By Chris Nelder
Posting in Design
Energy futurist Chris Nelder explains why you can run from EROI, but you can't hide, and why savvy energy investors should use it as a guide instead of ROI.
One of the key deficiencies of "unconventional" fuels is their low energy return on investment relative to conventional fuels. Many analysts have ignored this factor because investment decisions are made on the basis of the financial, not energy, return on investment. But a growing literature suggests that the two are intimately related.
But before I get into that, a quick note on terminology. The financial return on investment is known as ROI. The analogue in energy, the energy return on investment or EROI (also expressed as EROEI, for "energy return on energy invested") is a ratio of the energy produced to the energy invested in its production. Some, including me, have also referred to EROI as "net energy," but that really confuses the terms. For parallelism with the language of finance, net energy should refer to energy produced minus energy invested, whereas EROI should refer to energy produced divided by energy invested.
How EROI begets ROI
The relationship between ROI and EROI is actually very simple and logical. The more energy you have to invest to produce a fuel, the lower your EROI will be. The energy you invest has a cost. Therefore, the profit on the same barrel of oil will be higher when it's produced from a high EROI source than when produced from a low EROI source.
This simple concept gets lost, however, in the complex accounting of fuels in the real world. The financial return on all unconventional fuels is distorted in one fashion or another by subsidies designed to encourage new development, debt acquired to finance the projects, and complex accounting of the investments and returns. For example, as I discussed previously, the accounting methods used in shale gas development allow operators to roll over gains and losses creatively and amortize them across older and newer wells, wet and dry wells alike. Initial development costs tend to be intermixed with long-term operational and maintenance costs, debt servicing expenses, and so on. Initial exploration costs and even production itself can be offset by tax credits. Ultimately, the profitability of production tends to resemble a picture of cash flow more than pure ROI, and the EROI of some fuels becomes very murky indeed.
Corn ethanol offers a fine example of the problem. More than $20 billion in subsidies over the past three decades have ultimately turned nearly 40 percent of the U.S. corn crop into less than 10 percent of the country's fuel needs by volume, and less than 7 percent by energy content. In 2009, the U.S. taxpayer subsidized 75 percent of the price of each gallon of gasoline replaced with ethanol. It has proven to be an expensive way to make a low-quality fuel (ethanol has about two-thirds the energy content of gasoline) which reaches its scaling limit at a fairly low level.
Careful observers who did the math on the EROI of corn ethanol knew it would run into cost and scalability limitations literally decades before legislators and investors did. With a generally accepted EROI of around 1.4 (also variously estimated between 0.8 and 1.6), it was just barely a net energy-positive fuel at best. In the pithy observation of veteran energy analyst Robert Hirsch six years ago, making ethanol from corn is a process in which a certain amount of energy in the forms of natural gas and diesel fuel are used to create an equivalent amount of energy in the form of ethanol, with the primary output being money from government subsidies (not to mention soil erosion). Such a low EROI would imply a low profit margin, thin enough to be swamped by the volatility of both corn and oil prices, as indeed it was in recent years. However, only the ROI, in the form of increased "energy independence," was taken into consideration in the politically-motivated push for biofuels.
With the tax credit finally expiring at the end of 2011, we should now see the real costs of producing corn ethanol begin to be priced in to the cost of gasoline. Its EROI has been "hidden away in the attic like a crazy aunt," as my friend Gregor Macdonald quipped to me this week. Without subsidies, the ROI of corn ethanol must begin to converge upon its EROI.
The EROI tipping point
A small cadre of academic researchers have calculated the EROI of various fuels and explored their relationship with the economy—most notably, Charles Hall, Cutler Cleveland, Robert Kaufmann, David Murphy, David Pimentel, Robert Costanza, Carey King, and Adam Brandt—and the body of research is growing rapidly. Some of their recent papers have attempted to describe in mathematical terms how EROI translates directly into price and profitability, and how it can inform policy.
