Why Is Helium Preferred Over Nitrogen in High Temperature Gas Cooled Reactors?

[Azael]

In most high temperature gas cooled reactor concepts it seems like helium is the preferred gas. But what makes helium a better chooise than nitrogen?

And when on the topic of gas cooled graphite moderated reactors, what guarantees that the graphite will never catch fire? Is there just not plausible way that enough oxygen get into the core for it to be a possibility?

 

[Astronuc]

I imagine it has to do with the thermodynamic properties (per mass basis), fluid dynamic properties and the fact that He is chemically inert. Also He-4 does not activate, whereas with successive neutron capture N-14 (inert) -> N-15 (inert) -> N-16 and N-16 has a high energy gamma. N-16 decay to O-16 which is more chemically active, especially at high temperature.

http://www.nndc.bnl.gov/nudat2/decaysearchdirect.jsp?nuc=16N&unc=nds

 

[Azael]

I imagine it has to do with the thermodynamic properties (per mass basis), fluid dynamic properties and the fact that He is chemically inert. Also He-4 does not activate, whereas with successive neutron capture N-14 (inert) -> N-15 (inert) -> N-16 and N-16 has a high energy gamma. N-16 decay to O-16 which is more chemically active, especially at high temperature.

http://www.nndc.bnl.gov/nudat2/decaysearchdirect.jsp?nuc=16N&unc=nds

I see, thanks for the answer!

The neutron capture didnt cross my mind because of the low cross section.

 

[Astronuc]

Interestingly, N-16 carryover to the HP turbine and turbine train is a problem for BWRs on hydrogen water chemistry (HWC). Apparently the reducing environment puts more N in the steam which obviously gets carried over to the turbine. Hence the motivation for noble metal injection - which introduced its own set of surprises.

HWC was introduced to lower the electrochemical potential (ECP) in BWRs in order to prevent stress corrosion cracking of the SS (primary high purity SS304L) reactor internals and upper guide structure.

 

[vanesch]

But moreover, there is a N-14 (n,p) C-14 reaction http://www.nndc.bnl.gov/sigma/getPlot.jsp?evalid=4220&mf=3&mt=103&nsub=10
with non-neglegible cross section, so you produce C-14.

We use this reaction for low-efficiency thermal neutron detection, btw (in monitoring
devices).

I guess another reason is the much better thermal conductivity of He-4 (because of its low mass).

 

[vanesch]

And when on the topic of gas cooled graphite moderated reactors, what guarantees that the graphite will never catch fire? Is there just not plausible way that enough oxygen get into the core for it to be a possibility?

I personally, but probably without enough experience to say something sensible, dislike graphite-based reactors exactly for this reason. The worst accidents that happened in the past with reactors were with graphite reactors and exactly because of that problem (that graphite caught fire): Chernobyl and Windscale. All other reactor accidents (TMI included) were jokes compared to these two accidents. Once there's a confinement problem and the graphite is burning at 2000 degrees or something, you're in deep s**t, because it generates a hot smoke plume full of radioactive material which you cannot do much about. This is entirely different from water reactors or so. Once a big pile of charcoal (graphite) is on fire, there's nothing much you can do about it. If you pour water on it, you can have a gigantic hydrogen explosion, if you apply CO2, you can have a C + CO2 -> 2 CO reaction...
One shouldn't build a reactor in a hot charcoal stove in my opinion. Maybe there are now ways to render them more safely, but then I don't know about it.

 

[Astronuc]

Well - certainly one should not have a water-cooled, graphite-moderated reactor. That is just asking for trouble a la Chernobyl. The loss of coolant or reduced density actually put positive reactivity into the system because water absorps neutrons. Then if the high temperature water reacts with C, the product is Hydrogen. Windscale was problematic because of the air-cooling used.

The objective in high-temperature gas reactors of the pebble bed design is to use He gas and a finely divided fuel. The He gas good thermal conductivity and the small fuel particles provide for a large area for heat transfer. Also, the core can be 'dumped' in addition to have control elements (rods) inserted.

Thermal conductivities of gases (W/m-K).
Helium (20°C) ... 0.138
Hydrogen (20°C) ... 0.172 (H is great thermally, but certainly has other problems because of its chemical nature, especially as temperature increases)
Nitrogen (20°C) ... 0.0234

Viscosity of gases
http://en.wikipedia.org/wiki/Viscosity#Effect_of_temperature_on_the_viscosity_of_a_gas
http://hyperphysics.phy-astr.gsu.edu/hbase/tables/viscosity.html#c1 - contains He, air, N₂, O₂ at 20°C

Earlier graphite reactors utilized block fuel which means the fuel has to be removed with crane or handling machine rather than a simply flow scheme. Similarly, AGR fuel, which has a similar geometry to CANDU fuel, must be moved as a string (set of assemblies) with a machine but this can be done online.

