Display:
Rough math check. Upper bounds on uranium use: Currently, nuclear is supplying 2 % of global primary energy use - Assuming that already existing large scale hydro/wind stays in use, and that electrification of everything only cuts total energy use by 30% (this is conservative, yes?) the nuclear power sector would have to expand to about 30 times its present size. Currently (well, in 2006. sorry, really should get newer numbers), the world goes through some 66 500 tonnes of uranium/year, so this is a burn rate of somewhat over 2 million tonnes/year.
Known conventional resources extractable at <130 dollars/kg are 5.5 million tonnes. Estimates of currently unknown, but geologically likely resources of the same grade are another 10.5 million tonnes, so using conventional reactors on this scale would burn through global ore supply of reasonable grade in 7 years. Okay, that takes that plan off the table, I suppose seawater extraction could supply a once-through cycle for significant spaces of time, but the entire idea is to avoid industrializing more of nature than we have to, and extracting 2 million tonnes/year from the sea would run counter to that.

So, really, for nukes to work as a global warming fix, breeders are necessary. - Plausible designs for rapid adoption would be the IFR and the Russian BN-800 design.
- This actually has hilarious consequences in the opposite direction, since the ore requirement for a breeder reactor in operation is roughly one tonne of  fertile, (not fissile!) material per year, so after the initial fuel load, the stockpile of already mined depleted uranium (1.5 million tonnes) could keep the world in power for about a thousand years..

by Thomas on Wed Feb 10th, 2010 at 07:51:49 PM EST
[ Parent ]
I suppose seawater extraction could supply a once-through cycle for significant spaces of time, but the entire idea is to avoid industrializing more of nature than we have to, and extracting 2 million tonnes/year from the sea would run counter to that.

Indeed, and it may not even be feasible: again size of supply and speed of extraction are different things. The Uranium2007 study says:

Seawater may also be regarded as a possible sourcc of uranium, due to the large volume of uranium contained (about 4 billion tU) and its almost inexhaustible nature. However, because of the low concentration of uranium in seawater (3-4 ppb), it is estimated that it would require the processing of about 350 000 tonnes of water to produce a single kg of uranium. Nonetheless, with the exception of its high recovery cost, there is no intristic reason why at least some of these significant resources could not be extracted from various coast lines at a total rate of a few hundred of tonnes annually.

So, really, for nukes to work as a global warming fix, breeders are necessary.

Yep. But, for that, one would need:

  1. design &build a full-scale prototype: min. 10 years
  2. gain sufficient operational experience with the prototype: min. 5 years
  3. get first commercial plants running: min. 5 years

...and we arrived in 2030 already. And this is a best-case scenario, supposing that

a) the design works without problems (unlike the Superphénix or the never even started Kalkar),
b) the breeder plant is paired with another novelty, a flawlessly working reprocessing plant (in contrast to the dirtiest branch of the nuclear industry: Sellafield, Hanford, Mayak, Le Havre, Hanau, Tokaimura...).

So, I'm not saying it's impossible, but I don't see a breeder future any less hypothetical than a fusion future.


*Lunatic*, n.
One whose delusions are out of fashion.

by DoDo on Fri Feb 12th, 2010 at 05:20:46 PM EST
[ Parent ]
There are working, commercial, fast breeder designs available today. They are Russian. Which is ahem cough problematic from a public relations standpoint in the west, but the Chinese do not seem to care, and the tech in general really is quite massively further along than fusion research since just about everyone with a nuclear industry has at least built a prototype that produces net power.  
- Note that the French plan for their nuclear sector appears to be "Build PWR's now, Fast breeders later, run the PWRs on Mixed oxide  fuel produced in the breeders"
by Thomas on Fri Feb 12th, 2010 at 08:24:43 PM EST
[ Parent ]
There's really no reason to go for breeders until a long time later, and then it's just an academic exercise anyway, as we have no idea on the distribution of low grade ores, and we have even less idea how much lower grade ores that can be mined than the ones we're looking at now.

