It seems that oxygen may be extractable from oxides in lunar regolith using "some variant of the" molten oxide electrolysis.
Assuming that solar energy is used in a lunar setting, what would require more power for a modest demonstration-scale implementation; keeping the material molten or performing the electrolysis separating the oxygen from the metal and silicon oxides?
I ask because the heat for melting could come from solar concentrators but for the electrolysis you need to convert the incident solar power to electricity first.