Nuclear Reactor With No Moving Parts

[sysreset]

There's an article in the May 12 2008 New Yorker Magazine by Malcolm Gladwell, in which he talks about brainstorming sessions by teams of inventors headed by Nathan Myhrvold. One of the ideas is a small Nuclear Reactor with no moving parts. Supposedly the core would be about 3x10 meters, enclosed in a sealed armored box. It would run for thirty years, and put out 3 gigawatts. The point was that with no moving parts there would be no possibility of human error, and nuclear accidents could be eliminated. It would also be designed to eliminate nuclear weapon fuel generation.

Are there some flaws in this idea?

 

[vanesch]

There's an article in the May 12 2008 New Yorker Magazine by Malcolm Gladwell, in which he talks about brainstorming sessions by teams of inventors headed by Nathan Myhrvold. One of the ideas is a small Nuclear Reactor with no moving parts. Supposedly the core would be about 3x10 meters, enclosed in a sealed armored box. It would run for thirty years, and put out 3 gigawatts. The point was that with no moving parts there would be no possibility of human error, and nuclear accidents could be eliminated. It would also be designed to eliminate nuclear weapon fuel generation.

Are there some flaws in this idea?

I guess you're talking about this kind of thing ?

http://www.science-direct.com/science?_ob=ArticleURL&_udi=B6V3X-4RWRGSF-2&_user=10&_coverDate=08%2F31%2F2008&_rdoc=7&_fmt=high&_orig=browse&_srch=doc-info(%23toc%235742%232008%23999499997%23681062%23FLA%23display%23Volume)&_cdi=5742&_sort=d&_docanchor=&_ct=113&_acct=C000050221&_version=1&_urlVersion=0&_userid=10&md5=27b6c4282709ce4634d1691b470c528d

 

[Astronuc]

One could use natural convection to drive the coolant, and that is part of the ESBWR design. I believe the no moving parts applies to the primary or recirculation system, and not the feed water system. On the balance of plant side, one would need pumps for the condensate system.

I'd like to see their idea for reactivity control.

 

[daveb]

I also have to wonder about the feasibility of this in regards to control systems. How would you be able to control the reactor without movable control systems?

 

[mgb_phys]

How would you be able to control the reactor without movable control systems?

Pebble bed reactors have decreased power output at higher temperatures so are self controlling.

 

[vanesch]

Pebble bed reactors have decreased power output at higher temperatures so are self controlling.

In fact, almost all modern reactors have this property - doing otherwise is considered quite unsafe (like Chernobyl's reactor). However, the excursion of temperature in certain reactors before the reactivity drops totally may be too large to be used as a practical way of controlling the reactor in normal circumstances. The pebble bed reactor, being made of graphite spheres and an inert gas, is able to withstand very large temperature excursions without being damaged (as long as no oxygen gets to the graphite...), and hence can be passively regulated entirely.

A PWR could also in principle be regulated by natural negative feedback, but the temperature variations would be too large to be practical. A BWR is regulated mostly naturally if I understand well. Some sodium-cooled fast reactors like phenix also have this property (by Doppler, and by dilatation of the set of fuel elements).

 

[mheslep]

I'd like to see their idea for reactivity control.

Old thread, but interest is high w/ company about ready to launch to commercialize. Here's the reactivity control from the paper (Lithium 6 and 7):

Located throughout the core's fuel-charge in a 3-D lattice whose cell constant is roughly a mean free path of a median-energy-for-fission neutron, each of these modules consists of a pair of metallic compartments, each one of which is fed by a capillary tube. The small thermostat-bulb compartment located in the fuel always contains Li7, whose neutron absorption cross-section is essentially zero for neutron energies of interest, while the relatively large one positioned in a cooler location on the wall of a coolant tube may contain variable amounts of Li6, which has a comparatively large neutron absorption cross-section. [Lithium, melting at 453 K and 1-bar boiling at 1615 K, is a liquid across the entire operating temperature range of the reactors of interest, and its two stable isotopes each have useful nuclear properties.] As the fuel temperature rises, the thermostat-bulb-contained Li7 expands, and a small fraction of it is expelled (not, vert, similar10−3, for a 100 K temperature change), potentially under kilobar pressure, into the capillary tube which terminates on the bottom of the cylinder-and-piston assembly located outside of the radiation shield and physically lower than the Li6 intra-core compartment. There the modest volume of high-pressure Li7 drives a swept-volume-multiplying piston, which pushes a three order-of-magnitude larger volume of Li6 through a core-threading capillary tube into an intra-core compartment adjacent to but cooler than the thermostat-bulb, which is driving the flow. There the Li6, whose spatial configuration is immaterial as long as its smallest dimension is less than a neutron mean free path, acts to absorptively depress the local neutron flux, thereby reducing the local fuel power density. When the local fuel temperature drops, Li6 returns to the cylinder-and-piston assembly under action of a gravitational pressure-head, thereby returning the Li7 to the thermostat-bulb whose now-lower thermomechanical pressure permits it to be received. A total loading of ≤ 104 mol – ≤ 60 kg – of Li6 is sufficient for the filling of the not, vert, similar103 thermostating modules of the reference 1 GWe-reactor.

