Ok, so what to think of nuclear power as a path forward? It’s an extensive and complicated subject, and one that doesn’t lend itself to definitive pronouncements. But, I’ve figured out quite a few things, much of it from a single long blog post by Tom Murphy, “Nuclear Options“, and from that page’s well-moderated and even longer list of comments from highly-educated professionals in the field. I highly recommend reading both if you’re interested in the subject, but if you don’t have a few hours to digest it, here are some important points, as I understand them (the unattributed quotes in this post come from his page)—
The biggest point of all—we can’t just take what we have and scale it up. After my post about Vermont Yankee quite a few people have expressed their opinion to me that we need to take France as a model and start scaling up nuclear power as a (relatively) carbon-free energy source. But, there simply isn’t enough uranium to power a full-scale switch to nuclear fission as we know it, even aside from safety and other concerns. Depending on how you figure it, the world’s 13 TW energy appetite would use up all the uranium in somewhere between six years and a few decades. This is because fissionable U-235 makes up only .7% (that’s point seven percent) of naturally-occurring uranium, which is nearly all “impotent” but fertile (as opposed to fissile) U-238. As such, something like a million tons of natural uranium a year would need to be mined.
Two ways out of this particular dilemma seem to be 1) to use breeder reactors, which can use the U-238 indirectly by converting it to plutonium, and 2) to use thorium, instead of uranium, as a fuel, in reactors (also breeder reactors) that can convert thorium into U-233 (U-233 being the second, but not naturally-occurring, fissile isotope of uranium). Thorium is several times more abundant than uranium in the Earth’s crust, and either of these approaches could extend the available fuel supply to an extent that it would become much less of an issue, or even a non-issue.
(Some other “ways out”—extracting uranium from seawater, or even common granite. At present I believe these options are quite theoretical, with no proven way to extract uranium on a large scale. There is potential there, because fissile materials have a million times the energy density of chemical fuels (oil), but at present these are not realistic options. Then, of course, there is fusion, which is so complex and difficult that I don’t think we’ll ever achieve it on a commercial scale.)
So, back to the “two ways out”—the problem with breeder reactors is mainly that they are expensive and less safe, due to proliferation and other concerns, and as such haven’t been fully developed for power generation. But breeders (sometimes called “fast reactors”) have some advantages. First, they can use both of the naturally-occurring isotopes of uranium, which would extend the available uranium supply by a factor of 140. They can also be designed as “burners”, which are reactors that are specifically designed to use up spent fuel. Because they use fuel differently, spent fuel from fast reactors is also less of a long-term hazard. This point deserves a few lines of explanation, because it’s one of the key advantages of breeder reactors. When a neutron hits a uranium atom and causes it to split, it splits into two “daughter” atoms of various elements, with atomic weights of about 95 and 135. These materials are also referred to as “fission products”. But, some uranium atoms absorb the neutron without splitting, and become transuranics (sometimes called “actinides”). Both of these, the fission products and the transuranics, remain in the spent fuel. But, they are vastly different in terms of their long-term hazards—fission products have much shorter half-lives, and are more-or-less fully degraded after a relatively short 300 years. The transuranics, however, have half-lives that make them dangerous for tens of thousands, or even hundreds of thousands, of years. BUT—breeder reactors can be built to burn up these transuranics. (I believe GE’s new S-PRISM designs, which are on the verge of being constructed in both the U.S. and Britain, are reactors of this type). So, therein lie the advantages of breeders—they can utilize nearly all of naturally-occurring uranium, they can burn up high-level radioactive waste from non-breeding reactors, and they leave spent fuel that is less hazardous over the long term.
There are disadvantages to breeders, as well. They make plutonium as part of their fuel cycle, and as such raise concerns over proliferation, especially as fuel is reprocessed. But, it also seems that this plutonium isn’t pure enough to be usable in weapons (though all nuclear material of this sort is “pure enough” to be used in dirty bombs); such weapons-grade plutonium is more often made in reactors specially built for this purpose. Breeder reactors are also more complex than standard reactors, and as such, even more expensive. They aren’t unworkable, but of the 430 or more nuclear reactors currently in operation worldwide, only a tiny handful are breeders, notably Russia’s BN-600 reactor, a 560 MW plant that has been operating since 1980. Many more breeder reactors have been built but later shut down, such as France’s Super-Phoenix, and Germany’s SNR-300, the latter a $19-billion plant that was completed, never started, and then decommissioned. Because such reactors aren’t common, a path forward using nuclear to make large proportions of our power would require further research and development (and quite a few countries are currently moving in this direction). It is expense that keeps fast reactors from being more common now, they would only pay off if the price of uranium was substantially higher than it currently is (the cost of fuel is NOT currently one of the big expenses in nuclear power).
