This will be the last in the series of blogs on EROI that started with John Morgan’s guest blog (November 4). The set of data in all three blogs was based on a paper by Weißbach et al. that suggested an economic threshold below which sustainable energy sources cannot be used in the substitution of fossil fuel, thus calling into question the whole issue of a feasible energy transition to non-carbon-based energy sources. One of these sustainable energy sources is biomass, as exemplified by ethanol (drinking alcohol) fermented from corn. Its EROI value (buffered or un-buffered), as put forward in the Weißbach paper, was 3.5 – well below the stated economic threshold. I have decided to focus this blog on the debate around this value, which started as soon as the United States legislated a requirement to include a statement of the amount of alcohol in all automotive fuels being sold within the country.
Some of the characteristics of the alcohol (biomass in the figure) are not obvious, but they are both interesting and important in the climate change debate. Many of the characteristics, once the details are explored, also shed considerable doubts (at least in my mind) on the main EROI issues that we discussed in the last three blogs.
Why is biomass among the assortment of sustainable energy sources?
The burning of ethanol generates 6,743Btu per 1Lb of emitted carbon dioxide, while the burning of natural gas generates 8,751Btu for the same amount of emitted carbon dioxide – these figures are not much different yet one is regarded as sustainable while the other is regarded as a polluting fossil fuel. The reason for this is that the designation refers to the full cycle of production and burning. The ethanol is fermented from a biological crop grown annually; as with all plants, the corn uses a solar-powered photosynthetic process that sequesters the carbon dioxide from the atmosphere. The burning process returns the same amount of carbon dioxide to the atmosphere while releasing the stored energy. That means over the full cycle, the energy used actually all comes from solar energy, and presents no changes to the chemistry of the atmosphere.
In the case of natural gas, a similar process takes place – except that fossil fuels were formed hundreds of millions of years ago from the decomposition of dead plants and animals. This time lag completely separates the sequestration process from the combustion process, thus making it unsustainable.
If the energy that we use in the combustion of biomass comes from solar energy why it is not considered an intermittent source of energy like the other solar energy forms (solar PV, wind, solar CSP and hydro) which require storage (the difference between the buffered and un-buffered in the graph)?
This is an excellent question with a somewhat hazy answer. The main reason is that in a place like the United States the alcohol is made mostly by fermenting (bacterial digestion of the sugar) corn. Corn is also used to feed people and animals. This forms a tight connection between the food supply and the energy supply. In a sense, the food supply acts as the storage component of the energy supply. Unfortunately, it creates major issues for the food supply, which are especially notable since the majority of the world’s population is much more dependent on food supply than those who live in rich, developed countries such as our own. This brings us back to ancient times and the biblical story of how Joseph got to such a high position in the Egyptian hierarchy because of his ability to regulate the intermittency of the seven bad years and seven good years (I previously explored this in the context of fresh water supply in my April 8, 2014 blog).
The idea of using ethanol as a partial replacement for fossil fuels didn’t start because of awareness of climate change. As I have mentioned before (October 29, 2012), soon after World War II, it became evident that dependence on fossil fuels as the main source of energy would have to end sooner or later (remember the Hubbert Peak theory). The main reasons at that time were our finite supply and the limited reliability of foreign suppliers. So the thought was that it would be great to be able to grow our energy supply. This move would be especially great for the agricultural sector, which would be charged with increasing supply, thus giving them leave to raise prices and increase their profits. There were fierce debates that took place on this issue, many of which focused on the EROI that was calculated for this new energy supply.
The debate started following some EROI measurements that showed that it takes more fossil fuels to produce the same amount of energy that can be extracted from the ethanol – in other words, it’s an EROI smaller than 1. In 2002 the US Department of Agriculture (USDA) took its own measurements and summarized them in a report on the topic.
I am quoting three short introductory paragraphs from the report and including a detailed table that summarizes the results:
Ethanol production in the United States grew from just a few million gallons in the mid-1970s to over 1.7 billion gallons in 2001, spurred by national energy security concerns, new Federal gasoline standards, and government incentives. Production of corn-ethanol is energy efficient, in that it yields 34 percent more energy than it takes to produce it, including growing the corn, harvesting it, transporting it, and distilling it into ethanol.
