Issues
In Focus.
Natural Gas Consumption in Canadian Oil Sands
Production
In recent years, extensive investment has gone into
the development of Albertas oil sands. In 2002, Canadas
crude bitumen production from oil sands averaged 790,000 barrels
per day, while conventional crude output was 2,140,000 barrels per
day (including natural gas liquids). Natural gas is used both to
extract the bitumen from the sand and to convert the bitumen into
syncrude. Currently, oil sands operations consume approximately
330 billion cubic feet per year of natural gas.
Canadian oil producers have announced a number of
new oil sands projects and expansions to existing oil sands facilities.
The question has arisen as to whether these existing and future
facilities will raise Canadas gas consumption by a significant
amount, thereby reducing the amount of Canada gas production, which
is available for export to the United States. This discussion will
briefly examine this issue.
Most of the existing and proposed oil sands projects
are located in the east-central portion of Alberta and are dispersed
along a roughly north-south axis of about 200 miles in length. The
Canadian oil sands consist of a mixture of sand, bitumen, and water.
Based on existing facilities, and project announcements for expansions
and new oil sands production facilities, EIA projects total oil
sands bitumen production to be 1.7 and 3.3 million barrels per day
in 2010 and 2025, respectively (Table 13). In 2010, about 52 percent
of the bitumen is projected to be surface mined, and the remaining
48 percent is projected to be produced through in situ production
[49]. In 2025, approximately 57 percent of the oil sands
bitumen is projected to be surface mined, and 43 percent is projected
to be produced through the in-situ production method.
To produce synthetic crude oil, the bitumen can be
either partly or totally petroleum coked or hydrocracked. Petroleum
coking requires less process energy than hydrocracking and does
not require a hydrogen feedstock, but 100 barrels of bitumen yields
only 79 barrels of syncrude. Hydrocracking, on the other hand, requires
both more process energy and a hydrogen feedstock, but 100 barrels
of bitumen produces about 106 barrels of syncrude.
There are three potential fuels that can be used
either exclusively or in part to produce oil sands syncrude, namely,
natural gas, produced bitumen, or petroleum coke, the latter of
which is a process byproduct. Depending upon an oil sands facilitys
design flexibility, the syncrude producer can change the slate of
inputs, such as natural gas, and the slate of outputs (e.g., syncrude,
petroleum coke) so as to maximize the profit margin associated with
the production and upgrading of bitumen into syncrude, based on
the cost/price of both the inputs and outputs. Consequently, the
consumption of natural gas in these upgrading facilities is expected
to change over time as relative prices change. Moreover, the input/output
flexibility of any particular bitumen upgrading facility can be
enhanced in the future, if prices warrant. Consequently, if natural
gas prices were sufficiently high and oil prices sufficiently low,
syncrude producers could theoretically eliminate natural gas consumption
entirely through the exclusive use of bitumen and petroleum coke
to provide the energy and feedstocks to produce and upgrade the
bitumen.
Carbon dioxide emissions might also play a role in
determining the proportions of natural gas, bitumen, and petroleum
coke used for oil sands production and processing. On December 17,
2002, Canada ratified the Kyoto Protocol, which obligates it to
reduce carbon dioxide emissions to 6 percent below their 1990 level.
Because petroleum coke and bitumen release more carbon dioxide when
burned than natural gas does, Canadas Kyoto Protocol obligation
could limit the use of petroleum coke and bitumen in the processing
of bitumen from Canadian oil sands.
If natural gas were to be used exclusively to produce
and convert bitumen into syncrude, then the following volumes of
natural gas would be consumed to perform each of the following processes:
- Surface mine 1 barrel of bitumenapproximately 131 cubic
feet
- In situ production of 1 barrel of bitumen1,000
to 1500 cubic feet
- Petroleum coking 1 barrel of bitumenapproximately 168
cubic feet
- Hydrocracking 1 barrel of bitumenapproximately 490 cubic
feet.
