How much land would be needed to power the world with solar energy?

Harvard University has announced that its massive endowment fund is divesting from fossil fuels. Many other universities had already done so under pressure from their students and alumni and others are likely to follow.

Academic endowments are entering a new normal after Harvard University, the richest school in the world, said it would divest from fossil fuels.

The decision wasn’t made lightly. The nearly $42 billion endowment succumbed to years of pressure from students and climate activists, a massive protest at a 2019 football game, and a string of legal efforts. President Lawrence Bacow in the past has said the endowment shouldn’t be used for political ends. Earlier this month he changed his tune.

A cascade of similar announcements has followed in Harvard’s wake, with Boston University, the University of Minnesota and the $8 billion MacArthur Foundation pulling the plug on fossil fuels. And there are more to come.

“We’re going to see this ripple out in the coming months,” said Richard Brooks, climate finance director at the nonprofit “The financial arguments have never been stronger, with declining demand for oil, gas and coal. The social acceptability has now shifted as well.”

Divestment activists now have turned their focus to Yale, the University of Pennsylvania, Princeton, Boston College and the University of Wisconsin-Madison, none of which have fully abandoned fossil fuels.

An interesting question is how many solar panels would be necessary to meet all our energy requirement and how much land would that require. The answer to that question is not a simple case of taking the total energy consumption, dividing by the amount of energy produced by a panel, and multiplying by the area of the panel. That back of the envelope calculation would give us about 8 million acres for the US alone.

There are many complicating factors involved, and a VFX artist uses his skills to explain the problem and get a reasonable estimate.

He concludes that to power the world, we would need an amount of land the size of France. Of course, these estimates will change with improvements in solar cell technology and changes in energy use.

One thing I learned from the video is something called a Sankey diagram, which is a type of flow chart in which the width of the arrows is proportional to the flow rate, providing an easy-to-see visualization of how much energy comes from what source and where it goes. I had seen some diagrams before but had not known that they had their own name.

Here is one that shows the state of energy usage in the US as of 2019. There is a lot of rich information in that single diagram. (The ‘quad’ as a unit of energy is defined here.)

While only about 1% of the energy use in 2019 comes from solar, it is growing.

In 2019, about 37% (37 quads) of the 100.2-quad total was used for power generation, markedly lower than the 38.2 quads used in 2018. In 2012, by comparison, the generation sector consumed about 40% or 38.1 quads of total energy.

Continuing a trend first observed in 2018, in 2019, the largest increases in energy supply came from natural gas, wind and solar energy, with jumps of 4%, 10% and 8%, respectively. “Solar energy supply exceeded 1 quad for the first time ever as more users turned to renewables. Wind and solar combined now produce more electricity than hydroelectric power, which dominated renewable energy for decades,” the lab noted.

“Solar and wind continue to show year-on-year growth, which is an impressive change for the energy system,” said A.J. Simon, associate program leader for Water Security and Technologies at LLNL. “Other energy resources, such as hydropower, nuclear energy and geothermal have trended flat over the past decade, despite visible annual fluctuations.”

We see that an astounding 67.5% of energy is ‘rejected’, i.e., is dissipated as heat. Some loss of energy as heat is inevitable and in general the more energy that is produced and used, the more that will be rejected. While some of the energy is rejected in the course of power generation, most of it occurs during the end use. But here too there is some improvement.

Last year, 67.5 quads was rejected—1 quad less than in 2018 (which is about the same amount of solar energy consumed in 2019). That bucks a trend in which over the past five years, rejected energy has grown in tandem with increased energy use. In 2014, for example, total rejected energy stood at 59.4 quads.

The 2019 drop in rejected energy could be partly contributed to the shift from coal to more efficient gas generation. Rejected energy from power generation stood at 25.8 quads in 2014; it fell to 24.9 quads in 2016, 24.7 quads in 2017, but grew again to 25.3 quads in 2018. In 2019, it stood at 24.2 quads.

The bulk of rejected energy, however, typically takes the form of waste heat, “such as the warm exhaust from automobiles and furnaces. The efficiency of the nation’s cars, lightbulbs and factories determines how much waste heat is produced and how much fuel and electricity can be put to productive use,” LLNL said.

The ultimate goal would be to have 100% of our energy needs met by renewable sources. Solar energy costs are dropping rapidly and that might be a bigger danger to the fossil fuel industry than the divestment movements.


  1. brucegee1962 says

    I have always been a bit skeptical about divestment as a climate strategy. As long as the world has an inexhaustible supply of rich a*holes (which it does), it seems to me that responsible institutions like Harvard divesting just means that some of the a*holes get to pick up their stock at a bargain price and make a bit of extra profit. Sure, if a substantial part of the market wants to steer clear of planet-destroying stocks that will drive their market price down a little, but does having their stock price go down really hurt these companies substantially? They can just buy some of it back or whatever.
    Even if divestment does hurt these companies, it seems to me that climate groups might put their limited resources into more effective campaigns. I’m supporting’s campaign to get the Federal Reserve to cut off fossil fuel financing:

  2. Venkataraman Amarnath says

    Only an economist talks about cost of energy in terms of dollars that has no real value. The true cost of an energy source is the amount of energy it takes to harness it (as Dr. Tim Morgan puts it, ECoE). Before 1970 it was less than 2% for petroleum and America was great. It has been raising steadily and stands around 8% that will not make America great again EVER. The ECoE of renewable energy is rarely less than 20%. Besides, you still need fossil fuels to construct solar panels, wind turbines, back-up batteries. The only real option is divest from the American way of life.

