Pebble Bed Advanced High Temperature Reactor at UC Berkeley – low cost nuclear?

When I visited California earlier this month, Tom Blees and I paid a visit to Prof Per Peterson and Prof Jasmina Vujic at the Nuclear Engineering Department of UC Berkeley. After chatting over lunch, Per took us on a personal tour of his lab, which was quite an experience. Per’s research focuses on development of a high-temperature reactor with an incredibly high power density. Why? In short, it’s about the money. Per’s argument — and a quite persasive one — is that if the costs of advanced reactors can be brought way down, below that of pressurised and boiling water reactors (PWRs and BWRs), then their scaled-up deployment is highly likely. The following post owes a lot to Per’s insights on this critical issue.

Currently, one the most frequently cited criticism of nuclear energy, especially with reference to Europe or North America, involves economics. High construction costs for Advanced Light Water Reactors (ALWRs) have emerged as the number one issue limiting near-term deployment, and it now appears that the $18.5 billion in loan guarantees now available will fund no more than 2 or 3 new plants. The major area of anti-nuclear emphasis today is on preventing an expansion of this loan guarantee volume to the $50 to $100 billion level that the nuclear industry believes could be productively used in the near term. Even with loan guarantees, cited nuclear construction prices in the US remain high enough that nuclear remains marginally competitive and most utilities are slowing down their plans for new nuclear construction. Really, nuclear is getting nowhere very fast in the US at present, despite its great promise. AREVA France is now facing similar issues. China, happily, is not.

The main issue with Generation IV reactors such as the IFR or LFTR is the general expectation that they will be more expensive than ALWRs — at least in the early stages of deployment. Increasing the cost of new nuclear construction can hardly be viewed as a winning strategy these days.

For instance, a lot of design work was done by GE on the S-PRISM, after Department of Energy support ended, to bring down the cost. But it still needs to be updated to take into account new construction technologies and requirements (including aircraft crash). It would be very helpful to be able to argue convincingly that IFR technology will be less expensive than ALWRs. If this could be shown to be the case, one could also expect more substantive commercial interest and investment, such as a willingness to cost-share the Design Certification and to construct a prototype reactor outside the federal appropriations process (for example, under loan guarantees with some federal contract for procuring fuel irradiation services for transmutation fuel development and demonstration). Members of SCGI are working behind the scenes on these key issues, and progress is being made, but it’s naturally a protracted process.

Per argues that fluoride-salt reactor technology (AHTR/LFTR) has a clear path to achieve substantially lower energy production costs than ALWRs. His expectation is that this evolutionary path will remain focused mainly on thermal-spectrum reactors, with efforts to push to higher temperatures and efficiency, and the introduction of thorium.  Sodium-cooled, metal-fueled reactors are intrinsically bulkier and lower temperature/efficiency than AHTRs and LFTRs, but are not intrinsically more expensive than ALWRs. IFR is more mature than AHTR and LFTR, so the big question is what will be the most practical route to commercial demonstration. IFR will be a tough sell, though, if the general perception remains that it is more expensive than ALWRs. This is a complex topic, which I will endeavour to do more justice to in later posts in the IFR FaD series.

The 410MWe PB-AHTR core (left) next to the 168MWe PBMR right.

So, back to Per’s lab. He has various engineering models set up to test movement of TRISO pebble fuel through a fluoride salt coolant, whereby the pebbles are inserted in the inlet pipes and rise up through the reactor module over time, and then are put back through 5 or 6 times. This allows for very high burnup — exceeding 50 %, high power density due to the heat capacity of the liquid salt, and high temperatures thanks to the durability of the pebbles. This is a big (potential) advantage over the current Pebble Bed Modular Reactor technology (PBMR), because in that design, the gas coolant has a very low power density. He’s flipped the problem on its head. The reactor also has various inherent safety design features, such as control rods that sink naturally in response to elevated coolant temperature, thereby passively regulating reactivity. Very safe!

His testbed lab units use analogue fluids, including water and oils, and synthetic pebbles made from a nylon-like material. The model core of the reactor stood about 2 metres tall, and I asked what the power output of a full-sized 4 m tall (2 m wide) reactor core unit would be. Try 400+ MWe. Wow…

I’ll end this post with something a little more technical, if the above wasn’t already too techie for you (apologies to some BNC readers). Below I reproduce Per’s summary of the PB-AHTR, which he wrote up late last year in response to some prompting from me and others on the IFRG (a nuclear energy mailing list the Per and I, and many others from SCGI, are part of). It’s a terrific summary of Per’s research, for those with an engineering background or nuclear science predilection. For those who lack either, the core message is this:

Per’s aim is to develop really compact nuclear units with very high power densities, based on mostly well-understood technology that is deployable on the time-scale of a decade or less. The driving aim is to get these units commercialised in the near term, and to bring down costs, thereby paving the way for later widespread commercial deployment of full Generation IV designs like the LFTR and IFR, which not only achieve high burnup, but also completely close the fuel cycle.

Here’s Per’s summary:


Pebble Bed Advanced High Temperature Reactor

The Pebble Bed Advanced High Temperature Reactor (PB-AHTR) is a liquid salt cooled, high temperature reactor design developed at UC Berkeley in collaboration with Oak Ridge National Laboratory and other national labs.

PB-AHTR reactor system schematic.

The annular Pebble Bed Advanced High Temperature Reactor (PB-AHTR) design has a nominal thermal power output of 900 MWth (and electrical output of 410 MWe). The PB-AHTR differs from conventional helium-cooled HTRs because its liquid salt coolant enables operation with a core power density of 20 to 30 MWth/m3, compared to the 4.8 to 6.0 MWth/m3 typical of modular helium reactors (MHRs).1 The PB-AHTR delivers heat with a core outlet temperature of 704oC, achieving 46% thermal efficiency with a multi-reheat helium Brayton (gas-turbine) cycle. The low-pressure, chemically inert liquid-salt coolant, with its high heat capacity and capability for natural circulation heat transfer, provides: (1) robust safety (including fully passive decay-heat removal) and (2) improved economics with passive safety systems that allow higher power densities and longer-term scaling to large reactor sizes [>1000 MW(e)] for central station applications.

PB-AHTR primary, intermediate, and power conversion systems

PB-AHTR uses conventional TRISO high temperature fuel in the form of pebbles slightly smaller than golf balls. The baseline PB-AHTR design uses the well understood beryllium-based salt flibe(7Li2BeF4) as its primary coolant, and flinak (LiF-NaF-KF) as its intermediate coolant. Metallic structures and components like the reactor vessel are constructed using Alloy 800H, a ASME Section III code qualified material, with Hastelloy N cladding for high corrosion resistance. The coolant loop of the ORNL Molten Salt Reactor Experiment 2 operated with clean fluoride salt, like the PB-AHTR, for over 26,000 hours without any detectable corrosion to Hastelloy N samples that were studied after the reactor shut down 3. The major components in the reactor core are fabricated from graphite, which is chemically inert to fluoride salts.

PB-AHTR fuel pebble

The PB-AHTR combines together technologies derived from earlier reactor designs to create a new high-temperature reactor design with a unique combination of features:

  • Modular helium reactors (PBMR): TRISO pebble fuel, nuclear-grade graphite; high-temperature metallic and carbon composite structural materials; helium Brayton power conversion.
  • Sodium fast reactors (S-PRISM/EBR-II): Pool-configuration reactor vessel; reactor building seismic base isolation; direct reactor auxiliary cooling system (DRACS) for passive decay heat removal.
  • Light water reactors (AP-1000/ESBWR): Integral effects test scaling and best-estimate safety code validation methods; modern computer aided design, manufacturing, and modular construction technologies.
  • Molten salt reactors (MSRE/MSBR): Liquid salt pumps, heat exchangers, corrosion resistant alloys; liquid salt corrosion test and thermophysical property data base.

