EIA Administrator's Presentations

North American Gas Strategies Conference
Calgary, Alberta

EIA Administrator Jay Hakes,
October 19, 1998

Outlook for Natural Gas Supply

These projections are from the Annual Energy Outlook 1998.

This talk will briefly cover supply, demand, and prices of natural gas through 2020. EIA doesn't represent the complete Canadian gas market in its model, but we include projections of Canadian consumption and exports to the United States in the outlook.

The Annual Energy Outlook is a product of the Energy Information Administration. EIA is an independent statistical and analytical agency within the Department of Energy.

Assumptions are critical. The projections in AEO98 are not statements of what will happen but of what might happen, given certain assumptions. The reference projections are business-as-usual forecasts, given known technology and technology trends, demographic trends, and current laws and regulations. EIA does not propose, advocate, or speculate on changes in laws and regulations. So, one of our key assumptions is that all current laws and regulations remain as enacted.

That means, for example, that no new initiatives are assumed for the control of greenhouse gas emissions. And no new legislation is assumed for the deregulation of the electric utility industry.

Natural Gas Production, Consumption, and Imports, 1970-2020

Natural gas supply and demand is projected to grow significantly over the next 20 years.

At 32.2 trillion cubic feet in 2020, natural gas consumption will account for 28 percent of total domestic energy consumption, compared to 24 percent in 1996. Gas consumption is expected to increase at 1.6 percent annually from 1996 through 2020, mainly because of growth in gas-fired electricity generation.

Gas production increases at an annual rate of 1.5 percent over the forecast, rising from 19 Tcf in 1996 to 27 Tcf in 2020. Growing production reflects rising wellhead prices, relatively abundant natural gas resources, and improvements in technologies, particularly for offshore and unconventional gas.

Net imports are expected to climb from 2.8 Tcf in 1996 to 4.9 Tcf in 2020. Imports continue to be dominated by pipeline imports from Canada over the forecast.

Natural Gas Consumption by Sector, 1970-2020

Gas consumption in the industrial sector, which includes cogenerators, increases by 1.2 Tcf over the forecast. Combined, by 2020, the residential and commercial sectors add 1.2 trillion cubic feet from 1996.

The big story in natural gas consumption is in electricity generation. Gas consumption by electric generators, not including industrial cogenerators, more than triples during the forecast, from 3 trillion cubic feet in 1996 to 9.9 trillion cubic feet in 2020.

Growth is spurred by increased utilization of existing gas-fired power plants and the addition of new turbines and combined-cycle facilities.

In 1996 electricity generators are the fourth-largest natural gas consuming sector. By 2020, however, the enormous growth in gas-fired generation makes electricity generators the largest gas-consuming sector.

Electricity Generation and Cogeneration Capacity Additions by Fuel Type, 1996-2020

Of the 403 gigawatts of new capacity needed from 1996 to 2020, 85 percent is projected to be combined-cycle or combustion turbine technology fueled by natural gas or both oil and gas.

More than 1,300 new power plants are expected to be built between now and 2020. More than 1,100 of them are projected to be gas-fired.

Restructuring of the electric generation industry is expected to open up new opportunities for gas-fired generation, as nonutilities move toward less capital intensive projects.

The reduction in baseload nuclear capacity also has an impact on the electricity outlook after 2010. Almost half of the new combined-cycle capacity projected over the entire forecast is expected to be brought on line in those 10 years, due in part to nuclear retirements.

Costs for New Electricity Generation Plants, 2005 and 2020

Gas plants are less capital-intensive than coal, nuclear, or renewable electricity generation plants and this advantage overwhelms the lower costs of the other fuels.

By 2020 efficiencies for combined-cycle units are more than 50 percent, compared with 38 percent for coal-steam units.

By 2020 construction costs for combined-cycle units are a little more than one-fourth of those for coal-steam plants.

