How resilient has US shale become?
Genuine gains have been made, but the industry will need further technological breakthroughs to overcome the geology
Discussions about the US shale industry's resilience through the oil price downturn and swift recovery inevitably turn to drillers' ability to innovate their way to lower breakeven prices—a line of argument often pushed by producers themselves.
But does this hold up to scrutiny and are the gains made through the downturn sustainable? The industry has taken a number of measures in the face of lower oil prices, some of which are genuine breakthroughs, but many will see their effects fade or dissipate altogether with higher oil prices.
For instance, improved geosteering, which allows operators to drill further in less time, will bring lasting improvement to the shale patch. Drilling longer laterals, another key insight gained through the downturn, will likely lead to long-term improvements, but its effect will fade as service costs rise along with higher prices and increased drilling activity. Many of the changes to completion strategies—particularly the increases in volumes of proppant—are simply cases of "going big" rather than developing something new, and are thus more at the whim of market forces like proppant prices.
In addition to these changes in practices, there is some "survivorship bias". In order to survive the downturn, producers have sought to maximise short-term revenue by tapping into only the best of the best of their acreage, while other areas have been rendered uneconomic by low oil prices.
Calculating how these factors balance out is the challenge. Tudor Pickering Holt, an energy investment bank, has shown that gains in the Permian are far more limited when controlling for lateral length, indicating lateral length as a dominating factor. This in itself points to the potential for diminishing returns—once laterals are as long as they can be within a lease, there is no further room for improvement on that front. This is not to say there has been no innovation. Research from PetroNerds, a consultancy, shows that, after controlling for survivor bias by comparing wells drilled on the same lease years apart, some substantial well productivity improvements have been made in the tight oil patch.
To examine gains over time, Petrologica, an energy market analysis consultancy, created a well comparison index. We indexed average cumulative production of wells by completion year to the average well profile from 2010. This allows a month by month comparison of ultimate recovery, rather than only focusing on gains in initial production rates-an often-used but potentially misleading metric.
This technique shows that improvements in well productivity have been much greater in the Texas plays than in North Dakota, albeit they have come from a lower baseline.
New drilling techniques, employed at Bakken wells through to 2014, appear to mostly have accelerated initial production, with only a modest impact on ultimate recovery, if any. It is too soon to draw conclusions regarding wells drilled from 2015 onwards. Gains in initial production rates are retained in the ultimate recovery rate to a much greater extent in the Permian and, to a lesser extent, in the Eagle Ford.
For example, in the Bakken, 2013's wells have now produced only 2.1% more oil than 2010's wells by the same month of production, despite having produced 10% more by the end of their first year. This contrasts with the Permian's 2014 average well, which had produced 26% more oil than 2010's average well at the end of the first year, and are now still 22% ahead.
A possible reason for this discrepancy is the difference in overburden pressure on the respective basins. The Bakken's tight oil formation is located some 6,000 feet deeper than the Permian's on average. While this means oil is under more pressure at the beginning of a Bakken well's life, it also means that proppant—the grains of sand that keep the rock fractures open and the oil flowing—is more easily crushed, so rock fractures are more likely to close over time and prevent oil from flowing.
In both the Permian and the Bakken, a relatively small number of (on average) high quality wells were completed in 2010, whereas the next few years saw much larger numbers of (on average) worse performing wells completed. In both plays it took until at least 2013 for the average well to exceed their 2010 benchmark. This in itself points to productivity improvements: by 2014, though ten times as many wells were completed in both plays as in 2010 as companies ventured outside their best acreage, average production rates were similar to when the sweet spots were defined in 2010.
Nevertheless, the Bakken started from a much higher baseline than the other plays, and continues to have a somewhat higher estimated ultimate oil recovery for its average well due to a slower decline rate. This is despite the gains in initial production rates in the Permian and Eagle Ford outpacing those in the Bakken. Still, factors like transport and infrastructure cost, as well as the promise of 'stacked pay'—further exploitable oil layers on the same acreage—continue to make the Permian more attractive to producers for the time being.
Longer-term, companies will have to work both harder and smarter to grow production as they are forced to move further out of the sweet spots. A recent study by Wood Mackenzie, a consultancy, suggests that the Permian Basin could reach peak production as early as 2021 if technological gains cannot continue to outstrip geological factors. If shale innovation is subject to diminishing returns, this will prove increasingly more difficult as time passes.
Graham Walker is a Research Analyst at London-based consultancy Petrologica