About the SA Blog Network

Plugged In

Plugged In

More than wires - exploring the connections between energy, environment, and our lives
Plugged In HomeAboutContact

Guest Post: Burning Buried Sunshine

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

Oil – The Least Efficient Source of Energy

By Scott McNally

Solar energy is often criticized for its inefficiency – that only about 10% of the sunlight that hits a common commercial solar panel will be converted into electricity. Similar criticisms are voiced against biofuels, which have a solar energy to biofuel conversion efficiency of less than about 2%.* But how do these efficiencies compare to other sources of energy, like oil? Turns out – solar and biofuels do pretty well.

First consider this; we truly have three primary sources of energy: nuclear, geothermal, and solar**. Solar energy is the original source of energy for wind, biomass, fossil fuels, and even hydroelectric. The sun creates temperature gradients on the surface of the Earth, which creates wind. Biomass photosynthesis is powered by solar photons, and sometimes that biomass falls to the ground, gets buried, and cooked into fossil fuels. Hydroelectric dams harness the potential energy of water, but the water has to get “uphill” somehow. This happens when the sun lifts the water through evaporation, which is later released as rain.

These ‘sources’ of energy – wind, biomass, and hydro – are not true sources; they are really just different ways to carry, or convert solar energy. When comparing these types of energies, it is useful to think about how efficiently each type converts solar energy into a useful form of energy. In the cases of wind, hydro and solar, the useable form of energy is usually electricity; with biomass and oil, we get some sort of liquid fuel that can be used to either move your car or generate electricity. For the sake of simplicity, we will say that all of these sources end up as electricity.

Now we know that solar panel conversion efficiency is about 10%, and the biomass solar conversion efficiency is about 2%. As it turns out, that’s pretty good.

Let’s look at fossil fuels, oil in particular. Now, fossil fuels are pretty great. They are very energy dense, and have enabled society to do some amazing things. If you think of fossil fuels as a source of energy, they are efficient. But, fossil fuels are actually a secondary form of energy, and when you consider how conventional fossil fuels are naturally made, the solar conversion efficiency is astonishingly low.

How low? Let’s have a look at how oil is made, and how efficient each step is:

  • First, you need sunlight to grow plants, but photosynthesis is only about 1.7% efficient.
  • Only about 2% of the biomass that grows is actually preserved, and ends up deeper in the Earth.
  • About 74% of the sequestered biomass turns into oil.
  • Only about 2.8% of that oil gets trapped.
  • Only about 25% of the trapped oil is actually recoverable by humans.
  • About 90% of the recovered oil goes to products, since about 10% has to be fueled (usually as natural gas) to run the extraction and refining process.
  • Finally, internal combustion engines are only about 20% efficient.

So if you total that all up, fossil fuels are about 0.000003% efficient at converting sunlight to energy. Put another way: If you want one Watt of energy from solar panels, you need about 10 Watts of sunlight. If you want one Watt of energy from biomass, you need about 50 Watts of sunlight. If you want one Watt of energy from oil, you need over three million Watts of sunlight.

The actual number for fossil fuels is 3.15 X 10^6 Watts of sunlight per watt of fossil fueled energy. For you math lovers, that is close to pi million watts of sunlight.

We can also think about this is terms of land area required. If you want to run your house (about 4 kilowatts) for a year, you need about 35,000 kilowatt-hours of electricity. If you want to get that energy from solar, you will need about 10 acres of solar panels. To get that energy from biofuels, you will need about 50 acres of farmland. But to get that same amount of energy from fossil fuels, you need the sun to shine on a swath of land about the size of Connecticut, for an entire year, for one house.

That means that if all the land in the United States was devoted to producing energy through continuous sustainable fossil fuel production using the sun and the natural conversion process, the entire United States could only power about 600 homes. But unfortunately for those 600 homes, they would have to wait a couple million years before their energy was delivered.

Fortunately, we currently don’t have to rely on the Earth’s natural fossil fuel formation processes. We have an abundance of oil stored in the crust, and if and when that runs out, we have more efficient ways of making synthetic fossil fuels using biomass than the way the Earth does it naturally. So oil is great, for now, but as it turns out, solar and biomass are much more efficient in the long run.

*The advantage of biofuels should be noted though – that the collector assembles itself. You don’t have to build anything, you basically just throw a seed on the ground and your solar collector grows out of thin air. With solar panels, you have to actually build and transport the collector, which is annoying.

**There may be one tiny exception, and that is tidal power, which is powered by the gravitational forces of both the sun and the moon.

Author’s note: The inspiration for this article came from the Fundamentals of Renewable Energy Processes Class at Stanford University, taught by Dr. Adam Brandt. His inspiration for this calculation came from Dr. Jeffery Duke’s publication, Burning Buried Sunshine: Human Consumption of Ancient Solar Energy.

About the Author:

Scott McNally is an energy engineer who has spent the past year working on national energy policy issues in Washington, DC. He has worked as an ORISE Fellow with the Department of Energy’s ARPA-E Program and an energy and climate researcher with the White House Council on Environmental Quality. Previously, Scott was a project engineer for Shell Oil Company and an environmental engineer for Valero. Scott has a B.S. in Chemical Engineering from the University of Texas at Austin, and is currently completing a Masters in Energy Resources Engineering at Stanford University. Scott was invited to be a guest blogger by Plugged In’s Melissa C. Lott. You can reach Scott via e-mail at scottmcnally at gmail dot com.

