U.S. Gas Production: Can We Trust the Projections?

Government (EIA) forecasts suffer in credibility when compared with geologic assessments.
Fortnightly Magazine - August 2001


Government (EIA) forecasts suffer in credibility when compared with geologic assessments.

In recent years, many have promoted natural gas as the "clean and safe" alternative to coal or nuclear energy in electrical power generation. This emphasis on natural gas should focus attention on the long-term forecasts of U.S. production and consumption of natural gas, as published by the Energy Information Administration (EIA) in the Annual Energy Outlook. In particular, what confidence should be placed in these projections?

First, consider gas consumption. The EIA predicts that U.S. natural gas consumption will increase by 62 percent from 1999 to 2020, while use of petroleum and coal will rise by 33 percent and 22 percent, respectively, over the same period. The predicted gas upsurge will occur primarily as a result of construction and operation of a large number of new base-load gas-fired electricity generation plants.

Second, consider gas production. The EIA predicts U.S. natural gas production will increase by 56 percent from 1999 to 2020, while coal production will increase by 17 percent over the same period, and oil will decline by 2 percent by 2020. () Now compare those predictions with past experience. As shown in Figure 1, U.S. marketed gas production reached a maximum of 22.65 trillion cubic feet (Tcf) per year in 1973 - a record that still stands after more than a quarter century of upheavals in gas markets.1 This fact should provide a context for the EIA forecast.

Falling Productivity: A Steady Trend

In justifying its forecast of natural gas production, EIA argued that sustainable increases in domestic production would occur because the steady decline in drilling costs would lead to increased well completions, and that new, but unspecified, technology would boost well productivity. EIA projected that annual gas well completions would increase from 10,300 in 1999 to 23,400 wells in 2020. EIA's predicted well-head price for 2020 is $3.13 per thousand cubic feet ($/Mcf) - below average well-head prices for calendar year 2000.

Why, then, has U.S. gas production dropped since that quarter-century peak recorded in 1973? The reason is a decline in proved reserves.

Simply put, additions to U.S. "proved" reserves from new discoveries and from more intensive development of identified fields (field growth) were insufficient to maintain reserves at levels that sustained production.2 Production from individual wells is physically limited by the amount of hydrocarbons that can be accessed at any given time. In many cases, no more than 10 to 15 percent of the proved reserves of individual fields can be extracted annually without risking reservoir damage and reducing ultimate field recovery. The "proved reserves" category limits annual production to an amount well below recoverable resource volumes. Reserve additions may fail to keep up with production because of a declining discovery rate (gas discovered per well), and because of low drilling activity. Low drilling activity is a signal that the industry is having difficulty finding and developing resources that can be produced commercially.

Data show that average U.S. gas well productivity has declined by at least two-thirds since 1973. Consider Figure 2. It shows the trends in average, annual gas well productivity for the entire United States, and for Texas and Oklahoma individually. Texas and Oklahoma, taken together, account for about one-half of the onshore conterminous U.S. (lower-48) gas production. During the last decade, with full deregulation, average U.S. gas well productivity has stabilized at just over 50 million cubic feet a year (MMcf/yr/well). The obvious long-term decline in annual productivity of operating wells imposes a severe constraint to large production increases in mature producing areas.

Proved Reserves: Where Credibility Suffers

EES North America

The credibility of EIA's production forecast suffers when compared to established geologic assessments of gas resources. Figure 3 shows components of EIA's U.S. gas production forecast by gas source. Specifically, it shows gas production from (1) onshore conventional3 lower-48 gas fields, (2) onshore unconventional lower-48 gas accumulations, (3) offshore gas fields, and (4) associated/dissolved gas produced in oil fields (onshore and offshore). Onshore conventional gas represents the largest projected increase of any source; it grows from 6.6 trillion cubic feet per year (Tcf/yr) in 1999 to 11.4 Tcf/yr in 2020. Gas production from unconventional gas and offshore gas fields is said to increase from 4.4 to 8.5 Tcf/yr and 4.5 to 6.2 Tcf/yr, respectively.

