Some time ago I wrote an essay on LFTR/Molten Salt Reactor safety from the prospective of a system of barriers to radiation release. My agenda was to argue that LFTR safety could be achieved through a system of barriers to the release of radioactive materials. This argument assumed that a fuel spill was the over riding safety issue. However, the classic texts on MSR safety (Gat and Dodds) do not examine MSR safety primarily in terms of a system of barriers. Gat and Dodds believed that
The Ultimate Safe Reactor (USR) is a special concept of a molten-salt reactor with prime and complete emphasis on safety. The USR uses a processing frequency, yet to be developed, that is about an order of magnitude higher from that contemplated for the molten salt breeder reactor (MSBR). The MSBR had a ten-day inventory turn around in the fuel processing. The USR uses a one day or less of turnaround of the fuel inventory. This rather fast turnaround reduces the build up of all fission products with half-lives of a few days or longer. The reactor is an epithermal spectrum reactor and uses no moderator per se in the core. The clean core consists solely of a low-pressure vessel. Freeze valves are used throughout. The prime circulating pump is sized to assure no critical cold slug accident can occur. Furthermore, the USR uses the Th-U fuel cycle with a breeding ratio of exactly one. Thus, the USR has all the safety benefits that are passive, inherent and non-tamperable and, in addition, has proliferation-resistant attributes and simplified waste that is free of fissile material, which can be transported in any arbitrary size or quantity from the processing part of the plant.
Beyond the ultimate safe reactor Gat and Dodd argued that there could be an absolute and ultimate safe reactor:
The absolute and ultimate safe reactor (A+USR) is a special concept of the USR which utilizes natural convection to transfer the heat from the core to the heat exchanger. The A+USR has no safety-related mechanical operating parts nor any externally-actuated controls, it becomes the ultimate in PINT-safety. The reactor responds internally and inherently to a change in power demand via its temperature response.
Frequent processing of the fuel increases the fuel inventory in the processing part and puts high demand on the performance of the processing units. The removal of the fission products from the fuel stream occurs at low concentrations, which requires precision and sophistication. In an actual plant, an optimization between performance, inventory and safety is needed.Thus Gat and Dodd saw MSR (and LFTR) safety in terms of reactor design features, that prevented accidents from happening, and prevented bad things from happening in the rare event of an accident. Gar and Dodds, argue, in effect that absolute and ultimate safety can be manufactured into Molten Salt Reactors, and can be implemented through low cost mass production manufacturing methods.
As a consequence of the Gat and Dodds argument is that an elaborate and costly system of barriers is not required. to assure absolute and ultimate nuclear safety. Mass produced, factory manufactured features can in most cases be low priced. Thus from the Gat and Dodds perspective LFTRs can be more safe at trivial costs than LWRs can be with the massive expenditure of money on safety features. This leads us to consider drastic, cost lowering changes in the way reactors are built.
Even the worst sort of reactor disaster, say an aircraft attack on a reactor, would not cause a massive release of radioisotopes, because the nuclear fuel would be continuously cleaned of radioisotopes. Since an attack on a reactor no longer poses great danger for a civilian population, the reactor holds little value as a target for terrorist.
Secondly, LFTRs can be air cooled. Meaning that they do not have to be sited next to water, and water shortages posed no difficulties for LFTRs.
i recently observed on Narry Brook's blog, Brave New Climate:
David LeBlanc has designed a very simple, low material LFTR that could easily mass produced. David tells me:
My work on the tube within tube will take very little material but I don`t have a number off the top of my head. Cost figures would be pretty much guesswork at this point but seems obvious that a simple tube should not cost very much. As for output levels, we could have a 1000 MWe tube within tube but I typically look around 200 MWe as a good size and this is about 1 meter wide (inner tube) and 6 meters long. This is surrounded by 60 to 100 cm of blanket salt and then an outer Hastelloy vessel. The tube material might be Hastelloy or Molybdenum (or many other things). David adds, “The heat exchangers will be a bigger user of metals like Hastelloy and that will be the same for just about any design.” in addition the LFTR would meed a couple of closed cycle gas turbine generators.David has discussed lowering reactor costs by building them with stainless steel. Using CO2 instead of helium we could get about 175 MWe from each. You could easily mass produce 4 per day, 400 if you wanted too. LFTRs are very safe, and all you need is a steel shed with prefabricated concrete radiation containment barriers and a cement floor to house the things. Thus not only would the mass manufacture of LFTRs allow for the timely deployment of huge ammounts of post carbon energy sources, but mass manufacture is entirely consistent with greatly enhanced nuclear safety, while lowering nuclear manufacturing costs. That safety in turn would allow for great cost savings in the construction of nuclear housing facilities.