King and Hall (2011) established some critically important principles:
- As EROI decreases, price increases.
- EROI implies both profitability, and a price limit. A few examples: At $61 a barrel, oil production can be profitable at an EROI of 5 but not at 2. When EROI is less than 10, natural gas prices must be above $6 per thousand cubic feet to be profitable. A realistic EROI of 3 to 4 for tar sands implies a price of at least $50 to be profitable.
- The decreasing net energy and increasing capital intensity of our energy production have likely contributed to the economic downturn.
- When EROI remains above 10, the relationship between prices and EROI is fairly linear and steady. But when EROI falls below 10, it can force prices to increase at a dramatic and nonlinear rate, to much higher absolute levels.
The latter point is key. One example King and Hall offer is that a 60 percent drop in EROI from 25 to 10 resulted in a 150 percent increase in oil prices, from $19 a barrel to $48. But a 60 percent drop in EROI from 5 to 2, likewise with a 150 percent price increase, causes the price of a barrel to jump from $96 to $240.
Heun and de Wit (2012) found similar results: EROI and producer prices are indirectly and inversely connected, and while the correlation is weak while EROI is over 10, prices can still be projected from EROI trends.
Declining EROI has a "nearly inconsequential" effect on prices until it reaches about 18, then has an increasing effect until EROI falls below 10, when prices jump dramatically.
This finding meshes nicely with the "net energy cliff" model proposed by geologist Euan Mearns, which shows an exponential decline in the energy available to society as EROI falls below 10:
And this should send a chill up your spine, because the EROI of domestic U.S. oil production is now approaching 10, having fallen from around 100 in the early days of oil (Cleveland, 2005). Even in the few prospects where we can still drill a well that will produce over 100,000 barrels of oil per day, like the deepwater Gulf of Mexico, the EROI varies from 4 to 14 (Moerschbaecher, 2012).
Hall and Murphy have also found that a given fuel must have an EROI of at least 3 to deliver a net benefit to society because of the associated infrastructure needed to support and use the fuel, and that an overall EROI of at least 10 may be required to sustain a complex society. It takes a significant energy surplus to support things like higher education, entertainment, personal vehicles, a middle class with health care, outsized amounts of credit, and yes, subsidies for low-EROI fuels.
You can run from EROI, but you can't hide
All of these studies come to a similar conclusion: As we continue to substitute unconventional fuels for conventional fuels and the overall EROI falls below 10, it's going to be very difficult, if not impossible, to continue running our complex society. Prices will go too high for the economy to tolerate and kill demand before unconventional substitutes can scale up to replace declining higher-EROI fuels. Biofuels can't do it; tar sands can't do it; oil shale (with an EROI between 1 and 2.5) can't do it. There is still too little work on the EROIs of shale gas and shale oil (like that produced from the Bakken Formation) to know how far they can take us, but given the growing acceptance of the notion that they can sustain us for decades, the need for academic inquiry is urgent.
In the words of Heun and de Wit, "There are not perfect and scalable substitutes for oil at the present time." In so many ways, conventional oil is special. We are losing the race between oil depletion and the pursuit of substitutes and better extraction technology. Drilling technology cannot overcome depletion, because depletion is giving us declining EROI and intolerably high prices for substitutes. The researchers conclude that a smooth transition away from oil is unlikely without a deliberate policy effort to steer us toward alternative energy sources and manage the economic effects of depletion.
None of these insights should be particularly startling, yet most observers and policymakers continue to miss them entirely. Bedazzled by the sheer magnitude of unconventional resources—trillions of barrels of oil equivalent!— they cannot see how the low energy return of some (not all) of those resources will ultimately force us to leave them in the ground as their cost of production proves intolerable.
ROI cannot escape EROI forever. The magic of credit, government subsidies and creative accounting can forestall the recognition of ROI for awhile, but eventually the true cost of producing energy, both in dollar and energetic terms, becomes a limit on production. As ROI converges with EROI and the profit picture worsens, investors start bailing out and production falls.
