There's little disputing that the U.S. is currently in an unprecedented oil-production boom, driven by tight-oil resource plays. But how big will the boom be, and how long will it last?

U.S. tight-oil production could reach 7 million barrels per day (MMbbl/d) beyond the 2011 level, based on drilling and production numbers provided by the oil and gas industry. Even conservative estimates still result in total U.S. tight-oil production reaching 4 MMbbl/d (all production numbers in this article are beyond the 2011 production level). For context, current U.S. oil production, including conventional and tight oil, is already about 7 MMbbl/d, the highest level in decades.

The industry expects cumulative tight-oil production of 72 billion to 93 billion barrels of oil, more than twice as much as current estimates of total proved reserves in the U.S. And, output from currently known tight-oil reserves is projected to remain high for about 20 years before going into decline.

These estimates are based on combining oil and natural gas industry drilling and production numbers for the main U.S. tight-oil resource plays, into the total for U.S. tight-oil reserves. The plays include the Eagle Ford oil- and wet-gas windows; Spraberry-Permian verticals; Permian horizontals; other horizontals; the Utica shale; the Granite Wash; the Niobrara; the Mississippi Lime; and the Bakken and Three Forks. The long-term forecast examines three scenarios to arrive at the projection of 7 MMbbl/d: two industry cases, and a conservative case. Results can change with each well drilled.

The industry case assumptions are summarized in Table 1, which is based on the U.S. Energy Information Administration's 2012 Annual Energy Outlook (AEO 2012). For key inputs for plays, see sidebar.

Forecast results and comparisons

The results for the Industry Case With Rig Reallocation are graphed in Figure 1 (space prevents showing the graph for all three scenarios). Table 2 summarizes production for each resource play in the three scenarios. Total U.S. tight-oil production reaches a maximum of 7.02 MMbbl/d in the Industry Case With Rig Reallocation, 5.17 MMbbl/d in the Industry Case Without Rig Reallocation, and 4.07 MMbbl/d in the Conservative Case.

Table 1 shows that that the industry expects to eventually produce a total of about 93 billion barrels of oil from tight-oil resource plays in the industry cases. Total eventual production in the Conservative Case is about 72 billion barrels of oil. This compares with about 22 billion barrels of proved reserves at the end of 2012, reported in the AEO 2013 Early Release, and about 30 billion of proved reserves in the International Energy Agency's World Energy Outlook 2013.

Translating cumulative production in the Industry Case into reserves moves the U.S. ahead of Russia for sixth-largest reserves in the world.

A comparison with other forecasts is shown in Figure 2. It is especially useful because a single forecast can be difficult to evaluate, and comparisons with other forecasts provide context. These comparative forecasts had different levels in the early years, which is why all forecasts are shown here as changes from 2011.

All the forecasts except AEO 2012 show rapid growth in the early years. Clearly, the U.S. is in an oil boom; the question is, how high does production go long term?

The three forecast scenarios developed here from industry data are among the highest forecasts long term. The CIBC Wood Gundy fore-

cast reaches the same level as the Conservative Case. The Industry Case forecasts are significantly higher than the others shown.

Some of the forecasts have a flat spot in the middle years. This is merely a function of how the numbers worked out. A problem might occur if the time horizon is stopped during this flat period, because it might be interpreted as meaning that production could go on forever (or a long time) at that rate. But the long-term outlook shown here shows that the currently known tight-oil plays will eventually be depleted.

It is particularly interesting to compare the IHS CERA forecast with these forecasts. IHS CERA emphasizes that production would decline as drilling moved from the current sweet spots to the more general parts of the resource plays. With IHS CERA forecasting about 3 MMbbl/d and these forecast cases at roughly 4, 5 and 7 MMbbl/d, the former indicates that estimated ultimate recoveries (EURs) will be about three-quarters of those in the Conservative Case, three-fifths of those in the Industry Case Without Rig Reallocation, and three-sevenths of the Industry Case With Rig Reallocation.

In the Bakken, for example, the Conservative Case EUR of 500,000 barrels of oil equivalent (BOE) would be expected to average more like 375,000 BOE over the entire resource play, and the Industry Case EUR would go from 603,000 BOE to about 362,000 BOE at three-fifths, or to about 260,000 BOE at three-sevenths. These lower EURs apply to individual oil and gas companies as well as resource plays.