Here is a nice overview of reactor technology

http://www.euronuclear.org/library/public/enews/ebulletinautumn2004/nuclear-reactors.htm

Here another interesting presentation
http://www.ornl.gov/~webworks/cppr/y2001/pres/124827.pdf

 

[vanesch]

Well - certainly one should not have a water-cooled, graphite-moderated reactor. That is just asking for trouble a la Chernobyl. The loss of coolant or reduced density actually put positive reactivity into the system because water absorps neutrons. Then if the high temperature water reacts with C, the product is Hydrogen. Windscale was problematic because of the air-cooling used.

Yes, that's correct, and was bad design from the start in both cases. This was the cause of the accident, and with enough brains, you can avoid such a flawed design. But the problem I'm having with (massive and bulky) use of graphite is that, if for one or other reason there is an accident, then the presence of hot and burning graphite aggravates seriously the consequences of an accident. I'm thinking of the post-accident phase where one should think of mitigating the consequences (the large-scale spreading of highly radioactive material in the atmosphere, which is ultimately the only serious potential problem with any nuclear activity). On that point, graphite scores badly.

The objective in high-temperature gas reactors of the pebble bed design is to use He gas and a finely divided fuel. The He gas good thermal conductivity and the small fuel particles provide for a large area for heat transfer. Also, the core can be 'dumped' in addition to have control elements (rods) inserted.

I don't know much about this design - I've heard a lot of good about it though.

Earlier graphite reactors utilized block fuel which means the fuel has to be removed with crane or handling machine rather than a simply flow scheme. Similarly, AGR fuel, which has a similar geometry to CANDU fuel, must be moved as a string (set of assemblies) with a machine but this can be done online.

I was indeed thinking of the "bulk" application of graphite ("a pile of charcoal"). Maybe better materials are available now, like SiC or so, which is not combustible.

 

[Morbius]

Well - certainly one should not have a water-cooled, graphite-moderated reactor. That is just asking for trouble a la Chernobyl. The loss of coolant or reduced density actually put positive reactivity into the system because water absorps neutrons.

Astronuc,

The positive coolant temperature reactivity coefficient is NOT endemic to graphite moderated
reactors - but is a design deficiency of the RBMK.

The problem with the Chernobyl RBMK is not that it is a water-cooled graphite reactor; but that
it was "over-moderated". If the graphite can do the entire job of moderation without assistance
from the water - then, yes; when you lose water you are losing absorber. This is the situation
when you have an over-moderated reactor.

However, if the RBMK had less graphite, i.e. if it was "under-moderated" - then the loss of water
would entail not only loss of absorber - but also loss of moderator.

After all, the power reactors in the USA are all LWR - light water reactors that are moderated by
water. When you lose water - the result is decreased reactivity.

The main problem with the RBMK is that it was a poorly scaled up version of a Soviet weapons
production reactor. The production reactor had the proper balance between fuel, water and graphite
concentrations.

However, the RBMK essentially scaled up this production reactor; which due to the larger size;
reduced the leakage of neutrons per volume of core. This reduced leakage should have been
compensated for by reducing the amount of moderation - that is reducing the amount of graphite
in the core.

The Soviet designers failed to do that. Hence the RBMK reactor was "over-moderated".

A water cooled graphite reactor doesn't have a positive coolant temperature feedback merely
because it is a water cooled graphite reactor. One can design a water cooled graphite reactor
with a negative coolant termperature coefficient.

The Dept. of Energy operated the "N Reactor" at Hanford from 1963 to 1987 and it was a
water-cooled graphite moderated reactor that was both a power reactor and production
reactor like Chernobyl.

http://www.hanford.gov/?page=345&parent=326

The problem with the Chernobyl RBMK was NOT the type of design; but the execution of the design.

Dr. Gregory Greenman
Physicist

 

[vanesch]

Actually, in as far as I understand these things, having a positive void coefficient, or a positive coolant coefficient, by itself, is not a problem, as long as this is compensated somehow by another negative feedback, such as enough Doppler effect. A sodium cooled fast reactor usually also has a positive void coefficient and can have a positive coolant coefficient, but, as has been shown in the past (I think it was the IFR where such a spectacular experiment took place), such a reactor can nevertheless be made passively safe.
Of course, you're more on the safe side if all individual effects work against reactivity increase, but you can make designs where one "bad" effect is compensated by more of another "good" effect.

Correct me if I'm wrong on this...

 

[Astronuc]

The response of the reactor does depend on the relative magnitudes of positive and negative reactivity involved, or otherwise the degree of over- or under-moderation as Morbius mentioned. Ideally, as a reactor departs from normal operation, the negative reactivity from temperature increase, or moderator density decrease would increase to terminate an off-normal power ascension.

With higher enrichments and larger batch sizes in LWRs, some core designs did have +MTC, which is generally not accpetable. So core designs have to employ design modificaitons (e.g. burnable absorber distribution and concentration) to achieve -MTC.

 

[theCandyman]

Wasn't the Chicago Pile also moderated by graphite? I'm not sure how it was set up though.

 

[Azael]

This certainly turned into a interesting discussion :)

Interestingly, N-16 carryover to the HP turbine and turbine train is a problem for BWRs on hydrogen water chemistry (HWC). Apparently the reducing environment puts more N in the steam which obviously gets carried over to the turbine. Hence the motivation for noble metal injection - which introduced its own set of surprises.

