NUMERAIRE BLUES.
Robert Herendeen
Illinois Natural History Survey
217/244-2137
herendee@uiuc.edu
7 November 2005
Consider this contradiction:
You buy a gallon of gasoline for $2.00. Let’s say the producer’s price is $1.25/gal. The energy intensity for “refined petroleum products” from EIOLCA is 25,800 Btu/$, so the total direct and indirect energy impact is 25,800*1.25 = 32,250 Btu. Accounting for the energy impact of the $0.75/gal transportation and trade margins will only add ca 10,000 Btu, for a total of ca 40,000-50,000 Btu/gal. But you know that the heat content of gasoline is ca.125,000Btu /gal., so there is a major mismatch here. What is going on?
In general, possible contributing factors are:
1. Aggregation
2. Rapid price change
3. Converting from producers’ to purchasers’ prices
4. Difficulties from using dollar flows as surrogates for energy flows, an example of the general “numeraire problem”.
These factors are not totally independent. In particular, 1 and 4 interact, and if enough disaggregation is done, 4 may cease to be a problem.
Here is a possible scenario of how factor 4 could cause trouble. Gasoline is just one product under the aggregated category “refined petroleum products.” If gasoline were much more expensive (measured in $/Btu) than the average refinery product, then its apparent energy intensity would be less than the average for all refined petroleum products. In actuality, gasoline is more expensive than the average, so this is a potential explanation.
As I have described the problem here, I have assumed that the energy out of the ground is “counted” in the EIOLCA intensities; the problem is just allocation. However, in EIOLCA I am not sure that the energy out of the ground has been included anywhere!
Below I describe how I have adjusted things to patch up this problem; I address the question of whether energy out of the ground has been accounted for properly anywhere in EIOLCA. Before that, however, I point out that this problem was addressed directly in 1975 (Bullard, C. W., and R. Herendeen. 1975. Energy costs of goods and services. Energy Policy 3:263278.) A mixed approach was used, in which outputs of energy sectors are expressed in energy rather than dollars in the direct requirements matrix. Then the mixed matrix is inverted. The energy intensities of energy commodities then are measured in Btu/Btu. Compared with EIOLCA, this approach requires:
1. Additional data on energy flows between energy sectors.
2. That one invert the matrix oneself, rather than merely using the existing inverse from BEA.
A POSSIBLE WAY TO APPROACH FACTOR 4 WITHOUT USING THE MIXED APPROACH.
If the EIOLCA energy intensities do not include the energy from the ground, then the total energy intensity (Btu/$producer) would be
eeiolca +1/pproducer ,
where pproducer is the producer’s price in $/Btu. This would mean the physical energy intensity (= ephysical, measured in Btu/Btu) is
eeiolca*epoducer + 1.
This is always >=1, which removes the concern above.
I have applied thinking like this to coal, refined petroleum, electricity , and natural gas, but it did not seem to give reasonable results in all cases. Electricity has the problem associated with the thermal efficiency of thermal power plants, and coal and refined petroleum also have the problem of converting from producers’ to purchasers’ prices.
EIOLCA is not explicit about how these issues are handled, and I am reluctant t second-guess them; I think I have said enough here about the problem(s) to indicate a call for more clarity. If researchers want to discuss details, they should feel free to contact me.