Like modern MHRs, the baseline PB-AHTR uses a conventional low-enriched uranium fuel cycle. But the PB-AHTR technology also supports advanced fuel cycle options:

  • Deep burn fuel cycle: the PB-AHTR can use deep burn TRISO fuels to destroy plutonium and other transuranics from commercial spent fuel
  • Once-through seed-blanket fuel cycle: the PB-AHTR can operate with a low-enriched uranium seed and thorium blanket fuel cycle that can reduce uranium consumption and waste generation while maintaining once-through operation.
  • Closed thorium fuel cycle: the PB-AHTR can operate with a closed thorium based fuel cycle with greatly reduced production of plutonium and other transuranics. Achievable conversion ratios are being studied now.
  • Liquid fluoride thorium reactors: The PB-AHTR provides technology that can be applied to future deployment of molten salt reactors using sustainable closed thorium fuel cycles.4
  • Fission/fusion hybrid reactors (LIFE): The PB-AHTR provides technology that can be applied for the future deployment of fission/fusion hybrid reactors that would operate sustainably without enrichment or reprocessing of their fission fuel.5


1. P. Bardet, E. Blandford, M. Fratoni, A. Niquille, E. Greenspan, and P.F. Peterson, “Design, Analysis and Development of the Modular PB-AHTR,” 2008 International Congress on Advances in Nuclear Power Plants (ICAPP ’08), Anaheim, CA, June 8-12, 2008.

2. “MSRE Systems and Components Performance” Oak Ridge National Laboratory, ORNL-TM- 3039, June 1973.

3. “The Development Status of Molten-Salt Breeder Reactors,” Oak Ridge National Laboratory, ORNL-4812, pp. 200-201, pp.207-211, August 1972.

4. Energy From Thorium

5. Laser Inertial Fission/Fusion Energy (LIFE)

Add to FacebookAdd to NewsvineAdd to DiggAdd to Del.icio.usAdd to StumbleuponAdd to RedditAdd to BlinklistAdd to TwitterAdd to TechnoratiAdd to Furl


By Barry Brook

Barry Brook is an ARC Laureate Fellow and Chair of Environmental Sustainability at the University of Tasmania. He researches global change, ecology and energy.

93 replies on “Pebble Bed Advanced High Temperature Reactor at UC Berkeley – low cost nuclear?”

Thank you, a very interesting post! I’m sure there are many such projects we don’t know anything of.

What do you do with used pebbles? If the lwr-waste is a problem, isn’t the pebbele-waste a PROBLEM ? How do you recycle it?

I have some doubts about the helium Brayton cycle. It has not been scaled up nor demonstrated in this size. There are no manufacturer. The technical diffiulties are not small, for example the needed tightness may be difficult to achieve. Of course, the main componets are the same as in normal gas turbines, so maybe I’m wrong. I hope so.

How about the helium? Is there enough helium in the world for a fleet of this devices? It’s expencive.

The Rankine-cycle is already up to ower 600 degrees C, and will soon bee up to 700 degrees. There are manufacturers all ower the world and the technology is well known. If you can achieve a bit higher efficiency with Brayton, is it really worth it?


It is great that you got a chance to visit with Per, who is about to emerge as an important figure for Generation IV systems because of his work on the Secretary of Energy’s “Blue Ribbon Commission.”


@Kaj Luukko – recycling TRISO particle fuels is not the problem that some have made it out to be. The coatings do not dissolve in acids like the metal coatings on light water reactor fuel, but they can be removed in grinders and separated from the actinide materials with simple density separation schemes.

This kind of recycling is practiced on a reasonably large scale in the facilities where TRISO fuels are manufactured. The coating processes are not perfect, and there is a reasonably predictable rate of particles that get rejected. If the coatings could not be readily removed, these particles would be expensive waste products. Instead, the rejected particles get ground up to remove the graphite and silicon carbide coatings so that the fuel material can be recoated.

I agree with some of you comments about helium turbo machinery. It was concerns like those that prompted me to shift my focus to nitrogen for Adams Engines. The concerns that some reviewers of that design have always expressed was a question about C-14 production when N2 is exposed to a neutron flux by running it through the core, but Per’s system seems to alleviate that issue completely.

I would suggest that an air or N2 Brayton cycle would lead to much lower development costs and probably lower capital costs than a helium turbo machine since already manufactured air breathing turbomachines could be adapted with little alteration.

One of the reasons that Rickover succeeded was that he built a heat source that could use conventional heat engines without a lot of modification. Building a new heat engine from scratch is a multi-billion dollar proposition.

Rod Adams
Founder, Adams Atomic Engines, Inc.


Rod, Closed Cycle CO2 Brayton cycle turbines offer attractive efficiency. High temperature Brayton cycle turbines offer a further possible efficiency, a bottom steam cycle turbine, that would use the waste heat from the Brayton cycle to heat water in a boiler. An even further efficiency is possible if the waste heat from the steam cycle can be used in a desalinization system.


I’m curious as to whether the second salt loop is regarded as a necessary feature of the design even in pure electrical generation application. I can see the utility in the high-temp application, I think, but it seems excessive on the turbine side, unless the complexity through the reheaters is such that it’s best to partition it off. Bearing in mind that the BWR reactor coolant drives the turbines directly, I wonder how much extra isolation equipment is sensible.

And like Rod and Charles, the use of helium in the turbines seemed odd to me.

Anyway, a great concept and a flexible path forward to advanced reactor design. I hope something like this gets built soon.


Also a benefit of this is that the Fluoride salts don’t have the same reactivity in air as Sodium based coolants do – that’s gotta be good, especially from a public perception/acceptability point of view.

Also, I think I’m correct in saying that high temperature electrolysis can produce H2 with efficiencies up to 60%+. Perhaps this H2 could be pumped underground, and then burned to load follow? If 75%ish CHP was used, that’s a total efficiency of about 45%. What are peoples thoughts?


@Tom – why add the complexity of H2? As long as they are not too large, nuclear plants can be designed for rapid load following.

As we used to say on the “boat” – power follows steam demand.


Joffan, I do not view the use of helium as odd. The issue is managing tritium that is produced from lithium fission in a Molten Salt cooled reactor. It is usually considered desirable to trap the tritium rather than allow it to escape the reactor in an uncontrolled fashion. Maybe that issue should be revisited, but a tritium can be trapped in a helium environment. There are some other complex issues that requite the second salt loop. i am not as familiar with the PB-AHTR, so i do not know if a second salt loop is required, but i suspect it might be. The PB-AHTR is very closely related to the MSR, and would share some common problems.


Thanks Rod. Seems like I have been misinformed about TRISO fuel.

I would suggest that an air or N2 Brayton cycle would lead to much lower development costs and probably lower capital costs than a helium turbo machine since already manufactured air breathing turbomachines could be adapted with little alteration.

An air Brayton cycle would be as easy as to replace the burning chamber with a heat exchanger. And no secondary heat exchangers are needed as a gas turbine breathing air could dump the excess heat directly to the atmosphere. Without any cooling towers. However, the efficiency would ge quite low.

Or it could be build as a CCPP, as Charles mentioned.


@Charles Barton; Rod Adams.

Damn, I thought I had an original idea there.

Would either of you guys mind pointing me in the direction of any scientific/technical papers on these reactors that I could use in my dissertation on NP? Thanks.

Also, if these reactors can be used in a load following mode, maybe there is a synergy between them and EPR/AP1000/ESBWR that will likelly be built first?


Does the use of a graphite based fuel form present any risk of combustion in this design?

