Factors affecting the relationship between crude oil and natural gas prices
Background
Over the 1995-2005 period, crude oil prices and U.S. natural gas prices
tended to move together, which supported the conclusion that the markets for the two commodities were
connected. Figure 26 illustrates the fairly stable ratio over that period
between the price of low-sulfur light crude oil at Cushing, Oklahoma, and
the price of natural gas at the Henry Hub on an energy-equivalent basis.
The AEO2010 Reference and High Oil Price cases, however, project a significantly
longer and persistent disparity between the relative prices of low-sulfur
light crude oil and natural gas on an energy-equivalent basis [51]. The
apparent disconnect in prices between seemingly similar commodities varies
over a wide range between 2010 and 2035 [52]. Over much of the projection
period in the Reference case, the crude oil price is about 2.8 times the
natural gas price on an energy equivalent basis—115 percent higher than
the historical average price ratio of 1.3 from 1995 to 2005. In the High
Oil Price case, the ratio widens to as much as 4.8; in the Low Oil Price
case, it narrows from nearly 3.0 in 2009 to 1.1 in 2035.
Such an apparent lack of responsiveness of natural gas prices to changes
in crude oil prices in all cases reflects the changes that have occurred
in the underlying uses of the two commodities. The divergence of crude
oil and natural gas markets also reflects the fact that opportunities for
the substitution of natural gas for crude oil products are limited by the
large infrastructure investments that would be required to allow substitution
on a significant scale and bring the prices of the two commodities closer
together in the U.S. market in the Reference and High Oil Price cases. In
the absence of such investments, EIA expects the gap between oil and natural
gas prices in U.S. energy markets to remain wide.
Opportunities to substitute natural gas for petroleum
In the United States, the capability to substitute natural gas supplies
directly for petroleum, particularly in the electric power sector, has
eroded over time. In 1978, 4.0 quadrillion Btu of petroleum was consumed
to produce electricity, representing nearly 17 percent of total energy
use for U.S. electricity generation, as compared with 14 percent for natural
gas [53]. In 2008, only 0.5 quadrillion Btu of petroleum was consumed for
electricity generation, representing 1.2 percent of total energy use for
generation [54, 55], while natural gas has grown to 17 percent of generation.
The trend has been similar in the commercial and industrial sectors where
there are a declining number of opportunities to substitute natural gas
for petroleum.
Still, there are potential opportunities for natural gas to displace petroleum.
First, direct use of natural gas in the U.S. transportation sector could
provide an opportunity for substitution. Second, natural gas could be exported
to countries where petroleum is widely used for thermal applications. Third,
natural gas can be converted directly to petroleum-like liquid fuels that
could be substituted for diesel and gasoline in the existing vehicle fleet
using the existing distribution infrastructure.
The physical properties of natural gas are such that it is more difficult
and costly than liquid fuels to transport and consume. As shown in Figure
27, the energy density of natural gas is much lower than that of most liquid
fuels. To match the energy equivalent of a 1-gallon container of diesel
fuel, a balloon of natural gas at atmospheric pressure would have to be
nearly a thousand times larger than the gallon container. At a pressure
of 3,600 pounds per square inch (psi), however, which is the pressure rating
for the fuel tanks used in CNG vehicles, only 4 times as much space is
required to match the energy equivalent of 1 gallon of diesel fuel. And
when the gas is converted to LNG by chilling to about -260 degrees Fahrenheit,
its energy density increases to the point where it requires only 50 percent
more volume to match the energy content of diesel fuel. However, the materials
used for the handling and storage of LNG differ significantly from those
used for CNG or petroleum-like liquid fuels.
An expanded market for CNG or LNG would require additional investment in
vehicles and infrastructure for compression and storage of CNG or for liquefaction
and storage of LNG. Some of the issues, challenges, and opportunities surrounding
the use of natural gas as a substitute for diesel fuel are described in
the Issues in Focus section, “Natural gas as a fuel for heavy trucks: Issues
and incentives.”
Barriers to U.S. exports of LNG
World crude oil and natural gas prices could converge if barriers to the
flow of natural gas between U.S. and world markets were eliminated through
the combined use of the existing pipeline network, existing LNG terminals,
and investment in new U.S. LNG liquefaction capacity (and possibly LNG
tankers) to allow exports of U.S. natural gas when it is economical. Currently,
there is one liquefaction facility in Alaska that exports LNG from the
United States. Investment in new U.S. liquefaction capacity would face
significant risk, however, because there are large quantities of “stranded
gas” in remote regions of the world that can be priced well below the expected
cost of resources in the lower 48 States.
Potential for production of liquid fuels from natural gas
Another opportunity to substitute natural gas for crude oil would be to
convert it to petroleum-like liquid products similar to gasoline and diesel
fuel, for use in the liquid fuel infrastructure and end-use equipment.
Such a transformation is possible through use of the GTL process.