Consider the projection for onshore, lower-48 conventional gas production from gas fields (that is, conventional non-associated gas). For purposes of this analysis, the 1999 reserve-to-production ratio (10.6 to 1) for this gas source is assumed during the 2000-2020 time period. For this time period, total forecasted production (Figure 3) from lower-48 onshore conventional gas fields amounted to about 186 Tcf of gas. So, just to keep proved reserves from declining, at least 186 Tcf must be added to proved reserves during that period. However, to support the production increase from 6.6 Tcf/yr to 11.4 Tcf/yr, proved reserves in 2020 must also increase by about 50 Tcf from 1999 levels. Accordingly, the forecast implies that 236 Tcf must be added to proved reserves in conventional fields in the onshore lower-48 by the end of 2020.

The U.S. Geological Survey (USGS) assessment of undiscovered conventional technically recoverable gas is presented in terms of probability distributions (Gautier and others, 1996).4At the mean, the conterminous U.S. conventional undiscovered onshore non-associated gas is 142 Tcf. At the low probability end, with less than 1/20 chance of being exceeded, the estimate is 172 Tcf. Based on the 1995 USGS National Assessment (Root and others, 1996), reserve additions for gas in gas fields for the lower 48 from field growth for 2000 to 2020 are projected to be 85 Tcf. So, at the mean estimate of onshore, undiscovered gas in the lower 48, the EIA forecast exceeds available resources by 9 Tcf. Even at the high resource estimate, EIA's forecast implies almost 90 percent of the undiscovered gas in onshore conventional gas fields, regardless of cost, is found and added to reserves during the next 20 years.

Evaluating the Forecast: Implausible on Its Own Terms

The EIA production forecast falls in the middle of projections by commercial forecasters and trade group forecasts (EIA, 2000, p.107). However, other forecasters do not specify the gas sources, making it difficult to compare with the EIA data. Will other gas sources offset a possible shortfall in production of conventional onshore lower 48 gas?

According to EIA's documentation, the gas production forecast from unconventional sources - tight sandstone reservoirs, shales, and coal beds - depends on development of new technology that increases well productivity and reduces overall costs of unconventional gas development. However, if the new technologies are not ready or fail to materialize, production is expected to be 4.9 Tcf/yr in 2020, rather than the projected 8.4 Tcf/yr. (Krusskraa, 2001).

Gas from U.S. offshore fields is now produced almost exclusively in the Gulf of Mexico. New deepwater discoveries have prompted a reassessment of the undiscovered resource in the Gulf.5 However, major new gas fields must be discovered to meet the EIA6 offshore production forecast (see Figure 3). Otherwise, gas production from the Gulf, including deepwater areas is likely to begin its decline before 2010 (Nehring, 2001).

The Alaska North Slope is offered as an area containing significant new sources of gas, but potential deliverability is limited. Marketing the already identified 35 Tcf of Alaska North Slope gas requires construction of a new pipeline from the North Slope to connect into Canada's gas pipeline network. The largest pipeline under discussion, however, delivers gas at only 1.5 Tcf gas per year (4 billion cubic feet per day).

The EIA base case gas production forecast for 2020 is implausible. If forecast levels are not likely to be realized, decisions based on forecasts must be modified. Concern for adequate gas supplies should be broadened to include all of North America, as gas suppliers will initially try increasing imports to make up for shortfalls in U.S. production. Imports of liquefied gas (LNG) also have limited deliverability. If existing U.S. LNG receiving terminals are all re-commissioned and expanded to maximum capacity, they could accommodate only 1.3 Tcf gas per year.