Kaj & Rod, thanks for the Q&A on recycling – I also had heard it mentioned this was a difficult fuel form to deal with.


Also regarding TRISO pebble recycling, if you are getting burn-up of <50%, the issue becomes less relevant over the short- to medium-term at least. Per also mentioned supercritical CO2 as another potential heat transfer gas for the Brayton turbine. This idea is also being explored by ARC systems for a sodium-cooled fast reactor.

John, the pebbles are extremely heat tolerant. I can't think of any plausible mechanism whereby the graphite moderator pebbles might combust. Can you? It would be an interesting point to discuss if you have some plausible scenarios.


No, I have no plausible scenario in mind. Its an issue I thought might be raised based on the graphite fire at Chernobyl, so I thought I would get in first and ask what engineering controls ensure isolation of the hot pebbles from air.


Very interesting! No need for inflammables such as liquid sodium.

This project will improve our understanding of high temperature materials in contact with molten salts. The experience will probably be directly applicable to LFTRs.


Could fluoride salts be used instead of sodium salts in an IFR (or similar breeder reactor)?

IFRs still have that attractive feature, in that they can consume what is currently viewed as waste.


Tom K, you need a coolant that provides no moderation for a fast reactor. Further, the sodium bond in the IFR fuel pins works superbly with the metal fuel to allow expansion without burst cladding. Also, the whole sodium ‘danger’ issue is vastly overblown — a view totally reinforced by my recent visit to Idaho. All practical experience says Na is the #1 coolant for fast reactors, bar none (and they’ve all been tried). I’ll explain more in later IFR FaD posts.


John Morgan,
The pyrolytic graphite used to coat the “Pebbles” has amazing properties. I remember using this material in electrically activated clutches many years ago. No matter how much the clutches were overloaded there was no measurable wear.

However, pyrolytic graphite will burn (cf. Chernobyl). The amount of carbon in PBMR core will be orders of magnitude less than the graphite moderator in the Russian RBMK, greatly reducing the potential for radioactive “fall out” downwind.


Tom Keen, It is possible to build a very safe Fast Molten Salt Reactor using chloride salts. The French think they can build a fast reactor with fluoride salts, although I suspect that the chemistry of that fuel formula has not been investigated by chemists. All of the French MSR investigators appear to be physicists.

John Morgan, graphite only burns in the presence of oxygen. The core structure would provide barriers between graphite and O2. Although Molten Salt Reactor coolant temperatures are typically several as hot as the water coolant of light water reactors, fuel temperatures in the Chernobyl reactor were much higher than the core temperature of a MSR. The high fuel temperature coupled with the breakdown of the barrier between graphite and the O2 in air, caused by a powerful steam explosion which ruptured core air barriers and lead to igniting the core graphite of the Chernobyl reactor.


From my limited reading, I have developed concerns over sodium fires of which there appear to have been several in experimental fast reactors. I am fully prepared to accept that these have been of no consequence in terms of safety to the surrounding public. However, I remain concerned that even an small fire might have severe economic consequences and thus inhibit the enthusiasm of potential NPP investors.

Could anyone with more technical knowledge comment on this issue?


Douglas, John Sackett said this to the IFRG recently:

It may be noted that there have been many sodium leaks and resultant fires in the history of fast reactor operation. Such events are detectable and manageable and their probability has been greatly lowered as experience has been gained. The same can be said for sodium-steam leaks and resultant reactions. I remember traveling to Kazakhstan in the late 70’s to visit their fast reactor BN 350 (as part of a US team) to find that they had extensive failures in their steam generators due to poor fabrication. This experience as well as others has demonstrated that sodium-steam reactions in these systems are not catastrophic as once believed and further, their probability has been greatly reduced by what has been learned. No injuries have resulted from either event in the nearly 60 history of fast reactor operation.

At the BN-350, leaks were cleaned up and the reactor was restarted the same day. It’s successor, the BN-600, is considered the most reliable reactor in the Russian fleet.

I’ll talk more about this in an future IFR FaD post.


I came across a “special feature” in The Korea Herald covering Korea’s Nuclear Technology. The fact that it comes in 26 parts is notable in itself as an indicator of the perceived national importance of nuclear energy.

Also interesting are the articles on R&D including Pyroprocessing with an “engineering scale” facility by 2016, programs for a very high temperature reactor for hydrogen production, a sodium cooled fast reactor and of course fusion. And of note was the capacity factor of existing PWR’s which is around the best in the world.

Anyway, I thought it was worth a read:


A bit OT but here I go again.How much power does one (more or less standard) size fuel rod produce?
My question, would it be technically feasable to built a tiny labsize practical NPP ?
Say, enough to power a homestead or small enterprise?
I mean , aviation powerplants (turbines) or even Stirling engines come in lots of capacities. Or am I totally nuts?


Thanks guys for the info.

According to the text book on NP I’m studying for my dissertation, the pebbles have such a small mass/surface area that they naturally dissipate any latent decay heat naturally and do not melt under any circumstance.


@Uncle Pete – NASA and the Russian space agency have done a lot of work on the issue of building the smallest possible fission power systems.

In each case, the physical size and power output (tens of kilowatts) would be appropriate for a small enterprise or a large homestead – if your idea of home is a bit more expansive than most common folks.

In such applications, the machines designed for space operations would need more shielding, but that is simple and cheap if you do not have to lift it out of Earth orbit. Space and weight are at a far lower premium in terms of cost and system limitation for ground based systems.


@ Rod Adams
If these reactors would be available off the shelf , then I am sure lots of business’s and farmers would buy them.
heck, i would be first in line :)


Douglas Wise,

I agree with Barry Brook that the safety issues related to sodium fires are overblown. However, a realistic perception based on science does not always translate into parallel conclusions in the political arena owing to widespread irrational fear of nuclear processes.

The canceling of the IFR (USA, 1994) and the Superphenix (France, 1997) are examples of political perceptions being at odds with the science. No matter what other reasons were given, many believe that fears of sodium fires killed these projects. Is the political climate more favorable today in either the USA or France? I can’t see any sign of that so the leadership in this technology will go elsewhere.

The increasing interest in various NPP designs based on molten salts offers alternatives if the political resistance to sodium cooled reactors cannot be overcome.


Thanks to those who replied to my query. Thanks also to Quokka for his link to the articles relating to Korean nuclear power.

In article 23, relating to Korea’s choice of fast reactor design, the author suggests that Korea’s decision to go with sodium cooling may have been precipitate and that it be revisited. He gives his reasons for preferring lead eutectic cooling and it would be very interesting to have the matter discussed on BNC by those with more knowledge than I.

Barry’s post was essentially making the point that the nuclear industry in the States would stagnate until such time as nuclear electricity could compete on cost with that from fossil fuels. This could be achieved through subsidy, a carbon tax or cheaper, but no less safe, designs. The first two options require that a government is totally committed (rather than playing lip service to) avoiding the threats of either climate change or peak oil. In theory, the last option doesn’t depend on supportive government, only the absence of a hostile one. However, the principal costs of a nuclear build, as I understand it, are not overnight costs, but finance costs. It seems, therefore, that the only way to reduce finance costs still requires
full government backing in order to expedite build time and reduce risk premium. This makes me wonder whether the approach favoured by Prof Peterson will, on its own, be sufficient to make nuclear competitive. However, it is nevertheless extremely interesting that he has chosen to favour a thermal, molten salt route in his search for economy. I imagine this will give more comfort to LFTR supporters rather than those who favour fast reactors, however cooled.

