Net Energy Trends

In writing Wednesday’s post ERoEI for Beginners, I prepared a number of charts that were not used and these are presented here. Where it has been measured and according to the literature, the net energy of oil, natural gas and coal is falling everywhere. Surface mined US coal has one of the highest energy returns of any fuel and is substantially higher than deep mined Chinese coal. In electricity equivalent (Eeq) form, Chinese coal is marching towards the Net Energy Cliff edge while US coal remains far from it. The image shows part of a 50 km long queue of coal trucks in China.

Estimates of ERoEI for solar PV are all over the place (1 to 12) because different analysts set different system boundaries, the energy return is latitude and site specific and its possible that the literature based on historic deployed panels is not up to date with most recent advances.

Sugarcane ethanol in Brazil has ERoEI of 8 to 10 at the refinery gate which at face value seems OK [1]. But to be equitable with its FF cousins this needs to be reduced to 3 to 4 in Eeq form and is barely viable. Temperate latitude biofuels are not viable in liquid form at the refinery gate and converting them to Eeq cripples them completely. But I suppose burning them to make electricity is no less crazy than burning them in an internal combustion engine sat idle at traffic lights.

Most of the data in the following charts comes from Hall et al 2014 who summarise ERoEI research for a variety of fossil fuels and renewable energy flows (see table below) [2].

Oil and Gas

Note that this chart combines crude oil and natural gas and the data are for production which I assume to mean at the well head. Note that there is also a wide range of dates, which is part of the point of this chart.

Note how the ERoEI of world oil and gas production is deemed to have fallen from 35 to 18 in just 7 years. I’m not sure this is credible. USA production is deemed to have fallen from 30 in 1970 to 11 in 2007. Canada from 65 in 1970 to 15 in 2010. It is very true that more men, machines and energy are being used to extract oil all over the world and this has pushed the cost of extraction higher as ERoEI has fallen. Or is it high price that has encouraged companies to expend more effort? 😉 And those thinking that the price has collapsed need to be aware that the cost of extraction has not collapsed and this translates to massive losses for producers. The trend of falling ERoEI is certainly real, the extent is open to debate. But if it continues, increasing amounts of our human resources are going to be spent on oil and gas production.

In my previous post ERoEI for Beginners I introduced the concept of electricity equivalence (Eeq). This is a first step towards trying to normalise different energy sources to a common datum. Crude oil at the well head is not much use directly to anyone. But it can be used to make electricity and in doing so roughly 62% of its energy will be lost as waste heat (BP convention). This normalisation enables direct comparison with renewables and nuclear, that have the advantage of producing electricity directly.

Following this convention, oil and gas production in the USA and China has already fallen off the net energy cliff while global production is getting close to it.

Natural Gas

In 2005 (pre-fracking) US natural gas had high ERoEI of 67. Freise (2011) charts the decline in Canadian gas ERoEI from 38 in 1993, to 26 in 2000, to 20 in 2009. This march towards lower ERoEI in Canada is sending Canadian gas Eeq towards the net energy cliff edge. Large quantities of natural gas in Canada are used in tar sands extraction and upgrading. High ERoEI gas is being traded for liquid fuel that has ERoEI of about 3 (own calculation based on published Canadian statistics).


Hall et provide ERoEI data for the USA and China. US coal has very high ERoEI of 80 in 1950 and 60 in 2007. The fall is only slight and that is because mining methods have not changed very much. US coal is mainly surface mined from vast surface deposits in the Appalachians and Wyoming. By comparison, Chinese coal had ERoEI of 35 in 1995 and 27 in 2010. The lower ERoEI of Chinese coal to large extent reflects underground mining compared with surface mining in the USA. The Chinese need to apply vast effort to extract and transport coal to drive their economy.

The open circles show the electricity equivalent values. The mine mouth values are reduced by 0.38 and no deduction is made for transport. We see that US coal Eeq is probably one of the highest energy value sources we have but it is being forced out because of concern over CO2 emissions. Part of the problem here is that US coal is too easy to produce making it a fuel of first choice for exploitation. Chinese coal Eeq is getting closer to the Net Energy Cliff edge.

Renewable Electricity

Hydroelectric power

There seems to be agreement that Hydro Electric power has high ERoEI. A large amount of energy is invested at the start in excavation, concrete and generating kit. But thereafter a dam may produce electricity for 100 years or more with little operation and maintenance energy costs. Although many dams today are getting fitted with new more efficient turbines, which means a new energy and economic investment.

