by Vaclav Smil
The two key questions about shale gas development that have yet to be satisfactorily answered are the extent of ultimately recoverable resources and the productivity decline of wells. In trying to answer the first question, Berman (2010) tried to discern some longer-term patterns by turning to Barnett shale, the first major development of its kind with cumulative experience based on some 8,000 horizontal wells. His findings surprised him:
…most reserve predictions based on hyperbolic production decline methods were too optimistic when compared with production performance. There is little correlation between initial production rates and ultimately recoverable reserves. Average well life is much shorter than predicted, and the volume of the commercially recoverable resource has been greatly over-estimated. (Berman, 2010, 1)
Berman’s revised projection of ultimately recoverable reserves was 30% lower than his first estimate made just 2 years earlier. The main reason is that wells may not maintain the hyperbolic decline typical of the first months or years of their production: there may be sudden, catastrophic decreases, typically during the fourth of the fifth year of production but sometimes after just a year. Additional stimulation may temporarily boost the flow, but more often, it is followed by an even steeper output decline. As for an average well longevity, Berman argues that it does not make sense to expect at least 40 years of production instead of choosing an economic limit of about 57,000 m3/month (2 million cuft/month) as the threshold below which costs get higher than revenues based on the price of $3.5/1000 cuft, 25% royalty and average operating costs disclosed in the SEC filings.
These are critical differences: while the USGS put the ultimately recoverable Barnett shale gas at 736 Gm3, Berman has it at less than 255 Gm3 from nearly 12,000 wells, and he believes that additional 23,000 wells (with the land leased and wells drilled and completed at a cost of about $75 billion) would be needed to reach 736 Gm3. His two final surprising findings are the following: horizontal Barnett wells do not have significantly higher ultimate recovery than vertical wells—the difference is 31%, but the cost is 2.5 times higher—and performance of horizontal Barnett wells has not improved with time due to gains in experience and technical advances.
And a number of new resource appraisals—done with a much higher spatial resolution than used in the USEIA studies (blocks of a square mile that is 2.6 km2 rather than by county, with some counties larger than 1000 km2)—are coming up with much lower production forecasts (Patzek, Male, and Marder, 2013; Inman, 2014). While the USEIA sees natural gas production (largely driven by shale gas) growing until 2040, these new, much more conservative assessments see shale gas production from the four largest basins (Marcellus, Barnett, Fayetteville, and Haynesville) peaking already in 2020, with the 2030 output only about half of the USEIA’s expectations and a steeper decline afterward. If true, the US energy outlook would become much dimmer in a matter of years. At the same, these new conservative assessments do not take into account any innovative exploration and recovery techniques: some of them will certainly make a significant difference during the decades to come.
General trajectory of gas production (and, similarly, oil) from horizontal wells stimulated by hydraulic fracturing is easily stated. A typical course begins with a steep hyperbolic decline for the first 2–3 years after the initial production peak, followed by a long exponential (constant annual percentage) tail decline until the production stops: estimated ultimate recovery is based on a 30-year lifespan, but it would be surprising if commercial life of most shale gas wells would be longer than 20 years. Initial flow rates from fractured shales are between 60,000 and 120,000 m3/day, but the flows follow a pronounced hyperbolic decline whose magnitude differs among the major plays.
Sandrea’s (2012) preliminary evaluation of production potential of mature US gas shale plays, based largely on data from Barnett and Fayetteville, showed how different these resources are from conventional gas. Shale gas recovery is characterized by unusually high annual field decline rates (ranging from 63% for Fayetteville to 86% for Haynesville), low ultimate recovery rates, and hence low recovery efficiencies averaging just 6.5% and ranging from just 4.7% for Haynesville, 5.8% for Barnett, and 10% for Fayetteville. In Pennsylvania’s Marcellus shale, wells may produce more than 10 Mm3 during the first year (average is nearly 6 Mm3), less than 2 Mm3 in the second year, and less than 1 Mm3 in the third year (Harper and Kostelnik, 2009; King, 2014). In contrast, recovery rates range from 75 to 80% for conventional gas fields. Analysis of production potential showed that shale gas plays peaked at rates 2.6 times those of conventional deposits of the same size.
