When evaluating energy technologies – whether conventional fossil fuels or alternative energy – one thing that I pay close attention to is the Energy Return on Energy Invested (EROEI). While there are legitimate criticisms of the methodology, it can serve as a useful tool for comparing and contrasting various alternatives.
To give a flavor for why this is, consider an example. Let’s say society as a whole produces 50 million barrels of oil equivalents (could be oil, nuclear, wind, solar, biofuels, or a combination). Consider a couple of energy options. Option A has an EROEI of 10/1 (Energy Output/Energy Input). Option B has an EROEI of 2/1. Option A has to consume 5 million barrels to produce 50, for a net of 45. This net is what would be left for powering transportation, heating homes, and fueling industry. Option B, however, requires an input of 25 million barrels, so the net from the initial 50 is only 25 million.
The implications of this are that as EROEI falls, society must produce a lot more energy just to stand still. Even if total energy produced is constant, a falling EROEI means that there is less net energy left over after the energy input bills are paid. And because the easy energy is produced first, as time goes by this is in fact what happens: EROEI declines, and then it takes more time, effort, and money invested across society to keep things running. (Or, as EROEI declines energy efficiency must increase at such a rate that what is lost from the decline is made up from increased efficiency).
That’s a very basic introduction to EROEI. For a much more detailed look, see Understanding EROEI. In that essay I look at a number of examples, and explain how the EROEI of Brazilian sugarcane ethanol is probably much less than the 8/1 that is generally claimed, but that model still works well because a large portion of the energy inputs are waste biomass left over from sugarcane processing.
Over the past few years, I have seen a lot of speculation about the EROEI of tar sands (also known by the more marketable term, ‘oil sands’). I had seen estimates ranging from as low as 1.5/1 up to 4 or 5/1. My own suspicion has been that the number was higher than that, and I once did a back of the envelope based on some industry energy usage numbers that put the number at about 8/1 (for just the oil production step).
But now I have a much better number, thanks to a recent discussion at The Oil Drum. A reader linked to the following story:
Q&A with Marcel Coutu of Syncrude
This is the best reference I have ever seen for the EROEI of tar sands. Here are the important bits:
Oilsands Review: How much energy do you consume for every barrel of oil you produce?
Marcel Coutu: About 1.5 gigajoules (1.5 MCF of natural gas equivalent) per barrel. That’s higher than 0.8 MCF, the number I mentioned earlier; that refers to purchased energy. The total energy we consume in our operations includes energy we generate as a by-product to our upgrading processes. It is largely electrical energy, in which we are more than self-sufficient.
We produce a lot of waste gas from our processes, and use that to fire gas turbines. We also have a lot of waste heat from our operations, and we raise steam with that heat and put that steam into steam turbines. This makes our operations more efficient.
So, what we have is that some of the energy that is used is produced by the process. This is the accounting that results in an 8/1 energy return for sugarcane ethanol. By sugarcane accounting the EROEI of tar sands is about 5.8 million BTUs (the value of a barrel of oil)/0.8 million BTUs (the approximate energy content of 0.8 MCF that was externally purchased), or 7.25. By true EROEI accounting – which includes the internally consumed energy as an input – the EROEI would be 5.8/1.5 = 3.9.
Of course then the oil has to be refined. For a light, sweet oil such as the output of a syncrude unit, that step is going to be 12/1 or better. Putting the two steps together, I calculate that I need to spend 1.5 million BTUs to produce the oil, and another 5.8/12 = 0.5 million BTUs to refine it to gasoline and diesel. Total process is then 5.8 million BTUs/2 = 2.9/1 for the production and refining processes. Conventional light, sweet oil is around 6/1 for the entire process of oil in the ground to gasoline in the tank.
Let’s look at one more example to understand the implications. Let’s say we want 10 gallons of gasoline equivalent for our car. We need to solve two equations: Net Energy = Energy Output – Energy Input; and EROEI = Energy Output/Energy Input. If we combine equations and solve, we find that for light, sweet oil at a 6/1 EROEI, the total energy that must be produced is 12 gallons of gasoline equivalent. Two gallons of gasoline equivalent were consumed in the process of producing the 12 gallons, netting 10 gallons for the end user.
If we wanted to produce gasoline out of tar sands at a 2.9/1 total ratio, then 15.3 gallons of gasoline equivalent must be produced. 5.3 gallons would be consumed in the process, netting 10 to the driver. What I conclude from that is the tar sands is more than 2.5 times as energy intensive to refine to gasoline than is conventional oil.
