Estimating and Presenting Power Sector Fuel Use in EIA Publications and Analysis
Robert Schnapp and Renee Miller
Energy Information Administration
Last March, Bob Schnapp presented EIA’s challenges on Priority #5, the EIA-860B
Quality Effort. Form EIA-860B, “Annual Electric Generator Report—Nonutility” was
EIA’s annual survey of nonutilities.1 We told the Committee our objective was to
improve the accuracy of data on electric generation, fuel consumption, and thermal
output. We also had as an objective to present data on fuel use for electric power and for
the end-use sectors in a consistent manner throughout EIA publications, beginning with
the Annual Energy Review 2001 (AER 2001).
Since there would be massive changes as a result of our goals of achieving consistency
and improving data accuracy, we asked the Committee for advice on alerting users.
Johnny Blair suggested that we prepare documentation that would be accessible on the
Web that would describe the changes all in one place. We are pleased to report that we
have prepared the documentation. We plan to place it on the EIA Website along with
selected tables from AER 2001 on October 29th. The documentation is attached
(attachment 1) and we welcome the Committee’s comments on it (we can make changes
to the documentation after it is on the Website). We are particularly interested in
comments on sections I and II (pages 1 –5) and on the discussion of differences between
the electric power sector and generation in section III (second paragraph, page 6).
Background. The inconsistencies we were trying to address arose due to changes
resulting from the restructuring of the electric power industry. In the early days of EIA, it
was easier to depict data for the electric power industry. Electric utilities were the only
entities that generated electricity and they used fuel solely to generate electricity. Now
there are other entities that generate electricity and some –the combined heat and power
plants – also consume fuel to produce thermal energy as well as electricity. How to
classify the fuel used by the combined-heat-and-power plants (formerly called
cogenerators) was our challenge.
While there was general agreement among EIA staff that the nonutilities that produce
electricity only should be classified with electric utilities in an electric power sector, it
was not initially clear where combined-heat-and-power plants should be classified. On
the one hand, they generate electricity and so the electricity they generate should be
counted in total generation with the generation from other members of the electric power
sector. On the other hand, they are often associated with industrial and sometimes
commercial facilities. On survey forms that collect data on deliveries to commercial and
industrial customers, fuel consumption by combined-heat-and-power plants for all
purposes (electricity generation and thermal output) has been reported in the industrial or
commercial sectors. In fact, natural gas pipelines and local utilities (the respondents to
natural gas surveys) have been unable to distinguish gas delivered to combined-heat-and
power plants for electricity generation from gas delivered for other purposes.
Form EIA-860B has been replaced by Form EIA-906, “Power Plant Report.”
Although there was agreement that the unregulated electricity-only plants should be in
the electric power sector, this decision required a change in how EIA was presenting data
because data on natural gas deliveries to the unregulated electricity-only plants were
being included in the industrial sector in the Natural Gas Monthly, Natural Gas Annual
and the AER.2 As part of the recategorization described in the documentation, this
category was removed from the industrial sector and included in the electric power
sector. This change was not quite as straightforward as it may sound, subtracting a
quantity from one category and adding the same quantity to another. EIA is now
obtaining the data on natural gas consumption by unregulated electricity-only plants from
surveys of unregulated electric power plants (EIA-906, EIA-860B) rather than from
surveys of natural gas suppliers. This change parallels the way EIA obtains its official
data on fuel consumed by electric utilities. All EIA publications use data from EIA’s
survey of electric utilities.
As for the combined-heat-and-power plants, they are shown in the sector that they are
associated with: commercial, industrial or electric power (if their primary business is the
sale of electricity). We have separate break-outs for consumption by combined-heat-and
-power plants for these sectors. For example, Table 6.5 from AER 2001 will present
data on consumption of natural gas and will show these break-outs. The version that was
in AER 2000 (without the break-outs) can be found at
http://www.eia.doe.gov/emeu/aer/pdf/pages/sec6_13.pdf . Other changes to our data
presentation are detailed in the documentation.
Timing. AER 2001 was initially scheduled to be released in August (in fact, that was the
date we had given the Committee last March). Due to complications in reviewing and
cleaning the data, the release date was moved to Oct 29th. In an ideal world, the monthly
publications (e.g., Natural Gas Monthly, Electric Power Monthly, Monthly Energy
Review) would be ready in the new format with the revised data and would be released
directly following the AER. Since we are still in the process of cleaning the monthly
electric power data, there will be at least at a few months lag between the release of the
AER and the monthly publications. We have tried to narrow the gap between the release
of the AER and the monthlies, but we have gotten requests from users for the AER and
feel that we should go forward with it. The question is how best to bridge the gap
between the AER and the other publications for users. We would welcome the
Committee’s ideas on how to present data to help the users in the interim.
We have identified the areas in which the changes will be largest-- industrial natural gas
and total natural gas consumption-- but have not as yet done a detailed analysis of
revision error. Perhaps by the time of the meeting, we will have gotten farther.
At the Census Bureau (and other agencies) when revisions are made, data for one year are
often shown the new and the old way to form a bridge for users. In EIA’s situation, we
have revised the entire series of nonutility data going back to 1989, the first year for
Unregulated electricity-only plants were included in the industrial sector in other EIA publications as
well, but not in the Annual Energy Outlook. This difference in practice caused confusion for users who
tried to use our forecasts with the historical data.
which we collected nonutility data. So we have replaced the data series and have a
consistent series in the AER.
Issues. During the course of the review and recategorization, some issues arose. We are
presenting them as they may be future topics for the Committee.
Resolving differences in data reported by fuel suppliers and consumers. Historically,
data reported by natural gas suppliers on deliveries to electric utilities have differed from
the data reported by electric utilities on how much they consumed. Natural gas staff
members have been tracking and publishing the differences in the Natural Gas Annual.
These differences have been difficult to resolve because there is not a one-to-one
correspondence between the respondents on the natural gas and electric power surveys.
EIA, for example, does not collect information on the specific utility to whom the natural
gas suppliers deliver. A similar situation has developed for the nonutility data; i.e., there
are differences between what the suppliers and end-users are reporting. There is a new
survey, Form EIA-423 "Monthly Cost and Quality of Fuels for Electric Plants Report"
that collects data on fuel receipts at nonutilities and also asks respondents to report their
fuel supplier. These data may be helpful in matching respondents and resolving reporting
differences. This ties in with the project Ruey-Pyng Lu will be presenting on using the
EIA-423 data to obtain industrial prices of natural gas.
End-use sector definitions (who owns or how used). The Committee has previously
heard about how natural gas pipelines often do not know what their customers are going
do with the gas. It is not important for their business to track end-use consumption,
especially since they no longer own the gas. During our review we realized that EIA has
a wider challenge with tracking the eventual disposition of energy. With the rise of
energy management companies and change in ownership of plants, questions have arisen
on how power plants should be classifying the primary purpose of their facility. We have
observed that when plants change ownership to an energy management company their
classification has changed from an industrial plant to an electric generator. The
classification affects the data that EIA presents on fuel used by end-use sector and raises
questions on how best to collect these data. Should the classification be based on how the
fuel is used (e.g., for manufacturing or generating power) or should the classification be
based on the category in which the owner of the plant would be classified?.