TXU has applied for nuclear.  Given that they have also applied to build 10 coal plants, I am cheering their nuclear efforts on.  Once those nuclear plants are switched on, they will be run, and will displace fossil fueled plant.

I think some US states (like Texas) will prove to be much more nuclear friendly than New England or New York.

http://en.wikipedia.org/wiki/Photovoltaics#Worldwide_installed_photovoltaic_totals

...shows what looks like about 1GWE (peak power) installed wordwide in 2005. That's about 1 reactor at peak, or maybe 30% of a reactor allowing for daily cycles and seasonality.

Are you numbers peak output or annual average?

Note that this source http://www.eia.doe.gov/oiaf/ieo/electricity.html gives world electricity consumption of 14800E9 kWh in 2003. There are about 8000 hours in a year, so let's say a nice round CONSTANT 24/7/365 load of 2000 gigawatts. So by 2020, on your projections, solar PV could be producing 4% of the 2003 average demand (or 2% of 2020 demand on the EIA's estimate). Assuming the solar cells never age, of course. And assuming the supply chain doesn't get bottlenecked somewhere.

(some of the people who post around here who are pro nuke use much lower numbers 4 cents even).

A massive factor is capital cost-- even in a gas fired CCGT, where fuel is half the lifecycle cost, capital costs matter.  Change the real interest rate (interest rate after inflation) and nuclear and wind look a lot worse, or a lot better.

This is why gas turbines took off in the 90s.  Gas was cheap, and capital costs were the most important costs.  There was also an important efficiency shift (the Combined Cycle bit) which raised efficiencies from c. low 40%s to nearly 60% now.  Gas is however no longer cheap.  

Conversely nuclear plants wound up being delayed by several years, and in the high inflation/ high real interest rate environment of the late 70s and 80s, that meant their total costs were several times initial estimates.

If we can assume a world where nuclear plants really do get built on budget, in 5 years, then nuclear will look a lot cheaper but you will still have the waste problem and cost.

Good point about the true cost of coal.  I think most US power coal now comes from the Powder River Basin Wyoming, which has low sulphur coal and high productivity open-pit mining.  One of the big costs is then rail transport, whereas ocean transport of Australian coal is relatively cheap.

I think Appalachian coal is a decreasing fraction of coal consumption except for coking coal and for power stations that are very close to source.

The reality is, ex environmental considerations, any US utility (virtually) would rather build a coal fired station.  The risk is low, the technology proven, and the economics straightforward.

Once the thing is built, and once the debt raised to finance it is paid off or discharged, then the operating costs are massively cheap.

This is especially true for nuclear (pure maintenance and fuel costs probably 1-2 cents/kwhr) and wind (free fuel, minimal maintenance, complete replacement every 20 years).

But is also true of coal (probably the operating costs are c. 2.5-3 cents/ kwhr).

I certainly can afford to pay that much. So can many others."

Unless they ban coal fired generation or impose carbon taxes, the voter has nothing to do with it.  It boils down to a straight economic evaluation.  Utilities have to install the most cost effective option, which will usually be coal.

Also, overwhelmingly, people will not willingly pay a single penny more per kwh, if they have a cheaper (dirtier) option.  Did you notice how poorly the green energy companies did trying to sell their higher priced, cleaner electricity?

I am a big fan of a carbon tax, but I see no chance of it happening in the US.