In short: EROI ultimately determines ROI, but because investors and policymakers don't realize it until late in the development cycle when ROI finally proves weak, we continue to invest in fuels with poor EROI. To navigate the future of energy, investors should be looking to EROI, not ROI, as a guide.
All researchers on EROI and economics point up the need for additional studies, particularly on shale gas, oil shale, and effects on the economy as a whole. The existing body of work suggests that as EROI falls, disposable income does too and leads to recession. To prove it, we need models that include both the economic effects of resource substitution and the geological effects of depletion across the entire energy sector. I only hope that the academy is up to the task, because so far, our official energy agencies have failed to build any such models. We are simply drifting ever closer toward the net energy cliff, in blissful ignorance of EROI limits, with visions of unconventional resources dancing in our heads.
Photo: The business model of the underpants gnomes, from the instant-classic South Park episode.
Feb 14, 2012
Wonderful article. However, I have a querry. Is there any academic inquiry carried out on the EROI and ROI on energy from wastes?
Your ERoEI chart appears to be ah, slightly biased: http://en.wikipedia.org/wiki/Energy_returned_on_energy_invested
EROI being a single number derived from a lot of data, doesn't reflect everything you need to know about the energetics of a particular technology. One critical factor is time - not only how much energy needs to be spent, but also WHEN it has to be spent compared to when you get your Energy Return. Ten years or so before a nuclear plant is commissioned, you have to build a large office and fill it with planners, engineers, lawyers, etc, desks, computers, lights, air-conditioning, etc. Some years later when construction starts, you need to buy lots of concrete and steel (both energy-intensive materials), and the usual load of trucks and bulldozers and construction workers (with their cars). As construction gets more advanced, the need for copper, aluminium and a whole host of exotic materials comes to the fore. Meanwhile somebody has to be digging up uranium ore and processing it into fuel rods, and that means finding twice as much Zirconium as Uranium for the tubes, and removing the Hafnium contamination from it. Eventually you get your reactor running and it produces a lot of energy, but it will still takes X years to repay the energy used to build it, so some of that energy has been spent (10+X) years before it gets repaid. Only then does the nuclear plant turn into an energy-positive project. But in the meantime, you have had to make a start on several other nukes, and they will still be in their energy-negative phase, so the industry as a whole is still energy-negative. And if you go on building more new plants, as you would have to do to scale up, there is a danger that you will NEVER reach the industry-wide energy-positive stage, before the general lack of energy collapses the economy. This applies whatever the technology, but affects low EROI and "long building phase" technologies most. More at www.peakoil.org.au/news/index.php?energy_profit.htm
I am not disputing the relationship of ERoEI to ROI. What I question is its basic relevance in judging solar and wind devices. These devices are extensions of the fossil fuel supply system. There is a vast infrastructure of mining, complex processing, massive machines, large manufacturing facilities, multiply means of transportation as well as the regular accounted ERoEI inputs. The actual direct energy contribution of this infrastructure may by small but the devices do not get made without it. How are they factored in? This makes these devices questionably renewable, alternative, clean, green or even sustainable. See: A story in pictures and diagrams: From Machines making machines making machines http://sunweber.blogspot.com/2011/12/machines-making-machines-making.html
I have not seen EROI reported for oil shale at 1, although I have discussed Adam Brandt's values for two type of oil shale system with him extensively. I would suggest that the uncertainty is far larger on the high end, given values reported by technology developers at annual oil shale symposia which I co-chair. While these have not yet been as rigorously tested, they provide suggestions that the EROI could be substantially enhanced once real systems are working. If only the values reported by a select few researchers are to be allowed, then perhaps the cited range is adequate. I will certainly be delving deeper into the literature, especially regarding the feasible EROI cutoff for the traditional fuels, as well as trying to understand the potential EROI for electric vehicles, which perhaps you have already commented upon.
