Thursday’s Guardian ran another article on the clean-up of Japan’s Fukushima nuclear site following the tsunami six years ago today. They point out, quite correctly, that two robots have now failed in trying to investigate the inside of the Reactor 2 containment vessel. I don’t see why this is such a surprise to everyone, or why quite so much recrimination continues.

Let’s be clear, again, once and for all. The containment at Fukushima did its job. It contained the reactor cores (admittedly only just) under stresses (earthquake and tsunami) way beyond its design specification.
What failed were the cooling systems. And they failed because of major shortcomings in the risk analysis, and therefore the placement and design, of the plant.
Yes, there was a radiation leak – small in comparison to Chernobyl – as a result of fractures in the buildings surrounding the containment vessels. And yes, this is a disaster for the 160,000 people who were evacuated – the disaster is their displacement and, medically, the psychological effects, rather than the risks due to the actual radiation encountered in the time between the leaks and their evacuation.
The tsunami killed around 19,000 people. The radiation, as far as I am aware, has caused zero direct deaths (although a handful have died in accidents during the clean-up operation).
Of course the clean-up is going to take a very long time and be hugely expensive. The radiation level inside the containment vessels is going to be incredibly high – high enough to kill a human within minutes. So without robots there is no way to find out what actually is happening inside; and they will succumb to high radiation levels and blocks in their access routes. And yes there is a huge quantity of contaminated groundwater to contend with. Why would we expect otherwise?
The current estimate is that the clean-up will take 30-40 years and cost $189bn, although many believe this a significant underestimate in both time and money. On that basis one has to ask whether the clean-up should continue, or whether the whole plant should be permanently encased as has been done recently at Chernobyl – but I’ve seen no-one even mentioning this as a possibility. I’d be interested to see some analysis of the possibilities.

A week ago IFLScience published a very long, and fully referenced, article on a forgotten nuclear power technology which is much more efficient and robust than the current Light Water Reactors (LWR). It is actually a breeder reactor (but one which doesn’t produce weaponable products) called a Molten Salt Reactor (MSR).

As usual what follows is a few extracts by way of the TL;DR summary.
According to the article MSR are not just a better nuclear technology but also beat most other power sources (including most renewables) into a cocked hat.

Today’s cheap, bountiful supplies make it hard to see humanity’s looming energy crisis … Fossil fuels could quench the planet’s deep thirst for energy, but they’d be a temporary fix at best … renewable energy sources like wind and solar, though key parts of a solution, are not silver bullets … Nuclear reactors, on the other hand, fit the bill: They’re dense, reliable, emit no carbon, and – contrary to bitter popular sentiment – are among the safest energy sources on earth. Today, they supply about 20% of America’s energy.
The good news is that a proven solution is at hand – if we want it badly enough.
Called a molten-salt reactor [it] forgoes solid nuclear fuel for a liquid one … in theory, molten-salt reactors can never melt down … It’s reliable, it’s clean, it basically does everything fossil fuel does today … [and produces] energy without emitting carbon … What’s more, feeding a molten-salt reactor a radioactive waste from mining, called thorium (which is three to four times more abundant than uranium), can “breed” as much nuclear fuel as it burns up.

MSR were developed in the early days of the Cold War and the technology was proven in pilot production. However they were never pursued because (a) they didn’t produce weapons grade materials and (b) “not invented here”.
The article follows with a brief analysis of the safety of nuclear energy compared with traditional power generation, and a very brief summary of how nuclear physics works. Followed by an explanation of how MSRs using thorium can “breed” and then use uranium 233 but not weaponable plutonium.

The concept of the breeder reactor was fairly straightforward. It would dramatically increase the chances for fission, boost the flow of neutrons, and breed more fissile fuel from a “fertile” material than the reactor burned up. Breeding U-238 into Pu-239 created an excess of plutonium. Meanwhile, breeding thorium into U-233 broke even, burning up just as much fuel as it made. The choice of fuel makes all the difference. The plutonium fuel cycle is a great way to make weapons. Meanwhile, the thorium fuel cycle can produce almost limitless energy. A fluid-fuelled design [would] eliminate the considerable difficulty of fabricating solid fuelled elements … Liquid fuel also made it easy to remove both useful fission products – for example, for medical procedures, and those that poison nuclear chain reactions.

OK, so what’s the downside? Basically, apart from the proof-of-concept pilot, the technology hasn’t been developed fully. But it could be developed, and probably relatively easily, probably as the Liquid-Fluoride Thorium Reactor (LFTR). And the article lists (some of) the advantages of LFTR:

• Fuel burn-up is extraordinarily high. LFTRs could fission about 99% of their U-233 liquid fuel, compared to a few percent for solid fuel.
• It’s easy to clean up. Solid fuels build up fission products, or new elements generated by the splitting of atoms, which poison fission reactions and often end up being treated as waste. Liquid fuels, meanwhile, can be processed “online” – and the fission products continuously removed, refined, and sold.
• There’s less waste and it’s shorter-lived. For the above reasons, hundreds of times less radioactive waste is left over from LFTR operation compared to LWRs. And what remains requires burial for about 300 years, as opposed to 10,000 years.
• LFTRs operate under safe, normal pressure. All commercial reactors compress water coolant to extreme pressures – upwards of 150 times that found at Earth’s surface. One small breach can lead to a catastrophic explosion. If a LFTR pipe breaks, however, molten salt will only spill on the ground and freeze.
• Environmental contamination is far less likely. LWRs can release gases, fuel, and fission products into the air and water. Molten salt freezes and traps most contaminants.
• LFTRs can be made small and modular. LWRs require giant, reinforced-concrete containment vessels that scale with their operating pressure. LFTRs require small containment structures, so they could be made small – possibly to a size that’d fit [on a truck].
• They should be much cheaper and faster to build. LFTRs don’t require many of the expensive safeguards that LWRs do. Their potential to be modular could also lead to mass manufacture of parts and reduced cost.
• LFTR is immune to meltdowns. Molten salt that overheats will expand, slowing down fission.
• The design is “walk-away safe.” No nuclear power plant today can claim this. LWRs require backup power systems to cool solid fuel at all times. If power is knocked out to a LFTR, a freeze plug melts and lets the molten salt fall into underground containment units, where it freezes and stops fission.
• Electricity output is better. LFTRs are so hot, operating at roughly [1000°C] they can use more advanced heat-to-electricity conversion technologies.
• The excess heat is very useful. It could boil and desalinate ocean water into drinking water, help generate hydrogen for fuel cells, break down organic waste into biofuels, and power industrial processes.
• The “kindling” to start a LFTR is flexible. Burning up old nuclear weapons material is possible, since fissile U-233, U-235, or Pu-239 can be used to start the reactor.

So if thorium reactors are so great, what’s the holdup?

It basically boils down to … The science is easy. The engineering is hard … [which] is true in many, many advanced systems, nuclear and nonnuclear for that matter, where the scientists’ proof of concept is everything to them … To the engineer, getting it to the commercial-viability stage is their goal. And those are two very different hills to climb.

So there is still a long road ahead, but given the apparent advantages isn’t this a technology we should be pursuing? Yes, India and China are already doing so.