Energy Returned on Energy Invested Thread pt 1 (merged) Arch

[go5star]

Anyone? Anyone? Beuller? No one has any comment?

[shakespear1]

While trying to locate the well spacing in Duri I stumbled on the follwoing thread which has some good discussion.

Indonesia and Oman
with Oil & NG Market and Bolivia Updates

Interesting Thread

Enjoy

[EnergySpin]

Goodmorning/noon/evening/night folks
Time to explore the "scientific" away of lying about EROEIs and how to get (or not) away with it. The culprit (as always) is D Pimentel and I will be discussing his newest study and the flaws that it had along with responses from both scientific and non scientific perspectives.
D Pimentel and Patzek published yet another study that tried to examine the energy balance of ethanol from various plant sources. This study published in Natural Resources Research, Vol. 14, No. 1, March 2005 and available for download here
examined the energetics of ethanol from corn, switchgrass and wood. The abstract of the study reads as :

Energy outputs from ethanol produced using corn, switchgrass, and wood biomass were each
less than the respective fossil energy inputs. The same was true for producing biodiesel using
soybeans and sunflower, however, the energy cost for producing soybean biodiesel was
only slightly negative compared with ethanol production. Findings in terms of energy outputs
compared with the energy inputs were: • Ethanol production using corn grain required 29%
more fossil energy than the ethanol fuel produced. • Ethanol production using switchgrass
required 50% more fossil energy than the ethanol fuel produced. • Ethanol production using
wood biomass required 57% more fossil energy than the ethanol fuel produced. • Biodiesel
production using soybean required 27% more fossil energy than the biodiesel fuel produced
(Note, the energy yield from soy oil per hectare is far lower than the ethanol yield from corn).
• Biodiesel production using sunflower required 118% more fossil energy than the biodiesel
fuel produced.

It is instructive to follow the switchgrass case (corn ETOH and biodiesel have been addressed by both the USDA and the National Biodiesel Board) really closely to see how D Pimentel reasons on data.
At page 6 of their publication Professors Pimental and Patzek reach the following conclusions:
1) The EROEI of growing switchgrass is close to 11 . In order to arrive at this number they used
a) Canadian and NOT USA or EU production data (which on average are higher by 30-50% for the former as well as the latter when swichgrass is grown in Southern Europe)
b) included the energy needed to feed the human farmers (as if humans were specifically born to do this job and would not have been born if it had not been for this industry)
c) failed to do a sensitivity analysis (i.e. how sensitive their results are to different agricultural yields, productions practises)
d) fails to mention whether the energy cost of the seeds did take into account the fact that switchgrass is a perennial plant with a lifetime of 10-15 years and can be harvested without reseeding for that period
e) never i referenced the original ONRL switchgrass study which had looked into the farm to wheels EROEI of ETOH from switchgrass
In spite of a) to d) they confirm previous studies which did conclude that growing switchgrass (note for use as fuel in pellets) does have a positive energy return of about 11 (the ORNL study EROEI range was 12-17)
Note:
Point e) by it self would have been enough to lead to a rejection (or at least a resubmission) decision by the paper's reviewers but this is an issue that the editorial board of Natural Resources Management will have to address with the reviewers when/if NREL decides to get really serious about defending their own data/conclusions (as they ought to).

2) Then Pimentel proceeds to calculate the energy return ratio of the swithgrass to ethanol process and reaches the conclusion that Ethanol production using the switchgrass previously grown required 50% more fossil energy than the ethanol fuel produced . Is this number surprising? Does this tell the whole story? What does this number mean?
For one thing it does not mean that the EROEI of growing AND converting switchgrass to ethanol has an EROEI of 0.66 but only that the second step wastes about about 1/3 of the energy gained in the first step. Is this surprising? Hardly. Ethanol is a product of switchgrass, the conversion process is less than 100% efficient and it requires an energy input. Hence when one produces a fraction or distills a particular fuel one always ends up with less energy than one started, a statement that is as surprising as the statement 1+1=2.
The key question is how much energy is gained or spent in going from switchgrass seed to ethanol, a calculation that Pimentel never did.
By not combining the data of these two processes Pimentel arrives at a rather (non-)interesting conclusion, but one that is likely to be recited in press releases and create the erroneous impression that growing switchgrass and processing it to ethanol is an energy loser. The un-interesting conclusion I alluded to is that when one is distilling or fractionating a fuel one ends up with less energy than one started.
However this truism is elevated to the status of important scientific finding, it appears in the abstract (the only part of the paper likely to be read by the press) and the one likely to stick to people's minds. IMHO The only real lesson learned from this is that the editorial board of Natural Resources Management should re-examine their pool of reviewers OR simply not publish papers they cannot review.

However if one is to combine the two processes then the picture changes dramatically.
Process A: 1 unit of energy invested results in a net gain of 11 units of embodied energy in switchgrass
Process B: 1 unit of energy produced in step A yields 0.66 units of energy in the form of ethanol
Hence the EROEI of process A+B is 11 x 0.66 = 7.26 !
Edit: Strictly speaking one should multiply this by the efficiency of the conversion process which for the case of swithcgrass is 50%. Pimentel calculated the EROEI on the basis of output of the process and not on the basis of the input; I realised that in the 4th reading of the publication. However even by dividing by two this number is close to 4 which is significantly > 1

