7. United States Economy
The information detailed in the past chapters demonstrates the difficulties of developing energy-intensity indicators. For sectors such as manufacturing, the choices of energy-intensity indicators are many. For transportation, the choices are few. Data are limited in some sectors and abundant in others. As was demonstrated, every sector, especially transportation, has complexities.
There are even more complexities across sectors. Energy is consumed differently for different reasons. Structural and behavioral effects manifest themselves in different ways. In spite of the difficulties, why is it important to attempt to develop an economy-wide energy-intensity indicator?
Environmental concerns and concerns about energy supply require knowledge as to how well we are doing in reducing the growth of energy consumption for the Nation as a whole. Basically, households, establishments, etc., change consumption patterns that, in turn, affect the amount of energy that flows or is delivered to them. Developing site energy-intensity indicators is appropriate at this sectoral level. Primary energy used to generate energy, including all of the losses in production, transmission and distribution of energy, must be included in the measurement of changes in energy-intensity for the economy as a whole.
This chapter first discusses primary or "total embodied" energy in definitional terms. Included is a discussion of the trends in energy consumption and the losses associated with energy production, transmission and distribution, and how these losses have grown over time. A proposed methodology for the development of economy-wide energy-intensity indicators and the limitations of such indicators are presented next. This presentation is followed by the development of economy-wide energy-intensity indicators using the methodology, including the development of conversion factors to adjust site energy estimates upward to account for losses. Finally, several economy-wide energy-intensity indicators are presented and compared.(86)
Primary energy includes direct use by end users of fossil fuels (petroleum products, natural gas, and coal) and renewable energy (biomass and waste products) and the indirect use of fossil fuels, nuclear, hydropower, geothermal, and other renewable inputs in the form of electricity.
Two economy energy-intensity indicators, energy per capita and energy per Gross Domestic Purchases(87) are shown in index form in terms of primary and site energy (Figure 7.1). Indices of energy per capita rose in the early 1980's and remained higher than indices of energy per Gross Domestic Purchases, as economic growth outpaced population growth during the sustained economic expansion of the 1980's.
The difference between the estimates of primary energy and site energy has grown overtime (Figure 7.1). The indices using primary energy were slightly below their site energy counterparts in the 1970's, but were higher by the mid-1980's. The indirect consumption of energy as the inputs into electricity generation has grown over time as electricity consumption to end users has grown by 47 percent from 1977 to 1993.(88) This suggests increases in aggregate losses associated with time generation, transmission, and the distribution of energy.
Production, transmission, and distribution losses occur among all energy sources, but the magnitude differs widely. The next two sections describe these losses for natural gas and electricity.
Natural Gas. The energy losses in the production of natural gas include losses from repressuring gas into the wells, removing nonhydrocarbon gasses, venting and flaring certain amounts of gas during the extraction process, "extraction losses" or the removal of liquid constituents from the dry gas. Losses also occur in the inter- and intrastate pipeline distribution system. Operation of today's processing and pipeline systems were significantly modernized in the late 1980's through utilization of remote-terminal units, microwave communications, and computerized-control systems.(89) This modernization has reduced losses.
Electricity. Losses in the production of electricity are much higher than those from the production of natural gas. Electricity is generated by several different processes, each using different raw resources and each involving different amounts of energy losses. Most of the energy losses in the generation of electricity occur when heat is converted into mechanical energy for turning electric generators. Other losses include power plant use of electricity and losses due to transmission and distribution of electricity from the power plant to the end user. The system losses do change from year to year due to changes in the mix of inputs used to generate the electricity: coal, natural gas, petroleum products, hydropower, nuclear power, wind, sunlight, and geothermal heat. Therefore, for each year, there is a different multiplication conversion factor.
As demonstrated in the previous sections, primary energy estimates including losses should be used when developing an energy-intensity indicator for the economy as a whole. Additionally, estimating primary energy consumption, sector by sector, provides more insight than simply using total primary energy data for the economy as a whole. Regional and sectoral differences can help explain changes in energy demand.
The first step in the development of economy-wide energy-intensity indicators is to develop energy-consumption estimates for each of the sectors. This methodology may be used to develop primary, unadjusted site, or adjusted site economy-wide energy-intensity indicators.(90)
Next, for each sector, develop sector-specific energy-intensity indicators.
Finally, as a last step, add each of the sector energy-intensity indicators together into one economy-wide composite.
Changes in energy-intensity overtime could be estimated using the economy-wide indicators. This would provide insight into changes in energy efficiency. Three composites can be created, based on primary energy, unadjusted site, and adjusted-site energy.
The next sections discuss each of the individual steps listed above using this methodology to develop such energy-intensity composite indicators for the U.S. economy. The section also discusses obstacles associated with the third step of the methodology and how they may be overcome.