Another interpretation is that there are about 30 E&Ps that are saying the IHS CERA forecast is too low. Still another interpretation is that the IHS CERA forecast and some of the others are based on historical well performance. As is shown here, well performance has been increasing over time. Forecasts based on historical well performance can be expected to be lower than forecasts based on current or expected performance. There is probably some truth to all these interpretations. Well performance can be expected to increase over time with drilling and fracing improvements, but it also can be offset by an expected decrease as drilling moves beyond the sweet spots.

Key tight-oil plays

Eagle Ford oil window. The Eagle Ford shale is divided into the oil window, the wet-gas window, and the dry-gas window (which is not included in this oil forecast discussion). The size of the Eagle Ford was estimated from the map in EOG Resources' September 2012 investor presentation to be about 2,100 square miles for the oil window and 1,200 square miles for both the wet-gas window and dry-gas window, for a total of about 4,500 square miles.

Baker Hughes reports that there were approximately 169 horizontal, oil-directed (HOD) rigs and another 61 horizontal rigs drilling for natural gas in the Eagle Ford on February 1, 2013. The 169 oil-directed rigs have been split about two-thirds (113 rigs) in the oil window and one-third (56) in the wet-gas window based on

reported land holdings by the E&Ps.

Eighty-acre well spacing is used in this analysis (eight wells per 640 acres in a square mile) based on multiple sources. EOG reports moving from 130-acre well spacing to 65 to 90 acres. Marathon Oil uses 148 acres to 74 acres per well and is testing 40-, 60- and 80-acre pilots. Swift Energy reports 80-acre spacing.

EOG's drilling plans work out to a little over 15 wells per rig. EOG reports that its current production is 78% oil, 10% NGLs and 12% gas. Anadarko Petroleum Corp. reports an EUR of 450,000 BOE, which is used in the Conservative Case. Marathon Petroleum Corp. reports an EUR range of 694,000 BOE in some wells and 463,000 BOE in other parts of the play. The simple average, 536,000 BOE, is used for the Industry Case.

The resulting Eagle Ford oil window production forecast is shown in Figure 3 and increases rapidly to a peak in 2019 of 1.07 MMbbl/d in the Industry Case and 0.89 MMbbl/d in the Conservative Case above the 2011 level. Production declines rapidly after 2019.

Other oil production forecasts do not show this steep decline, and most company presentations also do not show it or only hint at it. Producers will likely try to mitigate this steep decline by moving to tighter well spacing. When the industry reports EUR results for tighter well spacing, they can be included.

Another concern is that drilling has been concentrated in the "fairway" and has not been extended to the edges of the oil window. If drilling cannot continue over the entire oil window or is significantly less productive than in the fairway, then production will peak sooner.

Eagle Ford wet-gas window. Some 56 rigs are estimated to be drilling in the Eagle Ford wet-gas window. Many of the inputs are for the oil window. Anadarko indicates, however, that the mix of output in the Eagle Ford wet-gas window changes to 40% oil, 25% NGLs, and 35% natural gas. Anadarko's 450,000 BOE is used for the Conservative Case. Marathon reports about 1.126 million BOE in one presentation and 965,000 BOE in its December 5, 2012, presentation. Again, the simple average, 847,000 BOE, is used for the Industry Case.

The gas-rich plays often have higher EURs than the oil-rich plays. The resulting oil production peaks in 2019. Production is shown in Table 2 for each of the cases and will not be described for the remaining resource plays.

Spraberry–Permian vertical. Spraberry oil production in the Permian Basin began zooming up in 2008. The application of shale gas-type procedures to vertical drilling and going into deeper formations boosted production. This is the only resource play in the forecast that considers vertical drilling.

Pioneer Natural Resources reports its Spraberry vertical-well EURs have recently been 110,000 BOE (Conservative Case) and could increase to about 140,000 BOE by drilling deeper. Apache reports 127,000 BOE for Spraberry wells, with the average of the two higher numbers used in the Industry Case.

