At the recent Association for the Study of Peak Oil conference in Italy, there was an excellent presentation given by Charles Hall, entitled "EROI: The Key Variable on Assessing Alternative Energy Futures?" The presentation has many interesting features, but for the purposes of this discussion it is significant because it adds to the growing body of evidence that Energy Return on Investment is rapidly declining. EROI is "Energy Return on Investment," also called "Energy Return on Energy Invested," and measures the ratio of how much energy is required to produce a standardized unit of energy--whether the produced energy comes from Coal, Oil, Hydropower, etc. EROI is notoriously difficult to calculate, but Hall does an excellent job of providing quantifiable and verifiable data to support the decline in EROI--in fact, his presentation gains validity in my eyes because it appears that he concludes that his initial hypothesis (that EROI is not declining) is incorrect based on the data he collected.
In light of the growing evidence that EROI is declining, I would like to make a few observations based on my own take on an Energy Theory of Value.
1. EROI decline exhibits a "lag time." If we accept the assertion that the overall EROI of energy used by the global economy is currently declining (regardless of the rate of decline), a corollary of that assertion is that current energy used to manufacture the equipment necessary to extract or produce future energy enjoys the advantage of prior, higher EROI. 50:1 EROI oil from the 70s or 80s was used to create the factories that created the massive trucks or turbines that produce today's energy. But these massive capital assets have limited life-spans--right now they create a "lag time" where it buoys the apparent EROI for current energy production--but as these assets need replacing they will need to be replaced using lower EROI energy, which will increase the energy cost of replacement and lower the EROI of the energy that the new products will eventually produce.
2. EROI calculations for "alternative" energy sources are distorted by this "lag time." This same process described above also applies to current calculations for the EROI of "alternative" energy sources: PV cells, wind power, ethanol production, etc. This leads me to think that EROI of currently popular alternatives are lower than currently advertised, but more on that next...
3. A monetary ROI (return on investment) calculation using a re-formulated Energy Theory of Value is a more effective way of assessing actual EROI for alternatives, even though this calculation contains distortions due to government market distortions. The Energy Theory of Value, which is along the same lines as Ricardo & Marx's Labor Theory of Value, suggests that the value of any product can be best calculated by looking at the total energy required in its production--tracing all the way back to the energy required to build the tractor to harvest the crop to feed the laborer to build the factory that... etc. What this boils down--assuming that we discount market distortions, which is a big assumption--is that EROI can be calculated based on the purchase price ($) of an energy production asset (which theoretically incorporates all energy required for its production) in comparison to the expected value of energy produced by that asset discounted by the time value of money. This is complicated, but let's look at an example with a PV panel. Let's take the Kyocera 50W PV panel from Gaiam. It costs $310, which is probably high, but is a bargain if we assume all necessary batteries, mountings, and inverters are included in this price (they are not). In Colorado (a good Solar state), we get the equivalent of 6 hours of full sun per day on average all year round, so we get $2191.56 hours per year, meaning that our 50W panel will produce 110 KWh per year. As of March, 2006 (per EIA), the average cost of delivered electricity in the US was 9.8 cents per KWh, so the value of our hypothetical panel's annual production is $10.78. If we assume that the inflation-adjusted time value of money (based on the inflation-adjusted Fed bond) is 3.5% per year, and that the panel will last 40 years (generous), then an energy production of $14.40 per year is necessary to break even on the initial energy outlay. So our generation of $10.78 worth of electricity/year gives us an EROI estimate of 1078:1440, or 0.75:1. I'll be the first to admit that this is only a THEORY on establishing EROI, and it includes many assumptions, but it does permit an empirical calculation of EROI. Because of the various market distortions making energy CHEAPER in the US than it would be without those subsidies, etc., I think that, if anything, this calculation actually gives us a rosy picture of EROI for solar or other alternative sources... **IF** this theory of EROI turns out to be accurate, it presents an extremely pessimistic picture of the potential for alternatives to replace the future decline in oil and gas production.
4. The closed-loop thought experiment to validate the EROI of alternatives. This really boils down to one simple question: why don't ethanol plants use their own product to distill yet more ethanol? Instead, they tend to use natural gas. It would be a great advertising feature to point out that your ethanol plant is powered entirely by ethanol--which is why my inability to find any evidence of an ethanol-powered ethanol plant leads me to the *assumption* that they don't exist. Easy proof that an alternative energy source has at least a superficial EROI of greater than 1 is to use its own product exclusively to power its own operation (I say superficial because this belies the substantial energy invested in the infrastructure manufacture in the first place and only focuses on the ongoing energy required to operate--something more immediately important in the case of PV or wind power where the majority of the energy invested is up front than with Ethanol where there is substantial ongoing enegy requirements).