Talk from new CEO Gilleland, a nuclear industry grey beard, at Berkley:
http://www.nuc.berkeley.edu/files/TerraPowerGilleland.pdf

 

[Astronuc]

Something to bear in mind when thinking about core and nuclear fuel design.

The functions of conventional fuel are:

• Position fissile material in the reactor core in a stable and predictable manner to allow a controlled fission reaction

• Allow effective transfer of nuclear reaction heat from the fuel to the coolant (or heat transfer medium)

• Provide containment of radionuclides (fuel and fission products) for operational convenience and as a first barrier for safety

• Provide/allow a convenient means of loading fresh fuel into the core and removing and managing spent fuel

• Perform the above for normal operation and (most) Design Basis Events

The TWR is definitely a departure from this set of functional requirements. I have to wonder about the heat transport from core to working fluid, and subsequent handling of fission products. Perhaps the modularity helps with the later, but I haven't really looked into this design very much.

 

[joelupchurch]

I tracked down the New Yorker Article here:
http://www.newyorker.com/reporting/2008/05/12/080512fa_fact_gladwell?currentPage=3"

The reference to Teller is obviously the TRIGA reactor:
http://en.wikipedia.org/wiki/TRIGA"

I wrote about the Traveling Wave Reactor in my blog back in July:
http://www.anupchurchchrestomathy.com/2009/07/nuclear-reactor-that-nevers-needs-to-be.html"

 

[mheslep]

Something to bear in mind when thinking about core and nuclear fuel design.

The functions of conventional fuel are:

• Position fissile material in the reactor core in a stable and predictable manner to allow a controlled fission reaction

• Allow effective transfer of nuclear reaction heat from the fuel to the coolant (or heat transfer medium)

• Provide containment of radionuclides (fuel and fission products) for operational convenience and as a first barrier for safety

• Provide/allow a convenient means of loading fresh fuel into the core and removing and managing spent fuel

• Perform the above for normal operation and (most) Design Basis Events

The TWR is definitely a departure from this set of functional requirements.

From my read, it departs only for your last two. Because of the cheap fuel (²³⁸ U or Th) and high burn up design, the intent is to fuel the reactor at the start for its life time (15 to 60 years). No refueling, no control rod operation to attend to. In other words, the TWR authors intend to eliminate all but one step of the existing fuel cycle (only fuel fabrication remains - mining, refinement, enrichment all go away, as do the post operation steps). Otherwise it appears to me the first three of your reactor requirement still apply.

I have to wonder about the heat transport from core to working fluid,

Gas cooling, He, at least in the original concept. There's a graphic in the paper showing the fuel/cooling lattice they have mind. Also note they intend to run very hot - 1200K.

and subsequent handling of fission products. Perhaps the modularity helps with the later, but I haven't really looked into this design very much.

The authors intend for the depleted reactor to remain in place, buried underground at its operation point for at least a 100 years. Residual radioactivity power is dispersed by the same cooling mechanism used fission heat transfer when the reactor was operational. After a century or so remaining radioactivity would me de minimus according to them.

 

[mheslep]

Question: would all the criticisms of the fast breeder designs, valid or not, equally apply to this design?

 

[mheslep]

Proliferation comment and question on the TWR:
The authors make the fairly inane comment (in my view) in the paper that the the TWR design eliminates proliferation concerns because the fuel, buried underground with the reactor, never gets touched. I'll grant that reducing fuel handling lowers diversion risk, but its trifling to say stop there regards weaponization, at least I didn't see it if was obvious. That is, if a TWR reactor fueled by depleted U, was handed over to a want to be weapons state is it not possible that they could start the reactor, i.e. start the the U to Pu breeding wave, shut down the reactor, dig it up and chemically process out the Pu? Perhaps that process is not feasible due to reactivity levels, I wouldn't know.