Update 10/22/09: David LeBlanc disagrees with my assessment of the potential of low cost LFTR technology. I eat crow and go back to the drawing boards. Yesterday David wrote me: “There are a few options for cheap salts without tritium and still below Melting point 525. One is RbF-NaF-27%(Th,U)F4 (I think its 27, might be 22%) but that salt isn`t an option for a fuel salt of a Two Fluid (too much Th+U). The other is old fashioned NaF-ZrF4 which you can break even (with a bigger fissile load) and you can`t really get the melting point down much to use stainless steel.
I wouldn`t want to think of not using a containment building. All we need is something that is air tight and safe against aircraft crashes. It needs to be air tight for any gaseous leaks like Xenon. It doesn`t need to hold pressure or be a big volume so that makes it far cheaper than for LWRs.”
Posts
12 comments:
Charles, does David's design work as a two fluid and U233 inline reprocessing or it is one bed and a negative breeding ratio?
Obviously, costs go up at the two ends, that is, really tight containment is costly, but so is high capacity fuel processing. Seems to me that there should be a minimum cost point some where in between. The only question then is if that minimum in cost curve is shallow and broad, or if it is narrow and deep. If the former, than good arguments could probably be made for either design philosphy.
At least initially, it will probably be politically necessary to have the same level of containment as present light water reactors. However, since there is no danger of steam explosion or hydrogen explosion, the containment can be quite a bit smaller. It would still need to withstand the impact of an aircraft, but if anything, that should be easier with a smaller containment.
David, I think David L is playing around with the idea of a slightly negative breeding ratio, but there might be a few nuclear engineering tricks to improve it to to around 1 to 1. If they could get a 1 to 1 ratio on the Shippingport LWR, surely they could get it with a two fluid MSR.
And the function of containment in a reactor with one day turnaround on fuel cleaning is? It seems to me that the goal of nuclear safety is to reduce the casualty rate in the case of a nuclear accident to zero, and to reduce the likelyhood of nuclear accidents to something like once during the life of the universe. If we can do that without containment why do we need it?
As I Understand It (AIUI), the purpose of containment in a pressurized light water reactor is to contain a steam explosion, plus any release of radioactives.
That they are so robust as to be able to take a hit from an aircraft is a plus, now that we have to take terrorist attacks into account.
Me, I like the idea of "burning" up all of the actinide products so there is no radioactive waste issue. Even if some of the fusion research pays off, we'll still need to deploy deep-burn reactors at the existing reactor sites, to destroy the existing spent nuclear fuel.
Larry, the LFTR burns more deeply than anything else.
Non volatile fission products are a hazard only if there is a mechanism to eject them offsite, for example the power spike at Chernobyl resulting in a powerful steam explosion without benefit of a containment building. In an MSR the combination of a strong containment, low excess reactivity, lack of water, negative temperature coefficients, thermal spectrum and low fuel volatility eliminate all ejection mechanisms I know of.
Given these conditions, only volatile fission products are an offsite threat from an accident, and in any type of MSR they are continuously vented as they are formed. They would be bled off and converted to stable compounds where possible, and stored in deep well protected vaults. After a cooling off period they would be packaged for disposal or commercial distribution as applicable.
Rapid reprocessing for non volatile fission products does not materially enhance public safety, but it would make the plants more expensive, longer to build and less profitable, extending our reliance on fossil fuel.
The first generation plants should be dirt simple once through designs that can be designed tested and built in the shortest time at the lowest cost.
MSR’s are not a great deal safer than conventional reactors because conventional reactors are very low risk. Coal plants kill about 20,000 Americans each year. How many Americans have been killed by commercial nuclear plants in the last 50 years? Anti nuke’s say reactors are dangerous, we should not play into their misleading hand by implying it is true.
