Thread by @GBruhaug

Since the myth of limited nuclear fuel supply is back in the zeitgeist due to some popular YT videos, I figured I would finally dive into the true insanity of nuclear fission fuel resources. Let's see just how long we can burn rocks! 🧵 1/23

So first and foremost, fission power is the process of splitting heavy atoms like uranium (but also neptunium, americium, and more) in a chain reaction. We get ~200 MeV from this reaction, which is a LOT! Millions of times more than chemical reactions. 2/23

The common fuel in use is U-235, which is 0.7% of natural uranium. The rest is U-238 but we can use that in a type of reactor called a fast breeder (normal reactors also use some but not efficiently). This turns non-fissionable U-238 into fissionable Pu-239 and is old tech! 3/23

In theory we could get 100% burn-up of uranium that way, but no one is anywhere close. Current (non breeder) reactors barely burn any of the fuel at all (about 1%)! That is why we like to call it spent nuclear fuel, not waste. It is mostly usable fuel! 4/23

The other fuel option is thorium, which is about 4X more abundant than uranium. Thorium can only be fully consumed in thermal breeder reactors, but this process has been shown at small scale before (see Shippingport). 5/23

Neither of these breeding options are done at scale today (although there are two Russian fast reactors making power right now, they are more prototype than anything), but they HAVE been done! This is not future tech like fusion, it just need refinement. 6/23

Now what does this all mean for how much energy we can get? Well there are a couple sources of uranium. There are the high grade ores we know of, the uranium in granite, the uranium present broadly in rocks and dirt and the uranium in the ocean. Same is true of thorium! 7/23

So there is 6147800 tonnes of U considered economically available. Keep in mind these are the ones we know of! As the "Peak Oil" scares have showed us before, things change when people look... This translates to ~924667 TWh thermal in LWRs. 8/23

That only gives us 2.6 years of making the ~40 TW I was thinking we would want to power everyone at a high living standard! And it gets worse if everything has to be made into electricity at 30-40% efficiency of conversion! Is nuclear a bad bet? 9/23

Thankfully we have tricks! First, let's remember that the reserve estimate is economic and assured sources. So there is probably a LOT more! But also if we burn this uranium in fast reactor and get complete burns we get 131.7 million TWh! That gives us 375.7 years! 10/23

Now there is also a roughly equivalent amount of known thorium reserves, so we can double that time frame just with easy to mine stuff. But what if we want to go longer? Let's get into the harder to mine things! 11/23

We know there is at least 20,000 tons of U in phosphate deposits, which nets us another year or so of energy from fast reactors. Much better then that though, granite is 3-5 ppm uranium and 12 ppm thorium! 12/23 core.ac.uk/download/pdf/2…
researchgate.net/publication/24…

This means that every ton of granite processed into nuclear fuel gives us the energy of 50 tons of coal. Your countertop has a higher specific energy than coal! The above paper estimates that using 1950s tech we could get about 1/5th of that, which is still great! 13/23

Granite forms a major portion of the planets crust and as such estimating total reserves are hard. Needless to say the answer is "a lot"! Interestingly enough the crustal average on Earth is about 2.8 ppm of uranium and ~6 ppm of thorium. 14/23

This means that any random scoop of dirt has more energy than an equivalent scoop of coal! There is a lot of dirt.... However, just saying there is a lot isn't as fun. Let's look at one huge resource that we know the numbers on, the ocean! 15/23

The ocean has uranium in it at 2-3 ppm and thorium at 0.5 ppb on average. Although there is more thorium on the planet than uranium, it is less soluble so there is more uranium in water. 16/23 helda.helsinki.fi/bitstream/hand…

We end up with around 4.2 billion tons of uranium and 700 million tons of thorium. Extracting and burning this in breeders gives us 111.97 billion TWh. At 40 TW average we would have 319337 years worth of fuel! 17/23

The really fun part though, is the ocean will continue to pull uranium and thorium from the soil to maintain equilibrium. Thus all of those trillions of tons in the crust will be available as well! In theory we could get great than 1 quadrillion TWh. 18/23

We would have ~4 billion years (give or take) worth of fissionable resources and at that point the sun will destroy the Earth, so running out is a moot point! Even cutting this in half due to extraction and conversion losses isn't that big of a deal. 19/20

Couple of things to keep in mind though. 1) We NEED to move to breeder reactors to achieve these amazing numbers. That is not future tech, but it is not most of the reactors we see either. 2) Unconventional (water, granite, etc) extraction will be needed at some point. 20/23

Luckily all of this stuff IS being worked on! The first breeder reactor was built decades ago and companies like @TerraPower are bringing new ones to the US. Seawater extraction seems to be making strides and even granite extraction has at least been looked into. 21/23

The biggest thing to keep in mind though, is that fuel being a limit to nuclear fission power is a myth! We have the resources and the know how to use them. Even back in the 50s this was known... 22/23

Here are some extra resources to check out and learn more. 23/23 world-nuclear.org/information-li…
sciencedirect.com/science/articl…
pubs.rsc.org/en/content/art…
whatisnuclear.com/blog/2020-10-2…