Read what I wrote more carefully -- the reactor built in the 60s was a molten salt reactor, IT DID NOT USE THORIUM. It used U-233 exclusively as a fuel, derived from military or research reactors at a guess. Lots of folks have conflated the molten-salt fuel transport system demonstrated in the 1960s with the Powerpoint presentations of purely theroetical LFTRs which have never been built or operated; as far as I know no-one has even made a benchtop demonstration of the operating principles. It is meant to breed thorium (Th-232) into fissile U-233 and then fission that in the same reactor to produce energy. This is a lot more complicated than what was achieved fifty years ago in a tiny (by modern commercial standards) laboratory model reactor.

I'm not sure what the LFTR boosters are about, really -- they may be cultist true believers aiming for the Promised Land but mostly they're chasing something that has fits no commercial niche today or for the next fifty years or so, while uranium is cheap and abundant (and that's without lots more recycling of spent fuel) and while there is an atom of carbon left under the surface of the earth that can be extracted and burnt to provide energy for free.

The Indians are closest to using thorium in any scale as a fuel source in heavy-water PWRs. They're basically regular-geometry fuel rod assemblies with a lot of thorium as well as kickstarter 20%-enriched U-235 and some Pu, probably a mix of -239 and -240 derived from spent fuel. The theory is they spend neutrons to breed the thorium up into U-233 which then fissions and generates energy and releases enough fast neutrons to breed more U-233 as well as slower moderated thermal neutrons to fission it. This is complicated and messy compared to one-step thermal moderated neutron fission of low-enriched uranium in regular fuel pellets and even MOX. They think they can make it work but they're spending a lot of effort and money in the process. The reason is political; they've got very limited native sources of uranium and they're not signatories to the Non Proliferation Treaty (NPT) which means theoretically they can't buy it from regular sources since nobody's supposed to help out non-NPT nations with nuclear equipment, materials etc. They do have a lot of thorium though hence their efforts to use it no matter how much it costs per kWh.

As for the fuel-rod thing I've got really no idea what you're on about. Most nations with more than one or two reactors make their own fuel assemblies, indeed many of them enrich their own fuel. It's not a razorblade handle situation, fuel costs including assembly manufacture are a trivial part of operating a nuclear reactor -- about 0.75 cents US per kWh according to the IAEA. The real stumbling-block is the five billion bucks plus per reactor up-front cost before the first truckload of concrete hits the rebar. That funding has to be in place even before the licencing and contruction application paperwork gets started, a process that can cost $500 million in itself and take more than two years before approval.

Nuclear weapons cores made from Pu-239 don't really degrade in storage and the material can always be reformed into new warheads as demands require but only in expensive facilities. The equipment around the cores does degrade -- for example the chemical explosive lenses arranged around the cores are precision devices and close proximity to a large amount of radiation will degrade them over time as will simple ageing and they do need to be swapped out as necessary. It's part of the expensive business of owning deployable nuclear weapons.

Britain has something like 70 tonnes of weapons-grade Pu-239 surplus to requirements from the time we had nearly a thousand warheads (we now have less than 200) and the US and Russians have a lot more than that in storage. It's just sitting there in several expensive storage facilities "somewhere". There's a much bigger cost penalty to converting material like that into usable weapons, not including the missiles, submarines and aircraft required to deliver them, the personnel to operate them, the training, security, release protocols etc. so most of those decommissioned warheads from the 60s and 70s have been dismantled beyond the point where they could be quickly put back into service. The US maintains a second-string reserve of warheads, mothballed at great expense beyond its incredibly expensive front-line fleet ready for use sitting on top of Minuteman-IIIs in South Dakota or riding in Ohio boomers somewhere in the Pacific. Notice the use of that word, "expensive". You tend to see it turn up a lot in discussions about nuclear weapons and materials. Keeping a secret stash of nukes and/or Pu-239 costs a lot of money, it can't simply be parked in a shed on an army base somewhere with a padlock on the door.

The Russians have sold the US a lot of highly-enriched U-235, some of it from weapons cores which has been downblended into nuclear fuel for power reactors, the "Megatonnes to Megawatts" project. The Russians surplus Pu-239 stockpile is more of a challenge but they are looking at using it in their BN-series fast reactors as well as MOX fuel for PWRs and the like. The US still hasn't licenced any MOX operations in its own commercial reactor fleet although there's a good deal of operational experience with it elsewhere in the world. This is the obvious way to use up surplus weapons-grade Pu-239 but the security of moving such materials around to downblend it is problematic -- commercial MOX with pure Pu-239 is a very great security risk.

