Tilting at windmills

24 02 2014

pedroI have ‘known’ Pedro Prieto online for many many years, and have featured his work here, and here on DTM.  Pedro is an expert on renewable energy, and is one of the few engineers I’ve ever read who understands ERoEI (Energy Return on Energy Invested), and has practical experience in deploying both wind and solar energy system in his native Spain.  Pedro has recently dropped a bombshell in a book he co authored with Charles Hall  (EROEI is the ratio of energy output over energy input, a measure that was developed by Professor Charles Hall).  This book, titled “Tilting at Windmills, Spain’s disastrous attempt to replace fossil fuels with Solar Photovoltaics”, is the first in-depth look at the ERoEI of large-scale PV in any developed nation. And the results do not bode well……

This is the first time an estimate of Energy Returned on Energy Invested (EROI) of solar Photovoltaics (PV) has been based on real data from the sunniest European country, with accurate measures of generated energy from over 50,000 installations using several years of real-life data from optimized, efficient, multi-megawatt and well oriented facilities.

Other life cycle and energy payback time analyses used models that left out dozens of energy inputs, leading to overestimates of energy such as payback time of 1-2 years (Fthenakis), EROI 8.3 (Bankier), and EROI of 5.9 to 11.8 (Raugei et al).

Prieto and Hall added dozens of energy inputs missing from past solar PV analyses. Perhaps previous studies missed these inputs because their authors weren’t overseeing several large photovoltaic projects and signing every purchase order like author Pedro Prieto. Charles A. S. Hall is one of the foremost experts in the world on the calculation of EROI. Together they’re a formidable team with data, methodology, and expertise that will be hard to refute.

Prieto and Hall conclude that the EROI of solar photovoltaic is only 2.45, very low despite Spain’s ideal sunny climate. Germany’s EROI is probably 20 to 33% less (1.6 to 2), due to less sunlight and efficient rooftop installations.

Here is what Gail Tverberg has to say on ERoEI…

Commenters frequently remark that such-and-such an energy source has an Energy Return on Energy Invested (EROI) ratio of greater than 5:1, so must be a helpful addition to our current energy supply. My finding that the overall energy return is already too low seems to run counter to this belief.

Adequate Return for All Elements Required for Energy Investment

In order to extract oil or create biofuels, or to make any other type of energy investment, at least four distinct elements described in Figure 1: (1) adequate payback on energy invested,  (2) sufficient wages for humans, (3) sufficient credit availability and (4) sufficient funds for government services. If any of these is lacking, the whole system has a tendency to seize up.

EROI analyses tend to look primarily at the first item on the list, comparing “energy available to society” as the result of a given process to “energy required for extraction” (all in units of energy). While this comparison can be helpful for some purposes, it seems to me that we should also be looking at whether the dollars collected at the end-product level are sufficient to provide an adequate financial return to meet the financial needs of all four areas simultaneously.

My list of the four distinct elements necessary to enable energy extraction and to keep the economy functioning is really an abbreviated list. Clearly one needs other items, such as profits for businesses. In a sense, the whole world economy is an energy delivery system. This is why it is important to understand what the system needs to function properly.

Source of the EROI 5:1 Threshold

To my knowledge, no one has directly proven that a 5:1 threshold is sufficient for an energy source to be helpful to an economy. The study that is often referred to is the 2009 paper, What is the Minimum EROI that a Sustainable Society Must Have? (Free for download), by Charles A. S. Hall, Steven Balogh, and David Murphy. This paper analyzes how much energy needs to be provided by oil and coal, if the energy provided by those fuels is to be sufficient to pay not just for the energy used in its own extraction, but also for the energy required for pipeline and truck or train transportation to its destination of use. The conclusion of that paper was that in order to include these energy transportation costs for oil or coal, an EROI of at least 3:1 was needed.

Clearly this figure is not high enough to cover all costs of using the fuels, including the energy costs to build devices that actually use the fuels, such as private passenger cars, electrical power plants and transmission lines, and devices to use electricity, such as refrigerators. The ratio required would probably need to be higher for harder-to-transport fuels, such as natural gas and ethanol. The ratio would also need to include the energy cost of schools, if there are to be engineers to design all of these devices, and factory workers who can read basic instructions. If the cost of government in general were added, the cost would be higher yet. One could theoretically add other systems as well, such as the cost of maintaining the financial system.