There are several GTL processes, the best known using a Fischer-Tropsch
reactor. The reactor produces a paraffin wax that is hydrocracked to form
liquid products that resemble petroleum liquids. Distillates, including
diesel, heating oil, and jet fuel, are the primary products, making up
50 to 70 percent of the total volume produced, and naphtha usually represents
about 25 percent of the volume. The process efficiency is about 57 percent
(43 percent of the energy content of the natural gas is lost in the process)
[56]. Thus, the price ratio of liquid products to natural gas would have
to exceed about 1.8 to justify operation of the plant, excluding consideration
of other operating costs and the cost of capital investment. To appreciate
the price risk faced by investors, one can consider the effects of recent
fluctuations in energy prices on investments in U.S. natural gas turbine
and combined-cycle generating units and ethanol production facilities [57].
Indeed, AEO2010 examines the potential impacts of lower energy prices in
the Low Oil Price case, which shows the ratio of crude oil prices to natural
gas prices declining to 1.1 in 2035, indicating that if any GTL plants were
built they would not be operated under those price conditions.
The technologies and equipment used in the best-known GTL technology are
similar to those that have been employed for decades in methanol and ammonia
plants, and most are relatively mature; however, the scale on which previous
GTL plants have been implemented is relatively small. The newest GTL plants
have been expanded to much larger sizes, including one in excess of 100,000
barrels per day, to take advantage of economies of scale, but recent attempts
to build projects at those larger sizes have encountered technology or
project execution risks [58]. Currently, there are four GTL plants in operation
worldwide, with 96,200 barrels per day of total capacity [59]. In addition,
two projects with 174,000 barrels per day of capacity are under construction
or ready for startup [60]. However, the construction of GTL plants at sites
with available stranded gas reserves has been limited, indicating investor
reluctance to pursue this option fervently, especially when investments
in less capital-intensive LNG capacity are possible. Indeed, some GTL projects
have been canceled or deferred in the past few years [61].
The overnight capital costs for a new GTL plant situated on the U.S. Gulf
Coast would range from $50,000 per barrel-stream day of capacity [62] to
an estimated $104,000 per barrel-stream day [63]. Accordingly, a relatively
modest unit with a capacity of 34,000 barrels per day represents an estimated
overnight capital cost [64] of $1.7 billion to $3.5 billion. With financing
included, the estimated total investment would be $2.2 billion to $4.4
billion. In addition, construction of the facility would take 4 years or
more, imposing further market risk. The risk-adjusted discount factor used
by investors will be critical to determining whether investors would proceed
with GTL investments.
Figure 28 shows the maximum “breakeven” average price of natural gas that
could be tolerated over a 10-year plant operating period [65] in order
to justify the risk associated with investing in a GTL facility, based
on the range of capital costs discussed above and a 10-percent hurdle rate
[66]. Profitable cases lie below the line. At $100 per barrel for crude
oil, the breakeven price for natural gas that would justify investment
in a GTL facility is -$1.20 to $5.80 per million Btu. At higher crude oil
prices, the range of the breakeven natural gas price also rises. At a crude
oil price of $200 per barrel, the breakeven price for natural gas is $10.20
to $17.30 per million Btu. At a crude oil price of $60 per barrel, the
breakeven natural gas price ranges from -$5.80 to $1.30 per million Btu,
illustrating the substantial impact of oil price uncertainty on the profitability
of investment in a GTL facility.
Figure 28 also shows how investment in a GTL facility would fare with the
natural gas and crude oil price projections in the AEO2010 Reference, Low
Oil Price, and High Oil Price cases. With the prices in the Low Oil Price
case, GTL is a poor investment. With the prices in the Reference case,
GTL is a marginal investment. Only with the highest prices in the Reference
case and the low end of GTL plant costs do the breakeven economics favor
the project. In the High Oil Price case, however, the combination of higher
crude oil prices and lower natural gas prices implies that investment in
a GTL plant on the U.S. Gulf Coast could be profitable.
A large investment in GTL would be needed in order to produce an appreciable
effect on worldwide prices for crude oil and U.S. natural gas. Construction
of sufficient new GTL capacity to affect world crude oil prices, about
1 million barrels per day, would require a total investment between $50
billion and $135 billion. That level of capacity would still represent
only 1.2 percent of the 85.9 million barrels per day of the world’s estimated
total liquids production in 2007 [67], and less than 1 percent of projected
2035 production in the Reference case [68].
Another option is the potential use of stranded natural gas in Alaska to
produce GTL. Because of Alaska’s severe weather conditions, construction
of GTL (or any other) facilities is likely to be much more expensive than
the construction of GTL plants on the U.S. Gulf Coast or in the Middle
East. Some estimates suggest that doubling the construction costs and extending
the construction period by at least 2 years would be reasonable assumptions.
Construction of GTL facilities in Alaska, therefore, seems unlikely given
the cost uncertainties mentioned above and the crude oil price projections
in the AEO2010 Reference case.
Looking forward
A large disparity between crude oil and natural gas prices, as projected
in the AEO2010 Reference and High Oil Price cases, will provide incentives
for innovators and entrepreneurs to pursue opportunities that, in the longer
term, could increase domestic or international markets for U.S. natural
gas. For example, a scenario with relatively high oil prices would tend
to increase the value of CO2 used for EOR as well as GTL production. Because
GTL processing plants can accommodate natural gas feedstocks with relatively
high CO2 content and can target fields smaller than those required for
LNG production, such circumstances would provide incentives for the development
of smaller GTL systems that produce both liquid products and a valuable
CO2 co-product. Because EIA cannot predict whether or when such innovations
might arise, they are not included in the AEO2010 analysis cases.
