Cassandra has moved. Ugo Bardi publishes now on a new site called "The Seneca Effect."

Wednesday, September 11, 2013

If we have to use fossil fuels to manufacture renewable plants, doesn't it mean that renewables are useless?

In this post, Marco Raugei makes a fundamental point about an often raised question: if we have to use fossil fuels to manufacture renewable plants, doesn't it mean that renewables are useless? Raugei's answer is a resounding "no". In fact, the EROEI of fossil fuels acts as a multiplier for the final EROEI of the whole process. It turns out that if we invest the energy of fossil fuels to build renewable plants we get an overall EROEI around 20 for a process that leads to photovoltaic plants and an even better one for wind plants. So, if we want to invest in our future, that's the way to go, until we gradually arrive to completely replace fossil fuels! (image above from "The Energy Collective")

The EROI and promise of PV (and other renewables). Trying to avoid unnecessary inconsistency and confusion, and to keep an open mind and a balanced viewpoint.

by Marco Raugei

The energy return on energy investment (acronym: EROI or EROEI) provides a numerical quantification of the benefit that the user gets out of the exploitation of an energy source, in terms of “how much energy is gained from an energy production process compared to how much of that energy (or its equivalent from some other source) is required to extract, grow, etc., a new unit of the energy in question” [1].

As straightforward as this definition may sound, when dealing with the diverse range of existing energy sources and technologies, the devil is in the details.

It goes without saying that, in order to ensure comparability, it would be wise to at least approach all EROI studies of different energy technologies by applying a strictly consistent methodology, including the all-important aspect of system boundaries (i.e. what should be included and what not). Otherwise, a reported lower extended-boundary EROI for any given new energy technology may (artfully or artlessly) be taken out of context by readers who have their own axe to grind (or who are just too eager to oversimplify), and then used to incorrectly single out that particular technology as a worse performer vis-à-vis more conventional ones (typically, fossil fuel-fired electricity production).
An instance of such a potentially tricky situation has recently arisen with the publication of a book on the EROI of the photovoltaic (PV) sector in Spain [2].

One aspect of the controversy is rooted in the fact that the interdependence of PV and fossil fuels is not ‘symmetrical’ – no-one in their right mind could claim otherwise - and hence the EROI of PV is affected by the EROI of the fossil fuels (oil, coal, and gas) that underpin it. Additionally, fossil fuel technologies are much more ‘mature’, and much of the necessary infrastructure for their operation (rigs, pipelines, roads, etc.) was developed long ago and has largely been amortized already. As a result, it may well be that extending the boundaries of the EROI analysis for fossil fuel-based technologies may end up making a smaller difference vs. doing the same for newer technologies such as PV.

Yet, it seems that this argument is too often brought up to imply that, since PV development and deployment is currently (largely) underpinned by fossil energy, and hence PV is not (yet) a fully independent and truly 100% renewable energy technology, then "why bother" in the first place?
Actually, this kind of critique is aimed at countering the incurable technological optimists' view that "there is nothing to worry about: we can continue unabated in our reckless business-as-usual overconsumption of energy (and resources) because soon PV (and other renewables) will seamlessly step in and take the baton from dirty fossil fuels, and all will be well".

Such through-rose-tinted-glasses optimism is most likely wrong-headed and should probably be tamed. But it is also worth looking at the issue from another angle. Let us assume that the average EROI of the current mix of fossil fuels (which still represent our main sources of primary energy, globally) is some value X > 1. And let us also agree that we (as a society) need a large and ever-growing share of our energy budget in the form of electricity (to power our computers, telecommunications, trains, home appliances, etc).

Broadly speaking, we therefore have two options:

1) keep using all the oil (and other fossil fuels) directly as FEEDSTOCK fuel in conventional power plants. In so doing, we would get out roughly 1/3 of the INPUT energy as electricity (electricity production efficiency in conventional power plants being ~0.33). This would be the "quick and dirty" option, that maximizes the short-term (almost instantaneous, in fact) "bang for the buck".

2) Use the same amount of available oil (and other fossil fuels) as (direct and indirect) INPUT for the production of PV plants.

Building and deploying a modern crystalline silicon PV system requires approximately 3 GJ of primary energy per m2 (note that this value takes into account the conversion to electricity at ~0.33 efficiency prior to use in the PV manufacturing operations which are carried out using electric power). When installed in southern Europe (irradiation = 1,700 kWh/(m2*yr)), such system, operating at an average efficiency of 13% (reference) * 80% (performance ratio) = 10%, will produce approximately 5 MWh (= 18 GJ) of electricity per m2 over its 30-year lifetime [3,4]. What this means is that the c-Si PV system would provide an output of electricity roughly equal to 18/3 = 6 times its primary energy input, which corresponds about 6/0.33 = 18 times the amount of electricity that we would have obtained, had we burnt the fuel(s) as FEEDSTOCK in conventional power plants (option 1 above), instead of using them as INPUT for the PV plant.