The way I understood the 5:1 ratio was that it was more or less a lower bound, below which even looking at an energy product did not make sense. Given the diversity of what is needed to support the current economy, the small increment between 3 and 5 is probably not enough–the minimum ratio probably needs to be much higher. The ratio also seems to need to change for different fuels, with many quite a bit higher.

So there you have it folks…….  solar will never keep civilisation as we know it going.  But you already knew that. And before Eclipse jumps in, I found this on Nuclear Power…:

The seemingly most reliable information on ERoEI is quite old and is summarized in chapter 12 of Hall et al. (1986). Newer information tends to fall into the wildly optimistic camp (high EROI, e.g. 10:1 or more, sometimes wildly more) or the extremely pessimistic (low or even negative EROI) camp (Tyner et al. 1998, Tyner 2002, Fleay 2006 and Caldicamp 2006). One recent PhD analysis from Sweden undertook an emergy analysis (a kind of comprehensive energy analysis including all environmental inputs and quality corrections as per Howard Odum) and found an emergy return on emergy invested of 11:1 (with a high quality factor for electricity) but it was not possible to undertake an energy analysis from the data presented (Kindburg, 2007). Nevertheless that final number is similar to many of the older analyses when a quality correction is included.

Notice this was written in 1986.  As the quality of Uranium ores worsen, (they’ve worsen rapidly since 1986…), nuclear will be no more able to keep Business as Usual running than solar.

As extraction and depletion have operated over time, the average ore grade has decreased and the uranium has become more and more dispersed within the background substrate, plus the total amount of uranium we can extract can decrease as well. Leuwen (2005) argues that the empirical extraction yield declines much more sharply than the hypothetical one, which could come into play if there is a large increase in nuclear capacity in the coming decades.

Figure 6 – % of Uranium Extracted from Ore as a Function of Ore Grade (Leeuwen 2005).
Click to Enlarge.An increasing portion of the world’s uranium comes from in-situ leaching (ISL) (Hore-Lacy 2007).

Just enjoy life in the quiet lane…….  it’s not that bead, really…..



19 responses

24 02 2014
Nathan A

I tend to agree with Nicole Foss that most renewables make sense on a smaller, local scale, but lose out once all the infrastructure costs of large-scale power generation and supply are applied. I think we need to descale things, but not all the way back to private homesteads, rather to relocalised communities.

re: nuclear, my understanding is that the ore extraction plays a relatively small role in the overall cost of power generation, as compared to say coal. So, while an increase in the cost of that part of the process will still raise the overall cost of the end product (nuclear-generated electricity), it would not have the same devastating effect as similar rapid declines in coal ore grade might.

24 02 2014
Eclipse Now

“Notice this was written in 1986. As the quality of Uranium ores worsen, (they’ve worsen rapidly since 1986…), nuclear will be no more able to keep Business as Usual running than solar.”
You have merely asserted this! Also, does that ERoEI study include the hundredfold INCREASE in energy one gets from breeding uranium? Given that today’s nuclear waste could run the world for 500 years if in breeders, more work needs to be done on nuclear ERoEI. Even after deducting a whole factor for reprocessing, breeders are probably up around 100:1 not 10:1!

24 02 2014

You keep telling us about these breeders, but as far as I know only two were ever built, and they were both lemons. The French one, Phoenix, was closed years ago after strings of problems and incidents. Ditto with the Japanese one. I reckon they are just pie in the sky. Just like 100% renewables.

24 02 2014

Mike, Never say never. Extrapolating current tech energy EROI ahead without a reasonable rate of cost improvement should remind you avoid making a blanket statement like, “Solar will never keep civilisation as we know it going.” I am aware of a thermoelectric technology right now that will change the entire equation for solar, as well as for electrical storage. Even if I wasn’t aware of this technology, I would not accept such a nihilistic view. And neither should you. The history of tech is lowering input costs over time and it would be a mistake to claim that solar costs cannot follow that same guideline.

