. The Utah project will have storage for 300 gigawatt hours of energy.
...
The salt domes for storing the hydrogen will be 3,500 feet underground and will be as deep as the Empire State building is tall — about 1,500 feet.
Ducker said the caverns will enable long duration storage of energy and prevent monthly curtailments of solar and wind energy.
Mitsubishi delivered the gas turbine generators for the project a few days ago:
This[1] is the Australian energy market operator dashboard. Note the demand curve. It is not sudden in anyway - it is highly, highly predictable. Nuclear reactors can handle that sort of curve just fine - you roll the control rods in when it's low, pull them out when it's high.
The "inability" of nuclear reactors to handle variable loads is to do with the thermal mass of the reactor pile which can't be changed rapidly, but electrical load generally doesn't change rapidly - it changes very, very predictably at large scale.
Nuclear reactors can handle normal electrical demand flows just fine.
[1] https://aemo.com.au/en/energy-systems/electricity/national-e...
I wasn't claiming that nuclear power plants couldn't technically ramp up and down. I will happily stipulate that they could. I was arguing it was economically ludicrous to do so. That's because most of the costs of nuclear are fixed: capital cost, financing costs, fixed manpower costs. If you operate the power plant at low capacity factor, the cost per unit of energy produced increases inversely, just because these fixed costs are being spread over less output.
Nuclear either makes sense for baseload or it doesn't make sense at all. Trying to retreat to an application for which it isn't suited, like covering for intermittent renewables, is a losing game. There are any number of alternatives that would be much cheaper.
So in a standard 24 hour day, there is at minimum 6GW of capacity which is always available and always demanded. And for most of year that's 2/3rds of the total demand which ever applies.
The question you are not answering is whether the cost of building a proportionally larger nuclear plant - i.e. one which can meet the upper ends of this scale - is substantially more expensive then a smaller one.
The answer is pretty obviously no: nuclear plants are front-loaded in capital and construction costs, but their relative size has very little impact on the cost of building them, or their maintenance needs, or even fueling costs.
The reality is renewables haven't got anything on that except from an electricity market which mostly doesn't care about them. Start asking renewable generators to well you guaranteed kilowatt-hours throughout the year and watch the "cheap" power skyrocket in cost.
But many components of a nuclear plant have size that scales with power. There are some economies of scale, but they're fairly marginal. It's mostly to amortize fixed operating costs (like personnel) over more output.
but it hurts the economic efficiency massively, as costs for nuclear are almost entirely fixed at construction time
https://www.nytimes.com/2009/02/18/us/18nuke.html
> "The rule, approved by the commission in a 4-to-0 vote, requires that new reactors be designed so their containment structure would remain intact after a plane crash, cooling systems would continue to operate and spent fuel pools would be protected."
You can't risk a failure in the primary cooling system, and since reactors need active cooling in the event of a regional grid power failure just to avoid core meltdown, you need onsight power generation capable of running the cooling loop 24-7 (failure in this system led to the Fukushima explosions). These systems (from cooling loops to steam generators) are under constant stress and have relatively high maintenance costs (a major factor in the closure of California's San Onofre reactor).
Then you have to add in the cost of the uranium fuel rods, which is a complex supply chain issue in many countries (the recent coup in Niger has shut down 1/3 of France's uranium ore supply chain for their reactors, say news reports). Uranium supplies are limited and historically uranium prices get volatile when it seems a reactor boom is coming (look at right before Fukushima). Then you have the long-term costs of spent fuel treatment and secure storage, and eventual reactor decommissioning.
I really don't see anyway to reduce these costs such that nuclear will be anywhere near cost-competitive with today's solar/wind/storage complexes, that are entirely capable of producing reliable 24/7 grid power at costs well below that of a comparable nuclear power plant in most locations.
> How can costs be reduced?
Build more reactors, and re-learn how to build them. Note that this doesn't touch on economies of scale, which will likely never really apply to nuclear power. Nuclear is likely to always have immense up-front costs, but it shouldn't cost this much.
Stop building nuclear reactors.
https://ieeexplore.ieee.org/document/9837910
"The main conclusion of most of these studies is that 100% renewables is feasible worldwide at low cost."
"Even former critics must admit that adding e-fuels through PtX makes 100% RE possible at costs similar to fossil fuels."
pumped hydro is not a winner. the locations that could be used are few and far between and require massive amounts of water and wreak ecological nightmare on a wide area.
https://www.cityofelynv.gov/pdf/CityCouncil2021/cc1-28-21/Wh...
https://www.whitepinepumpedstorage.com/
(the whole thing could be sped up; that's a general problem in the US)
Electricity is a vital service: completely vital. Without it, modern civilization halts. It might be annoying being unable to make a cup of coffee, but municipal water and sewage need electricity to work. You go without power for a week, and the entire wastewater infrastructure will start shutting down. Refrigeration and food storage fails. Even backup fuel storage becomes a liability because you need electricity to pump it around.
So the question is, how low can you let the reservoir get? Because it's not about how long you could run going from 100% to 0% - it's how much of it can you use. And we have a model for this, in the form of another service: city townwater supplies.
In Australia, water restrictions go into effect when we hit <50% water capacity in the dams. That's the level at which usage cuts are applied to try and ensure we don't run out. At <40% we increase the severity. But this sort of resource exhaustion is also slow - we lose storage capacity over the course of months, not days.
And this is a resource which is dependent on electricity to supply (we also have a desalination plant, so we have some guaranteed capacity).