Photo Credit:

1. Photo of solar panel by Andreas Demmelbauer and used under this Creative Commons License


Melissa C. Lott About the Author: An engineer and researcher who works at the intersection of energy, environment, technology, and policy. Follow on Twitter @mclott.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 9 Comments

Add Comment
  1. 1. sault 11:16 am 02/14/2013

    10% efficiency is the low end for thin-film panels. Crystalline panels have a low-end efficiency of 15% and some are punching over 18%. The world record for efficiency is around 45%, although this setup is not practical for rooftop application. Solar cells in commercial production are getting 24%:

    But I understand that the 10% makes for easier calculations AND keeps the clean energy haters from throwing accusations of bias against you. However, some people might be misled by the 10% figure and think “That’s it? This whole solar power thing is only 10% efficient?”

    And looking at your energy use calculations, I’m looking at my electricity bills and I’m using under 2,400 kWh per year. That’s less than 10% of the usage you’re quoting! So even if I upped my usage by almost 50%, I’d be up to 3,500kWh a year and have an average load of 400W. A 2 kW array would cover close to 100% of my demand, requiring 7 of these panels to do the job:

    The size of the array would be about 11.5 square meters, or A LOT smaller than 10 acres.

    Energy efficiency and conservation will play a big role in the clean energy revolution.

    Link to this
  2. 2. sethdayal 11:34 am 02/14/2013

    As usual Sault is unaware that sun doesn’t shine at night. Most of the time he is relying on the grid fueled by fossil sources to keep him going 24/7 and on cloudy and winter days. As well he is using fossil sources to fuel his transportation and heat/hot water needs.

    With the tiny amount of solar on the grid today its no big deal to cater to his fantasies, but as solar/wind moves to renewable energy standards of 15% he becomes a real problem.

    Link to this
  3. 3. sault 12:33 pm 02/14/2013

    seth, please post any links to the (hopefully) unbiased sources that say 15% renewable energy is some sort of barrier to overcome. I merely calculated what was necessary to cancel out all of my yearly consumption and compare it to the 10 acres of solar panels calculated in this article. Nice strawman approach you got there by accusing me of not knowig it gets dark an night, though! Classy!

    Link to this
  4. 4. jerryd 7:41 pm 02/14/2013

    I’m going with Sualt here as the author’s numbers are way off.

    Just a 100×100′ lot gets 5MWhrs/day in solar energy US average. So even at the stupid low 10% rate means 500kwhrs/day. Facts are most buildings need less energy than falls on their roofs so no extra land it needed.

    Seth is being his biased self lying about solar as always. Why do you try to misinform people Seth?

    Time and time again you put up your rigged numbers while many others are saving and even making money off their solar arrays, show clearly they are wrong. So why do you keep putting the lies up Seth?

    By now even you know they are wrong and rigged even if you didn’t to begin with. Why do you hate them, other RE so much?

    Link to this
  5. 5. sethdayal 1:59 pm 02/15/2013

    Finally a reasoned response from Sault – who knew he was capable?.

    Jerry not so much – as usual. Jerry has told us that he is unable to use the innernet to do research – he claims he just doesn’t get it. That and his job as chief biofuels tester in his family biofuel plant in a hut on stilts in the middle of the everglades explains his bleary posts.

    For Sault we have 25% name plant requirement in Calif with most of its res required for fairly reliable desert solar, and 80% assuming an all wind situation in Australia. German/Rest of US solar/wind would be 100% due to the less reliable wind and solar there.

    I also included work showing that wind/solar saves no GHG’s in a high res situation.

    Google “la-me-unreliable-power-20121210″

    ” Wind and solar energy are called intermittent sources, because the power they produce can suddenly disappear when a cloud bank moves across the Mojave Desert or wind stops blowing through the Tehachapi Mountains. In just half an hour, a thousand megawatts of electricity — the output of a nuclear reactor — can disappear and threaten stability of the grid.”

    “California now maintains a margin of 7% to 8% above projected daily demand, in case a nuclear power plant goes offline or outages occur. But when 33% of the state’s power comes from renewables, that margin will have to rise to 15%, said Stephen Berberich, the firm’s chief executive.”


    “After analyzing the data, the authors stated wind energy cannot be used for base load, and that the installed capacity of required back-up must be at least 80% of IWT installed capacity.”

    ” In Eastern Australia the required back up is OCGTs which are far less efficient than CCGTs. As CCGTs are less quick-reacting than OCGTs, the latter need to operate in synchronous spinning mode 24/7/365 to instantly provide energy when wind energy is ebbing.”

    Google “Hughes-Windpower”
    Hughes effect of RES Why is wind power so expensive?

    ” more wind capacity leads to increased – not reduced – emissions of co2. indeed, the situation is much worse if wind generation displaces nuclear power with minimal co2 emissions.”