High altitude wind

As explained in my earlier post, high altitude wind has potential to multiply the ERoEI of ground based wind turbines. The argument here is rather complex and will be explained in greater detail in a separate post. Suffice to say that the mass of the KiteGen, and hence energy embedded in the materials, is a fraction of that in a large turbine. This is specific to the design concept.

Wind turbines

The ERoEI of a wind turbine is site dependent. A good windy site will produce more energy over the life of a turbine in a calm site. The wind industry has tended to focus on sites with good wind resource and so site specific factors are less than for solar. A large number of studies places the primary ERoEI of wind turbines in the ballpark 15 to 20 and there doesn’t seem to be too much disagreement on that.

The contentious issue for wind is treatment of intermittency. Is there an energy cost associated with that? Of course there is. Broadening the ERoEI boundary to create dispatchable power substantially reduces the ERoEI. At present this economic cost is not paid by the wind producers but is borne by others. Weisbach et al [3] reduce the primary ERoEI of wind by a factor of 4 to generate their buffered ERoEI assuming that pumped storage hydro is used. But they rightly note that using FF balancing services would have a lesser impact.

Solar Photovoltaics

With a range of ERoEI from 1 to 12, anarchy reigns in the PV ERoEI business. There are a number of issues at play here. The first is that different energy boundaries are being used. I personally favour a wide boundary that includes direct energy use, materials and labour. And for intermittent technologies a reasonable energy cost needs to be apportioned to mitigating that intermittency. The second is that solar PV is site specific. A sunny tropical site may yield three times the lifetime energy of a cloudy high latitude site. The third is that the efficiency of PV is improving all the time. Mixing these factors to varying degrees underpins the anarchy. But adding battery storage to a good tropics-based system is going to substantially reduce the ERoEI. Proponents of Solar PV seem set to continue to promote optimum performance without backup while others will observe that normal performance is sub-optimal and that in the real world the sun does not shine at night.

Liquid biofuels

Sugar cane ethanol is made in the Tropics where there is more abundant solar energy. And sugar cane employs a more efficient photosynthetic route than maize to manufacture carbohydrate more efficiently. No fossil fuel based fertiliser is required and the bagasse (left over organic cane waste) is combusted to make electricity in the refinery. These conditions combine to give sugar cane ethanol a viable ERoEI of 8 to 10 at the refinery gate [1]. Too bad about the disappearing forest.

I have tried to develop an equitable way of comparing different renewable and fossil fuel based energy sources by reducing all to electricity equivalent. And I’m afraid that sugar cane ethanol cannot escape that net. In Eeq form, sugar cane ethanol falls off the Net Energy Cliff.

The temperate latitude biofuels (corn ethanol and biodiesel) are really not worth discussing again.

Concluding thoughts

The energy debate continues to be partly driven by emotion and not by the laws of physics and needs of human society or nature.

My own opinion is that understanding ERoEI is vital to the continuance of industrial society as we know it. That does not mean projecting economic growth infinitely into the future but managing the energy and human resources and natural resources we have in a rational and responsible way. One that optimises benefits for humans and nature.

Understanding the intricacies of the human energy web is enormously complex and requires substantial resource to fully understand it. But it is not nearly as complex as understanding the climate system and I would argue that understanding our energy web is far more important. And so here is a challenge for the United Nations. Establish 10 working groups globally to study the human energy web, deliberately choosing ones with opposing outlooks at the outset (fossil fuel v nuclear v renewables v collapse) and challenge them to model the human energy web and to explore the multitude of outcomes.

And to then use that information to answer if asking Africa to skip over the fossil fuel stage of their energy development is beneficial to Africans?


[1] Jose ́ Goldemberg , Suani Teixeira Coelho, Patricia Guardabassi: The sustainability of ethanol production from sugarcane. Energy Policy (2008), doi:10.1016/j.enpol.2008.02.028

[2] Charles A.S. Hall n, Jessica G. Lambert, Stephen B. Balogh: EROI of different fuels and the implications for society: Energy Policy 64 (2014) 141–152

[3] D. Weißbacha,b, G. Ruprechta, A. Hukea,c, K. Czerskia,b, S. Gottlieba, A. Husseina,d (Preprint): Energy intensities, EROIs, and energy payback times of electricity generating power plants

This entry was posted in Energy and tagged , , , , , , , , , . Bookmark the permalink.