Of course, I think, as you do, that our civilisation is pretty lame overall and I would never suggest that “keeping it as we know it” should be a goal for anyone. Improvement is possible and hopefully we’ll figure it out before we become extinct. Or, some of us may figure out how to build space colonies and occupy outer space, leaving the dregs of the species behind on Earth.

24 02 2014
Eclipse Now

Hi Mike,
What are his qualifications to comment on nuclear ERoEI? He might be an expert on larger Solar PV installations (and I’d love to see his claims peer-reviewed by other solar PV experts, because there’s thin film and different manufacturing techniques to analyse as well), has he overseen the construction of a nuclear power plant?
The peer-reviewed work seems to suggest he has not.

“Life cycle analysis for Vattenfall’s Environmental Product Declaration for its 3090 MWe Forsmark power plant for 2002 has yielded some energy data which is up to date and certified. It shows energy inputs over 40 years to be 1.35% of the output.”

“So if very low grade ores (0.01% U) were envisaged, the energy inputs for mining and milling would increase significantly, but total inputs would still only be under 4% of outputs for the full fuel cycle.”


Note: many nuclear power plants are now being built to last 60 or 80 years, so the infrastructure will last longer, raising the final payback and increasing the ERoEI.

24 02 2014

Google is your friend……….

24 02 2014
Eclipse Now

And while I don’t think Solar PV are ‘the’ answer, is this guy over-counting energy inputs or something? Other peer-reviewed energy studies seem to confirm Solar PV having quite the positive ERoEI.

24 02 2014

I think ERoEI is generally UNDER estimated for most systems….. the very point of this latest study that includes a whole of new inputs hitherto left out of calculations…….

24 02 2014
Daniel Koeppel

Shows again how important the clear stating of “system borders” is for everyone – pro and con. : )

24 02 2014
Susan Krumdieck

Isn’t there a simple bottom line? How big of a solar PV power facility plus a wind farm would you need to run a solar PV manufacturing factory, and aluminum smelter and silver mine and low-iron float glass manufacturing plant?

24 02 2014

Hello Susan, great to see you weighing in on my blog…! I’d love to hear what you think on many issues here.


25 02 2014

Right on cue, Eclipse Now tilting at their own personal windmills. Mate, I just don’t understand what your purpose in commenting here is. You do not seem to realise that you will not change anyone’s opinions. I think what you fail to appreciate is that the majority of readers are aware of your incessantly touted techno-hubris “solutions” already, we have just dismissed it as more nonsense by people who won’t do the maths!

25 02 2014
Susan Krumdieck

One of my former PhD students has been carrying on his research on EROI and energy costs of energy production technologies at Stanford University for the last 3 years. Several of his articles are available in open-source journals. He will be the new leader in the field.
Michael Dale.
He is very thorough in his data gathering and analysis. No need for armchair experts when we have Michael on the job.

10 03 2014
Graham Palmer

I came to same conclusion as Prieto and Hall, the main issue is one of boundaries. See


13 04 2014
The 5 key elements of sustainable transport | Damn the Matrix

[…] unless that car is 100% solar recharged.  And then I’m doubtful, because since we now know solar has a shockingly low ERoEI, it might be even closer than we think.  I’m also surprised cycling’s numbers are as […]

22 10 2014
The Energy Cliff Revisited | Damn the Matrix

[…] ERoEI somewhere around 20:1.  This is because it was made in 2009, and in the intervening 5 years, it has been established that it is fact less than 5…. maybe even less than 3!  This is displayed more accurately in the more recent chart […]

30 10 2014
How Solar PV Can NOT Power A Carbon-Free Energy Revolution, In 4 Charts | Damn the Matrix

[…] because such a short energy return implies an ERoEI of 25:1, when in fact Pedro Prieto and Charles Hall recently calculated that it was more like 2.5:1, but what’s one order of magnitude among […]

19 03 2016
Explaining the energy cliff | Damn the Matrix

[…] consider this……  to use the above primary energies to make PVs with an ERoEI of 2.45:1 gives us nett energy efficiency of […]

10 05 2016
Another study on the ERoEI of solar PV | Damn the Matrix

[…] posted by Euan Mearns on his blog Energy Matters, this study makes Pedro Prieto’s look very good….. the differences in ERoEI between the two studies must be a function of the […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: