[The need for stable power sounds like the sort of problem that could be engineered around. Maybe a modular solution would allow cells to be brought in and out of operation as energy availability changes, with some storage to smooth out the higher-frequency fluctuations?
I think the need for stable power is that, like blast furnaces, these plants are designed to remain hot. Blast furnaces take days/weeks to heat up and then remain hot for years. A blast furnace that cools down is probably rebuilt before heating up again.
Yep, pretty much once you start up a blast furnace you keep it running until it's end of life. This is why when one of the steel works had basically gone bust, it was critical it be rescued before they had to turn the furnaces off, else it basically destroyed most of the value of the furnace. The key difference between a blast furnace and the electrolysis cell, is you can stock pile a months work of coal on site. That's harder with electrons, where a gust of wind can be enough to switch off your supply, and there's no realistic stock pile...
J