Note that the original ORNL study did result in an EROEI of 4.34, hence Pimentel's/Patzek's data/calculations not only support but actually improve upon ORNL's findings yet he reaches the exact opposite conclusion. Is this an oversight or a deliberate misrepresentation? I believe the latter to be true given his track record of using outdated data, misrepresentation of studies, extremely wide system boundaries, selective data citation and circular references (usually of himself).
What it is more interesting is the fact that his conclusions fail to acknowledge the fact that the lignitic part of switchgrass (the one that cannot be converted to ethanol YET) is burnt in site to generate thermal energy and electricity which is used to "power" the conversion process. The latter point was addressed by John Sheehan here
JS writes:

Ethanol from Energy Crops—the ultimate alternative to fossil fuel

The most surprising suggestion in Pimentel and Patzek’s new publication is the claim that ethanol made from energy crops—trees and grasses containing cellulose and hemicellulose sugars—requires 50% more fossil energy inputs than the fuel energy it delivers. Pimentel and Patzek are uninformed about the technology for turning these new forms of biomass into ethanol. Studies by Argonne National Laboratory and the National Renewable Energy Laboratory have demonstrated that ethanol from energy crops and from agricultural residues like corn stover offer large fossil energy savings: savings of 90% or more in the case of energy crops like switchgrass and residues from corn production. Why the big difference? Pimentel and Patzek’s cursory review of the technology missed one very important design aspect for this new technology—the conversion of grasses and residues to ethanol is completely energy self-sufficient. That is, all of its energy needs are provided by the biomass, eliminating the need for the fossil energy that Pimentel and Patzek claim are needed to provide steam and power in the facility. It is unfortunate that such an uninformed claim has now been widely spread in the general media.

What this basically means is that process B's EROEI is marginally less than 1 and hence the EROEI of the combined process is approximately equal to the EROEI of process A (i.e. > 10)
I also find extremely interesting the following statement made by Sheehan:

Pimentel and Patzek fail to point out, for example, that gasoline and diesel fuel today actually do have a “negative” fossil energy balance.

which basically states that in distilling diesel and gasoline out of oil we spend MORE energy than we can recover in the process. Note that in
a) PO.com did try to discuss this issue here but no clear answer emerged.
b) in another discussion forum someone claimed that the conversion process EROEIs is 0.74 to 0.85 .
I cannot verify this statement but it in distilling gasoline out of the oil one always ends up with LESS energy than one started (after all gasoline is a fraction of oil) and hence the EROEI of the combined process of oil ground extraction AND distillation to gasoline is bounded from above by the EROEI of the oil extraction process (currently 8-10 for the lower 48 US states).
Note than in order for the combined process (growing switchgrass/distilling to ETOh OR extraction from ground/distillation to gasoline) to have an EROEI of less than 1, the following relations have to hold true:
EROEI(A) x EROEI(B) < 1 => EROEI(A) < 1/EROEI(B) =>
EROEI(A)< 1.5 (swithcgrass-ethanol)
EROEI(A) < 1.17 (oil-gasoline)
In the case of swithcgrass the EROEI(A) is approximately linearly related to the plant mass yields; hence it would have to drop to less than 10% of the estimate of 10 tons per hectare per year for the process to become an energy loser . Such conditions have never been obtained in switchgrass farms not only in North America but also in NW Europe (link) AND the Mediterranean (link) making this plant even more amazing (afterall it is not native to Europe).
BTW switchgrass does build organic soil content and actually sequesters approximately equal amount of carbon above AND below the ground .

I found the following white paper on bioenergy particularly interesting. It was written by David Morris of the ILSR who provides a checklist of the major differences between Pimentel's and USDA's/NREL's studies:

1. Crop Yields per Acre and Biofuel Yield per Bushel
2. Energy Used to Make Fertilizers and Seeds
3. Energy Used to Make the Ethanol and Co-products
4. The Energy Value of Co-products
5. Energy Used to Make the Machinery and Feed the Workers


Morris does say something interesting about Pimentel, an assessment that I agree with:

1. David Pimentel’s pessimism
about biofuels derives from a methodological
approach that leads him to a far
more sweeping and highly controversial
conclusion: the world’s population has
vastly exceeded its biological carrying
capacity.
Pimentel’s analysis leads him to conclude
that the world’s population of 6.5 billion people
has far surpassed the planet’s capacity to
feed that population. As he writes, “For the
United States to be self-sustaining in solar
energy, given our land, water and biological
resources, our population should be less than
100 million…”(the July 2005 population is 295
million).13 Pimentel further maintains, “the
optimum (world) population should be less
than…2 billion. 14
Pimentel’s pessimism about the world’s
capacity to feed its human population carries
over to his view about the limited potential of
renewable energy in general. In this he is
joined by Patzek, who with Pimentel recently
concluded that nuclear power may be the only
answer.
“We want to be very clear: solar cells, wind
turbines, and biomass-for-energy plantations
can never replace even a small fraction
of the highly reliable, 24-hours-a-day,
365-days-a-year, nuclear, fossil, and hydroelectric
power stations. Claims to the con-trary are popular, but irresponsible…new
nuclear power stations must be considered.”
15
Do two-thirds of us have to die in order to
allow the remaining third to live a comfortable
life on a sustainable basis? Must we rely on
nuclear power to provide us a reliable and sufficient
source of energy? These questions
dwarf that of whether the energy balance of
biofuels is slightly negative. One would hope
that reporters and others would attend to the
catastrophic predictions that result from the
full-scale application of Pimentel’s methodological
approach, rather than the tiny negative
impact predicted by its application to a
tiny slice of the world’s biological resources.

where the number 15 refers to another study by Pimentel:
Tad W. Patzek and David Pimentel, “Thermodynamics of Energy Production from Biomass,”
accepted by Critical Reviews in Plant Sciences, March 14, 2005.,
The PDF of the study can be read at Patzek's website (link)
and the statement appear's in the footnote number 64 on page 75.
Happy reading

[Barbara]

Not to spoil your very good post, but it seems to me that if something needs this much study to find out the EROEI, then it surely has a bad EROEI.