Policy Implications: Rethinking Fuel Supplies

Policies encouraging gas use as a base load fuel for power generation should be rethought. In the past, gas powered generators were typically serviced on an interruptible basis so that residential heating customers received priority service. However, base load plants cannot be supplied with gas on an interruptible basis without leading to electricity blackouts or brownouts. Furthermore, deregulation allows power plants and other industrial customers to contract directly with gas suppliers, bypassing the local distribution companies (LDCs), so the LDCs have no authority to curtail gas deliveries to industrial users in favor of their residential customers. Gas-fired plants currently account for more than 90 percent of the capacity of new power plants scheduled for construction, so the consequences of a gas production shortfall could be costly.

Decision-makers might be better informed if EIA presented forecasts as scenarios, with full documentation provided for each of the scenarios. Less emphasis might be attached to a 'base case.' Scenario analysis should encompass a broad range of extreme but plausible boundary conditions, so that the decision-maker could develop contingency strategies if such conditions are realized. Over the course of history, market prices and supply of natural resources have shown instabilities - instabilities that are better captured through well-thought out scenarios than through the standard types of sensitivity analysis involving a base case.


American Petroleum Institute, 2001, Basic Petroleum Statistics, Washington, DC Attanasi, E. D., Gautier, D. L., and Root, D. H., 1996, Economics of undiscovered conventional oil and gas accumulations in the 1995 National Assessment of U. S. Oil and Gas Resources: Conterminous United States, USGS Open-file Report 95-75H 50p.

Energy Information Administration, 2000, Annual energy outlook 2020, 258p.


Gautier, D. L., Dolton, G. L., Takahashi, K. I., and Varnes, K. L., eds., 1996, 1995 National Assessment of United States Oil and Gas Resources - Results, Methodology, and Supporting Data: U.S. Geological Survey Digital Data Series 30, version 2 corrected.


Kusskraa, V.A. and Kuck, B. T., 2001, A long term view for unconventional gas. Annual Energy Outlook Conference, March 27, 2001, Washington DC. Minerals Management Service, 2001, Outer continental shelf petroleum assessment, 2000, Herndon, Va., 12p.

Note 4, p.16

National Petroleum Council, 1999, Natural gas, meeting the challenges of the Nation's growing natural gas demand, Washington, vol. 1 Summary Report. 96p.


Nehring, R., 2001, The Gulf of Mexico: Rising star or over the hill? Annual Energy Outlook 2001 Conference, March 27, 2001, Washington DC

P.16 Root, D. H., Attanasi, E. D., Mast, R. F, and Gautier, D. L., 1997, Estimates of inferred reserves for the 1995 USGS National Oil and Gas Resource Assessment, USGS Open-file Report 95-75L. 29p.


  1. Natural gas prices were gradually deregulated during the 1980s and early 1990s. Prior to this, Federal regulators had set well-head and pipeline prices and local regulators set retail prices. Although new natural gas electricity generating facilities were prohibited by the Natural Gas Policy Act of 1977, that provision of the Act was rescinded during the 1990s.
  2. Proved reserves are estimated quantities of oil and gas that geologic and engineering data demonstrate with reasonable certainty to be recoverable in future years under existing economic and operating conditions.
  3. Conventional fields consist primarily of discrete accumulations delineated by hydrocarbon/water contacts. Data for Figure 3 were provided by EIA and relate to Figure 94 in the Annual Energy Outlook 2001 (EIA, 2000).
  4. Both the EIA (2000) and the recent National Petroleum Council (NPC) study claim to use the 1995 USGS assessment as the basis for their respective forecasts of onshore production (NPC, 1999, p. 37). See Attanasi and others (1997) for the province breakdown.
  5. The Minerals Management Service recently re-assessed the undiscovered oil and gas in the Gulf of Mexico. The mean undiscovered oil was increased by 29 billion barrels of oil (Bbo) to 37 Bbo and undiscovered gas increased by 97 Tcf to 193. At well-head prices of $30 per barrel oil and $3.50 per mcf, gas about 75 percent of the gas is economic.
  6. The EIA forecast assumes no opening of any offshore area where exploration is currently prohibited.

Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.