Fuel and enrichment requirements of EPR vs. LMFBR equivalent

[engineerpoet]

This is an analysis and contrast, and invitation for comment.

Back some 8 years ago, Barry Brook listed the fuel requirements for a couple Gen 3 nuclear reactors (AP1000 and EPR).  I can't find the total fuel inventory for AP1000 and have little interest in EPR, but over time the initial fuel load becomes irrelevant to the total consumption.  Ergo, the annual replacement fuel requirement is good enough for ballpark estimates over a period of decades.

EPR is more efficient than AP1000.  Brook gave the figure of 25.3 tLEU/yr at 5% enrichment for EPR.  My enrichment spreadsheet (see previous post) says that producing this fuel requires 237.7 tNU/yr at 0.2% tails.  EPR's stated output is 1650 MW(e) at 38% thermal efficiency, or 4340 MW(t).

LMFBRs operate at considerably higher temperatures, so should be able to achieve substantially higher thermal efficiencies.  A modern steam system fed by either a Fermi I or S-PRISM class reactor should be able to hit 45% efficiency without undue difficulty.  My own calculations suggest that 5000 psia steam with 2 reheats, which is typical of many fossil-fired plants, would suffice to get there.  I'm going to assume 45% efficiency for LMFBRs in this piece.

At 45% thermal efficiency, 1650 MW(e) is produced from 3667 MW(t).  This is 3 and 2/3 S-PRISMs or 18.33 Fermi I reactors.  I'll go with the Fermi I because its initial fuel load was 25.6% enriched uranium, not plutonium.  This makes the fuel requirements directly comparable to EPR.

Fermi I's initial fuel load had 484 kg fissiles at 25.6% enrichment, or 1891 kg total uranium.  Producing 1891 kg U at 25.6% enrichment and 0.2% tails requires 60.04 tons natural uranium at 0.711% U-235 feed, and 90422 SWU of enrichment work.  Fueling 18-1/3 such units would require 1101 tons NU and about 1.66 million SWU.  After the initial fueling the units would be a net producer of fissiles, and the DU tails would provide sufficient blanket material for several times the life of the plant.

The EPR requires 25.3 tLEU/yr at 5% enrichment.  Produced from NU at 0.711% U-235 and 0.2% tails, this requires 237.7 tNU/yr and 223,900 SWU/yr.

Something is rather immediately obvious:  in just 5 years, the EPR consumes more total uranium than the LMFBR equivalent.  In 8 years, the EPR uses more total SWU as well.  If there was a comprehensive effort to build out nuclear capacity to replace fossil fuels, the best PWR on the market requires substantially more raw fuel material and enrichment work in its first decade of operation than the equivalent in LMFBRs, and those being a design that's over 50 years old already.  Over a 30-year program (which an effort to replace fossil fuels would probably take) the difference in favor of FBRs is even more lopsided.

This does not include the effort required to reprocess FBR fuel, but the energy requirements of pyroprocessing are comparable to electrowinning of other metals and should be quite cheap by comparison to enrichment.

Conclusion:

Perhaps my spreadsheet is in error.  If it isn't, this is a rather convincing argument in favor of switching to FBRs as soon as suitable designs can be tested and standardized.  LWRs are okay but they are not the best we can do.  They are not the best for thermal efficiency, they are not the best for uranium consumption, and they are not the best for minimizing fuel supply effort.  We need to move on.

[Roger Clifton]

If there is to be a nuclear surge, a politically driven mass rollout of nuclear generation, and if the bottleneck is the provision of fuel, the scenario of fast neutron reactors would lag the slow neutron reactors by eight years. That is a significant lag, arguably the amount by which the Soviets lagged the US moon effort. Perhaps that is simply an argument in favour of laying out both reactor types in the surge.

[engineerpoet]

Jun 18, 2019 21:54:41 GMT 9.5 Roger Clifton said:

If ... the bottleneck is the provision of fuel

That's one aspect, but another is that LWRs have produced a vast amount of low-fissile material that is, if not an actual storage problem, a big political headache.  Even reprocessing this material as MOX does not truly get rid of it.  Nuclides like U-236 and higher isotopes of Pu do not fission well in a thermal spectrum and act as poisons, so there is a limit to how much can actually be consumed before it can no longer be used in LWRs.