[EnergySpin]

Not to spoil your very good post, but it seems to me that if something needs this much study to find out the EROEI, then it surely has a bad EROEI.

This much study was needed to find out that Pimentel is "twisting" the data.
The ORNL's study on ethanol from switchgrass did confirm that its EROEI is positive. I'm not touching the corn data even though USDA and NREL have replied to Prof Pimentel in a satisfactory manner IMHO.
Amory Lovins also wrote about ethanol production from swithcgrass in his recent Scientific American article and AL is an efficiency guru providing extra confirmation to the credibility of this switchgrass to ethanol process.
BTW I'm surprised by the (lack of) quality of Pimentel's publications. Circular references, citing his own outdated studies consistently ( a no-no as any college freshman knows). In my field, papers with these characteristics would have been rejected by the reviewers and yet he is able to publishe such crap with impunity. I guess it pays to be an Emeritus (Pimentel) or a Tenured Professior (Patzek)

[Caoimhan]

I have one word to describe why Pimental's work is so twisted:

Agenda

[nero]

b) included the energy needed to feed the human farmers (as if humans were specifically born to do this job and would not have been born if it had not been for this industry)

I disagree with this point. It is absolutely essential to include the energy to feed the farmers. And the energy to feed his barber and the energy to feed his politician.

Well no it quickly becomes meaningless to draw the system bounday too wide. But then on the other hand drawing it too narrowly is just as bad. In fact I would say that any system boundary is totally arbitrary and therefore the absolute EROEI is a meaningless number. I would propose that a relative EROEI be used. Set a standard system boundary definition that can be used for all energy producing processes, and then report only comparative EROEI ratios.(eg. regular oil extraction produces 3 times more energy per unit of energy invested than the tar sands given similar system boundaries.)

[backstop]

Energy Spin -

I appreciate your critique of Pimental's methodology, and I can only say that a reputable scientist would be ashamed of it.

Two things amuse me about his efforts.

1/. The really cogent critique of agribusiness biofuels has nothing to do with EROEI, (which is a bit of a moveable feast given varying desperation for liquid fuels) but rather with the inevitable depletion of vital topsoil by industrial-scale chemically-fed monoculture. It is unsustainable, and will, I mean will, lead to IIIW farmland being taken out of food production to raise fuel-supplies in wealthy nations.

Why Pimental did not wish to focus on this lack of sustainability seems intriguing.

2/. His underlying promotion of the nuclear option demonstrates not merely his partiality or 'agenda', I think the need to use such flawed reasoning to promote that agenda in fact shows some of the weakness of the nuclear case. Only by smearing other options can TINA (there is no alternative) be made to shine forth.

The nuclear industry's utter lack of sustainability, to the point that it can't even scrape into the cliche official category of "Renewable," means that it cannot afford to set its propagandists onto criticizing the lack of sustainability of other options such as agribusiness biofuels.

regards,

Backstop

[Sparaxis]

It appears that both sides of this debate twist the numbers to fit their pre-ordained conclusions.

Here's an example of the other side of the equation. This is Michael Wang's study of ethanol vs gasoline, done at Argonne National Lab.

On slide 4, he compares the "EROEI" of ethanol and gasoline, and determines ethanol is 1.35 and gasoline is 0.81! Now on the face of it, you'd ask, why do we produce gasoline if it is a net energy loser when ethanol is such a clear winner?

But note his methodology. He assigns all the energy used to produce crude oil, transporting it, refining it, and distributing it to gasoline (and no indication what proportion of the total goes to gasoline. This of course is an issue since gasoline isn't the only product of crude.) But he leaves out the energy profit of crude to begin with! Obviously, we use gasoline because its energy profit (and density, and mobility, and storability) is far superior to the alternatives. I think to properly compare the two you need to incorporate the profit of the crude (which we don't use directly) just as ethanol studies include the profit of the corn or switchgrass. Crudely speaking (no pun intended), the actual EROEI here would be more on the order of, say, 20 x 0.81 = 16.2. Then you clearly see why the market for transport fuels is dominated by gasoline and not ethanol.

Wang helpfully includes a discussion of energy quality (why we produce electricity at a loss and make batteries), and the system boundaries issues, and acknowledges that ethanol's energy profit does matter on where you draw those boundaries.

But I find the debate between Pimentel's 0.73 and Wang's 1.35 and USDA's 1.62 to be almost laughable since 1. ALL of these measurements are far far lower than conventional petroleum and require huge ramp-ups of energy INPUTS to even make them happen, resulting in even higher energy consumption after conversion than before and 2. we don't have the land to make a huge difference. Note the discussion starting on page 34 of this report, sponsored by the Energy Foundation (a leading US energy efficiency NGO in San Francisco), about whether we have enough land to grow switchgrass to supplant gasoline. Their own calculations (and the report is quite pro-ethanol) shows that in 2050, to replace US projected gasoline consumption using current yields and conversion technology, it would require 1.75 billion acres to be planted, compared to 1.9 billion acres of land in the entire US and 400 million acres of cropland. This is followed by a lot of "ïfs" and assumptions and technology improvement, and so forth to bring that down to 114 million acres (compared to 90 million acres now planted in corn). Then it blithely notes that we can just tell farmers to grow switchgrass instead of soybeans for animals since it's just "animal feed" and switchgrass would be more profitable.

Arg. In other words, by any way possible, we must project our current economic paradigm in the future and make whatever sacrifices necessary to maintain our ability to drive around.

By all means we will produce ethanol and biodiesel, so whether EROEI is less than 1 or more than 1 really doesn't matter. What matters is that none of these alternatives can be scaled to the level that would maintain our current consumption addictions.

[EnergySpin]

]
Backstop: using switchgrass IS NOT an unsustainable agricultural practise.
Switchgrass cultivation is used in the CRP program of the USDA as a method to build topsoil since they can fix both carbon and nitrogen in the soil. Lignocellulosic plants were investigated in the 90s as carbon sequestration methodologies. As far as NA is concerned switchgrass was the dominant species of the original prairies before they were converted to farm land. What does amuse me is that in spite of his critique of the unsustainability of the corn-ETOH cycle (he has written about it in various publications including the one I cited) he never addresses the fact that there are biofuels out there which behave differently, require minimal amount of fertilizer and irrigation and hence would be much more suitable for energy crops. Canada's REAP program did estimate that the use of 14% of the current farm land in NA could be switched over to growing perennial grasses in order to provide NA with 40% of its current liquid fuel needs while fixing carbon beneath the soil. The mean residence time for the carbon fixed in the soil compartment was 30-40 years on the basis of isotopic studies and this should be contrasted to the 7-10 years when carbon is fixed in tropical forests.
The corn-etoh seems to be a dead end IMHO; the return rate is so close to 1 that measurement error and annual variability in yields would likely make it a loser most of the time. As far as this aspect is concerned I would imagine that you, sparaxis and myself would agree on that.
As far as the wood to methanol/ethanol is concerned It would be instructive to follow his calculations: he did conclude that this is not worth it a finding that you especially would know it is incorrect. If I have time in the weekend I will run the calcs but If the swithcgrass case is an indication I would imagine we would reach similar conclusions


To Sparaxis:
The gasoline and corn EROEIs that are compared refer to different processes. The EROEI of gasoline is the energy required to distil one unit of gasoline from oil that is sitting in the refinery. It is not the EROEI of the process : pump oil from the ground + trasnport it to the refinery + refine it. The first two sides of the equation have a positive EROEI (8-10 for the US, close to 20 for Saudi) but the last process is an energy loser: only a fraction of the oil can be distilled and distillation requires input of energy.
In contrast the EROEI of 1.2-1.4 for corn ethanol refers to the whole process of going from seed to ethanol. If one believes the USDA it is marginally positive provided that there are no variabilities in agricultural yields/inputs etc. I have not read the 300+ page but till someone decides to do a Monte Carlo study to assess the variability in these calcuations I will not be convinced that corn-ETOH is a viable option.
As you have noticed I refrained from making calculations: that we can use XYZ acres to supply our needs in 2050 with ethanol/methanol/ammonia/hydrogen. I wanted to point out a practise of making unfounded inferences on the basis of existing data. There are more inaccuracies in his tables especially table 4.
E.g: the entry in the first line depicts the energy content of 2500 kgr of switchgrass is 694 thousand kcal. The last line in Table 3 however tells that 10000 kgr of switchgrass corresponds to 40 million kcals or that 2500 kgr has a thermal content of 10 million kcal (which is the correct). Anyone wants to combine Tables 3 and 4 now?
Hint: The truth is even uglier (i.e. Pimentel's data are off the mark AND NRM editors should really get rid of the paper reviewers) than my first calculation suggests especially since NREL is correct and modern plants do use the lignocellular component for heat and electricity generation ...
People that have access to a calculator should be able to figure this by themselves

[pstarr]

Energyspin,

Point 1. Switchgrass eroei is 11-- for stove pellets. You don't propose returning to steam cars post-peak? The Switchgrass still needs to be processed and fermented

So the rest of the lettered arguments for Switchgrass (a-e) really don't address Pimental's thesis at all.

The cheaper agricultural production cost of Switchgrass is offset by the higher fermentation costs. "The cost of producing a liter of ethanol using switchgrass was 54c/ or 9c/ higher than the 45c/ per l for corn ethanol production."

Your dismissal of the human cost to produce all crops is wrong. Once drilled, oil wells produce with practically no work. Farms are energy intensive and so is living as a farmer. Pimental did you a favor by not counting the cost to drive to church and entertain his kids. These are the energy costs of doing business.

You fault him for failing to do a "sensitivity analysis" Who cares? He uses standard accepted production values. Good enough for me

Who cares of Switchgrass is perenial. That saves the cost of seeds which Pimental keeps out of the equation.

I don't know the original ONRL switchgrass study but I have to assume that it too like the USDA study and the rest excluded the cost to develop, run, maintain, and repair and expensive modern industrial agricultural infrastructure that each and every farm is. Not to mention the fermenter and the delivery system.

You return to that phony eroei of 11 at the end as if that wraps up a provocative argument. It doesn't. I'm sure you can burn corn kernels in a Stanley Steamer and generate power also.

Point 2. I have no idea what this is about. It goes on and on and on comparing Switchgrass and Corn. The point is that corn and switchgrass are both crops that grow under the sun and require water, fertilizer, harvest, process, delivery, fermentation etc. You can nitpick but still not convince me that these are very different.

Both corn and switch grass are zero or next to zero as net energy producers. Add on all the other costs (environment, scale, food-replacment, economic justice, etc. etc. etc.) and you have to agree that there will never be enough liquid energy from these crops to power there own production, much less get mom and the kids to the soccer game.

[Antimatter]

Your dismissal of the human cost to produce all crops is wrong. Once drilled, oil wells produce with practically no work. Farms are energy intensive and so is living as a farmer. Pimental did you a favor by not counting the cost to drive to church and entertain his kids. These are the energy costs of doing business.

But don't these people have to eat anyway? I doubt the average farmer eats more than the average american, just look at all the fat people. They would still be eating if they were working as McMansion constructers or tanning hut managers instead.

Point 1. Switchgrass eroei is 11-- for stove pellets. You don't propose returning to steam cars post-peak? The Switchgrass still needs to be processed and fermented

You return to that phony eroei of 11 at the end as if that wraps up a provocative argument. It doesn't. I'm sure you can burn corn kernels in a Stanley Steamer and generate power also.

ES adressed that, including energy to ferment and distill EROEI comes out to about 7, and the process can be powered by burning the stuff that can't be fermented (yes some does have to be returned to the soil but switchgrass seems to build soil carbon anyway).

[Devil]

Of course, those with vested interests lie through their teeth to exaggerate in one direction to support their agenda. Those with opposing vested interests also lie through their teeth to support their's. Something as nebulous as the EROEI of ethanol can be cooked easily over two orders of magnitude, from EROEI = 0.1 to EROEI = 10, depending on what factors you choose to use and hundreds of other factors you don't choose. In any case, there can never be a fixed value of EROEI; can you harvest exactly the same tonnage of switchgrass or corn each year? Of course not, because you cannot factor in the weather, so the EROEI may drop a lot in a poor year.

Bio-EROEI is nothing but LIES, ALL LIES

[EnergySpin]

Energyspin,

Point 1. Switchgrass eroei is 11-- for stove pellets. You don't propose returning to steam cars post-peak? The Switchgrass still needs to be processed and fermented

This is not the point - I did not propose to return to steam cars (even though some people might disagree). And my intention is to show how he misrepresented HIS OWN DATA to reach an arbitrary conclusion.

So the rest of the lettered arguments for Switchgrass (a-e) really don't address Pimental's thesis at all.

Points a) to e) address both the thesis but most importantly the quality of the particular publication as any college graduate knows. Points a) to e) are standard evaluation criteria for a peer reviewed publication in any field. For example in my field any publication that addresses the benefits (or lack thereof) of a particular medical treatment should compare the results obtained with the results of other similar studies. If one does not do so, then there can be only a finite number of explanations:
1) one did not do his homework, hence the work is incomplete
2) one is deliberately "hiding" studies that have reached different conclusions. In this era of information explosion this particular strategy is used to create an inflated impression of the importance of the study. Usually this strategy works in journals only marginally relevant to the question under study OR when the paper did not pass the standard review process OR the review process is worthless.
The end result is that we have a worthless paper in a dubious journal which means that the study should be discarded AND usually but not invariably leads to the authors signing the study being discredited. Tough but true, this is how modern science works. One bad study and one's reputation is ruined. For example this particular study and a rather strange pattern of citations in Pimentel's work has previously led me to take whatever he writes with a lot of salt (tons of it), especially when there are obvious errors of numerical calculations and interpretetions in the majority of his (non-entomological) work. I do not care about his work as an entomologist and he probably does not give 2c about my work in auto-immunity but since this particular study does address issues of energy and society of which you and I are both members , I will dissect it to its bones.

The cheaper agricultural production cost of Switchgrass is offset by the higher fermentation costs. "The cost of producing a liter of ethanol using switchgrass was 54c/ or 9c/ higher than the 45c/ per l for corn ethanol production."

I did not address the financial aspects of this process; I would appreciate if we stayed focused in the energy issue. In any case ... if you want to do the divisions and honour us with the cost per litter of producing ETOH from switchgrass why don't you do so? Why do you focus on only one aspect of the process? Let me help you and do the calculations for you.
In table 3 Pimetel and Patzek (P&P) conclude that the 10000 (ten thousand) tons of switchgrass can be produced at a cost of 230$.
In table 4 P&P conclude that 2500 (twenty five hundred) tons of switchgrass can be converted to 1000 l of ETOH at a cost of $537. The $537 does include a $230 input for the production of switchgrass but this is obviously a numerical error: Table 4 lists the cost of the conversion of 2500 (and not 10000) tons of switchgrass to ethanol. Hence the $230 should be divided by 4. Equivalently one should subtract $172.5 from the total cost (3/4 of 230) leading to a final cost of $364 per 1000L or 36.5c per L or $1.38 per US gallon. Since the energy density of ethanol is approximately 2/3 of gasoline a simple division using P&P data suggests that in order to get the equivalent energy of 1 gallon of gasoline in ethanol produced by switchgrass one would have to pay $2.1 . His numbers not mine say that.

Your dismissal of the human cost to produce all crops is wrong. Once drilled, oil wells produce with practically no work. Farms are energy intensive and so is living as a farmer. Pimental did you a favor by not counting the cost to drive to church and entertain his kids. These are the energy costs of doing business.

One may argue that my dismissal is wrong. If the human that work on the energy farm were born for this specific purpose than by all means do include that cost. Otherwise one should leave it out. But let me honour you and Pimentel and include that cost. Did you read the table? Or are you simply quoting out of your head trying to create the impression that human labour is the major input that determines the energy budget of switchgrass production? Table 3 of P&P's study inform us that the human labour energy cost is 20000 (twenty thousand) kcals per ha per year. The total energy cost is 2755000 (two million seven hundred and fifty five thousand) kcals and thus the farmer energy cost is 0.7% of the total. I still maintain that this cost should not be included but I'm willing to concede. I have to point out that in discussions about the energy cost of gasoline one does not include the calories that people working on the oil wells, pipelines etc consume. One cannot say that they produce without energy costs, after all humans DO work on the oil fields (i.e. to do the watercuts ). If we are to maintain a consistent methodology then all these inputs should be taken into account. But this is beyond the swithcgrass point.