The free market has thrown up a vast array of different potential nuclear designs. It would seem likely that only a few will be built and it also seems that nobody is sure which will prove best until such time as, at least, demonstration units are built. This is a somewhat depressing state of affairs for a private investor with an interest and belief in the benefits of nuclear power. In order to avoid backing the wrong horse, he might be better to invest in wind and get a return guaranteed by government subsidy, even while having scant faith in the technology.

All these musings leave me with the feeling that nuclear power development may best be served by nationalisation and international cooperation, much as I normally distrust a nationalised approach. I would therefore be delighted if others could find fault with my reasoning. At the very least, perhaps there could be internatioal agreement to build a range of demonstration units as soon as possible on a cost sharing basis.


Barry, do you have any figures for $/kwh for all of these reactors, including the russian BNs?

Also, I just listened to your podcast with Ian Lowe, that you posted on Twitter. The interviewer was quite unfair to you I thought, giving the majority of time to Lowe and pretty much ignoring you.

Lowe I thought sounded almost convinced by your arguments and about ready to switch sides! His main grippes are still with the proliferation issue. Not only is this a non-issue to australia, but if he studied Nuclear Physics, he should know the isotopic quality of the Pu makes this issue a non starter.


when we say that IFRs are 160 or so times more efficient than lightwater reactors (LFTRs are similarly efficient, yes?) due to their full use of the uranium, how does this efficiency relate to both high burnup and closed fuel cycles?

I realize that burnup rate and open/closed fuel cycles are different things. You can have either one without the other. But what contribution do these two items separately and together make to greater IFR efficiency or LFTR efficiency for that matter?

And if my question is incorrectly put due to some confusion I suffer from, I’d like help sorting it out.

and Tom: what textbook are you studying? I’d like to read it too.


Charles, the point of the second heat-transfer salt loop in the MSR is reasonably clear, but since the flibe in Per’s design doesn’t contain fissionables I’d think the situation is significantly different. Admittedly it does flow through a high neutron flux zone, but I wouldn’t expect it to be significantly radioactive as the component elements are not strong neutron absorbers.

The helium is only odd because it adds a difficult and rare technology to the generaiton end when perfectly acceptable alternatives exist. In the design as sketched the possible) tritium and helium are not even in proximity, so I don’t see the relevance.


Rod Adams and Uncle Pete,
The Russians have designed some excellent power plants for use on space vehicles using plutonium as the heat source. I think you may have trouble getting one for use on a farm.

In 2002, Charlie Bowman built a prototype ADR on his farm in Virginia. The reactor shell was made of black pine (a neutron reflector) and it was about 8 feet high. This reactor was tested at the Triangle Universities Nuclear Laboratory (TUNL) and later at Los Alamos (LANL).


@Douglas Wise

In article 23, relating to Korea’s choice of fast reactor design, the author suggests that Korea’s decision to go with sodium cooling may have been precipitate and that it be revisited. He gives his reasons for preferring lead eutectic cooling and it would be very interesting to have the matter discussed on BNC by those with more knowledge than I.

I have read somewhere that there may be corrosion issues with lead, whereas sodium plays very nicely with steel (stainless steel?). Maybe somebody who knows what they are talking about might confirm this?


greg meyerson,
Reactors that burn almost all the fuel produce very little waste, thereby reducing the cost of dealing with waste. Many of these fuel efficient designs can use waste from early uranium reactors as fuel. Clearly a win-win situation.

Doug Wise,
You are so right about there being a confusing array of NPP concepts out there. Thanks to Barry, DV8 (when available) and many others we are getting all kinds of information on the pros and cons. Right now I have no idea which NPP concept will dominate a hundred years from now. Maybe several designs will survive.


@greg meyerson

I’m reading ‘The Future of Nuclear Power’ by John Lillington – very comprehensive and scientific. You can get it on Amazon but its unbelievably expensive:

I got mine from the local library.

I also read this:

Not quite as deep, and I felt it had a major ‘pro-nuke’ spin, but still very good and worth a read.



I think tom posted a small article on lead bismuth reactors in India. I think they talk a little about the solving the corrosion issue.

but you are right. sodium is non reactive with stainless steel. Blees talks about this. In addition, the sodium in the IFR is bathed in an argon blanket. argon is nonreactive with sodium and heavier than air, thus a very effective blanket, not likely to blow away so to speak.


galloping camel:

just to be clear: what precisely is the relation between “burning almost all the fuel” and high burnup on the one hand and closing the fuel cycle on the other hand?

They would both contribute to “burning almost all the fuel,” but one without the other would not “burn almost all of the fuel,” right?

high burnup is measured in terms of energy produced per mass of material, right? so hb would be a far more efficient way of doing once through reactors, but it’s still preferable to close the cycle if you are going to burn almost all the fuel.



greg meyerson,

You are pushing my expertise to the limit (where is DV8 when you need him?). However I will have a shot at it in the hope that more knowledgeable people will set me straight if I drift into BS. I am a physicist specializing in electro-optics rather than nuclear power!

The 100+ NPPs in the USA are operating on a “once through” fuel cycle owing to a ban on reprocessing that dates from the Carter administration. Without some form of reprocessing, less than 1% of the Uranium is “burnt”. The French have wet reprocessing that allows the fuel to be recycled for greater fuel efficiency and the extraction of plutonium for weapons use.

The French reprocessing at La Hague is expensive so various dry processes are being developed. When the reprocessing is carried out on site the goal is a “closed fuel cycle”, so as to “burn” a high percentage of the original fuel loaded.

Taking the LFTR as an example of a closed fuel cycle reactor, the fuel (metallic Thorium dissolved in molten salts) is not fissile (k<1) so fissile material such as U235 must be added to get the reaction started. Once operating, the Thorium is fertile so various new isotopes are produced including more U233 than is necessary to maintain the reaction. To reduce the Uranium concentration, Fluorine is blown through the fuel blanket allowing UF6 to be collected as a gas. There are complications caused by reaction products such as Pa233 and Xe135 but these are not show stoppers.

The Uranium extracted can be recycled with new Thorium fuel. The U233 produced is useless for weapons production owing to the unavoidable presence of U232 (a strong gamma emitter). LFTRs deal with Xe135 (a strong neutron absorber) more easily than solid fuel reactors do.

Not all advanced reactors are capable of burning all the original fuel. For example, I understand that the IFR was expected to burn 50-60%. Hopefully Barry will enlighten us on the fuel cycle for other advanced reactors such as the Russian BN series. Wikipedia is pretty vague on the issue

One might think that burning only 0.6% of Uranium is going to make US reactors more expensive to operate than the more efficient French ones. Did I mention that wet reprocessing is expensive? Even though the Carter ban on reprocessing has been removed, there will be no rush to build such plants here unless the price of Uranium skyrockets.


greg meyerson,

Rubia’s ADRs (Accelerator Driven Reactors) may have a role in fuel reprocessing especially when higher Actinides (destined for geologic storage) are present. In an earlier post I mentioned Charlie Bowman’s ADR or Sub-Critical Nuclear Reactor. While this work has not attracted significant funding it is not quite dead.

The idea is to dissolve high level radioactive waste in molten salts (as in MSRs) and then irradiate the the reaction zone with a high energy neutron beam. The reaction zone is surrounded by a spallation zone that multiplies the available neutrons by a factor of 10 or more. The spallation zone is surrounded by a graphite moderator to thermalize the neutrons and then a neutron reflector to prevent neutrons from leaving the reactor. Charlie Bowman used black pine as the neutron reflector owing to its high content of Hydrogen.

This type of reactor is capable of converting wastes that were destined for geologic storage into stable isotopes including some quite valuable ones. The challenge is to find a cheap source of neutrons! As a by-product you can generate substantial amounts of electric power.


gallopingcamel, Rubbia’s idea was to use a proton beam via a synchrotron, not a neutron beam (how would you make one of these?? A neutron amplifier I guess but these things are only theoretical). The protons are fired at the spallation target as you note, but it is this interaction which would then result in required neutrons.

