Battling with the Critics

Just when I thought I was out, they pull me back in.

I am trying to spend more time writing on topics other than ethanol. But I get a lot of e-mails on that subject, and often have 3 or 4 mini-debates going on at a time via e-mail. I just finished a debate involving a government official and some big names over the energy balance of gasoline versus ethanol. There still seems to be a lot of confusion surrounding this issue, so I asked for permission to publish the exchanges. I was reluctantly given permission, provided I deleted the personal information from the government official (name and government agency). The exchange involved myself, a government official that I will refer to as “Tom”, Michael Wang from Argonne, and Vinod Khosla. Tom’s responses are in black, mine are in blue, Wang’s (1) response is in green, and Khosla’s is in red.

It all started when I got an e-mail from Tom. It wasn’t clear to me which specific essay he had read that prompted his e-mail, but he wrote:

Mr. Rapier,

If your assessment of the ethanol fuel cycle energy balance (and its comparison with the petroleum fuel cycle energy balance) is right, then not only is Vinod Khosla wrong, but many others of us in the energy community — including the U.S. Department of Energy and Argonne National Laboratory (see attached summary) must also be wrong.

Attached was a summary of an Argonne National Lab report written by Michael Wang, who initiated the following claim (from the report):

As you can see, the fossil energy input per unit of ethanol is lower—0.74 million Btu fossil energy consumed for each 1 million Btu of ethanol delivered, compared to 1.23 million Btu of fossil energy consumed for each million Btu of gasoline delivered.

I must admit that appeals to authority don’t impress me much, especially when I know the person making the argument is completely wrong. Remember, this is coming from a government official involved in alternative energy. So, I responded:

Tom,

They are wrong. I have read all of the Argonne studies. I have exchanged e-mails with Wang at Argonne and Shapouri at the USDA. They know they are being misleading in these claims, but most people don’t dig into the details to see their sleight of hand.

Here is a very simple test that will demonstrate they are wrong. After people work through this, they always see the problem. Let’s say my goal is to make 1 BTU of liquid fuel. Will I consume more energy if I produce ethanol, or will I consume more energy if I produce gasoline? The implication from the Argonne et al. would imply that it should take more energy to produce the gasoline. However, that is not remotely the case. If I presume an energy balance for ethanol of 1.3, then I will consume 1/1.3, or 0.77 BTUs to make 1 BTU. My net is a mere 0.23.

If, however, I make gasoline, the efficiency is 80%. That is where the 0.8 number comes from. In this case, I only consumed 20% of the BTUs to make 1 BTU of gasoline. My net is 0.8 BTUs. What they have done is convolute energy return and efficiency, and act like 1.3 for ethanol is the same metric as 0.8 for gasoline, when they are actually 2 different metrics.

As I like to say, there may be some legitimate reasons for using ethanol. Efficiency of production is one of the most misleading arguments out there. It just isn’t true. And I will gladly debate Wang or anyone at the DOE in print regarding these misleading claims.

Tom responded, copying Michael Wang at Argonne and Vinod Khosla (they were copied on all messages from this point). I guess he felt he needed some backup.

Robert,

As I see it, the fallacy of your reasoning (similar to that of Pimentel’s and Patzek’s) originates, at least in part, from an “all Btus of energy are created equal” viewpoint. If continued /expanded use of petroleum was indeed feasible, sustainable, environmentally and politically acceptable, etc., then perhaps your conclusion, that petroleum is a more “efficient” energy option than ethanol, would be more valid — i.e., just keep burning the petroleum Btus and continue to accept the bottom-line energy result (albeit a continually worsening one in any petroleum-depletion scenario) that the luxury of stored fossil fuel deposits afford us: by reinvesting a fraction (1/5 today but steadily increasing) of the recovered petroleum energy, we can continue to harvest what’s left.

But the production of ethanol and other biofuels (which, by the way, should include a broader focus, encompassing other forms of pure and mixed alcohols, biodiesel-type fuels, bio-crude type fuels, etc.), along with other kinds of bioenergy, offers a means of harvesting Btus of solar energy and incorporating this contribution from solar energy into today’s transportation energy supply — an achievement that has thus far proved elusive via other means, such as electric vehicles or hydrogen.

The fact that today’s investment of 1 Btu of fossil energy in the ethanol fuel cycle delivers “ONLY” 1.3 Btus of ethanol to the vehicle fuel tank (the added 0.3 Btu being solar energy incorporated into the fuel cycle) is actually a very beneficial energy result, especially given that this result only gets better with technology advances, potentially including production from cellulosic biomass. Meanwhile, the energy reinvestment necessary to capture remaining petroleum resources promises only to become greater. Ask yourself this question: If producing and operating hybrid electric vehicles (which I suspect have their own underestimated trade-offs besides the obvious higher cost factor), in order to make petroleum Btus go about one-third further, makes good sense in today’s energy world, then why doesn’t achieving essentially the same result via ethanol production and use (with at least incrementally, if not fundamentally better results in store) offer at least as attractive an option?

While I don’t think I would personally try to argue that the ethanol fuel cycle is twice as efficent as the petroleum fuel cycle (i.e., by comparing a 1.3-1.6:1 ratio to a 0.8:1 ratio), neither do I find your analysis compelling from an energy standpoint; in fact, it appears even more misleading. I believe that most of us in the transportation energy community — along with many in the automotive industry, the oil and other energy industries, the environmental and global climate change communities, etc — have come to accept the results of Argonne National Laboratory (as summarized in the U.S. DOE webpage document I forwarded to you earlier) as the most authoritative and fair assessement thus far of ethanol’s net energy (and greenhouse gas) implications.

Michael Wang also weighed in, to say he wasn’t getting involved:

Dear Mr. Rapier,

Instead of wasting everyone’s time, let me just simply pointing out that I do not recall that I have extensive communication with you and I do not intend to do so, because of your statement “I have exchanged e-mails with Wang at Argonne and Shapouri at the USDA. They know they are being misleading in these claims, but most people don’t dig into the details to see their sleight of hand.”

You are entitled to have your opinion, but do not imply personal attack on my professional work.

Michael Wang

I answered both with my next response:

Tom,

There is no fallacy in my reasoning, and my arguments have nothing to do with Pimentel’s and Patzek’s. To suggest they do indicates that perhaps you still don’t understand my argument.

Unlike Pimentel and Patzek, I am using Argonne’s numbers to make my point. Your argument, “If continued /expanded use of petroleum was indeed feasible, sustainable, environmentally and politically acceptable….” is a different argument than the one you originally started off with. You are suggesting that there are other reasons for using ethanol. Fine. But you are not addressing the point of my argument, which is simply that ethanol is far less efficient to produce than gasoline, despite the proponent’s claims to the contrary. Argue the sustainability issues. Argue the environmental issues. But don’t mislead people by suggesting that it takes more energy to produce gasoline than to produce ethanol. That is an incredibly ludicrous claim.

My argument is not misleading at all. It does not convolute efficiency and energy return. It is a measure of the amount of energy that must be consumed to produce two different fuels: gasoline or ethanol. That is a very simple metric, and is not in any way misleading. Wang’s metric is misleading, and I am sure that he is well aware that people are misusing it. When people say “ethanol is 1.2, but gasoline is worse at 0.8”, they have compared two different metrics. When you write that you accept the authority of Argonne/DOE with respect to the net energy and greenhouse implications of ethanol, you are once again addressing a different argument. Please do not address Red Herrings, since I have accepted their net energy results for ethanol in my analysis.

Regarding Wang’s communication with me, I still have it if he would like for me to refresh his memory. I pointed out the same thing I have pointed out here, and his response was essentially “Yeah, but you are looking at the total energy inputs, and there are many different ways to look at this problem.” I do not regard the debunking of misleading claims as a waste of anyone’s time. I would think that Wang would want to defend his work against critics like myself, especially given that most of it has not been subjected to scientific peer review. Again, I will debate Wang, Shapouri, or anyone else who wishes to argue that it is more efficient to produce ethanol than gasoline. If you want to argue about something else, then you aren’t addressing the argument I am making. Yet this is exactly what you did in your second response.

Finally, I want to make it clear that my comments are not meant to defend the status quo. I want to see us move away from fossil fuels as quickly as we can. I am merely using the gasoline versus ethanol issue to show why these claims of higher efficiency of ethanol production are fallacious.

This response covers my biggest gripe about people who want to debate this issue. If I rebut a specific claim, they gallop off to a different claim. That is exactly what Tom did.

At this point, I also asked if they minded me publishing the exchange:

Incidentally, do you have any problems with me publishing this exchange? I will publish it without changing a word, and will include Wang’s statement that he doesn’t recall having extensive communication with me. I think the public can benefit from these exchanges. I understand your position quite well, however I hope it is clear that you didn’t actually address my arguments, but instead addressed other reasons for supporting ethanol.

I am confident that my argument as written is completely accurate and not in any way misleading, and I have no problem being judged by public opinion on its merits. I am a strong supporter of publicly debating these technical issues, and I have no interest in misleading anyone. But I also have no interest in allowing people to be misled.

Vinod Khosla weighed in next:

Robert’s argument would make solar cells a horrible source of energy at an efficiency of 0.15! And why would we ever use electricity?

Most modern ethanol plants being built have an energy balance of around 1.5 -1.6 as they try and minimize their energy use for cost reasons. That coupled with the higher use efficiency of ethanol energy than petroleum energy (25% less mileage even with 33% less energy is the accepted EPA rating for most flex-fuel cars – the SAAB 9-5 Biopower with Turbo is only 18% less mileage) gives an ethanol “fossil fuel efficiency” of about 2X per mile driven. The current California plants we are building don’t especially ship corn (they are built around cattle feedlots where the corn has been shipped in for years) and they don’t dry the distillers grain since they use it locally at the feedlot, does better than the 2X number. The E3 Biofuels plant in Mead Nebraska achieves an “energy balance” of five for CORN ethanol according to a report I saw from the National Commission ion Energy Policy.

It is time to stop asking the wrong question of “energy balance” or even the somewhat less wrong question of “energy balance relative to petroleum” but rather ask the two right questions (a) how much petroleum use can we displace per gallon of an alternative liquid fuel and (b) what is the green house gas reduction per mile driven.

For nuance we might add (c) at what cost of production per mile driven (to take away the short term price manipulation going on and (d) in what vintage of plant? Modern, average, old, coal fired, gas fired, with and without dry distillers grain, all the way to the E3 Biofuels model. Today the economics of reducing energy cost work.

I responded to Mr. Khosla’s argument:

The solar cell argument is not valid, as several people pointed out on The Oil Drum, because it confuses efficiency with energy return. The instantaneous efficiency may be 15%, but you can get that day after day. The total energy returned from a solar cell far exceeds the energy that went into creating it.

The reason we use electricity is because we convert coal, something not especially useful for doing work in its natural form, into a form in which it can do useful work. That is not the case with most of the fossil fuels that go into making ethanol. We turn natural gas, gasoline, and diesel, all perfectly good transportation fuels, into ethanol. We capture a bit of solar energy in the process, but grain ethanol is primarily recycled fossil fuel. And while this argument has focused on the marginal energy return, not included in those assessments (as Wang can attest to) are the secondary inputs, nor effects from soil erosion from growing corn, or herbicide and pesticide runoff into our waterways.

For the record, I fully support, and have advocated the E3 Biofuels model. In fact, I spoke with their project manager this week for an hour on the phone. I was also recently quoted in National Geographic endorsing the E3 process:

New Ethanol Plants to Be Fueled by Cow Manure

However, a couple of things need to be clarified. Their plant has not yet started up, so claims of energy return from this process are premature. It is definitely a step in the right direction, and I would prefer to see all new ethanol plants built around a similar model.

Regarding “wrong questions” and “right questions”, that misses the entire point of my arguments, which are quite simple. There is a horrendous level of misinformation out there surrounding ethanol. When someone claims that Brazil farmed their way to energy independence, or that it is more energy efficient to produce ethanol than gasoline, or that ethanol produces no greenhouse gases – those are claims that must be addressed. Ethanol policy should not be made based on misinformation like this. My agenda is simple, and that is truth in advertising. I am a skeptical scientist by nature, and I feel like these claims deserve critical technical scrutiny. It is not my goal to kill grain ethanol, unless it deserves to die. But we won’t know that without an honest debate, and too little of that is taking place. My goal is to separate hype from what the science actually indicates, and pursue those solutions that make the most long-term sense. Corn ethanol, which has been the primary target of my criticism, is not a very efficient use of our resources as it is currently produced. On this, I know that Mr. Khosla agrees with me, because we have spoken at length about this.

Tom indicated that he really didn’t want to have this debate in public:

I’m inclined toward Dr. Wang’s (and Mr. Khosla’s) viewpoints that it is somewhat of a distraction and probably unproductive to pursue this debate with you further or participate in your forum — especially in light of your unfortunate characterizations of individuals’ and organizations’ work (“sleight of hand”?). In any case, since you say you accept Argonne’s basic analytical results, then this entire debate is all about the interpretation and implications of these results (and who is “right” trying to answer the academic question “Which is the more “efficient” fuel, ethanol or gasoline”), which I don’t foresee being resolved in this forum.

I once again tried to convince Tom to take this debate into the public arena:

How else do you characterize the comparison of an EROI for ethanol to an efficiency of gasoline, other than sleight of hand? A straightforward assessment would be to consider either EROI to EROI, or efficiency to efficiency. Perhaps it wasn’t Dr. Wang’s intention to have this issue so thoroughly muddled, but the public has certainly muddled it. I have lost count of how many times someone claimed that it is twice as efficient to produce ethanol as to produce gasoline.

My impression then is that you do not want this exchange made public? If we posted this at The Oil Drum, it would be read by a tremendous number of people, and would have advocates on both sides. If your argument is correct, then you should have no concerns given that I will post this exchange verbatim. I think these are the kinds of open exchanges that need to take place so people can sort out hype from truth. My main objective is education, and I think it would certainly suit that purpose.

We exchanged 1 last pair of e-mails that I won’t entirely reproduce (because I told Tom I wouldn’t). Suffice to say that Tom agreed to publication, provided I removed some information on him and his organization. In his final response to me, Tom accused me of rancor (passion is not the same as rancor!), questioned whether my rancor explains my e-mail identity (tenaciousdna), and once again invoked the argument from authority, suggesting that my argument was subjective and merely my opinion, and he and all those other authorities couldn’t be wrong. Needless to say, my reply was “pointed”, but I offered to take up the matter with him at any time.

This exchange may help explain why my post count has gone down, which some have asked about. These things take up a bit of my time every day, so today I decided to kill two birds with one stone and make a post out of the debate. Let this also serve as a warning to those who want to bang heads with me. 🙂 If you want to win a debate with me, make sure you are arguing from a factual position.

114 thoughts on “Battling with the Critics”

  1. Vinod said, “Most modern ethanol plants being built have an energy balance of around 1.5 -1.6 as they try and minimize their energy use for cost reasons.”

    Robert,

    Vinod, Wang, Shapouri, and “Tom” should all know that if an ethanol plant actually returned 1.5 or 1.6 times the energy invested, the United States would soon be awash in energy and no ethanol plant would ever need to use natural gas or coal as a source of its thermal energy.

    Any production system that had a 150 – 160% return on energy invested would be a license to print money. (Which is perhaps what Vinod is hoping for.)

    Think of the possibilities:

    * Invest 100 units of energy and get 160 units back.

    * Re-invest those 160 units and get back 256.

    * Re-invest those 256 units and get back 410.

    * Re-invest those 410 units and get back 655.

    In only four cycles, one could increase the original energy investment by more than 600%.) If that were possible, we’d have energy multiplying faster than Tribbles on that famous Star Trek episode.

    If all those ethanol plants are producing 50% to 60% more energy than they consume, where is all that extra energy? And why can’t ethanol plants use a fraction of that surplus energy for milling and distilling corn to make more ethanol?

    Instead, being next to a high-capacity natural gas pipeline is one of the major limiting factors when locating new ethanol plants.

    Why is that a major LIMFAC if ethanol plants produce so much more enrgy than they consume?

  2. Gary, I think perhaps you’re forgetting about constraints other than energy return. The quantity of arable land, for example, is relatively limited.

  3. It is unfortunate that ‘Tom’ refuses to debate further. I am a proponent of cellulosic ethanol, but I very much believe that we should debate these things transparently and constantly…otherwise we might end up with another push for hydrogen, with billions of dollars and several years wasted.

    I hope not all government officials are like Tom, or you Yanks will never have a remotely productive energy policy.

  4. Quoting from your email

    “I have exchanged e-mails with Wang at Argonne and Shapouri at the USDA. They know they are being misleading in these claims,”

    They obviously do not think they are being misleading. This is awfully close to a personal insult, an allegation not just of deceitfulness, but of admitting to deceit.

    How can you honestly expect Mr Wang to reply to you after such a misrepresentation?

    ———————

    As to the substance, I largely agree with Vinod. “Efficiency” is a term open to definition. I think your definition of efficiency (process and extraction energy in compared to gasoline out) is not the best, but above all, it’s not the only one possible. A definition of “efficiency” other than yours (say fossil fuel energy in to transportation fuel out) is quite feasible.

    You should just accept that. After all, the actual facts are not in dispute. Process and extraction energy input per unit of transportation fuel output is much lower for gasoline than for ethanol, when comparing present gasoline and present ethanol production in the US, while fossil fuel input per unit of transportation fuel favours ethanol.

  5. They obviously do not think they are being misleading.

    That’s where you are wrong. They do know it. It was clear from Wang’s earlier e-mail response that they know it. What they have done is come up with a completely illegitimate metric in order to make ethanol look better than it is. More on this below. But I pointed this out to Wang, and he said “Yeah, but you are looking at all of the energy inputs.” Well, duh. Isn’t that what we should be looking at?

    How can you honestly expect Mr Wang to reply to you after such a misrepresentation?

