Why Might Beryllium Be Considered for Use in Nuclear Reactor Fuel?

[artis]

This has probably been asked before but why don't commercial LWR use beryllium in the core or more likely around the fuel or mixed within the fuel as an additive?
Beryllium gives off 2 neutrons if hit by a neutron that has on average more than 1.9 MeV and also produces neutrons when bombarded with alpha emission. Both cases are present in a reactor core, alpha emission from fission isotope decay and all neutrons being born as fast.
Wouldn't the use of beryllium or other similar neutron multiplier/source result in a more efficient core where a larger fuel burnup can be gained?
The most efficient moderator is heavy water IIRC, so what if one combined the original CANDU design that had heavy water both as calandria moderator and active cooling fluid in the channels with some beryllium in fuel cladding or mixed in as surface film, what would it do to the already neutron efficient reactor?
IIRC original CANDU much like the original RBMK design was able to run on natural Uranium with no artificial enrichment due to it's neutron economy and moderation to absorption values?

 

[Astronuc]

This has probably been asked before but why don't commercial LWR use beryllium in the core or more likely around the fuel or mixed within the fuel as an additive?

That has been considered for the purpose of improving the thermal conductivity of UO₂ fuel, which is the main ceramic fuel form in LWR fuel, by adding BeO to the UO₂. However, Be is very toxic, and it would make the bulk manufacture of UO₂ fuel much more complicated, e.g., enclosed production lines for powder and pellets, and folks around the fuel would likely have to wear respirators and protective covering, or risk berylliosis. In addition, BeO would displace some UO₂, so then the question would become, how much of the UO₂ would be replaced, e.g, 1%, or more or less, so one would need a slight increase in enrichment. Many fuel designs now are a maximum level of 4.90/4.95 % enrichment, although there is a current effort to

Beryllium gives off 2 neutrons if hit by a neutron that has on average more than 1.9 MeV and also produces neutrons when bombarded with alpha emission. Both cases are present in a reactor core, alpha emission from fission isotope decay and all neutrons being born as fast.

The 1.9 MeV value represents a threshold energy of the fast neutron for the (n,2n) reaction, so one doesn't gain 2 neutrons, only one, since the incident neutron is replaced by 2n. The cross-section at 1.9 MeV is very low, ~ 1E-4 b, but increases to about 0.3 b at 3 MeV and to 0.6 b at 6 MeV, which is still fairly low. See attached figure. But then one must consider that the fast neutron population falls for energy > ~ 1 MeV.

Wouldn't the use of beryllium or other similar neutron multiplier/source result in a more efficient core where a larger fuel burnup can be gained?

No, not really.

There are two more complications: 1) ⁹Be undergoes a photoneutron or (γ,n) reaction, in which a gamma with an energy greater than 1.643 MeV cause ⁹Be to disintegrate into n + 2α, which introduces complication 2) helium generation in the fuel, which would contribute to internal pressure of the fuel rod.

The most efficient moderator is heavy water IIRC, so what if one combined the original CANDU design that had heavy water both as calandria moderator and active cooling fluid in the channels with some beryllium in fuel cladding or mixed in as surface film, what would it do to the already neutron efficient reactor?
IIRC original CANDU much like the original RBMK design was able to run on natural Uranium with no artificial enrichment due to it's neutron economy and moderation to absorption values?

The use of natural U in CANDU limits the residence time and burnup of the fuel. Modern CANDU fuel designs use slightly higher enrichment to extend the residence time and burnup capability. Like RBMKs, the CANDU produce Pu isotopes, ²³⁹Pu and ²⁴¹Pu, which are fissile and contribute to the energy production. A Be layer on the cladding or fuel pellet would add to the complexity of the fuel design, and more so, would require significant changes to the production lines.

In conjunction with extended enrichment, the nuclear industry is considering extended burnup for current LWR fuel designs, from 62 GWd/tU to 68 GWd/tU and perhaps to 75 GWd/tU. That introduces other technical issues related to increased rod internal pressure, and concerns about fuel behavior/performance under RIA and LOCA conditions. Some experiments over the past decade or so have shown that high burnup fuel is potentially susceptible to "fuel fragmentation, relocation and dispersal" (FFRD) under certain accident conditions. This is a consequence of the development of the "high burnup structure" (HBS) on the periphery of the fuel pellet at high burnup. HBS involves the development of submicron grains and accumulation of Pu and fission products, notably Te, I, Xe, Cs, at the edge of the pellet during operation. Consequently, one would have to limit where fuel is placed in the core, e.g., at the periphery (outer row) of the core in order to limit maximum energy/power during a postulated/hypothetical accident condition (e.g., RIA or LOCA).

 

[artis]

The 1.9 MeV value represents a threshold energy of the fast neutron for the (n,2n) reaction, so one doesn't gain 2 neutrons, only one, since the incident neutron is replaced by 2n. The cross-section at 1.9 MeV is very low, ~ 1E-4 b, but increases to about 0.3 b at 3 MeV and to 0.6 b at 6 MeV, which is still fairly low. See attached figure. But then one must consider that the fast neutron population falls for energy > ~ 1 MeV.

Well a neutron goes in,capture happens the nucleus splits into 2 alphas +2n which are free to fly off is what I understand but it seems that is not the only reaction that can happen with Be9+n so one would need to look at reaction rates for each of them.

I guess from what you confirm is that Be9 doesn't really make that much of a difference in a LWR core in terms of producing more neutrons, but I suppose it does come in handy for a fast neutron prompt uncontrolled chain reaction like that in an A bomb, where IIRC from the diagrams , many bombs have used Be as primary tamper lining or so, where the goal I understand is to both reflect and multiply neutrons for a better reaction rate and larger yield?

