Is the nuclear renaissance dead yet?

[olivier]

The MIT "Future of Nuclear Power" paper, has this to say: "Typical LWR spent fuel today reaches a burnup of 50,000 MWD/MT"

There is a lot of information in the MIT study and it is definitely worth reading. A burn-up of 50 GWd/MTIHM (50 gigawatt.day per metric ton of initial heavy metal) is typical of a high burn-up LWR. It is the quotient of the thermal energy by the mass of enriched fuel that was inserted in the fuel rods before irradiation. It tells you how much energy you can get from a fuel element, and how long it can stay in the core before unloading, with an impact on reactor availability. It says nearly nothing on the efficiency of the reactor in terms of natural uranium usage. To get a higher burnup, you need a higher enrichment, hence more depleted uranium tails at your enrichment plant. The balance in terms of natural uranium usage is almost even (to be more specific, it is much influenced by the energy efficiency of the enrichment process). As an example, CANDU reactors have very low burn-ups since they use un-enriched uranium as a fuel. They need more uranium to pass through the reactor, but they do not use more natural uranium than LWRs.

A breeder should convert 99.9% of the fuel energy content to thermal power.

With the open fuel cycle practiced in the US and in many other countries, LWRs hardly burn (I mean, fission) a half-percent of the natural uranium withdrawal required by the process, two thirds U235 and one third U238 (transmuting into Pu in the core). Most breeder designs cannot obtain high burn-ups from a single load. They require many cycles of depleted fuel (blankets) and enriched fuel (inner core) loading, irradiation, unloading and re-processing, and so on. Not that it is theoretically impossible, but a 99.9% burnup would take centuries. Most people would be very happy with 50%, to be compared with the current 0.5%.
Now enough with uranium resources, let us think of the wastes. That is where the big deal is.

[jsbiff]

The MIT "Future of Nuclear Power" paper, has this to say: "Typical LWR spent fuel today reaches a burnup of 50,000 MWD/MT"

There is a lot of information in the MIT study and it is definitely worth reading. A burn-up of 50 GWd/MTIHM (50 gigawatt.day per metric ton of initial heavy metal) is typical of a high burn-up LWR. It is the quotient of the thermal energy by the mass of enriched fuel that was inserted in the fuel rods before irradiation. It tells you how much energy you can get from a fuel element, and how long it can stay in the core before unloading, with an impact on reactor availability. It says nearly nothing on the efficiency of the reactor in terms of natural uranium usage. To get a higher burnup, you need a higher enrichment, hence more depleted uranium tails at your enrichment plant. The balance in terms of natural uranium usage is almost even (to be more specific, it is much influenced by the energy efficiency of the enrichment process). As an example, CANDU reactors have very low burn-ups since they use un-enriched uranium as a fuel. They need more uranium to pass through the reactor, but they do not use more natural uranium than LWRs.

A breeder should convert 99.9% of the fuel energy content to thermal power.

With the open fuel cycle practiced in the US and in many other countries, LWRs hardly burn (I mean, fission) a half-percent of the natural uranium withdrawal required by the process, two thirds U235 and one third U238 (transmuting into Pu in the core). Most breeder designs cannot obtain high burn-ups from a single load. They require many cycles of depleted fuel (blankets) and enriched fuel (inner core) loading, irradiation, unloading and re-processing, and so on. Not that it is theoretically impossible, but a 99.9% burnup would take centuries. Most people would be very happy with 50%, to be compared with the current 0.5%.
Now enough with uranium resources, let us think of the wastes. That is where the big deal is.

Interesting stuff.