So within that context then - i.e. imagine you're planning a nation-state electricity supply, what are your risks? - how good does pumped hydro - or any storage-based solution - look, when your requirement is "the power cannot go off - ever". Put on your systems engineering hat, treat it like a software deployment - what level of redundancy and overbuild would you want when you're told "this is a mission critical, safety-critical system consuming an intermittently available resource". How much capacity and overbuild would you believe is necessary to have confidence, or even decision-making capability, when pressured?
This is a common problem in the anti-renewable arguments. You pick a particular design for an energy system, argue it doesn't work, then (wrongly) claim no renewable energy system can work. But to reach that conclusion, you have to show that no combination of elements can make a system that works.
It makes sense to have multiple storage technologies with different performance characteristics. You want efficient, if somewhat expensive, technologies for short term storage with large numbers of charge/discharge cycles. You want low capital cost systems for ultimate backup, even if those systems are not as efficient.
For example, one could back up the entire grid with combustion turbines burning an e-fuel like hydrogen. These are massively cheaper per unit of power output than nuclear. Because we are not using them very often, the low round trip efficiency doesn't matter much. You want guarantees this won't run out? Make the storage caverns larger. This is already what we do with natural gas -- we store a good chunk of seasonal demand and count this being sized large enough to not run out.
Why would hydrogen combustion plants - which don't exist at the moment, don't have turbines on the market, don't have a fuel supply pipeline - be cheaper then current coal fired powerplants?
So your cost of generation already is - at minimum - at least as expensive as a coal fired powerplant, in terms of fixed costs for maintenance (and investment - who's building these when they can't sell the power from them?)
There need not be any fuel pipeline, since the plant can be built at the hydrogen storage site. The electrolysers will be there also.
There have been industrial turbines that burn hydrogen for decades. It's not some sort of exotic technology.
Yes, there are many parts here. And it's still cheaper than nuclear. Nuclear is pathetic in that way.
But if you're electrical grid doesn't mostly have always on sources to backstop it, if you were all renewables and storage, then how much storage would you need to guarantee supply - 24/7/365 days a year.
Thanks for the observation, Mr. Obvious.
You also seem to be implying longer term storage of some form isn't feasible. If so, you are incorrect.
Note that I pointed to this project to debunk the falsehood that the locations for PHES are scarce, not to claim that PHES is good for long term storage.
Water use just has to keep up with evaporation on average.
I'm not sure why building a reservoir needs to be an "ecological nightnmare" except in the sense that it's a sudden change to an environment.
I wish this were true, but I haven't seen convincing evidence that it is. Up here in Minnesota, we heat our homes with natural gas. Once that's converted to electric, that's a _lot_ of energy to generate and store, and it has to be absolutely reliable for six straight months or you're talking mass death. Nuclear seems like a perfect fit for this scenario. I think it's a poor choice to take it off the table.
So having a stable source of cheap electricity that can be built in nearly any location is still a good idea. Even in a world with super cheap green hydrogen.
One can argue about costs, but costs are at least low enough to be viable, otherwise these countries couldn't exist as they do. You can say that the costs are being externalized to taxes or some other place, but these societies are being able to absorb these costs in aggregate. Nuclear might not be cheaper than gas and oil, but it's possible to build a modern industrial society with nuclear.
Now contrast with green hydrogen generation and battery storage, for instance. These approaches aren't working in country-level scales anywhere. We compare hypotheticals with systems that, despite problems, costs and limitations, are known to work.
Costs aren't the only problem, new nuclear reactors simply cannot be build fast enough to counteract the climate crisis.
Yet, France consistently produces half of Germany's CO2 per capita: https://data.worldbank.org/indicator/EN.ATM.CO2E.PC?location...
I would consider that a giant success of nuclear energy.
> A comparable number to Denmark that has almost no hydro power.
Denmark is not relevant - they import giant majority of their energy. Right now they barely produce at all: https://i.imgur.com/69SI5J9.png
> Costs aren't the only problem, new nuclear reactors simply cannot be build fast enough to counteract the climate crisis.
They could be build fast enough if we did it _seriously_, and not as vanity projects.
https://www.scmp.com/news/asia/article/2027347/south-korea-s... Is the suggestion to just import the energy?
In term of production a country can easily go above 100% renewable by selling a lot of it during periods of optimal conditions. Naturally, a country can not above 100% in terms of consumption. Denmark for example is a massive exporter in terms of production, but also a massive importer in terms of consumption and has a very large dependency on imports. They are not self sufficient despite producing more energy that they themselves consume.
You have the issue you talk about with all generation that isn’t load-following, including nuclear. France solves this by exporting subsidized electricity in time of low demand while importing in time of high demand, which is winter. In the surrounding countries like Italy and Germany you have gas-plants that jump in when needed.
That’s the thing we need to fix regardless of the power source.
That really casts doubt on your assertions. Nuclear carries significantly fewer risks than coal (which operates in the nuclear failure state all the time).
A Virginia class submarine in 2023 costs $4.3 billion (that's for the whole submarine), it has a 210MW S9G reactor.
The Vogtle Unit 3 was $17 billion over budget, for a $30 billion total budget, for 1,100 MW.
$204MM / MW for the US Navy vs $273MM / MW for Georgia Power et. al.
This project makes nuclear submarines look like a bargain!
Sources:
* https://en.wikipedia.org/wiki/Virginia-class_submarine
* https://apnews.com/article/georgia-power-co-southern-climate...
https://www.westinghousenuclear.com/energy-systems/ap1000-pw...
The naval reactor is more expensive per unit of output.
And that’s the optimistic scenario where it actually works well, scales, can be built out rapidly, and is not extremely expensive.