If energy prices do spike at this ratio, the just-in-time supply chains that maintain energy extraction, refining and distribution themselves will no longer be sustainable. Full stop. And rather sudden, I should think.
Thanks Chris, this gibes with and expands on some stuff I've heard before. But this is only about pipeline fuels, liquid and gas. To see the full picture don't we need to include other forms of energy including coal, nuclear, solar PV, solar thermal, wind, hydro of various types, geothermal and so on? I guess the electricity producing fuels. Twenty or thirty years down the road I believe transportation will increasingly be powered by electricity. All we need is the battery technology to make it happen (another idea for an article).
i'm sorry. i missed the externalities part of the analysis. This the same economic stupidity that has brought us CO2 levels, warming and weather trends that re frightening, and filty water, air, and the resulting illness. whoever suggested the atricle needs to go back to school.
Really good piece. One way to equilibrate market prices with EROI is end all the overt subsidies and clean up the tax code with a fewer options on depreciation. An overlooked fact is all the other distillates that come from a barrel of conventional oil. The refining companies need high gas consumption to subsidize the manufacture and sale of other distillates. If people drove less or even not at all, then gasoline would be a waste product from the production of all the other distillates. Thus gasoline consumption distorts the cost of the other distillates. The interesting implication of this is that government subsidized free roadways are a net subsidy to the production of other distillates. If roads were metered (tolled) per their actual usage by individual drivers, people would drive less and subsidize the other distillates less.
Thanks for the explaination and the charts to help visualize what is going on. Oil is used in a lot of things beyond fuel. There are the 6 or so gallons of oil to make a tire, use as fertilizer, cosmetic products, plastics and lubricants. There are replacement materials for these things but the cost may be higher and the quality will be different. It will be an interesting future with oil prices going higher and a frantic search for a substitute to make viable products that have high demand because of the low price of oil in the past made those things viable.
This is a pretty good post. There are a few key elements missing. With ethanol, almost everyone (including you) fails to recognize that the corn distilliation process does not produce ethanol exclusively. When corn is turned into ethanol it also produces distillers grain as a by-product. What's distillers grain used for? You can feed it to cows. What's corn used for, if not distilled into ethanol? In america, we feed it to cows. So, in the insane "feed corn to cows" world, the alternate strategy of "feed corn to cars and cows" is almost surely less insane (i.e. a net win). Strangely, the ethanol haters like to assume the distillers grain is thrown away, or that corn, if not turned to ethanol, is used to make corn totillas for poor people. Neither is even remotely true. The other missed point with oil sands is that it energy input is largely natural gas, and not the oil itself. (The processes vary). The input energy could very well be nuclear power, or even wind power, for that matter. A large part of the input energy is creating steam, which can be created with all sorts of ways. So the EROI for oil sands probably needs to be more holistic and ask where the steam energy is sourced from. If shale gas is very high EROI, or if even higher nuke EROI is used, then effective EROI of oil sands surely improves. (Think about it - if the oil sands operation was located next to the a massive hydro dam (say dating from a depression era government boondoogle) that was kicking out huge quantities of dirt cheap electricity, the price break even point for oil sands would fall by quite a bit, and the EROI would be silly without treating the hydro electricity as borderline free. This extreme thought experiment demonstrates why we need to back up a step and ask what the EROI is of the energy source that is creating the steam),
Chris, I can't say that I fully get all the mathiness here, but the implications are clear as a bell. A fantastic leap of communication that I hope millions eventually see. Thanks!
The impact on the quality of life in the US from the food for fuel mandate needs to be factored into why the corn for ethanol mandate has been a gross failure of our leadership. Bush started this mess and Obama who had the opportunity to kill it, but failed too because the people making money off it are on his list of finacial supporters that got him in office. By funneling nearly 40% of the US corn crop into this black hole we have seen food prices on corn connected food products, bread, beef, chicken, etc.. go up between 30 and 50% in the last 3 years. The loss of corn as an animal feed stock has had a ripple effect on the price of other feed stock commodities feeding the broader rise in food prices. For example dairy, rye, and soy based products.