Who cares of Switchgrass is perenial. That saves the cost of seeds which Pimental keeps out of the equation.

You are wrong sir about the costs of seeds. In table 3 P&P list the energy cost of seeds to be 100 thousand kcals per year and their monetary costs to be $3 per year. The fact that switchgrass is perennial is important because P&P never said whether they took this fact into account to arrive at this numbers. Switchgrass plantations are seeded once per 10 years on average hence the cost of seeds should be amortized over the same period. Since he never told us whether he did take this simple fact into account what am I supposed to assume? For the purpose of this discussion let's give him the benefit of the doubt and assume that he did take this into account. But this important distinction between perennial and non-perennial plants should be kept at the back of our heads for future reference.

I don't know the original ONRL switchgrass study but I have to assume that it too like the USDA study and the rest excluded the cost to develop, run, maintain, and repair and expensive modern industrial agricultural infrastructure that each and every farm is. Not to mention the fermenter and the delivery system.

Wrong again .. they included the same kind of inputs that P&P included and gave a minute credit for electricity/thermal cogeneration at the factory that ferments the alcohol. I did post the link at the beginning of this thread and I will provide it again here:
http://www.public.iastate.edu/%7Ebrumme ... -Walsh.pdf

You return to that phony eroei of 11 at the end as if that wraps up a provocative argument. It doesn't. I'm sure you can burn corn kernels in a Stanley Steamer and generate power also.

The EROEI is not mine it is Pimentel's, and it is phony as you will read below.

Point 2. I have no idea what this is about. It goes on and on and on comparing Switchgrass and Corn. The point is that corn and switchgrass are both crops that grow under the sun and require water, fertilizer, harvest, process, delivery, fermentation etc. You can nitpick but still not convince me that these are very different.

Corn and switchgrass are both plants correct. Their farming has much different requirements. Switchgrass is a natural plant of the North American prarie; growing it requires much less "fertilizer" input than corn. This is the reason why P&P have omitted various non nitrogenous fertilizer inputs; they are simply not needed OR are needed in such minute amounts to make their inclusion irrelevant.

Both corn and switch grass are zero or next to zero as net energy producers. Add on all the other costs (environment, scale, food-replacment, economic justice, etc. etc. etc.) and you have to agree that there will never be enough liquid energy from these crops to power there own production, much less get mom and the kids to the soccer game.

This is not a stamenent backed by data (even Pimentel's data). The process can power it self. The question of scale is a different question IMHO. One should first determine whether a positive gain is possible and then consider the absolute amount of energy that is obtainable under the constraints you mentioned. However I do fail to see how "justice" is a constraint; in addition for all your "environmental" concern you fail to see that growing switchgrass builds rather than destroys topsoil. That was the prime reason that it was investigated in the CRP and the carbon sequestration program. Note than at least in NA corn replaced swithcgrass when the native praries were converted to farm-land decades ago. I would expect the Europeans to be more concerned about swithcgrass growing in Europe since the plant is not a native species there but I'm sure they can find similar species.

You fault him for failing to do a "sensitivity analysis" Who cares? He uses standard accepted production values. Good enough for me

Pstarr you chose to misinterpret my thesis but then again ... the English is pretty clear. The fact that you dismiss the value of sensiticity analysis is totally wrong as any engineer/agricultural/medical researcher would tell you. In fact the reliability of the process in the long run does depend upon the variability of the outputs of the various steps in the process and sustainability judgements should be based on temporal averages. A link to the original ORNL study was provided in my original thread. You can read it at will.
I'm afraid that at the end of your post you are mixing science with politics. Science can only provide data to aid in the political decision making and it should do so in an objective manner. Even if the numbers add up there might be overriding reasons not to embark on this endeavour but one should not mess with science to to promote a certain agenda.


For the fun of it I will copy the data from Pimentel's table and do the calculations the way he did it and correct the numerical errors on the way.
I will not do the economics; the thread was and still is about the EROEI.


Table 3 Average Inputs and Energy Inputs Per Hectare Per Year
for Switchgrass Production
Input Quantity ( 10^3 kcal)
Labor: 20
Machinery: 555
Diesel: 1,000
Nitrogen: 800
Seeds: 100
Herbicides: 300
Output Quantity:
Total yield 40000 (corresponds to ten thousands tons of material)
Energy Input: 20+555+1000+800+100+300 = 2755
EROEI = 40000/2755 = 14.5

P&P write the following though:

The average energy input per hectare for switchgrass
production is only about 3.8 million kcal per yr
(Table 3). With an excellent yield of 10 t/ha/yr, this
suggests for each kcal invested as fossil energy the
return is 11 kcal—an excellent return.

They fail to cite theirown calculations correctly; the total energy input sums to 2.755 million not 3.8 million !!!!
In fact the EROEI for growing switchgrass is 14.5 which is lower than the average reported in the 1998 ORNL study and similar to the number found by the Canadians in their biopellet studies. This is hardly surprising given P&P's overreliance on the Canadian REAP data. What is also surprising is that they list the energy return of switchgrass production to be 11 while the energy return on pellets is 14.5. Pelletization is an energy intensive process and hence I find it this statement suspicious. Since P&P jave a tradition in misquoting people I decided to delve into the Sampson, Duxbury, Mulkins publication that was quoted by P&P.
The study is found in the following URL (note the url of the reference Samson, Duxbury, and Mulkins, (2004) report. that appears in the P&P Bibliographic Reference Section has changed) :
http://www.reap-canada.com/online_libra ... %20and.pdf
On Page 8 of this report we find the following text (actually summarizing the results of the PERD study done in 200 by S&D&M):

