This appendix provides information about the quality of renewable energy data presented in this report. Information pertinent to renewable energy source data quality, in general, is presented first, followed by discussion of electric and non-electric data sources by fuel type.
Renewable energy projects pose special challenges when attempting to collect complete information on them. One challenge is the dispersed nature of many renewable energy forms, such as a photovoltaic (PV) system for generating electricity that may operate in a "standalone" fashion in a remote location. If the facility is not connected to an electricity grid, there is no Federal regulatory requirement to report its operating information. Tracking down hundreds or thousands of such facilities, each with a small power output, can be extremely challenging.
Another challenge involves tracking renewable energy supplies. Conventional energy supplies, such as petroleum, are easily tracked because the distribution networks (usually pipelines) are limited and well-defined. This permits one to make reasonable assumptions about fuel consumption, assuming stocks can be reasonably estimated.(17) The same cannot be said for many renewable energy supplies. Often a large number of energy consumers must be surveyed in order to make reasonable inferences about renewable energy consumption. Wood, for example, is gathered by tens of thousands of entities--millions if residential use is considered--for fuel uses not reportable for regulatory purposes. Thus, obtaining accurate data on wood energy consumption would entail conducting large end use consumption surveys.
Finally, some renewable energy sources are byproducts (such as pulping liquor) of non-energy processes. To track such uses, information must be solicited from respondents not generally in the energy supply chain.
As noted in Chapter 1, 69 percent of renewable energy consumption measured by EIA is used to produce electric power. It is, therefore, important to examine the coverage quality of EIA renewable electricity data. EIA renewable electricity generation is derived from two principal sources: Form EIA-759, "Monthly Power Plant Report," and Form EIA-867, "Annual Nonutility Power Producer Report." Form EIA-759 is sent to all utilities, while the EIA-867 is required of all nonutility generating facilities exceeding 1 megawatt capacity. (This includes facilities which meet Federal Energy Regulatory Commission [FERC] standards as a "qualifying facility" [QF], as well as independent power producers [IPPs]). Therefore, off-grid electric applications are not captured here (although they may be covered in EIA's Manufacturing Energy Consumption Survey(19)).
Because electric utilities are easily identified, seldom change business status, and have mandatory regulatory reporting requirements, complete coverage of utility-generated electricity is virtually assured. In contrast, nonutilities (i.e., QFs and IPPs) are required only to file regulatory reports at the time of their intention to become a grid electricity-producing facility. Over time, QF ownerships and locations change frequently. These factors, combined with the large number of QF applications, make tracking these facilities difficult. Accordingly, EIA has developed a threshold below which nonutility units are not surveyed. Form EIA-867 is a mandatory survey of all existing and planned nonutility electric generating facilities in the United States with a total generator nameplate capacity of 1 megawatt or more.
An analysis of the Form EIA-867 universe indicates that the survey's capacity under coverage varies between 3 and 10 percent, depending on the fuel source (Table B1). Capacity and unit coverage are the most difficult for wind, where numerous small units exist. EIA has analyzed the differences between capacities reported for identical renewable units on Form EIA-867 and alternative sources. Capacity discrepancies were found to result from these factors:
In a follow-up study of capacity discrepancies, the EIA-867 was over four times more likely to have the correct value than the alternative source, which covered units of all sizes.
Industrial (Nonutility) Generation
Until this issue of this report, nonutility generation for the current year was based on preliminary estimates of nonutility generation, with final data appearing in the following year's issue. An examination of 1996 data reveals that the preliminary estimate overstated the year-to-year increase in generation by a substantial amount (Table B2). In this report, final 1997 data comes from the Form EIA-867, "Annual Nonutility Power Producer Report," eliminating the source of error associated with preliminary data.
Overview
The primary application for renewable energy other than making electricity is creating heat for industrial processes, buildings, or water. Most non-electric consumption data are gathered on two EIA consumption surveys: the Manufacturing Energy Consumption Survey (MECS), and the Residential Energy Consumption Survey (RECS). MECS is based on the U.S. Bureau of the Census' Census of Manufacturing. As far as renewable energy is concerned, MECS provides consumption estimates of total industrial energy and various categories of biomass, including wood. RECS is based on an area probability sample of households selected by EIA. For renewable energy, it provides estimates of residential wood energy consumption.
| ||||||||||||||||||||||
| Sources: 1996 Preliminary: Energy Information Administration, Annual Energy Review 1996, DOE/EIA-0384(96) (Washington, DC, July 1997); 1996 Final: Table 4 of this report. |
There are three other non-electric applications for renewable energy: solar heating, alcohol transportation fuels, and geothermal energy. Solar energy for non-electric applications is derived from the EIA Solar Collector Manufacturing Survey, Form EIA-63A/B (formerly CE-63A/B). The survey does not collect energy "consumption" data, but rather production statistics on various types of solar and photovoltaic energy units. EIA applies additional assumptions regarding their application to estimate the amount of heat energy derived from installed solar/PV panels. Alcohol fuel consumption information is provided by the Form EIA-819M, "Monthly Oxygenate Telephone Report." Geothermal non-electric energy information is taken from data provided by the Oregon Institute of Technology, Geo-Heat Center.
Biomass
Wood is the principal component of biomass energy. Information on non-electric wood energy consumption is derived from the MECS and RECS surveys.
