Monday, March 13, 2017

Why EROEI matters: the role of net energy in the survival of civilization

The image above was shown by Charlie Hall in a recent presentation that he gave in Princeton. It seems logic that the more net energy is available for a civilization, the more that civilization can do, say, build cathedrals, create art, explore space, and more. But what's needed, exactly, for a civilization to exist? Maybe very high values of the EROEI (energy return on energy invested) are not necessary.

A lively debate is ongoing on what should be the minimum energy return for energy invested (EROEI) in order to sustain a civilization. Clearly, one always wants the best returns for one's investments. And, of course, investing in something that provides a return smaller than the investment is a bad idea. So, a civilization grows and prosper on the net energy it receives, that is the energy produced minus the energy required to sustain production. The question is whether the transition from fossil fuels to renewables could provide enough energy to keep civilization alive in a form not too different from the present one.

It is often said that the prosperity of our society is the result of the high EROEI of crude oil as it was in mid 20th century. Values as high as 100 are often cited, but these are probably widely off the mark. The data reported in a 2014 study by Dave Murphy indicate that the average EROEI of crude oil worldwide could have been around 35 in the past, declining to around 20 at present. Dale et al. estimate (2011) that the average EROEI of crude oil could have been, at most, around 45 in the 1960s Data for the US production indicate an EROEI around 20 in the 1950s; down to about 10 today.

We see that the EROEI of oil is not easy to estimate but we can say at least two things: 1) our civilization was built on an energy source with an EROEI around 30-40. 2) the EROEI of oil has been going down, owing to the depletion of the most profitable (high EROEI) wells. Today, we may be producing crude oil at EROEIs between 10 and 20 on the average, and the net energy yield keeps going down.

Let's move to renewables. Here, the debate often becomes dominated by emotional or political factors that seem to bring people to try to disparage renewables as much as possible. Some evidently wrong assessments, for instance, claim EROEIs smaller than one for the most promising renewable technology, photovoltaics (PV). In other cases, the game consists in enlarging the boundaries of the calculation, adding costs not directly related to the exploitation of the resource. That's why we should compare what's comparable; that is, use the same rules for evaluating the EROEI of fossil fuels and of renewable energy. If we do that, we find that, for instance, photovoltaics has an EROEI around 10. Wind energy does better than that, with an average EROEI around 20. Not bad, but not as large as crude oil in the good old days.

Now, for the mother of all questions: on the basis of these data, can renewables replace the increasing energy expensive oil and sustain civilization? Here, we venture into a difficult field: what do we mean exactly as a "civilization"? What kind of civilization? Could it build cathedrals? Would it include driving SUVs? How about plane trips to Hawaii?

Here, some people are very pessimistic and not just about SUVs and plane trips. On the basis of the fact that the EROEI of renewables is smaller than that of crude oil, considering also the expense of the infrastructure needed to adapt our society to the kind of energy produced by renewables, they conclude that "renewables cannot sustain a civilization that can sustain renewables." (a little like Groucho Marx's joke, "I wouldn't want to belong to a club that accepts people like me as members.").

Maybe, but I beg to differ. Let me explain with an example. Suppose, just for the sake of argument, that the energy source that powers society has an EROEI equal to 2. You would think that this is an abysmally low value and that it couldn't support anything more than a society of mountain shepherds or not even that. But think about what an EROEI of 2 implies: for each energy producing plant in operation there must be a second one of the same size that only produces the energy that will be used to replace both plants after that they have gone through their lifetime. And the energy produced by the first plant is net energy fully available to society for all the needed uses, including cathedrals if needed. Now, consider a power source that has an EROEI= infinity; then you don't need the second plant or, if you have it, you can make twice as many cathedrals. In the end, the difference between two and infinity in terms the investments necessary to maintain the energy producing system is only a factor of two.