That is good information about the lengthy time from proposal to building the reactor. You should also include the decommisioning costs as well as the decades it takes to break down the reactor infrastructures. The start up costs and the decommisioning costs pretty much wipes out the cost effectiveness of nuclear power. It is a shame that nuclear power never lived up to the hype "Too cheap to meter" that was part of the sales appeal of nuclear power. Nuclear power will probably be in the mix for future energy needs. There are other types of reactors being considered, perhaps one of these would be cost efficient.
It's true that producing solar and wind devices does require material and energy input that is currently largely supplied by fossil fuels but once they are in place, other than normal maintenance, they don't require additional fuel input to produce energy. And potentially much the energy required for their production could be supplied by already built solar/wind. I wonder how much the fact that solar and wind have zero fuel cost affects their EROI.
Jeremy, I'd be very interested in seeing any data you may have on the EROI of oil shale. I am working through some of Brandt's papers. Please contact me through the contact form here http://www.smartplanet.com/search?q=chris+nelder
Possible, but I tend toward the "narrow ledge" view I've written about previously. We could teeter at the edge of an EROI of 10 for years, as emerging economies slowly outbid the OECD for free-market oil. But eventually--I can't say when--high prices brought on by EROIs below that threshold will kill their demand, too.
Quite so, Riverat1. That's a topic for another article, but Hall & Day 2009 had a handy chart showing the EROI of other fuels, here: http://www.condition.org/as95-6.htm
"There are the 6 or so gallons of oil to make a tire, use as fertilizer, cosmetic products, plastics and lubricants." Your comment is a bit misleading. Rifing oil is done through a process called "distillation". When you distil crude oil, you get the distillate, such as gasoline, diesel, jet fuel, etc, and the precipitate, such as plastic, asphalt, etc. The business of oil companies is the distillate. For them, the precipitate is a waste material. All the items you mention are from the distillate. If they were not recycled by other industries, they'd end up in land fills. The items you mention are precipitate, not distillate.
Knowing they have eliminated the competition, raw corn, from the feed market the distillers have been charging a pemium price for this allgeded waste product. Higher feed prices have forced many beef operations to reduce herd sizes to the point this years herd is at 60 year lows. The reduced herd size has reduced demand for distiller grain for feed to nearly half the of supply. Yet the prices of distillers grain do not reflect the glut. There is also a problem with the added cost of starch supliments for dairy cows as distiller grain as a feed by itself has a negative impact on the quality of milk verses corn feed. There is also emerging science indicating a higher incidence of E. Coli among healthy herds being fed distillers grain. This leads to increased beef processing costs. A physical limitation of a rail car shortage is also disrupting the distillers market as trains are doing double duty bringing corn to ethanol distillers and the waste out to feed processors. The lack of transportation increases the spoilage rate of this waste product. http://www.ethanolproducer.com/articles/4165/distillers-grains-symposium-to-address-feed-rations-animal-health/
Well, there was an across the board glut in commodity prices 10 years ago ... almost everything has risen over the last 10 years (with the exception of domestic natural gas). That said, Pork is cheaper than it was 10 years ago. I'm not sure if replacing beef with pork will represent such a terrible hardship for most people.
...using oil for things that do not require the oil distillate to be carried from place to place. Our vehicles carry both gasoline and a heat engine, neither of which is particularly light and for safety we place ALL these things as well as ourselves inside a steel body when the goal was to simply be somewhere else only to turn around and go home again. This is crazier then growing a lawn that you have to mow to enjoy.
Defining the boundaries of EROI analysis is difficult: how far back do you go when adding up the energy required to produce a form of energy? You can easily end up with a deep regression, making analysis impossible. Likewise the time of use issue. Dave raises some important points here, although staffing a planning office, for example, falls outside the typical boundaries. The authors of EROI research papers are usually careful to identify their boundaries of analysis, and readers should note those parameters carefully.
