Only GII produces a comprehensive energy projection with a time horizon
similar to that of AEO2006. Other organizations address one or more aspects
of the energy markets. The most recent projection from GII, as well as
others that concentrate on economic growth, international oil prices, energy
consumption, electricity, natural gas, petroleum, and coal, are compared
here with the AEO2006 projections.
In the AEO2006 reference case, the projected growth in real GDP, based
on 2000 chain-weighted dollars, is 3.0 percent per year from 2004 to 2030
(Table 19). For the period from 2004 to 2025, real GDP growth in the AEO2006 reference case is similar to the average annual growth projected in AEO2005.
The AEO2006 projections of economic growth are based on the August short-term
forecast of GII, extended by EIA through 2030 and modified to reflect EIAs
view on energy prices, demand, and production.
The projected average annual GDP growth rate for the United States from
2004 through 2010 ranges from 2.8 percent to 3.3 percent. The AEO2006 reference
case projects annual growth of 3.3 percent, matching the average annual
real GDP growth projected by the Office of Management and Budget (OMB),
the Congressional Budget Office (CBO), and the consensus Blue Chip forecast.
GII and Energy Ventures Analysis, Inc. (EVA) project real GDP growth at
3.2 percent per year. Two other organizations project somewhat lower annual
growth: Interindustry Forecasting at the University of Maryland (INFORUM)
at 2.9 percent and Energy and Environmental Analysis, Inc. (EEA) at 2.8
When the projection period is extended to 2015, the uncertainty in the
projected rate of GDP growth is reflected in the wider range of the projections
(2.5 to 3.2 percent per year). AEO2006 remains in the upper half of the
range, whereas the CBO projection shows a considerable slowing of GDP growth
from 2010 through 2015. There are few public or private projections of
GDP growth rates for the United States that extend to 2030. The AEO2006 reference case projection reflects a slowing of the GDP growth rate after
2020, consistent with an expected slowing of population growth.
World Oil Prices
Comparisons with other oil price projections are shown in Table 20. The
world oil prices in EIAs AEO2006 are generally toward the high end of
the oil price projections. Of the nine other publicly available long-term
projections, only twoPetroleum Industry Research Associates, Inc. (PIRA)
and Petroleum Economics, Ltd. (PEL)have projections of world oil prices
for specific years after 2010 that exceed the AEO2006 reference case projections.
Four of the nineGII, Altos Partners (Altos), Strategic Energy and Economic
Research, Inc. (SEER), and the International Energy Agency (IEA) Reference
Scenario have prices lower than those in the AEO2006 low price case for
at least some years. All the projectionsexcept for the price forecast
from Altos, which has not been revised since July 2003have raised their
long-term price expectations relative to last years releases.
The world oil price measure does vary by projection. In some cases, the
measure is the WTI spot price, Brent equivalent, weighted average U.S.
refiner acquisition cost of imported crude oil, or a basket of crude oils.
For AEO2006, EIA redefined its world oil price path to represent the average
U.S. refiners acquisition price of imported low-sulfur light crude oil
(see Issues in Focus for discussion). Those prices are considered comparable
to the WTI prices most often cited in the trade press as a proxy for world
oil prices. The different price measures used in the various projections
do not wholly explain the different price expectations among the projections.
For instance, GII publishes a WTI spot price forecast that is considerably
lower than the AEO2006 reference case prices, and PIRA publishes a WTI
spot price forecast that is considerably higher than the AEO2006 reference
case prices in most years (Table 20).
Recent variability in crude oil prices demonstrates the uncertainty inherent
in projections for crude oil markets. The oil price paths projected by
several organizations, including EIA, illustrate the uncertainty. For example,
for 2010, the price range in the projections is from a low of about $28
per barrel by Altos to a high of almost $48 per barrel projected by PEL.
The range in the projections widens in 2020, from a low of $32 per barrel
(GII and DB) to a high of $63 per barrel (PIRA). In 2030, the band of prices
represented by the published projections narrows to $23 per barrel, probably
in part because the PIRA forecast horizon ends in 2020.