Electricity Generation by Fuel, 1996-2020

As a result of the increase in natural gas plant construction, generation from natural gas-fired plants is expected to more than triple from 1996 to 2020, from 456 to 1,583 billion kilowatthours. Most of the increase is in utility generation, rather than cogeneration.

In percentage terms, generation from gas-fired power plants shows the largest increase in the forecast, growing from 13 percent of generation in 1996 to 33 percent of generation in 2020.

Gas-fired generation is projected to overtake nuclear power as the Nation's second-largest source of electricity in 2002.

Coal-fired power plants are expected to remain the dominant source of electricity through 2020, increasing from 1,796 to 2,304 billion kilowatthours. However, increased natural gas generation causes coal's share to decline from 52 percent in 1996 to 49 percent in 2020.

Natural Gas Production by Source, 1970-2020

The increase in U.S. natural gas production over the forecast comes primarily from lower 48 onshore conventional nonassociated sources. Conventional onshore production accounted for 40 percent of total U.S. domestic production in 1996 and is expected to increase to 45 percent in 2020.

Offshore production, mainly from wells in the Gulf of Mexico, also rises. Innovative, cost-saving technology and large finds, particularly in the deep waters of the Gulf, have encouraged interest in this area. Lower-48 offshore gross withdrawals at 5.6 tcf in 1996 were the highest yet recorded.

Unconventional gas production increases at the fastest rate of any other source over the forecast period, largely because of expansion of tight sands gas production in the Rocky Mountain region.

Alaska natural gas production rises gradually over the forecast. But Alaskan gas is not expected to affect the lower-48 markets, because prices are not high enough to cover delivery costs, including transportation, to the lower-48 States.

Lower 48 Successful Gas Wells by Technology, 1970-2020

Natural gas well completions are one of the important drivers of production. In the National Energy Modeling System, as in the real world, drilling produces reserve additions, which increase proved reserves. The larger the proved reserves, the easier it is to produce natural gas.

Rapid and slow technology cases were created to assess the sensitivity of the projections to changes in the economically recoverable oil and gas resources, exploration and development costs, and finding rates as a result of technological progress.

The rapid and slow technology cases were created by varying by one standard deviation from their statistically estimated values, parameters that represent the effects of technological progress on U.S. lease equipment, operating, and drilling costs. Statistically estimated values for U.S. finding rates were similarly varied. Parameters for growth in the U.S. undiscovered economic resource base were also varied.

With rising prices and generally declining drilling costs, lower 48 natural gas well completions in the reference case are expected to increase from 9,200 to 12,400 from 1996 to 2020.

Well completions are higher in the slow technology case, because of the higher prices in this case. Increased drilling is required to supply natural gas in volumes comparable to the reference case. The converse is true for the high technology case. In 2020 gas drilling is 18 percent higher in the slow technology case and 14 percent lower in the rapid technology case than the reference case. More than 14,600 wells are completed in the slow technology case and more than 10,600 in the rapid case.

Natural Gas Reserve Additions by Technology, 1970-2020

Increased well completions in the reference case are expected to lead to increased reserve additions, at least through the middle years of the forecast. Gas reserve additions peak at 25.3 trillion cubic feet in 2010 in the reference case. A couple years after the peak, production begins to exceed reserve additions, putting pressure on prices as proved reserves decline.

The relatively high levels of annual reserve additions through 2020 reflect an extension of recent technological improvements in exploration and development for a sustained period of unprecedented duration.

By 2020 more than 85 percent of the lower 48 nonassociated natural gas resources estimated to be technically recoverable from conventional resources using current technology are projected to be discovered.

Developments in the past year have shown the power of technological improvements. New field discoveries in 1997 were 2.7 tcf, up substantially from 1996 and twice the prior 10-year average. Total discovery volumes per exploratory well were up in 1997, more than four times the rates of the early 1980's.

In spite of more gas well completions in the slow technology case than in the reference case, reserve additions are lower in the slow technology case than in the reference case, because productivity is lower.