50 Responses to Net Energy Trends

    • gweberbv says:

      Here you find the links to the documents:

      Assuming a capacity factor of 50% is very optimistic. In particular over the full 20 years. Then they assume that 20% of the originally invested energy can be recovered by recycling. Might be realistic if the wind plant is repowered after 20 years, so that foundations/substation/etc. can be reused.

    • Euan Mearns says:

      If it was that high wind power would not need subsidy and would be so cheap that a portion could easily be cycled through a chemical storage loop to smooth intermittency.

      • Why would they? There is no legal requirement to store energy or to smooth intermittency.
        I don’t see how you can assume a low EROEI on that thinking.

        Really looking forward to that KiteGen article! Can you say when it will go online?

        • Euan Mearns says:

          There is no legal requirement to store energy or to smooth intermittency.

          UK energy secretary Amber Rudd has indicated that there may be a legal requirement for purveyors of wind and solar to meet the cost of intermittency in the UK some time soon.

          ERoEI and cost of energy are inversely correlated. If you had an oil field with ERoEI of 60 the government would be demanding a share and taxing you at 70% on the rest.

          • Doug M. says:

            “ERoEI and cost of energy are inversely correlated.”

            It’s a very loose correlation.

            Here’s a toy model. Imagine a wind turbine that has a lifespan of 20 years and an EROI of 10. Now imagine an otherwise identical wind turbine that has a lifespan of 200 years. It would have an EROI of 100, right?

            Okay, so would the second turbine cost ten times as much? Who knows? Probably not, though. Note that the average lifespan of wind turbines seems to have increased significantly over the last 30 years, even as costs have fallen.

            Well, would the second turbine be *valued* at ten times as much? No, it wouldn’t, because of the time value of money. A megawatt delivered today is worth more than a megawatt delivered a year from now, and a megawatt delivered 200 years from now has a value very close to zero. Ask an accountant and he’ll start talking about discount values and depreciation, but the long and short is that a sensible buyer might pay around twice as much for the 200 year turbine — probably not three times, and certainly not ten times.

            Toy models, but you get the idea. Letting price serve as a rough pointer towards the general direction EROI is okay. Thinking that price correlates closely with EROI will get you in trouble.

            Doug M.

          • @Doug M.
            I disagree, ROI and ERoEI are the same stuff, with a different currency.
            ROI is much complex because conventional currency are different per country, subjected to change rates, inflation, debt accumulation, and the nominal value change during the time etc.

            ERoEI is based to the Watthour currency and it unaffected by time, and the debt is meaningless.

            If the analysis is done in a precise and limited timeframe having a money relatively stable, it is meaningful to make all the updated conversions between Watthours and $,£,€ and the result is fully consistent.

            There are no difference in conversion between green Wh or black Wh and money, because when we trade energy, the energy mix is the reference. (i.e. a taxi driver when will pay his house (green) energy bill, he have already earned money through oil)

            A residual complexity is in the conventional exclusion of the human (endosomatic) energy in the ERoEI balance, but this is a complexity similar to the difference between ROI and IRR, then all energy projects are very capital intensive and the endosomatic energy is always negligible.

        • robertok06 says:

          “Why would they? There is no legal requirement to store energy or to smooth intermittency.”

          Leaving legalities aside, which do not belong to ERoEI, if one day intermittent renewables are expected to cover most of the energy production of the planet, then storage of their excess energy will be mandatory… and the LCA of any electricity-generating technology MUST keep that into account… if ERoEI is to be calculated.
          Germany’s wind and solar have covered, in 2014, 2.8% of the energy of the country… clearly storage and related energy costs and effects on ERoEI calculations are minimal… but it doesn’t take a rocket scientist to imagine that as soon as the penetration grows storage become mandatory.
          It is therefore scientifically naive (wrong, actually) to choose such technologies over others on the pretense that their ERoEI is high (wind, PV is already out of the question), when applying this conclusion would lead to the need of such a massive installation of them which would impact in a dramatic way their ERoEI.
          It is like the owner of an electric car living up a mountain, he could say “I don’t need any recharging to go to work, as I can use the downhill road to my office to recharge the batteries with the regenerating brakes”… leaving out the fact that he will need then a lift up home in the evening. So far, in this parody, the EV owner (intermittent renewables) have gotten a lift from the other baseload sources, FF, nuclear, hydro, biomass… but clearly IT CAN’T go on forever, it would go against several principles of physics.

          Mankind consumes electricity (energy in general) 24h/24, every day of the year, so it makes no sense to quote the ERoEI only when the wind blows (wind) or it’s not night or winter (PV, solar).

          The correct approach is the one utilized in Weissbach et al, the “buffered” ERoEI, and that kills both wind and solar.

          More worrying is the fact that intermittency and the lack of large-scale storage options with small losses and tolerable environmental impacts will keep on killing them, no matter what technological improvement they may have in the future.


          • werner says:

            Again such a boring and non funded comment. All non electricity needs of energy consumption allow a relatively big load shifting, and the effect of big grids is also always ignored, reducing the need for storages with rising grid size, so the storage needs falls below the size of existing storages at some size – which is really big, but far from being impossible to build, and far rom being too expensive.
            See e.g. and be aware, that losses can be reduced by using more aluminium to any degree wanted. And the t the life expectancy of a grid is very, very high, especially if the life expectancy is a design parameter.
            Or see this :
            or this or many other documents. the Mathematical effects which lead to a reduction of storage needs with growing grid sizes are inevitable.

          • Willem post says:

            It likely will not be feasible to store solar energy in the summer for use in the winter.

            During summer wind energy is much less than in winter, so it would need to be stored in winter for use in summer.

            Renewables is not just about electricity.

            Please see my below comment regarding the required acreage for replacing annual crude oil production with corn kernel production.

  1. Willem Post says:


    The lack of agreed on boundaries is like having traffic without rules. Hence almost all ERoEI values are ranges

    Adjusting for electricity equivalency introduces conversion issues which are technology and country dependent.

    Not all gas is burned on 60% efficient CCGT systems; in Ireland, that efficiency gets reduced to 50%, or less, due to peaking, filling-in and balancing services.

    Not all coals burned in 43% super-critical power plants.

    I made a comment on your earlier post. Here is in somewhat expanded from:

    Overlooked Usefulness of Fossil Fuels

    Eliminating fossil fuels from electricity production is one thing (expensive, but not all that difficult), it is quite another to eliminate fossil feed stocks from the thousands of industrial processes that use fossil feedstocks, and from the millions of products that consist of fossil feedstocks, such as plastics, drugs, etc. What would be the substitutes for the fossil feedstocks? Electricity? Plant material? That second, very important part, is usually ignored by most RE proponents, as they are much more interested in the pursuit of subsidies to build wind and solar systems using the slogans of “saving the world, fighting global warming, reducing climate change”.

    Plant material Replacing Crude Oil? To replace the Btu value of annual crude oil production with the same Btu value as corn kernels, about 2.949 billion acres would be required of the 12.14 billion acres currently used for food production, of which about 0.28 x 12.137 = 3.4 billion acres is in annual crop production, i.e., about 2.949/3.4 = 87% of crop land would be required for corn. Corn replacing coal and gas would require additional acreage. However, there is no equivalence, as crude oil, coal and gas are much more useful for various tasks than corn kernels.

    Corn crop: 160 bu/acre/y x 56 lb/bu x 7000 Btu/lb x (1 – 0.15) = 53,312,000 Btu/acre/y, equivalent to 9.6 barrels of oil/acre/y.
    World crude oil production = 80 million/d x (1 – 0.03) x 365 d/y = 28,324 million barrels/y.
    Land area in corn = 2.949 billion acres.
    World land area for food production = 18,963,881 sq mi x 640 acre/sq mi = 12.137 billion acres,

  2. I am not so sure about that.
    20 years is also rather pessimistic for new turbines. Some people in the industry have sayed already they rather estimate possible service life in excess of 30 years.
    With new turbines underated and optimized for lower cut in speeds CFs around 50% seem reasonable.

    I’d rather see some real data from 2014-2016 turbines of course. The wind industry is pretty innovative.
    Historical data is interesting in its own but not relevant for the current generation of turbines.

    That sayed I am a KiteGen fangirl from the start and would love to see them fly as soon as possible.
    There is so much going on in high altitude and wind recently, seems the developement took up speed the last 1,5years.
    Lots of new companies springing up recently. E-Kite just won an award at the Energy Fest Startup Event.

    • gweberbv says:


      the wind turbines that are optimized for weak wind locations are basicly huge turbines with small generators. Thus, they need more material (=energy) per kWh hour produced than a standard wind turbine. In that case, the very nice capacity factors that are possible with these turbines are misleading.

      • So where is the seeetspot for

      • So where is the sweet spot for rated capacity/real capacity/energy input and output?
        Do you know the turbines in the Siemens LCAs are not oversized?
        I guess turbines have improved much and EROEI has been rising with each newly introduced turbine.