But I agree with Luke, I really don’t see the point, and these things are pretty big. There’s a whole chapter about it in ‘Megatonnes and Megawatts’, including a neat picture of an elephant standing next to the core, looking rather dwarfed. These things are not tiny :)


Personally, I’m a bit skeptical about accelerator-driven reactors.

You need a *big* particle accelerator (something like 10-12 mA of proton beam at about 1 GeV, which is quite a big accelerator (10-12 MW of beam power)), so it adds to the cost of the reactor quite substantially.

But what does it give you as an advantage, that other competing advanced nuclear power, such as IFR and LFTR, don’t give you? There must be big advantages to justify that large added complexity and cost; advantages we can’t get from other reactor designs.

One possible advantage is that you might be able to start on a fertile fuel load of Th-232 or U-238 without any fissile startup charge.

However, I don’t think that’s a big deal, there is a lot of surplus fissile material in the world that can be consumed – and should be consumed – for startup of advanced power reactors.

Another advantage is that, yes, it will burn through all the transuranic actinides which are either formed in the reactor or introduced into it in the fuel, from existing LWR fuels. But IFR or LFTR will do the same.

Another claimed advantage is that you can shut off the accelerator and instantly stop criticality in the system.

But so what? The ability to effect a rapid emergency SCRAM has been a feature of almost every nuclear reactor ever built, starting from CP-1. So how does this offer you something other reactors don’t?

Peterson’s technology, mentioned by Barry in the original post, looks interesting.

TRISO pebbles immersed in molten fluoride salt?
This seems like an interesting alternative to a MSR with a graphite moderator core structure inside the salt, like MSRE.

It is possible to build an LFTR without graphite in the core, just salt, isn’t it? This is advantageous because you don’t need fuel irradiation testing and validation, which you need for Peterson’s TRISO approach.


Just reread this post again. Per’s concept looks really interesting – must have been great to get a tour of his lab. There’s a bunch of questions occurred to me, after having this idea in my head for a few days.

The main issue with Generation IV reactors such as the IFR or LFTR is the general expectation that they will be more expensive than ALWRs

Why? I’d have thought they’d be cheaper. The IFR does not require a large pressure vessel, so that major cost component is out. There’s no water in the vicinity of the core and no component under significant pressure, so the containment structure requirements should be much reduced relative to a LWR. The overall structure should be smaller, so construction costs come down. If you factor in the fuel cycle I’ll bet pyroprocessing is cheaper than wet reprocessing.

Similar arguments are made by the LFTR guys, and look reasonable to me. So excluding expensive initial units, what drives the expectation that gen iv is more expensive?

From your description of Per’s lab, it sounds like he’s working with reactor models for transport studies. Has anyone built an active core based on this concept? Maybe his buddies at ORNL? Or is core performance still living in simulation land?

The driving aim is to get these units commercialised in the near term, and to bring down costs, thereby paving the way for later widespread commercial deployment of full Generation IV designs like the LFTR and IFR, which not only achieve high burnup, but also completely close the fuel cycle.

Its not clear to me in what respect the AHTR design serves to pave the way for IFRs or LFTRS – perhaps experience in material properties and engineering of molten salts in the latter case? Or maybe the intermediate and power conversion systems are where the engineering development can cross over?

Finally, what are the near term options for commercialization of the AHTR?

Sorry for being such a question box.


John, the ‘general expectation’ words are definitely straight from the horse’s (Per’s) mouth, and I’m not necessarily endorsing the truth of that statement. But then again, Per is someone you tend to want to listen to (as is the most impressive figure of Chuck Till, but that is another story for another post :).

Per’s claim is that his PB-AHTR is more reliant on engineered systems that are already commercialised (so has a head start, even if it is only for the short term), and that because the power density is so high, the unit can be more compact than a sodium-cooled system (sodium has a lower heat capacity than molten salt or water, though of course all are far greater than gas, and lead has those troublesome corrosion and freezing issues [hence the use of the Pb-Bi eutectic for the latter problem]). The PB-AHTR will involve no steam cycle or high pressures either, so, like the IFR, would not technically need a large containment structure. The materials issues are less worrisome than the LFTR as there are no fission products in the salt. This (and the regulatory question I describe below) also relates to why he things it’s reasonably ‘near term’. 10-15 years at steady pace, 5-10 with an accelerated programme.

He sees the PB-AHTR as a pathway to LFTR via technical proving up of various subsystems that would be common to both (as you surmise), and a way of easing the US into an IFR in a broader sense, by getting the regulatory bodies (well, the NRC in the US) more comfortable with the licensing of reactor designs that are not your bog-standard water-cooled reactors (actually, for the US, designs that are not LWRs), are smaller, higher temp, and modular etc.

I’m not sure what you mean about transport studies. Yes, his model reactor is not an active core, but I think there have been experiments with earlier designs at ORNL. He’s been refining the concept using his physical, non-nuclear models at UC Berkeley (no radiation or fissile material regulatory ‘hassles’ etc. to worry about at that stage of R&D).

On the cost of wet vs dry recycling, I have some really interesting figures, but again, that’s for another day and another post.


Barry Brook,
Thanks for correcting my schoolboy “howler”. The proton beam used to test Charlie Bowman’s SCNR “Sub-Critical Nuclear reactor” came from a van der Graaf DC accelerator. The beam power was less than 1 Watt.

Charlie is working with L.N.Chang (Virginia Tech) on a scale up that proposes a 2 MW pulsed proton beam to produce 400 MWe while consuming nuclear wastes that would otherwise require geologic storage. While the 2 MW proton beam source proposed does not exist, the SNS at Oak Ridge is in the right ball park at 1 MW.

I am not seriously suggesting that ADRs will ever dominate the landscape when it comes to generating electric power. More likely they will be relegated to specialized waste reprocessing applications or isotope production.

I thought your readers would get a chuckle when they heard of a nuclear reactor built out of wood! I first heard about it while playing golf with a couple of my nuclear physics colleagues. I thought they were pulling my leg until they showed me the reactor under test in their laboratory.

Luke Weston,
While I agree with your comments on the superiority of other technologies when it comes to generating electric power, I still believe that there will be uses for ADRs with mega-Watt ratings.


Barry Brook,
Synchrotrons are not needed in ADR drivers.

While the Spallation Neutron Source at ORNL has a synchrotron section, it is used as a pulse sharpener, not needed for driving an ADR.


Thanks for that explanation Barry. By transport studies I mean transport of the fuel pebbles through the core, fluid through the core, heat out of the system, etc., the sort of thing you describe in the post and that can usefully be done in a non-active physical model.


Tom W:

thanks. yeah, those books look pretty expensive but David Bodansky’s (longer than yours) looks semi affordable, used.

I might get it.


John, then yes, it’s definitely used for transport studies. There are even different coloured nylon pebbles so that the movement can be carefully tracked, and they form interesting layers due to differential movement rates. Ahhh, but you’ve got to see it yourself!


Some more interesting comments on the PB-AHTR from George Stanford, Jan Van Earp and Dan Meneley, via the IFRG:

– The AHTR is a very promising concept, and might well find a place in the energy picture if the technology matures.

