Pages

November 04, 2010

David Leblanc explains why Thorium reactors need a lot less fissile nuclear material to start and for ongoing operation

Joseph Friedlander asked David LeBlanc about the starting fissile requirement for Thorium nuclear reactors.
David Le Blanc, Physics Dept, Carleton University, Ottawa, Canada and Ottawa Valley Research Associates Ltd

A 12 page paper by David LeBlanc , Denatured Molten Salt Reactors (DMSR): An Idea Whose Time Has Finally Come?

Take 1970s technology, add perhaps $3 billion of R&D over 15 to 20 years for a trillion dollar per year potential energy market.

Molten Salt Reactors are one of six next generation designs chosen by the Gen IV program. Traditionally these reactors are thought of as thermal breeder reactors running on the thorium to 233U cycle and the historical competitor to fast breeder reactors. However, simplified versions running as converter reactors without any fuel processing and consuming low enriched uranium are perhaps a more attractive option. Uranium consumption levels are less than 1/6th that of LWR or a 1/4th of CANDU while at the same time offering clear advantages in safety, capital cost and long lived waste production along with increased proliferation resistance. A review of previous work and potential improvements proposed by the author will be presented.

David answers are below.

Molten salt reactors can come in several forms with the two main factors being

(1) whether it is a Single Fluid (one salt for everything) or Two Fluid (one fuel salt for U233, one blanket salt for Thorium). There is also an intermediate called the 1 and 1/2 fluid that has a mixed fuel salt surrounded by a blanket salt. The second factor

(2) is how thermal versus fast the neutron spectrum is. This is what really determines how much starting fissile one needs.

The French Molten Salt Fast Reactor design with a faster neutron spectrum needs a great deal of starting fissile material (typically 5 to 10 tonnes per GWe) whereas a well moderated (lots of graphite) design is typically 1.5 tonnes and as little as 0.5 tonnes. As well of course, molten salt (liquid fluoride) reactors don't need to be breeders, they can act as very simple and very attractive converter reactors that only need a small fraction the uranium of LWRs.



Question - apparently there is a need to have Uranium 233 for purposes of the first loading of thorium power reactors? Or will any fissionable do?

Answer - Molten salt reactors require anywhere from 500 kg to upwards of 5 or more tonnes per GWe plant (whether one runs a more thermal neutron spectrum or a more fast one). 1500 kg is considered typical. Any molten salt reactor can start on bomb grade or reactor grade Plutonium (roughly the same amount needed as U233). Plutonium is expensive to isolate and of finite supply in spent fuel (600 to 1000 tonnes worldwide?) Any MSR can also start on highly enriched uranium U235 (i.e. old weapons supply). HEU (Highly enriched uranium) is also limited and would be very politically unattractive to be shipping around. Potentially the simplest startup fuel would be Low Enriched Uranium (LEU is under 20% U235 and useless for weapons) because we can produce this in great amounts (we already do) but this is where I have to qualify my answer.

There are two main classes of molten salt reactors that produce all their own fissile fuel after startup, the better known Single Fluid design has everything in one fluid has a very hard time starting on LEU but Oak Ridge National Labs proposed such a design in the late 1970s (DMSR Breeder). Another way to run is called Two Fluid which has separate salts for the fissile U233 and fertile Thorium (which greatly simplifies the removal of fission products). In this design, favored by many these days, startup on LEU is fairly straightforward. You simply run LEU in the central fuel salt for a few years while building up and saving U233 produced in the thorium blanket salt. Once you have enough U233 saved up, you remove (and sell) any remaining LEU in the fuel salt and replace it with your saved up U233 and from then on run on the pure Thorium to U233 cycle.

So the short answer is that startup fissile requirement is not a roadblock to building thousands of GWe of molten salt reactors like it is for something like metal cooled fast breeders that need ten to twenty times as much starting fissile material.

Question 2 - if we decided to go thorium LFR tomorrow to replace all coal/NG baseline generation we couldn't because we need to have X amount of 233 for the first fuel loadings? I would name this the "U233 first loading bootup problem" ; if it truly does exist I believe I might have thought of a way around it. But is it really a problem?

Answer I believe this was refering to smaller 100 MWe reactors and Robert is one of the many that prefers Two Fluid designs that can start on quite small amounts of fissile material. 100 kg of U235 for 100 MWe is possible but I'd probably up that number a bit to be conservative since U235 isn't as good a starting fuel as U233. The price of $50,000 per kg is correct (even a bit on the high side).