Of course, we cannot afford to switch to option 2 tout-court overnight, for a number of technical as well as systemic reasons [5]. First and foremost, we simply would not be left with enough energy output in the short term to sustain and power our complex society. But an almost 20x improvement in the efficiency with which we use our limited and dwindling endowment of fossil fuels must be worth at least some consideration.

A planned long-term investment might be advisable, for instance, aimed at bringing about a gradual transition. The latter is in fact what many have been advocating, often only to be met with rather negative ‘gloom and doom’ reactions by others on a number of prominent discussion forums. It seems as if, in the minds of the latter, the desire to show that ‘the emperor has no clothes’ (i.e. that PV and other renewables are not yet, and might never be in full, a real, completely independent and high-EROI alternative to fossil fuels) overrides all other considerations, and prevents them from realizing/admitting that, after all, it may still be reasonable and recommendable to try and push this slow transition forward.

To conclude, I would like to dispel all doubts and clearly state that I do agree with the aforementioned ‘pessimists’ that if we (as a society) do not come to grips with the notion that there is no such thing as infinite growth on a finite planet [6,7], and re-align our goals and ‘development’ strategies accordingly, then all the technological fixes in the world stand little to no chance of being enough to avert an ominous crash. But, why write off PV (and other renewables) and deny their value as useful tools to (hopefully) help us out on a safe slide along the slopes of a "prosperous way down" [8]?


1. Murphy D.J., Hall C.A.S., 2010. Year in review – EROI or energy return on (energy) invested. Ann. N.Y. Acad. Sci. 1185:102-118
3. Fthenakis V.M., Held M., Kim H.C., Raugei M., 2009. Update of Energy Payback Times and Environmental Impacts of Photovoltaics. 24th European Photovoltaic Solar Energy Conference and Exhibition; Hamburg, Germany
4. Fthenakis V.M., Kim H.C., 2011. Photovoltaics: Life-cycle analyses. Solar Energy 85(8): 1609-1628
5. Smil V., 2010. Energy Transitions: History, Requirements, Prospects. Praeger, ISBN-13: 978-0313381775
6. Meadows, D H., Meadows D.L., Randers J., Behrens W., 1972. Limits to Growth. Signet, ISBN-13: 978-0451057679
7. Bardi U., 2011. The Limits to Growth Revisited. Springer, ISBN-13: 978-1441994158
8. Odum, H.T., Odum E.C., 2001. A Prosperous Way Down. Colorado University Press, ISBN-13: 978-0870819087


  1. I'd blogged on this issue a while back.

    I suspect that hi-tech 'renewables' can only be a transition technology, supported as they are by fossil fuel subsidies and draw-down of Non-renewable Natural Resources in their manufacture.

    The links in my footnotes are worth a read.

  2. A journey begins with a first step and you really don't know where your going till you start. Writers, artists, creative types of all kinds know that when they start they don't know what they will wind up with when they finish. The creative act of renewing the world, of harnessing sustainable energies and creating new lifestyles will require the talents of the best of us.

    If carbon is managed there is no better use for fossil fuels than to manufacture renewable plants. If a renewable plant can't wear out (impossible thought experiment) the EROI of the plant would be absolutely the same as if renewable fuels had been used to build the plant in the first place. A fancy way of saying that if an energy plant lasts long enough its initial construction costs would be insignificant. - K-Dog

    And talk about not knowing what your going to end up with I deleted and reposted this comment twice all because I had 'not' when I needed 'no'. Ugh!

  3. renewables are a non starter without obvious measures such as down scaling population, the implementation of no growth economy and removal of luxuries like cars and air travel. the energy output of renewables [in terms of any likely scale] is just too pathetic to power both the rapaciously consuming industrial civilization, and the constant increasing of renewable energy needed to allow infinite economic growth to occur. i'm not convinced renewables could even create enough energy to replace themselves on a 1 to 1 basis when not powering anything else, taking the colossal energy needed in the complete industrial process, such as mining ores and transportation etc.

