Transatomic Power’s Molten Salt Nuclear Reactor | Kent Beuchert writes:
Leslie Dewan and Mark Massie, as Doctoral candidates in the nuclear department at MIT, produced a modified design of a reactor technology that first appeared many decades ago : the molten salt reactor, originally designed, built and tested at Oak Ridge Tennessee during the 1950’s 60’s and 70’s.
After her graduation, the two formed Transatomic Power in 2011, as a means of completing the design and commercializing their reactor design. Recently, they received a $2 million grant of venture capital from Founders Fund. The money will be used to test and verify the corrosion resistance of metals that their design employs in the reactor core and piping, as well as modeling the reactor design. One major purpose of the testing is to determine if the moderator’s lifespan (currently only vaguely estimated) will require periodic replacement, or will last throughout the reactor’s service lifespan. The answer to this question will not prove an obstacle to the design, only to the need for a design that provides for modulator access (for replacement).
Dewan believes one of the MSRs biggest advantage is the its ability to burn SNF (spent nuclear fuel – “nuclear wastes”) more or less completely, extracting 20 times more energy from uranium than a conventional reactor, producing a far smaller and far less radioactive final waste product, that will be much easier and cheaper to store, and will retain its radioactivity above background levels for only a few centuries rather than thousands of years. It also can be configured to burn Thorium, although that is not Dewan’s desired fuel, for several reasons : the greater perceived need to burn nuclear wastes, and the inferiority of a Thorium reactor’s proliferation safeguards, the lack of any need for an alternative to uranium fuel, as well as the current existence of a uranium fuel processing system.
The original Oak Ridge MSR design was modified in only a few ways : use of a different material for the moderator in place of the original space-consuming graphite, and slightly modifying the molten fuel salt (uranium dissolved in lithium flouride) being the most important. Together, these modifications allow for commercially competitive amounts of power to be generated, not possible from the experimental molten salt reactors built at Oak Ridge, and the ability to be powered by low level radioactive fuel, reducing proliferation concerns.
The Transatomic Power plant design has an estimated overnight build cost of $2 billion for a 520MWe unit. The lower costs are primarily due to the fact that no massive high pressure containment vessels or piping is needed for much of the plant, and also due to its higher efficiency output temps, which allow for smaller power turbines to be used. Power turbines constitute a major cost in any nuclear power plant design. With these build costs and the prospect of near zero fuel costs, there likely won’t be another power source that is cheaper, all things considered. Another advantage of the design is its ability to support load following – i.e. to alter power output quickly as demand changes. As of now only some fossil fueled and hydroelectric plants have such an ability.
This capability would allow for a larger percentage of nuclear power in any grid, which today only can exist as baseload power (although pumped storage does sometimes allow for nuclear power to be stored and then later available as hydroelectric, load following power). This plant would also likely reduce (realistically, probably eliminate) commercial prospects for the larger versions of small modular reactors, those that produce over 250 MWs. It achieves (actually, exceeds) the economies of scale of a conventional large reactor, something small modular reactors are totally incapable of. It also does not require shutdown for refueling – it is refueled at intervals and can be run continuously for decades, another cost advantage over conventional reactors.
Commercially, a $2 billion dollar 520MWe power plant can meet the needs of utilities facing only gradually increasing demand as well as those whose service area doesn’t require the 1000MW plus size of a conventional reactor, or those wanting to replace fossil-fueled load-following or baseload coal plants. Build time is estimated to be 36 months. The plant does not require any source of cooling water – the molten salt fuel liquid acts as its own coolant as it flows thru the primary loop, transmitting heat (but not radiation) to an intermediate loop. The lack of any requirement for cooling water dramatically increases the number of potential build sites. In fact, its inherent safety characteristics would allow for these reactors to be sited near large population centers, avoiding lengthy transmissions, reducing costs due to power losses and transmission line construction.
In considering the reactor’s characteristics with respect to safety, it’s hard to conceive of a situation that anyone would find threatening or dangerous. Every reactor state that one can reasonably imagine as conceivable, ends up with the reactor shutting down as the molten salt cools (slightly) and becomes a solid at which point no fission is possible (or heats up, reducing fission). No radioactive material (the molten fuel salt liquid) is ever subjected to anything other than slightly above atmospheric pressures, which essentially eliminates any radioactive blast issues, ( in fact, all pressures work towards forcing the radioactive material back into the reactor system), and no hydrogen emissions can develop to the point of posing an explosive danger.
