The Oil Drum | An EROEI Review

220 comments on An EROEI Review

Comments can no longer be added to this story.

| Show without comments | PDF version

Nate Hagens on March 12, 2008 - 12:38pm

Robert - a few thoughts on your nice summary.

1)In a pending paper, "A Consistent Definition for EROEI", Mulder, Lillies and Hagens point out that in situations where a co-product is used in the energy harnessing process itself (bagasse for sugarcane ethanol as an example), the traditional EROEI formulation does not account for the opportunity cost of the precursor energy input and thus overestimates the EROEI of the process.

The above graphics from that paper show the same process (cellulosic or sugar cane ethanol) measured in two ways. The bottom diagram shows traditional EROEI = energy out / energy in. The top graphic indicates that the intermediate step biomass K2 (the bagasse) has an energy opportunity cost as it could be used for other energy uses. This 'loss' of availability has to be considered, which translates to EROEI = (Eout +(E lost-Ein)) / E lost.
From our paper:

What does this imply for the EROI of cellulosic ethanol? Drawing on a review conducted by Hammerschlag [11] of four net energy studies, we averaged the energy inputs and outputs to produce estimates for system energy flows for cellulosic production. Flows follow Figure 3a with the exception that the available energy from Biomass A (the lignin) is higher than the required input to the ethanol processing system, thus yielding an additional energy output. Estimates are as follows on a per liter of ethanol basis:
Energy In = 5.3 MJ.
Energy from Biomass A = 32.5 MJ.
Energy into the biomass processing system = 29.0 MJ.
The surplus energy from Biomass A that is outputted = 3.5 MJ.
Ethanol production (Biomass B) = 23.6 MJ.
Using the intuitive definition, the EROI measure would be Eout ( = 23.6 + 3.5 = 27.1 MJ) divided by Ein ( = 5.3 MJ) for an EROI of 5.5, significantly higher than soy biodiesel or starch ethanol. However, using equation (1), we have:

EROI = 27.1 + (34.3-5.3) / 34.3 = 1.7

where E lost = 5.3 + 29.0 = 34.3 MJ. This value is only marginally better than reported EROI measures for starch-based ethanol [7]. A similar exercise shows that the high EROI numbers for Brazilian sugar-cane based ethanol, which uses the bagasse as an intermediate input, are also overestimations.

As you rightly point out Robert, the bagasse is probably being used for its best use, but in a different economy it might be used in many other energy technologies; heat, biogas, electricity, etc. that would have higher social efficiencies. So the bagasse has to be considered an energy opportunity cost, and if not counted as an input it will overstate the EROEI. This is relevant if we would change how society used energy. There would be less need for high energy surplus liquid fuels if things were done more locally, or with more electricity, etc.

2)You are absolutely right about declining EROEI forcing our hand. We can 'grow' with lower net energy, but only if the smaller energy surplus is offset by some mix of a)conservation b)efficiency or c)move to lower energy footprint infrastructure.

This all boils down to an acceleration of pulling in resources, both energy and non-energy inputs, in order to continue the ICE mode of transportation - a path chosen for efficiency and ease many decades ago. The more we try to generate liquid fuels with low energy gain, the faster we will use water, natural gas, corn, coal, etc. - the things that have NOT been in shortage heretofore.

I think as soon as there are more limiting inputs than just liquid fuels people will start to internalize the biophysical angle on this story. After all, dollars are infinite, but high density, easily transportable energy is not.

Comments can no longer be added to this story.

Petroleo on March 18, 2008 - 11:43am

So what you are saying is that if you burn biomass A and B together in a power plant and use the electricity to power a BEV, then that could (depending on the efficiency of the plant, transmission and batteries) be more efficient than using the energy in biomass A to convert biomass B to ethanol and run the car on that? Interesting.

Comments can no longer be added to this story.

Neven911 on March 18, 2008 - 4:44pm

Yes, This is the ridiculous thing about the obsession with liquid fuels. An internal combustion engine with an optimistic efficiency of 30% has an EROEI of 0.30 whereas you could expects a BEV to be 0.8+. The whole thing is based on the continuation of ICE based transport which are the "installed base" creating inertia in the re-engineering of the system. The justification is "Range", evidently though 80%+ of trips are under 40k's we require ranges which match liquid fuel cars, this is clearly untrue.

There is some truth in the old joke "God managed to make the world in seven days because (s)he had no installed base"

There are some signs of sanity in the asylum though, I see VW have totally dismissed Hydrogen and are delaying their entry into the hybrid fad as long as possible, VW and Toyota must be licking their lips at their prospects in the US in the coming years

Neven MacEwan B.E. E&E

Comments can no longer be added to this story.