  4. Curiously enough, the not referenced book, by someone as 'suspicious' as Mr. Hall and Mr. Pedro Prieto (who owns and rules some PV plants in Spain, and who gave a big amount of REAL data in order to do the calculations), calculate an EROEI in Spain for the PW <3 (IIRC, around 2.7).
    It seems to me that there is something wrong here, and usually only the PV is discussed and the main subject of many of the hot debates, while other renewable energies (hidraulic and wind, not only for electricity purposes) are totally forgotten, even when they have by far much better EROEI.
    To me, all this stuff looks like someone is crying because their faith in technooptimism and, probably, their investments/business are going to dire straits.
    PV, as many other electricity generation, only backs electricity based way of living, and that means batteries, electric cars and such. IMHO, BAU 100%.
    Of course, a growing population is something that even the most 'heavy believers in Oil Crash' avoid to speak about, but it is the real issue, not only the habits of (energy) consumption.
    Back to PV, just my thoughts: c-SI PV (as well as many other PV technologies) is somethng real complex (ok, it is simplier when compared to active electronics), not suitable to build in the backyard or in relatively simple societies. Eolic energy, as well as hydraulic energy, had been built by humans well before the industrial revolution, and gave a good EROEI, better than PV. Simplicity is another point, IMHO, when analyzing energy sources, not only EROEI, or the 'beautiness' of them.

  5. The above is RUBBISH!

    Our analysis shows the EROEI of Natural Gas Fired Electricity at the customer's meter to be ~ 1.2.

    Given a fair comparison, which we believe to be at the wellhead, minemouth for Hydrocarbons, and FOB plant for wind turbines and PV, and at the plant gate for Hydro and GeoThermal, Our analysis of PV mounted at Purdue University FOB plant is an EROEI of ~ 33.

    This has been vetted.

    Furthermore our NaSiS economic model in stark contrast to the current situation where 40 Quads of Energy produces 13 Quads of electricity at the Customer's Meter, and 30 Quads of Hydrocarbons produce 22 Quads of propellent in the Customer's tank

    Has PV & Wind & Hydro & GeoThermal producing 3O Quads of Electricity, 14 of which power Windfuels plants producing 3 Million Bbl/day of liquid fuels to power the Chemicals Industries, Military, and Ocean Transport netting 22 Quads of usable energy, with no CO2 emissions whatever.

    This model shows that ~ 2 Quads are required to maintain the infrastructure, in stark contrast to the existing situation where 30 Quads are so required.


  6. Interesting, thought-provoking essay. ERoEI is like one of those greased pigs at the county fair: you can see it perfectly well, but lotsa luck getting a good hold on it. Good entertainment, tho.

    As “K-Dog” expressed it, at the beginning of a development there is no clear view of how it will end up. (This applies to more than creative developments.) When my children were young, I would tell them a story that took shape in the telling — none of us, least of all myself, had any idea of the ending, or even if it would have an ending that night or whether it would need to be continued another night, the next day of the long trip in the car, whatever. And they would let me know if they did not like how it was going, and we would consider possible alternatives. Not surprising, the grandchildren like such stories, too. “What if… Let’s say” is the generic classification.

    There are songs in which some of the verses can be switched around with minor loss, if any, of sense. E.g. “Home on the range”, “Blow ye winds westerly”. In others, the verses must all be there, none left out, none out of order. E.g. ballads like “Golden Vanity”, “The streets of Laredo”. Songs such as the latter are meaningless if not followed through to conclusion. Each verse depends on the preceding verses.

    Fossil fuels figure in the first verses of the Energy ballads now being composed and played and, one way or another, will be gone with the wind at the end. How the ballads will develop is being determined even now.

    Separately, I would venture that Dr Raugei’s use of ERoEI could perhaps be sharpened by considering it as Energy Return on Fossil [or Nonrenewable] Energy Invested.

    David F Collins

  7. Comment received by Charles Hall.

    While I agree with many things that Marco Raugei says, and wish him and PV well, I wish to correct several issues that he keeps making about our study that are not true or where we have covered his points ourselves.

    1) It is not true that we are using different boundaries in our analyses for PV and fossil energies . While I do like the analysis Pedro and I did because of the detail he was able to provide, detail that is hard to find generally (at least at present levels of funding), we do in fact include, at least in theory, almost exactly the same boundaries for +/- all our fossil fuel analysis as for our Spain PV analysis since we include for ff all expenditures for indirect or input goods and services , usually based on government statistics (e.g Grandell et al; Guilford et al.) and imperfect energy intensities per dollaror Euro spent (but about which we undertake sensitivity analysis, not greatly changing our results because direct energies tend to be larger than indirect). Presumably this includes amortized indirect investments needed to do all the 30+ things we are able to list in detail in Prieto and Hall's book: e.g. needed roads, fences, financial services etc etc. (but not direct labor in either case). One may argue about differences in precision but not, I think, boundaries.