It is like that: the EROEI is a strongly non-linear measurement. You can see that in the well-known diagram below (here in a simplified version, some people trace a vertical line in the graph indicating the "minimum EROEI needed for civilization", which I think is unjustified)):

You see that oil, wind, coal, and solar are all in the same range. As long as the EROEI is higher than about 5-10, the energy return is reasonably good, at most you have to re-invest 10% of the production to keep the system going. It is only when the EROEI becomes smaller than ca. 2 that things become awkward. So, it doesn't seem to be so difficult to support a complex civilization with the technologies we have. Maybe trips to Hawaii and SUVs wouldn't be included in a PV-based society (note the low EROEI of biofuels) but about art, science, health care, and the like, well, what's the problem?

Actually, there is a problem. It has to do with growth. Let me go back to the example I made before, that of a hypothetical energy technology that has an EROEI = 2. If this energy return is calculated over a lifetime of 25 years, it means that the best that can be done in terms of growth is to double the number of plants over 25 years, a yearly growth rate of less than 3%. And that in the hypothesis that all the energy produced by the plants would go to make more plants which, of course, makes no sense. If we assume that, say, 10% of the energy produced is invested in new plants then, with EROEI=2, growth can be at most of the order of 0.3%. Even with an EROEI =10, we can't reasonably expect renewables to push their own growth at rates higher than 1%-2%(*). Things were different in the good old days, up to about 1970, when, with an EROEI around 40, crude oil production grew at a yearly rate of 7%. It seemed normal, at that time, but it was the result of very special conditions.

Our society is fixated on growth and people seem to be unable to conceive that it could be otherwise. But renewables, with the present values of the EROEI, can't support a fast growing society. But is that a bad thing? I wouldn't say so. We have grown enough with crude oil, actually way too much. Slowing down, and even going back a little, can only improve the situation.

(*) The present problem is not to keep the unsustainable growth rates that society is accustomed to. It is how to grow renewable energy fast enough to replace fossil fuels before depletion or climate change (or both) destroy us. This is a difficult but not impossible task. The current fraction of energy produced by wind and solar combined is less than 2% of the final consumption (see p. 28 of the REN21 report), so we need a yearly growth of more than 10% to replace fossils by 2050. Right now, both solar and wind are growing at more than a 20% yearly rate, but this high rate is obtained using energy from fossil fuels. The calculations indicate that it is possible to keep these growth rates while gradually phasing out fossil fuels by 2050, as described here


  1. The key here is self-replication. When you can build a Solar PV factory that gets all its inputs from Solar PV powered machinery all transported to the factory by EV Trucks and besides providing enough energy to do all of that ALSO provides excess power to run everybody's McMansions and Carz. Said factory also needs to replace these panels as fast as they wear out.

    When a demonstration factory is built that self-replicates with enough surplus and without any fossil fuel input, then I will start to believe it is possible to create a sustainable system this way.

    1. ouch . . . now there's the elephant in the room "green" politico's and company's are systematically ignoring . . .

    2. That argument could easily have been used to claim that it was impossible to transition to an oil based transportation system because coal was being used to make large numbers of products needed for oil extraction. There will always be some mix of energy supply, even if we return to 100% solar energy. Hydro, wood, animal fats and human and animal muscle power are all pre-industrial versions of solar power. We did just fine with them until a few hundred years ago.

      The only time that self replication becomes relevant is if it is assumed that only one energy source is available and it is also clearly shown that the output of that energy source makes it impossible to recreate the source. Since the output of PV is electricity and since it is possible, in theory, to use electricity to produce all other liquid and solid fuels, it is clearly theoretically possible to use PV to replicate PV. That no one has tried to do so is not to say that it can't be done.

      The key is not self-replication; the key is substitution. As pointed out in the article footnote, can all forms of renewable energy substitute for fossil fuels at a fast enough rate to save civilization and the climate? I think the answer is no, but it's not because of the issue of self-replication.

    3. " That no one has tried to do so is not to say that it can't be done."

      I didn't say it could not be done. I only said that until at least a demonstration factory did something at least CLOSE to this would I believe it is possible to do.