Pioneer also notes that drilling to other zones could add up to 100,000 BOE to the EUR, which is another source of upside not included in the forecast. Forty-acre spacing is used. Pioneer is looking to move to 20-acre spacing and indicates, but does not yet appear to be able to confirm, that its EUR will still be about 140,000 BOE with 20-acre spacing. Pioneer reports 70% oil, 20% NGLs and 10% natural gas.

Pioneer says that the Spraberry covers 1.7 million acres (2,656 square miles) but it could cover 5,000 square miles. The company also indicates that there are about 220 vertical rigs drilling in the Spraberry out of about 500 total rigs in the Permian Basin. It is drilling a little over 15 wells per year per rig.

Permian horizontals. Key resource plays for horizontal drilling in the Permian Basin are the Wolfcamp, Avalon, Bone Spring and Cline. These four plays identified are included separately because they have different proportions of oil, NGLs and natural gas; areal extent; and well spacing. Anadarko reports that its Bone Spring production is about 70% oil while its Avalon production is only about 15% oil. Anadarko says that both of these formations have EURs of about 400,000 BOE per well (used for both cases).

EOG reports that its Wolfcamp production has a slightly higher EUR at 430,000 BOE and

is about 42% oil, 30% NGLs and 28% natural gas (Conservative Case). Pioneer reports EURs of 575,000 BOE for its Wolfcamp production with 7,000-foot laterals with 90% liquids. Apache reports 598,000 BOE and 93% liquids. The average Wolfcamp EUR, 534,000 BOE, is used for the Industry Case. Pioneer also reports the potential for six horizontal intervals.

Apache is reporting EURs of 423,000 BOE (Conservative Case) in the Cline, while Devon reports 570,000 BOE. The average, 496,500 BOE, is used for the Industry Case. Apache has identified 2,321 well locations on its 520,000 acres, or about 224 acres per well. That works out to 28,000 wells in the 9,800 square miles of the Cline (22,400 wells after applying an 80% uncertainty factor).

Baker Hughes reports drilling for only the Permian Basin as a whole, however, rather than breaking out the individual plays. The newer Cline play is allocated 10 rigs. The horizontal rig count for the Permian Basin is allocated 10% (15) to Bone Spring, and the Avalon and Wolfcamp split the remainder in an attempt to roughly balance how fast the plays are drilled.

Other horizontal. The Other Horizontal category is included in order to match the total number of HOD rigs in the Baker Hughes February 1, 2013, report. Its total of 154 rigs includes rigs in basins with few HOD rigs: Cana-Woodford (13), Barnett (six rigs), Ardmore-Woodford (six) and Haynesville (two).

The 127 HOD rigs not assigned to a basin are in Texas (48), Oklahoma (28), Wyoming (25), Kansas (six), Louisiana (four), Utah (three), Mississippi (three), New Mexico (two), and Colorado and Florida (one each). The EIA shows large tight-oil potential in California, but only six HOD rigs are drilling there. The four HOD rigs drilling in Alaska are excluded.

An average of other plays is used for the Other Horizontal wells: 50% oil, 15% NGLs, and 35% natural gas. The EUR from horizontal wells generally ranges from 350,000 to more than 500,000 BOE. Assuming that the best production is in the named resource plays, the Conservative Case uses an EUR of 250,000 BOE while the Industry Case uses a rough average EUR of the Conservative Case for several other plays at 414,500 BOE. The number of Other Horizontal rigs is decreased by five rigs per year after 2020 in the Conservative Case and after 2030 in the Industry Case, because it is not clear that there will be resources to drill or move to in the long term.

Utica shale. Horizontal drilling for oil in the Utica is new, and there isn't much data yet. The first vertical, oil-directed drilling rig in the Utica was reported to Baker Hughes on July 8, 2011, and the first HOD rig was reported on October 14, 2011.

Chesapeake Energy Corp. reported that its Utica production was 20% oil, 15% NGLs and 65% natural gas. Gulfport Energy reported EURs of 455,000 BOE to 910,000 BOE. The Conservative Case uses the former number and the Industry Case uses the average of the two, 682,500 BOE. There were 21 HOD rigs in the play, and a generic 10 wells per rig was used for this stage of development. Because of the preliminary nature of these data, a 60% uncertainty factor was applied to the Utica.