Perhaps the most representative findings of productivity decline in shale gas extraction came out of a comprehensive study of natural gas wells in five major shale gas basins (Barnett, Fayetteville, Woodford, Haynesville, and Eagle Ford) conducted by Schlumberger (Baihly et al., 2011). The study analyzed 1,885 representative wells (the largest number, 838, in Barnett shale) in a uniform manner that allowed for clear comparisons of performance starting on the date of first production and that produced the best possible estimates of ultimate recovery and hence allowed revealing assessments of long-term economic feasibility.
In Barnett shale, the maximum decline trend for wells drilled between 2003 and 2009 showed relatively small fluctuations around the following curve: 50% decline at the end of the first year, approximately 70% decline after 3 years, and 80% decline after 5 years. But the study concludes that Barnett shale is an exception as it had a flatter production decline trend, and hence, it would not serve as an analog for estimating productivity declines in other major shale plays. Fayetteville shale had a 1-year decline of about 60%, Woodford a bit over 60%, and Haynesville about 80% (Figure 8.3). Well costs ranged from $2.8 million in Fayetteville to $6.7 million in Woodford ($3 million in Barnett), operating costs (in $/Mcf) had a 3.5-fold span from 0.7 for Barnett to 2.5 in Haynesville, and ultimate recoveries per well were as low as 72 Mm3 for Woodford wells drilled in 2008 to 172 Mm3 for Haynesville wells drilled in 2009.
Figure 8.3 Decline of shale gas well output in the United States.
There has been nothing really surprising about the findings concerning decline rates (many less comprehensive earlier studies found similar rates), well costs, and ultimate recoveries, but the conclusions regarding the economic break-even price are surprising: for Barnett wells drilled in 2008 and 2009, they were only $3.70–3.74; for Fayetteville wells drilled in the same years, they were even lower at $3.20–3.65; but for Haynesville, they were $6.10–6.95, for Eagle Ford $6.10, and for Woodford $6.22–7.35. Barnett and Fayetteville could make money with gas at $4/Mcf at 10% discount rate and be profitable under low spot gas prices—but Haynesville and Eagle Ford and Woodford would need at least $6/Mcf.
But it is important to note that many operators hedge some or all of their production at higher than spot price values and that accumulating experience and technical innovation will tend to lower the future costs of extraction. At the same time, between 2010 and 2014, even as the extraction was rising and productivity per well was increasing in nearly all shale gas formations, debt burden of America’s shale oil and gas companies had nearly doubled, while revenues grew at just 5.6% during that period (Loder, 2014a). Sandrea (2014) also notes this reality—high capital expenditures nearly matching total revenues with net cash flow declining and debt rising—and also points out a relatively large number of substantial write-downs as many companies found the extraction much less attractive than they initially believed.
Is the US shale boom akin to a treadmill of capital spending, with dire consequences well known from other boom-and-bust cycles? In early 2014, 75 out of 97 oil and gas exploration and production companies rated by S&P were below investment grade (Loder, 2014a). Will bankruptcies of smaller and most overextended companies soon lead to more rational production or will be there a substantial investment decline? Because shale gas has become such an important part of the US energy supply so rapidl
y, we still should withhold any definite judgments regarding its eventual cumulative production and environmental impact.
For every superlative appraisal and endorsement cited earlier in this chapter, I could supply opinions ranging from much more subdued appraisals to calls for outright bans of shale gas production using hydraulic fracturing. In October 2013, Matthias Bichsel, projects and technology director at the Royal Dutch Shell, suggested that the US hydrocarbon industry has “overfracked and overdrilled” and reminded the enthusiasts that not all fields are created equal: Bakken, Eagle Ford, and Marcellus get all the attention, but the industry has not been able to achieve the same success in the Green River Basin in Wyoming and Utah (Hussain, 2013).