The natural gas consumption estimates presented in
Table 13 assume that natural gas is the only energy and feedstock
source for the production and upgrading of bitumen into syncrude.
Table 13 assumes that the in situ production of bitumen requires
1,250 cubic feet of natural gas per barrel of bitumen. The first
estimate (Case I) assumes that the bitumen is exclusively petroleum
coked to create syncrude, while the second (Case II) assumes that
the bitumen is exclusively hydrocracked. Of course, if oil sands
producers were to extensively use bitumen and petroleum coke to
provide most of the process energy and hydrogen feedstock requirements,
then the actual natural gas consumed in future years would be considerably
less, potentially as low as zero.
In conclusion, given the potential fuel flexibility
of oil sands production facilities, the question of whether Canadian
oil sands production will consume significant volumes of natural
gas is not easily answered. The answer to this question will depend
not only on the relative prices of syncrude and natural gas, but
also on the degree to which oil sands producers build fuel-flexible
facilities. Consequently, the actual outcome could be as high as
1.3 trillion cubic feet per year or as low as zero.
Natural Gas Consumption in the Industrial
Sector
Natural gas consumption in the U.S. industrial sector
increased by 1.6 percent per year on average from 1990 to 2000,
fell sharply in 2001, and continued to decline in 2002. During the
1990s, the industrial sector accounted for slightly less than 37
percent of total U.S. natural gas consumption, peaking in 1997 at
8.7 quadrillion Btu or 37.5 percent of the total. In the AEO2004
reference case, industrial natural gas use is projected to return
to a path of steady increase after 2003, averaging 1.5-percent annual
growth from 2002 to 2025 (Figure 21). Total natural gas consumption
for industrial uses is projected to reach 10.6 quadrillion Btu in
20253.1 quadrillion Btu higher than in 2002based on
projected growth in industrial output and modestly increasing natural
gas prices over the forecast period.
Within the industrial sector, natural gas use for
combined heat and power (CHP) applications is projected to increase
by 2.6 percent per year, for feedstocks by 0.8 percent per year,
and for boiler fuel and direct uses by 1.4 percent per year from
2002 to 2025 (Figure 22). With total industrial output (value of
shipments) increasing by 2.6 percent annually over the same period,
the natural gas intensity of industrial output in 2025 is projected
to be 21 percent lower than in 2002.
As a result of the economic recession that began in
2001 and the rise in natural gas prices since 2000, some industry
observers have concluded that segments of the U.S. industrial sector
have permanently reduced output through closures of manufacturing
plants, and that the result will be a permanent reduction in demand
for natural gas. Others note that similar industrial reactions to
sharp increases in gas prices and to recessions are not unprecedented,
and that the recent drop in demand is likely to be temporary [50]
once industrial production growth resumes. A history of the recent
relationship between industrial production and natural gas consumption
is shown in Figure 23. In the absence of severe, multi-year recessions
in the industrial sector and sustained higher prices for natural
gas, it is reasonable to expect industrial output and natural gas
consumption to increase in the future.
AEO2004 projects little or no growth in industrial
demand for coal, and most of the projected increase in demand for
petroleum products is for asphalt and petroleum byproducts. Natural
gas remains the fuel of choice in the industrial sector and will
continue to fire most CHP applications. In the AEO2004 reference
case, industrial natural gas prices are projected to rise by 1.4
percent per year on average, to $5.00 per million Btu in 202560
cents lower in constant 2002 dollars than the 2003 price (Figure
24).
Some portions of the industrial sector, however, are especially
sensitive to natural gas pricesparticularly those that use
natural gas as a feedstock, such as nitrogenous fertilizer production,
organic chemical production, and petrochemical production. For example,
0.7 quadrillion Btu of natural gas was used for feedstocks in the
chemical industry in 1998 [51], accounting for about 10 percent
of total natural gas consumption in the manufacturing sector. Petroleum-based
products, however, were the largest source of industrial feedstock
(for organic chemicals, plastics, synthetic rubber, and petrochemicals),
amounting to 3.1 quadrillion Btu, more than four times the quantity
of natural gas used as a feedstock in 1998.
