Nuclear Power Still Doesn’t Make Much Sense
nytimes.com
nytimes.com
Is Nuclear older tech? Light Water Reactors (LWRs) are from 1950s and used by the Navy. In the same time, Molten Salt Reactos (MSRs) were developed by Alvin Weinberg in the Oak Ridge National Labs (ORNL) and proved very successfull.
Unfortunately, the US Gov't (Nixon) invested in Plutonium/Nuke tech and MSRs weren't developed further.
Not all energy uses are electrical. MSRs can also be used on thermal processes (above 700 degrees Celsius) and thermal energy is crucial to dozens of industrial processes, currently moslty feeded by coal and gas. Current light water cooled nuclear plants just do at 350 degrees and requires tons of protection to contain gaseous water pressure (> 150 thousands Atm). MSRs can operate at 3 Atm and higher temps => low explosion risk.
Radioactive waste: France is estimated to grow long term waste from 140,000 cubic meters (2019 - data from Statista) to 189,000 in 2040. Thats 2,200 40ft containers to 2,900. A 700 container size change in 20 years. All their current long term waste fills 11% of Evergreen, the infamous ship that clogged the Suez Canal. Compared to coal ash deposits, it is nothing.
Current nuclear waste retains 97% of usable energy. With MSRs, it is estimated to reduce it to under 5%, with half-times around 50 years (very short). Imagine the waste reduction after recycling all that into those reactors.
The most important component in solar panels is silicon with 99.999% purity. That consumes a lot or energy to produce. And China produces lots of them... fueled by coal plants. Nice for a carbon-free tech.
Concrete is a carbon intensive material but you´ll use a lot less in a nuclear plant than a hidropower one.
Concrete, aluminum, fertilizer are all thermal energy intensive industries and are carbon intensive with current tech. What do you think where their carbon footprint goes if you adapt them to use some nuclear thermo-electrical boost?
If we had never built nuclear power plants or wind turbines or PV fields, maybe trying to get a handle on the cost by such indirect arguments could be useful. But why resort to such an unreliable methodology when we have the actual cost data? We know that nuclear is considerably more expensive than renewables, which is why nuclear is shut out of new construction in the US. Even around the world, renewables are dominating nuclear for new construction.
You're still just pointing out categories where renewables win.
Glassless pv panels are available to anyone. Consider the Barrel 375W panel at 5kg.
That's a power density of 18W/kg which is about on par with the mass of steel in most nuclear reactors and dwarfed by the mass of concrete.
In a 25 year lifetime it has an energy output of 14GJ/kg. Compare to about 200GJ/kg of uranium fuel at typical burnup and thermal efficiency. It looks favourable for the uranium until you consider you have to have a huge amount of safety and shielding equipment around it. So this is fine unless you are arguing that LWRs have insufficient energy density?
Solar also requires a little bit of silver and copper, but that is vastly easier to get than all the exotic elements required to keep a nuclear reactor running.
> The most important component in solar panels is silicon with 99.999% purity. That consumes a lot or energy to produce. And China produces lots of them... fueled by coal plants. Nice for a carbon-free tech.
How do you think that uranium is mined? This is captured hy eroi which is in the 20-100 range for current generation panels.
I’m moving on.
Source: https://www.cjr.org/analysis/farhad-manjoo-nyt-unplug.php
Solar and wind require comparably MASSIVE amounts of raw materials because of their low power densities, and overall deliver a massively lower energy return on investment. Because of their intermittent nature they will also have to be paired with storage tech requiring still more raw materials and/or be built to a capacity which is far in excess of baseload requirements.
All those input raw materials are mined, processed, and finished using, by and large, fossil fuels. Ergo the perceived low cost of renewables is actually based on the relative low cost of fossil fuels. And that fossil fuel price point which is rapidly rising under supply shortages.
Moreover, EROEI is also a misleading metric. A technology with an EROEI of 100 is not 10x better than one with an EROEI of 10. The better metric is "energy after self-use" (say), which would be 1-1/EROEI. 0.99 is just 10% better than 0.9. This would be a minor advantage that could be easily swamped by other considerations.
Solar and wind require raw materials (for at least some materials), but nuclear requires much more labor. After all, nuclear is demonstrably more expensive than renewables; those $ have to be going somewhere.
I will add that solar actually uses much less concrete than nuclear, if mounted on the ground with steel earth anchors. Concrete production produces CO2 even in a fossil-fuel-free economy.
This is highly debatable, and very subject to change. Energy is the trump card of currencies, as the international currency market is now clearly demonstrating.
Let's see what that renewable cost comparison looks like when the renewables that were previously being subsidized by cheap natural gas look like a year from now.
This is not to say that solar and wind cannot have a meaningful place in the energy grid of the future. They absolutely have a place alongside of reliable base load generation.
But it is a pipe dream to assert that we can mine enough raw materials necessary to produce the wind and solar generation, install them (often in less than favorable geographies), install their requisite energy storage tech, in order to make a meaningful dent in CO2 emissions anywhere near as quickly as a renaissance in nuclear power would offer the world.
Solar module production is 240GW and wind is 100GW and both industries are growing at double digit percet yoy. Even at current production rates that will replace 15% of current fossil fuel energy before one new reactor like hinkley C will be built. At projected growth rates we'll see 2-4% of current primary energy in new capacity per year. The limiting resources (Silver and Copper) consume 2-5% of worldwide production.
Compare to the exotic metals like niobium, hafnium, beryllium and so on needed for a nuclear reactor which cannot scale to 100s of reactors a year.
Building a bunch of LWR plants won't solve the problem though because at those burnup rates world uranium reserves will only last around a century (and require just as much of that mining you complain about). Breeder plants and enrichment facilities will take even longer and cost even more.