Growth in ethanol production has provided an economic stimulus for U.S. agriculture, because most ethanol is made from corn. The increase in ethanol demand has created a new market for corn, and agricultural policymakers see expansion of the ethanol industry as a way of increasing farm income and reducing farm program payments, while helping the U.S. economy decrease its dependence on imported oil. Increasing ethanol production induces a higher demand for corn and raises the average corn price. Higher corn prices can result in reduced farm program payments.
Today’s higher corn yields, lower energy use per unit of output in the fertilizer industry, and advances in fuel conversion technologies have greatly enhanced the energy efficiency of producing ethanol compared with just a decade ago. Studies using older data may tend to overestimate energy use because the efficiency of growing corn and converting it to ethanol has been improving significantly over time. The estimated net energy value (NEV) of corn ethanol was 21,105 Btu/gal under the following assumptions: fertilizers are produced by modern processing plants, corn is converted in modern processing facilities, farmers achieve normal corn yields, and energy credits are allocated to coproduces.
The table is long and detailed, but I think that it is crucial to gaining some understanding of the complexity and subjectivity involved in the calculations of a concept such as EROI before accepting claims that such calculations prove or disprove the viability of doing things that we regard as essential.
Summary of abbreviations in the table: LHV-Low Heat Value, HHV – High Heat Value (the difference between LHV and HHV is often a source of confusion and it originates with exclusion or the inclusion of the water vaporization energy in the process; the difference amounts to about 10% in most cases), Bu – bushel, NR – not reported.
The debate started to have a wider audience after Congress passed the Renewable Fuel Standard (RFS) law in 2005. Here is how the EPA summarized this law and the mandated steps to be taken:
EPA is responsible for developing and implementing regulations to ensure that transportation fuel sold in the United States contains a minimum volume of renewable fuel. The Renewable Fuel Standard (RFS) program regulations were developed in collaboration with refiners, renewable fuel producers, and many other stakeholders.
The RFS program was created under the Energy Policy Act (EPAct) of 2005, and established the first renewable fuel volume mandate in the United States. As required under EPAct, the original RFS program (RFS1) required 7.5 billion gallons of renewable- fuel to be blended into gasoline by 2012.
Shortly after the law was passed (January 2006) a detailed article appeared in Science Magazine on the issue (“Ethanol Can Contribute to Energy and Environmental Goals”; Alexander E. Farrell,1* Richard J. Plevin,1 Brian T. Turner,1,2 Andrew D. Jones,1 Michael O’Hare,2 Daniel M. Kammen1,2,3 ; Science 311, 506 (2006)).
I am attaching the key figure from this article below.
Net energy and net greenhouse gases for gasoline, six studies, and three cases. (B) Net energy and petroleum inputs for the same. In these figures, small light blue circles are reported data that incommensurate assumptions, whereas the large dark blue circles are adjusted values that use identical system boundaries. Conventional gasoline is shown with red stars, and EBAMM scenarios are shown with green squares…
A few months later (June 23, 2006 – Vol. 312) an extensive comments section which focused on the article appeared in the magazine. The comments included both the issue of the competition with the food supply that I mentioned earlier and EROI calculations. Farrell et al. calculated the EROI of ethanol from corn as 1.2. This value means that the use of ethanol as an energy source leaves a replacement value of about 20% for other uses of fossil fuels aside from running the production of ethanol. This value was in agreement with the value that the Department of Agriculture quoted earlier and remains the most quoted value until today. The value that was given in the Weißbach paper, on the other hand, is 3.5.
These two values are not the end of the story. If you google the term “ethanol fuel,” you will find a Wikipedia entry that provides us with the following table:
These values are taken from an October 2007 article in National Geographic (“Green Dreams: Making fuel from crops could be good for the planet—after a breakthrough or two”; Joel K. Bourne, Jr., Robert Clark; National Geographic Magazine October 2007 p. 41). Everybody now realizes that the future of biomass – both in terms of the EROI and its disconnect from the food – supply rests in fermentation from cellulosic ethanol, which is defined as “biofuel produced from wood, grasses, or the inedible parts of plants.”
In the meantime, the production of corn-based ethanol in the United States went from 6.4 billion gallons in 2007 to 13.9 billion gallons in 2011. Not surprisingly, attempts to adjust the requirements are being met with considerable difficulties.