 

[Bob S]

Why don't you look at the Oklo reactor design.
http://www.ocrwm.doe.gov/factsheets/doeymp0010.shtml
It was self regulating (boiling water neutron moderator) without any moving parts (or EPA oversight). It ran about 1.7 billion years ago.
Bob S

 

[mgb_phys]

Why don't you look at the Oklo reactor design.
It was self regulating (boiling water neutron moderator) without any moving parts (or EPA oversight). It ran about 1.7 billion years ago.

Construction time was a bit of a problem.

 

[mheslep]

Why don't you look at the Oklo reactor design.
http://www.ocrwm.doe.gov/factsheets/doeymp0010.shtml
It was self regulating (boiling water neutron moderator) without any moving parts (or EPA oversight). It ran about 1.7 billion years ago.
Bob S

Construction time was a bit of a problem.

And the cost of the containment vessel is prohibitive.

 

[joelupchurch]

Question: would all the criticisms of the fast breeder designs, valid or not, equally apply to this design?

I don't know about all breeder reactors, but it seems to skip the electrometallurgical reprocessing that the GE Prism reactor requires, but that might be because the Prism is designed to process spent fuel.

Wikipedia actually has an article on the TWR now:
http://en.wikipedia.org/wiki/Traveling_wave_reactor"

 

[mheslep]

I don't know about all breeder reactors, but it seems to skip the electrometallurgical reprocessing that the GE Prism reactor requires, but that might be because the Prism is designed to process spent fuel...

So is the TWR. The point is it is somewhat fuel agnostic (U238, Th, Pu). If one doesn't much care about the content of spent fuel - the wave converts or burns whatever is present - then the separation step disappears.

 

[mheslep]

Slide 31 here states the TWR company TerraPower is using MCNPX-CINDER90 to simulate the burn.
http://www.nuc.berkeley.edu/files/TerraPowerGilleland.pdf

Also appears the Lithium control system is out. Now back to B4C control and safety rods, HT-9 fuel clad (Slide 34)

 

[joelupchurch]

Proliferation comment and question on the TWR:
The authors make the fairly inane comment (in my view) in the paper that the the TWR design eliminates proliferation concerns because the fuel, buried underground with the reactor, never gets touched. I'll grant that reducing fuel handling lowers diversion risk, but its trifling to say stop there regards weaponization, at least I didn't see it if was obvious. That is, if a TWR reactor fueled by depleted U, was handed over to a want to be weapons state is it not possible that they could start the reactor, i.e. start the the U to Pu breeding wave, shut down the reactor, dig it up and chemically process out the Pu? Perhaps that process is not feasible due to reactivity levels, I wouldn't know.

As I recall, they would have to cut it open to get to any Plutonium and even then it would only be the front few inches of the wave that would actually contain any Plutonium-239 that wasn't contaminated with 240 and higher isotopes. Cutting it open would make restarting it problematical.

It would be easier to build a weapons reactor from scratch like the North Koreans or use centrifuges like the Iranians.

I suspect that the TWR was developed from the SSTAR reactor concept. Some of the people working on the TWR used to work at LLNL.
http://en.wikipedia.org/wiki/Sstar"

 

[mheslep]

As I recall, they would have to cut it open to get to any Plutonium

Yes, but that is a small impediment.

and even then it would only be the front few inches of the wave that would actually contain any Plutonium-239

Yes, but in a ~100MW TWR , 300MW(t), that's likely to be plenty of Pu for weapon.
Edit: Or not? 300MW is about 1x10^19 fissions (200MeV) per second, a burn rate of ~4 milligrams of Pu-239 per second. That means about 6-7 milligrams per second of Pu-239 must be created by the breeder wave, the balance undergoing capture to Pu-240.

that wasn't contaminated with 240 and higher isotopes.

Pu-240 accumulation apparently is indeed the big barrier-to-entry for Pu weapons, i.e. it helps prevent proliferation. But I'm not sure of the mix here; I'm guessing a good answer requires a serious analysis. The pathways on Pu-239 are 64% fission, 36% neutron capture to 240, so 240 would certainly be in the TWR somewhere, but then at shut down all of the Np in the breed wave would be in place waiting to decay to pure Pu-239 over a few days. There might be a geometric slice of the reactor that had exceedingly pure Pu-239.

Cutting it open would make restarting it problematical.

That addresses someone trying to start an entire nuclear weapons industry, not someone trying to make 1 or 2 weapons per TWR.

It would be easier to build a weapons reactor from scratch like the North Koreans or use centrifuges like the Iranians.

It seems to me dumping the wavefront guts into a chemical PUREX bath might quickly recover high grade Pu 239. Possibly starting a new reactor from scratch is easier, I wouldn't know. The centrifuges however - there's nothing easy about them - they necessarily entail a massive, large scale facility and a lot of electricity.