The entire steam turbine system and its piping, which contains the only only material under significant pressure (water), is completely radiation free, meaning that any rupture in that system is, radioactivity speaking, a non-event. The fuel liquid operates at a much lower temperature than fuel rods in a conventional reactor and never contains the excessive reactivity potential possessed by a conventional reactor at the start of its two to three year run cycle.
The entire system is considered walk-away safe – no operator actions, or electricity, or pumps are ever needed in order for the system to shut down should an accident occur. The reactor will achieve a stable shutdown state in a fairly short time frame.
It’s hard to imagine anyone having any objections to this nuclear reactor design. For those attracted to the safety of Thorium reactors or concerned about future uranium fuel supplies (such as India) , this design is better all the way around. With its meager fuel requirements, uranium will be economically available from either SNF, terrestrial mining or ocean extractable uranium ( freely available to virtually every country) for many millennia, eliminating any conceivable concerns about future fuel sources. And, of course, it can burn Thorium as well, should anyone so desire.
The superior economics and flexibility and load following characteristics, lack of any need for refueling shutdown, elimination of any significant fuel cost increases, removal or reduction of nuclear waste storage requirements, a much lower build cost requirement and elimination of the economic danger of a multi-billion dollar nuclear accident, would certainly make these reactors the first choice for any grid operator. And the plants can easily be co-located with conventional nuclear plants without placing any additional water demands, and be located near the likely source of the SNF they will consume as fuel.
One finds it difficult to foresee any significant risks when buying into this reactor design, financial or otherwise. I see a real possibility for this reactor design to become not only the standard and universal nuclear reactor, but also the standard commercial power plant as well, rendering all others of this size and larger obsolete. It is everything one could reasonably ask for in a power plant.
A complete technical description, accompanied by economic and safety rationales can be found at the company’s website :
Click to access TAP_White_Paper.pdf
This design has my attention. Their strategy of concentrating on the spent uranium fuel rods as a resource is practical and pragmatic.
My problem with the Thorium reactor concept is that no matter how save the nuclear side of the engineering, it is an enormous chemical/metallurgical refining engineering and operations problem. I have seen estimates that 10% of a Thorium core needs to be continuously reprocessed per day
Chapter 8 of their White paper discusses Thorium reprocessing, particularly in regard to the removal of the poison Pr-232 which decays to U-233 (weapons grade) which can then be fed back to the core. But they do not discuss the reprocessing cycle.
I see no mention of amounts extracted per unit time or salt reprocessing throughput.
Correction: In regard to the Thorium reprocessing cycle, the poison is Pa-233, not
Pr-232Really just more FUD to support their favorite technology. Chemical processing is far easier than isotopic processing, so getting the 233Pr out of the salt isn’t that big a deal. 232Pr comes along with it (chemical not isotopic) which is important as we’ll see in a moment. The proliferation risk is relatively low (233 BOMBS KILLZ US ALL!!) since it is mixed with 232U which is a hard gamma emitter that is easily detected globally as well as very hazardous to anyone who wants to make said bomb. And we already touched on how difficult isotopic separation is relative to chemical separation.
All of that would say problem solved, but it’s not quite that easy. The half life of 232Pr is different from 233Pr, so if you’re willing to let your extractions sit for a while and pull out all of the Uranium (mostly 232U) for the first few half lives, you’re left with mostly pure 233Pr which you can allow to decay to 233U. The answer to that is to dump some 238U in the mix either to poison the extract if something suspicious is going on, or continuously if you’re willing to take the efficiency hit as it absorbs neutrons. It will be most entertaining to read how this couple thinks they can run their reactor with SNF containing a significant chunk of 238U that they will have to breed to 239Pu to burn and yet somehow THAT does not pose a proliferation risk…
Tsk Tsk: do you want to revise your reply to Rasey? As Rasey corrected, it is Pa-232 & Pa-233, not Pr-232 & 233 as you post. Pr is Praeseodymium, Z=59 and not a fissile material.