  3. says

    The true cost of an energy source is the amount of energy it takes to harness it (as Dr. Tim Morgan puts it, ECoE). Before 1970 it was less than 2% for petroleum and America was great. It has been raising steadily and stands around 8% that will not make America great again EVER. The ECoE of renewable energy is rarely less than 20%. Besides, you still need fossil fuels to construct solar panels, wind turbines, back-up batteries.

    Pfft. There’s not one way measuring the cost of energy. You could also measure it in the amount of labor hours it takes to harness it, or the number of human deaths per gigajoule or any number of other things.

    Petroleum is killing people, and it was killing people, as was coal. The rates of death and morbidity associated with extraction, collection, and use of all forms of energy today are much less than they were in the past, although fossil fuels continue to inflict more morbidity and death than other energy forms.

    In the meantime, who the fuck says that America was great before renewable energy tech became a priority. We literally had law enforcement murdering people to prevent efforts to register citizens to vote because of the color of their skin. America was great back then?

    fuck that shit.

    If there’s more than one way to measure the cost of energy, there are even more than that to measure “greatness”, and whatever you’re definition of US greatness you’re using, I reject it utterly.

    The only real option is divest from the American way of life.

    While I agree with this for certain definitions of “the American way of life” I don’t know how you could possibly assert this with any consistency. You’re the one who believes that the US was great in the past, presumably that way of life was an American way of life and it made the US great, in your mind.

    Besides, you still need fossil fuels to construct solar panels, wind turbines, back-up batteries.

    Pffft. No. You don’t “need” fossil fuels for this. Even if you need petrochemicals, these can be synthesized through energy intensive processes. The more we install clean capacity, the more we have the option to produce new solar panels etc. using the energy from existing ones without digging any fossil fuel out of the ground.

    I think it’s probably (though I don’t know for sure) still more efficient at this point to use fossil fuels to create these things if measuring “efficiency” as the amount of time it will take to get to carbon neutrality. It might even be “efficient” in terms of the amount of total CO2 it will take to build the infrastructure required to get to carbon neutrality. But there’s a difference between the statement

    “This is the most efficient path to our current goals, given that certain infrastructure already exists and the capital costs for those things are already sunk costs,”

    “This is necessary to reach our current goals.”

    In short, I strongly disagree with you.

  4. says


    You’re focused entirely on “hurting” bad companies and not at all at the difference a large investment fund can make by “rewarding” good companies. The divestment proposals don’t argue that Harvard’s capital should sit idle once certain stocks have been sold. The proposals argue for productive use of that capital in reforming the entire energy sector. 40 billion isn’t nearly enough to do complete that job, but it sure does have the power to make a difference, and it’s 40 billion better than nothing.

  5. garnetstar says

    Kind of off topic, I got solar panels on my house, and wow are they a good deal. For one, I haven’t had an electric bill since (and I live in the north, with winter snow on the roof.)

    They cost far less than half the price of a car, with financing that’s much better than cars. You get to take 1/3 of the cost off your federal taxes, and my state sends me kickbacks, a couple hundred dollars, now and then. I heat and cool my house with electric heat pumps, have an electric stove, and now I just have to get an electric water heater, and I’m zero fossil fuels, for almost no cost! I’m going to get a backup battery for the house, too, they are good enough and priced well enough now.

    Just to say, solar is a pretty good deal, at least individually.

  6. Mano Singham says


    One of my colleagues in Cleveland (where there is lots of winter snow) put solar panels on his roof many, many years ago soon after it became possible to do so, and he says that it paid off handsomely.

  7. raven says

    There’s not one way measuring the cost of energy.

    What is our planet earth worth? If we wreck it with fossil fuels producing green house gases like CO2, which causes global warming, and the sea level rises 3 feet, how much can we add to the costs of…fossil fuel sources?

    In 2011 it was estimated that up to 187 million people could be forced to flee their homes due to sea level rise by 2100. Contemporary figures now place that at 630 million people.Sep 30, 2020

    The impact of rising sea levels on mass migration -- The World …

    Can we charge the fossil fuel companies for relocation costs for 630 million people by 2100?

