One of the motivations for USSR's largely graphite-moderated fleet of reactors [likely] is plutonium production.
Wait, is that a thing? I remember that being a thing but I don't actually remember the reason behind it. Well, further reading then.
Pu239 is bred from U238, though I don't remember what spectrum of neutrons is required, and I don't recall offhand if graphite or water give different spectra. In any case, fuel needs to be changed frequently, so that the Pu239 doesn't capture additional neutrons up to the dangerously more fissile Pu244 or so. The heavier isotopes are generally considered unsuitable for bombs, due to the high spontaneous fission (SF) rate -- but I also recall reading that simulations have shown it's possible to make bombs including even heavier actinides e.g. americium, curium, so maybe it's more of a design issue than anything.
Tritium can be extracted from primary loop coolant, particularly of heavy water type reactors.
Plutonium doesn't decay very fast (by itself, when subcritical), so is fine to recycle (melt and recast) for quite a long time. Once in a while it might have to be dissolved and reprocessed, separating lower elements and byproducts. Chemical separation is relatively easy. Over time, the neutron content increases I think (neutrons from SF and short decay chains being absorbed to Pu240, etc.), which may then need to be phased out for new material (since isotopic separation is extremely difficult).
Tim