An analysis of the energy costs associated with switchgrass fuel pelleting is important in
identifying the greenhouse gas offset potential of the technology. The energy costs associated
with switchgrass production and delivery to a large industrial user have been estimated to be
approximately 0.91 GJ/tonne for an 8 tonne/ha yield (Girouard et al., 1999b). In the case of
pellet production, the hauling distance would be reduced from an average of 60 km to 20 km, as
the pellet conversion facility is much smaller than a pulp and paper industry. This reduces the
energy cost to 0.79GJ/tonne and creates an energy output to input ratio for the crop (assuming an
energy content of the crop to be 18.5 GJ/tonne) of 23:1 (Samson et al., 2000). This high level of
energy output to input compares favourably to grain production, which is typically in the 4-6:1
range.
Additional energy is required for pre-processing, pelletizing, marketing and delivery of
switchgrass for use as a pelletized product. The energy costs associated with switchgrass fuel
pellet production is estimated to be 1.27 GJ/tonne (Table 4). Surprisingly, production and
delivery of switchgrass represents 62% of the energy required in the entire switchgrass fuel pellet
production chain from field to delivery to the consumer. This is largely due to the energy
associated with fertilizer use and application which represents 36% of the total energy cost.
Nonetheless the net energy output to input ratio is 14.6:1 (assuming an energy content of
18.5GJ/tonne in the feedstock). Considering that this material can be used quite conveniently as

It appears that P&P did their calculations assuming a production path from switchgrass seed to switchgrass crop to switchgrass pellet to switchgrass ethanol.
This is absurd ... ethanol production out of switchgrass starts with bales of switchgrass NOT pellets. Since pelletization has substantial energy inputs and only a fraction of the crop material ends up in pellets there is no way for the distal process (pellets) to have higher energy gain than the proximal process (crop). The way that table was constructed is confusing. One should not mix production data from Canada AND US just because he wants to; a justification is needed which is never given. This practise runs against any scientific research standards but Pimentel consistently does this in his energy/society publications.
In any case even If I accept that the data of Table 3 correspond to the path from seeds to switchgrass bales and not from seeds to pellets the EROEI of the process is 14.5 (not 11) as is claimed by the authors. Their text contradicts their tabulation, which has happened more than once in this publication and countless others. Clearly P&P either have no clue about simple arithmetic OR they fuzzify the data to reach whatever conclusion they have decided is correct. In other words they "adapt" reality to their conclusions and not the other way around. In my part of the science world this practise can be either one of two things: gross incompentence or fraud.

Let's finish the EROEI calculation for the path from swithcgrass seed to ethanol now. In order to do so I will write down the number from Table 4 and we will all join hands and do addition and division as they taught us many many many years ago


Table 4. Inputs Per 1000 l of 99.5% Ethanol Produced From
U.S. Switchgrass
Inputs Quantities kcal × 1000a
Switchgrass 2,500 kgr 694
Transport, 300
switchgrass

Water 70
Stainless steel 45
Steel 46
Cement 15
Grind 100
Sulfuric acid 0
Steam production 4,404
Electricity 1,703
Ethanol conversion 9
to 99.5%
Sewage effluent 69
The first thing that is fairly obvious is that there is a discrepancy between
the energy content of the 2500kgr tons of switchgrass in the first line of Table 4 and the last line of table 3. 2500 tons of switchgrass have an energy content of 10 million and NOT 694000 kcals. I have a hard time understanding how they came up with this number and even more importantly what this number means. Lets examine two different ways of doing the EROEI calculations:
Calculation 1
One does not take into account the energy input in the form of switchgrass bales. Under this scenario 1000 lit of ETOH (representing 5.1 million kcal of energy are produced by investing 6.8 million kcals (the sum of all the numbers in Table 4 except the first one). This means in order to get 1 unit ethanol back one has invested 1.33 units of energy. Equivalently the EROEI process of the fermentation is 0.75. What does interest us though is the EROEI of the combined process from seed to bale to ethanol which includes the cost for fertilizers, machinery, transportation, etc.
Note that the industrial process assumed by P&P is 50% efficient i.e. 50% of ALL energy contained in switchgrass ends up in ethanol.
Hence the holistic EROEI calcaution is 0.5 x 0.75 x 14.5 = 5.4375 which is the close to the number 4.34 produced by the previous ORNL study.
Calculation 2
NREL maintains that 90% of all energy contained in switchgrass can be utilized either in the form of ETOH OR used to generate electricity/heat that is used to power the conversion process.
Hence the 10 million kcals in the 2500 tons of switchgrass bale are partitioned to 5 millions kcal of ETOH, 4 million kcal of electricity/heat and 1 million is lost in heat/residue (this would include the majority of non carbon non hydrogen non oxygen mass in the plant).
If one is able to utilize the 4 million kcals in the fermentation process then the energy inputs have to be reduced by 4 million kcals or alternatively that 5 million kcals of ethanol can be produced by 2.8 million kcals of energy. This fact alone makes the fermentation an energy winner even though we have lost close to 10% of all the energy gained by growing switchgrass. The calculations now read as: 0.5 x 14.5 x 5/2.8 = 12.9

[EnergySpin]

Something as nebulous as the EROEI of ethanol can be cooked easily over two orders of magnitude, from EROEI = 0.1 to EROEI = 10, depending on what factors you choose to use and hundreds of other factors you don't choose. In any case, there can never be a fixed value of EROEI; can you harvest exactly the same tonnage of switchgrass or corn each year? Of course not, because you cannot factor in the weather, so the EROEI may drop a lot in a poor year.