    Suggesting that someone is using “sleight of hand” is not an ad hominem argument. (Look it up). Nor is it an unfair characterization. Wang has given a number of presentations in which he compares 0.8 for gasoline, which is an efficiency, to 1.3 for ethanol, which is an EROI. This is an apples to oranges comparison, and making this comparison misleads the audience. I know the audience has been mislead because I hear them taking this information and using it to claim that it is twice as efficient to produce ethanol as it is to produce gasoline. Dr. Wang is directly responsible for this misconception, and that is why I hold his feet to the fire on this issue.

    As to the substance, I largely agree with Vinod.

    Then that would mean you are wrong, just like he is. The bottom line is exceedingly simple: Do I consume more energy in producing ethanol, or producing gasoline? The answer is that I consume 4-5 times as much energy in producing ethanol. Simple as that. And in this efficiency calculation, the only thing that matters is how much energy I consumed and how much liquid fuel I have available at the end.

    A definition of “efficiency” other than yours (say fossil fuel energy in to transportation fuel out) is quite feasible.

    You may not understand what they are actually doing, so let me explain. When they say that it takes 1.23 BTUs of fossil fuel energy to produce 1 BTU of gasoline, they are actually counting the contained 1 BTU of gasoline as an input. That is ludicrous. What this actually means is that 0.23 BTUs of fossil fuel energy were consumed to produce 1 BTU of gasoline. The net in their example is then 0.67 BTUs.

    On the ethanol side, when they say it took 0.74 BTUs of fossil fuel to produce 1 BTU of ethanol, they actually consumed 0.74 BTUs of fossil fuel, netting only 0.26 BTUs of ethanol. That is a legitimate, apples to apples comparison. What they have done is not legitimate, because the 1 BTU of gasoline they threw in with the gasoline comparison was not actually consumed. At the end of the process, it is still available as liquid fuel. Not so with the 0.74 BTU input for ethanol.

    It is not a personal insult for me to point out that this is a misleading metric, nor that Wang knows it is a misleading metric. He knows, because I pointed it out to him in an e-mail exchange.

    You should just accept that.

    Wrong again. There is no reason for me to accept that, and I won’t accept it. I don’t accept things unless I am actually proved wrong, and in this case I am right.

    …while fossil fuel input per unit of transportation fuel favours ethanol.

    As shown above, only via sleight of hand. If “input” means the energy that was consumed in production, ethanol is far worse off. The only way ethanol comes out on top is if you count as input something that just passes through the process, and is available as fuel at the end of the process. Yet if you do that, you are no longer comparing apples to apples, so proponents who suggest that ethanol is more efficient to produce than gasoline are not actually using metrics of efficiency. That is the point.

  6. There is no agreed definition of “efficiency” in this context. Therefore this is not a matter of being proved right or wrong, it is one of agreeing definitions, or I should say, maybe even just a matter of you simply acknowledging that Vinod prefers a different definition of “efficiency” than the one you like best.

    Vinod counts fossil fuel inputs only, and for each BTU of fossil fuel input we get 0.8 BTU of gasoline output, or 1.4 or thereabouts BTU of ethanol output. In the case of CTL, each BTU of fossil fuel input may yield 0.5 BTU of liquid output.

    This definition makes the most sense when we want to calculate how much fossil fuel has to gotten out of the ground, and is closely related to how much carbon dioxide will be added to the atmosphere.

    I think you are mistaken in thinking that Mr Wang has acknowledged his view being “misleading”, and he’s certainly strongly objected. In the interest of civil discussion, I would therefore just recommend that you do no imply otherwise.

    While I know what your preferred definition of “efficiency” is in this context, I haven’t seen a convincing explanation as to why it is relevant. What useful information does this efficiency number actually convey?

    Note that I’ve seen thousands of comments on the subject, in the literature, on the yahoo energy resource group where I used to post a lot at one stage, and on the oildrum (though most comments there are of abysmal quality).

    So, yes we could take 0.2 BTU of petroleum to get 1 BTU of petroleum (for a net return of 0.8 BTU), or we could take 1 BTU of fossil fuels to produce 1.3 BTU of ethanol (for a net return of 0.3 BTU), but how is this actually relevant?

    It doesn’t say anything about the key limitations (notably labour, land, environmental damage, fossil fuel availability).

    And as one commenter on the oildrum has said, why not consider the BTU’s of fossil fuel used to produce the fossil fuel going into ethanol. With coal, we then look at say 0.1 BTU of fossil fuel to produce 1 BTU of coal to produce 1.3 BTU of ethanol for a net output of 1.2 BTU.

  7. Robert&Chris: gary dikkers enthusiastic compounding of return was OK.

    Yes, if it went on forever then an infinite amount of land would be needed. But why assume something goes on forever? We don’t for any other argument.

    If we use a barrel of oil to pump five from the ground we don’t expect that to go on forever.

    We can even imagine the point where every inch of the earth was covered with solar cells. Thus solar cells won’t increase energy production forever either.

    I may have missed your point. These energy arguments are treacherous.

    sign me anon#2

  8. So, Heiko decided not to heed the warning at the end of my essay. 🙂 I see that you are also an engineer, so I guess your stubbornness is understandable. But is it supportable?

    Vinod counts fossil fuel inputs only, and for each BTU of fossil fuel input we get 0.8 BTU of gasoline output, or 1.4 or thereabouts BTU of ethanol output.

    Not true. “Input” in this context is used differently for the two cases. In the case of ethanol, it was actual fossil fuel BTUs that were burned to make the ethanol. In the case of gasoline, they are counting the crude oil as an “input”, even though it isn’t physically consumed. The gasoline is simply distilled off or reformed. Those inputs are not referring to the same thing, hence to call the comparison of 0.8 versus 1.4 an efficiency comparison is false. It is misleading, and not conducive to honest debate. But you don’t have to take my word for it. In a recent issue of Science, following some ethanol proponents who used this misleading metric, they were taken to task in a number of follow-up letters who pointed out the problems I have just mentioned. In response, the proponents admitted that they should have called their metric something else. I would imagine that you have seen the initial article and the follow-ups.

    I think you are mistaken in thinking that Mr Wang has acknowledged his view being “misleading”, and he’s certainly strongly objected. In the interest of civil discussion, I would therefore just recommend that you do no imply otherwise.

    You have not seen the e-mail I have from him, have you? He does not acknowledge that he is being misleading, but he knows that people are being misled. He knows that when someone says that ethanol is twice as efficient to produce as gasoline, that this statement is not true, even though they are using his work to support this claim. I am not one to parse words. Wang is being misleading. He has co-written a number of pro-ethanol papers, and they are constantly juggling numbers to exaggerate ethanol’s benefit. My comments on him being purposefully misleading are not based on this one incident. See:

    How Reliable are Those USDA Ethanol Studies?

    And

    Grain-Derived Ethanol: The Emperor’s New Clothes

    I demonstrate a pattern of creating special metrics and use of creative accounting in order to exaggerate ethanol’s benefit. Wang is one of those responsible, and I am calling it like I see it.

    What useful information does this efficiency number actually convey?

    It is important for several reasons. First, it demonstrates why ethanol will continue to rely on subsidies. Any process that burns up over 70% of the BTUs to make the product is going to have a hard time competing with a process that burns up 20% of the BTUs to make the product. It explains why the price of ethanol rises and falls with the price of fossil fuels. Since it is primarily a recycled fossil fuel, it must reflect the price of fossil fuels.

    It is also has profound implications for society. A society that has been running on fossil fuels with an EROI of 5 or so can’t transition to an alternative with an EROI of 1.2-1.4 without making enormous changes. Consider the amount of energy we would burn just to “create” the grain ethanol BTUs on a much larger scale. That is not a workable solution on a large scale without making big societal changes. The public is being lulled into complacency, though, by claims like “it is more efficient to produce ethanol than to produce gasoline.”

    ….on the oildrum (though most comments there are of abysmal quality).

    You mean like the following one, which you posted to your blog (in reference to one of Khosla’s presentations)?

    Lots of sensible answers there, let me just point out one. The energy balance is easily changed, a little more capital investment and an energy return of 5:1 for corn ethanol is no problem.

    That comment tells me plenty about why we are even having this discussion. You are far too quick to accept the claims of ethanol proponents at face value, as shown by that statement. 5:1 is no problem? Don’t you think someone should actually demonstrate this before claiming it is no problem? Khosla doesn’t seem to, because he has already started making this claim, that they “achieve this result.” Yet the people who are building the E3 plant have already told me that they think they will get 3:1, based on their most recent energy models. Optimism is fine, your comments (and Khosla’s) are something far beyond just optimism. The E3 plant hasn’t even started up, so claims of 5:1 or even 3:1 are not supported.

    So, yes we could take 0.2 BTU of petroleum to get 1 BTU of petroleum (for a net return of 0.8 BTU), or we could take 1 BTU of fossil fuels to produce 1.3 BTU of ethanol (for a net return of 0.3 BTU), but how is this actually relevant?

    It is relevant for the reasons I have pointed out. The net from the petroleum route is far greater than for the ethanol route. For the petroleum route, we burned 0.2 BTUs and netted 0.8 BTUs. The energy returned over energy consumed is 4.0. For ethanol, we burned 1.0 BTU to net 0.3 BTUs. The energy returned over energy consumed is 0.3, versus 4.0 for gasoline! This explains a great deal of why ethanol is at such a tremendous disadvantage, and further demonstrates how ridiculous it is to claim that it is more efficient to produce ethanol. It shows why, as long as there are fossil fuels, that fossil fuels will be favored. At least up to the point that the crude oil extraction step is substantially more energy intensive than it is now.

    It doesn’t say anything about the key limitations (notably labour, land, environmental damage, fossil fuel availability).

    All separate arguments from the efficiency issue, each of which should be examined based on its own merits.

    With coal, we then look at say 0.1 BTU of fossil fuel to produce 1 BTU of coal to produce 1.3 BTU of ethanol for a net output of 1.2 BTU.

    Or you could use the 1 BTU of coal to produce 10 BTUs of petroleum. The difference in my point and your point is that mine explains why we are so wedded to fossil fuels. Yours is “we could be doing something different”, but you ignore the societal change that will be required to massively scale up certain biofuels.

  9. Thank you, Robert, for your continuing efforts to bring fair analysis to this ethanol issue.

    It seems to me that a good argument for the general publich is that, if ethanol is so much more efficient than gasoline, why aren’t the oil companies making even more money by making ethanol instead of gasoline? Surely they could team up with Big Ag (ADM etc) to do it if it was really a winning proposition.

  10. Heiko:

    Efficiency is a well defined term in thermodynamics. Redefining a well-known physical definition is certainly dishonest. There are two definitions for efficiency, one based on the 1st law of thermodynamics, the other on the 2nd law.

    The 1st law establishes conservation of energy. Hence first law efficiency compares energy in to useful work out. Take for example an electric resistance heater. If I run 1 J through a baseboard heater I expect to get 1 J of heat out. By the first law of thermodynamics I have an efficiency of 100 %.

    The 2nd law establishes that in a closed system entropy always increases. The definition for efficiency by the 2nd law compares the exergy input to exergy output. If we go back to the baseboard heater analogy, let’s say that we want to heat air from 0 °C (273 K) to 20 °C (293 K). By an exergy argument, we would consider a Carnot heat pump to be an ideal method of moving heat from outside our environment (a house) to the interior. A Carnot heat pump over those temperatures would have a coefficient of performance of 13.65. The efficiency of the baseboard heater then, by the 2nd law metric, is only 7.33 %.

    Energy return on energy invested (EROEI) has nothing to do with efficiency per say. It is a metric that is used simply to represent our ability to leverage pools of exergy that exist in our environment. There are two sources of free exergy that we draw from, nuclear fission (geothermal and nuclear power) and nuclear fusion via the sun (fossil fuels, solar, wind, hydro). ERORI measures how much exergy we must spend to extract a given quantity of exergy from our environment. There is no relevant analogy for the baseboard heater example for EROEI, which is why it is dishonest to compare EROEI to efficiency.

  11. Yes I am a chemical engineer, and I work on fast pyrolysis, as you’ll quickly find through a web search.

    I don’t see why an input has to be something that is “consumed” rather than transformed. Again that’s more about the meaning of words than about any matter of substance. But in chemical reactor engineering, it is quite common to talk about reactants as inputs and products as outputs.

    I have indeed not seen your email exchange with Mr Wang, I can only go on what you’ve said about it, and what Mr Wang has said. And you both agree that he’s not admitted to being misleading.

    A net return of 1.3 BTU’s of ethanol per BTU of coal is no economic problem. Coal sells for something like $2 per MMBTU, gasoline for around $14 per MMBTU. So even if 2 BTU’s of coal were required (as they are in present CTL processes) per BTU of liquid fuel output, the process could be (in the case of CTL, presently is) economic.

    In your main post, you mention natural gas, gasoline and diesel as energy inputs to ethanol and say that those are all perfectly good transportation fuels. Gasoline and diesel are, but nat gas cannot be put into a conventional gasoline tank. This either requires capital investment (a high pressure storage tank and ancillary modifications to light duty vehicles) or it needs to be converted into a liquid fuel (it is used as a source of hydrogen for hydrocracking, for methanol production, with the methanol going into MTBE and biodiesel, and there are a number of massive GTL projects). Nat gas also sells for much less than gasoline/diesel (wholesale prices of nat gas and gasoline/diesel fluctuate, but over the last few months $7 per MMBTU for nat gas and $14 per MMBTU for gasoline/diesel are reasonable numbers)

    One of the great advantages of conventional ethanol production is that capital costs are so low. It compares very favourably there to CTL or GTL.

    So, why does net energy of present producers say little about a permanent need for subsidies:

    1. The 0.7 BTU of input may be worth a lot less than 1 BTU of output (presently they are)

    2. Even if the inputs and outputs sold for the same price per BTU, a 30% gain could be economic as long as the processing costs were small enough (certainly on the capital cost side they are pretty small)

    3. The system may not be optimised for minimal fossil fuel input (it definitely is not, not least because of 1.).

    Think of the option of using corn instead of nat gas for process heat. Corn has a net energy around 10:1. Convert 10 BTU’s of corn to 5 BTU of ethanol by adding 3 BTU of nat gas and we get a ratio of 4:5. Use the corn instead of nat gas, then we need 13 BTU of corn (3 to be burnt, 10 to be fermented) to produce 5 BTU of ethanol, and only 1.3 BTU of fossil fuel to grow the corn, and we’ve got a new ratio of 1.3:5.

    So, how about the impact on society?

    Suppose for now land etc. were no issue, we could produce lots of coal at high net energy return and low cost and feed ethanol and fertiliser plants with it. That may mean a lot of coal production (though less than would be required if liquid fuels were entirely CTL), but I don’t see any “societal impact” (I presume this means a larger fraction of the labour force being devoted to energy production or much lower per capita consumption of liquid fuels).

    The comment I posted on my blog wasn’t from the oildrum, but from energyoutlook.blogspot.com and I had made it myself. I read the oildrum, because there’s a lot of useful information there, there’s a lot of poor comments there (there are hundreds in the Vinod Khosla thread, and with such a flood it’s hard to expect quality in every comment), but I saw some very good ones too.

    I accept that a 5:1 or 3:1 return is a projection. But you can’t expect people to build high net energy return plants when there just isn’t an incentive. It’s money that matters in the end, and if building a 5:1 plant means spending a lot more on capital to cut energy expenditures so that the overall ethanol price comes out a little higher, nobody will do it. I misphrased my original comment (“a little more capital investment and an energy return of 5:1 for corn ethanol is no problem”), I actually meant substantially larger capital investment, but a rather modest increase in ethanol production cost.

    Zeolite membrane technology is one way to substitute capital for energy.

    What’s the production cost for E3’s ethanol? How much higher is it than conventional plant? Or is the same, but it’s merely more risky, because of the higher capital cost? Investors may even prefer building a plant producing ethanol at $1.10 per gallon, instead of $1 per gallon, if the payback period is 1 year rather than 5 years (ie most of the $1.10 is operating costs and most of the $1 is capital cost) .

    I believe (I think based on what you’ve said) that E3 still use nat gas, because they don’t get enough biogas. But they could instead use biomass gasification. I bet they don’t because the capital cost / energy costs tradeoff would be unfavourable (ie something like 10 cents per gallon less in energy operating costs compared to 20 cents per gallon more in capital depreciation costs).

  12. Robert McLeod,

    I know about the Carnot efficiencies of heat engines and heat pumps. They are well defined.

    This doesn’t apply in the context we are presently discussing, where there is no agreement on the appropriate inputs and outputs for the efficiency calculation (inputs/outputs).

    It is perfectly reasonable to say that fossil fuel BTU’s taken from the ground are the input, and liquid transportation fuel BTU’s are the output for a fossil fuel use efficiency.

  13. I don’t see why an input has to be something that is “consumed” rather than transformed.

    The problem is that the term is not consistent. In the case of transformation, the liquid fuel is still available at the end for doing work. In the case of consumption, it isn’t. The only thing that matters in the context of efficiency is how much energy was actually consumed, and how much is left as fuel.

    Here is an extreme example that might help drive the point home. Let’s say I can consume 1 BTU and get 1 million BTUs of oil out of the ground. Let’s say that the oil is essentially natural gasoline, such that very little processing is needed. Let’s say it takes an input of 9 BTUs to process the 1 million BTUs of oil into 1 million BTUs of gasoline. So, what does this look like according to the Wang metric? Well, we treat the 1 million BTUs of oil as an input, so that we say it took 1,000,010 BTUs of fossil fuel inputs to make 1 million BTUs of oil. Doesn’t sound very good, does it? But in reality, the 1 million BTUs just passed through the system. What mattered was that for an investment of 10 BTUs, we got back a million. Calling the 1 million an input is why this is a misleading metric.

    Think of it another way. Let’s say I invent a metric: BTUs of food required to make 1 BTU of fuel. Is that a good metric? It will show ethanol in a very poor light. But of what use is that metric, other than to further an agenda? That’s the way I view the Wang metric. As I indicated, he has a history of these sorts of accounting games.