        • gweberbv says:


          Siemens assumes CF of 50% and an average wind speed of 8 m/s. For a weak wind turbine you would expect a higher CF at such a marvelous location. See here:

        • Jenny,

          I’ve read about SWR and I agree with gweberbv.

          For structural dynamic reasons if a wind turbine grows in dimensions, it see the materials and the cost grow raised with the 3.6 power.

          A 20% increment in height imply about a doubling cost and a little CF improvement.

          The reason fall in two aspect:
          One, each scaled object grows with cube of a linear dimension.

          The the additional 0,6 in the exponent is justified by the extra structural material, that is needed to maintain the first modal frequency enough stiff or high or the safety factor.

          This is to avoid that the blades under erratic load hit the tower, or the tower accumulate energy during random wind gusts concerning the foundation fatigue.

          Around the classical design of 1,5MWn the HWT always diverges from the optimum possible ERoEI.

          I suppose that nothing has changed, apart the variable speed concept, about that in the last 15 years.

          • Euan Mearns says:

            Massimo, do you have historic data for wind turbines – height, rated capacity, mass of structure etc, for example a series of Vestas turbines?

            In England there was a big move offshore a few years ago where now I think over half the installed capacity is at sea. But I don’t think the fleet capacity factor rose. Perhaps one of those myths that the wind blows more strongly and more constant offshore? Or it could be that maintenance is more difficult offshore, so turbines sit broken for longer.

          • gweberbv says:


            you do not need to speculate if dozens of billions are invested in offshore wind because of a ‘myth’. Here is nice little webpage, that provides data for the Danish offshore fleet:
            Note that over the years the wind farms moved farer and farer away fromt he shoreline, achieving increasing capacity factors.

          • Stuart Brown says:

            Greater Gabbard is 504MWp (140*3.6MWp). RWE says:

            “The turbines generate around 1,749,000 megawatt hours energy per year”

            or a CF of 39.6% by my reckoning or “at least 40%” by theirs. Near enough.


            You can see the current output if you trawl around on the site, which claims a current peak capacity of only 252MW going flat out, so not sure what that means. I couldn’t find historical data. 🙁

          • Thinkstoomuch says:

            For gweberv,


            We now know the values for that general area. How does it apply ~1000 km to the west? Which is what Euan was talking about. Wonder if having a lot more distance an depth to build up waves might impact maintenance requirements?

            Or even how does that equate to what the predicted levels were before the Danish installations at those locations.

            I would think, absent data to the contrary, that it is a lot like comparing southern CA PV with southern Florida. Some reason they get 20% more sun. Despite South FL being farther south.

            A related article I found yesterday.

            Only deals with onshore but shows a lot of the variability over geographic distances.

            Have fun,

          • Euan,
            all the story about that is making me crazy. We must avoid to be fooled by CF details, this isn’t the main problem.

            Lets analise your new Beatrice offshore wind farm vs. coal.
            86 Siemens 7MW turbines, 4,59M€/MWn
            diametre 154m, 18600m^2
            a total of 588 MWn, €2,7 billions investment.

            about 5TWh is the energetic cost of the turbines manufacturing (30GJ/ton) and about 5TWh is the energetic cost of the lifelong maintenance.

            in the optimistic 20 years are expected 50000 hours realistic or 80000 claimed of availability, also said 30 – 47 TWh
            Then 20 -37 TWh of net energy totally produced.

            electrical production from coal today is less than 10M€/TWhe ($1/GJ) So 200M€-370M€ worth energy payed with an investment of 2700M€.

            the Scottish have to refund the difference between coal and WT that is 2500 – 2330M€.
            Due the energy intensity of the Scottish people, each £ earned it cost in primary energy 4kWh/€ then additionally a total of 10 – 9 TWh

            In the electric bill (the PPA + GC) Scottish people have to pay to Beatrice also the lifelong maintenance, other 2-3billions and 10TWh (3%/year), then the decommissioning….

            Finally, thanks only to Beatrice, the Scottish People renounced to more than 5000M€ of discretionary expense having the same energy and the same carbon footprint of coal.

            Check and adjust the figures as you want but nothing will change.
            how you say?
            Scottish people are screwed?
            Don’t worry Italians too.

          • gweberbv says:


            here you find data for UK offshore:
            See page 13 for capacity factors.

            For the German offshore fleet in the North Sea, one can roughly estimate capacity factors from the data presented here: (marking only solar and wind). I come up with (monthly) values between 75% to 40% for the first five months of this year.

          • Euan Mearns says:

            Thanks for the links Gunther, esp the Crown Estate report.