Engineer-Poet on March 19, 2008 - 12:31am

The whole thing is based on the continuation of ICE based transport which are the "installed base" creating inertia in the re-engineering of the system.

Which is silly, because the useful lifespan of a light-duty vehicle is ~17 years and 50% of the lifetime mileage is driven in the first 6 years. Designing infrastructure with a 50-year lifespan to suit ephemeral vehicles is letting the tail wag the dog.

The justification is "Range", evidently though 80%+ of trips are under 40k's

The PHEV is the solution to that.

I see VW have totally dismissed Hydrogen and are delaying their entry into the hybrid fad as long as possible

Good and bad, respectively. Hybrids are no fad; the PHEV is an evolutionary path to the pure BEV as battery technology improves and prices fall.

Comments can no longer be added to this story.

emalby on March 18, 2008 - 12:56pm

Well, the problem here is the boundary of the system. When you have a system and moving the boundaries you get different results for the same indicator referred to the same variable then it means that there is something wrong going on... In this case what is wrong is the indicator itself. EROEI is an easy indicator but should be used always the same way. In fact, it changes simply moving the boundary. Then this means you have to decide: using it as in the first case (arrows generated within the boundary and used in the processes within the boundary must remain in the boundary) or as in the second case (arrows must exit and then go back inside if they are meant to go in a process within the boundary).

If you assess a system using LCA, for example, you don't have these problems. Moving the boundary would generate the same results...

Comments can no longer be added to this story.

Sparaxis on March 18, 2008 - 3:28pm

Nate, I'd be very interested in reading your paper when it is released.

I'm curious how you derived some of the figures you have in your example. I can't replicate what is often offered as energy balance of a (theoretical) cellulosic ethanol plant when tracking the mass balance of the plant.

Take 1 kg of switchgrass for example.

It's 42% cellulose, 31% hemicellulose, and 27% lignin (including 0.7% ash).

From the cellulose, assuming 100% recovery, the stoichiometric ethanol yield of 51%, and 75% fermentation efficiency of glucose, you get 0.16 kg (0.20 l) of ethanol, 0.21 kg of CO2, and 0.05 kg of other mass (additional bacteria body mass; dilute solids)

From the hemicellulose, assuming 100% recovery, and 50% fermentation efficiency of xylose, you get 0.08 kg (0.10 l) of ethanol, 0.15 kg of CO2 emission, and 0.08 kg of other mass.

The balance is 0.27 kg of lignin, at 21 MJ/kg energy content, or 5.7 MJ. Biorefinery direct energy requirement for cellulosic ethanol production is 28 MJ/l-output (EBAMM 1.1), or 8.2 MJ to produce the 0.30 liters from the 1 kg of switchgrass input.

My question is, how does this 5.7 MJ of lignin per kg of switchgrass input provide all the processing energy in the plant (including drying the lignin, which is in solution when separated) and generate enough electricity to export 1.9 - 5.4 MJ/l of electricity? (Range in Hammerschlag)

If you zero out the lignin "credit" in the biorefinery in the EBAMM model, the EROI drops to 0.88, including the 4.8 MJ/l "credit" for some undefined byproduct.

Comments can no longer be added to this story.

thomas deplume on March 19, 2008 - 9:18pm

Anaerobic digesters use wet biomass to generate methane which can then be used by the biorefinery instead of natural gas(fossil methane). An Iowa engineer figured that a distillery's entire energy needs can be met by digesting the cellulose and lignin portions of the corn kernels.

Comments can no longer be added to this story.

Gail the Actuary on March 18, 2008 - 4:21pm

The problem with ignoring the energy costs of the biomass can easily be seen when you are working with wood. (I have heard people in the cellulosic ethanol field say that they prefer wood to switchgrass, because it is much easier to transport and store.) If you use a huge amount of wood to power your process for converting wood to cellulosic ethanol, you will have a high return on the fossil fuel inputs, but you are likely to have a very expensive process, since wood has other uses.

Comments can no longer be added to this story.

mdsolar on March 18, 2008 - 9:09pm

Nate,

It is customary not to count the renewable energy input in such calculations. About 2000 times more energy in sunlight falls on a cane field than ever comes out as ethanol. So, roughly, EROEI=0.0005 if the renewable energy in is counted. It seems to me that you are beginning to venture in this direction with your modification. I suppose you could say that the bagasse is processed so it is no longer just sunshine, but then, how do you count the wind that dries it so that it can be burned? I think you are OK counting ethanol burned in a truck to bring the cane to the plant as an energy input, but I'm not sure that counting the biomass after squeezing it is right.

Chris

Comments can no longer be added to this story.