    2) we do not include the energy costs of backup or of accommodation of the rest of the system to the intermittancy of solar (BUt note that when Graham Palmer does for rooftop PV in Australia he
    gets very much the same results we do). I think this is important in EROI studies.

    3) We have undertaken electricity-weighted analyses as Raugei does and reported EROIs of about 7:1 in our last chapter, which we compare to the low electricity yield of burning the input fossil energy directly. Since Pedro and I found that making the modules was about one third of the input energies I assume this is why Marco gets 18:1 from (apparently) modules alone. He should also mention that the returns come over 25 or 30 years whatever his results.

    I leave other issues to others.

    I too hope we can move to solar, but it will not be easy. I think the relatively low return over a very long time and society's discount rate (be that good or bad) has a lot to do with the slow growth of renewables. In fact since fossil fuels are still growing at last look we are not displacing fossil fuels with renewables but just adding them to the mix.

    Charlie (I am off line as I am going fishing)

  8. I am pleased and flattered that Dr. Hall has taken the time to read and comment on my post.

    To briefly address his points:

    1) Noted.
    First of all, I would like to make it clear that my post was rather more general in scope, and not meant as a direct critique of Prieto & Hall's book (which was only mentioned in passing as one possible instance where results might be extrapolated out of context ex-post). Specifically, it had been my impression that Prieto & Hall's analysis included a number of extended-boundary inputs that are not customarily included in many published EROI analyses. I am pleased that Dr. Hall himself has performed other studies with the same boundaries, though. As I said, I also agree that "it may well be that extending the boundaries of the EROI analysis for fossil fuel-based technologies may end up making a smaller difference vs. doing the same for newer technologies such as PV."

    2) True, and agreed. Even if some recent reputable studies have questioned whether the issue of intermittency may have been somewhat overblown [e.g. Gross R, Heptonstall P, Anderson D, Green T, Leach M, Skea J. The Costs and Impacts of Intermittency: An assessment of the evidence on the costs and impacts of intermittent generation on the British electricity network. UK Energy Research Centre 2006; ISBN 1-90314-404-3]

    3) Again, noted and true. But again, my post was more general, and not intended as a direct critique of that book. Besides, the book's last chapter is not given much prominence within the book as a whole, and my impression is that it may easily go relatively undetected. In fact, I have not come across a single instance where the electricity-weighted 7:1 ratio has been cited; everyone seems content with the simplistic take-home message of a too low ~2:1 ratio.
    As for the energy returns of PV coming over 25 or 30 years, that is precisely what I meant when I wrote "we simply would not be left with enough energy output in the short term to sustain and power our complex society." But in hindsight I agree I should probably have been more explicit about this specific point. Thank you for bringing it up.

    Overall, I am pleased to see that we are mostly on the same page.

    All the best (and happy fishing!)


  9. Hall believes that an EROI of at least 12 or 13 is needed to keep civilization going as we know it. Prieto and Hall state that even 100% efficient PV (impossible) would only raise the EROI of Spain’s sunny solar PV from 2.45 to 3.25

    The EROI is obviously lower than 2.45 – Hall assigns a zero EROI to the natural gas peaker plants and it’s also clear from his book that the PV is not going to perform for 25 years at the optimistic forecast levels for many reasons. A zero EROI figure is also given to events of Force Majeure, acts of god, wind, lightning, storms, floods, hail, etc.

    If you can’t afford the book or get it from a nearby university, I have a book review "Tilting at Windmills, Spain’s disastrous attempt to replace fossil fuels with Solar Photovoltaics" is at

    The financial issues alone are a showstopper, especially at this point in the largest credit bubble and most highly corrupt financial cycle in fiat currency history. (I have a book list about this at which would have a lot more than 40 books if that weren't the limit amazon places on book lists).

    There are so many issues with the electric grid that it doesn't matter if ANYTHING that generates electricity has a positive EROI. At my Electric Grid Overview
    there are links to articles about this, as well as book reviews of:

    "Lights Out. The Electricity Crisis, the Global Economy, and What It Means to You". Makanski.

    "From Edison to Enron: The Business of Power and What It Means for the Future of Electricity. Munson.

    “CYBER WAR. The Next Threat to National Security and What to Do About It”. Richard A. Clarke.

    “America the Vulnerable: Inside the New Threat Matrix of Digital Espionage, Crime, and Warfare”. Joel Brenner.

