[vagabund]

Here's my understanding:

The main electricity-generating fusion reaction is aneutronic -- 3He-D fusion yields regular hydrogen and a proton. But a fuel mixture of 3He+D will inevitably produce some D-D fusion, which yields 50% T+p (tritium and a proton) and 50% of He3 + neutron. The neutrons produced are slower than those produced in D-T fusion (2.45MeV vs 14.06MeV), and are similar to those produced by radiation therapy medical devices.

Since they need the products of D-D fusion anyways (tritium decays into 3He in 12.5 years) they can either run D-rich fuel in the main reactor and extract the yield for future fuel inputs or, if the slow neutrons of D-D fusion still prove too damaging to the device, they can lower the Deuterium ratio in the main reactor's fuel, and designate separate reactors to do the D-D fusion exclusively. That way they can concentrate the cost of neutron damage mostly to one fuel producing reactor, and not have to replace every electricity-generating reactor as often.