|End-Use Adjustment||Type of Housing Unit|
|Detached||Attached||2 to 4 Units||5 or More Units|
|a. Space Heat (Trillion Btu)||45||810||34||46||40||975|
|b. HDD Factor||1.251||1.251||1.251||1.251||1.251||1.251|
|c, Adjusted Space Heat (Trillion Btu) (a * b)||56||1,014||43||58||50||1,220|
|d. Air-Conditioning (Trillion Btu)||15||253||19||19||38||344|
|e. CDD Factor||.928||.928||.928||.928||.928||.928|
|f. Adjusted Air-Conditioning (Trillion Btu) (d * e)||14||235||18||18||35||319|
|g. Appliances (Trillion Btu)||38||659||38||40||60||835|
|h. Water Heating (Trillion Btu)||18||337||19||30||46||450|
Total Adjusted Consumption (Trillion Btu)
(c + f + g + h)
|Sources: Energy Information Administration, Office of Energy Markets and End Use, 1990 Residential Energy Consumption Survey, Public-use Data Files. Normal and annual cooling and heating degree-days provided by the U.S. Department of Commerce, National Oceanic and Atmospheric Administration.|
End-Use Estimate Calculation. In order to adjust for the influence of weather, end-use estimates were needed for the commercial buildings sector. The only CBECS end-use consumption estimates by energy source were available are for the 1989 CBECS. End-use data were needed for the other CBECS years. It was assumed that the percent shares for the end uses by energy source, principal building activity, and Census region were the same for all the CBECS years as they were in 1989.
Total energy consumption in each Census region by principal building activity was multiplied by these shares to obtain end-use consumption for the other CBECS years. Table A.2 provides an example for office buildings in the Northeast in 1992. As an illustration, the total consumption of electricity in 1992 was 125 trillion Btu. The percent share of electricity used for space heating in 1989 was 2.4 percent. One hundred twenty five trillion Btu * .024 = 3 trillion Btu used for electric space heat in 1992.
Degree-Day Adjusted Estimate Calculation
The following method is used to calculate the degree-day adjusted estimate:
Table A.2 also presents an example of the calculation of the degree-day adjusted consumption estimates. As an illustration, an estimated 8 trillion Btu were used by office buildings in the Northeast Census region in 1992 for air- conditioning. The CDD factor is 1.478. In 1992, 1.478 * 8 = 11 trillion Btu of degree-day adjusted electricity was used for air-conditioning.
All office buildings in the Northeast used 238 trillion Btu of energy in 1992, adjusted for weather effects.
Degree-day adjusted commercial buildings total energy consumption for the United States is the sum of the adjusted energy consumption by principal building type over all four Census regions.
Table A.2. Calculation of End-Use Energy Consumption in Northeast Office Buildings, 1992
|Energy Source||End Use||Total
|Space Heating||Air- Conditioning||Ventilation||Other||Total|
|Percent Share, 1989|
End-Use Consumption 1992
|Degree-Day Adjusted End-Use Consumption, 1992 (Trillion Btu)||
|Sources: Energy Information Administration, Office of Energy Markets and End Use, 1989 and 1992 Commercial Buildings Energy Consumption Survey, Public-Use Data Files. Normal and annual cooling and heating degree-days provided by the U.S. Department of Commerce, National Oceanic and Atmospheric Administration.|
Adjustments were made to eliminate vacant or mostly vacant buildings from the estimates for the survey years. Adjustments were made to both total site energy consumption and demand indicators for each principal building activity and Census region.
Demand Indicator Adjustment. Adjustments were made to the following demand indicators: floorspace, buildings, floorspace-hours and employees. Vacant buildings that were classified as vacant as well as those that were more than 50 percent vacant at least 3 months during the survey year were removed. An example is presented in Table A.3.
Table A.3. Occupancy-Adjusted Total Commercial Floorspace in the Northeast by Principal Building Activity, 1992
|Principal Building Activity||Total
(Million Square Feet)
than 50 percent
for 3 Months
During Survey Year
|Health Care (inpatient)||348||0||348|
|Health Care (outpatient)||38||0||38|
|Other, excluding Vacant||496||9||487|
|Source: Energy Information Administration, Office of Energy Markets and End Use, 1992 Commercial Buildings Energy Consumption Survey, Public-Use Files.|
Occupied Commercial Buildings Total Site Energy Consumption Adjustment. The total site energy consumption is adjusted by major fuel for each principal building activity (PBA) in each Census region by removing the site energy used in vacant buildings and buildings that were more than 50 percent vacant for at least 3 months (Table A.4).