Bill Hannahan
A couple of points
First. If a LFTR can be effectively and cheaply built at a size power, say, lower than 1000 MW thermal, some cogeneration low grade heat applications (besides electricity production) starts to become quite interesting, for example : seawater desalination (less than 100 °C), district heating/cooling (100-150 °C) and ethanol/biofuel production ( ~ 200 °C). Has David or others ever considered for LFTR low grade heat applications like these?
Second, I guess that a full breeder (and not simply a cheaper and simpler high conversion reactor) is politically and strategically very important for things like energy indipendence and full waste destruction, so I'd tend at a first sight to favour a full breeder vs an high conversion plant
So the question is, how much more costly can be a thorium full breeder vs a simple high conversion LFTR ? Assuming the O&M plus fuel is quite negligible, if the difference in costs to build a full breeder vs an high conversion plant is not too high, my choice is for the breeder
What do you think about?
Alex, I support your ideas for low grade heat applications, but achieving those goals should not in any way slow down the process of developing a replacement for the steam supply system at coal fired power plants. Coal plants are designed for a 40 year lifetime and the average age of a coal plant in the U.S. is over 40 years.
Charles wrote a good piece on Jim Holms concept to convert coal plants to nuclear.
Regarding your second point, that the breeder “is politically and strategically very important for things like energy independence and full waste destruction”.
For a conventional reactor we mine 58 lb of uranium to create 9 lb of reactor fuel assembly to generate 5.4 ounces of fission products and an 80 year lifetime supply of electricity in the U.S. This saves burning 1.1 million pounds of coal and the release of 2.4 million pounds of CO2.
The dirt simple MSR burning uranium can do the same thing on 20 lb of mined uranium, producing a few pounds of glass or rock like waste. The actinides can be recycled at end of life making the waste storage time much shorter.
See David LeBlanc comments following this post.
http://thoriumenergy.blogspot.com/2009/08/recent-two-fluid-work-at-rez.html
We have enough uranium and thorium to be energy independent for several hundred years without the breeder.
The fuel reprocessing system would be the most challenging part of developing a breeder MSR. It would be on the critical path. I believe it would extend the development time and cost substantially, it would add substantially to the operational cost of the plant and it may substantially reduce plant reliability and make the resulting kWh’s much more expensive than those from a simple once through system. If there is an accident in the reprocessing system the plant will be shut down.
That said, I see the reprocessing equipment in a breeder built into modules the size of a standard shipping container. They would be installed and removed like a cassette. Each plant would have 2 or 3 modules, any one of which would do the job. Technology improvements would be incorporated in the modules without the need to overhaul the plant.
We built the first commercial HTGR in the 70’s at Fort St Vrain in Colorado. The reactor worked great but the design of the circulators was terrible. The number of HTGR’s built since then is zero.
“So the question is, how much more costly can be a thorium full breeder vs a simple high conversion LFTR ?”
If the first model fails the cost could be measured in decades.
The Wright brothers did not build a super sonic transport. The first MSR plant should be as simple as possible. Anything we do not build will not fail.
Bill Hannahan
Bill wrote : " The fuel reprocessing system would be the most challenging part of developing a breeder MSR "
I think this is the most important point. My guess was that at the time of the MSRE the technology was improved so far that today the fuel reprocessing system can be considered a fully mature technology. A break even Th/U-233 breeder (or, at the beginning, any other fissile start-ups with thorium) has 1) no fuel problem, 2) no high level waste problem, besides some small actinides leakages (less than one kg per GWyear), 3) no graphite need as moderator (thus, no irradiated graphite waste/replacement) and stil an high level of semplicity (even if not like simple one fluid reactors), as far I understand
But if one fluid, high conversion MSR systems can do almost the same, besides maybe point 3) (it's impossible to avoid graphite use in that case) with less costs and much more simplicity, I am perfectly happy
Aside from the technical issues, there's also psychological issues here.
If you have a big honking physical barrier between the radioactive bit of a nuclear plant and the outside world, people will probably feel safer, for the simple reason that everybody can understand why a big honking physical barrier can protect them; the details of reactor physics and chemistry are rather more complex.
I suspect this would go double if you could bury it underground, as some of the proposals for small reactors seem to.
Perceptions matter!
Robert this is a thought experiment. I am looking at the limitations of the practical where. Public perceptions would have to change. Theis would only be possible in the absence of a public perception problem, or where public perceptions were not a problem.
Post a Comment