In the meantime, Happy Thanksgiving – I appreciate your continued readership.
This is the response to the last comment by John Morgan (posted November 29 – 11:20 PM):
Unfortunately, many discussions of what to do or not to do to mitigate climate change do not completely separate the science from politics. My blog on ethanol for fuel is a good example of such a close connection. When a paper appears that argues that based on EROI consideration, most solar based energy sources cannot be used because they will require more energy to operate compared to the energy that they will deliver; with all the authors connected with the German nuclear industry following Germany’s decision to discontinue nuclear effort, one cannot avoid suspecting a strong political motives. The close connection between the science and the politics is not something to discourage. Timely global energy transition to non-fossil sources will go through a productive, continues public debate based on multitude of arguments, sometime ferociously fought between winners and losers.
The data for the Weißbach paper and for John Morgans paper and guest blog came from LCA analysis of selective facilities. I am not dismissive of LCA, it is very useful for detailed analysis of selected facilities. The analysis is often targeted at detection of “hot spots” in these facilities to optimize energy use and minimize environmental impact. It is the first time that I saw it used as basis of EROI to delegitimize a whole industry. Morgan’s paper and blog emphasize storage. Weißbach’s paper ,on which it is based, uses storage as an add-on, even without storage most of these alternative energy sources should never be used. This cannot go without a response. Basing the argument on specific LCA data provides ample opportunity to select the data most suitable to serve the political message. This is a techniq1ue that is often been used by deniers, in spite of Morgan’s reluctance to be labeled as such.
Unfortunately, there seems to have been some mix-up in timing. My response posted below was addressed to an email that I got on November 26th that reads as follows:
I will respond separately the last comment in the one that will follow.
John,
I am really sorry that it came to this. My response will mix here the personal aspects with the professional aspects. I waited for your response to this email before posting a responding comment to the arguments that were posted on the site, but failing that am posting it now. You are more than welcome to continue the conversation in any form – privately through email, through comments on the site or through another guest blog on this site.
A bit of history:
The first time that we established any contact was when you responded to a short tweet of mine that described David Mackay’s paper about “ultimate” storage requirements in an energy transition with an emphasis on England.
Here is a copy of your tweets:
I regarded these tweets as big, authoritative statements that I wished to have explored further, so I invited you to write a guest blog on the topic.
I was familiar with the EROI arguments, but only in specific cases that relate to biomass and photovoltaics. I was not familiar with the Weißbach paper or any large, all-inclusive, statement that energy transition away from fossil fuels cannot use any intermittent, solar based, sustainable sources.
Once I got the details in your guest blog, my emphasis shifted to the EROI table from the Weißbach paper with the question of whether it justified your global responses. The emphasis within the Weißbach paper was not on the storage aspects of the solar sources but on the sources themselves. My conclusion was that it was not and I detailed my arguments in a series of three blog.
I do not share your statement in your email that climate change skepticism is an academic misconduct. If this were the case, the American Physical Society would have been in very deep trouble, as would many of my colleagues. The specific statement in my blog that raised your ire was:
“This attitude fits perfectly with one of the shades of deniers that I have previously described – the skeptics (September 3, 2012 blog) and requires a detailed response that will follow in this and the next few blogs.” The “shade” that I had in mind was “fatalists.” Here is the description from the earlier blog:
In my opinion, your tweets, guest blog, and the Weißbach paper fit this description in terms of the energy sources that can be used in the transition. My translation was this: if the intermittent solar sources cannot be used, we have to rely fully on nuclear, but since sometimes nuclear can not be used for other reasons, it means that we must continue to rely on fossil fuels as well, and climate change is unavoidable. I thought that such a position required detailed response.
My responses were never personal. Beyond our correspondence I don’t know you. Some of the comments that came in response to my first blog did testify that you fully believe, like I do, in preventive action to be taken now. I am delighted but, at least for me, this attitude didn’t come across in your blog or in the Weißbach paper.
Micha has responded to my EROI article in three posts; I’ll consolidate a response here.
The core thesis of my article is that: energy storage cannot back up wind and solar for primary energy supply, because storage degrades EROI below a viable level.
In his three posts, Micha discusses a range of issues, but does not challenge that core thesis about storage, which I believe stands. There are now over 500 comments on this piece at The Energy Collective and Brave New Climate that directly interrogate that conclusion at a range of technical levels, and while many qualifications can be elaborated the conclusion appears robust.