  3. Ajay Gupta says:

    Concerning the trends in oil production, which is actually oil and gas as stated, don’t forget that the IEA and EIA changed the definition of oil to include shale and tar sands. Might explain some of the rapid drop in EROI in such short time periods for Canada and US. I write this in haste without checking the numbers behind the graphs. Just a thought.

  4. Euan: A fundamental question.

    Why is there an “energy cliff”? Why don’t resources just get gradually less viable with time as the higher-quality ones are used up? Why should they all abruptly decrease to zero?

  5. Beamspot says:

    Just an stupid question, that I answered other times, with very few responses.

    What about ERoEI of direct action renewables?

    Solar heating being a good example. China is plenty of SWH systems, probably more than 10 times the installed PV, but it seems only PV catches the eye of the green dreamers, while SWH (as hydro) is deprecated or simply forget.

    I wonder why, as well as I wonder they probably have higher ERoEI (and by far) than PV, while supplying the form of energy more demanded.

    • Greg Kaan says:

      The conversion of energy to electricity equivalent makes SWH look ineffectual. I’m not saying this is correct – the electricity equivalent measurement seems designed to favour PV and wind turbines.

      • Willem Post says:


        Oversizing a household, roof-mounted, PV system for space heating/cooling and DHW with heat pumps and charge EVs, is the way forward for very energy-efficient, single-family houses.

        About 10-15 kW of panels would be required, and a 250 to 300 gallon, well-insulated DHW tank, if grid connected.

        If off the grid, a battery system and 3-5 kW generator would be required.

        Using roof-mounted, solar thermal heating systems would be significantly more expensive, due to the low cost of panels.

        Whereas, in the past, this made sense in China, when panel prices were high, it is no longer the case.

        • Graeme No.3 says:

          10-15kW of panels? That is a big roof area. What size house are you considering?

          • werner says:

            15kW, 150W/m² to 200W/m² means 66-100m². East-West roof, roof 0,5m bigger in each direction than walls, roof fully covered with panels (more usual today, looks better), this means something like 9×8 to 7×5 m dimensions of the house on the ground, so a rather small house.

          • Willem Post says:


            10 kW @ 300 W/panel = 33 panels
            3 rows of 11 panels
            up it to 12
            panels 2.5 ft wide X 5.5 ft high
            10 ft width for 4 panels
            30 ft wide x16.5 ft high panel area

            My house roof area is 44 ft wide x 22 ft peak to eave
            If garage roof were aligned with house roof, more area would be available.

          • Beamspot says:

            I wonder how this will fit onto that high amount of people that lives in big towns like Madrid or Barcelona. Even in my old flat in a Little village in Spain, we have roughly 70m2 allowed for 5 families. Clearly a no go for a big part of population in our sunny country.

      • Beamspot says:

        Interesting point. Let me do some sanity check, as I did one year before.

        This is one SWH I can use, with all HW and 150l tank. I’m beginning to think to install one at my new home.

        1185€ for the whole pack, less installation (I’m capable to install it, though).

        Now, let’s try the same with in the same supplier, but with PV:

        Back of the envelope calculations, 2m2 for the previous heater, with an estimated 60% efficency (probably higher), we have about 1200Wp.

        We need five of those:

        That is 1175. Almost the same than the whole pack, but with less things (one year ago they were more expensive).

        Then we have to add the invertir:

        And the heater tank:

        Plus hardware (fittings and such), this is below 2000€. The last time I’ve checked, that was in the 2700 – 3000€, so certainly there is something to re-think on that.

        But I wonder if I have 2m2 of SWH, and about 10m2 (more or less, but that depens on delta temperature, thus latitude), so we need five times more glass, aluminium and such. And I wonder solar cells are more expensive than the aluminium/copper fins into the SWH.

        Plus the inverter has lots of ‘exotic’ elements inside, about 70 of the 92 elements of the periodic table, so it is more resource intensive than SWH, thus the price seems misleading to me.

        Hence the question why nobody never calculated the ERoEI of SWH or other direct action renewable energies.

        The conversion of all kinds of energy to electricity is a no brain for me. But I’m a quite biased, strange nerdy guy that tends to take showers with warm water instead of electrons… After all, what we want the energy to? And, which kinds of energy we really need?

        Since I’m a nerd, I took all my bills, so I calculate my energy consumption for my house, and I found that about 70:30 or even 75:25 of heat;electricity, so I can’t understand why this obsession with electricity.

