The New Nuclear Age is Coming

Since I filmed this video, a number of THINGS have happened, including the re-election of Donald Trump. One thing about Donald Trump you never really know what he means by anything he says, but my overall analysis is that this probably slows this down because one thing that is bad for very large-scale projects is a LACK OF LONG TERM CLARITY, and having a couple of knuckleheads threatening to eliminate a third of the government isn't great for long-term clarity.

I don't think it changes the politics or the science, but regulations and approvals are the biggest concern here, and unless someone is actually focused on that, it's going to get kicked down the road.

But things do continue to move...here's a bunch of nuclear energy news just from the last week:

[Fryguystudios]

I'd like to add that Fukushima wasn't just the result of a massive earthquake, but the operating company deliberately ignoring government regulations. The Onagawa powerplant, which was closer to the epicenter, took both the earthquake and tsunami head on and even served as a refuge for survivors.

[LeakyTrees]

Hank’s inability to immediately know which month is which, and his anger at Rome for fucking the months up, is something I relate to so deeply in my soul

[Lawsonomy]

I am so glad to here him mention Geothermal power. As a chemical engineer I get so disheartened that Geothermal gets left out of the conversation. Even among clean energy activists and science popularists. It's clean, it's easy to maintain, there is enough of it to last BILLIONS of years (even with ever increasing demand), the main problem is the initial amount of money to lay down is so large. It is the perfect example of something that needs government subsidy, but no one in power has even heard of it!

[tjg]

Quick note about Microsoft restarting TMI, they are NOT starting the partially melted down reactor (unit 2) back up. They are restarting the other reactor (unit 1) that continued to run until 2019 when it was shut down due to lack of economic support from the Pennsylvania government.

[matthewsteve]

As a Canadian, I can confirm that I am, in fact, imaginary

[alang]

Hi Hank, Great video, but I wanted to comment on some things you mentioned as a nuclear engineer myself:
First, is that there have been multipe nuclear accidents, not just the big three. Those are only really talked about because of their impact on the actual population and environment, but nuclear is not quite as squeeky clean as that (absolutely still better than everything else). Davis-Besse had a very bad near-accident, and early accidents like SL-1 are actually rather gruesome. We have learned our lesson over the years and it is much better than in the past, but still not perfect. Its also important to note that a lot of accidents happened due to improper training or a clear disregard to regulations/procedures.
Second, is that coal power plants can defintely be repurposed, but not nearly as easily as it seems. Coal turbines, for example, spin at much higher rates and pressures than nuclear turbines, so they absolutely cannot be reused no matter what type of reactor is used (BWR/PWR). They are also not up to the standards of the NRC. Switchfields and cooling towers are two great examples of things that can be reused, but they will definitely need to be looked at to see if they are still suitable; its not a 100% guarantee that they will be reused. But the most important thing is the land. If a company can get away with reusing the land, it can definitely be enticing, but nuclear plants need a lot more of it to store their spent fuel and as a security radius around the plant.
And lastly, nuclear fuel itself isn't the issue with proliferation, its the reprocessing. About 95% of a fuel's uranium is U-238. When struck with a neutron, it transmutes into Pu-239, which is the stuff used in nuclear weapons. Countries can absolutely reprocess spent fuel into weapons, and is the reason why reprocessing is banned in the US. So China, if they are building a lot of nuclear reactors, can absolutely be able to produce more weapons and faster with more nuclear plants, since they will be able to produce more Pu-239. Reprocessing itself is not a bad thing and France is famous for reprocessing a lot of their spent fuel, but its only a good thing when its not used to potentially end the world.

[TKHaines]

The university I work for uses geothermal power, so it was a surprise last week when the power went out.
Turned out a squirrel knocked out a transformer or something like that. Squirrels, they will be our downfall.

[Blakexyn]

It's so sad that projects like these only get the ball rolling when we fear our competition getting more advanced than us as opposed to doing it just because it's better (for both our power needs and the environment).

[reaganharder]

Hi, certified Canadian here. Not only Canadian, but Saskatchewanian. Saskatchewan is quite motivated to promote nuclear around the world, as we have some of the largest and best uranium ore deposits in the world. Saskatchewan is also currently looking at deploying SMRs into our power grid.