One sector particularly sensitive to higher natural
gas prices is the nitrogenous fertilizer industry. Natural gas costs
account for 70 to 80 percent of the cash cost of fertilizer: production
of a ton of ammonia uses 33.5 million Btu of natural gas [52].
At the average industrial natural gas price during the 1990s, the
embodied cost of energy per ton of ammonia equates to about $120.
At the estimated average industrial natural gas price in 2003 ($5.60
per million Btu), the embodied cost of energy is $188 per tona
57-percent increase. This significant increase in cost, if passed
through completely, would amount to only 9.9 cents per bushel of
corn, or 4 percent of the total average price of $2.35 per bushel
in 2002 [53]. Large percentage increases in costs for ammonia
production do not, therefore, necessarily result in proportional
increases in the price of agricultural products.
Higher production costs tend to be passed through
quickly to the price of ammonia [54], although the amount
of the pass-through can be reduced by competition from imports.
Imports of ammonia historically have accounted for about 20 percent
of U.S. demand. Their impact on reducing the amount of pass-through
costs can, however, lag over time.
The demand for natural gas as a feedstock to produce
ammonia is determined largely by the quantity of ammonia produced,
because petroleum-based fuels are not generally a viable economic
alternative [55]. In 1998, the nitrogenous fertilizer industry
consumed 338 trillion Btu of natural gas as a feedstock [56].
An additional 234 trillion Btu was consumed for process heating.
In principle, the portion of the industrys natural gas consumption
used for process heating could be switched to another fuel; however,
in 1994 (the most recently available data for fuel switching), the
nitrogenous fertilizer industry reported that only 3.1 trillion
Btu (1 percent) of its natural gas use was switchable [57].
For at least two decades, the nitrogenous fertilizer
industry in the United States has been consolidating [58].
From 89 plants with an average annual capacity of 171,000 metric
tons in 1970, the number of plants fell sharply after 1980, and
the average capacity of the remaining plants more than doubled.
In 2002 there were only 37 plants operating, with an average capacity
of 451,000 metric tons. Total industry capacity in 2002, at 16.7
million metric tons, was only slightly higher than in 1970 (15.2
million metric tons).
The consolidation, or even permanent closure, of nitrogenous
fertilizer plants has no meaningful impact on U.S. natural gas markets,
because the plants account for only a small portion of total U.S.
gas consumption (0.5 quadrillion Btu out of 21.1 quadrillion
Btu total in 1998). In addition, permanent closure of fertilizer
plants in response to a temporary increase in natural gas prices
is unlikely. For example, several producers temporarily idled their
plants in the first quarter of 2002, but most of the idled capacity
was back on line by the fourth quarter of the year [59].
Also, the largest U.S. producer of nitrogenous fertilizer (Farmland
Industries, an agricultural cooperative), which declared bankruptcy
in early 2002 [60], continued to operate most of its plants.
In the AEO2004 reference case, industrial sector
output is projected to grow by 2.6 percent annually from 2002 to
2025, the same growth rate experienced in the 1990s. The bulk chemical
industry is projected to grow by 1.6 percent annually, slightly
below its 1.8-percent growth rate during the 1990s. Agriculture
is projected to grow by 1.2 percent annually, leading to a projected
0.9-percent annual growth rate for agricultural chemical production,
of which nitrogenous fertilizer is a part[61]. In 2025, the
value of agricultural chemical shipments is projected to be $24 billion,
approximately equal to their 1997 value (Figure 25).
Natural Gas Consumption for Electric Power
Generation
Data from EIAs Form EIA-860 survey, Annual
Electric Generator Report, show a dramatic increase in additions
to U.S. electricity generation capacity over the past 3 years. In
2000, 2001, and 2002 more than 141 gigawatts of new generating capacity
was constructedfar more than in any previous 3-year period.