I should mention, I guess, if I wasn't clear, that I was just trying to figure out how much total energy can be got out of the 'spent fuel' if you were able to burn it all (e.g. in something like an Integral Fast Reactor). Basically, I took that 'Burnup' figure from the MIT study as a base starting point for how much energy an LWR extracts from the fuel, then, because everyone I've read or watched video of, has said that a reprocessing reactor can (eventually - not in one loading, of course) get 100 times more energy out of the fuel, so basically I did some conversions to tranform that GWD/MT burnup figure into kWh/MT, then multiplied it by the number of MT of waste we have stockpiled (approx. 65,000 I believe), which should have given me the total number of kWh of thermal energy we can get out of the spent fuel, then I multiplied THAT by a thermal efficiency factor of .3 (because converting thermal energy to electricity with steam turbines is inefficient, of course), which gave me approximately how many kWh of electrical energy we'd get from all that, and finally multiplied all that by $0.01 to see how much money that electricity might be worth, if it were valued at 1 cent/kWh (which is probably too low, but I was trying to be conservative in any assumptions I make).

So, the question is, did I do something wrong, methodologically speaking, by using that 'Burnup' figure as the basis for comparison? That is, when people writing/talking about Gen IV nuclear reactors which extract 100X more energy from the fuel, is the 'Burnup' figure NOT the amount of energy current LWRs extract?

[KitemanSA]

Personally, I would like to see the thorium cycle used. The burnup approaches 100% and few higher actinides are formed since the cycle creates 3 fissile materials on the way (U233, U235, Pu239). I am not sure what happened to the U237 that SHOULD be bred, but I never see it in the cycle. Since it is an "odd" atomic weight, it should be fissile too.

Further, with the appropriate reactor type (MSR) the fuel preparation costs are next to nill (refine the metal, dump it in) and with accelerator enhancement, no significant reprocessing is needed except to remove the fission products (distill the salt; the fission products have lower vaporization temperatures IIRC so they can be removed without futzing with the actinides).

Just seems to make sense to me.

[olivier]

So, the question is, did I do something wrong, methodologically speaking, by using that 'Burnup' figure as the basis for comparison?

No, you did not, but burn-up has differents meanings in different contexts. You cannot compare the 50 GWD/THM where the divisor is a mass of enriched uranium with the 99.9% of an ideal breeder cycle where the divisor is a mass of natural uranium.
See the end of my previous post and compare the 50% of a realistic breeder cycle with the .5% of the current open LWR cycle and you'll get your x100 factor.
Here is a back-of-the-envelope calculation (I love SI ):

• 1 GWd is 1E9 x 365 x 24 x 3600 = 3.15E16 J.
  1 fission provides 200 MeV but you loose 10% as neutrinos and have only 40% thermal-electric conversion efficiency.
  So you need to break 3.15E16 / (2E8 x 1.6E-19 x .9 x .4) = 2.7E27 atoms of uranium.
  Each atom weighs 238 x 1,67E-27 = 3.97E-25kg (no big difference between isotopes for what I am doing here).
  The mass of uranium to be fissioned during a year is then 2.7E27 x 3.97E-25 = 1070 kg.

[WizWom]

Kiteman, I haven't seen anyone claiming > 80% burnup for any fuel cycle; that was an advanced Thorium design.

The trouble with high burnup is that your K gets more and more unsure, and your residual heat load goes up also (as fission products are typically much higher lambda).
Not to mention managing reactor fuel structural stability becomes an important consideration.

Personally, I'm working on using a contained non-structural MOX fuel, similar to a pebble-bed design, but more on the order of sand sized.

As for people's estimates of current spent fuel energy available, if you use 20 m^3/GWe/yr, or about 650 m^3 per year for the US. Currently, the spent fuel pools hold about 80x the amount of fuel in the cores of the reactors. It would be very advantageous to the economies of reactors to be able to use this major source of high level waste. Accelerator driven waste burners with some small multiplier for net energy gain have been suggested for at least 20 years.

[jsbiff]

Kiteman, I haven't seen anyone claiming > 80% burnup for any fuel cycle; that was an advanced Thorium design.

The trouble with high burnup is that your K gets more and more unsure, and your residual heat load goes up also (as fission products are typically much higher lambda).

I guess I should ask something here. In a reactor design like the IFR, where you burn for awhile, refine the fuel, burn again, refine, burn again, etc, I assume a couple things are going to be happening:

1) Each refining/reprocessing cycle, a little bit more fuel (in the form of spent fuel, or depleted uranium tailings from enrichment) is added to the remaining fuel (while unusable waste is removed), before the reprocessed+'new' fuel mixture is formed into new fuel rods/assemblies.