Correct that is why a sensitivity analysis is needed. The EROEI should be conceptualized as a random variable with its own distribution. And to calculate such a distribution we need longitudinal data, observations that last for years and an understanding of reseeding requirements , variability in agricultural yields etc. And choosing the boundary is key. For example any process that includes the solar input in the input part of the equation will have an EROEI < 1. Refining oil (switchgrass) to gasoline(ethanol) WILL ALWAYS have an EROEI of < 1 if the oil that entered the refinery is used in the calcuation.

The same can be said about any process that converts energy though.
I think that the cry: "lies lies lies" is jumping the gun. The proper response is to mandate full data transparency and proper statistical analysis, including stastical simulation and queue modelling. These are standard tools of process engineering and it is surprising that no one is doing it in this field.

[Devil]

...stastical simulation and queue modelling. These are standard tools of process engineering and it is surprising that no one is doing it in this field.

Not at all, because if they did, they would have tell even bigger lies to further their agenda

[EnergySpin]

...stastical simulation and queue modelling. These are standard tools of process engineering and it is surprising that no one is doing it in this field.

Not at all, because if they did, they would have tell even bigger lies to further their agenda

That goes both ways I'm afraid as you pointed out. And unfortunately worthless papers like the one by P&P make the debate extremely difficult. If his contribution and objections are not backed by robust arguments then the other side will not even bother to strengthen their numbers and their methodology and the rest of us will be left in the limbo.

In any case I'm surprised that no one has nitpicked my Calculation 1 and 2. There is something interesting hidden there ... but the Numeracy Literacy Lesson will stop here. People should be able to do divisions, additions, subtractions, multiplications and dimensional analysis on their own and stop parroting others.
A few points before I sign off on this thread (since my ROI has dropped to < 1 at this point): scaling up the procedure will lead to improved EROEIs (but I'm not telling which is the correct baseline ) .... after all one does not expect a biorefinery to refine only 1 hectare or that water will not be reused or that a portion of the switchgrass will not be burnt directly to generate heat and electricity in order to maximize the yields (hint: this requires a simple application of linear programming) ?

It would be interesting to see figures for the refining process of oil to gasoline ... after all distillation is an energy intensive process and one has to account for all this infrastructure. How much energy do we need to produce one galon of gasoline? Even if we do not include the energy content of the oil that is refined, this number will propably be < 1. Maybe one of the geologists/petroleum people here has the answer.

Regarding the process from switchgrass (or other perennial grasses) seed to switchgrass bale or switchgrass pellets. This has been looked into by various groups in 5 different countries in 2 continents. The return rate (accounting for equipment, fuel, electricity and even feeding people) is definitely positive. Even in the highly energy inefficient US sector the return rate ranged from 12 to 17 (taking into account the variable agricultural yields). The ORNL study is extremely revealing and by reading it one can highlight what is hidden in Calculations 1 and 2 in a previous post on this thread.

[EnergySpin]

It appears that both sides of this debate twist the numbers to fit their pre-ordained conclusions.

Here's an example of the other side of the equation. This is Michael Wang's study of ethanol vs gasoline, done at Argonne National Lab.

Thanks for the link Sparaxis. Even though I started the thread using Pimentel it will be interesting to delve on the other side as well. No promises though that it will be done this weekend since the GREET model will need to be used. Note that Wang makes an assumption that is valid IMHO : since we are not responsible for the thermonuclear reactions that occur inside the sun, reactions that are the source of energy for growing corn (along with fossil fuel inputs) then this energy input should not be used in the calculation of the holistic EROEI. But as you said he failed to give credit to the drilling operation that allow us to get 10-20 barrels of oil for every barrel we spend on drilling .

Crudely speaking (no pun intended), the actual EROEI here would be more on the order of, say, 20 x 0.81 = 16.2. Then you clearly see why the market for transport fuels is dominated by gasoline and not ethanol.

Actually in order to use this calculation one should know what is the fraction of the energy content of oil that ends up in gasoline.
For example if this % is 40% then the EROI for gasoline is 16.2 x 0.4 = 6.48. If it is 80% gasoline's EROEI is 0.8 x 16.2 = 12.96

Arg. In other words, by any way possible, we must project our current economic paradigm in the future and make whatever sacrifices necessary to maintain our ability to drive around.

I hope not .... this is the time to change things around. But we should be pretty honest about our data and not mix science with politics. Science answers technical questions, but the public decides hence the quality of data should be suberb.

By all means we will produce ethanol and biodiesel, so whether EROEI is less than 1 or more than 1 really doesn't matter. What matters is that none of these alternatives can be scaled to the level that would maintain our current consumption addictions.

True ... the REAP people were much more conservative that's why they put a bound of 14% (or current farm land) to provide 40% of the fuel needs in 2001.
But the point remains: bad calculations do infest one side's arguments more than the others hence destroying the opportunity of meaningful debate as Sheehan said. We should not fantasize about miraculous new technologies but at the same time we should use current data (reflecting what is done now rather than 20 years ago) to assess any new technology.

I found slide 16 to be particularly revealing ... most of the authors (i.e. Wang and Pimentel) appear more than once. Are these separate , indepedent studies or are they recycled products?
From the view point of non energy scientist I find it interesting that no formal meta-analysis has not been published. This is how we solve issues in the medical field .... which makes me proud (for once!) for the quality of research work in Medicine.

[EnergySpin]

Wang's GREET model can be found in the following website:
http://www.transportation.anl.gov/softw ... index.html
It might be instructive to run a few sims on the processes involved since the model seems to be able to accomodate different boundary assumptions.
And being an EXCEL spreadsheet with a GUI anyone should be able to use it.
Note: Registration is required prior to download