I'm all for solar PV and I get the scale argument, but we need to think a bit more about this. To supply current electrical needs in the US, solar would need to cover 10,000 sq MILES of area with high solar flux (say Arizona John McCain...) In general, the US has the least favorable solar subsidies of any country which has serious deployments, and the states like CA are picking up some of the slack. But 'grid parity' in power generation is still at best 5 years away, new amorphous silicon plants in China are poisoning their workers and large swaths of the environment, and still we are doing NOTHING about nighttime storage and long distance transmission infrastructure.

A comprehensive renewable program would include a combination of mandates, feed in tarrifs, and infrastructure support combined with competitive R&D packages which were technology independant. Add a carbon tax and it will happen even faster! Phase out the subsidies and phase in the carbon taxes over a 20 year ROI life and you might make it economic for companies to make business decisions to invest in renewables. But seriously, papering the AZ desert with solar cells will take many decades, and in the end, that would only address 20% or so of our energy needs. Unless we convert all transportation, heating, and hot process manufacturing to electricity, which would take 50,000 sq miles BTW, then we still have a global warming problem.

I find it curious that they are using photovoltaics for this.   I would have thought thermal solar a much better bet, particularly since it is amenable to being retrofitted with short term thermal storage to give a flatter production curve.

db-

We could find the 50,000 sq miles just by covering the roof of every Wal-Mart in America with pv cells, while eliminating the long-distance transmission problem, as those things are everywhere! Well, maybe not enough Wal-marts, but there are plenty of warehouses, distribution centers, and other large box stores that have prime real-estate for PV sitting over their heads. A subsidy/tax-break for installing pv (with wind turbines in the parking lots?), plus the opportunity for businesses to advertise as good corporate citizens, could equal a win.

Some e perspective on that 50,000 square mile figure is in order: it's less than 7% of the combined area of Texas, Colorado, New Mexico, Arizona and Nevada - which contain the best solar real estate in the country, overall.  By way of comparison, 65% of Iowa alone planted to corn and soybeans this summer, and yet 3 million people found the state a very comfortable place to live, notwithstanding.

Or, perhaps more telling, 50,000 square miles is roughly equal to 1/3 of national corn acreage, and thus only about 50% more than  the land we're already devoting nationally to biofuel production.  The payoff for that acreage is less than 5% of our liquid transportation fuel use.  If we could get anywhere near our national energy budget for a like use of land for solar production, it would the the bargain of the century.

A different kind of big solar, reproducing photosynthesis to produce hydrogen for fuel.  From 'down under' at Monash U scientists can mimic photosynthesis and instead of producing carbohydrates as a plant does, produce hydrogen:

www.eurekalert.org/.../mu-mtl081408.php

Hi Jet, that is cool. I had seen the recent claim from MIT for the same thing but it is really just hydrolysis. But the sq mi problem exists here too. My point isn't that that is a lot of sq miles (tho it is), my point is that it is a lot of square miles to cover with a manufactured substance. Both for the industry to scale to that point and for us to install and maintain. Planting doesn't require nearly as much effort since the plants are self-assembling. When solar power today already has more cost in the balance of system than in the solar cells, and it is still not at grid parity, this presents some issues. And note, these are sunny sq miles, not rooftops in Seattle for instance.

Speaking of rooftops, you are off by orders of magnitude. According to Wikipedia, we are talking about (generaously) 650m sq ft of walmart stores. Assuming all single story then we are talking about 28 sq miles of rooftop. Total in the US. Another way to look at it is the DOE solar initiative: 1m rooftops. Which they claim will save the equivalent of 800,000 cars emissions. To get to 100m cars, we would need 125M rooftops. Another stat people throw out is all the roads in the US. However, DOT estimates we have in the neighborhood of 4M miles of roads, at an average of 45 ft width still only equals 34k sq miles. Not bad, but a solar cell is a LOT harder to make than a road.

My point in all this is the scale problem is massive. Traditional sources have had 100+ years to develop the infrastructure for the scale we have today, we only have decades to do the same in renewables.

Now for some idle thinking, which I don't know the answer for: Many renewables (biofuels, solar PV or thermal, algae ponds, etc) all implicitely hit the sq mile problem. Some are much less efficient that solar PV (e.g. corn ethanol). Perhaps we really need to look at renewables that can gather energy from an area substantially larger than the installation. Wind and wave come to mind. It would be interesting to compare the amount of material v. the energy harvested for various technologies.

"Not bad, but a solar cell is a LOT harder to make than a road."

An interesting assertion, but not an obvious one.  Solar cells, or thermal solar ( which I think is potentially more efficient, and longer lasting or at least more repairable), are inherently amenable to mass production; roads much less so.

I don't dispute that building 50,000 square miles of solar facilities is a massive undertaking (and I certainly don't think we need a completely solar economy), but if your target were to have a totally carbon-neutral energy infrastructure in 30 years, I strongly doubt that it is beyond achieving at least half of this with pure solar of one form or another.

SD, I'm actually a big fan of thermal solar. IT is still a lot of shiny metal, but I don't think we are going to run out of aluminum. Infrastructure is still an issue, CSP only works were you can focus the sun (no cloudy days) so we'll need a transport infrastructure. And don't get me wrong, solar will be part of the mix and we need to come up with a way. I just get concerned when people underestimate the scale of thee problem so badly. Blindly assuming it can be done doesn't mean it can. To get a sense, one power plant on Colorado's front range burns 100 railroad cars of coal PER DAY. Thats an enormous amount of energy to replace with solar flux capture. Half our energy in 50 years would be pretty optismistic I think, but there are a lot of analyses which give various results. Here's one particularly skeptical one: www.abelard.org/.../replacing_fossil_fuels.php