Virtually all of that new capacity uses natural gas as the primary
fuel for electricity generation (Figure 26).
Given the recent pace of capacity additions, it is
not surprising that the amount of electricity produced from natural
gas has increased substantially; however, natural gas consumption
in the electric power sector has not increased as rapidly, because
the efficiency of gas-fired generation has improved significantly
(Figure 27). From 1995 to 2002, natural-gas-fired generation in
the power sector increased by 43 percent, but natural gas consumption
increased by only 31 percent. Notably, the gap between growth in
natural-gas-fired generation and natural gas consumption by power
producers began to appear in 2000, when the first wave (27 gigawatts)
of the recent surge in capacity expansion occurred.
The role of natural gas in the electric power sector
is expected to continue growing for the foreseeable future. At the
same time, the disparity between increases in gas-fired generation
and in the amount of natural gas consumed by power producers is
also expected to continue growing. In addition to the amount of
new gas-fired generating capacity added, other factors that will
affect the amount of natural gas used to generate electricity over
the coming decades include: the rate of growth in electricity sales;
the efficiencies of new gas-fired plants relative to those of older
plants; and the price of natural gas relative to the prices of other
fuels, particularly coal.
Relative to the amount of generating capacity operating
in 1999, additions over the 2000-2002 period amounted to an increase
of 18 percent. Over the same period, electricity sales grew by only
5 percent. Consequently, many of the plants added in recent years
are unlikely to be used at full capacity in the early years of their
operation. Moreover, an additional 45 gigawatts of new capacity
is expected to be added in 2003, all but 2 gigawatts of which will
use natural gas. With growth in electricity sales expected to continue
at a much more modest pace, the recent disparity between generating
capacity growth and sales growth is expected to widen in the near
term, and it could be many years before much of the newly added
capacity is used intensively.
Where new natural gas plants are used, their generation
will often displace generation that would have come from older,
less efficient oil- and gas-fired generators. The natural-gas-fired
plants that have been added in recent years are much more efficient
than older plants. For example, new combined-cycle plants have operating
efficiencies between 45 and 50 percent, whereas the efficiencies
of older steam plants generally are 33 percent or less. Accordingly,
a new plant could generate the same amount of electricity as an
older plant while consuming 27 percent less natural gas, or could
use the same amount of gas as an older plant while generating 36
percent more electricity [62]. The efficiency gap
between old and new natural-gas-fired power plants is expected to
lead power companies to retire many of their older plants, because
it will no longer be economical to maintain them. The newer plants,
using substantially less fuel, will provide the power that the older
plants were generating.
In the AEO2004 reference case forecast, natural
gas consumption in the electric power sector is projected to continue
to increase; however, the gap between the growth in natural gas
generation and natural gas consumption in the power sector is also
projected to widen (Figure 28). In 2025, the amount of electricity
generated from natural gas is projected to be 166 percent greater
than it was in 1995, but the amount of natural gas consumed for
electricity production is projected to increase by only 98 percent.
Over the same period, the average efficiency of all generators using
natural gas is projected to increase from 33 percent to 45 percent.
Finally, in the later years of the forecast, rising
natural gas prices are expected to make new coal-fired capacity
economically competitive. When new coal-fired generating plants
are added, they will be less expensive to operate than gas-fired
plants, including those currently coming into service, and they
are expected to be used for baseload generation, meeting customer
needs around the clock. The capacity factor for all oil- and gas-fired
capacity is projected to decline initially (Figure 29) because of
the surge of capacity additions in 2002 and 2003, then rise to about
28 percent in 2018, and then decline as new coal-fired plants
come on line. In the AEO2004 forecast, the end result is
that natural gas consumption in the electric power sector is projected
to continue growing more slowly than either additions of gas-fired
capacity or generation using natural gas
.
Notes and Sources
Released: January 2004
|