Sceptics see the output boom in the US about to bust
Chris Nelder reviews new analysis of US tight oil and shale gas production which suggests the output boost from unconventionals will wane before the end of this decade
According to the Energy Information Administration
(EIA), the data-gathering arm of the US Department of Energy, fracking has a bright future. Production of tight oil and gas from US shale formations is expected to remain robust in the EIA's Annual Energy Outlook 2014
reference case, with total crude oil production rising from 7.4 million barrels a day (b/d) in 2013 to 9.6m b/d in 2019, then slowly declining to 7.5m b/d by 2040. Natural gas production is projected to grow continuously from 66.5 billion cubic feet a day (cf/d) to 102.7bn cf/d over the same period.
Drilling Deeper, a comprehensive, extraordinarily detailed, and highly transparent new report prepared by veteran Canadian geoscientist J David Hughes for the Post Carbon Institute, comes to radically different conclusions, with huge implications for the US and the world. The report analysed the seven largest tight oil plays, representing 89% of present and 82% of projected US tight oil production through 2040, and the seven largest shale gas plays, representing 88% of present and projected shale gas production in the EIA's reference case.
The most likely models in the Hughes report suggest that the major tight oil and shale gas plays will peak around 2016-17. This forecast deflates the EIA model, since nearly all growth in US production is expected to come from tight oil and shale gas. According to Hughes's report, total US tight oil and shale gas production will peak before 2020 and decline much more sharply than the EIA projects.
Hughes's methodology analysed historical production, and recognised both wells in production and those drilled but not yet producing. The decline rates of horizontal and vertical wells were analysed separately. Drilling and production data were taken from the Drillinginfo well-production database, and was current through early- to mid-2014.
The average estimated ultimate recovery (EUR) of each play was determined by developing future production profiles based on assumed rates of drilling, average well quality by area, well- and field-decline rates, and the estimated number of available drilling locations remaining. Field decline rates and the number of available drilling locations were determined down to the county level for the Bakken and Eagle Ford tight oil plays, and for five other shale gas plays.
Future production of each existing well was modelled using average well-decline profiles, starting with the year each well was drilled. For future wells, first-year production was modelled using observed average 2013 first-year production rates. Average first-year production was also used to determine the number of wells needed to offset field decline each year, and to model future production at various drilling rates.
Existing trends in well productivity were also taken into account to model future production rates for newer plays: If productivity has been increasing recently, then the model assumed that it would continue to increase for a time, before declining as drilling moves away from the most-productive sweet spots and into lower-quality portions of the plays.
Various drilling rate scenarios were used to estimate future production for the Bakken and Eagle Ford, and for five shale gas plays. It was assumed that well productivity would decrease over time from that observed in the most-productive counties of those plays to that observed in the outlying counties, as drilling moves away from the sweet spots.
The productivities of all plays were also compared to each other to arrive at an overall assessment of the likely long-term sustainability of tight oil and shale gas production. Production histories and projections for all plays were then compared to the EIA forecasts.
The two largest plays, the Bakken and Eagle Ford, contribute 62% of US tight oil production. Their output will largely dictate the future of US oil production. Hughes's expectations for the peak rate of production from those two plays, and when that peak occurs, is similar to the EIA's expectations. However, lifetime production from those two plays will be 28% less in Hughes's model than the EIA estimates. Production from those two plays is likely to peak by 2017 at around 2.8m b/d in Hughes' model, then decline sharply to 80,000 b/d by 2040 - less than a tenth of the EIA's 1m b/d estimate.
Total US tight oil production by 2040 was not modelled, but the report implies that it will be a fraction of the EIA's forecast. The EIA's projection that US tight oil production in 2040 will be comparable to current production at 3.2 b/d is 'highly unlikely given a thorough analysis of the data', Hughes says.
To meet the EIA's forecast, the five lesser tight oil plays, along with other emerging plays, would have to produce twice the amount of oil that the Bakken and Eagle Ford will produce in Hughes's model by 2040. The EIA expects three Permian basin plays - Spraberry, Wolfcamp, and Avalon/Bone Spring -- along with the Austin Chalk and the Niobrara, to produce four to five times their historical production over the next 26 years. The report calls this forecast 'extremely optimistic' since those plays are already 40-60 years old, with tens of thousands of wells drilled, and half the well quality of the Bakken and Eagle Ford wells on average. Hughes's best-case estimate is that these lesser plays will produce about 10bn barrels between 2014 and 2040, just over half the EIA's estimate.