Question 3
Also, that implies we DON'T need the U233 for first bootup? Theoretically *we
could use the 1000 tons of HEU military stockpiles of U- 235 to boot 10000
thorium reactors tomorrow? **Are these numbers correct?

Answer - Yes, as I discussed in Question (1) and again this is smaller 100 MWe reactors.

(Question 4)*
(Obviously, if the designs and all parts and tools pre-existed-- speaking
only of nuclear materials)*
I am also trying to construct a realistic list of fuel loadings needed for various kinds of reactors contrasted for a hypothetical all nuclear world of 15000 1 gwe reactors (or 150000 100 mw reactors) (ie all thorium powered, chemical fuel needs synthesized from CO2 as in the AIM HIGH or Keith Henson plan)

http://www.nss.org/settlement/ssp/library/CO2andSpaceResources.ppt

In the Aim high example this seems to be about 15000 tons of Th a year plus 1500 tons of initial U-235 to start it off (after that, consumed *235 replaced by generated 233 is my understanding?) Is this correct*?*

Answer I think assuming 100% of world energy a little high (lets leave at least some space for hydro, biomass, wind and solar!). But yes, 15,000 GWe would only consume 15,000 tonnes of thorium per year, but you are missing a zero in regards to how much U235, by Robert Hargraves' numbers it is 15,000 tonnes, not 1500. Starting on U235 is not quite as efficient as U233 though so to be conservative I'd probably double that value to 30,000 tonnes.

(Question 5)For fifteen thousand 1 GWe DMSR (Denatured Molten Salt Reactor) reactors what are the initial and yearly loadings?

A 12 page paper by David LeBlanc , Denatured Molten Salt Reactors (DMSR): An Idea Whose Time Has Finally Come?

Answer The DMSR converter is my other passion as of late. It is a much simplified design that needs a lot less R and D and likely an easier sell to utilities, governments and regulators. It doesn't need to process the salts at all (except an optional single treatment after each salt batch of 10 to 30 years). Oak Ridge National Labs promoted this design in the late 1970s but was subject to little optimization (i.e. done with almost no funding). The 1 GWe design requied 3.5 tonnes of U235 (in LEU) to start and averaged about 150 kg of U235 per year of fissile top ups after that (at 75% capacity factor). Both these values could easily be improved, especially the annual requirement which could be halved. For LWR and CANDUs the starting load is more for LWRs (5 tonnes U235) and a less for CANDU (about 1 tonne U235). The annual needs for LWR and CANDUs are similar at roughly 1000 kg U235 per GWe- year (i.e. about 200 tonnes natural uranium).
A newly proposed embodiment of a Pebble Bed DMSR Converter based on a modification of design work from ORNL 4344.

Question 6 Regular lwr/bwr/candu typical loading is like 200 tons right? (first two use 3.5% LEU, last natural--) So we are discussing perhaps *3 million tons of Uranium* for them--- not sure that is doable.
* (I assume PU239 would work too as initial fissionable loading with some
reactor modification--*
*
Yes, no reactor modifications needed really.

Question 7 How much Pu 239 needed per 1 ton fuel loading thorium reactor?) How much seed U233 (235 was listed as 100 kg)
* (I am trying to get the ratios of 235-233-239 required for initial
fissionable loading in a 1 ton Thorium reactor in kg...*

Answer
There isn't a specific answer to this but a rule of thumb might be just slightly more Pu239 than U233 and 1.5 to 2 times as much U235 than U233.


Question 8 Is that U235 HEU price correct (I thought the Russians were selling it in the 90s for $5000 a kilo --PU239 for 25000 a kilo?*) Apparently it is 10 times that now?

Answer The Russians were likely selling it cheap to raise cash. I would say 30,000$ per kg of U235 is typical. For Pu for MOX use, it typically costs about 100,000 per kg to remove it from spent fuel and this is only 60% fissile (239+241). A sodium cooled fast breeder needs about 18 tonnes of reactor grade Pu to start($1.8 billion per GWe!). If we are removing it for molten salt reactors we can likely use methods that are less than half this price per kg and of course we need a lot less kg.


If you liked this article, please give it a quick review on ycombinator, or Reddit, or StumbleUpon. Thanks

Featured articles

Ocean Floor Gold and Copper
   Ocean Floor Mining Company
blog comments powered by Disqus