Occupied and Degree-Day Commercial Buildings Site Energy Consumption Adjustment. This methodology adjusts for both vacancy and weather. First, the occupied site energy consumption is determined as explained above. Then the occupied site energy consumption is adjusted for weather. An example of this methodology was presented in Table A.2.
Table A.4. Occupancy-Adjusted Total Commercial Site Electricity Consumption in the Northeast, by Principal Building Activity, 1992
|Principal Building Activity||Total
Site Energy Consumption
|All Buildings||Vacant Buildings||Occupied Buildings|
|Food Sales/Food Service||37||< 1||36|
|Health Care (Inpatient)||24||--||24|
|Health Care (Outpatient)||1||--||1|
|Public Order||8||< 1||8|
= No cases.
Source: Energy Information Administration, Office of Energy Markets and End Use, 1992 Commercial Buildings Energy Consumption Survey, Public-Use Data Files.
To determine the U.S. total occupied and degree-day total site energy consumption use the following method:
Domestic Air Energy Use for Passenger and Freight: Derivation
The available data for aviation energy demand is for both the passenger and freight transportation sectors combined. In many aircrafts, freight is carried in the hull of the craft while passengers ride in the cabin. Although the passenger-miles traveled by air cannot be separated into passenger and freight components, the revenues (in millions of dollars) received for different types of air travel can be separated.
The relative share of revenue dollars was used to estimate the portion of energy consumed for passenger and freight movements by air. Table A.5 presents the methodology.
Table A.5. Calculating U.S. Domestic Passenger and Freight Air Travel Energy Consumption
|U.S. Domestic Passenger|
|Revenue per Passenger Mile||Cents||12.21||12.31||13.22|
|Passenger Revenues||Million Dollars||33,924||41,151||44,695|
|U.S. Domestic Freight|
|Freight Revenue per Ton Mile||Cents||102.23||111.31||103.50|
|Freight Ton Miles||Million||6.71||9.33||9.96|
|Freight Revenues||Million Dollars||6,860||10,385||10,309|
|Total Operating Revenues||Million Dollars||37,629||50,155||56,165|
|Air Travel Energy Use||Trillion Btu||1,366||1,609||1,542|
|Air Passenger||Trillion Btu||1,231||1,320||1,227|
|Air Freight||Trillion Btu||134||289||315|
Passenger and freight revenues may not add to total operating revenues,
due to calculations and differences in data sources. Air passenger
energy use calculated as the multiple of its revenue ratio and total
air travel energy use, e.g., in 1985 the equation is (33,924/37,629)*1,366
= 1,231. Air freight energy use calculated as the remainder after
air passenger energy use is subtracted from total air travel energy
use, e.g., in 1985 the equation is (1,366 - 1,231) = 134.
Sources: Department of Transportation, Bureau of Transportation Statistics, National Transportation Statistics (September 1993), Tables 1, 4, and 6. Eno Transportation Foundation Inc., Transportation in America 1994, pp. 44 and 49.
Site Energy Consumption Conversion: Electricity consumption is reported for pipelines and passenger rail in the Oak Ridge National Laboratory Transportation Energy Data Book. The primary electricity volumes were converted to site electricity by applying conversion factors. See "Primary Conversion Factors" in the economy section in this appendix. Table A.6 presents the methodology.
Table A.6. Calculating U.S. Transportation Site Electricity Consumption
|Primary Electricity (Trillion Btu)|
|Electricity Conversion Factor||3.292||3.311||3.30|
|Site Electricity (Trillion Btu)|
Site electricity is calculated as primary electricity divided by
the conversion factor, e.g., in 1985 the equation for pipelines
is (239.0/3.292) = 72.6.
Sources: Department of Energy, Oak Ridge National Laboratory (ORNL), Transportation Energy Data Book (ORNL-6798), Editions 11 and 14, Table 2.6 and unpublished 1985 data from ORNL.
The capacity-utilization rate equals the seasonally adjusted index of industrial production divided by a capacity index (sustainable practical capacity, i.e., the greatest level of output a plant can maintain within a realistic work schedule). The Federal Reserve Board weights the capacity indexes by value-added proportions.