    And you both agree that he’s not admitted to being misleading.

    But we also agree that people are being misled. Whether he admits to being misleading is irrelevant, given that his work is misleading people.

    Gasoline and diesel are, but nat gas cannot be put into a conventional gasoline tank.

    Neither can ethanol. Hence, we have a big political push for E85 vehicles and E85 pumps. What we should be doing is simply converting more of our fleets to natural gas, and skipping the ethanol middle-man. It does not take an extreme conversion to build a natural gas vehicle. Interestingly, Brazil, for all the ethanol hype, has almost 10 times the natural gas fleet that we do here in the U.S. That never seems to get any airtime in the U.S.

    Even if the inputs and outputs sold for the same price per BTU, a 30% gain could be economic as long as the processing costs were small enough

    You do realize that the 30% gain includes animal feed byproducts valued as BTUs, right? As far as fossil fuels in and ethanol out, it is about 1:1. You could make a case for burning the animal feed for process heat, but then it has to be thoroughly dried out.

    I accept that a 5:1 or 3:1 return is a projection.

    The problem is, people are already running with it as fact. Look at Khosla’s comments, and maybe you will begin to understand my point of view. It is very bad energy policy to make such extravagant claims, because they lull society into thinking everything is just peachy. Peak Oil? No problem, we’ll just switch to ethanol.

    I believe (I think based on what you’ve said) that E3 still use nat gas, because they don’t get enough biogas. But they could instead use biomass gasification.

    They could, and I am a fan of biomass gasification, but then they wouldn’t have anything left to feed the cattle.

    It is perfectly reasonable to say that fossil fuel BTU’s taken from the ground are the input…

    Not in my opinion, because it is an arbitrary starting point. It ignores the very small BTU investment required to extract the oil from the ground. The starting point for both processes should be with zero BTUs in hand. Now, tell me how many it’s going to take to produce gasoline, versus ethanol. That’s what this debate is about.

  14. If it weren’t for other things than usable vehicle fuel being counted as energy output for ethanol, I would regard Wang et al.’s metric as almost legitimate, though potentially misleading.

    The only problem in that case is that they are presumably only counting the fuel actually used in the plant and in fertilizing, etc., rather than the total fossil fuel that had to be mined to get it.

    However, since they are counting coproducts, their metric is total BS, since the corn input already could be used for such things, and the energy content of the corn wasn’t counted in the energy inputs.

  15. Chris said, “I think perhaps you’re forgetting about constraints other than energy return. The quantity of arable land, for example, is relatively limited.”

    Chris,

    Of course my example is hyperbolic, but that’s just the point. The 160% return they are talking about doesn’t even work for one or two cycles while we still have arable land available.

    I know my example wouldn’t work because that 160% return on energy invested in each production cycle represents a geometric progression — and Wang, Shapouri, and “Tom” should realize that.

    * They key in my argument is that an ethanol production process that actually returns 160% of the energy invested would not need to be located near a high-capacity natural gas pipeline — which is a prime criterion when deciding where to build ethanol plants.

    The point that Vinod, Wang, Shapouri, and “Tom” all miss is that no one has yet demonstrated that growing corn and turning it into ethanol is not utterly dependent on a contunual injection of fossil fuels.

    If ethanol production actually returned 160% of the energy invested, the ethanol industry would have long ago set up a practical demonstration showing that they can grow corn and turn it into ethanol without fossil fuel inputs, if for no other reason than to quiet the naysayers such as Robert and me.

    They haven’t tried to do that, because they know too much would be riding on an attempted demonstration. It is easier to wave around a paper study from Wang and Shapouri than to lay everything on the line and attempt to prove it.

    If their demo failed (which I think it would), their industry would be turned upside down.

  16. In your example the process and extraction energy requirement is negligible and the fossil fuel getting to the tank is essentially the same as that taken out of the ground.

    Not bad, but we could also say add hydrogen from renewables (say for hydro-cracking), and have more energy in the fuel getting into the tank than we had to take out of the ground.

    I have no fundamental problem with your newly invented metric of BTU’s of food to BTU’s of fuel (it’s a little over fifty percent I believe, ie 10 BTU of corn give about 5-6 BTU of ethanol). That is in fact useful information (a little quible though, not all of the heating value of corn is actually available as food, some will be fibre and therefore go right through the digestive system).

    I don’t think it’s that hard to tell people that process and extraction energy requirements for ethanol are quite high for current ethanol production (several times as high as for gasoline), that this results in substantial fossil fuel consumption, and greenhouse gas emissions, but still less so than for conventional gasoline (by nearly a factor two), and even less than for CTL (by nearly a factor three).

    That paragraph neatly summarises a lot of information without much of an agenda either way, … well or so I hope at least.

    As for people being mislead, I think this cuts both ways really. It often happens that people confuse net energy with net gasoline, ie they think that a gallon of ethanol production requires a gallon of gasoline and so there is not net addition to the liquid transportation fuel pool. They often don’t realise that the vast majority of the input is nat gas and coal and the proper comparison for those I think is CTL and GTL.

    Ethanol can be added to conventional gasoline, as a fuel additive, and while I don’t have the exact figure, I should think that well over 90% of ethanol is used as an octane enhancing fuel additive in the United States. In truth I think for that function it should be compared with alternative ways to enhance octane, and the fossil energy inputs required for those (though I am not aware of anybody having done so, it might be too complex a task).

    E85 is only really required, if the US wanted to go above 10% ethanol in the gasoline pool. The modifications required to make vehicles E85 capable are very minor. Nat gas conversions are more capital cost intensive and hit range (and/or cargo capacity) more than E85 conversions. I am well aware of nat gas as a vehicle fuel, there’s a big push in Argentina and a few other countries.

    It is in fact one of the ways of dealing with “peak oil”, displace nat gas for heating and electricity (by coal/nuclear/renewables/heat pumps/insulation/efficiency), and then use the nat gas as vehicle fuel. The main issue with this approach is that nat gas is extremely convenient and low capital cost for heat and electricity, while it is not as a vehicle fuel.

    From a pure efficiency point of view, it would be better to burn corn in stoves, replacing nat gas, and then use the nat gas in vehicles, than to turn the corn into ethanol to displace vehicle fuel. But taking capital costs and convenience (storage space on board, storage space for corn in single family housing) and operating costs (eg maintenance for corn fired boilers) into account, I think you’ll find that less of a subsidy is required (well, none at the moment) for the ethanol route than for the corn stove/nat gas route.

    It’s fair enough to include by-product BTU’s. I believe Wang includes them based on displacement value (ie fossil fuel saved by not having to grow corn/soybeans as an alternative cattle feed), rather than on their actual heating value (which would be considerably higher).

    For biomass gasification I was thinking about corn stover, straw, switchgrass or plain wood, rather than distillers grain.

    When you write about peak oil and net energy and ethanol, I think you imply that a ratio of 1.3 in and of itself would be a deciding factor. But the real issues are things like land and water and environmental impacts. The net energy of 1.3 could just be dealt with by having coal (or woody biomass and wind, which also have a net energy around 30:1) as the process and extraction energy input.

    The energy problem is one of liquid fuels, and of primary energy. As long as energy sources other than ethanol are available to provide process and extraction energy (with high net energy ratios, eg coal, nuclear, hydro), then ethanol’s low net energy (even assuming for the sake of argument that ethanol’s net energy ratio was not improvable) is not a key limiting factor.

    How to best address peak oil is really a different, though closely related question. I think ethanol, based on present technology, would only be a small part of the solution. But not because of net energy, but because of its land requirements and the concerns too much land use for ethanol would raise about biodiversity/monocultures/environmenntal impacts.

    CTL, GTL, BTL, nat gas cars, hybrids and renewables/nuclear for process energy/hydrogen going into CTL, GTL, BTL, I think, based on present technology would carry the majority of the load.

  17. Hi anonymous,

    the basic idea is to count all the fossil fuel going in, including the fossil fuel required to produce the fossil fuel required to produce the (etc.) fossil fuel inputs. That’s not too difficult, a BTU of coal gets counted as 1.11111… BTU’s of fossil fuel (1 the coal itself, 0.1 the energy required to produce the coal, 0.01 the energy required to produce the energy to produce the coal etc..).

    They do neglect plant construction, saying it wouldn’t materially impact the overall results (and it doesn’t, because that input is small) and getting good data isn’t worth the bother then.

    The displacement method (ie counting the fossil fuel saved by not having to grow an alternative cattle feed) for co-products seems a good one to me.

  18. Gary,

    there are two principal reasons why your logic does not hold. Firstly, the energy inputs and outputs are not equivalent. Coal sells for $2 per MMBTU, gasoline for something like $14 per MMBTU. A plant using 1 BTU of coal to produce 1 BTU of a gasoline replacement can be very profitable. This is an energy conversion more than an energy production process.

    Secondly, energy is not the only input into energy production. Most of the cost, particularly so for high net energy sources like wind or nuclear power is in land, labour, steel and the like.

    It’s therefore not enough to just reinvest the energy output, the reinvestment needs to include land, labour, steel etc., and on that basis reinvestment will nearly always look much less favourable than implied by net energy (which for wind is above 30:1).

  19. I believe it was MonteQuest over at peakoil that stated you don’t make gasoline but rather you refine oil. And I believe that is where more people go wrong when talking about ethanol versus gasoline.. I could be wrong though..

  20. They do neglect plant construction, saying it wouldn’t materially impact the overall results (and it doesn’t, because that input is small) and getting good data isn’t worth the bother then.

    Actually, the USDA reports say they neglected secondary inputs because they didn’t have any good data on them. But even though the input is small, so is the net energy produced. Therefore, even a small secondary input can cause a large percentage decrease in the net energy.

  21. I believe it was MonteQuest over at peakoil that stated you don’t make gasoline but rather you refine oil. And I believe that is where more people go wrong when talking about ethanol versus gasoline.. I could be wrong though..

    That’s actually a good way of putting it.

  22. I suspect you have probably misunderstood what they mean by secondary inputs, because otherwise your comment does not make sense to me. After all, I explicitly said that energy costs for the construction of farm machinery or ethanol plants were excluded by Wang as being negligible.

    They do account for the energy used to produce the fossil fuel inputs going into ethanol production though.

    http://www.transportation.anl.gov/pdfs/AF/265.pdf#search=%22Wang%20ethanol%20net%20energy%20an%20update%22

    “The estimates in table 4 also include the energy used
    to mine, extract, and manufacture the raw materials
    into the final energy product. The sum of these
    energy values was included in the estimates to derive
    the total energy associated with each farm input
    required to produce a bushel of corn. Input efficiencies
    for fossil energy sources, which were estimated
    with Argonne’s GREET model, were used to calculate
    these additional energy input values. In particular,
    GREET estimated the energy efficiency of
    gasoline (80.5 percent), diesel fuel (84.3 percent),
    LPG (89.8 percent), natural gas (94.0 percent), coal
    (98.0 percent), and electricity (39.6 percent).”

    As for small inputs making a big difference percentage wise. The difference in net energy between a ratio of 1.01 and 1.1 is 1000%. Between 1.01 and 1 it goes infinite.

    When the ratio is 1 or less, the process will only make sense as an energy conversion (or storage) technology. In that case, it’ll be coupled with a high net energy source, like say coal and even a net energy ratio of 0.8 wouldn’t be a problem.

  23. Reno,

    and nobody makes ethanol then I suppose they all refine corn.

    So, yes, the crude oil is refined into gasoline and other products and the original C is still (mostly) there in the refined gasoline and hasn’t been burned yet, it merely gets burned in the car and in the refining is only transformed/distilled/separated.

    In the case of ethanol, none of the fossil fuel carbon input is in the finished ethanol product. All the fossil carbon going into ethanol is burnt before it can get to the car. The ethanol burnt in the car only releases renewable carbon.

    And so what?

  24. I explicitly said that energy costs for the construction of farm machinery or ethanol plants were excluded by Wang as being negligible.

    I know that’s what you said. But that’s not why they were excluded. From their 2004 report (Shapouri, Wang, and Duffield):

    Energy used in the production of secondary inputs, such as farm machinery and equipment used in corn production, and cement, steel, and stainless steel used in the construction of ethanol plants, are not included in our study. Available information in this area is old and outdated.

    They didn’t include the information because they didn’t feel like the information was current. They didn’t say a word in that report about the numbers being negligible. So, their 1.27 energy return (which includes co-products) is actually less by whatever is the amount of the secondary inputs. Given that the EROI is already so marginal, they need to include all inputs.

  25. …it merely gets burned in the car and in the refining is only transformed/distilled/separated.

    Look back at the exchange that started this. When Tom wrote to me, he was writing to defend this statement:

    As you can see, the fossil energy input per unit of ethanol is lower—0.74 million Btu fossil energy consumed for each 1 million Btu of ethanol delivered, compared to 1.23 million Btu of fossil energy consumed for each million Btu of gasoline delivered.

    I am sure you will agree that this is not only misleading, but it is flat out false. The clear implication is that we are better off using our BTUs to make ethanol, since it is more efficient to do so. That is my complaint. It is a false argument.

  26. I think when he says “consumed”, this is supposed to be synonymous with “taken out of the ground”. I don’t think he means “consumed” as in “burnt to make the gasoline”. Anyways, that’s semantics really.

    ——–

    I only looked at the last paragraph of the post you wanted me to read. Yes, using coal rather than nat gas makes ethanol less green (as in means higher CO2 emissions).

    But I mentioned coal as a possible input, because it’s one of many (others would be nuclear or wind or wood or hydro) that have a high net energy ratio and could be used to provide the (currently) relatively high process energy input for ethanol production. It was meant to show that a net energy ratio of less than 1 isn’t a problem, if we are in effect talking a conversion process (such as nuclear energy + corn to ethanol).

  27. I think when he says “consumed”, this is supposed to be synonymous with “taken out of the ground”. I don’t think he means “consumed” as in “burnt to make the gasoline”. Anyways, that’s semantics really.

    When a word is taken and the meaning is completely changed, this is not semantics. If I fill up a 1 gallon jug with water, and you had 1 gallon of water, but drank it, we did not each consume 1 gallon of water. I still have my gallon available for consumption. You don’t.

    This argument has caused all manner of people, including Vinod Khosla, to claim that it is twice as efficient to produce ethanol as it is to produce gasoline, when it is in fact 4-5 times more efficient to produce gasoline. You apparently condone this misinformation, given that you think this is merely a semantics argument, and you would rather defend what you think they mean, as opposed to what they actually said, and how people are using their argument. But when an argument sows this much confusion and causes so many ethanol proponents to exaggerate the benefits of their product, it is not merely a semantics argument.

    Here is something else they did that shows a pattern of this kind of behavior. Let me know if you condone this. In a previous essay, I discussed their 2002 report (Shapouri, Wang, and Duffield):

    The actual energy inputs into the process according to the authors are 77,228 BTUs per gallon of ethanol produced (using the higher heating value, or HHV). The BTU value given for a gallon of ethanol (HHV) was 83,961. Therefore, excluding co-product credits, the EROI would appear to be 83,961/77,228, or 1.09. They include a co-product credit of 14,372 BTU, which should raise the overall value of the BTU products to (83,961 + 14,372), or 98,333 BTUs. This would imply an EROI of 98,333/77,228, or 1.27.

    However, they then perform a completely illegitimate accounting trick to exaggerate the EROI of ethanol. They use the 14,372 co-product credit to reduce the energy input of 77,228 and assumes an energy input of just 62,856 BTUs/gallon. Since the co-products are not actually used as inputs in the process, this is invalid. But that is not the most serious issue. When they use the co-product credit to offset the energy input, it should be removed from the product side. They include it on both sides of the equation – reduce the inputs with the co-product credit, and increase the BTU output with the co-product credit.

    Consider this analogy. I invest $100, and I get a return of $20 and another $40 worth of goods (co-product). What is my return on investment (ROI)? Most people would say that I got a total return of $60 on an investment of $100, for an ROI of 60%. If we utilize their accounting practices, we would use the $40 co-credit to offset our initial investment. We would then argue that we only invested $60 to get a return of $60, for an ROI of 100%. So, the answer to the question – “When does a $60 return on a $100 investment amount to a 100% return on investment?” – is “Whenever the USDA is doing the accounting.”

    So, is this once again semantics? What I see is a pattern of deceptive accounting designed to exaggerate ethanol’s appeal. You may not have a problem with it. I do.

  28. I can’t believe that some people have a hard time understanding why using a word like “consume” or “consumed” in different ways when referring to corn ethanol and regular petroleum production is a real problem. When you “consume” something, you use it up.

    When you “consume” 1 BTU of energy to produce 1.4-1.6 BTU’s of ethanol, you are left with only a surplus (net gain) of 0.4 to 0.6 BTUs of energy and even that return relies on counting the “energy content” of byproducts like animal feed in addition to the fuel.

    When you “consume” 0.23 BTU’s of petroleum/fossil fuels to produce 1 BTU of end product such as gasoline, you are still left with a net energy of more than three times the input.

    If in fact it took 1.23 BTU’s of “consumed” energy to produce a single BTU of end product with fossil fuels, you’d have a process with a net negative energy balance and the best thing that could be done to improve our energy situation would be to stop producing any fossil fuel/petroleum energy at all. In fact, if we must “consume” 1.23 BTUs for every BTU produced, where in the name of God has all that excess energy come from to power our “net energy losing” oil based economy?

    Sorry, but this is not just semantics. The people making these claims MUST know that they are misleading. If they don’t know that these claims are misleading (and in fact just wrong), then they are too stupid to be engaged in this debate, much less engaged in anything to do with setting energy policy.