To construct the world oil price cases for AEO2006, EIA employed input
from a Delphi group of energy analysts. In August 2005, an informal, nonrandom
sample of expert oil analysts from outside DOE were invited to participate,
with the stipulation that the responses were to reflect the analysts personal
views and not necessarily the views of the organizations with which they
were affiliated. In addition, the analysts were told that their responses
would be anonymous. Seventeen analysts were surveyed, and eight responses
were received. The median response from the Delphi group was generally
higher than any of the other published projections, though still falling
within the range defined by the AEO2006 low and high price cases (Table
20). The group expected oil prices to continue rising through the 2005-2030
time period, to nearly $73 per barrel in 2030more than $20 per barrel
higher than the nearest alternative, the Deferred Investment Scenario published
Total Energy Consumption
The AEO2006 projects higher growth in end-use sector consumption of petroleum,
natural gas, and coal than occurred from 1980 to 2004 but lower growth
in electricity consumption (Table 21). Much of the projected growth in
petroleum consumption is driven by increased demand in the transportation
sector, with continued growth in personal travel and freight transport
projected to result from demographic trends and economic expansion. Natural
gas consumption is expected to increase in the residential, commercial,
and industrial sectors, despite relatively high prices. Natural gas is
cleaner than other fuels, does not require on-site storage, and has tended
to be priced competitively with oil for heating. Coal consumption as a
boiler fuel in the commercial and industrial sectors is expected to decline
slightly, with potential use in new boilers limited by environmental restrictions;
however, the projections for industrial coal consumption include its use
in CTL plants, a technology that is expected to become competitive at the
high oil prices assumed in AEO2006.
While strong growth in electricity use is projected to continue in the AEO2006 projections, the pace slows from historical rates. Some rapidly
growing applications, such as air conditioning and computers, slow as penetration
approaches saturation levels. Electrical efficiency also continues to improve,
due in large part to efficiency standards, and the impacts tend to accumulate
with the gradual turnover of appliance stocks.
The AEO2006 projections are generally consistent with the outlook from
GII; however, GII projects slightly faster growth in petroleum and natural
gas consumption and slightly slower growth in electricity consumption and
losses. The differences can be attributed largely to the higher oil and
natural gas prices assumed in AEO2006. Differences between the AEO2006 and GII projections for coal result from an increase in coal use for CTL
The AEO2006 projections for the electricity generation sector assume that
new generating capacity will be built by independent power producers rather
than utilities. Retail price projections are based on average costs for
electricity supply regions that are still regulated; marginal costs for
regions that are competitive; and a mixture of average and marginal costs,
weighted by the amounts of load, in regions with a mix of regulated and
competitive markets. As of 2005, only four electricity market regions had
fully competitive retail markets in operation; seven had mixed competitive
and regulated retail markets; and two had fully regulated markets. The AEO2006 cases assume that no additional retail markets will be restructured,
but that partial restructuring (particularly in wholesale markets) will
lead to increased competition in the electric power industry, lower operating
and maintenance costs, and early retirement of inefficient generating units.
Comparison of the AEO2006 and GII projections shows some variation in electricity
sales (Table 22). The projections for total electricity sales in 2030 range
from 4,828 billion kilowatthours (AEO2006 low economic growth case) to
5,854 billion kilowatthours (AEO2006 high economic growth case). The rate
of demand growth ranges from 1.2 percent (AEO2006 low economic growth)
to 1.9 percent (AEO2006 high economic growth). All price projections reflect
competition in wholesale markets and slow growth in electricity demand
relative to GDP growth, exerting downward pressure on real electricity
prices through 2030. Rising natural gas and coal prices balance some of
the downward pressure and tend to push electricity prices up in the later
years of the projections.