Capacity-Adjusted Value of Production Method. This method adjusts the value of shipments for changes in capacity after the value of shipments has been adjusted for changes in inventories (value of production). See "Inventory Adjustment" in this section of the appendix. The method is as follows:
Constant-dollar Capacity-Adjusted Value of Production t
Capacity-Utilization Rate t
|Constant-dollar Value of Production|
Table A.7. Capacity-Utilization Rate: 26-Year Average and Annual for 1985, 1988, and 1991, by SIC
Major Industry Group
|Capacity Utilization Rates|
|20||Food and Kindred Products||82.3||81.0||81.8||81.4|
|22||Textile Mill Products||86.2||83.0||88.7||83.3|
|23||Apparel and Other Textiles Products||81.1||80.4||82.8||77.6|
|24||Lumber and Wood Products||83.1||84.6||89.7||79.3|
|25||Furniture and Fixtures||81.7||79.8||84.2||74.2|
|26||Paper and Allied Products||89.7||89.7||93.2||88.4|
|27||Printing and Publishing||86.5||87.0||89.9||79.7|
|28||Chemicals and Allied Products||80.0||77.1||83.9||80.6|
|29||Petroleum and Coal Products||85.5||78.6||85.2||86.0|
|30||Rubber and Miscellaneous Plastics||83.6||85.3||87.7||80.3|
|31||Leather and Leather Products||81.9||75.6||79.5||78.5|
|32||Stone, Clay, and Glass Products||77.9||75.6||82.2||73.2|
|33||Primary Metal Industries||80.1||74.0||87.5||77.9|
|34||Fabricated Metal Products||77.2||74.9||81.1||73.2|
|35||Industrial Machinery and Equipment||80.8||73.4||80.5||72.6|
|36||Electronic and Other Electric Equipment||80.4||80.7||83.7||78.0|
|38||Instruments and Related Products||82.0||83.6||80.5||77.2|
|39||Miscellaneous Manufacturing Industries||75.6||67.7||79.5||74.6|
= Not Available
Sources: U.S. Department of Treasury, Federal Reserve Board (Table provided by Charles Gilbert, 10/12/94). Federal Reserve Statistical Release (August 15, 1994), Table 3 (average).
Changes in inventories need to be considered when using a demand indicator such as the value of shipments. If inventories are being drawn down, the value of shipments will overestimate the actual value of production. If inventories are being built, then the value of shipments will underestimate the value of production.
Inventory-adjusted Value of Shipments or Value of Production. The inventories used in the adjustment are year-end inventories at cost or market value, deflated to 1987 constant dollars using value of shipments implicit price deflators reported by the U.S. Department of Commerce, Bureau of Economic Analysis. The following steps were followed to adjust the value of shipments for the effects of changes in inventories:
Value of Shipments Deflator t = Constant-dollar Value of Shipments t / Current-dollar Value of Shipments t
Constant-dollar Value of Production t =
Value of Shipments t
Value of Shipments
|-||(Value of Shipments Deflator t-1||*||(Current-dollar) Inventories t-1|
Table A.8. Value of Production Methodology Example
|Calculations||Preceding Year (t-1)||Current Year (t)|
|a. Value of Shipments (million 1987 dollars)||307,345||319,212|
|b. Value of Shipments (million current dollars)||255,723||303,270|
|c. Implicit Price Deflator (a/b)||1.20||1.05|
|d. Year-End Inventories (current million dollars)||24,397||24,023|
|e. Year-End Inventories (million 1987 dollars) (c*d)||29,276||25,224|
|f. Value of Production (million 1987 dollars) (319,212 + 25,224 - 29,276 = 315,160)||--||315,160|
|Note: This example is for illustrative purposes only. Although, MECS-weighted value of shipments data (adjusted to 1987 SIC) were used throughout Chapter 6, "Industrial Sector," confidentiality does not permit EIA to release value of shipments data that have been revised by MECS weights.|
One way for removing effects such as geography and housing unit type effects is to index from a "two-way" disaggregation of characteristics to develop an index for total U.S. housing.
As an example, for the Northeast Census region in the residential sector, start with consumption per household for each of the five housing types indexed to 1 in the first year as follows:
Northeast Census Region 1984 1987 1990
Mobile Homes 1 x x
Single-Family Detached 1 x x
Single-Family Attached 1 x x
Multifamily (2-4 Units) 1 x x
Multifamily (5 or More Units) 1 x x
Compute an energy-weighted index (preferably Tornqvist index) of the individual energy-intensity indices for 1987 and 1990 for each of the Census regions. This index will be devoid of mix issues relating to housing type and geography. While there will still be other behavioral and/or structural effects in the five disaggregated indices, a layered procedure that moves down the chain to include measures that disaggregate end-uses (for example, an index of space-heating consumption per square foot, not household, since housing sizes are changing, or an index of water heating per occupant rather than per household, since persons per household are changing, etc...) and to exploit all of the detail available from the energy consumption surveys is a "good measure" of energy intensity. These effects might be relatively minor over the 3 surveys years being compared, but without the calculations such as these, only qualitative judgments as to the potential effects can be made. This proposed approach is for all intents and purposes has been used for the transportation and economy composites. Indices can be developed for each sector and weighted by shares of total energy consumption in the economy.