– If the AHTR’s enrichment is only 4-5% fissile, like LWRs, then 50% burnup would be some ten times what is achieved by the LWR. But if the AHTR fuel is initially 100% uranium, 90% of the ore’s energy remains in the DU, so the resource utilization would be around 5% — considerably better than 0.5 %, but still very far from the 99% utilization achievable in a truly closed fuel cycle like the IFR’s. However, if the enrichment needed is 10%, then about 96% of the ore’s energy would be left in the DU, and the resource utilization of a once-through AHTR cycle would be closer to 2%. Clearly there are many possible loadings involving spent LWR fuel or thorium, whose resource utilization would be different. I don’t know enough about the AHTR to say whether there are any proposals for fuel cycles that would approach the IFR’s 99%, but it doesn’t seem likely — unless, of course, it turns out to be feasible to consume the spent AHTR fuel in IFRs.

– Regarding the “very high burnup” (which is achievable because the pellets are so stable under irradiation that they can cook for a long time without losing their integrity, as long as there’s enough fresh fuel being added to maintain criticality). A single-pass burnup of 50% is indeed quite “deep” in comparison with the 4-5% of an LWR. But what counts is the overall use of the resource. The single-pass burnup in an IFR is relevant to the IFR’s economics and the completeness of actinide consumption, but fundamentally it’s just an engineering detail: the lower the single-pass burnup, the more frequently the fuel is recycled. I think a single-pass burnup of at least 20% is anticipated, but it is not meaningful to compare that with 50% burnup in a once-through cycle. It’s the total burnup, if you will, that counts — i.e., 99% vs 5%.

– I would be more optimistic about a major role for the AHTR if it had been demonstrated that the remaining 50% of the actinides could be economically recycled — either in AHTRs or in IFRs. As it is, some of the pebble-bed people tend to cite the difficulty of reprocessing as a proliferation advantage (although no proliferator would use such crap for a weapon anyway).

– Rod Adams points out that the pebble manufacturers are already recycling some defective fuel pellets. But those are unirradiated, and pellets with high-burnup fuel would be a different ball game. It’s my impression that the actinides extracted from such high-burnup fuel would be so alpha-active that fabricating new fuel with them would be very difficult — not from a personnel-exposure viewpoint, but because of severe alpha-contamination of the processing facility.

– Thus it looks as though the AHTR (and PBMR) are inherently once-through systems, and the million-year concern (spurious, in my opinion) that some people have regarding Yucca Mountain would not be appreciably allayed — with uranium fuel, that is.

– With thorium as the main fuel, the production of higher actinides is considerably reduced, but I don’t know to what extent.

– Per Peterson is a smart guy and a careful thinker, and I’m not going to try to read his mind. He’s keeping his ideas re the IFR and what should be done very close to his vest — and rightly so, since he’s on the Blue Ribbon Commission. But he knows very well that his AHTR idea is not as ready for commercialization as the IFR, and I’m sure he will take that into account in his deliberations.

– Although not prominent in current thinking, the ability to breed more fissile than is consumed can be a vital consideration if we get serious about replacing fossil fuel with nuclear. I don’t hear thorium enthusiasts mentioning a breeding ratio that is much greater than unity. That indicates that the rate of deployment is potentially constrained by the availability of startup fissile — which will come first from spent LWR fuel. When that runs out, fissile for new reactors will have to come either from continuing and expanding mining and enriching of uranium, or, alternatively, from production of plutonium in U-238 blankets around (a) accelerator-driven spallation targets, or (b) as-yet-undeveloped fusion reactors.

– While the initial rate of deployment of thorium reactors could be greater than that of IFRs, the latters’ shorter doubling time would permit them to grow as long as necessary, without requiring outside sources of fissile material. As of now, putting fissile material into a break-even system is like depositing money under a mattress — and, of course, putting it into a “burner” would be like putting money into a shredder.

– Seems to me that the IFR should be given a chance to establish whether it can be economically competitive, with the realization that it’s not necessarily the last word for baseload power generation.


Among the various aspects that should be considered in connection with High-Temperature Gas-cooled Reactors (HTGRs) are the following:

(1) HTGRs do not seem to offer a long-term solution to the world’s energy supply requirements and therefore do not meet the sustainability criterion.

(2) HTGRs are graphite moderated, requiring relatively high fissile content for currently proposed commercial systems (graphite is not the best moderator for thermal reactor systems).

(3) Most proposed HTGR systems have an initial fissile content of up to 20% and an average fuel discharge fissile content of about 6% (higher than 20% initial fissile content is not recommended, among others on the ground of proliferation concerns).

(4) HTGR fuel is difficult to reprocess because of its being intimately mixed with graphite. Reprocessing of HTGR fuel has not be done commercially up to now and it is questionable that it can be done economically (it would require burning off the graphite with its associated environmental concerns).

(5) Most studies advocate a once-through fuel cycle, requiring long-term storage of the used fuel elements.

(6) Long-term storage of HTGR fuel with 6% average fissile content would constitute a serious concern because of the possibility of criticality accidents, among others due to in-leakage of water.

(7) The primary perceived advantage of HTGRs is the high temperature which could be used for applications requiring high-temperature process heat .

(8) One of the main proposed applications of HTGRs is hydrogen production by chemically “cracking” of water. However, this process requires operating temperatures of about 900C which is outside the range of presently available structural materials for high-pressure systems.

(9) Hydrogen production can just as well be carried out by means of electrolysis. This can be done at any temperature with higher efficiency at higher temperatures. Temperatures in the range of 500C are achievable by means of LMFBRs. Also, temperatures could be further “boosted” by combustion of a small stream (a few percent) of the generated hydrogen, thus increasing the efficiency of the process */.

(10) HTGRs have been around since the 1950s (the Dragon Project in the UK, the Kartoffel Reactor and the HTR in Germany, Peach Bottom and Fort Saint Frain in the U.S). One may well ask why to dedicate valuable and limited resources to the development of a system that does not appear to offer sustainability and that has been around for some fifty years while not achieving successfully the commercial stage ?

(11) The copper – chloride process is now the leader in the race for chemical separation of water. The operating temperature of that process is about 550 C

(12) If you have very cheap nuclear fuel (e.g. CANDU) you can boost the product temperature using electricity, as well.

*/ A study of this was performed by AREVA for PWRs, as reported at the International Meeting on Nuclear Production of Hydrogen, organized and conducted in 2008 by NEA, IAEA and ANL at Oakbrook, IL, USA.





The copper chloride water splitting process looks very achievable. How does it compare costwise to hydrogen generation from methane? The latter is of course dependent on the cost of methane and the free dumping of CO2, but just to get a feel for the difference…


Barry Brook,
Thanks for that link. It is not the “poisoning” issue that bothers me.

LFTRs can easily vent the Xe135 but I was wondering how the gas gets out of TRISO pebbles. The Xe135 is converted to Xe136 by neutron absorption, so you will have a gas pressure build up inside the fuel pebbles unless there is some way for it to get out. Does one make tiny holes in the pyrolitic graphite or is the material porous?

No big deal, just curious!


Some useful follow-up comments from Per Peterson on the IFRG:

A quick note on the fuel design for the PB-AHTR.

The PB-AHTR is definitely a converter reactor, like ALWRs. The current neutronic and depletion calculations show a reduction of 40% in uranium consumption, and 20% in enrichment SWU, compared to ALWRs, when operated with a Radkowski-type seed/blanket core. The fluoride salt coolant, which occupies 40% of the core volume, provides over half of the moderation and thus the heavy metal loading in the fuel is more than double that in typical HTGRs, with a corresponding reduction in spent fuel volume.

In this design, the LEU seed pebbles are enriched to 19.9% and reach a discharge burn up of 21% in approximately 210 days, and the thorium blanket pebbles reach a nominal discharge burn-up of 30% in approximately 30 years with about 2% residual U-233 (probably not economic to recover). If non-fertile LWR-TRU seed pebbles are substituted for LEU, they reach discharge burn up levels of 60% in under one year.