The AEO2006 reference case shows a slight decline in real electricity prices
over the full period of the projection (except for the industrial sector),
although average prices increase slightly during the last several years
as capacity margins tighten and natural gas prices climb. In contrast,
GII projects a decline in electricity prices over the second half of the
projection, as lower delivered natural gas prices to generators ($5.08
per million Btu in the GII projection, compared with $6.26 in the AEO2006 reference case in 2030) contribute to a small decrease in average electricity
prices, from 7.6 cents per kilowatthour in 2015 to 7.4 cents per kilowatthour
in 2030. The higher natural gas price in the AEO2006 reference case leads
to an increase in average electricity price, from 7.1 cents per kilowatthour
in 2015 to 7.5 cents per kilowatthour in 2030.
Both the AEO2006 reference case and GII projections include some planned
capacity additions in the near term, with the AEO2006 reference case expecting
about 29 gigawatts through 2006 and GII expecting about 25 gigawatts. Virtually
all the projected capacity additions are natural gas fired. Both projections
show electricity prices falling in the near term as a result of excess
Except for GII, all the projections for electricity demand show the fastest
growth in the commercial sector, and more additions of cycling and baseload
capability than peaking units. All the projections show significant net
additions to coal-fired capacity, including 167 gigawatts through 2030
in the AEO2006 reference case and 136 gigawatts through 2030 in the GII
projection. Both GII and the AEO2006 reference case project no nuclear
retirements; however, each of the three AEO2006 cases (reference and high
and low economic growth) projects 6 gigawatts of nuclear capacity additions
by 2030 as a result of the incentives in EPACT2005.
The fuel mix in the EVA projection differs from that in the AEO2006 reference
case and the other projections. Except for EVA, all the projections show
coal meeting about one-half and natural gas about one-quarter of the growth
in electricity generation capacity over the projection period. The EVA
projection assumes that legislation similar to the Clear Skies Actincluding
further restrictions on SO2, NOx, and mercury emissionswill be in effect
by 2010. The EVA projection also includes a tax of $5 per ton on CO2 emissions,
beginning in 2013. AEO2006 includes the impact of the EPAs new CAIR and
CAMR regulations, which have environmental effects similar to those of
the Clear Skies Act; however, AEO2006 does not assume any tax on CO2 emissions.
In the EVA projection, the combination of further environmental restrictions
and a tax on CO2 leads to greater growth in hydroelectric generation. In
the AEO2006 cases, no additional hydroelectric generation is built, both
because its economics are unfavorable and because of the difficulty of
siting new hydropower facilities.
Published projections of natural gas prices, production, consumption, and
imports (Table 23) differ considerably. The differences highlight the uncertainty
of future market trends. Because the projections depend heavily on the
underlying assumptions that shape them, the assumptions made in each should
be considered when they are compared.
The AEO2006 reference case in general projects lower total natural gas
consumption than in the other projections, and it is the only one showing
a period of decline. The exception is in the early part of the projection
period: in 2015, PIRA and Deutsche Bank AG (DB) project lower natural gas
consumption than the AEO2006 reference case, but by 2025 the AEO2006 reference
case projects lower consumption than any of the others. The primary reason
is that AEO2006 expects a stronger demand response to higher natural gas
prices, particularly in the electricity generation sector.
The highest projected level of total natural gas consumption is in the
EVA projection, due to strong growth in natural gas consumption for electric
power generation. Altos projects the strongest growth in residential and
commercial sector natural gas consumption through both 2025 and 2030, whereas
the GII and EVA projections have the lowest projected consumption levels.
The AEO2006 reference case projection for residential natural gas consumption
in 2030 is lower than all but the EVA projection, but its commercial sector
projection is higher than the GII, EVA, and SEER projections. Natural gas
consumption in the industrial and electric power sectors is more difficult
to compare, given potential definitional differences. The combined total
of industrial and electric power sector natural gas consumption from 2004
to 2030 is projected to grow the fastest in the EVA and Altos projections;
the DB projection shows much slower growth but still faster than is projected
in the AEO2006 reference case. The DB combined total in 2030 exceeds the AEO2006 reference case by less than 10 percent, whereas the GII, EVA, SEER,
and Altos projections all exceed the AEO2006 by more than 25 percent.