Primary Conversion Factors for Total Site Electricity
Primary energy estimates include losses in the generation, transmission, and distribution of electricity. In this report, conversion factors are developed to account for these losses. Total site electricity estimates are multiplied by these conversion factors to obtain primary electricity estimates.
The methodology for developing the conversion factors and obtaining primary electricity estimates is shown in the following steps:
Step 1. Calculate gross inputs:
Convert utility-site generation by energy source and region in kWh to equivalent gross-generation estimates in kWh (including generator or shaft losses) by multiplying site generation by the appropriate gross/site ratio plus the transmission and distribution losses estimated at 8 percent by the Department of Energy, Office of Energy Management.
For each year, the gross-generation estimates, by Census region, are multiplied by the appropriate annual heat rate for each energy source to obtain gross inputs for electricity generation by utilities such that:
Gross Inputs = (Net Generation * (Gross/Site Ratio + T&D Losses) * Heat Rate) /1000.
Data on energy sources used by nonutilities and net exporters to produce electricity purchased by U.S. utilities are not available. Since electricity from these sources is primarily produced from either fossil fuels or hydro resources, the heat rates of fossil-fueled steam generators are applied to the purchased energy. Table A.9 presents an example of the calculation of gross inputs for the Northeast Census region in 1992.
Table A.9. Example of Calculating Gross Inputs for the Northeast in 1992
|Calculation Inputs||Generation Energy Sources|
|Fossil Fuels||Nuclear||Hydropower||Geothermal||Other||Nonutility Purchases||Net
|a. Site Generation (Billion kWh)||218.4||144.4||30.6||0||0.5||52.5||12.1|
|b. Gross/Site Ratio||1.07||1.06||1.01||1.06||1.07||1.07||1.07|
|c. T&D Losses||0.08||0.08||0.08||0.08||0.08||0.08||0.08|
|d. Heat Rates (Btu/kWh)||10,302||10,678||10,302||20,955||10,302||10,302||10,302|
|e. Gross Inputs (Trillion Btu) (a * (b + c)* d)/1000||2,588||1,757||344||0||6||622||143|
Fossil fuel heat rate is used also for hydropower, nonutility
purchases, net imports, and other.
Sources: Energy Information Administration, Office of Coal, Nuclear, Electric, and Alternative Fuels, Electric Power Annual 1993 (DOE/EIA-0348(93), Tables 13, 61 and 62. Energy Information Administration, Office of Energy Markets and End Use, Monthly Energy Review (DOE/EIA-0035(94/08)), Table A8; Production Annuelle Brute et Nette d'Electricity, p.16.
Step 2. Add utility-plant use to electricity sales for each sector.(99)
Step 3. All regional estimates of electricity sales and plant use in kWh are converted to Btu using 3,412 Btu per kWh.
Step 4. Divide gross energy inputs by the total site electricity (sales and plant use) for each Census region to obtain conversion factor.
Step 5. Site electricity consumption for each end-use sector and Census region are multiplied by the primary electricity conversion factors to obtain the corresponding primary electricity estimates. The data source is the consumption survey data as developed in each of the sector chapters.
There are three main advantages to using the method described above:
Regional Manufacturing Estimates: The 1985 estimate of total inputs for heat, power, and generation was based on EIA's revised 13,631 trillion Btu total. This total was distributed regionally by using regional shares based on Table 3 in the EIA publication, Manufacturing Energy Consumption Survey: Consumption of Energy, 1985 (DOE/EIA-0512(85).
The analysis of the manufacturing sector used revised 1985 estimates to match the revised 1987 SIC standards. Electricity was not considered separately. Therefore, electricity consumption for heat, power, and generation in Btu were calculated from the kilowatthour estimates in Table 3 of Manufacturing Energy Consumption Survey: Consumption of Energy, 1985 and Manufacturing Energy Consumption Survey: Consumption of Energy, 1988 (DOE/EIA-0512(88)) by multiplying the estimates by 3,412 Btu per kWh. Natural gas consumption for heat, power, and generation in Btu were calculated from the cubic-foot estimates in the same tables. The natural gas estimates for the respective surveys were multiplied by 1,031 Btu/cubic feet.