  29. Robert and The Ox,

    you are trying to pin the word “consume” to net energy, rather than fossil fuel use. Anybody who reads the factual information will grasp the difference, and people who don’t will easily be led to other unintended misunderstandings, and as I mentioned, particularly prevalent is the notion that net energy for ethanol is equal to net gasoline/liquid transportation fuel production. That’s what happens when you only emphasise the low net output, without mentioning that the net is calculated by comparing input BTU’s of coal and nat gas with output BTU’s of ethanol.

    The Ox,

    for CTL 2 BTU of coal input will yield 1 BTU of transportation fuel output, no problem because the coal itself is produced at very high net energy. Whether the coal is “consumed” by CTL, I think is really semantics. It’s no longer there, the carbon in the finished synthetic fuel will come from the coal, the hydrogen may largely be water derived. For that matter, the crude oil will also no longer be there after refining, and refining involves more than mere crude oil component separation, particularly for gasoline, where operations such as hydro-cracking (with nat gas derived hydrogen) will be performed that result in different chemicals being in the finished gasoline product than in the crude oil “consumed” to make the gasoline.

    But as said, that is semantics, it’s clear to me that “Tom” referred to fossil fuel input, not net energy. And I might say that your whole post is written in a way that one might be led to believe you’d actually believe yourself that net energy=net liquid transportation fuel, without realising that ethanol production is also a conversion process with inputs other than ethanol (particularly coal and nat gas) being converted into a liquid transportation fuel.

  30. Robert,

    http://www.usda.gov/oce/reports/energy/aer-814.pdf

    I don’t see any double counting there.

    The net energy ratio including co-products is 83961/62856=1.34. They did not include the co-product credit twice. If they had double-counted, as you suggest they would have gotten a ratio of 98333/62856=1.56.

    They subtract the co-product credit, because they allocate the input energy to ethanol and co-products, ie some of the input energy goes towards producing co-products rather than towards producing ethanol.

    That makes sense, because the co-product credit is calculated on a displacement basis, ie how much energy would be required to produce alternative cattle feed, if the dried distillers grain was not available.

    If the actual heating value of the co-products was considered, a better net energy ratio would result, even with the co-product credit being added to the output side. For heating value I think that would be appropriate, because the co-products would then be considered as an energy product. If the co-products are not considered as an energy product, they can only be included by considering how much energy inputs went into producing them, or would have gone into producing an alternative non-energy product (such as soybeans as cattle feed).

  31. The net energy ratio including co-products is 83961/62856=1.34. They did not include the co-product credit twice. If they had double-counted, as you suggest they would have gotten a ratio of 98333/62856=1.56.

    My reference was to their 2004 report. They did double-count, and they reported an energy ratio of 1.67. So, after that report, ethanol proponents everywhere started claiming that ethanol returns 1.67. Google ethanol and 1.67 and you will see for yourself. Then let me know whether you admit they double-counted.

  32. My reference was to their 2004 report.

    Oh, sorry, I did say their 2002 report. But I meant their 2004 report. That is the one in which they claim the 1.67 energy return by double-counting. They also do some questionable accounting in their 2002 report. They report an energy return of 1.34, but they did that by reducing the inputs by the amount of co-product credits. The real ratio of energy out over energy in is 1.27.

  33. http://www.biomass.govtools.us/pdfs/net%20energy%20balance.pdf#search=%22ethanol%20and%201.67%20%22

    You said 2002 actually, and the numbers were the same as in the link I gave to their actual report. The 2002 report has an energy input of 77,228 and there is no mention of a net energy ratio of 1.67.

    In the above report, while the net energy ratio of 1.67 turns up, the input is not 77,228.

    You seem to be confused about what report you are actually talking about.

    And there is no double counting in this report either. They get a higher co-product credit, because they use a different methodology for dividing up the energy input between ethanol and co-products (see page 4).

  34. We seem to have posted at the same time (blogger is a bit slow today too).

    Subtracting the co-product credit on the input side is the most appropriate option when we are talking co-production of an energy product (ethanol) and a non-energy product (cattle feed). The non-energy product is dealt with by considering how much energy would be required to produce an alternative to the non-energy co-product (say soybeans instead of dried distillers grain).

    Conceptually, the input energy is split, as if there were two processes, one yielding cattle feed, and another yielding ethanol.

  35. I strongly agree with Heiko here. Robert’s methodology is absurd.

    For the ethanol case, we will start with 1 barrel of “fossil energy”, take it to the ethanol factory, and end up with 1.3 barrels of ethanol energy.

    For the gasoline case, we again start with 1 barrel of “fossil energy”, but instead of taking it to the “gasoline factory” (ie. refinery), Robert ships it back to the Middle East! He uses that 1 barrel in Saudi Arabia to produce 10 barrels of fossil energy. He then ships those 10 barrels back to the U.S. where they go to the gasoline factory and produce 8 barrels of gasoline. Presto, he has got much more gasoline from his starting 1 barrel than he did for ethanol.

    Surely the sleight of hand here should be obvious to everyone. You can’t add an extra step of fossil production when accounting for gasoline, if you’re not doing so for ethanol.

    The bottom line remains that to produce enough ethanol to drive a car 100 miles takes LESS fossil energy out of the ground than to produce enough gasoline to drive a car that far. If Robert disagrees with this last fact, I’d like to hear him say it.

  36. I strongly agree with Heiko here. Robert’s methodology is absurd.

    Hal, if this is your best argument, I strongly advise you to heed the warning at the end of my essay.

    For the ethanol case, we will start with 1 barrel of “fossil energy”, take it to the ethanol factory, and end up with 1.3 barrels of ethanol energy.

    Your mistake is in presuming we are talking about literal barrels of oil. A barrel of oil is only a metaphor for the BTU content. You don’t make ethanol out of oil, and you don’t extract oil from the ground using oil. We use various forms of energy that can be represented as BTUs. If I have X BTUs, then I can invest it to make 5 BTUs of gasoline, or 1.3 BTUs of ethanol. That is the bottom line on the efficiency argument.

    You can’t add an extra step of fossil production when accounting for gasoline, if you’re not doing so for ethanol.

    Actually, the full life-cycle is being analyzed for ethanol, but not for gasoline. The full life-cycle for gasoline doesn’t start with a barrel of oil waiting to be refined. It starts with oil buried under ground. When you compare life-cycle to life-cycle, efficiency to efficiency, or EROI to EROI, gasoline comes out ahead. The only time ethanol comes out ahead is if you mix and match.

    The bottom line remains that to produce enough ethanol to drive a car 100 miles takes LESS fossil energy out of the ground than to produce enough gasoline to drive a car that far. If Robert disagrees with this last fact, I’d like to hear him say it.

    Do I disagree? Of course I do. Your statement is absurd. Let’s say it takes 1 BTU to drive the car the preferred distance. To net 1 BTU of ethanol, I am going to have to physically consume (1/0.26) = 3.84 BTUs of fossil fuel. Remember, when making ethanol, you have to burn 0.74 BTUs of fossil fuel just to make 1, for a net of 0.26. If your desire is 1 net BTU of energy, then you will burn 3.84 BTUs of fossil fuel to end up with 4.84 BTUs of ethanol. If you are merely concerned about the gross, then you still burned 0.74 BTUs of fossil fuel to produce 1.0 BTU of ethanol.

    To net 1 BTU of gasoline, I am only going to have to physically consume 0.26 BTUs of fossil fuel to get the oil out of the ground, refine it, and get it to your tank. I had to pull 1.26 BTUs of fossil fuel out of the ground to net 1.0 BTUs of gasoline. To net 1.0 BTUs of ethanol, I had to pull 3.84 BTUs of fossil fuel out of the ground (actually more, since it would have to be refined). Again, ethanol only wins here if comparing apples and oranges (net energy of gasoline to gross energy of ethanol).

  37. Robert – Let’s see if we agree with the following facts.

    You take 1 unit of energy into the ethanol plant, and walk out with 1.3 units of ethanol energy. This is a net gain of 0.3 units.

    You take 1 unit of energy into the refinery, and walk out with 0.8 units of gasoline energy. This is a net loss of 0.2 units.

    Aren’t these the agreed-upon facts that we are starting with?

    Let’s say it takes 1 BTU to drive the car the preferred distance. To net 1 BTU of ethanol, I am going to have to physically consume (1/0.26) = 3.84 BTUs of fossil fuel.

    I am not sure you are using the term NET correctly. In this example, you don’t want or need to NET 1 BTU of ethanol in order to HAVE 1 BTU of ethanol. NET measures energy out minus energy in, by definition. What we want is 1 BTU of ethanol out, because that is what we need to do our little drive. We don’t need 1 BTU net. If we did as you suggest and consume 3.84 BTUs of fossil fuel, then at 1.3 BTUs out per 1 BTU in, we’d end up with 1.3*3.84 = 5 BTUs of ethanol, enough to drive 5 times the required distance. I think your use of NET here is confusing you.

    To net 1 BTU of gasoline…

    Again, NET gain is not the issue. The issue is how much fossil fuel energy comes out of the ground in order to produce a certain amount of gasoline vs ethanol energy. Do you still claim that it takes MORE fossil fuel energy out of the ground to produce a certain number of BTUs of ethanol, vs that same number of BTUs of gasoline?

    If so, please tell me exactly. How many fossil fuel energy units come out of the ground to produce (i.e. walk out of the ethanol factory with) 1 BTU of ethanol, and how many come out of the ground to produce (i.e. walk out of the oil refinery with) 1 BTU of gasoline?

    Isn’t it clear that the answers are 1/1.3 for ethanol and 1/0.8 for gasoline?

  38. [Note: I replaced the original comment with this one. I included the wrong well-to-pump ethanol figures in the first version]

    Well, we’ve got quite a debate here. Thanks for taking this to a public forum, Robert (and thanks to Heiko for actively engaging in this debate).

    First some background: I recently completed a well-to-wheels analysis of several dozen alternative fuel/vehicle pathways including ethanol (from corn and cellulosic biomass) and reformulated gasoline. The full report can be found here.

    I performed a fairly extensive survey of the extant debate on ethanol’s net-energy return on investment and did my best to utilize the best and most accurate data from throughout the literature to develop the inputs for my well-to-wheels modeling. (I utilized Argonne’s GREET model, so I’ve got to admit, I’m a bit indebted to Michael Wang et al. at Argonne).

    First off, I agree that comparing two different metrics (efficiency and EROI) can be very misleading. Furthermore, the use of the EROI metric itself has the potential to be quite misleading.

    Focusing solely on net-energy ratio can result in misleading results, particularly when the metric is considered ‘in a vacuum’ and not compared to the fuel that the alternative fuel is likely to replace – i.e. gasoline. In particular, a net energy metric ignores the fact that not all fossil fuels ‘are created equal’ – that is, there are vast differences in the energy, environmental, and policy implications of the use of various fossil fuels (coal, petroleum and natural gas) that a simple net energy metric ignores.

    Furthermore, a net energy ratio does not provide a sufficient environmental metric either, as it is not an accurate indicator of emissions of GHGs or criteria pollutants, or of other environmental factors including soil erosion or deforestation.

    Additionally, focusing on a net energy ratio for a given fuel obscures the fact that not all forms of energy are equally valuable. For example, electricity is clearly more valuable than the potential fossil energy in coal, natural gas or petroleum, which is why we routinely accept ‘negative’ net energy ratios for electricity generation. Likewise, liquid fuels for transportation are considered more valuable than the various feedstocks that are used to produce them. Thus, the direct comparison of various fuels for use in specific contexts using multiple energy and environmental metrics yields the most valuable insights into the relative benefits and costs of these fuels.

    And finally, the issue of which energy inputs to include in EROI can further confuse the use of the EROI metric. For example, some studies go so far as to include energy in solar radiation absorbed by plants, while others restrict their focus only on fossil energy inputs.

    For these reasons, my study provided several different metrics to compare alternative fuels and vehicles, both with the baseline fuel (gasoline) and with each other. I presented results for well-to-wheels total, fossil and petroleum energy use, as well as emissions of the three main GHGs and five harmful pollutants. It is my hope that the several metrics included in my study (i.e. total, fossil and petroleum energy, GHG and criteria pollutant emissions), as well as the easy and objective comparison of each fuel to gasoline and the other alternative fuels will provide a more accurate analysis of the merits of the various fuels included in the study than a simple net energy metric.

    However, I did provide net well-to-pump (fossil) energy ratios for the various fuel production pathways analyzed so as to allow comparison with other literature.

    [A note on the study: all results are modeled for the year 2025, not for the present, so some efficiency improvements in ethanol production are assumed while some decrease in petroleum recovery efficiency are assumed.

    Estimates are conservative. Credit for ethanol co-products are allocated based on the displacement method – i.e., energy used to grow/produce products displaced by the ethanol co-products is subtraced from ethanol’s inputs.

    Energy contained in the fertilizers, pesticides, herbicides, etc. are included, as are fuels for farming machinery. Embodied energy in infrastructure and machinery is excluded.

    Energy input calculations are recursive – i.e., the energy inputs to produce the coal used to produce ethanol are included, and the inputs to produce those inputs and so on up to about 100 iterations].

    Here’s what I found on a Well-to-Pump basis – that’s all the fuel production and transportation stages on through to the fueling station – for ethanol and reformulated gasoline:

    Ethanol (E100 from corn) – Btus per million Btus fuel available at fueling stations: fossil energy inputs: 701,834 Btu; petroleum energy inputs: 72,945; EROI (fossil energy inputs): 1.42.

    E85 (85% ethanol from corn and 15% reformulated gasoline by volume) – Btus per million Btus fuel available at fueling stations: fossil energy inputs: 618,425 Btu; petroleum energy inputs: 83,966; EROI (fossil energy inputs): 1.62.

    Reformulated Gasoline (30-ppm sulfur content, 5.7% ethanol from corn by volume as oxygenate) – Btus per million Btus fuel available at fueling stations: fossil energy inputs: 301,759 Btu; petroleum energy inputs: 125,831; EROI (fossil energy inputs): 3.31.

    So, on a well-to-pump basis, gasoline clearly yields considerably more energy per Btu of input (fossil or total) than ethanol. These results are largely consistent with Wang, et al.’s own results in their Well-to-Wheels studies. The use of the 0.8 efficiency figure in comparison with ethanol’s EROI is clearly misleading, and their own study’s results show that gasoline’s EROI is considerably better than ethanol.

    However, that’s not to say that there is no benefit to ethanol. Ethanol achieves moderate reductions in petroleum energy use as well as GHG emissions on a well-to-wheels basis (i.e., total well-to-wheels inputs/emissions per vehicle mile travelled fueled by ethanol).

    On a well-to-wheels basis, I don’t compare E100 to gasoline, as I could not find suitable vehicle emissions figures for E100 vehicles. I instead compare to E85 which is what would most likely be used anyway.

    So, on a well-to-wheels basis, here’s how E85 and gasoline compare [the vehicle is a 22 mpg-gasoline equivalent spark-igntion vehicle meant to represent the average stock vehicle in 2025 as predicted by the EIA]:

    Reformulated gasoline (30-ppm sulfur content, 5.7% ethanol from corn as oxygenate) – Btu or grams/vehicle mile travelled: fossil energy: 6,523 Btu; petroleum energy: 5,616 Btu; total GWP-weighted GHG emissions: 519 grams-CO2 equiv.

    E85 (85% ethanol from corn and 15% reformulated gasoline by volume) – Btu or grams/vehicle mile travelled (% change relative to gasoline): fossil energy: 4,264 Btu (-34.6%); petroleum energy: 1,507 Btu (-73.2%); total GWP-weighted GHG emissions: 392 grams-CO2 equiv. (-24.6%).

    So, on a well-to-wheels basis – and that’s what really matters anyway (i.e., how much do we put in for the same amount of work done, one vehicle mile travelled) – ethanol from corn provides a moderate reduction in fossil energy use and GHG emissions and a pretty sizable reduction in petroleum use, per vehicle mile travelled. (The reduction in petroleum use can’t get much better than it is since gasoline makes up about 21% of the E85 by energy content).

    I feel like we should focus on well-to-wheels comparisons of specific metrics of interest between two fuels, rather than a focus purely on EROI. If our goals are to reduce petroleum energy use, E85 from corn is a winner. If our goals are to reduce fossil use or GHG emissions, E85 from corn can help somewhat (although there are much better alternatives).

    The Well-to-Wheels metrics I focus on still don’t capture soil erosion, nitrogen run-off, and other environmental effects of ethanol production/corn farming, and these ought to be carefully modelled and considered as well. Nor do the capture scalability issues: even if E85 does do a good job reducing petroleum use, how many gallons of gasoline can it displace/how much can we grow/is it going to make much of a difference, etc.?

    Ethanol from corn is unlikely to ever be anything but a fuel supplement/additive, and in that job, it helps. It is in no way a ‘solution’ to our transportation energy problems though, and other alternatives must also be pursued.

    Still, while it’s EROI may not be as good as gasoline, ethanol from corn does offer clear advantages to gasoline based on well-to-wheels metrics.

    I hope that this (long) comment doesn’t simply confuse the debate further. I simply wanted to illustrate how a focus on EROI obfuscates more important metrics that can more objectively and informatively compare alternative transportation fuels.

  39. Hal, you write, “How many fossil fuel energy units come out of the ground to produce (i.e. walk out of the ethanol factory with) 1 BTU of ethanol, and how many come out of the ground to produce (i.e. walk out of the oil refinery with) 1 BTU of gasoline?”

    See my long comment above for the total well-to-pump and well-to-wheels energy inputs for ethanol and gasoline as that answers your question. Based on my findings, it takes about 700,000 Btu of fossil energy inputs to get 1,000,000 Btu of ethanol (E100) at the fueling station (or .7 Btu to get 1 Btu). It takes about 300,000 Btu to get 1,000,000 Btu of gasoline at the fueling station (or .3 Btu to get 1 Btu).

    Those figures include all of the fossil inputs needed to extract/farm the primary resources (crude oil or corn), transport that feedstock to refinery/ethanol production facility, refine/produce the fuel and transport and distribute that fuel to fueling station. It also includes the fossil energy inputs to make the agricultural chemicals and fertilizers used in corn production, as well as the secondary inputs needed to make the fossil fuels used as inputs in all of these stages. Basically, it includes the whole shabang, excepting energy embodied in machinery and infrastructure.