      In each "energy transition" we did so far, there was always a BIGGER source of energy to access. Wood--->Peat--->Coal--->Oil--->Nuclear, always climbing UP the ladder there in energy density. Nobody claims solar is as energy dense as any of the last 3 AFAIK. If it cannot support itself and provide enough excess for the population, it's just not sustainable, unless of course the Italians get their Cold Fusion system running.

  2. Both the Ellen MacArthur Foundation and the physicist Robert Ayres have estimated that the efficiency of private automobiles, in terms of the work accomplished by moving humans from Point A to point B, is 1 percent. Ayres states that he has never seen a study of diesel trucks, but thinks the efficiency is probably higher.

    Now assume that the energy cost of producing the gasoline and diesel which power the private automobile system increases by 5 units. Also assume that there is really no substitute for the gasoline and diesel. Then the private automobile system that we see in the world today cannot continue to exist.

    Can the private automobile system morph into something else? For example, could motor scooters replace automobiles? And the answer is 'probably so'. And motor scooters are similar enough to automobiles that we might claim that the system has simply morphed, and not collapsed. Or we might envision public transportation systems.

    Coming out of left field, the ETP model (which you despise), claims that when the EROEI measured at the well-head is around 7, the oil industry will no longer be able to reproduce cannot afford to replace the barrel of oil which is burned. Such a claim rules out motor scooters as any long term solution. It also rules out buses and trucks as long term solutions.

    Whether a society can be built on PV panels remains to be seen. But the ETP model would claim that PV panels cannot be pasted on to a society built on oil which is rapidly approaching its dead state of EROEI 7.

    It is also possible to envision a society based mostly on photosynthesis with PV panels pasted on to it. Such a society would probably look much like Dmitry Orlov's Homesteader society.

    Don Stewart

  3. Euan Mearns article on the electric grid in the UK in January is relevant.
    It seems that wind can be pasted on to a system which relies on fossil fuels, but cannot be a stand-alone substitute for fossil fuels. At least in our current economy. Regardless of the EROEI of wind.

    If oil really is approaching the dead state with a well-head EROEI of 7, then coal transportation to the coal fired power stations will not be available.

  4. I would enjoy living in a society experiencing infinite growth. Growth of the mind. More novels, more plays, more poems, more operas, more philosophy, ... And none of that relies on fossil fuel. The greatest civilisation in history - the Hellenistic Age - was built on muscle power, which will always be with us.

    Yes, our obsession with material growth - more skyscrapers, more freeways, more aeroplanes - must end. But renewables are falling into a similar trap: the grid. I believe the electric grid will have to go: it is too inefficient, too complex, too fragile, to be supported by renewables. So our PV panels, our offshore windmills, are just expensive boondoggles that cannot run any feasible society. As for their EROEI, I think it is grossly exaggerated, but in any case, current renewables are sustained only by a tenfold larger fossil fuel base, and in the long run that cannot work.

    1. Well, whether one chooses the Hellenistic Age as "The Greatest Civilization in History" or not is a reflection of ones culture and upbringing. There is some pretty good competition for that sobriquet.

      But remember my friend, that the muscle power for that "Greatest Civilization" was provided for the most part by slaves.

  5. I agree that the economic precepts our society currently operates under can't be sustained with renewable energy sources. I also don't necessarily disagree with your arguments about societal disruption due to EROEI. However, "The sower’s way: quantifying the narrowing net-energy pathways to a global energy transition" and the above narrative completely ignore the necessity of resources and the energy density and specific energy required for resource extraction, transport and processing. Our Caterpillar 777's and other heavy equipment don't run on batteries, fuel cells or fairy dust. Iron ore, copper, silver, limestone, lithium, etc. in almost all cases can't be mined with electricity. The thought of using electricity for resource extraction becomes even more absurd considering the remoteness of many of the best mines. Until resource acquisition, life cycle energy costs, energy type and energy substitution are considered in the models, these dreams of a wind and PV electric savior are disingenuous at best.