[liamgauge]

God Hank is such a special creature. He is incredibly intelligent and the single greatest science communicator I've ever witnessed and yet his struggle to figure out which month is the 11th (The month it is currently) and watching him count up the months from January was just too wholesome.

[goodgamer-yfil]

This is an oversimplification, but the reason that Nvidia chips are good for AI is down to the kind of calculations needed for AI models. Graphics chips (the kind Nvidia makes) are good for a very high throughput of (relatively) simple calculations. They are designed to do lots and lots of simpler calculations in parallel, because that is what is needed for rendering graphics. And it just so turns out that this kind of computation is used a lot for AI. Obviously, it's more complicated than this, but that's the general idea. If you want to understand more there are lots of good sources on this topic.

[emu]

Something to remember is that the calculation for nuclear power highly depends on where you live. Here in Australia we are far better off going with renewables and grid level storage (e.g. batteries and pumped hydro) because we get significantly more sun than most of the USA. One of our southern states (the ones closer to the south pole that doesn't get as much sun as the more northern parts) actually produced 107% of the entire state's demand via roof top solar just the other day. All we really need is for our government to pull its thumb out of its proverbial and actually get on upgrading the grid to support the generation and storage of rooftop solar.

That said, one of our political parties is dead set on not only stopping the build out of renewables but rather just building a building a bunch of nuclear power plants which, even by their most optimistic timelines, won't even have the first one online until 2034...

[splashybard]

As far as I know, the issue with putting a nuclear plant into an old coal plant is that the coal plants are far too radioactive to be certified, because coal is slightly radioactive, and the laws governing nuclear plants are so strict.

[avmtg]

I'm surprised it wasn't also mentioned how much energy usage will go up due to trying to keep people comfortable through more intense weathers due to climate change.

[gigabyte]

I'm still sceptical. Hear me out.

First: there are other solutions to the mismatch between renewable generation and demand.
- For intra-day arbitrage, I see enormous potential from V2G. Hyundai/Kia, for one, are already selling EVs with bidirectional on-board chargers, which only need a software update to become V2G-ready. California is already considering whether it's possible to mandate that all new EVs are V2X-capable. At 7kW peak each, you don't need a high proportion of EVs on V2X to give the grid serious support.
- For inter-day storage, there's a technological arms race already underway. Compressed/liquid air, high-density pumped hydro, gravity storage, iron batteries, the list goes on. There will be winners and losers, but there's a lot of interest and demand for the winners to capitalise on.
- Inter-season storage is genuinely very tricky. This may be a use case for hydrogen - the round-trip efficiency is rubbish, but if demand fluctuations through the year are high enough, it can still be profitable, particularly if there are other markets for that hydrogen (e.g. industrial processes). I've seen some interesting-looking thermal energy storage concepts, similar in efficiency, using sensible heat in rocks or latent heat in scrap aluminium.
- The better solutions for inter-seasonal fluctuations, though, are interconnects and diversity of supply. Solar is terrible in the winter, but wind productivity increases (offshore, especially). Spread the wind farms over a large enough area (say, an entire continent) and the magnitude of the supply fluctuations goes down. Hydropower is more productive in the winter and, while a lot of the low-hanging fruit has already been picked, I think there's still a lot of potential for run-of-river projects. There are new ideas for geothermal coming through, including projects which extract lithium salts from the geothermal water, adding another revenue stream to projects that might otherwise struggle for profitability. Tidal power still hasn't hit its stride yet, but it's currently heaving itself out of the valley of death, with more tidal stream projects in the pipeline than ever before. Wave is, like, a decade behind tidal, but there may be more opportunities to combine it with floating wind. Combine these all together, with a bit of storage, and the darkest days will no longer be quite so dark.