    So yes, it takes more fossil energy to produce a given amount of usable energy in ethanol as it does to produce a given amount of gasoline. In fact, it takes roughly twice as much fossil energy to produce ethanol as it does to produce gasoline.

    However, this is a misleading figure, as the energy that remains in the fuel, and is consumed when the vehicle moves, is largely fossil (petroleum) based in the case of gasoline and largely biomass (corn) based in the case of the ethanol. That’s why a full well-to-wheels metric makes more sense and allows a more objective comparison than a EROI or well-to-pump metric.

    As shown above, ethanol offers reductions in fossil energy use as well as petroleum energy use on a per vehicle-mile-travelled well-to-wheels basis.

    So, to drive your car one mile on E85 it will take 4,264 Btu of fossil energy, while it takes 6,523 Btu of fossil energy to drive one mile on gasoline.

    I hope that answers your questions, and again goes to show how fossil EROI is a confusing metric.

  40. It takes about 300,000 Btu to get 1,000,000 Btu of gasoline at the fueling station (or .3 Btu to get 1 Btu).

    This is thermodynamically impossible.

    You can’t take 300,000 Btu of fossil energy out of the ground and get 1,000,000 Btu’s of gasoline. The amount of fossil fuel energy you take out of the ground is GUARANTEED to be greater than or equal to the amount of gasoline energy you end up with. This is a consequence of the first law of thermodynamics. Add the second law and you can eliminate “or equal to”.

    Obviously you were not counting the total fossil energy that came out of the ground in your calculation.

  41. Hal, you are right, the well-to-pump figures do not count the energy still contained in the fuel. That’s why the well-to-wheels figures are what we should be concerned with. How many Btus of fossil energy does it take to move our vehicle the same distance when fueled with ethanol (E85) or gasoline? That’s what the well-to-wheels figures answer.

  42. You take 1 unit of energy into the ethanol plant, and walk out with 1.3 units of ethanol energy. This is a net gain of 0.3 units.

    If you draw your boundary at the ethanol plant, you walk in with 1 BTU of corn, and walk about with about 0.25 BTUs of ethanol. The rest of the energy went into cooking the mash and distilling off the ethanol from the (mostly) water solution. That is your apples to apples comparison to walking into the refinery with 1 BTU of energy and coming out with 0.8.

    I am not sure you are using the term NET correctly.

    Net is simply the energy outputs minus the energy inputs. If your purpose is to produce ethanol, you consume 0.74 BTUs to produce 1 for the full life-cycle. If your purpose is to produce gasoline, you consume 0.23 to produce 1 for the full life-cycle. By full life-cycle, I mean starting from nothing in both cases.

    If we did as you suggest and consume 3.84 BTUs of fossil fuel, then at 1.3 BTUs out per 1 BTU in, we’d end up with 1.3*3.84 = 5 BTUs of ethanol, enough to drive 5 times the required distance.

    Net is everything. When we input BTUs into the process, these are BTUs that could have been 1). Used in another process, or 2). Used directly as transportation fuel. So, in the case that we are actually trying to produce energy to run our society on, the net is the key. If we each use 1 million BTUs a day, then if all we care about is the gross then it wouldn’t matter if it took 10 million BTUs to make 1. But if we care about net, then no way would we waste 9 BTUs, unless there was a special circumstance in which the 10 were very cheap relative to the 1, or very abundant.

    Do you still claim that it takes MORE fossil fuel energy out of the ground to produce a certain number of BTUs of ethanol, vs that same number of BTUs of gasoline?

    You are confusing 2 arguments. How much fossil fuel energy came out of the ground has nothing to do with the efficiency of the process. That is a sustainability argument. But from an energy investment point of view (in other words, I have X BTUs to invest in any process) then it takes far more fossil fuel to make ethanol than to make gasoline. QED.

    Again, don’t confuse yourself by getting hung up on literal barrels. I may have to extract 1.2 BTUs of oil out of the ground to wind up with 1 BTU of gasoline. But I now have that gasoline available as transportation fuel, and I only consumed 0.2 BTUs to do so. I multiplied the BTUs I consumed by a factor of 5.

    If I want to make ethanol, your argument suggests that I could instead extract 1 BTU of oil and produce 1.3 BTUs of ethanol. In that case, I multiplied the BTUs I consumed by a factor of 1.3. But from an efficiency point of view, it is far better for me to reinvest that 1 BTU of oil into extracting 10 more BTUs of oil. That is most efficient usage of my available BTUs.

    From a societal standpoint, all that really matters (efficiency-wise) is how much I net on my BTU investment.

    The fatal flaw in the Argonne analysis is to treat 1 BTU of oil input into a refinery as equivalent to a BTU of input into the entire ethanol life-cycle. These things are not the same. I could instead take that BTU from the ethanol input and multiply it into 5 times as much gasoline by extracting and then refining oil.

  43. Hi Jesse (also known as Watthead),

    The figures for ethanol and gasoline petroleum energy inputs are interesting. They are lower for ethanol than for gasoline, ie 1 BTU of ethanol actually displaces more than 1 BTU of petroleum (based on your figures about 1.05 BTU of petroleum, because 1 BTU of ethanol requires 0.07 BTU of petroleum, but 1 BTU of gasoline requires 1.12 BTU of petroleum).

    The use of E85 for your well to wheel figures kind of hides that, particularly so, because E85 contains 21% by energy gasoline, which of course also diminishes the percentage savings for fossil fuel and greenhouse gases.

    Have you tried to incorporate differences in fuel economy between ethanol and gasoline? It has been suggested that due to ethanol’s very high octance higher compression ratios could be used and therefore higher efficiency obtained.

    “If our goals are to reduce fossil use or GHG emissions, E85 from corn can help somewhat (although there are much better alternatives).”

    I am not particularly happy with this sentence. How do you define “better” here? If you mean “percentage GHG reduction using corn based ethanol with similar production methods as currently employed”, then I readily agree there are better options, say battery cars charged from renewable energy sources like wind.

    If “better” refers to cost per tonne of GHG emissions reduction, then ethanol comes out very well indeed, because at the moment it’s cheaper to produce than gasoline, and the abatement cost is therefore negative (less than zero Dollars per tonne of GHG emissions reductions saved).

    The other part of the sentence I don’t like is the word “can”. Using present process inputs (ie nat gas or coal for distilling the alcohol), GHG reductions on a percentage basis are modest. But corn based ethanol can be produced differently. That’ll add cost, but on a cost per tonne of GHG emissions reduction basis, that cost is actually not all that high.

    I know that abatement cost estimates depend enormously on relative energy prices. They’d soar for ethanol, if gasoline wholesale prices came down, because the abatement cost depends on the difference between ethanol productions costs and gasoline wholesale prices.

    But, I do think abatement cost (ie the cost per tonne of GHG emissions reduction) is still a very important figure to include.

  44. To further clarify, Hal, the well-to-pump figures means that it takes 300,000 Btu additional energy inputs to get 1,000,000 Btu of fuel at the fueling station (i.e., you have to invest 300,000 Btu to yield 1 million Btus of fuel). If the primary feedstock for the fuel is a fossil energy source (i.e., petroleum) then there is still quite a bit of fossil energy embeded in the fuel. That’s how you end up with positive EROI and this again illustrates how the EROI differs from the efficiency of a process. You are correct that the efficiency of the process cannot be greater than 1, thermodynamically speaking (although the fossil energy efficiency of the process can be, if the process uses non-fossil energy inputs and yields more energy than the fossil inputs it requires)

    For example, if we consider the breakdown of the total well-to-wheels energy expended to move a 22-mpg gas equivalent vehicle 1 mile fueled with gasoline, about 76% of the fossil energy and 88% of the petroleum energy consumed over the well-to-wheels pathway in order to move the vehicle one mile is consumed during the vehicle operation stage and is not included in the well-to-pump stage’s figures.

    Sorry, this is all a bit confusing and it looks like I may have simply added more confusion to things, rather than clearing things up.

    My point in all of is that what we ought to do is focus on well-to-wheels comparisons, as these provide an objective means of comparing two or more transportation fuels and vehicle technologies with one another using a number of metrics. On these metrics, ethanol (E85) shows a moderate reduction in fossil fuel use and GHG emissions and a significant reduction in petroleum energy use.

  45. They are lower for ethanol than for gasoline, ie 1 BTU of ethanol actually displaces more than 1 BTU of petroleum (based on your figures about 1.05 BTU of petroleum, because 1 BTU of ethanol requires 0.07 BTU of petroleum, but 1 BTU of gasoline requires 1.12 BTU of petroleum).

    The other side of the coin is that ethanol production uses far more natural gas than gasoline production. The overall fossil fuel displacement is not very great.

    If “better” refers to cost per tonne of GHG emissions reduction, then ethanol comes out very well indeed, because at the moment it’s cheaper to produce than gasoline…

    I suspect this claim is in the category as “it’s twice as efficient to produce ethanol than to produce gasoline.” I hear people make the claim all the time that it’s cheaper to produce ethanol, yet the average annual rack price for ethanol has been significantly higher than gasoline for 25 years in a row. Either ethanol producers are the ones who are gouging us, or someone is lying about production costs.

  46. Robert,

    what’s the basis for the conversion efficiency? (0.25 BTU of ethanol per BTU of corn)

    Currently, in a petroleum refinery inputs other than petroleum (electricity and nat gas notably) are used in addition to petroleum.

    And in an ethanol plant, in addition to corn, nat gas and electricity are consumed, and ethanol is not the only product.

    We can sort this out in several ways, A) force all the inputs to be corn/petroleum respectively, or B) allocate the inputs and outputs, or C) do a straight comparison corn in / ethanol out, petroleum in / gasoline out (neglecting other inputs/outputs).

    For 10 BTU of corn input to an ethanol plant, something between 5 and 6 BTU end up in the ethanol, around 3-4 BTU are the dried distillers grain, and 3-4 BTU are losses.

    For 10 BTU of petroleum input into a refinery, something like 4-5 are in gasoline (in the US at least), 5-6 are in other products and there’s maybe 1 BTU of losses.

    If we forced ethanol to be the only output (eg burn the dried distiller’s grain for heat and electricity) and corn the only input, and otherwise used similar processing, 10 BTU of corn might yield 5-6 BTU of ethanol.

    If we forced petroleum to be the only refinery input, and gasoline the only output (eg by extensive hydrocracking, and production of hydrogen, process energy and electricity from the petroleum itself or distillation cuts unsuitable for direct incorporation into gasoline), 10 BTU of crude oil might maybe yield 6-7 BTU of gasoline.

  47. Heiko wrote: “Have you tried to incorporate differences in fuel economy between ethanol and gasoline? It has been suggested that due to ethanol’s very high octance higher compression ratios could be used and therefore higher efficiency obtained.”

    No, I assume that E85 achieve the same fuel efficiency (on a Btu/mile travelled basis) as a gasoline vehicle.

    How do you define “better” here?

    Better is certainly an ambiguous and subjective term, I’ll give you that. What I meant to imply is that there are other alternatives that offer larger reductions in fossil and petroleum energy use and GHG emissions, have greater potential for scalability, and thus have far greater potential to reduce GHG emissions.

    That being said, you are probably correct that there are not too many better options for reducing GHGs when considering the abatement price – although fuel economy improvements would certainly be better than ethanol on a cost/ton of GHGs reduced, right?

    However, ethanol from corn can only go so far, and while it may be a good interim step, it cannot be a major contributor to reducing GHG emissions in the U.S. I’m afraid that we will be blinded into thinking we’ve got a solution on our hands and are doing enough by investing in the development an ethanol infrastructure, when in reality, we must pursue other alternatives if we truly seek to slash our GHG emissions to the levels that have been deemed necessary by the scientific community – i.e., 60-80% below 1990 levels by 2050.

    You are correct, though, Heiko, that I do not focus on costs or economic analyses here, and economics will clearly be an important factor.

    Finally, as for your comment about 1 Btu of ethanol displacing more than 1 Btu of petroleum, I’m not quite following you. Can you rephrase what you mean?

  48. Jesse,

    “My point in all of is that what we ought to do is focus on well-to-wheels comparisons, as these provide an objective means of comparing two or more transportation fuels and vehicle technologies with one another using a number of metrics.”

    I think well-to-wheels studies are useful, but they don’t include things like cost, which can make the results near meaningless “in vacuum” as you so neatly said for net energy.

    So, we get a 30% GHG reduction for ethanol say, and a 50% GHG reduction for plug-in hybrids say. Which is “better”?

    That’ll depend on what we want to achieve. If we want to compare marginal cost of greenhouse gas reductions today, ethanol may come in at -$20 per tonne of GHG, and plug-in hybrids at $3000 per tonne of GHG.

    If we want to reduce US emissions to 10% less than 1990 levels by 2015 at least cost say we might find that building 150 GW of wind turbines and shutting down lots of coal fired generation is 50% of the solution and expansion of ethanol to 20 billion gallons per year 10% of the solution, and the cost to achieve that is $100 per metric tonne of carbon, and that that justified substantial capital investments into more efficient ethanol plants which give us well-to-wheel GHG reductions of 70% rather than 30%.

  49. Heiko wrote, “If we want to reduce US emissions to 10% less than 1990 levels by 2015…”

    Yes, if we are talking about reducing GHGs by 10% below 1990 levels by 2015, then ethanol might provide 10% of that reduction at least cost. And it should. But we’ve got to be thinking about a 60-80% reduction by 2050, and there’s no way ethanol from corn scales large enough to help in a significant way in achieving that goal.

    So yes, if we want a ready-to-go alternative fuel that offers least-cost GHG reductions, up to a certain point, then we should go for it. But let’s not delude ourselves, or the public, into thinking that ethanol from corn can do much more than contribute some small portion towards an overall strategy aimed at drastically reducing GHG emissions to levels 60-80% below 1990 levels. We’ve got to be looking for alternatives that can scale to that level, and we’ve got to be doing more than just looking; we’ve got to be developing strategies for implementing these alternatives as soon as possible to maximize the GHG reductions they can achieve.

    Ethanol from corn has it’s (small) part to play. But it’s not the solution. Some combination of fuel efficiency first, then a mix of cellulosic biofuels and electrification of transport is a solution that can scale and has a much better chance at achieving the targets we must set. (clean diesels and hybrids have their role to play too, but they can also fall under the fuel efficiency heading, as they basically make do with less petroleum-based fuel).

  50. what’s the basis for the conversion efficiency? (0.25 BTU of ethanol per BTU of corn)

    That was just off the top of my head. It was simply meant to put the argument on an apples to apples basis. But here is the actual calculation.

    In the 2004 USDA report, which was based on actual plant surveys, production of 1 gallon of ethanol consumed 50,000 BTUs of fossil fuel in just the ethanol plant. That ignores all other energy inputs, which means the boundary is drawn around the plant (just like the case of the oil refinery). The BTU value of ethanol is 75,000 BTUs. The co-product credit they took from the previous study was 14,000 BTUs. So, for an output of 89,000 BTUs, they had an input of 50,000 BTUs, meaning they consumed 56% of the net energy in the starting material (corn). Therefore, walking into the ethanol refinery with 1 BTU would result in a net of 0.44 BTUs of ethanol plus product. Walking into an oil refinery with 1 BTU of oil is going to result in a net of about 0.9 BTUs (the Argonne estimate of refinery efficiency is too low) of gas, diesel, fuel oil, etc.

    Bottom line, when the boundaries are drawn around the plants? It takes 0.56 BTUs of fossil fuels in the ethanol refinery to produce 1 BTU of ethanol and co-products. It takes 0.1 BTUs of fossil fuels for gasoline refining to produce 1 BTU of gasoline and co-products.

  51. Hi Jesse,

    we got both our replies in at the same time.

    Firstly, your final question. Net energy often gets twisted into suggesting that 1 gallon of ethanol will displace very little gasoline.

    At the gas tank the displacement value is straightforward, as you said in your reply, you basically assume 1 BTU of ethanol in the gas tank replaces 1 BTU of gasoline. But there’s also the petroleum going into the upstream processing of ethanol and gasoline respectively. And there ethanol comes out on top based on your analysis.

    On the upstream side, 1 BTU of ethanol in the tank requires 0.07 BTU of petroleum, and 1 BTU of gasoline in the tank requires 0.12 BTU of petroleum.

    Subbing ethanol for gasoline in the tank, leaves an upstream surplus of 0.05 BTU of petroleum, which could be refined into gasoline.

    In other words, subbing one gallon of gasoline equivalent at the gas tank with ethanol will free enough upstream petroleum to produce another 0.05*0.8=0.04 BTU of gasoline.

    ———————

    The abatement cost can be terribly dependent on energy costs. Say, consider a hybrid that displaces gasoline at a cost of $4 per gallon. If gasoline sells for $4 per gallon the abatement cost is zero, if gasoline sells for $3 per gallon, it’ll appear very high, as a gallon is very roughly about 1/250th of a tonne and we’d be looking at an abatement cost of something like $250 per tonne.

    Other abatement options may have very little dependence on energy prices (say tree planting).

    This I think is the biggest problem in using abatement cost for comparing different options. A small change in energy prices can make say hybrids look exceedingly expensive compared to tree planting, or exceedingly cheap.

  52. Robert,

    “So, for an output of 89,000 BTUs, they had an input of 50,000 BTUs, meaning they consumed 56% of the net energy in the starting material (corn).”

    This is a strange way to derive the energy in the corn. Why not go about it this way:

    Corn yields 2.6 gallons per bushel. Each bushel has 390,000 BTU. 2.6 gallons of ethanol have 200,000 BTU. And therefore the conversion efficiency is just over 50%.