And, somewhere in this post I need to point out that there are scores of different reactor designs, and they don’t all lend themselves to easy characterization. All reactors are breeder reactors to some extent (about 30% of the power from even a “standard” thermal reactor comes from U-238 that gets converted to plutonium and then split), and the degree of breeding can vary along a spectrum. Likewise, the differences between “Generation II” (Gen. II) and Gen. III and IV designs is often just a matter of degree. This smooth spectrum is a potential advantage, though, as designs can be continually modified in directions that are more efficient and safe.
The Holy Grail in nuclear power (short of fusion, which is the real nuclear-power Holy Grail) seems to be a “thorium molten-salt reactor”, or MSR. (Or, more specifically, an LFTR, or Liquid Fluoride Thorium Reactor). The thorium fuel cycle is less prone to proliferation (its route to fission goes through U-233, and not plutonium). While a few experimental reactors have been cooled with molten salt, this one is fueled with molten salt; the fissionable materials being blended with the salt and then pumped through the reactor core, where they reach criticality. The design would be passively safe, as a loss of power to the reactor would cause the salt to run out into a sub-critical arrangement and cool on its own. The problem is that we are apparently a good long way from achieving anything of this sort, and economic and safety concerns loom large. (Finding materials to contain highly-corrosive and blisteringly-hot molten salt reliably for decades being one of them). Several nations are discussing efforts to build such reactors, including China. Many who follow this effort, however, seem to be quite skeptical that working reactors of this sort can be built any time soon, saying that there are “real physical limitations”, others implying that we are “1% of the way there” in terms of development.
But, all these technicalities aside, my bigger issues with nuclear power as a path forward have less to do with these specifics, and more to do with general issues that are inherent to the field as a whole. The list of downsides to nuclear power is long. It is expensive, reactors are targets for terrorism, uranium production involves continual mining, and we would need thousands more nuclear plants if we were attempting to replace fossil fuels with nuclear (my rough calculation—about 8,000 more reactors). This would increase the likelihood of incidents by a factor of twenty or more. And when things go wrong at a nuclear plant, they can go very wrong, witness Fukushima (my post, “Oh My“). To switch completely to nuclear power would also require reactors all over the world, even in unstable places like Somalia and Afghanistan. Nuclear power is extremely complex and technical, and “incompatible with hard times”. Hopefully humankind will navigate the path forward smoothly, but if not, then nuclear power could be a tremendous liability (imagine untended, or even poorly maintained, nuclear reactors across the country). In either case, all of those thousands of plants would need decommissioned after forty years, and replaced. The materials in a used nuclear plant are largely radioactive, and as such can’t just be recycled. And, of course, nuclear plants produce radioactive wastes that are hazardous for centuries or more.
Some of these issues led Tom Murphy to conclude in his post that nuclear power might help meet future energy needs, but will play only a modest role. The Union of Concerned Scientists has a similar opinion, referring to nuclear power as “not a silver bullet“.
In general, I might be even less sanguine. The quest for something like thorium reactors seems to be a quest for humanity to avoid hard choices as we go forward, a way to breathe a sigh of relief over a bullet dodged, and go on with our trajectory of ever-increasing consumption, without dealing with the limits to growth (last week’s Economist article notwithstanding, post here.) I alluded to this in my post about Vermont Yankee, when I said that we needed a hard path forward. Nothing we do is decoupled from its environmental effects, and this will apply to thorium reactors, too. The power will have an environmental price, and the consumption that such power would engender would have an environmental price as well.
One last quote from one of the comments on Tom Murphy’s page—”World population has grown so far beyond what could be supported absent the machinery and workings of technological civilization that any breakdown could get messy very quickly…we need to step away from the precipice.” There are a lot of truths worked into that short statement. We can’t go backwards; we couldn’t feed ourselves, and things would unwind. But, we have to be really careful going forward, too, so that we can work in some resilience to our system, and “step away from the precipice”. Solar power, wind, sustainable biomass production, permaculture agriculture systems (post: “An Important Piece of the Puzzle“) , reduced consumption, and zero-population growth are all moves toward a safer, more stable and sustainable future. Incredibly expensive, complex, centralized nuclear power, dependent on prosperous times, seems to be doubling down on a questionable bet. Charging forward willy-nilly, so that we can keep our huge houses and extravagant ways, is part of that dangerous path.
So, I suppose I haven’t changed my opinion much. I’ll just reprint my conclusion from last month here, “we might as well think big, and work toward a planet powered by clean, renewable power, with reduced consumption and a reduced focus on material things; a world of wind turbines, solar panels, permaculture, highly-efficient buildings, and more intentional living. That’s the clean, safe, healthy future we need.”