Granite Wash. Only the Granite Wash is included here, but there could be more potential in the region. Linn Energy says there could be eight horizontal zones, Apache identifies 12 and Forest Oil shows 16 possible horizontal zones. These zones could also cover different areas than the 4,800 square miles estimated for the Granite Wash alone. Baker Hughes reported 41 HOD rigs on Feb 1, 2013, a significant drop from 65 at the end of November. Apache and Unit Corp. reported drilling about 10 wells per rig, and Unit Corp indicates 120-acre spacing.

Apache reported EURs of 606,000 BOE for its 2010 wells, 895,000 BOE for its 2011 wells, and 1,146,000 BOE for first-half 2012. These increases over time indicate why forecasts using historical production are lower than forecasts using recent or future production levels.

The Conservative Case rounds this to an EUR of 600,000 BOE while the Industry Case uses 800,000 BOE based on a later Apache presentation. The production mix is estimated to be 25% oil, 25% NGLs and 50% natural gas, based on Apache and Unit Corp. information.

Niobrara. Anadarko owns a great deal of potential Niobrara property in Colorado as a result of its railroad grant lands. Anadarko estimates the range on Niobrara EURs to be 300,000 to 600,000 BOE. Anadarko and Noble Energy Corp. say 335,000 BOE to 435,000 BOE. An EUR of 350,000 BOE is used in the Conservative Case. The Industry Case uses an EUR of 400,000 BOE, which is still towards the low end of the range. Anadarko estimates 60% oil, 10% NGLs, and 30% natural gas.

The size of the Niobrara play is from AEO 2012. Because much of this land has not been drilled, a 40% uncertainty factor is applied. To be conservative, 160-acre spacing is used, although Noble reports its best results with 80-acre spacing. It also sees potential for 30 wells per section in three Niobrara horizontal zones and another in the Codell. Baker Hughes reports 17 HOD rigs, which results in modest oil production reported in Table 2.

Mississippi Lime. The Mississippi Lime is one of the hottest plays in the U.S. The number of horizontal rigs drilling for oil in the Mississippi Lime grew from only three rigs in the February 4, 2011, Baker Hughes data to 84 rigs in the February 1, 2013, report. The rapid increase in the drilling rig count in the Mississippi Lime and its areal extent make it one of the five most important resource plays in the continental U.S., along with the Bakken, Permian, Eagle Ford and Marcellus.

The data shown here indicate that the Mississippi Lime covers a larger area than the Bakken (12.5 million acres compared to 9.6 million) and has greater total recoverable reserves (26.5 billion BOE compared to 24 billion). Apache Corp. is testing whether it extends into Nebraska in addition to Oklahoma and Kansas. This might extend its aeral extent to 17 million acres.

SandRidge Energy and Encana reported well spacing of four wells per square mile (used in this forecast). Apache and Range Resources reported eight. SandRidge reported 46% oil and 54% natural gas (used here). Chesapeake is similar at 39% oil, 12% NGLs and 49% gas while Range Resources reports 67% liquids.

SandRidge reports EURs of 456,000 BOE while Chesapeake and Encana reported 425,000 BOE (Conservative Case). Range Resources noted 485,000 BOE for its 2009-2011 wells with 2,197-foot laterals on average, but says that its 2012 wells with laterals averaging 3,648 feet are producing at rates that imply EURs of 600,000 BOE. The Industry Case uses 475,000 BOE, which is the rounded average of the other five EURs.

Bakken, Three Forks. The Bakken and Three Forks in North Dakota have 188 HOD drilling rigs working, making it the most active play in the U.S. Total size is a question. Continental Resources estimated the Bakken at 14,300 square miles with 82% derisked in 2011, at 13,000 square miles with derisk not mentioned in 2012, and 15,000 square miles with 87% derisked in 2013 (15,000 and 80% used here).

Continental's 2011 estimate of EUR was 500,000 BOE (Conservative Case); recent presentations show 603,000 BOE (Industry Case). It assumes four wells per square mile with wells in both the Bakken and underlying Three Forks. Continental is testing 160-acre spacing and two more horizontal zones in the Three Forks. Drilling has recently been at an annual rate of nine wells per rig, according to data from the North Dakota Department of Mineral Resources.