Very rapid fuel qualification is possible because seed pebbles reach full discharge burn up in under 1 year (HTGR fuel takes 3 to 6 years to reach full discharge burn up because the core power density is 6x smaller). This means that new fuels can be tested and qualified in a very short period of time compared to conventional fast and thermal reactor fuels. Both ATR and HFIR have test positions with the necessary thermal neutron fluxes to perform testing, and B&W now has the full capacity to fabricate appropriate TRISO fuel at commercial batch scale. So the U.S. has the full domestic capacity to produce, irradiate, and examine these fuels.

Electricity generation costs for AHTRs will be less sensitive to future uranium prices than ALWRs, but solid-fuel thermal spectrum reactors are extremely difficult to design to achieve a sustainable conversion ratio of one or greater so that uranium prices no longer matter at all. To do this, either fast spectrum (IFR), or fluid fuel with continuous xenon removal (LFTR) is required. Structural materials for fluid fuels require additional development including long-term corrosion testing, but we know from the 26,000 hours of operation of the MSRE intermediate loop that clean fluoride salts cause no detectable corrosion, just like we demonstrated with sodium in EBR-II.

Since there is a rapid path to qualify fuels and materials for AHTRs, compared to other reactor technologies, the major question is whether the compact primary systems (smaller primary volume than PWRs), high thermal efficiency (~45%), and compact reactor and turbine building structures compared to ALWRs (due to the lack of stored energy sources to pressurize the containment and the use of gas Brayton power conversion) can lead AHTRs to have significantly lower capital cost than for ALWRs.

AHTRs are clearly not a mature technology, in contrast to LWRs, HWRs, SFRs, HTGRs, and lead fast reactors where actual test reactors as well as commercial/military reactors have been operated. But if lower capital costs are possible, this would worthwhile to know because capital cost dominates nuclear electricity generation cost and thus (along with availability) is the most important variable that utilities consider in selecting reactor technologies. AHTRs are still in viability-phase R&D, so understanding capital cost is a major viability R&D issue to address.

…and from Neil Brown:

As I recall not only could fuel qualification been done more quickly than for HTGRs but it can be constructed using current code qualified materials. On another point. The Japanese have been operating their HTGR test rector for extended periods at 950C and have plans for running a developmental H2 production loop off of it in the future.


George Stanford then asked: “Can I conclude that you see AHTRs as potential competitors for ALWRs (Gen-III technology), rather than for SFRs (Gen-IV technology)?” and Per answered: “You are correct. AHTRs do not address the fuel cycle issues that SFRs do, although they do provide some uranium resource extension compared to ALWRs.”

I hope people are finding this ongoing dialogue interesting. I think this page will be a useful reference source for PB discussions.


There are even different coloured nylon pebbles so that the movement can be carefully tracked, and they form interesting layers due to differential movement rates.

Just like a magic gumball machine.

A 400 MW gumball machine ..



This post and your subsequent comments in the thread seem to me to have acknowledged that there is some validity in at least two of the criticisms levelled against nuclear technology by its critics, namely cost of all types of fission reactors and sustainability with respect to once through reactors.

I don’t recall that these issues have generally been taken seriously by most of the nuclear proponents who contribute to BNC. Finrod and several others, for example, have done their best to rubbish the peak uranium/sustainability issue, even with respect to once through reactors. I found this reassuring, but wondered why others, such as David MacKay for whom I have respect, apparently disagreed. Now you have seemingly displayed your concern over sustainability consequent to your visit to the States and discussions with nuclear experts. This pushes my layman’s thinking further towards acknowledging the relatively urgent need for breeder reactors.

Next comes cost. Peter Lang is an exemplar of the position that nuclear power must be cheaper than that of coal if it is to become deployable. It would, indeed, be highly desirable were this to be possible, but I have come to the conclusion that, under current free market conditions, it is not possible in the short term. Your new post seems to lend credence to this view.

Neither of these criticisms should give solace to the anti-nuclear brigade, given that it seems to be clear from basic scientific principles that renewables will prove a far more costly solution and neither would they sustain even current population level, let alone the projected 2050 level that is already in the pipeline, barring massive die off.

If my reasoning is correct (I’m sure others will object to it), I am forced to the conclusion that our best future hope is for governments to take charge, as they would in a war-time situation, and to take nuclear power development forward as quickly as possible. This requires leadership, leadership that acknowledges that we are in an emergency situation and that strong action aimed at weaning ourselves off fossils fuels as soon as possible is all that will secure any hope of other than a bleak future.


gallopingcamel, I didn’t know, so I asked. Mike Lineberry says that: “The “story” on TRISO particles is that they are their own first level of “containment”. Nothing is supposed to get out. I’m not sure anybody outside the gas reactor community completely accepts that, but that’s the claim or the intent. I’ve never been able to see that this is really any different than with LWR fuel rods. In that case the leakage rates have evolved to near-zero… I think that it is true today that one out of 50,000 to 100,000 fuel pins become “leakers” and leak fission product gases to the primary coolant system.”

Huw, no, I’ve not talked about then on BNC. The idea is that they’re good (potential) sources of neturons for breeding more fissile from fertile. There was an interesting article in Science magazine recently on them. Here is an interesting sceptical assessment from some MIT guys:


There’s one other thing that has bothered me about pebble reactor designs, and that is that there are a number of core parameters that are not deterministic – not under direct engineering control, or controlled by design. These are:

1. Residence time of pebbles in the core. If I understand correctly the residence time of an individual pebble is determined by the rate at which it migrates through the bed. I expect a wide distribution of residence times due to a somewhat random path through the bed. Your comment about stratification confirms this is happening. You can’t pick and shuffle the pebbles around like you can fuel rods in a conventional core. So there will be a distribution of individual pebble burnups.

2. Is there a problem if the burnup of a TRISO element is too large? If there’s a range of burnups, the fuel efficiency will be less than optimal – you’d need to pull down the average burnup to avoid the high burnup tail.

3. Core density. The density depends on the packing of spheres, with packing density between 52-74%, with random close packing somewhere in the middle of that range, but variable. Does having a core density that might be +/- 5% create a problem?

4. Spatial profile of the core density. Edge effects force a different pebble packing density at the bed walls compared to the interior bulk of the bed.

So we have an uncontrolled transport process through a core of uncontrolled density with uncontrolled residence times of pebbles with uncontrolled burnup rates in zones of uncontrolled reactivity. Does this cause a problem? Two possible problems spring to mind, which are that very high burnups might cause fuel integrity problems like the camel mentioned. The other is poor fuel utilization, to avoid the tail of the burnup distribution.

I’m not suggesting these really are problems, I assume the system must be robust to these issues to have sustained interest. But I’m curious about how they play out.


John M, I think the two specific problems you mention at the end there are somewhat addressed by the multiple trips through the core that each pebble makes. If each pebble makes about 6 trips through the core, say, then a high burn-up pebble might make only 5 trips before it is rejected and a low-burn pebble might make a seventh trip – both ending within a close approximation of the optimum value. The higher the number of trips, the closer each pebble will be to that value (but of course the handling overhead for the reactor would go up in that case too).


Thanks a lot Barry, that article is very interesting. I was thinking hybrid systems might be a good ‘kick start’ to a global IFR roll-out.


As I understand pebble-bed reactors, and I’m guessing this one is the same, the pebbles move slowly through the core and are removed, then suitable pebbles are cycled back into the reactor. Presumably in some suitable combination of old and new pebbles, and every pebble makes several trips through the core, in due course.


Discussion above seems to suggest that the PB AHTR , at best, will be an interim solution which may produce power a bit more cheaply than current NPPs . Its pricipal value would appear to be as a proving ground for emergent closed fuel cycle technologies. Is this the best and most economic way of achieving this objective?