Another mistake you make, U-232 is an alpha emitter, not a hard gamma emitter. U-232 emits some gammas but the intensity (number of emissions from lots of decays) is very low and the energies are relatively low. The four primary gamma emissions energies and intensities are* : 58 KeV/0.21%, 129 KeV/0.082%, 270 KeV/.0038%, and 328 KeV/0.0034%. You would have to be handling a LOT of U-232 to receive any harm from these levels of gamma emissions and they are easy to shield and most would be easily shielded by the uranium where they are produced. To help explain intensity further: with a given quantity of U-232, you will only see these gamma emissions at any one time in accordance to the percent noted. So, if at the same moment a thousand U-232 atoms decay (alpha decay), you will only see the about two 58 KeV gamma emissions along with the 1000 alpha emissions. The other gamma emissions are much less.
*data from an old handbook and there may be slight difference from newly measured data.
… It’s hard to imagine anyone having any objections to this nuclear reactor design. …
Not so hard when you’ve met folks who are either afraid to open a microwave, or refuse to use one because the microwaves might “escape.” No joke.
I wonder if/how they recover if an incident occurs and the freeze valve causes the fuel to drain into the containment tank. They indicate the reaction stops nicely and the tank cools automatically. Seems like the containment tank then contains a huge blob of solid salt. Is this the end of the reactor or can it be restarted?
Gas cooled reactors can be quenched quickly by injecting boron dust into the coolant gas. However, my understanding is that it is generally impossible to remove enough of the dust to ever restart the reactor. So pushing the panic button on a gas reactor costs several billion…
There is also something called a Traveling Wave Reactor (TWR) being investigated by the Bill Gates-chaired company TerraPower. They are also looking into MSRs. However, it could be a couple decades or so before they (either TWR or MSR) become commercially viable.
Universally acceptable with the word “nuclear” in the name? No way. Rebrand it.
That’s a good point. Let’s start calling it evergreen. Evergreen Power, powering the world for eons to come!
RE: Conrad’s “Reuters” link.
Reuters starts out xenophobic and then moves into the actual relationship.
Our countries are working on more near term relationships in nat. gas & COAL “OMG” as well.
Here is one of the joint presentations. As noted in my first response, at last we have a horse race. So, I’m hopeful that acceleration will occur. Utility scale programs require partnerships, I.P. sharing and contracts to drive prices down. No country or company has a silver bullet, winning solution.
I apologize up front. I do not understand why WordPress automatically linked to the previous post, but didn’t in this case.
Very knowledgeable comments here immediately–impressive.
Incidentally (speaking of breakthroughs), here’s a new fuel cell design in testing that runs on natural gas (hit page-down five times once there) (story is from June 26, 2014):
http://www.dailytech.com/Microsofts+New+Fuel+Cell+Partner+is+Ready+to+Blow+Away+the+Bloom+Box/article36118.htm
Redox Power believes it’s ready for its first serious commercial test in the wild. The startup is a spinoff from the University of Maryland Energy Research Center (UMERC). Launched in Aug. 2013, the company continues to collaborate with the Univ. of Maryland.
Redox Power’s founder, Professor Eric Wachsman, is an instructor at the university and is director of UMERC. He holds key patents on the technology which he claims will offer 100 times the density per cost of current cells, including Bloom’s Energy Server. He claims his cells are 1/10th the cost of commercial alternatives and are also 1/10th the size.
One strength of Redox Power’s cell design is flexibility. It is designed to primarily run off natural gas, but can also generate power using propane, gasoline, biofuel, and hydrogen. At its maximum efficiency, when processing natural gas and doubling as backup heaters, the cells can output heat and electricity at 80 percent efficiency (and 70 percent efficiency for electrical generation only).
That’s a good deal higher than Bloom Energy Servers, which are 60 percent efficient at optimal conditions.
But will it power my car?
Cheers
Roger
A quick tutorial for those not familiar with neutron interactions with matter. Neutrons do not interact electromagnetically with matter like other radiation emissions do. They have to bang into something to dissipate their energy. As a neutron crashes into molecules, eventually it loses its energy and either becomes a hydrogen molecule (by collecting an electron) or it can get captured by the nucleus of an atom. In the latter case it changes the atom, is called neutron activation, and is the source many radioactive contaminants in a reactor.
Besides heat being generated during fission of a Uranium atom, the interaction of a thermal neutron with other atoms will also generate heat. Heat is the layman’s term for the excitation of molecules which we experience as warmth (or lack thereof). In the TAP reactor, fission produces heat which is transferred to the LiF based salt. Additionally, thermal neutrons generate heat by interacting with the salt.