Domestic natural gas production provides a decreasing share and net imports
an increasing share of total natural gas supply in all the projections.
The EVA projection shows the greatest increase in the net import share
of supply, at more than 41 percent of total supply in 2030. More than 34
percent of supply is projected to come from imports in 2030 in the DB projection,
and GII and SEER both show net imports providing about 30 percent of total
natural gas supply. The AEO2006 reference case and Altos project that net
imports will meet the smallest share of total supply21 percent and 22
percent, respectivelyin 2030. Most of the projections show a notable decline
in pipeline imports over the forecast period. Only DB shows an increase
from 2015 to 2025. EVAs pipeline import projection, although significantly
greater than the rest, also declines after 2015. Much of the variation
in imports reflects different projections of net LNG imports in 2030, ranging
from a low of 4.4 trillion cubic feet in the AEO2006 reference case to
10.5 trillion cubic feet in the EVA projection.
The AEO2006 reference case projections for wellhead natural gas prices
in 2025 and 2030 fall within the range of the other projections, with the
EEA, EVA, and Altos projections higher than AEO2006 and the others lower.
In the earlier years, however, all the projections with the exception of
Altos show wellhead natural gas prices exceeding those in the AEO2006 reference
case. Of the three projections that project end-use prices for 2030 (AEO2006,
GII, and SEER), the AEO2006 reference case and SEER show the highest end-use-to-wellhead
margins for the electric power sector ($0.50 and $0.51, respectively).
The AEO2006 reference case shows the lowest end-use-to-wellhead margins
for the industrial sector. While GIIs margins for the electric power sector
are the lowest, some of the difference may be definitional. For the residential
and commercial sectors, the projected margins in the AEO2006 reference
case exceed the other projections by more than 15 percent.
As discussed earlier in this report, crude oil prices in the AEO2006 reference
case are substantially higher than they were in earlier AEOs. They are
also considerably higher than those in most of the other projections. The AEO2006 reference case shows the weighted average refiners acquisition cost
of imported crude oil (the price basis used in most of the other forecasts)
ranging from $43 to $50 per barrel (2004 dollars) between 2015 and 2030
and the average refiners acquisition cost of imported low-sulfur light
crude oil (the reference price used in AEO2006) ranging from $48 to $57
per barrel (2004 dollars) over the same period. DB assumes that the refiners
acquisition cost of crude oil will average $31.75 per barrel from 2010
through 2030; GII assumes that the refiners acquisition cost of crude oil
will be between $28 and $31 per barrel from 2015 through 2030. PIRA gives its oil price forecast
in terms of WTI, a low-sulfur, light crude oil, assuming prices of $50
per barrel in 2015 and $63 per barrel in 2020.
Despite much lower crude oil price projections, GII and DB project gasoline
consumption levels that are essentially the same as those in the AEO2006 reference case (Table 24). The GII and DB projections for gasoline demand
are within 1 percent of the AEO2006 reference case from 2015 to 2030. PIRA
sees slower growth in gasoline demand, 14 percent below the AEO2006 reference
case in 2015, due to more rapid improvement in vehicle efficiency.
In comparison with the AEO2006 reference case, projected distillate consumption
is about 2 percent lower in the DB and PIRA projections in 2015 and 5 percent
lower in 2030 in the DB projection. GII also projects lower levels of distillate
consumption than the AEO2006 reference case, 6 percent less in 2015 and
13 percent less in 2030. Most of the variation is accounted for by the
projected level of highway diesel consumption.