David Hughes, a Canadian geoscientist now at the Post Carbon Institute, concludes that “although the extraction of shale gas and tight oil will continue for a long time at some levels, production is likely to be below the exuberant forecasts from industry and government” (Hughes, 2013, 307). But some of these published expectations have not been excessive: the USEIA estimates that by 2030, shale gas will supply 7% of global natural gas production, hardly an excessive share. Similarly, the latest long-term price forecasts do not appear unrealistically low, assuming that by 2030 the gas for industrial consumer will be 85% more expensive than in 2012 and that residential consumers will pay about 30% more (USEIA, 2014i), an increase that might cover most, or all, higher costs of future shale gas extraction.
Environmentalists object to shale gas because its large-scale development and construction of pipeline and LNG infrastructure would be locking-in carbon emissions—some even claim at a level comparable to coal combustion (and, certainly, new gas-fired electricity-generating plants built recently will operate at least until 2040 to recoup the investment and make expected profit)—and making it more difficult to lower carbon emissions by adopting new reduction or capture techniques unless, of course, governments were to offer even greater incentives. As I explained in some detail (in Chapter 6), others object to any extensive development mainly because of the potential impact on freshwater resources.
On March 18, 2014, leaders of 16 environmental organizations sent a letter to President Obama asking him to stop any exports of LNG derived from shale gas because
we are disturbed by your administration’s support for hydraulic fracturing and, particularly, your plan to build liquefied natural gas export terminals along U.S. coastlines that would ship large amounts of fracked gas around the world. We call on you to reverse course on this plan and commit instead to keeping most of our nation’s fossil fuel reserves in the ground. (McKibben et al., 2014, 1)
Writers of the letter also make a direct link between LNG from fracked gas and catastrophic climate change:
Emerging and credible analyses now show that exported U.S. fracked gas is as harmful to the atmosphere as the combustion of coal overseas—if not worse. We believe that the implementation of a massive LNG export plan would lock in place infrastructure and economic dynamics that will make it almost impossible for the world to avoid catastrophic climate change… President Obama, exporting LNG is simply a bad idea in almost every way. We again implore you to shift course on this disastrous push to frack, liquefy, and export this climate-wrecking fossil fuel. (McKibben et al., 2014, 1–2)
Of course, such a stance would essentially eliminate any further combustion of all fossil fuels, an impossibility in a civilization where they now supply about 86% of all primary energy.
And factors other than productivity of shale gas wells and their environmental impacts will affect the extent of future recovery: these will include, above all, competition from other fossil fuels, renewable conversions, and nuclear electricity generation. Because of China’s and India’s demand surge, coal’s share in global primary energy use has been actually rising since the beginning of the twenty-first century, and despite the intended goals of reducing coal combustion and consuming more natural gas, cost advantages of coal may make this substitution slower than envisaged. Hydraulic fracturing is, of course, also a major new source of crude oil: as a result, the United States became again not only the world’s largest producer of natural gas but (when crude oil and natural gas liquids are combined) also the leading producer of liquid hydrocarbons in June 2014, the primacy it lost to the former USSR in 1975.
A new wave of nuclear power plant construction would reduce the demand for natural gas used for electricity generation, but given the parlous state of commercial fission in all Western countries and in Japan (stagnating, declining, on the way out, banned outright), this would matter only in Asia (China, South Korea, India), and even the greatest imaginable progress would be unlikely to displace large volumes of natural gas. Hence, the opposite consideration is much more pertinent: how much will cheap natural gas contribute to (once again) postponed renaissance of nuclear electricity generation, particularly in the United States, a phenomenon we have been promised since the mid-1980s?