    All right, maybe you want to include the fossil fuels. Make that 2.6 gallons times 50,000 then or 130,000 BTU per bushel of corn input, and corn and fossil fuels together come to 520,000 BTU for 200,000 BTU of ethanol + co-products. If we consider those an energy product and grant them 50,000 BTU, then the total input is 520,000 the output 250,000 and the efficiency just under 50%.

    The Argonne number of 0.8 is not for a refinery efficiency. It’s for the fossil fuel input to gasoline. And I should think, that they’ll do an allocation between the different products (gasoline, diesel etc.) that allocates a disproportionate share of the energy inputs to gasoline. I expect that, because gasoline gets the most processing (things such as hydrocracking).

  53. I want to go back to my first statement here, my 11:49 AM post:

    The bottom line remains that to produce enough ethanol to drive a car 100 miles takes LESS fossil energy out of the ground than to produce enough gasoline to drive a car that far. If Robert disagrees with this last fact, I’d like to hear him say it.

    And Robert replied, in the next posting, at 12:49 PM:

    Do I disagree? Of course I do. Your statement is absurd.

    I will note that Watthead confirmed my comment in his 3:35 PM post:

    So, to drive your car one mile on E85 it will take 4,264 Btu of fossil energy, while it takes 6,523 Btu of fossil energy to drive one mile on gasoline.

    He agrees with me that it takes LESS fossil energy out of the ground to drive a given distance with ethanol than to drive that same distance with gasoline.

    Then Robert replied at 4:34 PM. (I’m putting in these times because I can’t link to the individual posts, and I want to indicate which messages I am quoting from.) He echoed my question:

    Do you still claim that it takes MORE fossil fuel energy out of the ground to produce a certain number of BTUs of ethanol, vs that same number of BTUs of gasoline?

    and Robert replied:

    You are confusing 2 arguments. How much fossil fuel energy came out of the ground has nothing to do with the efficiency of the process. That is a sustainability argument.

    Note that he no longer calls my claim – the SAME claim I made above – “absurd”. Instead he tries to argue that it is irrelevant. I don’t know if it is or not, but my point is whether the claim is true.

    I still believe it is true. Watthead agreed. Robert earlier said it was absurd; now he seems to have backed off from that without actually saying so. I’d still like to know if Robert will acknowledge the factual truth of this claim.

  54. This is a strange way to derive the energy in the corn.

    No, it is extremely straightforward, especially since it is derived from actual ethanol plant energy usages. Nothing strange about that. It just means that my number is supported from the actual USDA study.

  55. I still believe it is true. Watthead agreed. Robert earlier said it was absurd; now he seems to have backed off from that without actually saying so. I’d still like to know if Robert will acknowledge the factual truth of this claim.

    The answer is the same as the one I gave you earlier. If your objective is to create 1 net BTU of energy, then you have to extract far more fossil fuel to make ethanol due to the poor efficiency of making ethanol. If you are only looking at gross BTUs, then it takes more BTUs from the ground to make gasoline, but you end up with a far greater percentage of BTUs after processing.

    Look at the difference:

    Gasoline: Extract 1.2, burn 0.2, gross almost 1.2, net 1.0.

    Ethanol: Extract 1.0, burn 1.0, gross 1.3, net 0.3.

    The gross gives more BTUs of ethanol, but that ignores all of the BTUs that went into refining the ethanol. That’s why the net is important.

  56. Robert,

    you don’t seem to have read what I actually said beyond the one sentence you quote.

    You don’t calculate the energy in the corn anywhere for your efficiency number. You just use the fossil fuel input and the ethanol and co-product output numbers.

  57. The use of the 0.8 efficiency figure in comparison with ethanol’s EROI is clearly misleading, and their own study’s results show that gasoline’s EROI is considerably better than ethanol.

    Frankly, that is the simple bottom line as far as the efficiency argument goes. That was the whole point of my challenging the Argonne argument. It is not more efficient to produce ethanol than to produce gasoline, and those who claim that it is are wrong. The other issues must be argued on their own merits.

    By the way, I have been running GREET simulations as well. I have come to pretty similar conclusions as what you presented. I may have some questions for you at some point.

  58. You don’t calculate the energy in the corn anywhere for your efficiency number. You just use the fossil fuel input and the ethanol and co-product output numbers.

    Well, do the mass and energy balance. We know the BTU outputs in the form of products. Therefore, we know the BTU inputs. Are you suggesting that there are other BTU outputs that are unaccounted for? The corn coming in ends up as ethanol and co-products. We know the BTUs for those. We also know what the USDA said it took to process the ethanol. What else is missing?

  59. Heiko wrote, “In other words, subbing one gallon of gasoline equivalent at the gas tank with ethanol will free enough upstream petroleum to produce another 0.05*0.8=0.04 BTU of gasoline.

    OK, now I see what you were getting at. Yes, that is correct, based on my analysis.

    The abatement cost can be terribly dependent on energy costs.

    Yes, again we agree. Adding to the confusion is the difficulty of forecasting the costs of abatement strategies like cellulosic ethanol or plug-ins 10, 20, 50 years down the line, given the nascent stages these technologies are in, and the uncertainties about what kind of price curve they will follow as the technologies are refined and economies of scale are reached.

    That’s the main reason I left out any kind of economic analysis from my well-to-wheels study (that and I can only do so much in an undergraduate thesis!).

    __________________________________________

    Hal, you are correct that it takes less fossil energy (on a well-to-wheels basis) to move a car a given distance than it does to move a car the same distance using gasoline.

    The EROI figures obscure the fact that the energy embeded in the fuel that you end up with is still fossil based in the case of gasoline, while it’s predominately biomass-based in the case of E85 (with the gasoline portion being petroleum-based). EROI doesn’t figure in what kind of energy you get in return for your energy investment.

    Robert, would you care to engage this part of the discussion directly? I’m curious what you have to say.

    (too many metrics… too much confusion…)

  60. Robert,

    losses. Some of the energy in the corn + fossil fuel input will end up as useless low grade heat.

    If this wasn’t so, your energy balance for the corn should be:

    ethanol out + co-products out – fossil fuels in = corn in

    Or using your numbers

    75000 + 14000 – 50000 = 44,000.

  61. Robert wrote, “Frankly, that is the simple bottom line as far as the efficiency argument goes. That was the whole point of my challenging the Argonne argument. It is not more efficient to produce ethanol than to produce gasoline, and those who claim that it is are wrong. The other issues must be argued on their own merits.

    OK, I think we’re on the same page here. Feel free to email me with any questions regarding GREET [jesse.d.jenkins[at]gmail.com]. I’ve gotten pretty familiar with it at this point.

    (Also, please note if you are reviewing my thesis at any point that I have modified portions of GREET to add more detail in some areas, and have extensively modified the inputs based on my survey of relevent literature and in order to model the year 2025. If you are interested, I can send you a copy of the modified GREET model I used).

  62. it seems you are having great difficulty in getting Robert to give you a straight yes/no answer here.

    If you have nothing better to do now than troll, perhaps you should move along. I have given him a straight answer twice. If you are merely interested in sound bites, I can’t help you out. These issues don’t all reduce to sound bites.

    But then again, I am still arguing with someone who thinks if I show that the efficiency of an ethanol refinery is 44%, he should argue that it is 50%. At least you have backed off of your earlier argument:

    For 10 BTU of corn input to an ethanol plant, something between 5 and 6 BTU end up in the ethanol, around 3-4 BTU are the dried distillers grain, and 3-4 BTU are losses.

    Baby steps.

  63. losses. Some of the energy in the corn + fossil fuel input will end up as useless low grade heat.

    The process heat will. The corn will not end up as useless low grade heat. All of the corn ends up as ethanol plus co-products.

  64. Robert, would you care to engage this part of the discussion directly? I’m curious what you have to say.

    Actually, I am about to call it a day. I worked 12 hours today, running simulations and doing calculations. I have gotten close to 100 e-mails today (after getting 67 yesterday), many on this debate. Finally, I have written around 25 responses on this topic here and at The Oil Drum. I a bit spent at this point.

  65. Jesse,

    do you accept that the GHG reduction of corn based ethanol could be significantly raised (eg by using biomass gasification instead of natural gas to provide process heat – other options include zeolite membrane processes to save on distillation heat)?

    I accept that scalability is an issue for corn absent major technological breakthroughs, because too much land would likely be required with current yields if corn ethanol was to supply all transportation fuel demands.

    There’s a lot of buzz about cellulosic ethanol, but projections about costs and fossil fuel consumption are highly speculative. For example, it is assumed that lignin will be burnt for process energy. But suppose a low lignin feedstock is chosen (grasses can have lignin below 5%), then there wouldn’t be any lignin to be burnt for process energy and we’d be in the same place as for corn ethanol (ie a broth of sugar and water, that gets fermented to alcohol, which then needs to be distilled / or separated by zeolite membranes).

  66. Robert,

    I don’t think you’ve given him a straight answer. That would be a simple yes.

    ———

    I haven’t backed off my earlier numbers, and I think your calculation is nonsensical, even if it by pure coincidence gives vaguely the right answer (and incidentally twice what you claimed earlier).

    Some of the energy in the corn gets consumed by the micro-organisms doing the fermenting, not all of it will end up in the ethanol and co-products.

    You also used the co-product credit as if it referred to the actual heating value of the co-products, when it the co-product is actually calculated based on displacement.

    You are welcome to have a good night’s rest and answer tomorrow.

    If you don’t wish me to comment any more, I shall not bother you again.

  67. Robert wrote, “Actually, I am about to call it a day.

    No worries, Robert. Take a breather.

    Heiko wrote, “do you accept that the GHG reduction of corn based ethanol could be significantly raised

    Yes, I do agree that that is a possibility. I have not, however, heard of any demo or pilot projects attempting to do so. Still, there is a lot of potential for improvement by replacing the natural gas/coal inputs at ethanol refineries with biomass.

    An integrated ethanol facility that utilized corn and corn stover would make alot of sense. The corn kernels could undergo the typical starch fermentation process, while the corn stover could first either undergo pretreatment to seperate out the lignin and then ferment the cellulose (a la Iogen-style processes), or it could be gasified and then converted into ethanol via microbes (a la BRI-style processes). The cellulosic ethanol process could probably provide process heat for both sides of the process, or at least a large portion of it, from the corn stover (either by burning the waste lignin, or from the gasification process). I havent’ done the math though to see if there is enough energy in the corn stover to provide process heat for the whole integrated process. However, both Iogen and BRI’s cellulosic ethanol processes co-generate a significant quantity of electricty for export to the grid, and the heat could presumably be used for process heat more efficiently (and cheaply) than it could be converted to electricity…

    OK, I’m done for the day as well.

  68. I don’t think you’ve given him a straight answer. That would be a simple yes.

    Not if your objective is to produce net energy (and that should be the objective). Then it is a simple no.

    I haven’t backed off my earlier numbers, and I think your calculation is nonsensical, even if it by pure coincidence gives vaguely the right answer (and incidentally twice what you claimed earlier).

    Well, your earlier numbers gave well more than 50% efficiency. And my off the cuff 25% was just for the ethanol, not the co-products. In that, it is pretty close.

    Some of the energy in the corn gets consumed by the micro-organisms doing the fermenting, not all of it will end up in the ethanol and co-products.

    Very little. The bottom line is that you are arguing that an ethanol refinery is 50% efficient. My calculation put it at 44% efficient. What’s the difference? As far as I am concerned, that’s essentially the same answer, and it wasn’t by coincidence on my part. The vast majority of the BTU content comes from the ethanol anyway, so the energy consumption by the microorganisms, etc. doesn’t affect the bottom line by much.

    If you don’t wish me to comment any more, I shall not bother you again.

    I would just prefer that your comments add something to the discussion. The one I responded to didn’t. Actually, I think this has been a pretty good discussion, despite your occasional jab. But I am prone to jab back when jabbed.

  69. Jesse,

    you’ll see an example in Robert’s long blog post.

    http://news.nationalgeographic.com/news/2006/08/060818-ethanol.html

    They use biogas from cow manure, which is from integrated feed lots (which also saves on drying the distiller’s grain).

    There’s enough energy in the stover, but that isn’t used by any current plants, not even the ones discussed in the link above. Collection of the stover is expensive, and it’s a difficult to handle fuel.

  70. Robert, I’m no expert, but it seems to me that there are two fundamentally different types of energy “production” processes: extractive and conversion. In my view, extractive processes are most naturally characterized by EROEI while conversation processes are most naturally characterized by efficiency. Gasoline production in this model is an extractive process, while ethanol production is a conversion process. To measure each of the processes with the metric for the “other” process is difficult to say the least. I’m an electrical engineer, good with numbers, and I became very confused trying to follow the discussion in this thread. My question/request for you is, could you provide a post in which you precisely define the equations for EROEI and efficiency for both extractive and conversion processes, and then tabulate in a two-by-two table all four calculations, i.e. EROEI and efficiency for both gasoline and ethanol production? Thanks.

  71. “For 10 BTU of corn input to an ethanol plant, something between 5 and 6 BTU end up in the ethanol, around 3-4 BTU are the dried distillers grain, and 3-4 BTU are losses.”

    Going back to my original statement. Another roughly 3 BTU of input is fossil fuel.

    13 BTU in gives roughly 13 BTU out.

    I now see what you were trying to do with your calculation and I’ll accept that you can neglect the energy needs of the micro-organisms for a very rough back of the envelope.

    You can’t use the co-product credits based on displacement value though, not even for a back of the envelope. The actual heating value of the co-products is much higher than their displacement value.

    And I now see how you tried to calculate efficiency:

    (Ethanol out + Co-Products out – Fossil Fuels in) over Corn in

    And efficiency for ethanol alone:

    (Ethanol out – Fossil Fuels in) over Corn in

    I would remark though that if you leave out co-products and allocate all the input process energy to gasoline, gasoline looks pretty poor too:

    (Gasoline out – Process Energy in) over Petroleum in

    That would be the apples for apples and would give something well below 40% for gasoline.

    On a

    100 – Process Energy / Petroleum input

    basis refineries ought be at the 90% efficiency level.

    A similar definition for ethanol would then be

    100 – Process Energy / Corn input

    And efficiency on that basis would be around 70%.

    What I thought you meant with 25% efficiency was that 10 BTU of corn would be turned into 2.5 BTU of ethanol, and your calculation made no sense to me whatsoever for that (as actually a bit over half ends up in the corn, something like 40% end up in the co-products and a little gets consumed by micro-organisms).

  72. (as actually a bit over half ends up in the corn, something like 40% end up in the co-products and a little gets consumed by micro-organisms).

    That should read, a bit over half ends up in the ethanol. Sorry.

  73. Anonymous electrical engineer,

    efficiency can be defined as input/output, and people are pretty free to decide what to count as inputs and outputs.

    If you count the corn and fossil fuels at the ethanol plant gate, and the petroleum and other fossil fuel inputs at the refinery gate and compare with the total heating value of the products leaving, petroleum beats ethanol.

    It also beats ethanol on net energy, nearly no matter how you define net energy.

    Ethanol is better on a fossil fuel inputs only conversion basis.

    As Watthead says there are lots of metrics, you need a few to capture the breadth of the issue.

    All these metrics have the potential to confuse, as has been discussed here at nauseam, because people may assume a different definition than the one intended.

    A good way of describing ethanol I think is to say it’s a solar leveraged conversion of nat gas and coal into petroleum.

  74. To measure each of the processes with the metric for the “other” process is difficult to say the least. I’m an electrical engineer, good with numbers, and I became very confused trying to follow the discussion in this thread.

    The level of confusion over this issue has been surprising to me. The problem seems to me like it should be straightforward.

    I can treat BTUs as an investment. Where can I invest my BTUs so that I end up with the highest BTU return? If I invest 1 into ethanol production, I end up with 1.3 BTUs. If I invest 1 BTU into oil and gas, I end up with 4-5 BTUs of gasoline, diesel, etc.

    Or, let’s say my objective is to make 1 BTU of liquid fuel. Will I burn up more fossil fuel energy in making ethanol, or in making gasoline? Again, I will burn up 3-4 times the fossil fuel energy in making 1 net BTU of ethanol as opposed to 1 net BTU of gasoline.

    That’s it for my contribution today. This consumed far too much of my day yesterday, and I am working 12 hour days this week as it is.

  75. Robert wrote, “I can treat BTUs as an investment. Where can I invest my BTUs so that I end up with the highest BTU return? If I invest 1 into ethanol production, I end up with 1.3 BTUs. If I invest 1 BTU into oil and gas, I end up with 4-5 BTUs of gasoline, diesel, etc.

    This has been your focus throughout this discussion, Robert, hence the focus on EROI. This is all fine and good, but what this kind of thinking obscures is what exactly are you getting for your investment, and what are you investing.

    Not every Btu is created equal, and not every resulting fuel is equal either. Yes, in the case of ethanol and gasoline, if all we look at is Btus, then gasoline is a better investment.

    But let’s look at what we get out of these investments:

    ->in the case of gasoline, you are using your investment to pump hydrocarbons out of the ground and what you end up with is a petroleum-based fuel that when burnt emits plenty of CO2 into the atmosphere that was previously sequestered in the ground;

    ->in case of ethanol, your investment is spent converting corn into a biomass-based fuel that when burnt releases CO2 into the atmosphere that was previously absorbed out of the atmosphere by the growing corn.

    Thus, in the first case, you invest some fossil fuels to get more fossil fuels, and thus plenty of CO2 ends up in the atmosphere. In the latter case, you invest more fossil fuels than the fomer, but what you get is a biomass-based, carbon neutral* fuel.

    Focusing on fossil EROI thus obscures what kind of Btus you get for your investment of fossil fuel inputs (and what happens when you use them). A full well-to-wheels metric includes the emissions and energy use associated with the combustion/use of the fuel during vehicle operation as well, and thus factors in the qualities of the resulting fuel ignored by EROI.