There are four basic neutron classifications: thermal neutrons (a very low energy of approximately .025 to .038 electron volts or eV); then slow, intermediate and fast neutrons indicating higher energy ranges. As a uranium (or other fissionable atom) fissions, it will release neutrons in any of the classification ranges. Because heating occurs principally in the slow and thermal ranges, for a power reactor the idea is to slow all of the neutrons down to the thermal range. This is where the moderators come into play. Some materials are quite good at slowing down neutrons, and other materials are nearly opaque to higher energy neutrons. For example, hydrogen, water, heavy water, boron, and beryllium are good at slowing neutrons. Lead (Pb) is poor at slowing down neutrons. Physicists use the term cross-section to indicate a material’s ability to slow down neutrons; a high cross section material is a desirable material to slow neutrons. Apparently Zirconium Hydride (I’m unfamiliar with it) is fairly good at slowing down neutrons.
Further fission will occur if a uranium atom captures a neutron. The moderators are adjusted to increase or decrease the number of neutrons being slowed down thus increasing or decreasing the number available for uranium to capture.
I hope the above helps a few. Understanding the basic underlying principles can help to understand the implications of the TAP reactor. The principles have been applied to molten salt reactors for a long time, but the engineering problems associated with the salts (principally Sodium) have been daunting. The use of LiF based salts looks most promising if it solves the corrosion problems. The press release and white paper are correct in pointing out the advantages of a molten salt reactor, and those advantages are huge.
Sodium cooled reactors are not MSR’s. They have had lots of problems and are a bad choice given Sodium’s reactivity. The salts used in MSR’s are chemical stable and basically inert, so even if they were to leak out of the reactor, there would be no explosions are fires. You would still have a mess to clean up, but there’s no risk of a hydrogen explosion like there would be with a Sodium cooled reactor (a remarkably BAD idea would be to put such a thing on something like a submarine like the Soviets did).
These reactors are breeder reactors so the moderator exists to ensure that the neutron energies are appropriate to the cross section of the fuel. Pu breeders require fast neutrons. Thorium requires slow (thermal) neutrons, which is a point that this press release fails to note.
I hadn’t looked up the LiF salts’ reactivity, so thanks for filling me in that it is inert. Concur that Sodium reactors were a bad idea. I did get to visit the Fast Flux Test Facility at Hanford and despite the sodium problem, it was pretty darn cool.
There’s definitely some bias in the white paper. The discussion about not using Be salts because it’s toxic to 10% of the population is just silly. CuBe is used in lots of places and Be is used in other alloys. So that doesn’t help their case.
Then we have this example when being critical of Thorium:
“Therefore, even with U-232 mixed in with U-233, it may be possible to chemically extract any decay products produced from U-232 before they become gamma emitters, thereby leaving weapons-grade uranium that is not protected by high energy gamma radiation.”
So chemically extracting 233Pr (with a half life measured in weeks) is hard, but chemically separating the hard gamma emitters from the 232U decay chain, some of which have half lives measured in minutes or seconds is feasible. Yeah, your bias is showing.
Finally they make a claim that a thermal spectrum reactor can function just fine with 238U. If it were that easy we could run our PWR/LWR’s today in thermal mode, but we can’t. CANDU’s can do that, but that’s a very special environment. I’m skeptical of their claims as a result and think they really need to be running in a fast mode if they want to stay with a Uranium breeder fuel cycle contrary to their claims, or they simply get more efficient burn up of the fissile rather than fertile materials. That is certainly plausible in an MSR.
There is no such thing as a universally acceptable power plant. The communists who commandeered the environmentalist movement don’t want safe, clean energy. They want to destroy capitalism. A safe, clean energy source doesn’t help them.
The fundamental problem here is the same problem trying to commercialize anything other then a light water reactor.
The thorium fans, fast breeder fans and fans of this design all face the same daunting hurdler.
There is a fuel fabrication manufacturing facility to be developed and someone will have to foot the bill to establish the fuel cycle fabrication facility.
(Justifying government funds was easy for light water reactor fuel fabircation…we needed the plutonium waster for bombs)
This is a difficult financial case to be made unless an order for ‘dozens’ of a reactor type are pending.
Of course…no one is going to order ‘dozens’ of anything until the first one has a bit of a track record.
Nice design…I wish them well. They problably should chat with the Chinese and Indian’s about maybe some ‘knockdown’ pricing for an order of a dozen in exchange for fuel fabrication facility funding.
Hey, throw a few billion this way, Obama!
No. It might work
Hell this is old news, if you wish to LEARN about LFTR technology check out Kirk Sorensen on Youtube and watch the long versions, he tells it all and has been advocating this for years.