PV solar has a very poor EROI, the next worst after ethanol. This is due to the energy needed to refine the silicon to ultra-pure standards. The ultra-clean, air-conditioned factory that makes those panels has to be built first (with fossil fuels) and it will operate for maybe 30 years if you're lucky. So some of the Energy Input must wait for 30+ years to be repaid with Energy Output. Wind turbines in windy places do much better, but even then they have limited lives. A wind farm on Hawaii ran for 20 years and then needed a major refit, but they scrapped it and built another one instead. This sort of thing is only possible in an energy-plentiful economy. When we change to an energy-scarce economy, these projects will be un-doable.
That assumes that all driving is discretionary, when in fact it's not. People have to get to work, and if they live in the suburbs they are unlikely to be serviced by public transport, so driving is effectively compulsory. Commercial vehicles, likewise, cannot just choose to use a bit less, because the amount of transporting of goods is directly related to profit. If a business can't make a profit, it goes out of business, and THAT is where you will see the reduction in consumption coming from - mass lay-offs. During the price rises of early 2007 to mid-2008, the price of WTI Crude Oil rose 72%, and consumption remained flat (instead of growing at 1.6% like the previous 20 years). Not much price feedback there.
Electric cars don't run on electricity, they run on the fossil fuels that fired the electricity generating plants. There are the usual losses due to inefficiency in this process, plus the losses in transmission, plus the losses in charging the batteries, plus the losses in drawing energy from the batteries, plus the energy cost of lugging the batteries around. And to make things more ridiculous, to overcome the limited mileage of the batteries, hybrids have a gasoline motor and a gas tank to lug around as well !
I knew that plastics are made from what was left over from distilling fuel out of crude but I didn't know that the other products came out of the precipitates. The point I'm trying to make is that the same crude oil is used for more than just fuel and what effects the cost of oil effects all the products derived from crude oil.
That said, I read your link and it certainly doesn't look to me like the distillers grains are, as a whole, failing to find an economical niche. With re: to pricing - all feed products are expensive, no? Relative to corn, distillers grain appears to be cheaper (again, inferring from your link). That said, there are going to be hiccups incorporating distillers grain into the food cycle, but I think it's safe to say food prices would be higher overall if this product was merely discarded instead of marketed for consumption. Distillers grains is not some new fangled frankenfood. Pretty sure its been fed to livestock for a long time. There is some credible evidence that distillers grains are, on a whole, better for cows than corn, because it's more fibrous and less starch. That might be speculative though.
EROI is typically measured at the wellhead, mine mouth, or other point of production. Net energy at the point of use--like in a car--is a whole 'nother topic.
Electric cars run on whatever source of electricity they use. Here is the Pacific Northwest where I live we get a significant amount from hydroelectric and some from wind. Even if it is fossil fuel powered electricity it is probably more efficient than the approximately 20% efficiency of an internal combution engine. If some of the battery technology that's being developed comes to fruit batteries and hybrids won't be an issue.
When a process produces more than one product which is valuable, it complicates the EROI maths, but it can be overcome. Essentially what you do is apportion the Energy Inputs to the various Energy Outputs (co-products) based on the proportions of energy the outputs contain. This is standard practice among EROI mathematicians.
Distillers' grain may still contain enough calories to be usable as feed, but it definitely will not contain as much as the raw corn (the energy in the ethanol has to come from somewhere). You are right that it should be factored into the EROI of corn ethanol, but you would have to think a bit about how you would measure that. One way might be in terms of calories of meat produced divided by calories of grain fed (which is, AFAIK, pretty terrible -- something in the region of 0.08 for beef). If you use that calculation, it barely changes the overall EROI of the ethanol. Of course this raises a whole separate issue -- that most of the US-produced corn not going into ethanol goes into meat production, which is a terribly inefficient way to feed people. But one thing at a time...