What kind of problem does nitrogen cause when it reaches the turbines?

 

[Azael]

The objective in high-temperature gas reactors of the pebble bed design is to use He gas and a finely divided fuel. The He gas good thermal conductivity and the small fuel particles provide for a large area for heat transfer. Also, the core can be 'dumped' in addition to have control elements (rods) inserted.

Roughly how high can the temp of the fuel particles get during a complete loss of coolant?

 

[Morbius]

Wasn't the Chicago Pile also moderated by graphite? I'm not sure how it was set up though.

Candyman,

Yes - but it was "air-cooled". Actually it didn't have a forced cooling system at all.

So there was no problem with voiding of water coolant.

Dr. Gregory Greenman
Physicist

 

[Astronuc]

This certainly turned into a interesting discussion :)

What kind of problem does nitrogen cause when it reaches the turbines?

It's that 6 MeV gamma ray from N-16.
http://www.nndc.bnl.gov/nudat2/getdecayscheme.jsp?nucleus=16O&dsid=16n bM decay&unc=nds

Roughly how high can the temp of the fuel particles get during a complete loss of coolant?

That I'm not sure about because it depends on the decay heat and the heat removal system - whether is by forced convection with gas (and at what flow and pressure) or by conduction. I've not seen any calcs on these systems.

 

[Morbius]

Roughly how high can the temp of the fuel particles get during a complete loss of coolant?

Azael,

Actually, the limiting consideration doesn't have to do with the fuel but the cladding.

You want to be sure that the cladding surrounding the fuel isn't compromised; the cladding
is the first line of defense against the release of radioactive material, and it also holds the
fuel in a coolable geometry. As long as cladding remains intact; then any accident consequences
will be rather minor.

The zirconium cladding undergoes an oxidation reaction with water starting at a temperature of
about 1750 F; if memory serves. If the peak cladding temperature is held to less than about
2200 F; then no more than about 15% of the cladding thickness will be oxidized. That's the
limit that is specified in the reactor's "tech specs".

Caveat: the above values are only approximate; I'm dredging this up from memory of 30 yrs ago.

Dr. Gregory Greenman
Physicist

 

[Astronuc]

The zirconium cladding undergoes an oxidation reaction with water starting at a temperature of about 1750 F;

The key is 'rapid' or 'breakaway' oxidation.

Zr alloys will oxidize slowly at operating temp in a BWR (~285°C) or PWR (~285-330°C).

By the end of 4-6 years of operation, one only wants the cladding to have about 40-60 microns of oxide.

The number referenced by Morbius apply to accident conditions in which there is a risk of the cladding material rupturing as a result of the Zr + 2H₂O -> ZrO₂ + 2 H₂ reaction which was the problem at TMI-2. The fuel cladding is the first barrier to contain fission products. The second and third barriers are the primary system and containment building structures, respectively.

 

[Morbius]

The key is 'rapid' or 'breakaway' oxidation.

Zr alloys will oxidize slowly at operating temp in a BWR (~285°C) or PWR (~285-330°C).

By the end of 4-6 years of operation, one only wants the cladding to have about 40-60 microns of oxide.

The number referenced by Morbius apply to accident conditions in which there is a risk of the cladding material rupturing as a result of the Zr + 2H₂O -> ZrO₂ + 2 H₂ reaction which was the problem at TMI-2. The fuel cladding is the first barrier to contain fission products. The second and third barriers are the primary system and containment building structures, respectively.

Astronuc,

If memory serves; the reaction you quote above is exothermic. So if it gets started, then you
have the "breakaway" oxidation you refer to.

However, the reaction has a threshold activation temperature.

Am I remembering this correctly, Astronuc?

Dr. Gregory Greenman
Physicist

 

[vanesch]

Astronuc,

If memory serves; the reaction you quote above is exothermic. So if it gets started, then you
have the "breakaway" oxidation you refer to.

However, the reaction has a threshold activation temperature.

They have a fun test facility to blow up claddings (without real fuel) in Karlsruhe, in the QUENCH facility: http://cat.inist.fr/?aModele=afficheN&cpsidt=17652994

There seems to be a kind of instability at a certain point: or the cooling sets in first, or the exothermal H2 production sets in.

 

[Astronuc]

Astronuc,

If memory serves; the reaction you quote above is exothermic. So if it gets started, then you
have the "breakaway" oxidation you refer to.

However, the reaction has a threshold activation temperature.

Am I remembering this correctly, Astronuc?

Dr. Gregory Greenman
Physicist

Yes, it is exothermic, but it is rate controlled by the dissociation of the water and diffusion of oxygen to the metal-oxide interface. At slow oxidation rates, Zr builds up a protective oxide, but at high rates, the oxide is porous and can flake away, thus providing no protection.

There is a temperature threshold below which the reaction proceeds slowly, but at a certain temperature, the reaction takes off violently. I'll have to dig around for that temperature, which may depend on composition. Pure Zr oxidizes very rapidly, while alloys like Zr-4 and Zr-2 much more slowly, but that depends on temp and nature of the oxide film.

Combustion of fine Zr-wire used to be the basis of flash bulbs. It is very pyrophoric.