    Furthermore, EROI ignores what use the fuel is put to and how efficiently it is used. For example, we may invest a given amount of fossil energy into making a liquid fuel for an internal combustion engine, or an equal amount of fossil energy making half as much electricity. EROI obsucres the fact that the electricity can be utilized by a plug-in hybrid or EV five times more efficiently than an ICE utilizes the liquid fuel, so using an EROI metric, the liquid fuel looks like a winner (just as gasoline looks like a winner next to ethanol using EROI).

    A well-to-wheels metric takes into account the efficiency at which a given fuel is converted into actual useful work done (i.e., vehicle miles travelled), which is what we should honestly care about. Unless we know that two fuels can be converted to work at the same efficiency (as is probably more or less the case for gasoline and ethanol), EROI is pretty useless.

    I know a well-to-wheels figure is more difficult to obtain than EROI, but I do feel that it provides a more accurate and insightful comparison than EROI, for the reasons I have discussed in this and previous comments on this thread.

    Thanks again Robert for facilitating this discussion. Go get some real work done (I know it’s time for me to do some real work…).

    (*carbon neutral referring to the fuel itself here, not all of the inputs which make the well-to-wheels net carbon output for ethanol positive. This simply means that all of the carbon contained in the ethanol itself is biomass-based and thus was previously absorbed from the atmosphere.)

  76. It is shocking that khosla et al are wasting time with you. OF COURSE gasoline requires less energy to produce than corn ethanol. But that is a static view of the world. They are looking to the future and what may happen. I only stumbled upon this website because of khosla. you should be honored that they wasted time with you.

  77. I can treat BTUs as an investment. Where can I invest my BTUs so that I end up with the highest BTU return? If I invest 1 into ethanol production, I end up with 1.3 BTUs. If I invest 1 BTU into oil and gas, I end up with 4-5 BTUs of gasoline, diesel, etc.

    The only way that works with oil is if you start with your 1 BTU, and instead of taking it to the gasoline refinery, you go out to the oil field, use it to produce even more BTUs of oil, and only THEN take it to the refinery. This extra step of going to the oil well amplifies the BTUs you have available to take to the refinery.

    You could have done a similar step with ethanol. Most of the BTUs of fossil energy that go into ethanol are from natural gas, I understand. You could start with 1 BTU of NG, and instead of taking it to the ethanol factory, you go back to the NG field, use your 1 BTU to produce even more NG, then take your many BTUs of NG to the ethanol factory, and produce many BTUs of ethanol. This gives you a similar amplification of your starting BTUs before you start the conversion to ethanol, as you did in the case of conversion to gasoline. Presto, now you can get 10 BTUs of ethanol for 1 BTU of fossil energy input.

    Does that seem like a reasonable way to do the calculation? I don’t think so, but it’s exactly what you’re doing in the case of gasoline.

  78. Robert,

    people do get confused into thinking that net energy = net petroleum. They are also confused into thinking other factors don’t matter.

    But if net energy was the only thing that mattered, we wouldn’t be producing oil, we’d just be producing coal (which on the Argonne figures used by Jesse has a net energy of 50:1, we could also go for wind energy which is well above 30:1).

    And if conversion efficiency of a particular step in isolation was all that mattered, we’d never turn crude oil or coal into electricity, we’d always turn it into space heat.

    As Hal said in practise, we can’t just take a BTU of oil in the US and invest it in Saudi Arabia and get back 5 BTU. The Saudis aren’t interested in our oil, they want money and they are certainly not going to give the US 5 BTU of oil next year in exchange for 1 BTU of oil this year.

    When you say “I can treat BTU’s as an investment”, you neglect that BTU’s alone are not enough, we also need access (eg to petroleum deposits), labour and loads of other inputs, and you also neglect that not all BTU’s are created equal. BTU’s of coal for example currently sell for a fraction of what gasoline BTU’s sell for. And thirdly, you don’t consider the possibility that several re-investment cycles may be possible for one source (say biomass, which can be grown using last year’s biodiesel) in the time in which another (say a nuclear power plant that is constructed over four years) produces no energy yet in spite of a stellar net energy ratio.

    Net energy is not policy relevant in isolation. What does matter is:

    How much imported gasoline can be displaced by ethanol (slightly over 1 BTU’s worth per BTU of ethanol),

    how much fossil fuel has to be taken out of the ground to have 1 BTU in the gas tank (nearly half the amount required for gasoline),

    how much CO2 gets added to the atmosphere by using a BTU of ethanol rather than a BTU of gasoline (about 30% less).

    how costly is ethanol production compared to imported gasoline (substantially cheaper, it does not matter that Saudi Arabia can produce gasoline for much less than the US can produce ethanol for)

  79. But let’s look at what we get out of these investments:

    But as soon as you did this, you veered off the point of this essay, which was merely to challenge the claims that ethanol production is more efficient than gasoline production. It looks like we both agree on this basic point. Whether ethanol has benefits over gasoline in other areas is a separate matter.

  80. OF COURSE gasoline requires less energy to produce than corn ethanol.

    Dear Anonymous,

    Thank you for missing the entire point of the essay. Khosla et al. claimed that gasoline requires more energy to produce than ethanol. I am glad to see that you agree with me. Any other points about ethanol’s relative merits are completely different subjects than the topic of the essay.

    They are looking to the future and what may happen.

    Again, thanks for completely missing the point, and for throwing out red herrings. Looking to the future is fine. Making false or misleading claims while describing this future is not.

    you should be honored that they wasted time with you.

    Honored? Sorry, I don’t suffer from such an unfulfilled life that I feel honored by discourse with public figures.

  81. Does that seem like a reasonable way to do the calculation? I don’t think so, but it’s exactly what you’re doing in the case of gasoline.

    What you are doing is a cradle to grave analysis on each case. If we consider the BTUs that must be consumed to bring oil from the ground to gasoline in your tank, versus planting and harvesting corn, turning it into ethanol, and then bringing it to your tank, the gasoline efficiency is about 5 times the ethanol efficiency.

    It is clear to me that the real hang-up people have with this is that they envision a literal barrel of oil, and presume that you can either turn it into ethanol at a 1.3 return, or into gasoline at a 0.8 return. But that is a complete straw man. What you really do is invest BTUs into the process, and for gasoline the small BTU investment of getting the oil out of the ground is being ignored. You have to define your processes on an apples to apples basis. The only time ethanol comes out on top is when a full life-cycle for ethanol is compared to just the refining process for oil to gasoline.

  82. They are also confused into thinking other factors don’t matter.

    To be certain, this debate has wandered into many areas. There has been a lot of confusion, and I am tempted to try to write one short post summarizing both sides. However, I want to make it clear that I recognize that other things do matter. But the topic I was addressing were the claims the it is more efficient to produce ethanol than gasoline.

  83. I would argue that to a point, net energy itself is an apples to oranges comparison, namely between inputs and outputs that cannot be compared merely on their BTU content.

    Efficiency is really a pretty vague term, it doesn’t mean much more than the quotient outputs/inputs with people free to define what they consider the relevant inputs and outputs.

    When Vinod talks about efficiency, he presumably means liquid fuel outputs / fossil fuel inputs, not process and extraction energy inputs / liquid fuel outputs .

    See when we talked about the efficiency of an ethanol plant, I thought you had a completely different definition in mind than the one you actually meant, which completely threw me and made me think you were doing a completely nonsensical calculation.

    I agree that the kinds of definition you have for conversion efficiency and net energy will make ethanol less good than gasoline, I disagree that those are the metrics people like Vinod are referring to, and that they cannot use their preferred metrics because people might be confused into thinking that the kind of efficiency you’ve got in mind is meant.

  84. Here is a follow-up letter that I sent today:

    —————————–

    Dear Tom, Dr. Wang, and Mr. Khosla:

    First of all, let me apologize for the offense you took at my usage of “sleight of hand.” Never in my life have I considered that phrase insulting, but clearly you were insulted by it. I have used that term on many occasions, and had that term used against me. For me, it just means that things are not as they appear to be. So please do not presume that I was being intentionally insulting, because I was not.

    Second, I have been stunned at the response from publishing our exchange. Between my R-Squared blog and The Oil Drum, the exchange received well over 400 responses to date, and I got around 200 e-mails. And while you may consider me combative and stubborn, I am also open-minded and very analytical. I engage in this discourse as much to learn as to convey information, and I was able to understand through those responses just why people are so confused about this issue of gasoline efficiency versus ethanol efficiency.

    The reason I am engaged in this debate is that it is very important to me that we pursue the correct energy policy. While I have argued in favor of certain solutions, I have also spent a lot of time debunking certain claims. I don’t believe we do ourselves any favors, nor do we help ourselves make educated decisions by allowing myths to persist.

    I agree with Mr. Khosla that maybe there are other questions that are better asked. We can debate many different angles over whether or not we should be advocating ethanol from corn. But this particular point of contention is about whether the claim “the efficiency of producing ethanol is better than the efficiency of producing gasoline” is accurate. I have lost count of how many times I have heard some variation of this claim. Tom, in your initial response to me, you included an attachment which made the claim:

    “As you can see, the fossil energy input per unit of ethanol is lower–0.74 million Btu fossil energy consumed for each 1 million Btu of ethanol delivered, compared to 1.23 million Btu of fossil energy consumed for each million Btu of gasoline delivered.”

    That is simply a false claim. Dr. Wang will probably acknowledge that this claim as written is incorrect, and yet it is derived from his work. That is why I say people are being misled as a result of his work. Perhaps it is unintentional, but when people make a claim such as the one above, they have misinterpreted what is being said, and used this misinterpretation to promote the ethanol agenda.

    I am writing a short follow-up essay after seeing the issues that people found most confusing. I will send the link when I am finished. The real critical point when comparing the two processes is to make sure the boundaries are drawn in exactly the same place and definitions are consistent. I think it will become clear why the above claim as written is incorrect. But please don’t misinterpret this into thinking that I am trying to completely rebut all ethanol arguments. I am addressing a single issue.

    Incidentally, I have also recently written some essays on Prop 87. Because I am a big believer in open debate, I have asked Ana Unruh Cohen, the Director for Environmental Policy at the Center for American Progress, if she would like to write a rebuttal/pro-Prop 87 piece. She has agreed to do so, so I will be placing that on The Oil Drum in a few days. Based on the traffic I have received from my previous Prop 87 essays, it will undoubtedly be read by a lot of interested voters.

    Again, please accept my sincere apologies for offending you. That was not my intent.

    Sincerely,

    Robert Rapier

  85. I may be weighing in too late here, but I’m going to take a stab at simplifying the discussion, and I’d like whatever feedback folks are inclined to give.

    I’m going to leave aside the questions of who is trying to fool whom, definitions of efficiency and EROEI, and the difference in performance between alcohol and gasoline.

    It seems like the simplest form of the question that we really ought to be asking is: For one BTU of extracted fossil fuel, how much usable liquid fuel do we get?

    Using numbers already cited (and apparently accepted):
    gasoline: extracting 1 BTU of oil makes available 0.8 BTU of gasoline.
    ethanol: extracting 1 BTU of misc. hydrocarbons makes available 1.3 BTUs of ethanol.

    So, by this metric, ethanol is a winner, providing 1.63x as much BTUs as gasoline for the same starting amount of fossil fuel. If one accepts the 1.6:1 ratio figure, then one sees a 2x multiplier on starting fossil fuel energy.

    Note that I am avoiding using the word “input” here. The point is not “what’s an input?” vs. “what is passed through?”, etc. The point is, “what do you start with, and what do you finish with?”

    Looked at this way, ethanol looks pretty good. But I am not a fan of ethanol. Why? Because these calculations ignore the additional costs (infrastructure, soil damage, greenhouse gas, etc) required by the ethanol production process. Because there are much easier and cheaper ways to achieve factor 2 improvements. And because this way of looking at things still rests on the assumption that oil is reasonably cheap and available. In a post-peak scenario, oil and natural gas are absurdly expensive, Half of absurd is still absurd.

    Bottom line is, corn ethanol is better in some (limited) ways than current practice, but it’s not good enough. If we were using sugarcane (EROEI ~8:1), or if we had exhausted our options for low-cost efficiency, then it would be a different story. But we’re not.

    There are better, cheaper ways to get the same benefit. That’s the story that the anti-ethanol crowd should be promoting. Don’t split hairs over the definition of efficiency, or whether the EROEI is 1.3:1 or 1.6:1. Instead, focus on the fact that it isn’t anywhere close to being a real solution, for reasons that are obvious but are readily obscured by the “ethanol is marginally positive” vs. “ethanol is marginally negative” debate.

  86. Greenengineer,

    one other conclusion to draw, and I think Robert is basically drawing it, when you look at his latest post, is that maybe it’s time to look at ways to improve ethanol, which make ethanol more than a solar leveraged conversion of nat gas and coal into liquid transportation fuels, but mostly, and eventually entirely, a renewable energy source.

    Also, maybe there are lower cost ways still to be had than ethanol, but there’s also the point that there are lag times and the need to prove a technology and show that it can scale, and be it only to the point where it can meet half of an 80% reduced demand in a universe dominated by plug-in hybrids.

    In the end, that’s why we have invested so heavily in wind and PV, even though they are not, or at least were not (in the case of wind) the most cost effective at the time.

    (I have a somewhat different perspective on peak oil than you, as far as I can tell, but that aside, I agree with pretty much everything you say in your comment)

  87. “As you can see, the fossil energy input per unit of ethanol is lower–0.74 million Btu fossil energy consumed for each 1 million Btu of ethanol delivered, compared to 1.23 million Btu of fossil energy consumed for each million Btu of gasoline delivered.”

    Sounds reasonable to me, assuming that by “consumed” we mean “consumed out of earth’s store of fossil fuels”, i.e. “taken out of the ground”.

    Robert uses a weird definition whereby fossil fuels are not ‘consumed’ when being turned to gasoline but are ‘consumed’ when being turned to oil. That is a superficial distinction which does not go to the root of the question, which is, given our diminishing store of fossil fuels, what is the most effective (“wells to wheels” as they say) way to turn it into transportation miles?

    (And BTW, I agree we are over-investing in ethanol and that it cannot be the salvation many people claim to believe it is, but that doesn’t change the fact that this particular argument of Robert’s looks to me to be off base.)

  88. Robert uses a weird definition whereby fossil fuels are not ‘consumed’ when being turned to gasoline but are ‘consumed’ when being turned to oil.

    Sorry, I meant:

    Robert uses a weird definition whereby fossil fuels are not ‘consumed’ when being turned to gasoline but are ‘consumed’ when being turned to ethanol.

  89. heiko,

    I do agree with your statements about ethanol generally; i.e. it could be a viable and sustainable replacement for gasoline after we reduce our demand by factor 5 through greater efficiency measures.

    However, I’m very skeptical that corn ethanol in particular can every be sustainable. Corn is the darling of the industrial agribusiness world because it can produce enormous amounts of sugar and starch per acre. But that high level of productivity is reflected in a proportionally high level of nutrient demand. Industrial corn cultivation is soil mining.

    I am willing to believe that, through careful attention to closing nutrient loops (as E3 is attempting to do), the negative ecological impacts of corn ethanol could be minimized. But I suspect that that energy could be better invested by starting with an agricultural crop that isn’t so heavily reliant on external inputs and externalized costs.

  90. Robert uses a weird definition whereby fossil fuels are not ‘consumed’ when being turned to gasoline but are ‘consumed’ when being turned to ethanol.

    No. I am using a conventional definition of consumed, where consumed means actually used up. When you “consume” oil, you didn’t actually use it up. You ended up with gasoline. What you consumed was only the 20% fraction that was needed to turn the oil into gasoline. That is consistent with the 100% you consumed when you turned it into ethanol. And in fact, Michael Wang wrote back to me last night to agree that the statement as written is incorrect. So, it didn’t look reasonable to him, even though you say it looks reasonable to you. I will be posting our most recent correspondence in a few days.

    The problem here is that the process boundaries are being drawn differently for the two processes. Proponents think that the options for our BTU of oil are either to turn it into gasoline, or to turn it into ethanol. The difference is that when we turn it into oil, we are looking at only 1 step in the life-cycle. On the other hand, we looked at the entire life-cycle for ethanol. So, the other option, and the apples to apples comparison that I am making, is to look at the entire life cycle for oil to gasoline. If you look at a part of one process and compare it to a different part of another process, you are going to get nonsensical answers.

  91. “As you can see, the fossil energy input per unit of ethanol is lower–0.74 million Btu fossil energy consumed for each 1 million Btu of ethanol delivered, compared to 1.23 million Btu of fossil energy consumed for each million Btu of gasoline delivered.”

    But Robert, you would agree that this statement is true if by “consumed” we mean, “taken from earth’s store of fossil energy”?

    Your definition of “consumed” is such that we do not “consume” food when we eat (since it is mostly transformed and incorporated into our body.)

  92. But Robert, you would agree that this statement is true if by “consumed” we mean, “taken from earth’s store of fossil energy”?

    Yes. Unfortunately, that is not what the word consumed means in the English language. If you told someone you consumed a barrel of oil, they presume that you actually used it up. They don’t envision that you now have a barrel of gasoline, diesel, etc.

    The problem is that the word in being used inconsistently for the two cases. I no more consume oil taken from the ground than I consume water when I fill up my glass. The consumption takes place when I drink it. In the case of oil, you refined it. The energy is still there ready to be tapped, with a minimal consumption (BTUs that were actually burned). In the case of ethanol, all of the BTUs were burned, such that burning 1 BTU netted out 0.3 BTUs.

    Your definition of “consumed” is such that we do not “consume” food when we eat (since it is mostly transformed and incorporated into our body.)

    Consumed food is no longer available to be eaten. Consumed oil is no longer available to be burned.

    You have to understand that the proponents’ choice for the barrel of oil – refine it, or invest it into the life-cycle for ethanol – is a false dichotomy. There is a 3rd choice, and that is to invest it into the life-cycle for gasoline production. That is the relevant efficiency comparison to investing it into the life-cycle for ethanol production.

