[pfdietz]

Cost has been the problem with fission. The first nuclear buildout in the US ended because of market forces, not because of TMI. Costs were higher than promised, a constant problem then and now. Electricity demand growth slowed, and with the passage of PURPA in 1978 non-utility sources started to be added to the grid. Nuclear could not compete with them even then.

Nuclear's long term prospects are poor, for one big reason: it has never shown good experience effects. This means its cost has not come down with time. Any technology like that is doomed, if it has competitors that are improving. PV started off orders of magnitude more expensive than nuclear (levelized cost of energy) but now is much cheaper, because PV improved 20% for each doubling of cumulative production and nuclear didn't.

[danpalmer]

> Nuclear's long term prospects are poor, for one big reason: it has never shown good experience effects.

Good point, I hadn't considered this. I would suggest that a possible cause for this is lack of investment though. We had the 1st generation. Most current reactors are gen-2 or gen-3. Gen 4 hasn't really started use, still in R&D I think. There are ideas for Gen 5 but no progress. Each generation seems to bring moderate energy improvements, and significant cost, safety, and waste management improvements, but the interest just isn't there.

[pfdietz]

I think the problem is not lack of investment, but how nuclear power plants are built. Experience effects come from people doing a task over and over, learning to do it faster and modifying it so it can be done faster and better. But there's little opportunity for that in nuclear construction. Most workers will not work on many NPPs, and the designs are not easily changed.

It might be better if a string of smaller but identical NPPs were built with overlapping construction timelines, so groups of workers could specialize on one part of the construction of each and switch to the next in line to repeat the process. This would be a kind of assembly line without a factory. But this would require construction of many reactors, which means they'd have to be small. Renewables benefit from having huge opportunity for this kind of pipelining, because individual units are so small.