For example, an almost unknown mining project in northern Sweden (Myrviken), an vanadium, uranium, molybdenum and nickel deposit, is big enough to fuel all our nukes for 20 years! And this from a country which the IAEA classifies as having "zero" uranium reserves. The uranium ore is not rich enough to mine by itself, but the other minerals make it possibly profitable. There are lots and lots and lots of big low grade deposits like this scattered all over the world.

Total known world recoverable uranium reserves are 5.5 million tonnes. In spite of this, there's 1 million tonnes of the stuff lying around in a single small Swedish mountain (Billingen) which is not included in any data because of the low grade. At what cost is this ore profitable? No one knows. Is it possible to mine it? Yes, it was mined to get uranium to our weapons program.

Oil reserve data is bad, gas reserve data is awful, coal reserve data is horrible and uranium reserve data is nonexistent. The only thing we can say about uranium reserves with any measure of safety is that there's really plenty of the stuff around, and whenever people head out and look for it they find a lot more than is consumed any given year.

Peak oil is not an energy crisis. It is a liquid fuel crisis.

by Starvid (arvid.hallen at gmail.com) on Sat Feb 13th, 2010 at 12:27:43 AM EST
[ Parent ]
Oh, and then there's another 4.2 million tonnes of uranium in the Storsjön area. This is where the Myrviken deposit (about 30.000 tonnes) is located.

So, in just these two areas there's as much uranium as in the official numbers for the entire world. Think about that for a while. On the other hand, the low concentrations and the thin deposits (3 metres) means that most of these deposits would have to be mined in a way which looks a lot like open pit coal mining. Myrviken is an exception, because here the alum shale is not 3 metres thick, it's 200 metres. Myrviken would look very much like and have about the same dimensions as the Aitik copper mine in northern Sweden.

       

Peak oil is not an energy crisis. It is a liquid fuel crisis.

by Starvid (arvid.hallen at gmail.com) on Sat Feb 13th, 2010 at 12:45:01 AM EST
[ Parent ]
There are working, commercial, fast breeder designs available today. They are Russian.

There is

  1. Beloyarsk-3, a working BN-600 lower-scale prototype: without containment dome, burning uranium only; and with a history of accidents and consequent (checking) 45-78% utilisation rate, though neither as bad as for Superphénix);
  2. Beloyarsk-4, a much-delayed demonstration unit in construction: it will pioneer MOX use in the BN series, I don't know about containment dome;
  3. the export to China you mention, which will be another BN-800 demonstration unit, to be built from next year, which China wants to use to develop its own type.

That's not commercial yet -- the Russian commercial version (BN-1600, Beloyarsk-5) would be twice the size, and use lead rather than sodium as coolant.

Which is ahem cough problematic from a public relations standpoint in the west

It may also be problematic considering the power sector safety culture there, especially considering the liquid sodium pool. I shudder at the thought of a commerical-scale Chinese fast breeder reactor operated with the same care coal plants or indeed photovoltaic factories are. The same goes for large-scale reprocessing.

Anyhow, though I think that my 10 year minimum applies, for purposes of a timescale, it is wholly uncertain when Beloyarsk-5 (or a Chinese equivalent) will be up and running (I haven't found any officially stated concrete dates). But looking further East, here is a timescale:

Mitsubishi to develop Japan's next fast breeder reactor

Mitsubishi to develop Japan's next fast breeder reactor
18 April 2007

The Japanese government has selected Mitsubishi Heavy Industries (MHI) as the core company to develop a new generation of fast breeder reactors, in an initiative promoted by the Japan Atomic Energy Agency (JAEA).

...The company plans to establish a new unit by March 2008 to orchestrate engineering activities and carry out development, looking towards construction of a demonstration FBR by 2025 and a commercial reactor for introduction by 2050...



*Lunatic*, n.
One whose delusions are out of fashion.
by DoDo on Sat Feb 13th, 2010 at 03:53:21 AM EST
[ Parent ]

Display:
Login
. Make a new account
. Reset password
Occasional Series