Only fast-spectrum reactors can consume these nuclides.  This requires a reprocessing industry to keep them fed, but we appear to know how to do this and it was only politics that stopped it from being demonstrated at full scale in the 90's.  The world inventory of used LWR fuel has to be in excess of 200,000 tons by now; AFAIK only France has been reprocessing everything and a great deal of its recovered Pu is sitting in storage.  It makes sense to build out the reactors to put it all to work.

Once you have a fast-reactor buildout in progress, you have to ask:  does it make sense to build more LWRs as well, or has their time passed?  You know what I think the answer will be.

[Roger Clifton]

Jun 19, 2019 1:11:18 GMT 9.5 engineerpoet said:

a vast amount of low-fissile material ... a big political headache

It will not be the biggest political issue when the angry climate hits every home. Rather, it will be the dangerously vast quantity of accumulated waste CO2 that will (politically) drive a global surge of nuclear generation. The greenhouse and ocean have about 1000 gigatons too much of the stuff. Blocking escaping heat are two kilograms per square metre of excess CO2. It is increasing by more than 40 g/m² per year. Everywhere. In contrast, the equivalent slow reactor fleet would generate a relatively negligible quantity of used fuel, sitting in someone else's backyard. And yes, it can be recycled, eventually. Then in good time only the fission products need be buried, an even tinier one gram per year per kilowatt. That eventuality is already in the Chinese plan for nuclear up to 2100. The rest of us can plan likewise.

[engineerpoet]

Jun 19, 2019 6:58:01 GMT 9.5 Roger Clifton said:

It will not be the biggest political issue when the angry climate hits every home. Rather, it will be the dangerously vast quantity of accumulated waste CO2 that will (politically) drive a global surge of nuclear generation.

That's closing the barn door after the horses have gone.  By the time the effects become that obvious, considerably more change will already be "locked in" even if emissions halted immediately.  Building more nuclear only when the need breaks through the walls of denial is far, far too late.

[Roger Clifton]

Yes, it is already too late to preserve the climate of the IGY, 1957, when the experts blew their whistle. It has been getting worse ever since. It will stay bad,too, but it will never be too late to stop it getting worse.

[Roger Clifton]

Jun 19, 2019 8:19:36 GMT 9.5 engineerpoet said:

Building more nuclear only when the need breaks through the walls of denial is far, far too late.

There is a problem here of who is "we". I could have said, "we will only be able to build a surge [massive rollout] of nuclear generation when the need breaks through our denial". That breakthrough requires commitment to arise in all sections of the community, including the singers and marchers. What the technocrats can do in the meantime is prepare the designs for mass production.

The NuScale design appears to be the leading LWR, already on its way to the factory floor. Can you see a lead contender for mass production among the fast reactor designs? For that matter, is there a mass producible MSR in sight?

[David B. Benson]

Roger Clifton --- Thorcon clearly is designing for ease of production, ship yard style. Moltex Energy is also aiming for a $2/W design. I don't see that either company is well funded the way Nuscale is.

[engineerpoet]

Jun 19, 2019 17:07:08 GMT 9.5 Roger Clifton said:

Can you see a lead contender for mass production among the fast reactor designs?

Rosatom has the BN-800 in operation already.  GE Power has offered to build two S-PRISM units on spec for the UK to get rid of their unwanted plutonium, with the only proviso being a power purchase agreement (no takers so far).  That puts BN-800 way out front for readiness, but with S-PRISM as a serious contender.

S-PRISM is air-cooled for emergency shutdown, making it walk-away safe.  Dunno about BN-800.

Since I'm an American chauvinist, if I had the national credit card I'd order up at least 3 S-PRISM units to be built serially.  The first two would go in at the former NRTS in Idaho, and would be used to prove things like walk-away safety and other features.  After that, I'd use them to test things like solar-salt heat storage to permit load-following without changing reactor power.

There's more than just the reactors, of course.  Dunno what I'd do about the S-PRISM fuel supply, but emptying out the fuel pools at Columbia Generating Station and San Onofre would be quite popular with at least a few people.  The pilot fuel plant would go in at NRTS, of course.

[David B. Benson]

NRTS, National Reactor Test Station, is now named INL, the Idaho National Laboratory.