U-233 -- no, the US does not have hundreds of tonnes of the stuff. It has two tonnes, no more. There are no real thorium reactors planned, granted Construction and Operating Licences (COLs) or pouring concrete now or in the forseeable future (unlike the US where there are four new-generation PWRs under construction and several more COLs have been issued). Assuming a series of financial, regulatory and licencing miracles occurs the earliest a molten-salt thorium reactor would be starting up anywhere would be fifteen to twenty years from now, and even that's optimistic as long as gas is cheap, coal is cheaper and yellowcake is $35/lb at the minehead. Storing that bomb-grade U-233 is expensive (oh look, there's that word again!) and it's not necessary to use U-233 to start up a thorium breeder, this can be done using U-235 and Pu-239 as the Indians plan to in their heavy-water PWRs in a thorium/MOX fuel cycle.

President Nixon was the one who started the nuclear arms reduction talks as I remember, so why did the White House want or need more bomb-grade plutonium by the late 1960s? A quick check on Wikipeida suggests US stockpiles peaked about 1965 or thereabouts at a bit over 30,000 warheads of all types. By the early 70s it was down to 20,000.

The purpose-built reactors at Hanford and elsewhere had already produced as much bomb-grade material as the US ever needed, building dual-purpose commerical reactors was pointless. The UK did go down this road with the flexible-fuel-cycle Magnox designs but again by the time they came into operation the UK already had as much Pu-239 as it needed for its own stockpile of warheads.

The Soviet-era BN series fast reactors have an iffy safety record with classical sodium leaks and fires -- the rumour mill was that the BN-600 had two turbine halls, one would burn down when the sodium leaked out of the heat exchangers and caught fire but it was easy to get the leak fixed and they could use the other turbine hall until the next leak and fire while they rebuilt the first turbine hall.

IFRs are basically breeder reactors and they have generally proved to be uneconomic in the world energy markets even if the technical problems many of them have suffered from in the past could be overcome. Generating electricity at 20, 30 or 50 cents US per kWh with IFRs or LFTRs is pointless when gas is cheap, coal is cheaper and no-one cares enough about CO2, acid rain, particulates, mercury and all the other ills of burning billions of tonnes of carbon-based fossil fuel each year in a planetary atmosphere.

No, there was no molten-salt thorium reactor in the mid-60s. There was a small (7MW thermal maximum output, never generated any electricity) molten-salt reactor fuelled with U-233, a testbed that ran only for a few weeks total over its lifespan and at maximum output for a very small part of that time. Other people have worked on breeding thorium into U-233 for use as a nuclear fuel but no-one's done it in a molten-salt reactor.

It depends on the operation being carried out. Reactor 4 at Fukushima Daiichi was famously empty of fuel rods with the entire core load stored in the adjacent Spent Fuel Pool of Doom Doom Doomity DOOM! (sorry, channeling Arne Gunderson there for a moment) but the Japanese nuclear authority requires a complete inspection of each reactor structure every 13 months and it was usual to empty the entire reactor core of fuel to allow this to be done during a refuelling operation, but as you say some of the less-spent fuel rods would be returned to the core before restart. Inspection of the inside of the vessel and the core structures is, I understand, carried out with cameras and remote probes underwater while it remains flooded as the core and inner lining are noticeably radioactive.

The outside of the pressure vessel (RPV) is also checked at this time with engineers entering the void between the vessel and the inner containment structure to do so. Once fission had stopped there was some residual radioactivity there (less because of the screening effects of 20cm of steel) but not enough to trip personal exposure limits for the inspectors, as long as they didn't make a picnic of it.

Without a moderator there is no criticality to worry about. Molten-salt reactors operate by moving a salt stream carrying fissile fuel through a moderator core where neutrons are slowed down, fission occurs and heat is generated. The fuel then leaves the core, fission stops and it enters a heat exchanger which produces steam or hot gas to drive a turbine and thus generate electricity. The LFTR designs add a breeding stage to the basic fuel transport system, converting Th-232 into U-233 which fissions in the moderator core. Thorium by itself is useless as a nuclear fuel. The often-touted experimental salt reactor run in the 1960s never actually used thorium, it was fuelled with U-233. As far as I know nobody's ever run a thorium-to-uranium breeder using molten salt. Thorium can be bred into fissile U-233 in other conventional reactors, commonly heavy-water PWRs and the like but it might also work in regular PWRs. There's no real demand for it today though since uranium is plentiful and incredibly cheap.