Natural Gas Reserves by Type, 1995-2020

Total U.S. proved reserves of natural gas were up in 1997 for the fourth year in a row to 167.2 trillion cubic feet. This upward trend is expected to continue through 2012. After that, reserves begin to decline, as depletion begins winning the race against technological improvement.

Most types of lower 48 reserves are expected to grow for at least part of the projection period. Offshore reserves grow the fastest, from 34 tcf in 1996 to 49 tcf in 2020, peaking at 50 tcf in 2017. Conventional onshore reserves grow the second fastest, and unconventional onshore reserve growth is third.

Associated-dissolved onshore gas reserves, which depend on oil supply activity, are expected to decline over the forecast as oil drilling declines in the short term and becomes less productive as the years go on.

Natural Gas Economically Recoverable Resources, 1990 and 2020

Gas production is limited by the available resources. In NEMS we model economically recoverable resources, which are a subset of the technically recoverable resources. Technically recoverable resources, in turn, are less than the in-place resources.

Resources include not only proved reserves, but inferred reserves from known fields, and undiscovered resources from new fields. Estimated resources come primarily from the assessments done by the U.S. Geological Survey for onshore regions and by the Minerals Management Service for the offshore.

As technology improves, economically recoverable resources grow. Economically recoverable resources are expected to grow from 1,040 tcf in 1990 to 1,355 tcf in 2020 in the reference case.

Technology has a significant influence on the development of resources. The rapid and slow technology cases assume resource variation of more than 100 tcf plus or minus the reference case values in 2020.

Resource levels have a significant influence on prices. The Gas Research Institute assumes significantly more resources, for example, and predicts significantly lower wellhead prices.

Natural Gas Net Imports, 1970-2020

Natural gas imports fill the gap between production and consumption. Over the forecast imports continue to be dominated by pipeline imports from Canada.

Gross imports from Canada are expected to reach 4.9 Tcf in 2020, compared with 2.9 tcf in 1996. Gross imports grow at the rate of 2.2 percent a year--faster than the growth in U.S. production.

Since 1984, U.S. natural gas trade with Mexico has consisted primarily of exports from the United States to that country. That trend is expected to continue through 2020, despite the North America Free Trade Agreement and the substantial size of Mexican natural gas resources.

Some of the unknown factors that affect the Mexican market are economic growth and development, environmental concerns, the role of Petroleos Mexicano as regulatory reform proceeds, political and policy risks, and the overabundance of heavy fuel oil in Mexico.

Another potential source of gas imports is liquefied natural gas (LNG). However, given the low natural gas prices in the lower 48 markets, LNG imports are not expected to become a significant source of U.S. supply. LNG imports are projected to increase from 40 billion cubic feet in 1996 to 360 billion cubic feet in 2020.

Natural Gas Consumption in North America

How much gas is available for export to the United States depends partly on local natural gas consumption.

The International Energy Outlook 1998 shows a 52 percent increase in North American gas consumption from 1996 to 2020, from 26 to 39.4 tcf. As in the United States, strong growth for electric power generation is expected in Canada and Mexico.

While U.S. gas consumption is expected to grow from 22 tcf in 1996 to 32 tcf in 2020, Mexican gas consumption is expected to grow even faster in percentage terms, because it starts from such a small base. Mexican gas consumption is expected to grow from 1 tcf in 1996 to 2.8 tcf in 2020, leaving none to spare for the United States.

Canadian gas consumption is expected to grow more slowly than U.S. and Mexican consumption, and more slowly than exports to the United States. Canadian consumption is expected to grow from 3.1 to 4.4 tcf from 1996 to 2020.

Lower 48 Natural Gas Wellhead Prices, 1970-2020

The average wellhead price in the lower-48 States is projected to reach $2.54 per thousand cubic feet in 2020. This is an average annual growth rate of of 1 percent from 1996 through 2020. This rise in prices reflects rising demand and increasing supply costs.