I have worked in power systems and [worked] with the NRC.
Mr laarrry Geary’s post is quite correct.
The cost to commecialize [this] technology would be immense and very time consuming,having nothing ot do with other then paper studies and safety reviews of every [conceivable] and even inconcievable failure possibilities. Plus a substantial number of the voting commissioners are characterized as rabid Green eco-loons placed there by politics and the green community, to ensur [another] test or study be conducted to delay the process. Starting today, and nobody is really doing so, I’d wager that a Fusion [powerplant] not even begun to design yet will be adding power to the grid before such an MSR plant could do so.
The primary technical problem has always been the [materials] to contain and transport the corrosive MS working fluid. Then there is the problem of contamination from leaks between the MS and the turbine working fluid. Leaks will develop.MS and H2O do not mix well.
I recall reading of problems in early MSR reactors, with the MS as pockets of unevenly mixed actinides could result in occasional hot spots of enriched fuel leading to spots or regions of increased output and possible critical mass being achieved. All these problems could eventually be solved
But the real problem is politics and delay.
As he so aptly says:
“I’m concerned that the political/regulatory regime will effectively kill this like they have conventional nuclear reactors. They could take decades to approve the design. (Visit the website of the Nuclear Regulatory Commission and see their level of urgency for getting anything done.) Environmental lawsuits could prevent start of construction for another decade, and continuing lawsuits mandating construction halts and design changes could stretch out construction to 10-12 years beyond that. Reason does not rule in this area. The average age of a nuclear reactor in the US is 39 years, and the last one was built in the 1990’s. The same people complaining about CO2 and coal emissions work their tails off to prevent nuclear power from solving those issues. And most people seem to think nuclear = bomb, so they are easily swayed by green propaganda. The company may have to go to Russia or China to actually get anything built.”
Terrestrial Energy is working on a simplified MSR based on the ORNL design. China has a crash program underway. The problem is our US government. The NRC hasn’t allowed a new reactor design in 40 years. MSRs are the way to go and will be built anywhere but here in the US where we invented it…and I am pissed!
Leadership does matter. Hopefully the U.S. is in the process of changing ours. We’ll see.
Someone should tell Gov. Brown and spend the money for his $25b Delta tunnel and make these reactors to power desalination plants for So Cal.
Molten salt reactors have been around for quite a while. EBR2 was used in a 1986 loss of cooling accident demonstration, and it passed with flying colors. Another descendant of EBR2’s design was the Integral Fast Reactor. IFR never got built, but its design had many of the passive safety features that this new one has, plus it had a built in reprocessing plant on site. The idea was that all the fuel it would ever need could be delivered at the end of construction, and it would breed and recycle its own fuel for at least 40 years with minimal waste.
@Tsk Tsk 8/27 6:14 pm
Finally they make a claim that a thermal spectrum reactor can function just fine with 238U. If it were that easy we could run our PWR/LWR’s today in thermal mode, but we can’t….. they really need to be running in a fast mode if they want to stay with a Uranium breeder fuel cycle contrary to their claims, or they simply get more efficient burn up of the fissile rather than fertile materials. That is certainly plausible in an MSR.
On page 14 of their White Paper, they very briefly mention that they use two zones in their reactor, no doubt with the salt flowing between them, with different neutron spectra.
I’m just speculating here, but imagine a donut shaped reactor, with fluid salt flowing clockwise. in one part of the donut is moderated for fast neutrons (the Zr-H) for fission and actinide removal. It is this area in which the poison production is a problem (hypothesis). But in time the flow moves into the unmoderated section where (and here I’m really speculating) the fast-neutrons poisons are less of a problem and have time to decay or transmute in the thermal and epi-thermal neutron spectrum.
Rather than attempt to reprocess 10% of a Thorium LFTR core per day to remove poisons by chemical and electrochemical brute force, You move the TAP salt into a different neutron environment to breed the U-238, and wait out the poisons. All that mess stays within the hot donut. Elegant! In fact, there is no reason to limit yourself to two zones if a third zone neutron spectra could be advantageous.
Mind you, I’m not sold yet. They mention the need for poison removal. Perhaps their two-zone methods have greatly reduced the processing volume rate. Maybe they have reduced the number of elements and isotopes they need to remove — all good. But they are relatively silent on the details. It is what people don’t tell you that you must watch for.