[rcxdude]

That paper was analysing the system with a temperature difference of 40C (a temperature difference which doesn't allow you to extract energy efficiently anyway). Nuclear power plants at minimum produce steam at about 300C (all the way up to 900C). It doesn't seem like it the paper would particularly support its use for nuclear power. (Basically all plausible uses for this involve using low-grade power. But any system which attempts to do this needs to be really, really cheap compared to conventional systems to be worthwhile).