Hi Robin,
You said: According to
http://atom.kaeri.re.kr/cgi-bin/nuclide?nuc=H-2&n=2
the (n,2n) cross section for 14 MeV neutrons on D is 177 mb
That sounds about right - but this is very low - and if memory
serves this is actually about the same as the capture
cross-section of oxygen for thermals - and you know how rare that
is - plus since you get many more collisions at thermal energy -
you will actually loose neutrons to the oxygen of heavy water
faster than you can make extra neutrons via (n,2n)
IOW we should get about 1.5 thermal neutrons out for every fast
neutron going in, and nearly all of the thermal neutrons are
going
to produce T. This should still be a breeder.
Taking the ratio of the two - as you did is just a ballpark calc -
and not the correct way to figure it - but I will have to get the
experimental data out of some hard copy files tomorrow - suffice
it to say that the best TBR you can get with current technology
using beryllium and highly enriched 6Li is about 1.3 in real
tests - and the actual neutron (n,2n) ratio for any reasonable
amount of heavy water is slightly underunity - for the reasons
mentioned above.
Millibarn cross sections allow for much unpredictability, in
general, and it is tough to build a breeder on that kind of
rarity - you could actually get a full neutron decay before a
(n,2n) multiplication in "just" heavy water - with no lithium.
BTW - almost forgot - you CANNOT use dissolved lithium hydroxide
in the blanket at all - as the T will preferentially displace a D
in D2O and you will loose all of your precious tritium to
tritiated heavy water - and it becomes such a small percentage
that it is "effectively" lost.
You must capture the bred T as a gas - from a porous metal,
preferably and cannot let it even get close to water of any kind -
as it immediately displaces a lighter hydrogen and is effectively
lost.
Jones