Proponents suggest that the LFTR gives the promise of producing electricity at least cost, provided that corrosion issues can be addressed. Its slower breeding rate is compensated for by its lower start charge requirements and there is time to address long term sustainability issues in a few decades. Because of its economic potential, therefore, it would seem at least as or more urgent to get to grips with LFTR than IFR research. I understand, however, that the IFR is closer to its research endpoint prior to commercial demonstration. I also understand that the majority of private NPP producers don’t make as much money from plant construction as they do from fuel assembly, another reason why it may be better to leave them to get on with their Generation 3 activities and leave governments (in collaboration) to develop and build Generation IV plants.

I understand that there was a plan for the Russians to build an IFR with the States constructuing a pyroprocessing plant. Is this plan still extant and, if so, how far towards initiation?

If nuclear power in the West is perceived as being 25-50% more expensive than power from coal, then CCS coal may prove viable. However, it would seem that coal mining and associated fuel transport are often not fully factored in when considering it a clean energy source.

Peter Lang argues that nuclear energy must be cheaper than coal energy if it is to be worth pursuing, but seems to have a very hazy understanding as to how this is to be achieved, other than by lowering safety standards. I don’t think the Indians, Koreans and Chinese would accept that their plants were less safe. However, they may be less burdened by the conflict between free market competitors and regulating bureaucrats and be less constrained by NIMBYs and lawyers.

I apologise if this is a mish mash of reflections from a layman when it comes to nuclear technology. It is a sign of frustration at the apparent lack of concrete progress in the West. Unless something happens soon, we’ll be left so far behind that we’ll be lucky if our next generation survives to make cheap goods for the BRIC nations – the boot will be on the other foot .

In the financial section of the UK Daily Telegraph yesterday there was a plea for the States to undertake a Manhattan 2 approach to develop a thorium based energy future, claiming that it was probably the only way of protecting its position in the world economy. Whether LFTR or IFR would be the best choice, I don’t know. It should come down to costs. However, the Manhattan approach seems to be the key factor that is required in expediting either.


John, whilst welcome in a general sense, the Telegraph piece is also misleading — to say that thorium can provide 200 times the energy of the same amount of uranium, without any further explanation. He’s talking about two different technologies, and of course if you ran the uranium through a fast reactor, you’d get the same energy yield (the “200 times” or thereabouts). Lazy reporting, or else the reporter has only been told half the story. I suspect the latter, actually.


Douglas, I am not sure what MSR corrosion problem you are referring to. I am aware of only one corrosion issue that has not been definitively solved, and the hold up has more to do with funding for tests than a lack of insight into the problem. There is enough RGP in nuclear waste stored in the United States to start enough LFTRs to provide most electrical power used in the United States right now. It is not clear how close to commercial production a high breeding ratio IFR is right now.


On the question of fissile for start-up..
The official ‘Red Book’ estimate for economically recoverable global uranium reserves is 6.3 million tonnes. Enrichment leaving 0.2% U235 behind in the depleted uranium would yield about 0.5% of this, or 30,000 Te of fissile U235. LFTRs need 1 Te of fissile inventory per GW(e) of output, so there is enough fuel to give everyone (10 billion people) 3kw(e). The estimate of power requirements posted a few months ago on BNC was about half that. Low breeding rate therefore doesn’t exclude LFTRs – or any other concept that can ‘just about’ breed and has low fissile inventory. Whether mining and enriching uranium on such a scale is a good idea is another question, but it is possible. Starting the first 1000+ reactors on existing spent fuel, or pulling uranium out of seawater, help the argument for LFTRs, but neither is required.


Luke, by your name, I’m guessing your from the UK. I’m planning on starting some kind of ‘Environmentalists for NP’ type pressure group to counter the demonstrations against the new build, thats taking place next year.

I’m thinking leafleting, turning up at anti-Nuclear demos (outside build sites) to show the press that they are not the only environmentalists, etc

Interested in joining me?


Shouldn’t clutter BNC with this, but sure, worth a try. I live ~8 miles from the proposed Hartlepool C EPR site, so I’ve been wondering when Greenpeace etc were gong to show up to try to rouse NIMBY protest. Be nice to have something to give them other than just an argument. Contact Luke_dot_Collie_ta_boltblue_dot_com



Just sent you an email, let me know if you got it.


Sorry for cluttering up BNC with this stuff, but I’m sure you don’t mind when its organising for a pro nuclear thing!


Charles Barton:

You queried my comment relating to corrosion problems with LFTRs. Possibly I expressed myself clumsily.

Prof Peterson, as quoted by Barry, stated that “structural materials for fluid fuels require additional development including long term corrosion testing”. I accept that this is not to suggest that corrosion will be a problem when and if suitable materials are developed. However, it was the “long term” comment that struck me as having the potential to be problematic.

How is one to define long term, given that one would wish, for economic reasons, to get a 60 year life from a plant? How long would it take to satisfy regulators if the testing were to start today with money no object?

As a layman, I have found that the propagandists for LFTRs make a very compelling case, particularly with respect to their potential to produce electricity more cheaply than by any other method. Corrosion could be the potential show stopper. Thus, if one considers that it is urgent to close the fuel cycle, perhaps sodium rather than salt cooling is the way to go, even if one has to accept that it has lower potential for reducing electricity costs.

Please realise that I don’t really know what I’m talking about and am not advocating one technology over another. I am trying to acquire a more detailed understanding from informed commentators.



You suggested that fires were really not a big concern when considering sodium cooled reactors. You went on to suggest that the Russian BN600 was the most reliable design in its nuclear fleet.

Could you explain why the Russians would apparently prefer to adopt IFR technology rather than continuing with their own sodium cooled fast reactor. I appreciate that the former uses metal fuel and the latter metal oxide. Does pyroprocessing have something to do with it? if so, does this mean that the IFR has greater potential to produce electricity more cheaply than that from the BN600? Finally, setting aside its good reliability record, does the BN600 currently produce electricity as or more cheaply than the more conventional water cooled reactors in the Russian fleet?


Douglas, I would argue that even a severe corrosion problem is nothing like the show-stopper that it might be for water-based reactors, because those plant components are required to hold high pressures, whereas MSRs are roughly at atmospheric. So even replacing major portions of the reactors should not involve the expense traditionally associated with nuclear power.

Having said which, most corrosion problems are amenable to the correct materials and operating conditions, along with a program of inspection and maintenance.

I’d disagree, though. that it is “urgent” to close the fuel cycle. Important and useful, certainly, but not urgent.


The preferred Milton salt mixture for pebble bed reactors is toxic. Beryllium fluoride is highly toxic and very soluble in water. Lithium flouride is toxic. Also these salts emit toxic fumes when heated.
Thankyou for your comment. As a new commenter to BNC you may be unaware that, as a science based blog, BNC Comments Policy requires you to give links to refs to support your assertions. Please supply refs for your comments above.


That the molten salts are toxic if you eat them isn’t something anyone is denying.

It just so happens that it is possible to deal with the toxicity by not eating them (industry uses toxic water soluble chemicals all the time, there are ways to handle them safely).

If there’s a leak they’ll solidify thereby preventing them from getting out of the reactor (so it’ll just be a local hazard requiring the clean-up crew to wear protective clothing).


Pete: If we don’t use any of those chemicals we have a much worse disaster, namely lower living standards (a lot of them involved Ammonium Nitrate, you really don’t want to be without that though, at least if you care about not having the majority of the people on this planet starving).

That isn’t to say that we should be reckless, just that there isn’t a justification for outright banning them (nastier chemicals than them are used all the time in industry).


Leave a Reply (Markdown is enabled)

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s