  6. I can certainly envision an eventual group, mounted on what might be descended from thoroughbreds raised on Kentucky ranches, waving swords that might have been axle leaf springs and lawnmower blades, impinging upon homesteader society in a manner not showing much in the way of empathy. In fact, I wrote a novel with just that scenario.

  7. Yo made a valid point. It's not EROEI itself important (with the exception of EROEI<1 or even <2).
    It's the time of selfreplication. (Time to double the power).

    Energy sources based on extraction has very high time of selfreplication, because the time of extraction and consumption could be very, very high. The limit of the speed of consumption a liquid fuel it's so high that the real limits are normally imposed by other factors. See a rocket, for example. In matter of minutes the rocket consumes a insane big tank of fuel. Fortunately build rockets (not used them) take a lot of time.

    In renewables, this is very different. They can have high EROEIs, but the time of return is high too.
    We use "biofuels" in the past, using animals like vehicles today or farm soil, with low EROEI. But even biofuels with low EROEI ( perhaps 5 for a agrarian society with very low consumption) allow better time of return because the farm cycle is annual.
    That's the reason because we farm. Seed a tree and wait the result could be very high EROEI. The effort to put a seed in the soil is very low. But you have to wait a lot of years to have a return, even if it's a lot greater than farming (higher EROEI). So for us is more important fast return (time of selfreplication) than higher return on long term (EROEI).

    To work at this level with solar and wind, these energy sources needs to have high EROEI levels to compensate their long periods of time return.
    Fortunately, for this purpose, we are speaking about pure EROEI and not extended EROEI, as each energy source depends for selfreplication only on it's own energy and not about the energy of the society around it.

  8. There is a paradox here. The "Red Queen effect" predicts that a low EROI will increase the rate of growth rather than limit it. At an EROI of 1, all energy is used to produce more energy, with no net (the economy is doing nothing except 'spinning its wheels'). But all that energy use is counted in measures of economic activity. So as EROI approaches 1 the rate of growth will approach infinity.

    1. Honestly, Sam, that doesn't sound right.....

    2. i think thats wrong sam, and correct me if im wrong ugo, but at an eroei of one, an economy or any living thing wouldnt have the energy to work any more. so it wouldnt be 'spinning its wheels'. a living thing has to get more energy back from foraging than it expends in foraging, to survive. the ratio depend on the metabolism of the creature. an economy needs to more than break even to function, which is why fossil fuels are so much more addictive to humans than subsistence farming

  9. I know it doesn't "sound right". It doesn't feel right, either. That's why it is a paradox. As I see it, it has to do with how we measure "growth".

    Another example of paradoxical "growth" measurement is that both the manufacture of weapons and the health care for injured soldiers are positive additions to GDP. So maximum positive GDP would be for everyone to buy a weapon and wound someone.

    More seriously, the problem, at least as I see it, is that we have no metric for well-being. Are the Bhutanese onto something with their "Gross National Happiness" metric?

    1. Genuine Progress Indicator is quite well developed, has been calculated for most major economies, and treats the deficiencies of GDP while remaining an economic indicator.

  10. Ugo
    The historical European expansion and resulting Diaspora and spread of industrialisation over several hundred years of growth can be described as a race for global resources. ( Globalization and the Race for Resources, Bunker & Ciccadell, 2005)

    Your point about growth could be made more generally – ‘renewables’ are the next resource and must in large part over next decades substitute for existing energy resources. Growth (expansion) of industrial output has been continuous but distributed very unevenly. Indeed we can see the legacy of ‘uneven’ growth even now in remnants of historical agrarian economies (old civilisation) within Europe. (You have mentioned in the past historical adaptive change in 19th C Italy without local coal resources). Perhaps the most homogenously industrialised country, the USA, even now demonstrates both its own internal unevenness and a continued critical dependence on global resources.