This forecast is in constant 1996 dollars. In nominal dollars, the price of gas rises from $2.24 to $5.21 per thousand cubic feet.

Natural gas prices are highly sensitive to changes in assumptions about technological progress. For the first decade of the forecast, both price and production levels are almost identical in the reference case and the two technological progress cases. By 2020, however, natural gas prices are 29 percent higher (at $3.28 per thousand cubic feet) in the slow technology case and 24 percent lower (at $1.92) in the rapid technology case.

Although we don't do a full Canadian market forecast, we do project Canadian border crossing prices. We expect those prices to rise from $1.96 per thousand cubic feet in 1996 to $2.55 in 2020. That rise is faster than the general increase in lower 48 wellhead prices, largely because of the expected increase in the marketability of Canadian gas with the planned increases in pipeline capacity into the United States.


In summary, over the next 20 years we expect the U.S. natural gas market to be largely driven by the demand for electricity. From now through 2020 we expect gas consumption by electricity generators to more than triple. Total gas consumption is expected to rise to 32 tcf. U.S. production is expected to increase to 27 tcf and net imports to 5 tcf. Prices are expected to rise very slowly to $2.54 per mcf.

Natural Gas Consumption by Carbon Reduction Case, 1996-2020

Ten days ago EIA released a major study on the impacts of the Kyoto Protocol on U.S. energy markets and economic activity. We found that efforts to reduce greenhouse gas emissions would have a significant impact on electricity generation, and thus, the natural gas market.

We examined six scenarios with different reduction targets for energy-related carbon emissions, although only three of them and the reference case are shown here for clarity. The reductions ranged from 24 percent above to 7 percent below 1990 carbon emissions levels.

Natural gas consumption is expected to rise more rapidly in all the carbon reduction cases than the reference case, because of increased consumption by electric generators. Although electricity generators would produce less electricity in the carbon reduction cases than in the reference case, they would consume more natural gas, because relatively high-carbon coal would be replaced with relatively low-carbon natural gas.

Natural gas consumption in the residential, commercial, industrial, and transportation sectors is expected to be lower in the carbon reduction cases than in the reference case, because gas prices are higher and fuel switching opportunities are limited in these sectors.

Natural gas prices in 2010 range from 5 cents to 70 cents per thousand cubic feet higher in the carbon reduction cases than the reference case. Increased consumption is responsible for higher prices at the wellhead.

Natural Gas Production by Carbon Reduction Case, 1996-2020

Natural gas production in the carbon reduction cases is higher than the reference case, because of higher consumption and prices. In 2010 gas production is 400 billion cubic feet to 2.1 trillion cubic feet higher than in the reference case.

The production patterns are not quite symmetrical across the cases, because of the penetration of conservation and renewables in the high priced cases. In the 3 percent below case production and prices rise sharply after 2005, when carbon reduction begins. The lower-priced 9 percent above case shows a steadier, but ultimately bigger increase in production.

Imposition of carbon reduction targets in 2005 cause a sharp increase in natural gas production. The projected increase between 2005 and 2006 ranges from 390 billion cubic feet in the reference case to 1.75 trillion cubic feet in the most extreme case. High levels of carbon reduction would pose a challenge to the oil and gas industry, and careful planning would be required.

EIA's Internet Site

Everything I've talked about--and more--is available to you electronically on the World Wide Web at www.eia.gov.

We'll continue to print the Annual Energy Outlook, but more is available electronically than in hard copy. In addition to the AEO, we have year-by-year forecasts, regional tables, and assumptions available only the Web. We even have an interactive query system available for regional forecasts.

Our customers are making increasing use of our web site. When we released the Kyoto report many of our customers found out about it through a "listserve," in addition to the more traditional print and broadcast media. Hundreds of customers read and downloaded the report within hours of its release on the Web.