Development of LFTR equipment technology, testing of the design and construction, and construction of factories to produce them: ~$5Billion.
Build 100MW LFTRs on assembly lines: ~$200 Million.
Fuel for 1GW electricity in a LFTR or any MSR: $10,000/yr. (Compare to LWR: $50-60 Million/yr.)
Generating 1 GigaWatt electricity in a Molten Salt Reactor generates $595 million to $690 million of electricity per year (varies based on the current local price of electricity), plus:
- 150kg xenon @ $180,000.
- 125kg of neodymium @ $150,000
- 15Kg Pu-238 (only Pu-239 is fissile) for radioisotope power @ $75M-150M
- 20kg medical molybdenum-99, plus 5g Th-229 (decays to Bi-213 for cancer treatments)
- 20Kg radiostrontium for remote heating units.
Radioisotopes and Medical Isotopes from LFTR
Low estimate of those stated amounts: $670,330,000/GW-year.
So, at 20% markup for the 10 x 100MW reactors, producing 1GW electricity for a year, plus an additional reactor “on standby”, the buyer would be making money in under 5 years. Have an additional reactor “on standby” for maintenance or any problems, so the city would have essentially zero downtime in electricity production (no needing electricity from long-distance sources).
Get additional reactor capacity to use the heat for industrial processes. One use would be more energy-efficient (at higher temperature) desalination of water. Another use would be making CO2-neutral vehicle fuel from H2O + CO2; the heat of the reactor would, for some MSR designs, be enough to directly split water and carbon dioxide, which could then be used for making gasoline with no pollutants (depending on the purity of the water and carbon dioxide).
Capital costs are generally higher for conventional nuclear versus fossil-fuel plants, whereas fuel costs are lower… because the construction, including the containment building, must meet very high standards; the facilities include elaborate, redundant safety systems; and included in capital costs are levies for the cost of decommissioning and removing the plants when they are ultimately taken out of service. The much-consulted MIT study The Future of Nuclear Power, originally published in 2003 and updated in 2009, shows the capital costs of coal plants at $2.30 per watt versus $4 for light-water nuclear. A principal reason why the capital costs of LFTR plants could depart from this ratio is that the LFTR operates at atmospheric pressure and contains no pressurized water. With no water to flash to steam in the event of a pressure breach, a LFTR can use a much more close-fitting containment structure. Other expensive high-pressure coolant-injection systems can also be deleted. One concept for the smaller LFTR containment structure is a hardened concrete facility below ground level, with a robust concrete cap at ground level to resist aircraft impact and any other foreseeable assaults. Hargraves, American Scientist Vol 98, July 2010
Limited decommissioning cost vs. LWR systems, infrastructure & waste. Total development cost for Th-MSR may be less than ‘true cost’ of decommissioning current LWR. Th-MSR’s value in burning existing waste may offset its total Build & Operational Cost. Kennedy TEAC3
The U.S. nuclear industry has already allocated $25 billion for storage or reprocessing of spent nuclear fuel. FLiBe Energy. [Perhaps the company that builds LFTRs will get the contract? Development of LFTRs and construction of manufacturing plants for the entire world would cost less.]
For economics of solid fuel reactors, including construction, fuel, waste, decomissioning, see Doty Energy – Fission.
Would you happen to know offhand how much U-233/yr could be realistically bred in radial & or axial Th-232 blankets surrounding an ALMR-PRISM reactor? I understand that LFTR fissile requirements are vary low (~100kg) and that the ORNL U-233 stockpile is about 500kg; is this correct?
Generally a LFTR would be designed to breed barely over what it would consume. It can be tweaked by changing the salt slightly, by adjusting the size of the core, etc. The “Fast Spectrum Molten Salt Reactor Options (ORNL)” goes into this in more detail. I don’t know any details about an ALMR reactor.
Flibe-Energy.com says “LFTR can also be used to consume existing U-233 stockpiles at ORNL ($500 million allocated for stockpile destruction)” but I’m not sure how big the stockpile is. Let me know if you find it, okay?
How much does the building of a GW reactor cost? I’m not entirely sure what the $5 billion implies in the first sentence.
$5 billion estimate for all the materials testing, reactor design, design of fission product handling equipment, pilot reactor construction and testing, and factories to mass-produce the reactors. I don’t know we’d ever build a single reactor that generates 1 giga-watt electricity, since 200MW reactors can be shipped in standard trucks and used in more locations; of course multiple 200MW reactors can be at a site.
100MW @ $200 Million –> $2 Billion for each gigawatt capacity, at current prices for electricity would pay for itself in under 5 years.
(I don’t know how much of that $200 Million would be profit or recovering development expenses, and how much is materials and labor etc. Keep following Thorium Energy Alliance conference talks.)
Thanks for the reply!
Another question, would the statistics for a 1 GW LFTR above be about the same for 5 200MW LFTRs?
Almost. One big reactor would be harder to ship and install than several small ones (small can be completely assembled in factory, shipped in one piece and connected to heat transfer unit and electric generator and vehicle fuel maker and water desalination; big ones would need some on-site assembly), but big would have slightly less total metal. The fission and waste is identical, per GW-Yr.
Is the 1 gigawatt reactor produce 1 gigawatt a day or 1 gig watt a year?
[George – 1 gigawatt-hr per hour, or 1 gigawatt-year per year. A Watt is defined as one “joule per second and can be used to express the rate of energy conversion or transfer with respect to time” [Wikipedia]. A gigawatt is equal to one billion (10^9) watts. A gigawatt reactor would be expected to be able to produce a gigawatt of power at any moment; actual energy output required would of course fluctuate, so over a year it would probably average less than the rated 1 gigawatt. The “capacity factor” of MSR would be very high, not needing to be stopped for refueling, but would likely have some maintenance. Wikipedia “Wind Power” says “German nationwide average wind power capacity factor over all of 2012 was just under 17.5% (45867 GW·h/yr / (29.9 GW × 24 × 366) = 0.1746), and the capacity factor for Scottish wind farms averaged 24% between 2008 and 2010.”, I would expect MSR to be capable of around 99%, electric power demand likely lower than that, so we’ll want to have MSR do more than make electricity. We’ll likely make 200-250MW reactors, since these can be shipped in standard trucks, and install them throughout an area.]
Does anyone know that the US had a LIFTER [no, a MSR, without converting thorium to fuel; LFTR is a specific design of MSR] up and running just fine for about 5 yrs in the sixties, I believe at Livermore Labs [nope, Oak Ridge National Laboratories]. Despite solar costs approaching Moore’s Law, it simply isn’t practical every where. I think LIfTErs are an ideal “bridge” solution which is being held up by the existing fossil fuel interests and politics. I’d be interested in seeing the life cycle cost comparisons of solar vs. LIfter. Any engineers out there looking at this? Kirk Sorensen maybe? As we all know, fusion power will be only 50 yrs away, in another 50 yrs.
Are these figures still accurate or have they moved since?
The principal benefit of a LFTR is that the design already exists. You can scratch that 5 billion dollar design cost as it has been designed and tested already. It has added advantages of very little waste, medical isotopes that can be sold, thus lowering operational cost. LFTR’s don’t require tons of toxic chemicals to make like solar panels do. Best of all, zero chance of a meltdown.
[George’s reply: The 1960s Molten Salt Reactor Experiment demonstrated the type of reactor works, with much greater safety and much better fuel use than the Light Water Reactors we have been using. However, we would not use the exact design of the MSRE. We would use a design optimized for manufacturing many reactors, not a design for scientific testing; we would probably use a design using improved modern materials; we would use a design with modern sensors and monitoring. That money would also be for designing the factory.]