Random Thoughts from the High Desert
by Sig Silber
EROEI = Usable Acquired Energy / Energy Expended
It seems so simple. If the amount of energy produced relative to the amount of energy utilized in producing that energy tends to decline, at some point as the ratio approaches 1.0 (or perhaps even becomes a fraction less than 1.0) there is little if any return on the energy invested and society will collapse. But is this concept really workable and useful?
There are many issues related to how this ratio (sometimes abbreviated as EROI) is calculated. This affects both the numerator and the denominator of the ratio. The first problem is that this equation is usually interpreted as being the useful acquired energy divided by the useful energy expended.
Energy expended is usually limited to something you would miss if it became unavailable. This means you do not count the energy from the sun that is used to make energy since sunshine is not a scarce resource. Thus solar insolation in the creation of biofuels via photosynthesis is not included. In some cases the decision to include or not include may be controversial. If you utilize natural gas that would otherwise be flared or not even extracted, should that energy be included in this calculation? This becomes very important when evaluating Canadian Tar Sands.
Similarly in the numerator, what constitutes “useful energy”? Is it useful if it is not at the location where it is to be used? Is it useful if it is produced at a time when it is unable to be utilized? Waste energy produced, e.g. heat, is not included even though it is available to be utilized as either economics or natural evolution of engineering progresses. A good example of this is the progress towards using flare gas as an energy source for oil and gas operations.
This is one of the reasons that for every source of energy, there are many different calculations of the EROEI.
In some cases, EROEI calculations ignore factors that determine the quality of an energy source and in other cases elaborate steps are taken to convert to units of equivalent quality. Converting all energy inputs to common energy units using only heat equivalents assumes implicitly that a joule of oil is of the same quality as a joule of coal or a joule of electricity. Since this is clearly not the case, some estimates of EROEI account for differences of energy quality within EROI analysis when this is possible. This means adjusting the thermal equivalents to economic value taking into account attributes unique to each fuel such as scarcity, capacity to do useful work, energy density, distance to place of intended use, amenability to storage, safety, flexibility of use, negative impacts on the environment and so on. One of the most important of these is the preference for liquid fuels and the flexibility of electricity when the quantity able to be provided is responsive to levels of demand. So without a quality adjustment, two sources with the same EROEI might appear to be equally attractive or the one with the higher EROEI might be considered more attractive than one with a lower EROEI but one might result in a liquid fuel such as gasoline and the other in something else such as intermittent electricity. The quality adjustment is intended to convert apples and oranges into something comparable but involves adjustments that might be considered subjective.
Both energy produced and energy required to produce the output present challenges with respect to how far in the value added chain one looks. The more obvious problem relates to the energy in the denominator but there are similar issues that apply to the numerator. This can be thought of as a number of different levels in the value added chain such as:
- Direct inputs at the source of where the energy is produced
- Inputs at various stages of upgrading the energy utilized into a form that is usable which is similar to the energy quality issue that applies to the numerator.
- Energy inputs required to produce the equipment used in the production of the energy at the source and everywhere along the way in improving the quality of the output including the distribution network.
- Energy required to maintain the labor force from transportation costs to housing and food.
All of these tend to reduce the EROEI. The adjustments to the numerator can be considered the quality adjustments. Obviously no two researchers are going to come up with the same estimates of the EROEI.
Here is one approach suggested by Murphy et al in their article “Order from Chaos: A preliminary Protocol for Determining the EROI of Fuels” which may be one of the most accepted if not the most accepted resource for those who find merit in this approach.
Table 1. Two-dimensional framework for EROI analysis. The system boundaries, which determine the energy produced from a process (i.e., the numerator of an EROI calculation) are across the top, while the boundaries that determine the energy inputs (i.e., the denominator of an EROI calculation) are listed down the left. The shaded cells represent those with boundaries that favor economic input-output analysis while the other cells favor process-based analysis.
The subscripts are really simply a reference to the cell identification. Notice the author recommends calculations include direct and indirect energy and material inputs but be limited to extraction (Cell Row 2 Column 1 with the abbreviation stnd for standard). I certainly do not agree with this choice as I think it is too limited a boundary, but if any such choice were adopted it would make the comparison of EROEI estimates a lot more comparable. To some considerable extent I believe it is commonly used but not always. So one has to be careful to understand how EROEI calculations have been made.
EROEI is problematical if it does not take into account the time sequence of inputs versus outputs. Many energy inputs occur prior to production and during the reclamation phase. The energy outputs in many cases follow a decline curve but not always as a coal mine works differently than an oil or gas well. To convert to present value one has to decide on a discount rate. That is always challenging and would likely make estimates from different sources incompatible.
In many cases, there is some leeway in substituting energy sources in the denominator and these substitutions are based on many considerations one of which is the relative price of the alternative energy sources. But prices and relative prices vary over time. So this can lead to a lack of stability of EROEI calculations.
Relationship to Net Energy Gain
EROEI and Net Energy Gain are two formulas that look at the math in a slightly different way. Net energy gain describes the amounts, while EROEI measures the ratio of the amounts. You can convert from one to the other by considering that:
Net Energy Gain = Energy Produced – Energy Used.
EROEI = Net Energy Gain/ Energy Used + 1
For example given a process with energy produced being 10 units, expending 2 units of energy yields a net energy gain of 8 units with an EROEI of 5. Although equivalent to EROEI, it seems that EROEI has more appeal and appears more often in the literature.
Typical current EROEI values
The following table comes from two sources as indicated. It is illustrative of the differences and sometimes similarities in the values of EROEI calculated by various authors. Although it is not always shown in this table, in some cases you can see trends over time.