        Of course, our greedy concept of business selling electricity to the grid, plus our knee bending in front of electronics may bias our perception. After all, I’m an electronics engineer that worked as R+D engineer in Electric Vehicle in Germany, so maybe I’m vaccinated about that…


        • Willem Post says:


          About 5 years ago, an installer recommended a vacuum panel, solar hot water system with 250 gallon storage tank (electric heater in tank for back up) that would last about 15-20 years, at about $7000 turnkey cost.

          Recently, I told him I would like to use PV panels to heat a 250 gallon DHW storage tank.

          He recommended a 1.2 kW PV system, 4 panels @ 300 watt each, with tank at $7000 turnkey cost that would last at least 25 years.

    • gweberbv says:


      I agree that at first sight the ERoEI of solar thermal is probably significantly higher than that of solar PV.
      But it is very hard to really profit from this high ERoEI as you need to use exactly this amount of heat that was just generated (or was generated a few hours ago). In Germany it is now nearly impossible to build a new house with a heating system only relying on FF. You must add solar thermal (or PV). For solar thermal as a rule of thumb only 50% of the harvested energy (=heat) is used in the end. And if the need for warm water is unusally low, then this can go down even more.

      In contrast, electricity offers much more possibilities to use it. In particular, one can feed it into the grid.

  6. Will Fischer says:

    Roger Andrews, it looks like a cliff because of the definitions, because of what you’re plotting, and because you have a linear x axis. It would look quite different if the x axis were plotted on a logarithmic scale. The graphs above do not include a time axis, but what is perhaps most important is how quickly over time the EROI declines. Unfortunately, the real decline rate for most of civilizations energy inputs with time depends on many factors and is not well understood.

    If the decline rate with time were to slow markedly as the absolute value of the EROI decreases, the cliff would not be so worrisome. On the other hand, it’s easy to imagine situations in which the EROI would decrease abruptly to zero. (That’s what happens in the closed universe EROI of my driving my automobile: up to a certain point a very small investment of energy for my foot on the gas pedal produces a very large increase in the car’s kinetic energy return, but when the car runs out of gas, that EROI immediately becomes zero (and the net energy ratio therefore becomes infinitely negative). How fast will the oil run out? How much more costly will it become to mine iron or rare earths or uranium or whatever we happen to need at the time? It all depends. And no one knows.

    • Willem Post says:


      I remember gasoline at 20 c/gallon in the US in 1955.
      Now it is about $2.50/gallon, or 2.50 *1/(1.02) to the power 65 = 2.50/3.62 = 69 c/gallon, inflation adjusted.

      The 69 c is not quite comparable to the 20 c, due to increased efficiency of producing crude and processing it over the past 65 years, i.e., the 69 c would be higher, if no efficiency increases.

      The energy items can be analyzed on the same basis.

      • Willem Post says:


        Revision to above comment:

        Gasoline was 23 c/gallon in 1955,
        Inflation factor was 8.93 for the 1955 – 2016 period
        Inflated price 205 c/gallon
        Actual price 250 c/gallon
        Excess 22%*

        Even though crude oil production has greatly increased to about 80 million barrels per day since 1955, so did production/processing efficiency, which kept the excess to 22% for that period.

        Even though crude oil ERoEI significantly decreased, that is barely reflected in the price at the pump.

        *Cars had an average mileage of about 15 MPG in 1955, about 26 MPG in 2016.

        Most farm products, such as eggs, coffee, milk, decreased in price.

        • Willem post says:

          Addition to comment:

          The above indicates, whereas the world average ERoEI (a dubious, fallible academic construct) for crude oil production declined, A to Z efficiency improvements kept the economic impact at the gasoline pump to about 22% over a period of 65 YEARS.

          This is truly incredible.

          Please take note, before listening to rediculous, scare-mongering about “falling off the cliff”.

          No such thing is about to happen for a long time.

    • sod says:

      ” The graphs above do not include a time axis, but what is perhaps most important is how quickly over time the EROI declines. Unfortunately, the real decline rate for most of civilizations energy inputs with time depends on many factors and is not well understood. ”

      I agree, this is an important point.

      in the past, we had massive increases in use of electricity. this required a fast expansion of production, which requires high ERoEI.

      Let us assume a 1:3 ERoEI system, in a environment, that is doubling or tripling electricity use fast. The electricity needed to build the new systems, come from the old system, which only allows the building of 2 extra systems (if all electricity is used to build more electricity production).