    Link to this
  6. 6. sault 3:40 pm 02/15/2013

    From your first article:

    “The state’s electricity system can handle the fluctuations from existing renewable output…The state now gets 20% of its power from renewables.”

    So yeah, the 15% barrier is a myth. Why can’t you just provide the link instead of telling people to “google” gibberish that EVENTUALLY leads them to this article:

    And continuing on in the same article, you conveniently left out this part:

    “On the day last month when wind energy provided just 33 megawatts of power statewide, a brilliant sun spiked solar plant output.”


    “The problem with reserves is an outgrowth of a separate decision by the powerful California Water Quality Control Board. The board last year essentially ordered 19 gas-fired generating plants up and down the coast to close by 2020 because their cooling intake pipes suck in and kill fish.”

    So, it seems like the whole reserves issue is caused by something ENTIRELY DIFFERENT than renewable energy, but you just jump to the conclusion that it does. And the article IN NO WAY provided ANY proof to back up its sensationalist headline…Still Classy…

    And Andrew Miskelly drawing such SWEEPING conclusions about power generation based ON ONE MONTH OF DATA IN SOUTH AUSTRALIA is dubious at best.

    Did you manage to read this article about wind energy in Australia?

    “Australia wind beats new coal in the world’s second-largest coal exporter

    Sydney, 7 February 2013 – Unsubsidised renewable energy is now cheaper than electricity from new-build coal- and gas-fired power stations in Australia, according to new analysis from research firm Bloomberg New Energy Finance.

    This new ranking of Australia’s energy resources is the product of BNEF’s Sydney analysis team, which comprehensively modelled the cost of generating electricity in Australia from different sources. The study shows that electricity can be supplied from a new wind farm at a cost of AUD 80/MWh (USD 83), compared to AUD 143/MWh from a new coal plant or AUD 116/MWh from a new baseload gas plant, including the cost of emissions under the Gillard government’s carbon pricing scheme. However even without a carbon price (the most efficient way to reduce economy-wide emissions) wind energy is 14% cheaper than new coal and 18% cheaper than new gas.

    “The perception that fossil fuels are cheap and renewables are expensive is now out of date”, said Michael Liebreich, chief executive of Bloomberg New Energy Finance. “The fact that wind power is now cheaper than coal and gas in a country with some of the world’s best fossil fuel resources shows that clean energy is a game changer which promises to turn the economics of power systems on its head,” he said.”

    See, this is what’s called a LINK. You don’t have to tell people to “google” stuff when you post them.

    Link to this
  7. 7. sethdayal 3:55 pm 02/15/2013

    I course the Ca prediction of a 25% nameplate fossil backup for solar/wind is nonsense when you consider, every now and then climate events like this one that caused it to rain for 43 days straight.

    Google “California Megaflood: Lessons from a Forgotten Catastrophe”

    In that case Ca’s generation would be all wind bringing it back to the 80% level.

    Then we need to allow for regular climate events when peak load occurs in the summer evening with no wind or solar to speak of. In that event wind and solar require 100% fossil backup.

    Link to this
  8. 8. sethdayal 9:05 pm 02/15/2013

    Once again our resident troll Sault with his laughable claim of a MS and BS in Engineering, spewing more horsepucky keeping up his title as the stupidest commenter on SCIAM.

    Has everybody notice how this stupid fathead spews without ever reading or looking for background first.
    Ca 2011 generation 287 TWh, solar/wind 8.7 TWh. Since the troll is too stupid to do the arithmetic that’s 2.9%.

    More idiocy – if one month of data shows a need for 80% fossil backup, you’d think the halfwit would get than adding more months is only going to make it worse. Whadda twit.

    And as always never doing any research he quotes a greenie news release that doesn’t give any details on where they got the wacky numbers. No indication of when or if ever the study will be available for review and debunking. In fact indications are that the results are based on some way out their assumptions on boring costs for the different options.

    My advice Sault is refrain from commenting ever. Each time you spew you make yourself look even stupider, if that is possible.

    Link to this
  9. 9. yoohsiu 4:46 pm 10/25/2013

    Rerunning the “10 acres of solar panels required” calculation again.

    According to the U.S. Energy Information Association, in 2011, the average homeowner in the U.S. used about 11280 kWh per year.

    Assuming, conservatively, that panels have even just 4 hours of peak (1000kW/m^2) sunlight per day, to generate all the power you need, you would need a 11280kWh/ (4*365) = 7.8kW array, which would be

    - for 20% efficient SunPower eSeries solar panels: 39.4 square meters of panels, or 24 320W solar panels at 1.63 square meters each
    - for, at the low end, 8.1% efficient Sharp NA-V115H1 thin-film solar panels: 95 square meters, or 68 115W solar panels at 1.4 square meters each

    Neither of which approach the 10 acre (40000 square meter) requirement that the author here has set out.

    Estimates of sunlight/day vary depending on where you live. LA has an average of 5.62 hours of sunlight/day, for example, while Seattle averages 3.57 hours/day over the whole year. Obviously there are sunnier places in the U.S. where you’d get more value for your solar installation.

    But the author’s 10 acres calculation is at least 500x too large. Please rerun your calculation…

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article