  8. tuatara says

    We dont need to fill empty land the size of France. There are millions of buildings that are ready-made platforms for PV installations. Here in Australia approximately 25% of all homes have a rooftop PV array with almost all of them being grid-connect embedded generators.
    World Nuclear Industry Status Report 2021

    The annual Levelized Cost of Energy (LCOE) analysis for the U.S. last updated by Lazard,
    one of the oldest banks in the world, in October 20201241, suggests that unsubsidized average
    electricity generating costs declined between 2015 and 2020 in the case of solar PV (crystalline,
    utility-scale) from US$64 to US$37 per MWh, and for onshore wind from US$55 to US$40 per
    MWh, while nuclear power costs went up from US$117 to US$163 per MWh. Over the past five
    years alone, the LCOE of nuclear electricity has risen by 39 percent, while renewables have now
    become the cheapest of any type of power generation

    My own PV system has generated 21,252kWh since it was first commissioned on December 4, 2018. The vast majority of that production has been exported into the grid with most of that export used by my immediate neighbours (those without their own embedded generators at least).

    The calculated embodied energy of my polysilicon array from mineral extraction to manufacture including distribution, plus the inverter embodied energy, plus cabling, switches, etc, is approximately 3,700kWh/kW. I have a 4.8kW array. That is a total embodied energy of 17,760kWh.

    If I include the rejected energy of the coal generation used to supply my local grid I need to add 3kWh of energy burned for every 1kWh of power offset by my array. That makes my array net positive to the tune of 60,000kWh!

    That 17,760kWh of embodied energy in my PV system is repaid within a year. With a projected lifespan of 25 -- 30 years I think it is a good investment.

    Even if the system needed curtailing during times of low demand (frequency shift appears to be gaining ground as the preferred PV inverter rapid shutdown method) it would still be a positive investment. Our local network operators are working on other methods -- Volt-VAR and Volt-watt power response is now mandatory for all new grid-connect PV generators, and a demand-response (DRM) functionality is also mandatory even though not yet implemented. I haven’t looked into the DRM in depth but believe it will operate in a similar fashion to a ripple control to rapidly shut down or ramp up PV production as required. The grid is therefore coming along for the ride.

    My two concerns however are risks of manufacture (PV cells use some nasty chemicals in their manufacture) and disposal at EOL (hopefully into a mature recycling market).
    But I would rather be battling these than trying to figure out how to dispose of millions of litres of radioactive sulphuric acid (a byproduct of uranium mining) or secure disposal of nuclear waste. Check out the documentary called “Into Eternity” directed by Michael Madsen.

  9. alanuk says

    So much is revealed by the Sankey diagram (they teach these in schools in England, seldom seen elsewhere) yet so much is hidden. Apples and oranges in, pears and bananas out.

    If you subtract the fossil fuel from the input, you are left with 20 Quads (horrible unit). How are you going to run America on that? I suspect that all of the cheap hydro is already being used and some of the nuclear is coming from reactors reaching their end of life.

    When electricity is produced from thermal sources: fossil fuels, biomass, and nuclear, it is not surprising that twice as much goes to waste as the amount converted into electricity -- the laws of thermodynamics dictate it and there is only so much that can be done to improve it owing to the limitations of the materials used.

    Compared to the UK, rather a lot of electricity goes for residential use compared to natural gas. I presume this is due to the use of air conditioning. Again, the laws of thermodynamics prevail; removing heat from air and dumping the heat outside is an inefficient process and, of course, the dumped heat is itself wasted energy. Double plus ungood!

    None of this matters (apart from asthma and pneumosilicosis) if it was not for the CO2 produced. As can be seen, burning fossil fuel to produce (nice clean) electricity actually produces more CO2 from the wasted output than from the useful output.

    It would be good to see the Lawrence Livermore National Laboratory’s Sankey diagrams for 2029, 2039 etc.

  10. GerrardOfTitanServer says

    The thing is, solar and wind won’t work. They can’t work. The intermittency is currently an unsolvable problem. At least nuclear can work.
    Also, nuclear waste isn’t as bad as you think it is. Even if it leaks, the odds of anyone getting harmed is miniscule. People exaggerated its effects by a million fold. Get your information from reliable sources instead of liars like Green sources.

  11. John Morales says

    Gerrard, I was expecting your comment, though its form is a tad surprising.

    The thing is, solar and wind won’t work. They can’t work. The intermittency is currently an unsolvable problem.

    Why you disbelieve those claims is beyond me.
    I myself have a 6.6kW system (5kW inverter) and send well over 4 units of energy into the grid for every unit I consume myself. Also get zero electricity bills (actually, I’m in credit).

    So, solar and wind evidently do work — they generate power. This is indisputable.
    Intermittency is a solved problem ( scientifically and technologically, just not yet implemented at a sufficiently large scale.

    Again, to pre-empt you, I’m all for nuclear power. It works.
    But just like grid-scale, it’s a solved problem scientifically and technologically, just not yet implemented at a sufficiently large scale.

    Thing is, unlike you, I don’t think those two things are mutually exclusive.

  12. KG says

    John Morales@11,

    As I’m sure you know, GerrardOfConspiracyTheories has to believe that “solar and wind won’t work”. If they will, it means that he has been wrong, and his hatred of greens is unjustified. That’s simply intolerable.

Leave a Reply

Your email address will not be published. Required fields are marked *