Two months ago Rystad Energy argued there were no signs that increasing well performance in the Permian had reached an inflection point. This assertion was primarily formulated by analyzing the average production per new horizontal well, which is the average daily output on the second production month, typically the first full production month. With two additional months of production data now available, we confirm that our previous conclusion holds true, as seen in Figure 1. In fact, the average well each month is slightly more productive than average well drilled three to six months ago. It should be noted that the basin-wide average productivity of new wells has been observed above 1,000 barrels of oil equivalent per day (boepd) for five months in a row on a two-stream production basis.
Some market participants argue that although new well productivity on the second month (which we consider a proxy for peak month well productivity) continues to trend upwards, decline rates are also getting steeper. We have certainly documented a significant number of empiric examples where this phenomenon can be observed, however we see that the impact of steeper declines on longer term recovery rates at basin level is limited. Figure 2a shows the evolution of cumulative two-stream recovery per horizontal well over an initial three, six and twelve month period. The month on the X-axis corresponds to the month a well is turned-in-line, while production is again measured on a two-stream basis. The Figure clearly illustrates that the average recovery per well continues to trend upwards regardless of the length of the aggregation period. Even the cumulative first-year recovery trend is positive with a frequency of 200,000+ barrels of oil equivalent (boe) average recovery, a frequency which increases as we move from 2017 to 2018.
It could be argued that the productivity increase is no longer visible after accounting for changes in perforated lateral length. This criticism is valid; if the cumulative recovery per well is linearly normalized to a 9,000 feet lateral design, as in Figure 2b, a flat or even marginally negative trend indeed emerges, for all recovery metrics from late 2016.
But what does flat normalized productivity actually mean? It is important to remember that since 2016, flat development in normalized productivity has occurred parallel to the simultaneous increase in the frequency of long (2-mile) laterals. The average lateral length in some months of 2016 barely exceeded 6,000 feet, moving systematically above 7,000 and 7,500 feet in 2017 and 2018 respectively. As of 2Q19, the average lateral length in the Permian is around 8,500 feet. As we have previous argued, it is difficult to maintain flat normalized productivity when an operator switches from short to long laterals. The benefits of making the switch are realized on the cost side of the equation. Certain well cost components are not related to the length of horizontal sections (e.g. the most tangible costs exhibit some correlation with lateral length, but they might be better correlated with the total measured depth or pad design). As a general rule, this means that long laterals exhibit lower drilling and completion cost per foot. Can we see this long-term trend if we look at the evolution of normalized well cost in the Permian? Figure 3 illustrates this trend, however the reality of the situation is more complicated that it appears, as the long-term structural increase in the average lateral length in the Permian is accompanied by cyclical changes in service costs. For example, when service costs collapsed in the first half of 2015, the impact of longer laterals on cost per foot accelerated, with D&C cost per 9,000 feet falling from $12 million in 2014 to less than $8 million in mid-2016, despite the significant increase in proppant intensity and the number of frac stages per lateral over the same period. On the other hand, the continuous increase in lateral length helped to mitigate a significant share of the service cost inflation effects in the second half of 2016 through the first half of 2018. As a result, well capex per unit of first year recovery increased from $30 in mid-2016 to only $35 to $36 in mid-2018, and additional improvement in the future is likely amid the new cost deflation trends which emerged in 2H18.
Therefore, despite the flat development in normalized productivity, Rystad Energy expects normalized well costs will keep trending downwards as long as the service cost cyclicality is ignored. With this in mind, the long-term improvement in Permian well economics will persist as operators continue to optimize lateral length and other design parameters.
Outside the Permian, major liquid basins exhibit similar development. We have recently seen some publications argue that productivity deterioration has occurred in US Shale liquid basins in 2019. We cannot confirm any of these observations with the use of empiric data. In the Bakken, Eagle Ford and DJ Basin, well productivity is observed at or close to all-time highs as of 1H19. Normalized productivity has exhibited rather flat development in most of these basins in recent years, but similar to the Permian, this trend was accompanied by a significant increase in lateral length in the Eagle Ford and DJ Basin, which allowed the most recent wells in these basins to achieve the best economics in the history of unconventional development.