2) That such a reprocessed fuel rod has essentially identical characteristics when burned in the reactor, that the fuel previously did when it was first put into the reactor before being burned.

3) That by this process of burning part, reprocessing, mixing in new fuel, reusing old+new fuel in the reactor, eventually all the fuel which was in the 'first rods/assembly' that you put in the reactor, after 10 or 20 cycles or somesuch, is all burned up, and your 'current' rod is essentially all the material you've been adding for the last 10 or 20 reprocessing cycles (sort of like if you have a lake with a river flowing into it, and another river flowing out of it, eventually all the water in the lake will have flown out the outflow river, and all the water now in the lake is 'new' water which came in from the inflow river).

4) So, you don't get 100% fuel burnup in any one fuel cycle, but you will eventually, over the course of many fuel cycles, achieve 100% burnup.

Is that correct with regards to reprocessing based designs like the Integral Fast Reactor?

[WizWom]

Kiteman, I haven't seen anyone claiming > 80% burnup for any fuel cycle; that was an advanced Thorium design.

The trouble with high burnup is that your K gets more and more unsure, and your residual heat load goes up also (as fission products are typically much higher lambda).

I guess I should ask something here. In a reactor design like the IFR, where you burn for awhile, refine the fuel, burn again, refine, burn again, etc, I assume a couple things are going to be happening:

Is that correct with regards to reprocessing based designs like the Integral Fast Reactor?

Yes, Reprocessing can help - some. You have the problem with unfissionable isotopes being formed. So, once you get too much Pu-240 or U-234, or whatever in your used fuel, you can't put it into the core again, because it will poison the reaction, and ALSO contaminate any additional breeding. The calculations for how soon this happens are nth order Products, which gets hairy pretty quick.

Once you have to do isotopic separation, you're outside the realm of cycle reprocessing.

So - Separation, Pu from U, Cf/Am from U/Pu, Th/U, etc are all reprocessing. Separating U238/U235/U234 or Pu240/Pu239 are considered refining fuel.

[Enginerd]

So, once you get too Much Pu-420 or U-234, or whatever in your used fuel, you can't put it into the core again, because it will poison the reaction, and ALSO contaminate any additional breeding.

If you get too much Pu-420, I'd venture a Nobel Prize for physics is in order...

[jsbiff]

Yes, Reprocessing can help - some. You have the problem with unfissionable isotopes being formed. So, once you get too much Pu-240 or U-234, or whatever in your used fuel, you can't put it into the core again, because it will poison the reaction, and ALSO contaminate any additional breeding.

Isn't the electrorefining system which is proposed to be paired with the IFR supposed to seperate out those poisons? Or is it the case that it seperates out MOST of the poisons, but not some of the ones you mentioned?

I guess what you're saying is, that some things formed during fission are different elements, which are pretty easy to seperate out (relatively speaking), but when you are talking about different isotopes of Uranium and Plutonium (which, by and large you want to keep), the weight differences are so small, and the chemical properties so similar, that it becomes very difficult to to that isotopic seperation (as you called it)?

The ones it can't seperate out then, are they pretty rare, so that you can go maybe 10 or 20 cycles before the poisons have accumulated to a level where you have fuel which can no longer be reprocessed?

Is it likely that anyone might come up with a process for taking the 'poisoned' fuel, and seperating out those poisons?

[KitemanSA]

Kiteman, I haven't seen anyone claiming > 80% burnup for any fuel cycle; that was an advanced Thorium design.

From everything I've read, the burn-up of thorium in a Molten Salt Reactor will approximate 100%. There are no "K" factors, "structural fuel stability", etc. as with solid fuel systems. To add fuel, put a scoop of thorium fluoride into the molten salt and there you go. Want more fuel, add more thorium salt. I've even seen where folks have said that you can just stick a rod of thorium into the molten salt until the right amount dissolves off the rod! You burned 5kg today? Add 5kg of rod. The ultimate in simple refueling!!