Second: I'm still not convinced that the cost of nuclear will go down. I haven't seen any compelling evidence that SMRs will actually be cheaper than conventional nuclear, particularly while every country has their own little player they're holding a torch for. I see them as being useful for reducing the risk of complications and over-running on time and budget, and possibly useful as a route to develop new technologies faster, not for reducing overall cost. I'm prepared to be surprised by cost reductions from deregulation, but I'm not holding my breath. Thorium is one of many technologies that has promised a lot, never to deliver (I see more potential from the molten salt design, once a proper demonstrator has proven out the claimed benefits). Speaking ideologically, too, I'm concerned about high-level, long-lived transuranic waste. As a species, we've done a very good job of sticking our head in the sand and ignoring this problem. We need to either reprocess it for fast reactors or actually start building long-term geological storage for it.

Third: I'm convinced the AI bubble is going to burst. Oh, don't get me wrong, that generative AI genie is out of the bottle and isn't going back in, but I'm 100% convinced there is a financial bubble and it's going to burst, probably in the next year or so. There's too much money in it for results that currently don't go far beyond curiosities. This meteoric rise in energy demand isn't going to keep going forever. I don't know what will cause the AI bubble to burst - I hope it's not a humanitarian tragedy, but I can't really see anything else doing it - but I'm 100% convinced that it's going to happen one way or another.

Finally: the nihilist in me fails to believe in democratic government and capitalist greed sticking to a plan that needs commitment and investment for over half a decade before seeing significant benefit. Chinese dictatorship, sure, absolutely. The last decade or so has proven that a 'stable' dictatorial government can achieve long-term goals exceptionally well (while trampling on its people). But the West - investors, politicians and the electorate - have grown as greedy and self-obsessed as ever and this past month has proven more than ever that, given the choice between optimistic delusion and harsh reality, the dream world tends to win.

Footnote: I'm a career scientist/engineer, PhD, in the UK, as well as a clean energy ultra-nerd. I've worked in power electronics and motor drives, and am now back at my old uni, running its cleanroom. Oh, and I also once had an interview with a nuclear fusion company that taught me just how far away fusion power really is. I've seen, from the inside, the West give up on semiconductor manufacture over the last few decades, then be caught with its pants down after COVID. I've seen two nuclear projects run wildly over time and over budget, and I've seen politicians still insist that nuclear is the future while Hinkley Point C and Sizewell C continue to drink taxpayer money. I've watched our gird's energy mix like a real anorak, through wind power gluts that nearly gave us a net-zero grid and through weeks-long lulls. My opinion is not unbiased, but I hope it doesn't come across as ill-considered.

[carterjanssen]

Albertan here. Can confirm Canada is not real. It's just 20 resource extraction corps in a trenchcoat.

[GTaichou]

Hi John! As someone who previously worked a nuclear power facility, I am so happy to see this video!
One VERY IMPORTANT note; the isotope of uranium used for energy production is DIFFERENT from the one used to make bombs. The enrichment process is different. This is one of the biggest hurdles with public opinion we dealt with (most of the signs we saw from protestors after Fukushima Daiichi seemed to think we had bombs on site and we very much did not)

[Sagealeena]

Here in Australia we’re dealing with issue of oversupply, we have so much solar power that when demand isn’t high we just don’t need other power from the grid. Solar power is still greatly increasing, and we’re starting to get off-shore wind, so it’s really about creating more demand at times when there isn’t current demand (eg. Manufacturing on the weekend during the day, or timing appliances to run at night). In many ways we’re realising we don’t need baseload power and we can instead focus on the demand side.

[darknewt]

Problem with reusing infrastructure at a coal-fired plant: Coal stations aren't shutdown overnight. They are run down over years. And during those run-down years, the maintenance schedules of the infrastructure are wound down too - it's part of the business case. This wind down would have to be prevented, otherwise everything up to the power lines themselves (and maybe even some of those) would have to be replaced.

Another one: Coal-fired power stations don't need the sort of secure perimeter a nuclear plant needs - there are likely to be a lot of coal-plants that are simply too close to other properties to expand their boundaries enough to make them nuclear secure.

I'm not saying it makes no sense to repurpose some coal-plants, but it isn't a no-brainer by any stretch.

[sabrinaoxford]

As a Canadian who lives in Ontario, I always find it wild that not more places use nuclear. The majority of our power is nuclear or hydro. I believe we have 3 nuclear power plants and we rely on hydro so much that we refer to our electricity bills as hydro (very confusing because you would think that means our water bills).