mgb_phys said:
vanesch is right - even 'free' power costs more than nuclear, unless you have some local customer that can use power whenever it's available.
Well, this is not totally unthinkable of course. For instance, hydrogen production by electrolysis might (I'm no expert) probably be such kind of flexible load, which could be powered by an erratic source.
The point is that we DO have a big load which is not flexible, and which requires adaptation. And it is this part which cannot economically be expected to live on a majority of renewables.
Again, I'm not against renewables. But I think that in the current state of technology, they do not play in the same ballpark as does nuclear, which, in my eyes is the only *realistic* alternative to fossil fuel on a majority basis. As I said, if you plan to do 10% nuclear, you can just as well plan to do 10% wind/solar. If wind/solar replaces nuclear, well, then you've won 10% on fossile, if wind/solar ADDS to nuclear, well, you'll have 20% gain in fossile, and you will be cheaper on average.
You've pretty much installed what you can concerning wind/solar, but you can go up to 80% nuclear if you want to. So the REAL replacement for fossile on LARGE scale, is nuclear.
Now, the (distant) future might be different, but you cannot plan a POLICY on something that hasn't yet been demonstrated.
It still might make business sense though if you can get your customers to pay more for 'green' power than they would pay for nuclear.
Although this might make microeconomic sense, it is stupid on macro-economic scale (you put several times the resources in just for sake of ideology), and I'm also not convinced that it makes objective ecological sense. I'm really not convinced that compact nukes are environmentally less friendly than HUGE installations, of which the impact has not been considered seriously.
It's also worth having a couple of pumped storage schemes to both handle unexpected peaks (superbowl ad breaks) and to provide grid load -balancing and restarting.
Of course, that's why you cannot even go to 100% nuclear either. Nukes are (contrary to what people think maybe) flexible, but not on a minute-scale. That is, they do not need to be static baseload (as was their use in the 70ies-80ies), they can follow the consumption, but their "slew rate" is limited, for security reasons (not technologically: Chernobyl went from 200 MW to 30GW in 7 seconds ... ok, this is bad taste

)
So you need a small "fast responder" capacity to take over the very strong rises and drops in consumption, during the few minutes it takes for the nukes to adapt.
The only "fast" responders on a minute scale are hydro and gas turbines.