While I don’t know what the ‘real’ EROEI is of sugarcane ethanol, it is probably in the vicinity of tar sands. However, as stated the big difference is that the bulk of those energy inputs are waste biomass, which dramatically boosts the sustainability of that option. Sugarcane ethanol – even if it has a lower energy return than tar sands – far exceeds tar sands in the sustainability category. This is one of the weaknesses of EROEI accounting; accounting for energy inputs from diverse sources – some more sustainable than others.
I too like the EROEI concept, but I wonder if the good ol’ price mechanism is more important.
It seems to me that low EROEI methods would fail naturally in free markets. Of course, the external costs (pollution, vulnerability to blackmail) of a product should somehow be included in the price. Usualy taxes can accomplish that.
And sometimes, I wonder if EROEI is even calculable. For example, palm plantations are discovering that much of their waste biomass can be used in the making of medium-density fibreboard (mdf). What sort of EROEI credit do they get for that?
The price mechanism is a wondrous thing.
I’ve got to admit, I’ve always wondered if we’d produce oil at unity EROEI. ie 1:1 EROEI
Consider the case of a nuclear plant. The only way for it to produce storable energy is to produced electricity at circa 35% efficiency and then use this to electrolyse hydrogen at perhaps 70% efficiency.
This would be a EROEI of 0.24:1 or about 1:4
In effect you’d only get one quarter of the upper heat value of the steam (from the reactor) in storable hydrogen.
But consider the case of using nuclear power to run, for example stripper wells requiring large pumps.
You might only get 1 BTU out for each BTU put in, but it sure would beat a 1:4 process…..
A lot of folk on TOD confidently exclaim that oil will not be extracted once its EROEI reaches 2:1 or worse.
I’m pretty sure they’re wrong, mainly because even at unity EROEI oil is still worth obtaining as a storable fuel.
Andy
My understanding was that ‘oil sands’ is the more accurate term. Wikipedia says: “Tar sands is a colloquialism for what are technically described as bituminous sands, and commonly known as oil sands or in Venezuela, extra heavy oil.”
Consider the case of a nuclear plant. The only way for it to produce storable energy ….
Another way is to have the electricity flow into 200 million car batteries. Smart chargers would only charge when there is excess energy available, creating a “dispatchable load”.
You know, I hate to risk a Godwin’s Law violation, but I can’t think of a more clear example of a calculation that better approximates an attempt to estimate the efficiency of the trains ferrying Jews to camps in Eastern Europe during WWII.
Whether you call them tar sands or oil sands, what you are really talking about is a fantastic effort to turn a vast area into the pits of hell to add a few drops to the massive quantities of energy waste squandered daily, while producing prodigious amounts of greenhouse gas, while burning equally prodigious quantities of the best of the fossil fuels (natural gas).
I realize that the calculation is simply on EROEI and that there’s a certain appeal to the engineer in us, but we have a duty as humans living in time of Peak Lite and on the cusp of the real thing to remember that unidimensional thinking is how we got into this mess in the first place, and that gross violations of environmental responsibility need to be stopped, not analyzed for their efficiency.
I’m pretty sure they’re wrong, mainly because even at unity EROEI oil is still worth obtaining as a storable fuel.
You are correct, with the following qualification. You would never use a transportation fuel to produce another transportation fuel with an EROEI of unity (unless the market was distorted by subsidies).
You might use nuclear or coal to produce a transportation fuel, or electricity, at even less than unity. The economics might dictate that this makes sense. It could make sense to use nuclear to electrolyze water and then use the hyrdogen in a fuel cell. So this is the caveat that must be applied when doing EROEI analysis. You need to compare apples to apples. When doing so, EROEI can be very useful. But EROEI alone does not tell the entire story.
Cheers, RR
I realize that the calculation is simply on EROEI and that there’s a certain appeal to the engineer in us,
I almost put a disclaimer at the beginning of the essay stating something to the effect of “This should not be taken as an endorsement of tar sands…”
The calculation was done merely to fill a need: People wonder about the energy return of tar sands. I think this is a pretty definitive example. But I think the latter part also showed the futility of thinking tar sands will ever displace a major fraction of our oil production. Let me just say that even if it did, the environmental impact would be far greater than for an equivalent net production of conventional oil.
Cheers, RR
A nuke can produce hydrogen directly without going through the electricity step.
How much more could the energy return be increased if petcoke or other oil waste was gasified to generate hydrogen for the extraction process?
Robert:
Does your 6/1 ratio for light sweet crude include transportation? Or is transportation negligible?
Jerry Unruh
My point about producing storable fuel from nuclear was more about truck/ship/aircraft fuel than car fuel….