UO2 fuel, which is the main ceramic fuel form

Maybe not directly on topic but I've always wondered why UO2 fuel is ceramic? It's main ingredient is still Uranium which is a metal, or do they classify metals as ceramics whenever there is a certain metal to oxygen ratio in the oxide, because in nature many metals oxidize naturally in contact with oxygen like the surface of Aluminum and yet we don't say the surface of Aluminum has become ceramic?

Like RBMKs, the CANDU produce Pu isotopes, 239Pu and 241Pu, which are fissile and contribute to the energy production.

This does happen in any LWR fuel irrespective of core design right?
I could say in any U238/235 fuel mass capable of a chain reaction?

 

[Astronuc]

Well a neutron goes in,capture happens the nucleus splits into 2 alphas +2n which are free to fly off is what I understand but it seems that is not the only reaction that can happen with Be9+n so one would need to look at reaction rates for each of them.

Yes, one would have to look at the other reactions and determine how each would contribute to the neutron population, or not. Consider that ²³⁵U releases on average slightly more than 2 neutrons per neutron absorbed, and more like 2.2 to 2.3 neutrons (on average, so usually either 2 or 3), and that would be more preferable than 2 neutrons per neutron absorbed.

I guess from what you confirm is that Be9 doesn't really make that much of a difference in a LWR core in terms of producing more neutrons, but I suppose it does come in handy for a fast neutron prompt uncontrolled chain reaction

Be has been used in neutron sources in commercial nuclear reactors. Early primary sources would contain an alpha emitter mixed with Be, whereby an (α,n) produces a neutron, while secondary sources would use ¹²³Sb mixed with Be, in which ¹²³Sb + n => ¹²⁴Sb => ¹²⁴Te + β^- + γ (1.64 MeV), which induces a (γ,n) reaction in Be. Primary sources are now ²⁵²Cf.

https://www.frontier-cf252.com/antimony-beryllium/
https://www.osti.gov/servlets/purl/4275578

I will not comment on nuclear weapons technology.

Maybe not directly on topic but I've always wondered why UO2 fuel is ceramic? It's main ingredient is still Uranium which is a metal, or do they classify metals as ceramics whenever there is a certain metal to oxygen ratio in the oxide, because in nature many metals oxidize naturally in contact with oxygen like the surface of Aluminum and yet we don't say the surface of Aluminum has become ceramic?

Ceramic usually refers to a compound of a metal, or metals, and non-metal, usually C, N, or O. For example, U ceramics are uranium carbide (UC), uranium nitride (UN) and uranium dioxide (UO₂). One can substitute Pu or Th for U in these compounds. We also find other carbides, nitrides and oxides. In the case of U, the ideal ceramic is UO₂, but there are forms such as U₄O₉, U₃O₇, U₃O₈, and UO₃, which are undesirable for fuel since the U density is lower and the thermal conductivity is lower. UN and UC have lower melting points than UO₂, but they have much greater thermal conductivity. There is also the concern about chemical compatibility with water in the event the cladding is breached which has occurred with some regularity in commercial LWRs, even with attempts to reduce failure rates to zero. We also have U(C,N), a carbonitride, and UCO, a carboxide. Except for special test/experimental fuel rods, commercial nuclear fuel contains UO₂.

There are other forms for U, since U₃Si or U₂Si₃, which are kind of an intermetallic, and U-10%Mo and U-Zr alloy or intermetallic forms. On has to consider fission product dispositions with each system, in addition to coolant compatibility.

With respect to oxides of aluminum, we would refer to Al₂O₃, or alumina. And one may find alumina pellets in some fuel designs, or HfO₂ (hafnia), or ZrO₂ (zirconia). With respect to an oxide layer on a metal, we might say the surface is passivated, depending on the environment. In the case of metal like Al, Ti, Zr, Hf, Cr (in stainless steels), very thin submicron layers (or nanometers) of oxide form that protect the underlying metal for further oxidation/corrosion, unless the layer is disturbed/compromised.

In some fuel, one may find neutron-absorbing burnable poisons, or burnable absorbers, e.g., Gd₂O₃ (gadolinia) or Er₂O₃ (erbia) blended into the UO₂, or ZrB₂ coated on the outer (circumferential) surface of the UO₂ fuel pellets.

This does happen in any LWR fuel irrespective of core design right?
I could say in any U238/235 fuel mass capable of a chain reaction?

Yes, in a thermal reactor, or epithermal reactor, ²³⁸U will absorb fast and epithermal neutrons, to become ²³⁹U, which undergoes beta decay to ²³⁹Np, which undergoes beta decay to ²³⁹Pu. However, some of the 239 isotopes of each element can also absorb neutrons to become 240 and 241, which also undergo beta or alpha decay, or in some cases fission. As burnup increases, more and more ²³⁹Pu, ²⁴⁰Pu, ²⁴¹Pu, ²⁴²Pu, are produced as well as isotopes of Am and Cm, with quite a lot on the outer surface of the fuel pellets.

With respect to ²³⁵U, some of the neutron captures do not result in fission, but instead a gamma ray is emitted which leads to a more stable ²³⁶U, which can absorb a neutron to become ²³⁷U. Those nuclides undergo beta decay forming ²³⁶Np and ²³⁷Np, respectively, and some of those nuclides can absorb a neutron, emit a beta particle, and form Pu isotopes, or emit an alpha become a lower mass nuclide.

Edit/update: There are also compounds UCl₃ and UF₃ usually mixed with other chlorides or fluorides, which are considered salts.