      In many first world countries today electricity use is growing less fast (or might even sink). So high ERoEI systems from the past can easily build a new generation of lower ERoEI systems.

  7. Colin says:

    If the ERoEI of US coal is so high then how does it fail to compete with shale gas? Bear in mind that shale gas receives no subsidies and is vilified by the all powerful green lobby. It still competes with coal. The decline in American CO2 emissions has also been attributed to a switch from coal to gas fired electricity. No one disputes that it takes more energy to extract shale gas than conventional. That an even lower ERoEI source of gas is driving out coal suggests that the true ERiEI of gas and oil is higher than made out here.

    • sod says:

      “If the ERoEI of US coal is so high then how does it fail to compete with shale gas? ”

      If a kite has fantastic ERoEI, while wind turbines have a basically useless one, why are there thousands of the one and none of the other? Even subsidies should not distort this so badly.

      ERoEI obviously has some problems.

      • Euan Mearns says:

        I think you have to look at the time line of technology development. Ground wind power has been around for several hundred years as a source of energy for Man. Wind turbines have been around for perhaps 50 years. I think Massimo was perhaps the first Man on Earth to generate significant electricity from a high altitude kite perhaps 10 years ago. And commercial deployment is dependent upon state of the art motion control technology as deployed in drones.

        • I would like to take the opportunity to tell the challenge facing the concept of tropospheric wind energy exploitation.
          The control isn’t any more an issue, it is done and it is scale independent.

          The common opinion is that KiteGen should gradually scale up prototypes it has already made in the past in order to achieve step by step a significant performance.
          This approach could be quite easy with respect to the ground machine.
          The large driving robots, for electrical generation and take-off/landing operation, is simply a special machine with clear requirements and continuously scalable specifications.

          But there is a technological barrier, a deep step, in passing from small fabric wings with low efficiency to the semi-rigid composite wings. A due evolution,likely out of any more discussions.

          First, there is a problem of balance between the density of the materials adopted and the size of the wing.
          The composite skins have been taken to have a minimum thickness below which it is currently impossible to go.

          So this also obliges to a minimum wing dimension that can only be what we have already calculated and manufactured.

          We are talking of at least 130 square meters of wing to have a good weight/aerodynamic ratio, mandatory for easy take-off and landing operations.

          This wing appears to have also a very high aerodynamic efficiency, certainly more than 28.

          This technological monster is able to fly over 80 m/s and to develop axial forces higher than 300 kN with relatively weak winds.

          This barrier, that consist to evolve from toys directly to have a powerful thruster, was met by all the labs in the world that have reproduced our approach starting from the textile wings.

          In fact, the maximum power that we all have been able to achieve was a hundred kW, often followed by the fabric wing or some other piece of equipment breakage.

          this movie show an unfortunate 50 sqm fabric wing:

          the link and the description is available here:

      • Coal:
        The problem of coal is a communication problem, because is the currently greenest method to massively produce electrical energy even including externalities, Ferroni/ Hopkirk just remebered us that: “modern coal plant emits fewer greenhouse gases per unit of electricity than does a photovoltaic plant.”
        Despite this fact easy to check autonomously, solar PV is subsidised everywere and Coal heavy penalised.

        Kite energy:
        Sorry, we just “invented” it. Is true that we talking of this opportunity since 2003, but at the beginning was only an hobby aside other jobs and without serious resources involvements. I was professionally a researcher in IT and NMP (info tech and new material and production)
        with several tens of running public funded research projects in different fields.
        Due this experience we answered also to several call of proposal about base research on renewable energy, unfortunately no one was accepted both in Italy and in Europe, here one of the first:

        online reader:


        We continued to study the issue and realize and operate prototypes without public support but gaining an experience and a clear view tanks only private support.
        Only few months ago we declared the industrial and the technological research completed (including subsystem validation at full scale TRL8), having satisfactory answered to all open issues.
        Unfortunately the problem is still to explain and disseminate this brand new domain, that seem to be much more tricky than rocket science.

        Our current need to disseminate is to call fresh forces around the project that is become quite complex and require an wider organization, and is indeed one of most exciting multidisciplinary technology never seen.
        But we have to start again from the basis: Why we need cheap and abundant energy?
        Even fighting on this to Paul Ehrlich’s bizarre position.

  8. Pingback: The Energy Return of The Three Gorges Dam | Energy Matters

  9. Pingback: Kite Wind / KiWiGen (generatore eolico alimentato dal movimento di aquiloni): QUI TUTTE LE DOMANDE E DUBBI - Pagina 51

Comments are closed.