Primary Energy Estimates. Primary conversion factors are calculated. These vary regionally according to the mix of fuels used directly or indirectly for electricity generation. For each of the sectors, these conversion factors convert site-adjusted estimates of energy consumption into primary energy estimates.
Adjusted-Site Estimates. Using a sectoral approach site energy is adjusted by weather, behavioral, or other structural effects that may increase or reduce actual energy consumption.
In the residential sector, site energy is adjusted by heating and cooling fluctuations relative to a normal 30-year average expectation for weather, in order to compare what energy demand might have been had the weather been normal. In this way, excess demand could be explained in terms of unusual weather patterns.
In the commercial sector, site energy is adjusted by both weather and occupancy, taking into account the fact that energy demand is nominal in a vacant building. By eliminating vacant buildings, commercial energy data more realistically reflects actual per building usage.
In the industrial sector, manufacturing value of shipments is adjusted by inventory changes and capacity utilization. By eliminating stock build-ups or depletions in any year, industrial production can be more realistically measured. By considering capacity utilization, periods of low or excessive product demand can help explain variations in energy demand.
In the transportation sector, the demand for passenger travel (vehicle miles traveled) is adjusted by occupancy. A vehicle traveling 100 miles with four passengers represents twice as much demand as a vehicle traveling 100 miles with two passengers.
Using adjusted site energy in the development of energy-intensity indicators as a basis for evaluating changes in energy efficiency is only one concept of energy efficiency based on the effects they adjust for.
Activity in each sector of the economy is measured uniquely. The demand indicators for each sector are parameters that influence energy consumption such as: the number of buildings, operating hours, the number of workers in a building, and the size of a building as measured by its floorspace. For instance, the residential sector measures its activity in terms of the number of households owned and rented, whereas the freight transportation sector measures activity in terms of ton miles traveled. There is no demand indicator common to all sectors, so sectoral energy-intensity indicators will be unique to each sector.(91) This leads to the major obstacle in Step 3--adding unique sectoral-intensity indicators together to obtain an economy indicator of energy-intensity.
The actual sectoral energy-intensity indicators cannot be added together, but an alternative is changes in an economy-wide energy-intensity built up from the changes in the sectoral energy-intensity indicators. For any given period of time, an economy composite energy-intensity indicator can be developed by weighing each of the sectoral changes in each of the indicators by the percent share of total consumption that each sector holds.
Box 7.1 provides an example of the development of such an economy-wide energy-intensity composite. The advantage of this methodology is that this single economy-wide energy-intensity indicator reflects the uniqueness of each sector, and can be adjusted for some of the weather, structural, and behavioral effects that can affect measures of energy-intensity such as those listed in Step 1 above. The next section attempts to develop economy energy-intensity composites using the methodology presented in this section.
This section will demonstrate the use of the methodology presented above. An economy energy-intensity composite (hereafter called economy composite) will be developed using changes in energy-intensity indicators based on adjusted site energy reflecting weather, behavioral, and structural changes. This economy composite will be compared to an economy composite developed based on changes in energy-intensity indicators using unadjusted site energy. Although an energy composite will be developed based on changes in intensity indicators using primary energy, adjustments to reflect behavioral and structural changes will not be developed in this report.
Two major obstacles must be overcome in the development of economy composites. The first is the limited availability of comprehensive energy consumption data for the two comparison years. This is an obstacle whether the economy composite uses changes in intensity-indicators based on primary or site energy (unadjusted or adjusted). The second major obstacle is the necessity of developing primary conversion factors. These are needed to convert site energy into primary energy so that losses in production, transmission, and distribution can be included in an economy primary energy composite.
The best consumption data available are from EIA's consumption surveys for the manufacturing, commercial, residential, and residential transportation sectors. Transportation sector data are available Oak Ridge National Laboratory for other modes of transportation. The major obstacle is time coverage. EIA's consumption surveys do not cover the same years. Therefore, the only solution is to create an, e.g., "1985-1988" economy composite instead of a true 1985-1988 economy composite. This solution is not perfect, but it is the best solution until further research detects another solution.
Information on losses is not available for all energy sources.(92) Primary conversion factors are available or can be developed for natural gas and electricity.
Electricity. Electric utilities, and by association, nonutility generators,(93) can fully measure their generation and transmission and distribution (T&D) losses by fuel input (i.e., fossil fuel, nuclear, hydropower, and geothermal). In the development of an economy composite using changes in energy-intensity indicators based on primary energy, annual primary conversion factors for electricity by region are developed from the losses. These standard, useful measures of the efficiency of electricity generation and T&D are multiplied by regional site electricity requirements for each sector of the economy in order to estimate primary electricity consumption. The primary conversion factors developed for electricity vary by year and Census region (Table 7.1).(94)
Natural Gas. Natural gas T&D losses are more difficult to measure since they are pipeline specific. Losses on the total amount of natural gas passing through the entire system vary with the volume of gas and distance traveled in the pipeline. Industry experts within EIA and the American Gas Association came to the conclusion that 1.02 was a reasonable estimate to be used as a primary conversion factor in the development of primary energy estimates. This conversion factor for primary natural gas used in this chapter is less sophisticated than the one developed for electricity. It is a single multiplier, regardless of year or region.