Almost all current reactors have fixed solid fuel elements and flow coolant (water, steam, gas, sodium, lead/bismith alloy etc.) around them to extract the heat of fission. Sometimes the coolant is also the moderator (PWRs and BWRs), sometimes a separate moderator is used, like graphite in the British AGRs and the ex-Soviet RMBK-4s. I think most of the many LFTR designs being promoted rely on graphite cores for moderation. If there's no moderating material in the LFTR fuel stream then dumping it into tanks has no bearing on whether the salt/fuel mixture can go critical or not. It will have a large payload of fission products and the dump tanks will have a lot of decay heat to cope with and the fuel salt will be intensely radioactive for a long time, long after it has cooled down enough to solidify; centuries or millenia perhaps, depending on the radiochemistry and amounts involved. It might be necessary to make the tanks removable using remote-handling equipment but that simply moves the problem, it doesn't deal with the dumped salt itself.

Actually the core of a regular PWR or BWR and even CANDU, Magnox, AGR or even the dreaded RBMK-4 graphite moderated reactor designs don't get very radioactive thanks mostly to careful choices of the steel alloys and other materials used in their construction (no cobalt, for example). The vessels can be removed from the containment after shutdown during decommissioning within a year or two with minimal shielding or after forty or fifty years of Safstor on site they're no more radioactive than, say, granite and can be treated as low-level waste. It is common for the inside and outside of a BWR/PWR reactor vessel and its core structures to be manually inspected during refuelling outages, for example.

The really intense radioactivity in a conventional reactor is contained in the spent fuel rods which, if undamaged, can be easily handled, transported and after a few years dry-casked for storage or shipped to a reprocessing plant to be recycled. It's done all the time in hundreds of reactors around the world during refuelling operations and has been for decades.

The LFTR concept involves moving intensely hot radioactive fuel in a salt stream through a carbon moderator for decades with no capability to repair or even properly inspect this part of the reactor as the piping will be mindbogglingly highly radioactive even if the fuel stream is removed to permit inspection.

Germany's burning coal and gas like there was no tomorrow. Its per-capita carbon emissions curve is on the rise again as it starts to shut down its non-CO2-emitting nuclear reactors while supertaxing the ones still operating to help pay for the construction of new coal-fired and gas-fired power stations from its climate change fund -- they can't put consumer electricity prices up any more to pay for these new fossil plants as they're already the highest in Europe, double that of 80% nuclear France next door which has half the carbon footprint of its solartopian neighbour.

Meanwhile states like Ontario are moving away totally from fossil-fuel for electricity generation, having embraced nuclear generation along with hydro and closing down their main fossil-fuel plants. Germany will still be burning tens of millions of tonnes of brown coal and lignite a year in 2050 by the best hopes of the supporters of renewable generation, not to mention Russian gas if there's any left.

On one hand there are about 400 "conventional" nuclear reactors generating power around the world, nearly all boiling-water or pressurised-water reactors with a few other types like the heavy-water CANDUs, the British gas-cooled AGRs and the infamous ex-Soviet RMBK-4s. They all use water or gas to cool solid-fuel elements fixed in place in a core structure. They have a typical uptime of 90% between refuelling outages and repair/inspection cycles and mostly sit there generating away and keeping the lights on.

On the other hand there are a few experimental power reactors in existence that use liquid metal cooling because they run much hotter thermally (700 deg C and higher) and with very high neutron economies (incredibly high fluxes in a small volume, thermal moderated neutrons for fission and also fast-spectrum neutrons for breeding and waste destruction). Most of the worked examples are hangar queens, breaking down repeatedly due to the thermal stresses and neutron flux damage to core structures, leaking coolant and catching fire and generally being unreliable. Because of this a lot of them have been shut down permanently as uneconomic to operate. A few are still running despite fires and leaks and the Russians are maintaining some interest in further developing their BN series of sodium-cooled fast reactors, possibly with investment from the Chinese.

In addition there are a host of new reactor designs which are basically paper exercises, grad student presentations and the like, dragging a wing around the academic world and hoping for a bite from one of the Big Guys who will drop a few billion bucks on bending metal and pouring concrete for their Precious. Probably not going to happen -- the only concrete pours going on right now are for more PWRs and BWRs which have a proven track record of producing electricity and making money for their operators.