Footnotes:
51. Low-sulfur crude oil priced for delivery at Cushing, Oklahoma, and natural gas priced at the Henry Hub.
52. While simple price comparisons assume the same point of sale in retail markets, crude oil and natural gas price comparison reflects unprocessed prices at a supply node. To describe the ratio in terms of the retail market, multiple delivered petroleum product prices would have to be compared to delivered natural gas prices in the same markets. Because making the com-parison at the detailed retail level would require a far more complex set of comparisons involving different tax structures and processing costs, the comparison on the supply side is a useful, if somewhat oversimplified, comparison that accounts for most of the price diver-gence described.
53. U.S. Energy Information Administration, Annual Energy Review 2008, DOE/EIA-0384(2008) (Washing-ton, DC, June 2009), Table 8.4a, web site www.eia.doe. gov/emeu/aer.
54. U.S. Energy Information Administration, Annual Energy Review 2008, DOE/EIA-0384(2008) (Washing-ton, DC, June 2009), Table 8.4a, web site www.eia.doe. gov/emeu/aer.
55. Consistent with that reduction has been the abandon-ment of large-scale storage by electric utilities of petroleum products for generation.
56. U.S. Energy Information Administration, Assump-tions to the Annual Energy Outlook 2010, DOE/EIA- 0554(2010) (Washington, DC, April 2010), p. 137, web site www.eia.gov/oiaf/aeo/assumption/pdf/0554 (2010).pdf, p. 137.
57. Favorable natural gas prices in the late 1990s led to a surge in investments in turbine and combined-cycle units, to a level that could not be supported when nat-ural gas prices increased. Many of those purchases were resold at discounts, or installation was postponed for years.
58. J. Macharia, “Sasol Oryx GTL Plant Has Problems, Shares Hit,” Reuters (May 22, 2007), web site www. reuters.com/article/idUSL2206595120070522.
59. Sasol I (2,500 barrels per day), Mossel Bay (45,000 bar-rels per day), Bintulu (14,700 barrels per day), and Oryx (34,000 barrels per day).
60. Qatar’s Pearl GTL (140,000 barrels per day), which is currently anticipated to begin production in 2011, and Nigeria’s Escravos GTL (34,000 barrels per day), which is also slated for a 2011 startup. See “Pearl GTL Sets Milestone as Steam Boilers Start Up,” Gulf Times (not dated), web site www.gulf-times.com/site/topics/ article.asp?cu_no=2&item_no=345533&version=1& template_id=48; and J. Macharia and M. Whittaker, “Update: 2-Sasol’s Nigeria Project Costs Up, Loses on Oil Hedge,” Reuters (July 29, 2008), web site http:// uk.reuters.com/article/idUKL921344320080729.
61. Including Exxon’s Palm GTL in Qatar (154,000 bar-rels per day), which was cancelled in 2007. See National Petroleum Council, “Facing the Hard Truths About Energy, Topic Paper #9, Gas To Liquids (GTL)” (July 18, 2007), page 2, web site www. npchardtruthsreport.org/topic_papers.php.
62. U.S. Energy Information Administration, Assump-tions to the Annual Energy Outlook 2010, DOE/EIA- 0554(2010) (Washington, DC, April 2010), p. 137, web site www.eia.gov/oiaf/aeo/assumption/pdf/0554 (2010).pdf, p. 137.
63. S. Reed and R. Tuttle, “Shell Aims for ‘New Nigeria’ as $19 Billion Qatar Plant Starts” Bloomberg Business Week (March 3, 2010), web site www.businessweek. com/news/2010-03-03/shell-aims-for-new-nigeria-as- 19-billion-qatar-plant-starts.html.
64. Overnight capital costs exclude financing costs during construction.
65. A 10-year operating period is assumed as a maximum private-sector investment horizon for such a project. The 10-year period was chosen based on input from an EIA workshop held in 2007 that looked at capital investment decisionmaking. The papers resulting from that workshop can be found at www.eia.gov/ oiaf/emdworkshop/model_development.html. It is pos-sible that a longer operating period would be appropri-ate with public financing or loan guarentees. This would have the effect of lowering the effective breakeven levels discussed in the article.
66. hydrocarbons-technology.com, “Pearl Gas-to-Liquids Project, Ras Laffan, Qatar” (not dated), web site www. hydrocarbons-technology.com/projects/pearl.
67. U.S. Energy Information Administration, Annual Energy Review 2008, DOE/EIA-0384(2008) (Washing-ton, DC, June 2009), Table 11.10, web site www.eia. doe.gov/emeu/aer.
68. U.S. Energy Information Administration, Short Term Energy Outlook (March 9, 2010 Release), Table 3C, web site www.eia.gov/emeu/steo/pub/contents. html. |