    The bottom line is that no EROI study can possibly discover all the oil embedded in the entire process, which I think is captured best in Leonard Read's 1958 "I, Pencil My Family Tree".

    Alice Friedemann

    1. To get higher EROIs, it isn't neccesary to get better power/surface ratio but increase the "cost" (energy to build a panel)/power ratio.

  10. Imagine for a minute a society, where 80 % of the electricity generation - is from renewables. Then of course the EROI of any new renewable installation would be a considerable factor higher than nowadays. This is where we are hopefully heading.

    In the meantime the sacred cow of "irreduceable" energy needs of so called "developed" societies should be slaughtered. Give us a really working, comprehensive cap-and-trade scheme and you see, how much entropy production can be reduced even here...

    1. The cap and trade scheme is really only a small facet of the overall problem. If you make specific elements (carbon) a subject of law, you end up with too many ways to circumvent the law and lose the true effect it needs to be addressing: consumption.
      In other words, instead of cap and trade for carbon credits, we need a universal sales tax on ALL goods that reflects the cost of human development and consumption of resources at the point of decision: the cash register. This is where the real public opinion votes are counted.
      At the same time, we need to eliminate all of the schemes that encourage consumption and debt (promises to consume in order to pay), such as income taxes, loan programs that encourage economic growth without a sustainability clause, etc.
      I do like your main point of thinking in terms of a society with 80% renewables. It is rare to divert the fossil vs. renewable discussions from their predictable responses.

  11. In addition to what Charlie said, there are a few more false assumptions being made in the article.
    - All the energy used in the construction of a PV module is electric
    - All the electricity used come from FF power plants with a very low efficiency of 33%
    - Once installed PV systems last 30 years without any additional energy input

    Most likely there are more but these jumped out at me straight away. Getting those numbers right would give a lower EROEI.

  12. Just to set things straight regarding these three particular comments:

    " there are a few more false assumptions being made in the article.
    - All the energy used in the construction of a PV module is electric"

    Not true. The cited primary energy requirement was calculated through a complete Life Cycle Assessment, and included both inputs of electricity and direct inputs of fuels (for transport, infrastructure, etc.), each duly expressed in terms of the primary energy ultimately required to generate and deliver all those different types of energy.

    "- All the electricity used come from FF power plants with a very low efficiency of 33%"

    That much was not stated, nor implied, anywhere. The 33% efficiency figure was only used as a benchmark value for 'typical' FF power plants.

    "- Once installed PV systems last 30 years without any additional energy input"
    Some known and predictable maintenance inputs were actually included in the underlying LCA (and hence in the calculated ~3 GJ of primary energy per m2).
    Additional fossil energy may be required over the PV plant's lifetime, but surely nowhere near as much as the amount of FEEDSTOCK fuel consumed by a conventional thermal plant...


  13. I am basically in agreement with what Andyuk says further above.

    However I also think that renewables can be one part of an overall more comprehensive solution / strategy nvolving several other elements or components. (such as ending population growth and reversing it) So if renewables is presented as "THE solution" it almost certainly will do more harm than good because it will detract from the focus and attention that needs to be placed as well on other components or elements of an overall strategy. (an overall global strategy for survivability and sustainability)...But if it is seen as one of several component (and its relationship to the other needed components is also clearly spelled out) than it is probably a very good thing to (also) focus on. And to do that, working out and spelling out some of the quantitative aspects, as the post above does is also quite useful.

  14. Too often (practically always) these types of articles zip past us without a second thought to the real picture of energy use. Instead they focus on generation, and everyone just either assumes energy use will not change, or they assume that what we are using is OK because it is what people want, and that means we have to meet that demand. People also demand porn, but we don't provide it to every teenage kid in public school. We set limits. We manage.
    The issue of renewables comes down to costs. If it costs more to produce energy in the future, people will have to change their habits. Turning off their personal 'grid' at night, for example. Having batteries as an essential part of every household system is another example.
    We can produce diesel and gas from biomass. From that, it's just a matter of quantities of energy that are being used that are necessary vs. the quantities that are basically wasted for entertainment (such as the internet servers running porn and redneck TV).
    The issue has never really been whether renewables can replace fossil fuels. The issue is whether human beings have the ability to manage their affairs with logic and understanding instead of just mindless consumerism exploited by a few who profit from maintaining the mindlessness.



Ugo Bardi is a member of the Club of Rome, faculty member of the University of Florence, and the author of "Extracted" (Chelsea Green 2014), "The Seneca Effect" (Springer 2017), and Before the Collapse (Springer 2019)