Then what? The reactor operators can't just leave this mindbogglingly-radioactive boiling-hot slurry in those tanks, they have to clean it up. How do they intend to do so? It will be a requirement of the licencing of such a reactor design that they have plans and procedures ready if it ever does and equipment on standby just in case. "...and then a miracle occurs." is not going to pass scrutiny anywhere in the modern world's nuclear regulatory environment.

BTW the dump tanks don't need to be of sub-critical volume -- in fact they can't be. The molten salt stream carrying the fissionable materials only goes critical when it passes through the carbon moderator in the reactor core. Outside that core no fission can occur unless something goes really badly wrong and moderating material gets mixed into the molten salt stream (say if the graphite moderator core gets badly damaged) at which point you really don't want to be within a thousand miles downwind of this "safe" reactor -- one of the commonly posited cost-saving points of molten salt reactors is that like the Soviet RMBK-4s they don't need an expensive containment structure because they're "safe". Honest.

The molten-salt reactor could have produced weapons-grade plutonium (just add U-238 and continuously extract Pu-239 from the molten salt flow) but by the time it was up and running the US had as much plutonium as it wanted or needed for its thousands of in-service nuclear warheads, created in purpose-built breeder reactors running in Hanford and elsewhere in the 50s and early 60s.

As for "just drain(ing) the liquid nuclear fuel from the reactor" then what? How do you clean it up afterwards? You can't just leave it there. Mop and buckets, or a big sponge?

Going back to the original article there are some fun things folks have been doing recently with experimental reactors but the usual result has been expensive messes that are difficult to clean up afterwards. Commercial breeder reactors, for example, most of which have been shut down as either uneconomic or easily broken (or both). Gas-cooled pebble-bed designs; the Germans are still waiting for the radioactivity in their one to decay sufficiently so they can finally defuel it, including all the bits of fuel pebbles that fractured and jammed the mechanisms. It's been 25 years now and counting. Gas-cooled graphite-moderated son-of-Magnox designs like the British AGRs have high thermal efficiency but fuel is cheap and they were expensive to build and operate so the extra efficiency didn't help them proliferate in the world markets. We'll pass quickly over the RMBK-4 graphite moderator designs... CANDUs are doing quite well in some markets but they're expensive for the amount of generating capacity they provide and heavy water reactors present all sorts of proliferation risks. The Russians are doing some interesting things with compact fast-spectrum reactors which have very high burnup rates, effectively closed-cycle breeders with a possible sideline in isotopic waste destruction but they are very very experimental -- liquid sodium coolant, say no more.

Actually no. The oldest reactor at Fukushima Daiichi, no. 1 was about 40 years old but still operating problem-free and it was likely to get a ten-year licence extension after inspection. The average working life of 1970s-era reactors looks to be about 50 years or more; in a few cases economics and the dash for gas are getting some smaller facilities like Vermont Yankee (a 600MWe single-reactor plant) shut down. Reactors 2, 3 and 4 at Fukushima Daiichi were more modern designs and had at least ten years life left in them before the tsunami hit. Reactors 5 and 6 were built in the 1980s and were totally undamaged but they will be decommissioned as the site is considered inoperable in toto.

As for new modernised reactors designs Japan has three new reactors under construction at various stages of completion and its newest complete reactor came on-line a few years ago (2005, I think). The delay in building new reactors is due to the fact the older models are still operating safely and that reduces the demand for new ones. Gas is cheap at the moment as is coal and that also cripples the case for new nuclear builds because of the very large upfront costs of licencing and building any new reactors (which are expected to have a service life of 60 years).

International laws about dumping all sorts of waste materials at sea stop the Japanese from dumping the contaminated waste water there. They're making efforts to stop contaminated ground water escaping the site into the ocean, with variable results hence all the large storage tanks full of water being built on the site. In fact much of this water is actually safe to swim in or even drink by radiation standards adopted from the World Health Organisation and similar groups but it has enough measurable contamination to make it against the law to simply pour it down a drain into the sea.

Two reactors at Ohi were restarted back in 2012. Japanese nuclear regulations require a shutdown and inspection of all reactors every thirteen months, usually done as part of a refuelling operation. The Ohi reactors have been shut down again after operating for thirteen months but are not restarting after inspection and refuelling for various reasons, mainly bureaucratic and local-political.