> But it is a pipe dream to assert that we can mine enough raw materials necessary to produce the wind and solar generation, install them (often in less than favorable geographies), install their requisite energy storage tech, in order to make a meaningful dent in CO2 emissions anywhere near as quickly as a renaissance in nuclear power would offer the world.
Wind turbines can be built from abundant materials, but there is plenty of copper for current and future production (the same copper a steam turbine would use). Silicon cells are made of sand, Cu, Al, Ag, and Fe. There is only around 1g of silver per kW (and this proportion is decreasing faster than production is increasing) and a kg or so of copper for wiring -- most of which can be subbed out for Al.
All this nuclear shilling is just directing energy away from the thing that will make fossil fuels irrelevent and helping the coal plants stay on for a few more years.
All nuclear plants rely on government subsidy and externalization of decommissioning costs. In other words -- they're not on the balance sheets. What legit business gets away with that?
The reality is that nuclear power is a 1950's technology -- one from the past, not the future. We never fully understood the implications of running or taking apart a nuclear plant at commercial scale and instead of addressing the real problems, we just pretend they don't exist.
It's time to put the effort and expense once reserved for nuclear into something that doesn't create waste that lasts for 240,000 years (plutonium) or require supervised swimming pools to store (used fuel rods). We need power plants that can be turned on and off quickly -- at small scale -- not ones that require 20 years to build and 50 to tear down.
The time for nuclear is over... and it has been for a while! We've just been in denial.
It's certainly very economically dubious, though.
Lots. Name one business that has compensate for all of the CO2 their activities have produced when they are going out of business.
The old generic argument "nuclear because CO2 is bad" doesn't work, because there's now a superior non-fossil alternative. Nuclear rested on its laurels, its supporters thinking it was the only alternative when fossil fuels had to be abandoned, and it's adrift now that it's no longer entitled to be fossil fuel's successor.
Using 2011 weather data and 2030 cost assumptions (some of which are already met), we get 71 euro/MWh. That's much better than what Flamanville 3 is going to be delivering. Inclusion of hydrogen reduces that cost from 93.1 euro/MWh. Many arguments against renewables assume batteries are to be used instead for long term smoothing, which is a poor choice.
2030 cost assumptions are appropriate because any new nuclear plant started now isn't going to be available before then. Ignoring the cost trajectory of renewables and storage is another common mistake of nuclear advocates, a mistake that nuclear investors cannot allow themselves to indulge in. Indeed, they likely are looking even further ahead, as that nuclear plant has to operate until 2070 or so.
The companies that operate nuclear power plants usually pay into a fund that finances the decommissioning no?
> The energy source that was supposed to be "too cheap to meter" turns out to have a net negative energy value
This is obviously nonsense. Nuclear plants produce far more energy than it takes to construct and manage them, for decades at a time. Nearly all of France's energy comes from nuclear; if this were true they would need to import vast amounts of fossil-derived energy just to cover the costs of building the plants. Completely inane. Obviously nuclear power makes power.
> When you add together the true costs of building, running, and decommissioning a nuclear plant it comes to more than you can sell the electricity for.
I'm highly skeptical of this claim - like the above, it doesn't pass the order-of-magnitude sniff test. But even if it were true, it wouldn't necessarily tell us much in the current world where the energy market is distorted by cheap fossil fuel power, which is both a) subsidized and b) under no requirement to pay for its own cleanup, the expense of which would be prodigious.
>All nuclear plants rely on government subsidy and externalization of decommissioning costs.
That's not true, many pay into their own decommissioning fund.
> In other words -- they're not on the balance sheets.
If anything, the externalities of nuclear are the most meticulously accounted of all the power sources.
> What legit business gets away with that?
All fossil fuel businesses, who spew their waste into the atmosphere and pay nothing for it.
>The reality is that nuclear power is a 1950's technology -- one from the past, not the future.
If we could use technology from the future, that would certainly solve a lot of problems. But if we're ranking energy technology by age, then the ordering goes: fossil fuels -> solar panels -> windmill generators -> fission.
> We never fully understood the implications of running or taking apart a nuclear plant at commercial scale and instead of addressing the real problems, we just pretend they don't exist.
...We didn't? Again, France seems to manage okay. I'm not even sure what you're trying to say here.
> It's time to put the effort and expense once reserved for nuclear into something that doesn't create waste that lasts for 240,000 years (plutonium) or require supervised swimming pools to store (used fuel rods).
What's your suggestion? What should we build right now? We can't just not build power plants because someone might invent a better technology in the future.
> We need power plants that can be turned on and off quickly -- at small scale -- not ones that require 20 years to build and 50 to tear down.
Not only is this is a weird false dichotomy - switching time is orthogonal to construction time - but it's also wrong on both sides of the dichotomy: nuclear power plants can be switched very quickly (just raise and lower the control rods), and also don't take 20 years to build. Gravelines Nuclear Power Station (not cherry picked, just the one I happen to have the Wikipedia page open for) is the largest nuclear power station in France, and took... 8 years to build.
> The time for nuclear is over... and it has been for a while! We've just been in denial.
What denial? People have been irrationally fighting nuclear power since the early days of the Cold War, largely due to its association with nuclear weapons.
Wind and solar (then when variability limits from existing hydro and nuclear are reached, add H2, ammonia, and battery storage). It comes online in years instead of decades and doesn't wind up with 20% offline due to unexpected downtime when it's needed most like france.
It is already possible to build a mix with at least x watts 24/7/365 for lower price than nuclear (which has 60-90% availabity) and price goes down as you build more instead of up. The R&D required to reach these efficiencies has been possible for decades, but we chose to spend those billions elsewhere.