Already, the effect has been undeniable as applications for construction of 12 new units have been either withdrawn or were suspended (USNRC [US Nuclear Regulatory Commission], 2014). Competition by wind and solar electricity generation could be much more important if judged simply by the total output of these renewable conversions—but, in reality, expansion of those intermittent renewable capacities would almost certainly increase the need for backup and peak demand gas-fueled generation during calm and cloudy days and (in the absence of mass-scale electricity storage) also for nighttime supply.
Global future of shale gas is even more uncertain. As already noted, even after nearly a decade after the US extraction began to take off, only Canada has become engaged in extensive shale gas production, while there is no sign of any similarly aggressive development of this new resource either in Europe or in Asia. In Poland, one of a few EU countries with interest in developing shale gas, detailed reevaluation of resources found them to be only a fraction of the estimates in EIA/ARI report (Kuuskraa, 2013). Looking ahead, it is not unreasonable to think that the combination of lower-than-expected endowment, environmentally based opposition, lack of readily available drilling, and well completion skills as well as the availability of requisite fracking capacities, regional shortages of water, and surprisingly high cost of extraction from deep, thin, and complex deposits will result in only slow rates of shale gas developments, not in replications of Barnett or Marcellus experience.
Yet another uncertainty comes with China’s determination to use its abundant coal in a cleaner manner. China’s first coal-to-gas project (Hexigen in Nei Monggol) began deliveries to Beijing at the end of 2013, starting with first 1.33 Gm3/year and eventually rising to 4 Gm3/year; in 2013, construction also began on three other projects, each with annual capacity of 4 Gm3; and preparations started for five additional plants: if all completed, they would deliver annually 36 Gm3 of gas (China News, 2013). If these projects were to prove successful and economically acceptable, the coal-rich country may emphasize coal-to-gas conversion rather than shale gas development, a combination of techniques where it has no deep technical expertise.
8.3 GLOBAL LNG
LNG has been the fastest-growing segment of global gas supply, but, once again, it would be unwise to overlook those factors that might moderate its further expansion. Perhaps the most important general consideration is the cost of new projects: there is no prospect for cheap LNG. Actually, the latest plants require considerably higher capital expenditures than the facilities built during the first years of the twenty-first century: upstream capital cost index of LNG developments has more than doubled between 2003 and 2013, and this rise has not been satisfactorily explained (IHS CERA, 2014; Songhurst, 2014). Typical rates during the 2000–2005 period were between $200 and 500/t of installed capacity. In contrast, even the latest low-cost projects (some involving conversion of regasification sites to liquefaction plants) range between $600 and 800/t of capacity; higher-cost projects (in Australia, Angola, and Papua New Guinea) range from $1,000 to
1,800; Norwegian Snøhvit (on a small Arctic island) cost $2,000/t (Figure 8.4); and among Australia’s dozen new projects, one is above $2,000/t and five are above $3,000/t.
Figure 8.4 Snøvhit LNG plant.
Reproduced with permission from statoil.com.
This means that a new high-cost LNG plant producing 5 Mt/year will cost at least $5 billion, development of the requisite gas supply will run (at $400/t of capacity) $2 billion, five LNG tankers ($220 million each) will add $1.1 billion, and regasification facility will require about $600 million, for the total cost of $8.7 billion. With construction delays, that can easily increase to $10 billion. That is why conversion of receiving terminals (so-called brownfield projects including Sabine Pass, Freeport, Lake Charles, Cove Point, and Cameron in the United States) is so appealing: converting the US regasification facilities to LNG plants would cut capital expenditures by at least 10%. Break-even price for most new LNG projects (be they in the United States, Africa, or Australia and based on conventional or nonconventional gas) is $10–11/million Btu (Fesharaki, 2013), and it might be possible to deliver the US shale gas (priced at $4/MMBtu) at less than $8/million Btu to Europe and at less than $10/million Btu to Asia (CB&I (Chicago Bridge and Iron Company), 2011).