  93. R^2,

    With all due respect, I think your definition of “consumed” is unnecessarily constrained. In the production of gasoline, crude oil acts as both energy source and chemical feedstock. I think it’s perfectly reasonable to treat 100% of the crude oil as being “consumed”, in much the same sense that pelletized plastic is consumed in an injection molding process: the material is not actually destroyed; only its form has been changed.

    I don’t mean to keep harping on this, but I really don’t understand your perspective, and I would like to. I do agree that it’s important to compare the two processes on an equivalent basis, but I don’t see why that requires treating the oil-for-feedstock differently from the oil-for-energy.

  94. You have to understand that the proponents’ choice for the barrel of oil – refine it, or invest it into the life-cycle for ethanol – is a false dichotomy. There is a 3rd choice, and that is to invest it into the life-cycle for gasoline production.

    But if you invest the oil in the life-cycle of gasoline, in order for your investment to be productive, you must also invest 5x as much oil in the process, as feedstock. Otherwise, where’s your gas?

    In the case of ethanol, you’re investing hydrocarbons, in part, to produce your feedstock. The feedstock itself (corn) never appears in this calculation, since it is presumed to be accounted for by the hydrocarbon investments that produced it, and is also presumed to be greenhouse gas neutral (with respect to itself, not with respect to the invested hydrocarbons).

    Honestly, the only basis on which your position makes sense to me is if you presume an unlimited supply of oil-as-feedstock. And I know that’s not where you’re coming from. So I don’t understand your refusal to consider the feedstock oil as an input to the process.

  95. OK, one more time. This is an argument about the efficiency of 1 process versus the efficiency of another. In order to compare the two processes, no matter how we define our terms, we must define them consistently.

    The barrel of oil is really just symbolic of the BTUs you have to invest into the process. If you have BTUs to burn – in other words they are going to be used up in the process and no longer available as energy – then 1 BTU burned in the production of gasoline will produce about 5 BTUs of gasoline. That’s because 0.5 BTUs can extract 5 BTUs of oil, and another 0.5 BTUs can turn it into 5 BTUs of gasoline, diesel, etc. If we invest our BTU into ethanol, we will end up with 1.3 BTUs of ethanol.

    I can’t really make it any simpler than that. This is an argument about efficiency. It is not an argument over whether oil is finite resource, etc. It is an argument that should help explain why ethanol can’t compete with oil in most cases: It is just not nearly as efficient to produce ethanol as it is to produce gasoline.

    And as you will see just as soon as I post the recent e-mail from Michael Wang, he agrees with me on this point.

  96. OK, I now see that what we are having here is a nearly pure semantic arguement over the meaning of “efficiency”.

    Normally I stop arguing when I realize that I’m in a disagreement about definitions. I’m not going to this time, because I think that there’s an important point to be made.

    To be clear:
    You are using the term in the strictest thermodynamic sense:
    eff = (useful energy return)/(energy burned to heat)

    I am using “efficiency” in a total energy sense:
    eff = (useful energy return)/(total energy input)
    In this case, “total energy input” very specifically does include chemical potential energy.

    It is my understanding that we are having these conversations because we are all interested in figuring out what a sane energy policy would look like, or if one is even possible given the energy demands of our technological civilization.

    In that context, I would assert that the definition of “efficiency” that I (and apparently Wang, etc) am using is the more relevant one. I assume the reasons why are obvious, given the presumed context of the conversation.

    The reason that I’m still arguing about this is because I respect your work, and I believe that we hold fundementally similar positions on the major issues around energy, and on the viability of corn ethanol in particular. I think it’s important to define terms in a way that is meaningful in context of the discussion. More to the point, I think that it actively hurts our position to get drawn into fights with our opponents over questions that are, in the end, purely academic.

    There are plenty of reasons to oppose corn ethanol as a basis for energy policy. We can argue against the energy return figures that they provide. We can point out the unaccounted externalities. We can take exception to the highly political nature of the decision to focus on corn as a source of energy.

    If we want to advance our point of view effectively, we need to focus on the real points of disagreement. Getting tied up in definitions wastes time. Worse, it puts us in the position of accepting the other side’s assumptions and conclusions (like the 1.6:1 EROEI figure).

    Ultimately, this is a matter of framing the debate properly. And on that subject, I will defer to the master, George Lakoff. If you haven’t read Don’t Think of an Elephant, you should. It’s very short, and utterly crucial. An excerpt can be found .

  97. Sorry about the deleted comments. I was trying to use an href link, but I finally gave up and decided that Blogger doesn’t handle the label properly in the comments (even though they displayed correctly in the preview; go figure).

    Anyway, the period at the end of my last post is the link I was trying to create.

  98. Don’t think this is just a semantic quibble over the definition of efficiency. It is not. It is a quibble over evaluating 2 processes via the same metric. I can’t compare the life-cycle for ethanol to just the refining step for gasoline, and then declare that the former has a higher efficiency on the basis of that. That would be like me claiming that a mouse is bigger than a cat from comparing the cat’s whisker to the mouse’s tail. Apples must be compared to apples.

    Here is your reformatted link:

    Don’t Think of an Elephant

  99. I have a very hard time believing an engineer or anyone else is being honest when he says he doens’t understand how Khosla and Wang’s math is wrong. Anyone in doubt should just look at Khosla’s own words:

    It is time to stop asking the wrong question of “energy balance” or even the somewhat less wrong question of “energy balance relative to petroleum”

    If you still don’t get it then I think the odds of you being one of those paid pundits from DCI to spread misinformation on blogs is extremely high. I honestly don’t think anyone can be that dumb and still figure out how to get to Robert Rapiers weblog on the internet.

  100. “It is a quibble over evaluating 2 processes via the same metric.”

    The metric is: fossil fuel out of the ground per available BTU in the gas tank. That is apples to apples.

    I don’t find Wacki’s contribution particularly to the point.

  101. The metric is: fossil fuel out of the ground per available BTU in the gas tank. That is apples to apples.

    It isn’t, because you are comparing primarily petroleum to primarily natural gas. And the statement in dispute was not “out of the ground”, it was fossil fuels consumed. The only fossil fuels that were consumed were those expended in getting the fossil fuels out of the ground.

    Yours would be an apples to apples comparison above if we could literally turn a BTU of oil into 1.3 BTUs of ethanol. But then it is no longer an efficiency metric.

    Since we don’t turn oil into ethanol, the best possible metric is BTU inputs into the life-cycle of a process versus BTUs returned. That is a metric that gives a true evaluation of the efficiency of a process. That metric explains why ethanol can’t compete with gasoline unless it receives massive subsidies.

  102. I don’t find Wacki’s contribution particularly to the point.

    I’m saying that Khosla is telling you to ignore “energy balance” or EROEI which is the most important question of all when it comes to costs and sustainability. He’s literally asking you to ignore what matters most. What he calls “the right questions” are:

    1) replacing oil

    ok, helps avoid war but unless we can replace the entire production of the middle east this won’t make a difference. So this point is meaningless.

    2) reduce greenhouse gases

    well once natural gas runs out we will have to rely more on coal. Then we have to tell Khosla the real question is “is there lower hanging and more cost effective fruit than subsidizing ethanol?” It seems pretty obvious to me that there is.

    So I honestly don’t think Khosla has a leg to stand on.

  103. Yours would be an apples to apples comparison above if we could literally turn a BTU of oil into 1.3 BTUs of ethanol.

    As far as I understand, we could do this. We wouldn’t, for economic reasons (oil is too valuable). But, on an energy-balance basis, it’s a valid simplification.

    But then it is no longer an efficiency metric.

    Just because we switched BTU sources? How’s that?

    the best possible metric is BTU inputs into the life-cycle of a process versus BTUs returned.

    Yes, I agree. But if we’re going to talk about energy balance issues, then we need to include all non-renewable BTU inputs. Gasoline production requires oil as a feedstock, which should be counted as a non-renewable BTU source. Ethanol requires corn as an input, the BTUs for which are ignored because they are derived from solar energy. The cost of gathering and processing those solar BTUs is paid in non-renewable energy, but that energy is counted.

    That metric explains why ethanol can’t compete with gasoline unless it receives massive subsidies.

    I disagree. The energy metric really only gets at the question of total energy balance. It does not address economics, except very indirectly. If we want to talk about economics, then we need to account for many factors that are ignored by the (relatively simple) energy balance discussion.

  104. If we do switch to coal won’t that make grain ethanol far more damaging to the environment?

    Vastly more damaging, both in terms of greenhouse gasses and “conventional” air pollution (sulphur, mercury). I don’t know the numbers offhand, but if you look around you can probably find figures for the carbon release per BTU of coal. Compare that to oil, and to natural gas, and you’ll get a sense of the relative impacts.

  105. WattHead said, “Energy contained in the fertilizers, pesticides, herbicides, etc. are included, as are fuels for farming machinery.”

    Did you include the energy needed to make the seed corn? Growing the hybrid seed from which to grow corn is even more energy-intensive than growing corn to make ethanol? (And one of the unfortunate side effects of hybrid corn is that farmers can’t save a portion of their crop to use for growing next year’s crop — they must buy expensive (and energy intensive) seed corn each spring.)

    WattHead said, “Embodied energy in infrastructure and machinery is excluded.”

    Why would you exclude that energy? Making corn ethanol is just as dependent on the energy needed to build (and embodied in) a tractor or ethanol plant, as the energy used to run the tractor or to mill and distill corn into ethanol?

    The energy used to build an ethanol plant or tractor should definitely be considered as part of the energy investment in making corn ethanol, and part of the EROEI equation.

  106. Wacki said, “I’m saying that Khosla is telling you to ignore “energy balance” or EROEI which is the most important question of all when it comes to costs and sustainability. He’s literally asking you to ignore what matters most.”

    Absolutely correct Wacki.

    If Vinod wanted to do something worthwhile with his money with regards to ethanol, he should fund a legitimate, neutrally-audited, practical demonstration to settle once and for all the question of the EROEI of corn ethanol.

    Something on the order of this:

    Start with 100,000 gallons of ethanol.

    Then using only the energy in that ethanol, see how much corn can be grown and turned into more ethanol.

    If after the demonstration he ends up with more than 100,000 gallons of ethanol, the question is settled. We all go home, buy ethanol stock, and thumb our noses at the oil shieks.

    The only kicker is that his demonstration would have to use only the original ethanol for every energy input into growing corn, and milling and distilling it into ethanol.

    Some examples:

    * Ethanol would have to power the lights and computers at the ethanol plant.

    * If the corn farmer needs to run into town to have a broken bracket on his corn picker welded, that trip to town has to be powered by the orginal ethanol as well as the power for the arc welder.

    * Some energy from the original 100,000 gallons has to be invested in making the seed corn, fertilizer, pesticides, fungicides, and herbicides the farmer uses.

    * Some of the energy in the original 100,000 has to be used to power all the farmers ag equipment, as well as transporting the corn to the ethanol plant, and transporting the finished ethanol to a filling station.

    * Yadda, yadda, yadda. (I’m sure you all get the drift.)

    Start with 100,000 gallons. Use only the energy in that ethanol with which to grow corn and make ethanol.

    At the end of the cycle, measure how much new ethanol has been made.

    Vinod could do the entire country a favor by funding such an experiment. The problem: The ethanol industry woudl never let him try. There would be too much at stake. If the experiment failed, their industry would evaporate overnight.

    It is easier for them to wave around the Wang/Shapouri paper studies rather than to put it all on the line with an actual practical test.

  107. As far as I understand, we could do this.

    You couldn’t do it without having your EROI plummet. Turning oil into fertilizer and fueling boilers with it would be far less efficient than doing so with natural gas. So, it would take substantially more BTUs of oil than BTUs of natural gas.

    Just because we switched BTU sources? How’s that?

    No, not because you switched BTU sources. You are once again comparing part of a process to a whole process. It’s the tail of the mouse versus the whisker of the cat. What you are saying is that for ethanol, we are going to count the captured solar energy from growing the corn. For oil, we are going to ignore the millions of years of captured solar energy. We are going to ignore that nature has already done the heavy lifting for us, that we are trying to replicate on an annual basis with ethanol. What you have is a metric, but it isn’t an efficiency metric. I could identify all kinds of metrics that would not be efficiency metrics when comparing the two processes. I could compare food input per available BTU in the gas tank. That would favor gasoline by a huge margin. But if I then say, on this basis, that gasoline is hugely more efficient than ethanol, I would be misleading people – especially when I don’t explain what I have actually compared.

    The cost of gathering and processing those solar BTUs is paid in non-renewable energy, but that energy is counted.

    See above. In the case of ethanol, you are saying “let’s count solar BTUs.” In the case of gasoline, you are saying “let’s not.” The renewable/nonrenewable aspect is an important argument, but not related to the efficiency argument.

    I disagree. The energy metric really only gets at the question of total energy balance.

    No. Energy inputs for any process have a huge impact on economics. If I could produce ethanol at a 10/1 EROI, I have now essentially divorced ethanol from the rise and fall of natural gas and oil prices. But when the EROI is 1.3/1, ethanol will always be more expensive to produce than those fossil fuel sources (with the exception that we are using coal) and that’s exactly why ethanol must rely on subsidies. It is a direct result of the poor EROI.

    Think about what we have done when we make ethanol. We have taken 1 BTU of fossil fuel, and turned it into 1 BTU of ethanol plus 0.3 BTUs of animal feed co-products. But we don’t subsidize ethanol on the basis of created energy; we subsidize on the basis of the entire gallon. Here is an interesting exercise. I have done this calculation many times, but I recommend you do it for yourself. We subsidize ethanol at $0.51/gallon, plus state subsidies that are typically in the $0.20/gallon range. Calculate how much we are paying per MMBTU of energy created, and per gallon of gasoline displaced just from these 2 subsidies. I think you will be shocked, especially when this doesn’t consider other negative externalities.

  108. Leaving aside the debate on definitions for the moment, two points.

    Firstly, with a net energy ratio of 1.000000000000000000000000000000000001

    you’ll find that it’ll take more than the whole federal budget to produce 1 gallon of gasoline equivalent of net energy.

    When net energy falls to 1 exactly, you’ve got a division by zero and the subsidy becomes undefined or infinite.

    I happen to think that the subsidy is there for a particular liquid fuel, mostly used as an octane enhancing blending component, and that the subsidy per unit of net energy created is nonsensical.

    On my preferred measures the subsidy looks pretty modest.

    Secondly, a net energy of 1.3 of course doesn’t say anything about economic viability. Some wind projects with a net energy of 30 can be economically entirely unviable, and a process that has coal at $2 per MMBTU as its input and gasoline at $14 per MMBTU as its output can easily have a ratio below 1 and be very profitable indeed.

  109. I’ve read many of these insightful posts and both sides have great arguments. However, I must take a different approach and question the debate itself. Many years ago I did a paper on the use of alternative fuels (1983). I set out to prove conspiracy and manipulation had caused our reliance on fossil fuels. I found out through honest research that I was naïve. I discovered that it’s no accident or conspiracy we’re addicted to fossil fuel. In terms of energy yield and availability it’s a clear choice by a wide wide margin. So in my opinion it is pointless to debate whether any alternative fuel stacks up against fossil fuel as a viable alternative.

    If you look the fact that carbon dioxide levels are higher than they have ever been in the history of mankind, and Co2 levels correspond directly to global temperatures, the argument changes to something futile. Futile, like arguing what’s the best thing to wear if you’re jumping into a volcano. Alternative refers to choice and I don’t think we have one.

    Back to basics, we need to manufacture a portable energy rich substance that will fuel vehicles so they can operate remotely, an energy that doesn’t take carbon dioxide that has been trapped beneath the earth for millions of years and systematically reenter it into our balanced life sustaining ecosystem. This process may very well be at a net loss. This loss can be off set if we use renewable energy in the production. Since ethanol contains stored energy and it only produces Co2 it absorbed from the atmosphere it starts to look a little more viable. Unless we want to break out the horses, we have to accept the efficiency issues, whatever they are. I suspect energy from wind, photovoltaic, geothermal, hydro and farming methane will have to be used to produce this portable fuel. So if your corn field has wind turbines which power your bio fuel plants you have found a way to convert non-portable energy to portable (Co2 neutral) energy.

    Now the good news about biofuels, is due to the lean forward optimism and financial support of some, we are starting to develop an infrastructure/foundation for their production and use. And although we are currently using fossil fuels to facilitate their production, I have no doubt that we will continue to improve these biofuels and convert them to Co2 neutral by the use of other renewable energy sources in production.

    The truth is if we are going to survive the future and live life as we know it there isn’t just one solution. The future will probably use efficiency and numerous renewable energy sources, biofuels, hydrogen, wind, hydro, solar, etc. combined with Co2 capture and storage techniques.

    One other thing that is over looked in the energy calculations is the considerable energy it takes to capture Co2 released from fossil fuels and store it permanently. That changes the formula greatly.

    cheers

  110. Came to this blog via Rolling Stone, whose most recent issue promised some sort of exchange of opinions via their letters forum. Damned if I can find that. My letter to them ought to see the light of day, because I raised the key fact about energy conversion via eethanol, which follows:

    According to Vaclav Smil, in his 1983 book Biomass Energies, one liter of ethanol produces roughly 21 megajoules of heat value. To produce that one liter of ethanol from the mash liquor in the fermentation tank via fractional distillation requires 19 megajoules of energy, which has to come from coal or electricity or somewhere. (Brazil’s ethanol production is semi-successful in this regards because they burn the sugarcane bagasse to distill the ethanol. The heat value of bagasse is something like 7 MJ/KG [coal’s is something like 29MJ/KG, depending on type, which shows that bagasse’s heat value is better than dirt but not by much] and it burns dirty as hell.)

    Ran this by Rolling Stone, and they didn’t bother with it. Too many confusing numbers, I guess. Vaclav Smil–sounds like a furriner, too. But dammit, this basic set of numbers ought to make plain to anyone who can count the insanity of ethanol as a public policy.

    Best Regards–Daniel White

Comments are closed.