The projected pattern of growth in jet fuel consumption varies significantly
by projection, and the basis of the variation is not always clear. Relative
to the AEO2006 reference case, PIRA projects slightly higher jet fuel consumption
in 2015, whereas GII projects higher jet fuel consumption only toward the
end of the projection (25 percent higher in 2030 but 4 percent lower in
2015). DB also projects lower jet fuel consumption in the middle years,
9 percent below the AEO2006 reference case in 2015, but is nearly identical
with the AEO2006 reference case in the later years of the projection.
The projections also differ substantially on the projected future use of
residual fuel oil. PIRA and GII project a steady decline in residual fuel
oil consumption, but DB sees some growth in the future. In the GII projection,
residual fuel oil consumption is 3 percent below that in the AEO2006 reference
case in 2015 and 18 percent below in 2030. Both GII and PIRA project deep
declines in residual fuel oil consumption for electricity generation. The AEO2006 reference case projects more modest reductions through 2015 and
then slow growth for the remainder of the projection. The DB projections
are 14 percent and 17 percent above the AEO2006 reference case in 2015
and 2030, respectively.
Domestic crude oil production declines in all the projections, but at different
rates. As compared with the AEO2006 reference case, domestic crude oil
production declines more rapidly in the earlier years and much more slowly
in the later years of the GII projection. GII projects domestic crude oil
production 14 percent lower than in the AEO2006 reference case in 2015
but essentially the same in 2030. DB and PIRA project a much more rapid
decline in domestic crude oil production: both are about 15 percent below
the AEO2006 reference case in 2015, and DB projects a further decline,
to 19 percent below the AEO2006 reference case in 2030.
The projections do not agree on domestic production of NGL. The AEO2006 reference case projects NGL production slightly above current levels in
2015 and 2030, with peak production in 2020. DB is bearish on NGL production,
projecting 19 percent lower levels than in the AEO2006 reference case in
2015 and 42 percent lower in 2030. GII, on the other hand, is bullish on
NGL production, projecting domestic production 24 percent above the levels
in the AEO2006 reference case in 2015 and 38 percent above in 2030. EVA
and DB project the lowest totals of domestic crude oil and NGL production
in 2015 and 2030.
Declining domestic production of crude oil and rising petroleum product
demand imply greater dependence on imports in all the projections. The
decreases in crude oil production are offset somewhat by projected increases
in NGL production in the AEO2006 reference case and GII. DB projects substantial
declines in crude oil and NGL production and therefore projects the highest
levels of net imports of crude and petroleum products. DB projects import
shares 9 percentage points above the AEO2006 reference case in 2015 and
14 percentage points above in 2030. GII projects import shares that are
8 percentage points above the AEO2006 reference case in 2015 and 2030.
AEO2006 also includes alternative price cases. The AEO2006 low price case
assumes that the average refiners acquisition cost of imported crude oil
will remain at about $28 per barrel from 2015 to 2030 (2004 dollars), and
that the average refiners acquisition cost of imported low-sulfur light
crude oil will remain at about $34 per barrel over the same period. The AEO2006 low price case is somewhat lower than GIIs crude oil price path.
The AEO2006 high price case assumes that the average refiners acquisition
cost of imported crude oil will range between $72 and $90 per barrel from
2015 to 2030, and that the average refiners acquisitions cost of imported
low-sulfur light crude oil will range between $76 and $96 per barrel over
the same period. The crude oil prices in the AEO2006 high price case are
well above PIRAs projected levels. The AEO2006 low price case shows the
highest levels of total petroleum demand in 2015 and 2030, and the AEO2006 high price case shows the lowest. The projected demand reduction in the AEO2006 high price case also results in the least reliance on imports to
meet petroleum demand in 2015 and 2030. The DB projection shows the greatest
reliance on petroleum imports, because it assumes the lowest levels of
domestic crude oil and NGL production.