The storage data presented is for pumped hydro. According to the Stanford solar paper I cite, batteries require an energy input about 10x higher than pumped storage. So if pumped hydro is not viable storage, we can certainly be sure that batteries are not viable, even if there were quite generous errors in favour of the EROI of solar or wind.
Micha focusses on the Weißbach paper and carefully points out the authors work in a nuclear physics department (why?). In fact I cite four sources, and the other three are from solar and renewables and biophysical economics researchers, including a respected Stanford renewables team. The Weißbach paper happens to present its conclusions most clearly, and is the easiest to discuss in limited space (this article originally appeared in print, with a limited word count). The Stanford paper is particularly obscure in its presentation. But they all arrive at much the same place in respect of storage EROI, and are themselves part of a larger literature.
Micha is dismissive of the EROI threshold because it is determined economically, but this does not somehow invalidate either its reality or its importance. The EROI itself is not economic, its a purely physical energy balance. The absolute threshold requiring EROI > 1 is also purely physical. That there exists some threshold above 1 that is a minimum requirement for a given mode of organization of society is also physical. The exact value of this threshold is very difficult to establish, but one exists. For a modern technological society Weißbach et al. estimate a value of 7. This is in fact a lowball estimate; the more recent work I cite in my Postscript pushes it up around 12-14. Modern batteries, wind turbines and solar components are at the pinnacle of human technological achievement. Societies capable of producing these components operate at a high EROI threshold, almost certainly well beyond that yielded by stored energy from low EROI sources.
He’s equally dismissive of lifecycle analyses (LCA) for estimation of EROIs. Certainly, these are difficult measurements to make, and there is variability among the different attempts. This doesn’t mean we ignore this work. It means we intelligently discriminate between different studies, with better or worse methodologies. Micha for instance points to wide variability in nuclear EROI, from less than 1 to greater than 100. But the <1 values are clearly absurd. The other low range values come from calculations that include some fraction of very energy intensive diffusion-enriched uranium. But my understanding is there are no longer any diffusion enrichment plants left operating anywhere in the world. That leaves centrifuge enrichment, which gives a nuclear EROI of ~40-60. This is near enough to Weißbach's value of 75 to accept that its in a reasonable ballpark. Adopting more recent values for plant life, for instance, would likely rationalise the differences.
In fact the Mason Inman EROI review Micha cites has very similar numbers to the Weißbach paper – 6 compared to 4 for solar, 16 compared to 20 for wind, 49 compared to a range of 40 – 250 for hydro, nuclear 40-60 compared to 75. These are all very close for such a difficult-to-measure quantity. There's lots of work to do here, but you can't sustain the idea that the Weißbach values are outliers or otherwise unreliable. They're consistent with the literature and can be taken as representative of the current state of the art. In preparing his review Inman interviewed the authors of three of the other papers I use – these are authoritative voices in the field.
In his third post on biomass Micha observes the difference between the USDA's and Farrell's value for the EROI of corn ethanol of 1.2, and Weißbach's value of 3.5 for corn biomass. But they should be different. Weißbach is referring not to corn ethanol but to corn biogas – natural gas produced from corn, and burnt in gas turbines. They are different fuels produced in different processes, so there is no expectation they should have similar EROIs.
So, putting it together, we have: credible values for EROI for a variety of energy sources which are consistent with the current literature, a lowballed estimate of the minimum societal EROI for a technologically advanced civilisation, and a calculation of storage impact on EROI that is also lowballed because it uses the storage technology that consumes the least energy (pumped hydro). This is a very conservative calculation. If the result is that stored wind and solar PV power have EROIs too low to power society, it is very likely to be true, because any correction to the calculation makes it worse for these power sources. Stored solar thermal is marginal on these numbers, and fails if the minimum societal EROI is slightly higher (for instance).
The implication is that energy storage does not help wind and solar variability. They can contribute in the long run only in their direct, unbuffered form. This limits their penetration in the energy mix to a modest share. Other energy sources must fill the gap which are both high EROI, and dispatchable. Of those available with these characteristics all have high greenhouse emissions, except for nuclear.
This may be unpalatable, but no compelling challenge to this thesis has emerged, here or elsewhere.