A Solar Array on Every Roof, and a Chicken in Every Pot
How large of an undertaking would it be to replace US residential power consumption with solar?
I do most of my window shopping for solar at Wholesale Solar, which has excellent, if volatile, pricing for all the components you need to build a grid-tied or off-grid solar installation.
The average US home uses 940 kWh of electricity per month, $103 at the national average rate of $0.11/kWh. The size of the solar installation needed for 100% grid offset varies based upon a number of factors including longitude, installation angle, cloud coverage. And of course, solar only produces power when the sun is shining. A typical grid-tied setup produces more power than the house needs when the sun shines and pushes the excess power back to the grid for storage or third-party consumption. At night the house will draw from the grid.
NREL published a study recently on this topic titled The Regional Per-Capita Solar Electric Footprint for the United States.
An example 7.8 kW gridtie system (US-manufacturered Helios panels, Fronius inverter) costs $13k, perhaps $18k installed. 30 panels in a 15 x 2 arrangement would occupy a 11 ft x 49 ft section of a south-facing roof when installed. Wholesale Solar says this system produces up to 1054 kWh per month. This is somewhat over-optimistic for a typical US home – probably 1/3 of homes are sited poorly for solar (shaded for the best part of the day or no major south-facing roofing surfaces). For the remaining 2/3 this would probably generate 60-80% of their year-round energy consumption.
The 2010 US census recorded 132 million homes. I’ll assume below that all 132M houses are well-sited.
132M * $18k = $2.4 trillion dollars. A princely sum, but close to the $2.2 trillion we’ve already spent on war in Iraq and Afghanistan and less than the projected total cost of $4 trillion.
132M * 70% * $103/mo * 12 mo/year = $114 billion/year saved, not far off the claim. That’s about a 20 year payback without any consumer subsidy. Pretty sweet, considering the panels are typically guaranteed for 25 years.
A good friend responds:
If we tried to produce solar panels in sufficient quantity to do this, would the price go down as a result of economy of scale, or are there scarce commodities used in the manufacture that would cause the price to skyrocket?
First, understand that we’re talking about 0.75 TWp (photovoltaic terawatt) in this scenario (7.8 kW x 132M x 2/3 x 110% for breakage, inventory, servicing). Worldwide production is around 60 GWp per year currently .. or about 12 years consuming the entire global production capacity.
The National Renewable Energy Lag (NREL) answers a similar question in a FAQ (PDF) with a larger scope: can ALL energy used in the world be produced with solar? Note that this includes transportation energy, industrial energy, etc. The figure in question now is not ~0.75 TWp but 75 TWp.
What’s the bottom line—will we really have enough materials to produce 75 TWp of PV?
The silicon-based materials technologies are all unconstrained by feedstock supplies and could individually produce 75 TWp of clean energy. CIS and CdTe are constrained, but could still contribute significantly to the goal for 2065. Although establishing a realistic limit on the use of rare materials will require further analysis, new technologies that employ other materials or use rare materials more efficiently are likely to be developed, and this could also expand the potential contribution of PV to the global energy system. A mix of PV technologies should be able to meet, or indeed exceed, the “TW challenge.” In doing so, PV would provide a uniquely attractive contribution to the world’s economic growth, environmental sustainability, and energy security.
What does 75 TWp really mean?
The idea of generating 75 TWp of electricity from PV is unprecedented. Production on this level would not be “business as usual.” Entire industries would be created or revolutionized.
Using $1/Wp as a round number, 75 TWp of PV would mean $4 trillion of annual revenue in 2065, or 10% of the world’s gross domestic product (GDP) today. This seems like a large number, but if the world GDP grows 3% annually between now and 2065, $4 trillion would only represent 2% of GDP. Additional energy will be needed to meet expanding energy needs, whether or not it comes from PV. Trillions of dollars will be spent on infrastructure and fuel, no matter how we obtain our energy.
Some 75 TWp would require about 500,000 square kilometers of 15%-efficient modules. In terms of land area, we would need 2.5 times that area for PV installations, or a square 1,120 km on a side. This is about 0.8% of Earth’s land surface. Although large, it is about the same, on a percentage basis, as the amount we already use for national defense. It should certainly be acceptable for such a complete transformation of our energy and environmental infrastructure. For more on land space issues, please see How much land will PV need to supply our electricity? in this PV FAQ series.
More realistically, energy inputs in the future will come from a combination of fossil fuels, nuclear, wind, hydro, and solar much as today, just hopefully with more renewables and less fossil fuels.
Here’s some “light bedtime reading” discussing an energy grid design using hydro (baseload but virtually tapped out), solar, wind, and EV grid storage), my preference.