    The global economy and the 50% of the population which is urbanised, is large without precedence and needs vast daily resources for maintenance in addition to delivering daily functions. It seems to me we are already testing whether resources can supply maintenance for existing economies given the difficulties in further growing the resource base. This is clearly still a race for resources. One might expect global industrial overall ‘growth’ to slow to a trickle from now on. Although renewable energy is widely distributed, the resources needed to build the units are not and can only be found within a global economy.


  11. Excellent essay. I agree that a society driven by renewable energy will have very little growth.

    What then are the implications of low growth? I argue in this essay that the main implication is very little credit. If true, how can a society build out and maintain renewable energy with very little credit? I don't think it is possible, but I would love to be proven wrong.

    Assuming it is possible to switch to renewable energy with little credit, what will replace the diesel that powers our critical life support network of combines, tractors, trucks, trains, and ships?

  12. Ugo:

    Thought-provoking. But I find my eye drawn back to "pyramid". I am thinking that this kind of thought exercise needs to be looked at and unpacked carefully.

    I think that Professor Hall's pyramid reflect the categories in terms of what they were/are during the halcyon days of oil leading up to this uncomfortable moment we are currently facing.

    14:1 for Art. Really? Is that a Shakespeare play performed at the New Gobe Theatre with the audience flying in for opening night, or is is a quality regional play held down in Ashland. i think that the presuppositions for this part need to be examined closely.

    Health Care at 12:1. Positron Emission Tomography scans and Big Pharma, yes. Four Thieves Vinegar Collective...Maybe a lot less, and X-rays are pretty easy to make too.

    Education at 10:1. Maybe in today's system of overeducating and underteach, but return to land colleges and teachers colleges might shave this down a lot.

    At so on through the "pyramid"

    When people look at little bits of eye-candy like Professor hall laid down, they seem to come up with the idea that when we drop below 14:1, Art will fall out of our lives....Not true...But it will change.

    The same can be said of all the other stages of the pyramid. We will always have all of these things....Just in what form will we have them?

  13. The equations and graphs here treat all energy types as equally useful but they are not. The higher EROEI renewables both produce electricity, useful but not concentrated and portable like liquid fuels, especially diesel. The ships used to build or even maintain all those offshore windmills, for instance, will need safe liquid fuel. How much energy would be lost making it from electricity. I wonder too if a share of the embodied energy in those ships is included in the EROEI calculation of the off shore wind turbines. Will those ships be recycled and rebuilt using wind power?

    Robert Firth points out that we have built civilizations on muscle power in the past. Indeed we have but most, perhaps all of them, have depleted other resources such as farmland or forests in the process. Our descendants will be facing a whole planet depleted and poisoned as never before.

  14. the above article goes into some depth on oil as our energy resource, but misses the point that our ultimate energy resource is food. It also seems to support the fallacy that somehow we will continue with BAU with some minor inconveniences.

    Energy , in whatever form can be reduced to its calorific value

    currently. in our industrial infrastructure and commercialised system, our food requires 10cal of input for every 1 on the plate. therefore we need at least a 10:1 return in order to survive. Those extra 10 calories have to come from somewhere, and there is only one source---oil.
    We eat oil
    the only factor that will drive people back to growing their own food is the prospect of starvation.
    A few will not succumb to that, but most will, leaving our population once again at a sustainable (pre-oil) level

    the alternative is to expect that everyone grows their own food on their own land ---which is of course nonsense, there isn't enough land to do that.

    In addition, 80m new mouths present themselves to be fed each year.
    And already the mothers of the next 2 bn are alive and waiting to reproduce

    Add to that the uncertainties of the effects of climate change on food supplies and the above pyramid starts to look a bit shaky

  15. You may be interested in this

    A study looking at a rapid built of renewable power . During the build out the eroi available to society dips significantly, as the energy needs to be "plowed into" further build out.