Although some questions about MECS coverage have been raised, no formal analysis of current data exists to support this concern. According to 1983 U.S. Forest Service statistics on wood harvested for fuelwood, the Pulp and Paper Industry subgroup of the Forest Products Industry group consumed only 42 percent of total sector wood energy, not including black liquor (a byproduct fuel). MECS surveys the smaller-populated Pulp and Paper Industry intensively but only randomly samples the larger-populated remainder of the Forest Products Industry. For a variety of reasons, it is difficult to trace wood energy supply to wood consumed for energy. RECS covers wood consumption only for the primary residence of those surveyed; thus, wood consumption by second homes is omitted. This causes residential wood energy consumption to be understated by about 5 percent, if not adjusted. Beginning this year, EIA is adjusting RECS wood consumption estimates for second home use.
Cross-checks of Form EIA-819M information on alcohol fuels with data from the Bureau of Alcohol, Tobacco, and Firearms and the U.S. Department of Transportation have not revealed any major deficiencies in the Form EIA-819M data.
Geothermal
EIA does not collect data on non-electric applications of geothermal energy such as crop drying and groundwater heat pumps. A study prepared for the U.S. Department of Energy by the Oregon Institute of Technology, Geo-Heat Center, indicates that non-electric uses of geothermal energy amounted to nearly 16.2 trillion Btu in 1997 (Table B3). Sixty-four percent of this energy was provided by geothermal heat pumps.
| ||||||||||||||||||||||||||||||||||||||||
| Source: John Lund, Oregon Institute of Technology, Geo-Heat Center (Klamath Falls, Oregon, March 1998), unpublished data. |
Wind, Solar, and Photovoltaics
EIA does not collect information on direct energy uses of wind (e.g., water-pumping). No comprehensive source of such information is known.
The data collected on Forms EIA-63A and EIA-63B are subject to various limitations: (1) coverage (the list of respondents may not be complete or, on the other hand, there may be double counting); (2) nonresponse (some of those surveyed may not respond, or they may not provide all the information requested); and (3) adjustments (errors may be made in estimating values for missing data).
EIA collects solar data only on terrestrial systems; it does not collect data on satellite and military applications. The total value of U.S. photovoltaic shipments in 1997 according to the Forms EIA-63A and EIA-63B was $175 million. Based on anecdotal information for 1998, shipments ranging from about $150 million to $160 million went for satellite applications. Military applications cannot be estimated due to classified information and budgetary accounting. These figures do not include possible inventories held by distributors, retailers, and installers.
The universe of solar/PV survey respondents is a census of those U.S.-based companies involved in manufacturing and/or importing solar collectors and photovoltaic cells and modules. Care has been taken to establish the survey frames accurately. The frames of potential respondents are compiled from previous surveys and from information in the public domain. However, because the solar collector and photovoltaic cell and module industries are subject to sporadic entry and exit of manufacturers and importers, the frame may exclude some small companies that have recently entered or reentered the industry. From 1993 through 1997, EIA received reports from all known potential respondents.
In 1997, the EIA began collecting information on geothermal heat pumps using its new survey the Form EIA-902, "Annual Geothermal Heat Pump Manufacturers Survey." The principal data collected are the number and type of heat pumps and their capacity ratings.
The data collected on Form EIA-902 are subject to various sources of error. These sources are: (1) coverage (the list of respondents may not be complete or, on the other hand, there may be double counting); (2) nonresponse (all that are surveyed may not respond or may not provide all information requested); (3) respondents (respondents may commit errors in reporting the data); (4) processing (the data collection agency may omit or incorrectly transcribe a submission); (5) concept (the data collection elements may not measure the items they
were intended to measure); and (6) estimation (errors may be made in estimating values for missing data). Because the survey is a census survey, the estimates shown in this report are not subject to sampling error. Although it is not possible to present estimates of nonsampling error, precautionary steps were taken at each stage of the survey design to minimize the possible occurrence of these errors.
Follow-up contacts with EIA-902 respondents suggest that the most important data quality issue is whether manufacturers are reporting ARI-320 units accurately, because these units can be used in multiple applications. ARI-320 units may be connected either to a "boiler / cooling tower" configuration or ground / ground water. Ground / ground water connections are geothermal applications, while boiler / cooling tower configurations are traditional water-to-water heat exchange uses. Some respondents apparently interpreted any water-to-water connection as geothermal and therefore reported all ARI-320 shipments as being units used in geothermal applications. The potential error is very large, possibly exceeding 50 percent of the ARI-320 estimate. Also, most manufacturers do not have records indicating the application type for individual ARI-320 units shipped. Therefore, manufacturer data reported on the EIA-902 for ARI-320 units are estimates and not counts.
Another data quality issue involves the distinction between ARI-325 and ARI-330 units. Many ARI-325 units are dual-rated to qualify as ARI-330 units, also. Which rating is appropriate depends upon the installed application, another factor not known when the manufacturer ships the unit. Therefore, while the sum of ARI-325 and ARI-330 units may be regarded as an accurate total, manufacturers estimate the number in each category based upon heuristic information.
EIA will be addressing these data quality issues in two steps. For the upcoming survey being sent out in 1999, the EIA-902 instructions and form will be clarified to emphasize the type of ARI-320 applications manufacturers should report as geothermal. Subsequently, EIA will propose modifications to the year 2000 survey to address the above issues.
Endnotes
17 Even if stock data are only approximate, conventional energy stocks are normally a small percentage of production.
18 Information in this section is based on the report, "Renewable Energy Frame Review Updated Report: Survey Sampling Frame and Electricity Discrepancy Estimates," by Decision Analysis Corporation of Virginia (Vienna, Virginia, August 1993).
19 Because the MECS is based on the Bureau of the Census' Annual Survey of Manufacturers, EIA does not know the identity of MECS respondents.