Andy
You and RR should get a room.
Seriously, I took your advice and looked RR previous posting.
“When evaluating energy technologies – whether conventional fossil fuels or alternative energy – one thing that I pay close attention to is the Energy Return on Energy Invested (EROEI).”
The correct way to evaluate the environmental impact of energy technologies is using LCA per ISO 14000. EROEI is popular Monday morning quarterbacks, bloggers, and college professors at community colleges.
LCA would include energy to be sure but should include other significant factors such as water pollution.
Sure, LCA is a good way to evaluate these technologies as well, but it can get pretty hairy trying to do a COMPLETE LCA and account for all direct and indirect pathways of likely significance, witness the current debate about having EPA quantify GHG from indirect land use changes for US biofuels. Like RR said recently, if they can’t even agree on NEV (or EROEI?), how are they ever going to develop consensus on LCAs??
Even just on the water side, it would be interesting to evaluate oil sands vs corn ethanol. US will use about 25 million acres of corn this year to produce 8-9 billion gallons of ethanol (or equivalent of 5 or 6 billion gallons of gasoline). What are water quality and ecosystem impacts from that (pesticides, fertilizers, soil erosion etc.)? What is equivalent area of oil sands extraction (surface mining) and related water impacts that would provide an equivalent amount of fuel over 30 years, 60 years?? Then maybe compare that with ANWR’s suggested ‘2,000 acre’ footprint and related impacts. Think EPA will try this soon.
Does your 6/1 ratio for light sweet crude include transportation? Or is transportation negligible?
Hello Jerry,
The energy return has a margin of error of +/- 1 or so. It is based on a worldwide oil extraction ratio of 17/1 (a reference from Charlie Hall). In places it is much higher, and in places lower. The refining step is more consistent: 12/1 to 15/1 for light sweet crude down to 10/1 or so for heavy sour crudes.
The transportation energy is definitely included in the crude production step, but maybe not for getting the gasoline from refiner to gas station. But that piece would be down in the fractions. It wouldn’t materially impact the 6/1 ratio.
Cheers, RR
The correct way to evaluate the environmental impact of energy technologies is using LCA per ISO 14000.
Hello Kit,
I have been involved in the development of several LCAs. If I was being paid to do this blog, and I only had to put out a post a month, then I might take your advice and do an LCA for every energy technology. Of course nobody would read my blog, and even then people would fight about various assumptions in the LCA. But the purpose of this blog is discussion, not to put out boring LCAs for a specific technical audience. That’s part of my real job. But regular people don’t read LCAs.
EROEI is popular Monday morning quarterbacks, bloggers, and college professors at community colleges.
While I can certainly appreciate your attempt at derision, the concept of EROEI was invented by (and is still heavily used by) Professor Charlie Hall of SUNY. The concept is very useful in context. Just because you may not understand the context, don’t presume that nobody else does.
LCA would include energy to be sure but should include other significant factors such as water pollution.
If the title of the essay was “The Environmental Impact of Tar Sands” you would be correct; the article would have been very lacking. It would have also taken me a month to write, when the objective was to simply provide a reference for a frequently asked question. Perhaps you can take up challenge of doing the LCAs and post them on your blog? Personally, I don’t expect to see people providing something like LCAs as a free service to the world. They take a heck of a lot of work (and are still riddled with assumptions), and I have an actual job. I have written elsewhere about various other environmental impacts – including the water usage/pollution question of tar sands – but once again I won’t be providing LCAs for free. If that’s what you are looking for, then you are at the wrong place.
RR
Even just on the water side, it would be interesting to evaluate oil sands vs corn ethanol.
That would be an interesting evaluation. I know that the water usage from oil sands operations is pretty intensive, but then the produced water is also a lot dirtier and more difficult to deal with than the effluent from a corn ethanol plant.
Cheers, RR
OPTI/Nexen’s Long Lake project will use essentially zero external energy. They gasify part of the bitumen for process energy and hydrogen. They project something like 58k net bpd of premium crude out of 72k bpd bitumen, so 14k is consumed in-process. Ignoring the electricity they export, that comes to roughly 4:1 before refining.
Yes, an LCA takes care of EROEI as well as pollutants and also resource depletion. So is a good standards, as far as single assessments go.
I wonder if it would not make more sense to use nuclear cogeneration for ethanol thermal inputs (eg distillation). Simply site the ethanol plant within a few miles of a nuclear powerplant, and build an insulated pipeline from the nuke to the ethanol plant. Perhaps decision makers and utilities can keep this in mind, so as to build their new nuclear projects within major feedstock areas.