Running the U.S. on Solar Power
How much land would it take for solar power to satisfy the electricity demands of the U.S.? I made some attempts to calculate this before, but a recent story may enable me to calculate some more reliable numbers if the solar is provided via solar thermal power:
Two bits caught my eye:
Abengoa Solar, a subsidiary of a similarly named technology company based in Seville, Spain, and Arizona Public Service on Thursday announced plans to build a 280-megawatt solar thermal power plant about 70 miles southwest of Phoenix.
So we know the planned capacity of the solar thermal plant. In case you are unfamiliar with solar thermal:
Solano will use parabolic mirrors to follow the sun across the sky and concentrate its energy, heating a fluid to 700 degrees Fahrenheit, and using the fluid to make steam that will spin turbines to generate electricity. The plant will use an unspecified heat storage technology so the plant can continue generating electricity for six hours after sunset.
So, how much area to produce 280-megawatts?
The project will bring economic benefits, too. During three years of construction, it will employ 1,500 workers at the 1,900-acre site near Gila Bend. After completion, 80 permanent employees will work at Solano.
OK, let me say before running through this calculation that I have no idea how it is going to turn out. And if someone spots an error in math or logic, please bring it to my attention. I am going to scale this up to produce all current U.S. electricity demands.
Peak U.S. demand, according to the EIA, is almost 800,000 megawatts. Actual available capacity is 900,000 megawatts. So let’s make our solar capacity equal to today’s total installed electrical generating capacity.
Assuming the entire 1,900 acres is needed for the plant (maybe not a good assumption, but all I have), then this breaks down to (280 megawatts)/(1,900 acres), or 0.147 megawatts per acre. This of course includes all of the land associated with support functions, and it may include area for future expansions. So the calculation may be conservative.
The second assumption is that the areas in which will put our solar plants will be as productive as this one in Arizona. That is not a conservative assumption, and will somewhat offset the previous conservative calculations.
Then to get 900,000 megawatts is going to take (900,000 megawatts)/(0.147 megawatts per acre), or 6.1 million acres. How large of an area is this? I don’t know. I have to get out my calculator.
My calculator indicates that 6.1 million acres is an area of 9,531 square miles, which is equivalent to a square of just under 100 miles by 100 miles (which would be 10,000 square miles). That’s a large area, to be sure. But the possibility is there.
A couple of caveats. First, this calculation does not make a provision for a mass migration to electric transport. That would clearly require (a lot) more power. On the other hand, we already have a lot of installed electrical capacity in the form of hydroelectric (78,000 megawatts), other renewables (24,000 megawatts), and nuclear power (100,000 megawatts). This lessens the power requirement from solar.
How does this compare with my previous calculation for solar PV? I don’t know. Let me check.
OK, I checked. Not too bad. In A Solar Thought Experiment, I had assumed a slightly lower power requirement and only included the actual area of the solar cells. I came up with an area of about 50 miles by 50 miles of PV panel surface area. So it was in the ballpark. The 100 by 100 mile number is probably more realistic (and is for solar thermal – a different animal), given the need for the real estate for supporting infrastructure.
Other conclusions from the previous essay remain the same. For solar PV, there are around 100 million houses in the U.S., so there is quite a bit of surface area readily available, right where the power is needed. Your results will obviously vary depending on whether you live in Maine or Nevada. The cost is still a staggering $6 trillion. However, to put that number in perspective, at $100/bbl, the U.S. would spend $6 trillion on oil in less than 8 years.
What is the limiting factor? Are there particular components that are critical, but not available in large enough quantities to make this work? Possibly, but I don’t know what those might be. I actually believe that this could be our Manhattan Project, and it could be done. But it doesn’t even have to offset all of our current electrical capacity. We just need to start chipping away, and substituting solar in place of fossil fuels and new capacity that is needed.
Can we afford it? The key question to me is, “Can we afford not to try?”
2015 EIA Energy Conference
June 15-16, 2015 - Washington, D.C.
Platts North American Crude Oil Summit
February 26-27, 2015 - Houston, TX