  16. Norman hit the "nail" on it's "head", "renewables" produce only ELECTRICITY erratically, unreliably & NONE of the essential RAW MATERIALS we now get from OIL!
    This is a very important point that most people who are pushing renewables overlook.
    We do indeed "eat" oil, it powers our transportation system, our manufacturing, it powers our highly mechanized agricultural system, it provides the feedstock for our fuels, medicine, lubricants, fertilizers, pesticides, herbicides, plastics, synthetic fibers & millions of other products.
    OIL powers our mining equipment, those huge machines we now need to mine the low quality ores to produce those "renewables" & "renewables" can never produce enough energy to power those machines, transport that ore or process it into the Parts & products we use.

    There is no way for "renewables" that collect about 20% of the weak, dispersed energy from sun or wind can replace the concentrated energy & raw materials found in OIL.

    "Renewables" are made WITH the raw materials & with the energy of OIL,COAL & NATURAL GAS, without these fossil resources, they like most of us would not even exist & they won't exist for very long after the oil age.

    We will collapse both economically & our population will collapse from it's high of 7.6 billion down to perhaps only 500 MILLION.

    Now use your imagination to figure how this will happen, then add climate change to declining oil, not a pretty picture is it.

  17. Surprised I don't see more mention of emergy analysis ( per Odum) here and elsewhere as a better way to understand why the EROEI of PV and other technologies isn't the best measure of whether they can be self sustaining past the fossil era. Granted that the calculations are complex, but the full accounting of the embodied energy of these artifacts will show that EROEI only partially tells the story.

    I also second Anonymous's point above that the extractive and heavy transport industries run on liquid fuels, and electric power does not replace them. All the other commodities that we need to run our industrial system must first be freed up for use by liquid fuels, the indispensable resource.

  18. If selfdriving cars can prevent accidents, we could lighten cars to little more than thin shelled electrical bikes. Most city travel is under 35 mph anyway. Too busy on Facebook and it won't matter how long the drive takes. We don't need much energy for art. We need a health culture. No technology will provide that. Here in the good old USA tv is our art and having a game show host as President is the height of culture. We have a sickness solar panels can't fix.

  19. All this article does is demonstrate that ERoEI is not a good measure - it is too simple, as you would expect from a single number.

    The most important factor missed is the TIMING of ER and EI. The EI has to be spent first, before you get ANY ER back, and then ER trickles in over the next 25 (Lifetime) years. So you cannot use 10% of your ER to be your EI - you have to already have the EI spare from other sources.

    When you recognise that, you will realise that if we had started the transition to renewables 30 years ago, it could have been completed. But given that Crude Oil has passed its peak, and Coal has passed its peak on an Energy basis (as opposed to a tonnage basis), the ability to complete a transition to renewables has now gone. At some point in the future the problem will become whether you keep investing increasingly valuable energy in more renewables infrastructure, or whether you keep the lights on.

    And since ALL energy-producing industries have to operate at a profit or go bust, the future is going to be financial chaos and collapse.

    1. All complex technologies including energy require a significant up front investment. Which means they need plentiful credit. Which means they need economic growth. Which means energy must have a high EROEI.

    2. The thing is, credit can't always buy you all the energy you want - that is pre-Peak Energy thinking. Post Peak Energy, there has to be some demand destruction somewhere, and that means there must be losers in the competition for the remaining energy, no matter how much credit you have.

      Hours after oil reached $147 /b in NYMEX trading on 11 July 2008, the news broke that Fanny Mae and Freddy Mac were to begin talks that led to their receivership/"special management" - too much housing credit had turned bad in an environment where an extra $1.75 trillion over 4 years had been sucked out of the financial system to buy more expensive oil. When markets re-opened, the oil price began its crash to $34 /b.

      The next oil price cycle took 6 years, and the NEXT cycle will be even shorter. Ugo's diagram of the cycle's feedbacks is correct, but it looks like it will go on and on for ever - it won't.

      The way to get the time factor into the analysis, is to drop the simplistic ERoEI in favour of the lifetime energy budgets from which ERoEI is calculated - two arrays EI(t) and ER(t) for t=0 to Lifetime.



Ugo Bardi is a member of the Club of Rome and the author of "Extracted: how the quest for mineral resources is plundering the Planet" (Chelsea Green 2014)