Here is another set of estimates from the open source document “EROI of different fuels and the implications for society” Charles A.S. Hall, Jessica G. Lambert, Stephen B. Balogh some of which are reflected in the above table since those who work using EROEI or EROI are a small somewhat close-knit group.
Even if the absolute value of EROEI is difficult to calculate in a non-controversial way the trend in the values over time as certain energy sources deplete and others experience improved ratios due to innovation may indeed be very useful. The following are also taken from EROI of different fuels and the implications for societyCharles A.S. Hall, Jessica G. Lambert, Stephen B. Balogh
Time series analyses of oil and gas production within the US including several relevant “oil related” historical events. Each analysis demonstrates a pattern of general increase then decline in EROI with an additional impact of increased exploration/drilling.
Gagnon et al. (2009) estimated the EROI for global publicly traded oil and gas. Their analysis found that EROI had declined by nearly 50% in the last decade and a half. New technology and production methods (deep water and horizontal drilling) are maintaining production but appear insufficient to counter the decline in EROI of conventional oil and gas.
Time series data on EROI for oil and gas for Norway, Mexico and the Daqing oil field in China based on several papers published in the 2011 special issue of the journal Sustainability and works in progress.
Two published studies on the EROI of dry (not associated with oil) natural gas: Sell et al. (2011) examined tight natural gas deposits in western Pennsylvania in the US, and Freise (2011) analyzed all convention natural gas wells in western Canada.
EROEI Compared to Traditional Economic Analysis
Profit has traditionally been the standard for assessing technologies. If you can make a profit producing an energy source, it has a place in the market. Some feel that EROEI is only a partial measure of potential profitability and is redundant to traditional methods. In many ways the adjustments that are or can be made to the EROEI in its calculation are attempts at including the factors that would be included in a standard analysis of the competitiveness of a given energy source.
So one is then inclined to come to the conclusion that EROEI has its most utility in terms of:
- Screening energy sources
- Monitoring energy sources over time.
The case of corn ethanol illustrates the above. On an EROEI basis, corn ethanol would appear to be a losing proposition as essentially all the EROEI estimates of ethanol production from corn in the U.S. cluster around 1.0. From an energy production perspective it is a wash.
One can identify many negatives. It uses land that could be used to produce food and it uses water often from non-renewable sources such as aquifers with limited recharge.
There are some pluses. The oxygenation argument for ethanol seems to have been debunked but there may be some merit to the argument that ethanol allows refineries to produce gasoline with an octane rating of 87 (which appears to be the minimum allowed for regular grade gasoline at most altitudes) more economically by producing 83 octane gasoline and blending it up to 87 with ethanol. But I am not sure if this is based on refinery economics or tax credits for use of ethanol. Ethanol has created a thriving farming sector and corn processing sector during a period of time when job creation has been difficult.
So it seems that the EROEI has played no role in the decision to expand corn ethanol production and continue to produce it at high levels. At one time there may have been a national security issue as corn ethanol reduced imports of crude oil from unstable sources but that argument is declining in significance as shale oil and gas and Canadian production of hydrocarbon products has increased. It is now difficult to reduce corn ethanol production without causing a lot of economic disruption.
At the other extreme, coal has the highest EROEI. In theory the environmental negatives of coal could be incorporated into the EROEI calculation. But that does not seem to be what has happened. So decisions on attempting to switch away from coal have also been totally unrelated to its high calculated EROEI.
Renewables especially those associated with solar tend to have low values of EROEI and suffer from the reality that the infrastructure required to produce solar energy utilizes higher EROEI energy sources.
I am forced to conclude that EROEI is not at this point in time a very useful measure for either an entrepreneur or public policy. However, EROEI may have more merit than Club of Rome and other NeoMalthusian approaches since it is based on the change in practical availability of energy resources rather than a concept that finiteness means that per capita availability must decline as population increases.
Relationship to GDP
Extracting from Hall et al. the ratio of energy costs to GDP tend to vary within a range of 5% to 10% with the lower end correlated with economic expansion and the higher end related to deficiencies in aggregate demand due to diversion of purchasing power into energy and with rapid increases in this ratio creating shocks to the system. I have not verified that relationship but it makes sense intuitively but that still does not provide a link from either EROEI to GDP or EROEI trends to GDP growth.
I may simply not have come across a study on this or such a study may not exist. And yet that is the key question. To what extent does the EROEI values or the trend help us predict economic growth, standard of living, or other economic measures?
Intuitively one would expect that low or declining EROEI is negative for the economy. But what is intuitively obvious may not turn out to be confirmed by the data. We know that the ratio of GDP growth to energy use has been declining for various reasons including improved energy efficiency but probably more importantly the growth of service sectors where energy intensity tends to be low.
So it could well be that the declining contribution of energy to GDP may be occurring more rapidly than the rate of decline of EROEI. Again I do not know if that relationship has been studied and reported. I have as yet not found such a study but I have not made an exhaustive search. So at this point I see this question as remaining at least for me unanswered. That would appear to be a key question to be answered if we are to place a lot of value on EROEI current values and trend.
I suspect that EROEI is less useful than the proponents believe but it is not something to be ignored. I am not an energy expert but I have a background in hard rock mining and this reminds me of the issue of declining ore grades and increasing stripping ratios. In theory those two factors should lead to a real rise in metal prices but they have not. Technology for extracting copper as an example has advanced more rapidly than copper resource quality has declined plus the miniaturization of electronic devices has reduced the amount of copper required per unit of GDP.
We may have the same situation with respect to energy resources.