Both the 1991 electricity and natural gas estimates in Btu were provided in Table A.4 of Manufacturing Energy Consumption Survey: Consumption of Energy, 1991, DOE/EIA-512(91) report.
Regional Passenger Transportation Estimates: Passenger transportation data are not available by Census region. On the advice of Oak Ridge National Laboratory, it was decided that since household vehicles account for over 70 percent of the total energy for passenger travel, regional shares based on the Residential Transportation Energy Consumption Survey (RTECS) would be an adequate proxy for passenger transportation energy estimates.(100)
There has been very little change in the RTECS regional distribution across survey years. For 1985, 1988, and 1991, the total U.S. passenger transportation energy estimates were multiplied by the percent shares by Census region to obtain regional energy consumption estimates for passenger travel. The percent shares were based on the 1988 RTECS. The 1988 shares were: 17 percent for the Northeast; 25.2 percent for the Midwest; 36.0 percent for the South, and 21.8 percent for the West.
Regional Freight Transportation Estimates: Freight transportation data are not available by Census region. On the advice of Argonne National Laboratory, it was decided that since most of the energy source used for freight transportation was fuel oil, EIA data could be used instead of Department of Transportation, Bureau of Transportation Statistics data.
EIA's Petroleum Marketing Division annually surveys State-level distillate fuel oil consumed by trucks, rail, and marine vehicles and residual fuel oil consumed by rail and marine vehicles, bench marking to the petroleum product supplied data published by EIA's Petroleum Supply Division. The report, Fuel Oil and Kerosene Sales, provides State-level data on the number of gallons consumed, which were converted to trillion Btu using Oak Ridge National Laboratory conversion factors, in order to derive percent shares.
Since there has been little fluctuation in regional distribution year-to-year, it was determined to apply 1988 regional percent shares to all annual estimates. The regional percent shares used were: 12.5 percent for the Northeast; 20.2 percent for the Midwest; 42.0 percent for the South, and 25.3 percent for the West.
Quantity Index. A quantity index measures changes in quantity over time. The Index of Industrial Production, developed by the Federal Reserve Board, is a quantity index.
The weighted aggregate quantity index is computed similar to the weighted aggregate price index.
The Price Index. The price index is a weighted average of expenditures, as a percentage of expenditures existing in a base year. A price index may be calculated for a single good or can be calculated as an aggregated price index for a "basket" of several goods. Price indices can be unweighted or weighted. The unweighted aggregated approach is heavily influenced by those goods with higher prices which dominate the index. To reduce this sensitivity of the unweighted index, a weighted price index is used. Each good in an weighted aggregate price index is be weighted according to its importance.
One way to weight a price index is to use base-year quantities, the Laspeyres price index.
Laspeyres price indices are calculated by comparing the current and base year cost of a basket of goods of fixed composition. As an example, the "basket" can be several "goods" such as energy, clothing, food, housing, etc. that we find in the "basket" used to calculate the Consumer Price Index (CPI) or one "good" such as the major energy sources that is used to calculate the energy component of the CPI. The Producer Price Index (PPI) developed and maintained by the Bureau Labor Statistics, is also a base-weighted price index.
Laspeyres base-weighted price index (ratio of today's cost using base-year quantities to the base-year cost of the goods) equals:
ptiqoi / poiqoi * 100
where the base-year quantities of the various goods = qoi ,
the base-year prices of the various goods = poi , and
present prices of the various goods = pti .
Quantity Index. Similar to the Laspeyres price index, quantities for each item are measured in the base year o and year t with qoi and qti, representing these quantities for item I (e.g.,end use or energy source). The quantities are then weighted by a fixed price (wti) such as value added, value of shipments, etc. where the quantity index I equals
In some quantity indexes, the weight for item I is the base-period price (poi).
98Heating degree-days, cooling degree-days, and normal degree-days are defined under "General Terminology" in the Glossary.
99Electricity sales data are obtained from EIA's Electric Power Annual 1993 (DOE/EIA-0348(93), Table 26. Plant use electricity is the difference between gross inputs and site electricity.
100Business fleets are operated very differently from household vehicles, but adequate data are not available on fleet vehicles.
File Last Modified: October 17, 1999