Other Energy Sources. Energy losses in pipeline, marine, and truck transportation as well as in bulk storage and distribution facilities have not been quantified for either petroleum or coal products.
Primary energy can be estimated by summing the site energy consumption for each sector, multiplied by the primary conversion factors at the regional level.(95) This represents approximately 86 percent of the primary energy reported by EIA in Table 2.1 in the Annual Energy Review 1993 (AER 93). The only energy estimates omitted by this method are primary energy estimates used in mining, agriculture, forestry, recreational boats, and military transport vehicles. Comparing percent changes over time for the derived primary energy and AER93 primary energy consumption data for 1985-1988 and 1988-1991 reveals the following:
For 1985, 1988, and 1991 economy composites have been developed using site energy, adjusted site energy, and the primary energy estimates described in the last section. Figure 7.2 shows the changes in these composites during both the intervals of growth/growth and growth/recession. Additionally, this figure also shows the changes in the two energy-intensity indicators depicted earlier in Figure 7.1.
During both intervals, Gross Domestic Purchases grew faster than primary energy consumption, resulting in declining energy per dollar of Gross Domestic Purchases. During periods of economic growth, the primary energy per dollar Gross Domestic Purchases ratio overestimates the true reduction in energy-intensity, whereas in periods of recession, the primary energy per dollar Gross Domestic Purchases ratio underestimates the reduction in energy-intensity. In the latter case, the energy per Gross Domestic Purchases ratio may simply eliminate the impact of weather, behavior, or other structural effects that may have occurred during the recession rather than representing true efficiency improvements.(96)
Population growth remained very steady, averaging approximately 3 percent over both intervals. The major contraction in primary energy growth during the growth/recession interval (down to 1.1 percent from 8.4 percent in the growth/growth interval) resulted in very distinct percent changes in primary energy per capita ratios. Another interesting finding is that the economy's primary energy composite registers the same percent change as the primary energy per capita ratio during recessionary times. This could be interpreted to mean that without strong economic growth, population is the strongest variable influencing energy-intensity changes. Population is an important demand indicator in residential and passenger transportation energy consumption.
Energy-intensity changes are the smallest economy-wide when examined in terms of primary energy, especially during the growth/recession interval, highlighting the effects of increases in electricity consumption by end users.
Site energy-intensity indicators need to be examined without weather, behavior, and other structural effects that camouflage the true efficiency improvements in the sectors. In Figure 7.2, note how adjusting for weather in the commercial and residential sectors, occupancy in the commercial sector, and inventory changes and capacity in the manufacturing sector reduces the magnitude of the change in the energy composites during growth/growth interval and increases it during the growth/recession interval.
Building a composite from the specific energy-intensity changes in each sector provides far greater insight than energy per GDP or energy per capita. A composite using population is inappropriate since not all sector activity responds to changes in population; e.g., the industrial sector's output cannot be related directly to population growth. Likewise, it would be difficult to build a composite using GDP since the estimation of each sector's contribution to GDP is not an exact science.
Using demand indicators such as ton
miles allows each sector to be measured uniquely. A limiting factor is that the
availability of consumption survey data used to create the composite are available only
every 3 years. Additionally, the available 3 years are not the same 3 years for each of
the surveys. Thus, the composite developed in this chapter is far from perfect, but indeed
is a "beginning" in the development of energy-intensity indicators for the
economy as a whole.
87Gross Domestic Purchases is used instead of Gross Domestic Product since it includes all personal consumption, gross private- domestic investment, and government purchases (Federal, State, and local) including imports, but not exports.
91Sectors, in this report are defined by the standard energy convention: residential, commercial, industrial, and transportation. This is a different accounting framework than reflected in the National Income and Product Accounts and the Input-Output Accounts, where sectors reflect SIC categories. Creative use of the Gross Output by SIC sector may provide the common denominator for the residential, commercial, industrial, and transportation sectors.
92Losses occurring in the extraction and production phases for coal, petroleum, and natural gas are accounted for in mining, which pertains to the nonmanufacturing industrial sector. Energy losses at refineries are also generally included in the industrial sector.
96 If Gross Domestic Product was used as the demand indicator instead of Gross Domestic Purchases, the percent changes would not be too different. From 1985 to 1988, Site Energy/Gross Domestic Product decreased by 1.1 percent and Primary Energy/Gross Domestic Product decreased by 1.6 percent. From 1988 to 1991, Site Energy/Gross Domestic Product decreased by 1.7 percent and Primary Energy/Gross Domestic Product decreased by 1.0 percent.
File Last Modified: October 17, 1999