The coal projections for the AEO2006 reference case and economic growth
cases (Table 25) incorporate CAAA90, CAIR, and CAMR. EVAs forecast assumes
legislation similar to the Clear Skies Act but also includes a fee of $5
per ton on CO2 emissions, beginning in 2013. The AEO2006, PIRA, and GII
projections do not include assumptions about reductions in CO2 emissions
for the United States. In addition to environmental assumptions, differences
among the AEO2006, EVA, PIRA, and GII projections reflect variation in
other assumptions, including those about economic growth, the natural gas
outlook, and world oil prices.
While all the projections show increases in coal consumption over their
projection horizons, the AEO2006 reference case projects the highest level
of total coal consumption. Given its more restrictive environmental assumptions
after 2012 and an average economic growth rate of 2.5 percent per year
from 2004, EVA projects lower levels of coal consumption (11 percent lower
in 2025) than the AEO2006 reference case. The EVA and PIRA projections
for total coal consumption in the 2015-2020 period most closely resemble
those in the AEO2006 low economic growth case. GIIs projection, which
does not include a carbon tax, has the lowest projection of total coal
consumption. Although the GII projection shows 21 percent less total coal
consumption than the AEO2006 reference case in 2030, GIIs outlook for
coal consumption in the electric power sector in 2030 is virtually identical
to that in the AEO2006 low economic growth case.
In contrast to the AEO2006 reference case, the other projections show natural
gas with a larger share of electricity generation than coals. GII, PIRA,
and EVA expect imports of LNG to be greater than projected in the AEO2006 reference case. Although EVA and the AEO2006 reference case project similar
levels of generation in the electric power sector, the AEO2006 reference
case also projects 19 gigawatts of generation capacity at CTL plants by
2030, representing 11 percent of total coal consumption in 2030.
For coke plants, both GII and the AEO2006 reference case project declining
consumption of coal. EVA differs from the other projections and projects
an increase in coal consumption at coke plants, peaking at around 30 million
tons before falling to 26 million tons in 20252 million tons higher than
2004 consumption. In the GII projection, coke plants consume only 14 million
tons of coal in 2030, compared with 21 million tons in the AEO2006 reference
case. The AEO2006 reference case shows no change in industrial/other coal
consumption, whereas EVA projects a drop in industrial/other consumption,
to 60 million short tons in 2025.
In the GII projections, minemouth coal prices (electric power sector only)
appear to peak by 2010 and then fall below 2004 levels by 2030 (in real
dollars). The AEO2006 reference case shows a similar downward trend after
2010 in its average national minemouth price (all sectors combined) through
2020; however, prices rise after 2020, in response to substantial growth
in coal demand from the electric power sector and CTL. GIIs average delivered
price of coal to the electric power sector in 2030 is 19 percent lower
than the AEO2006 reference case (on a Btu basis). The average delivered
price of coal to the electricity sector in the GII forecast is still 13
percent lower (on a Btu basis) than the AEO2006 low economic growth case,
despite comparable levels of coal consumption in the electricity sector.
All the forecasts reviewed meet coal demand primarily through domestic
production. AEO2006 projects the largest increase in production over the forecast horizon, 51 percent
higher in 2030 than in 2004. As with consumption, the PIRA and EVA projections
for coal production most closely resemble those in the AEO2006 low economic
growth case. GII projects coal consumption levels for 2015, 2025, and 2030
that are more than 100 million tons less than projected in the AEO2006 reference case.
In all the projections, gross exports of coal represent a small and declining
part of domestic coal production. EVA projects the most exports, 30 million
tons in 2025, and the other projections are around 20 million tons. The AEO2006 reference case shows coal exports falling to 17 million tons in
2030, and GII projects 22 million tons. In the AEO2006 reference case,
the export share of total U.S. coal production falls from 4 percent in
2004 to roughly 1 percent in 2030. Currently, coal is the only domestic
U.S. energy resource for which exports exceed imports. All the projections
expect the United States to become a net importer of coal over the projection
period. GII projects the lowest level of coal imports, only 14 million
tons in 2030. The AEO2006 reference case projection for coal imports in
2025 is 23 million tons higher than the EVA projection, which is the next
Forecast Comparisons - Tables