Promoted for discussion by Brendan

Nuclear power has some very good and vey bad aspects and partisans on both sides argue passionately about these issues.  Ultimately though it is the pedestrian issue of fuel availability that should preclude us from embarking upon the type of massive scale nuclear power plant building that some have advocated.

To get a sense of current nuclear energy production consider the following:

Total world electricity production as of 2005 was 18,235 TWh (terawatt hours).  Total world nuclear produced electricity as of 2005 was 2,768 TWh (~15% of the total world electricity).

Total world energy production as of 2005 was 11,435 Mtoe (million tons of oil equivalent).  Total world nuclear energy production as of 2005 was 720 Mtoe (~6% of the world total energy)

Data from here: www.iea.org/textbase/nppdf/free/2007/key_stats_2007.pdf

Nuclear power plants currently use an isotope of Uranium called U235; the number refers to the atomic weight of the isotope the most comon form of Uranium is 238 which has three additional netrons as compared to 235.  Less than 1% of all Uranium is 235.  Uranium235 is the only naturally occuring fissile material of any real quantity on Earth.  There are a few other sources which can be converted into fissile fuels which I'll discuss a little later.  Additionally "Breeder" reactors can make some use of the more commonly available U238 because as part of the process of operation it converts U238 into U235.

So how much Uranium is there?

Breaking down the categories-

RAR is basically U we know is there.  EAR-I is Uranium we have very good reason to believe is there, geologically speaking.  EAR-II is U that is probably there and SR is getting into wishful thinking.  The "secondary sources" exist but there are some issues in using them and the "unconvetional sources" are pipe dreams at this point.

The subtotal is is about 17,000,000 tons of Uranium but over half of that comes from SR resources and really shouldn't be remotely counted upon.  Additionally both the EAR-II and SR resources do *not* take into account milling and mining losses (whereas RAR and EAR-I do).

Given that much Uranium, how much power generation is there available?

(this data and additional explanations of the categories is here: www.nea.fr/html/pub/newsletter/2002/20-2-Nuclear_fuel_resources.pdf )

Looking at the chart they use two different assumptions; "conventional" means the 17 mtons from the subtotal above whereas "total" includes phosphates and 10% of all the uranium currently dissolved in the oceans.  Getting uranium from seawater has been demonstrated under lab conditions but has not been put into any kind of production scale.  Until and unless it does work in the real world we should absolutely not depend upon it working.  You can see the Japanese Atomic Energy agancy website on the topic here: www.jaea.go.jp/jaeri/english/ff/ff43/topics.html .  Here you can see a nice page that discusses the technology and costs: npc.sarov.ru/english/digest/132004/appendix8.html .  Uranium from phosphates is a more mature technology but it's not like the world phosphate resources are just sitting idle.  Fertilizers, for one, make extensive use of phosphates.  What amount of the total phosphate reserves will be made available to uranium extraction is questionable.  

Consequently we can and should ignore the "total" columns and focus on the "conventional" columns.  And even those are far from assured since, as above, over half of that total comes from well named "speculative resources."  If we cut the "conventional" column in half that's a more realistic evaluation of what we really have avaialable to us and can depend upon.  Conventional resources would allow for 326 years of electricity production at 1999 levels (since about 1993 nuclea power generation has been pretty static so we can use the 2005 levels above) through a one use LWR (light water reactor) type process .  Cutting that figure in half, so we aren't betting our lives on speculative resources, means 163 years.  But thats only providing 6% of our energy needs.  For nuclear to do anything about the upcoming fossil fuel crash it will have to provide an order of magnitude more energy.

And that's also assuming our demand doesn't increase.  The Department of Energy estimates 2030 energy demand will be 1.5x the 2005 level (www.eia.doe.gov/oiaf/ieo/world.html ).  So looking at the supplies of uranium we can really depend upon, and asking nuclear to provide just 75% of our electricity needs (just  5x what it currently provides percentagewise, and probably too litle to make any difference to peak oil considerations) results in a mere 40 years until total exhaustion of our nuclear fuel stocks.  And doing so would require the creation of some 2000 additional plants (the world currently, as of 2008, has 439 nuclear power plants producing about 3,300 TWh, we'd need to produce 20,514 TWh).

What about the lower lines in the chart above talking about recycling and fast reactor usage?  Currently the world uses LWR pretty much exclusively.  Those reactors can usually only use the fuel once and then it is waste.  Some LWR can use recycled fuels (the second line) which is very doable but doesn't result in that much additional fuel.  Breeder reactors can reuse the fuel multiple times because in the process of the reaction they end up converting more of the U238 into U235 than is actually used up.  Consequently with a breeder reactor the composition of the fuel convertsmore and more into U235.  This is why breeder reactors can be used to enrich uranium for weapons use (weapons grade uranium has a much higher percentage of U235 than natural or nuclear power fuels).  Because of this using breeder reactors means you can greatly increase the life span of a fuel stock.

Just one problem, as above nobody is using breeder reactors.  Worldwide there's one operating plant (the Russian BN-600 expected to reach end of life in 2010, en.wikipedia.org/wiki/BN-600 ).  France and the US both had test breeder reactors but both projects have been abandoned.  The Japanese have one breeder reactor that suffered a very serious accident and was taken offline (en.wikipedia.org/wiki/Monju ).  It was supposed to come back online this October but so far has not due in part to a huge public outcry against the plant.  That's the extent of breeder reactors in the world.  Obviously if no one is building and using breeder reactors then it is a moot point as to their potential utility- they aren't being utilized.  At some point uranium prices may climb to the point that they become commercially viable, but the question is how much of the world's uranium reserves will we have used and sequestered (as nuclear waste) in the meantime?  Also how many LWR will then be sitting around useless while we build the thousands of breeder reactors to try and scavenge the remaining uranium supplies?  In a scenario where a centralized government controlled energy production it might be possible to build a breeder reactor based energy infrastructure that was worthwhile, but so long as atomic energy is the province of for profit companies (or short sighted governments) it is simply not going to happen within a useful time frame.

Addtionally since breeder reactors are inherently capable of enriching uranium their wide spread use is a counter-proliferation nightmare.  Consider how much certain parties have freaked out about he Iranian nuclear project and that is not a breeder reactor.

Thorium as a potential source of fuel has some possibility.  Thorium itself is not useful as a fuel but can be converted into isotopes of Uranium that are.  Once again though we are talking about an unproven technology.  There are no current operating plants using Thorium derived fuels.  There's one experimental plant, meant as a test facility not a production model, that is hoped to be completed in 2015 (en.wikipedia.org/wiki/HT3R ).  At least thorium designs do not have the problem that we are eating into their potential fuel source already (as is the case with breeder reactors and phosphates).

Thus barring some huge change in the current scheme of things we are left to the first two lines of the chart as the real situation.

To sum up; there is no good reason to encourage the building of large numbers of new plants today.  Current technology simply will not generate enough energy for long enough to matter.  Advanced technology may change this, but has not as of yet.  Some of these advanced techniques would require different designs than our current models anyway, so once again heavy investment in building plants now is pointless.  Investment in research is certainly worthwhile, but we have to accept that unless a breakthrough occurs nuclear fission is relegated to providing only a small fraction of our energy needs.