The report also casts doubt on the notion that improving technology will result in far higher future production. Tight oil producers have aggressively promoted the idea that by continuously improving fracturing technology - using longer horizontal laterals, more fracturing stages per well, more sophisticated mixtures of proppants and other additives in the fracturing fluid, and larger injections - well productivity will be significantly increased, reducing costs.
But the increase from each new well has been declining since early 2012. In 2012, a new well in the Bakken added about 170 b/d to the field's production for that year. By mid-2014, a new well only added about 100 b/d. The average first-year production of Bakken wells in 2013 was only 7% higher than it was in 2011, and probably owed more to concentrating drilling in the sweet spots than improved technology. In the Eagle Ford, the average production of oil and gas over the first six months of a well has been essentially flat since 2011, averaging 432 barrels of oil equivalent a day.
The operators' assumption that wells will be in production for more than 30 years also remains to be proved, and seems dubious considering that production from Bakken wells falls more than 90% in five years. As of March 2014, only 1% of Bakken wells were over 10 years old, 82% were younger than five years, and the mean age was about three years. In the Eagle Ford, 89% of the wells are younger than three years, after which their production will have fallen by nearly 80%. Fewer than 1% of Eagle Ford wells are more than five years old, and the mean age is just 20 months.
The EIA is also overly optimistic about total recovery, says Hughes, given that it assumes very high (70% to over 100%) recovery rates for unproved resources, which have no price constraints and only loose geological constraints, as well as total recovery of proved reserves, which are defined as being recoverable with existing technology and economic conditions.
Shale gas production from the top seven plays will peak at nearly 34bn cf/d in 2016 in Hughes's 'most likely' model, then decline to less than 16bn cf/d by 2040. Total production through 2040 will be 39% less in Hughes's model than the EIA estimates, and fall to about one-third the rate in the EIA forecast. The EIA projects domestic gas production to reach nearly 38 trillion cubic feet per year by 2040 -- 55% above 2013 levels -with shale gas providing most of the production growth, but the Hughes report asserts that merely maintaining current shale gas production will be 'problematic'.
Four of the top seven shale-gas plays are in decline. Of these major plays, only the Marcellus and associated gas from the Eagle Ford and Bakken tight oil plays are still growing. Average well productivity is flat in all major shale gas plays except the Marcellus. Yet the EIA forecast calls for plays currently in decline to grow to new production highs at moderate future prices. Average first-year field decline rates range from 23% to 49%, so one-quarter to one-half of all production in each play must be replaced by new wells each year simply to maintain flat production.
Under Hughes's 'most likely' scenario, the top seven shale gas plays will produce 230 trillion cf of gas from 2014 to 2040 (about five times the amount recovered to date). Production from them in 2040 will be 14.8bn cf/d, assuming no capital constraints and no restrictions on access to drilling locations. In contrast, the EIA forecasts that 377 trillion cf of gas will be recovered over the same period from these plays, and that production in 2040 will be nearly three times higher, at 41.8bn cf/d.
The implications of Hughes' analysis are stark. While tight oil and shale gas production will continue to increase for a few more years as drillers continue to focus on sweet spots, it could decline much more rapidly than the EIA projects as drillers exhaust those areas and are forced into the periphery of shale plays. The recent decline in oil prices, which are now below the $85-90 a barrel threshold that some analysts believe is the marginal cost of tight-oil production, also makes it increasingly uncertain that capital markets will be willing to fund the enormously expensive future production the EIA projects.
Without considerably higher future prices for oil and gas that might justify drilling in progressively poorer rock, Hughes asserts, production will fall. He flatly dismisses as false industry assertions that shale plays are broadly similar in quality; that they are 'manufacturing operations' where tens of thousands of wells can be drilled with the same productivity; that technological advances can overcome steep decline rates and declining well quality; and that large remaining resources 'imply high and durable rates of extraction over decades'. "Rather than viewing tight oil and shale gas as an unlimited bounty," Hughes says, "it should be viewed for what it is -- a short-term reprieve from the inexorable decline in US oil and gas production."
Chris Nelder is an energy analyst and consultant who has written about energy and investing for more than a decade. He is the author of Profit from the Peak and the co-author of Investing in Renewable Energy. He consults with business and government on the future of energy and blogs at GetREALList.com