Ok, I am being a bit flippant here. Thorium isn't "fuel" per-se. But it is the fertile material that converts into fuel (U233). None-the-less, once you get the fuel salt set, 100% of all additional thorium is consumed. At the end of the reactor's useful life you would have effectively the same batch of salt/fuel/etc mixture you started with, ready to use in a replacement reactor. Total unused thorium, 0%. That means 100% burn-up to me.

Ok, I suppose there may be minor amounts of un-burned actinids that fall thru the re-processing cracks, so maybe 99.9% rather than 100%? 99.8%? Whatever!

[WizWom]

Yes, Reprocessing can help - some. You have the problem with unfissionable isotopes being formed. So, once you get too much Pu-240 or U-234, or whatever in your used fuel, you can't put it into the core again, because it will poison the reaction, and ALSO contaminate any additional breeding.

I guess what you're saying is, that some things formed during fission are different elements, which are pretty easy to seperate out (relatively speaking), but when you are talking about different isotopes of Uranium and Plutonium (which, by and large you want to keep), the weight differences are so small, and the chemical properties so similar, that it becomes very difficult to to that isotopic seperation (as you called it)?

Exactly.

The ones it can't seperate out then, are they pretty rare, so that you can go maybe 10 or 20 cycles before the poisons have accumulated to a level where you have fuel which can no longer be reprocessed?

Well, the half-life of Np-239 is 2.4 days; if during that time, it absorbs another neutron (about 20x more likely than a U-238 capturing a thermal neutron) it turns to Np-240, and quickly decays to Pu-240. The Pu-240 sticks, it's got a half-life of 6500 years.
Similarly for Th-233, the "bad chain" capture is 27 days with a 6x capture rate, to a 25000y half-life.

Is it likely that anyone might come up with a process for taking the 'poisoned' fuel, and seperating out those poisons?

The Molten salt reactors can pull out the "middle" products of the bad chain, and let them age into the good stuff. A travelling wave reactor won't be able to, and a solid-fuel system might do it with very fast cycling of the fuel, and measuring each pebble to get an idea of the amount of breeding chain production before feeding it back in or allowing it to "rest" to finish breeding.

Once the poisons are made, though, the best you can do is gaseous diffusion or gas cyclotron to separate the isotopes. I guess you could do an ionization/Mass spectrometer separator, too.

[KitemanSA]

Is that correct with regards to reprocessing based designs like the Integral Fast Reactor?

Yes, Reprocessing can help - some. You have the problem with unfissionable isotopes being formed. So, once you get too much Pu-240 or U-234, or whatever in your used fuel, you can't put it into the core again, because it will poison the reaction, and ALSO contaminate any additional breeding.

I've read somewhere that U233 puts out more neutrons than either U235 or Pu239 so the thorium fuel cycle can handle the minor accumulations of U234 and Pu240, etc. This however is getting a bit out of my depth.

From what I've read, the only significant issue with the thorium fuel cycle is the Pa 233 that is bred along the way. If it isn't removed and set aside to decay into U233, it will poison the reaction more than the good neutron economy can handle. Thus, either it needs to be removed or an augmentation of neutrons is needed.

The reprocessing is supposed to be very easy, but I am of two minds about that. "Easy" makes energy production more cost effective. It also makes the weaponeering easier too, I would guess. I've been told I shouldn't worry about that because of the U232 that comes along with the 233. I'm not TOTALLY convinced about that.

[Enginerd]

The reprocessing is supposed to be very easy, but I am of two minds about that. "Easy" makes energy production more cost effective. It also makes the weaponeering easier too, I would guess. I've been told I shouldn't worry about that because of the U232 that comes along with the 233. I'm not TOTALLY convinced about that.

Making weapons using a LFTR would be difficult. That said, I would certainly not be in favor of building one in North Korea or Somalia any time soon. Any neutron producing reactor could conceivably be used for weapons (no matter how dangerous and ridiculously inefficient) when you have a suitably insane and well financed dictator or terrorist group wanting to make a weapon...