How Frac Crew Efficiency May Impact Future Frac Fleet Demand

OPEC recently revised downward its outlook for global oil demand to an average of 90.2 MMBOPD (million barrels of oil per day) in 2020. That reflects an average 2020 demand contraction of 9.5 MMBOPD year-on-year, or close to a 10% reduction.

For 2021, it expects global oil demand to grow by 6.6 million bpd to an average of 96.9 MMBOPD next year. For the United States, EIA expects U.S. crude oil production to fall from an average of 12.2 MMBOPD in 2019 to 11.4 MMBOPD in 2020 and 11.1 MMBOPD in 2021.

US Shale represents a growing share of all US (crude and condensate) production. As we infer from the graph below for March 2019, 54% of all US production came from shale; 14% from deep water; and, 32% from conventional. 

This graph shows US conventional + deep declined from 5.3 to 4.9 MMBOPD between previous peak and trough of Feb 2015 to March 2016. At the same time shale went from 4.3 to 3.6 MMBOPD, so shale took ~2/3 of the reduction. As shale represents a bigger portion of US production today, the expectation is that the shale “share” of the 2020 reduction will be 75 - 85%.

That means US shale production is expected to fall from an average of 7.1 MMBOPD in 2019 to somewhere around 6 MMBOPD average in 2021.

What does this mean for frac fleet activity?

We developed a simple method to determine how these expected production levels tie to a frac fleet count to sustain it. To do so, we applied the following steps:

First, we use US production as an input parameter. This depends on many other things we cannot predict in this simple model – but other entities like OPEC and the EIA have hopefully done this for us.

Second, we determine US production by yearly vintage. The plot below has US shale production in black (from Liberty’s database of ~100,000 liquid-rich wells), as composed as the sum of yearly production by vintage of the completion. 

Then we simply divide yearly production and its subsequent decline for wells frac’ed in that year by average yearly frac fleet count. Note that the red line in the plot above shows a yearly average for fleet count that originates from public data provided by Primary Vision. The next plot below is a simple division of the yearly vintage production by the average number of fleets that were working in that year. 

This plot show the progress of the shale revolution in service company throughput improvements – year 1 peak production per fleet (for all wells that were completed throughout a calendar year) went from about 1,500 BOPD/fleet to about 12,000 BOPD/fleet. As more clearly indicated by the title plot for this article, yearly production per fleet appears to have plateaued. 

Why a plateau? This seems counter to our earlier assessment of massive efficiency gains in frac crew throughput since 2012: a 2x increase for oil equivalent production per oil & gas employee, a 3.5x increase in oil equivalent production per frac company employee and a whopping 8x increase in proppant mass pumped per frac company employee per year.

US oil production per liquid-rich frac fleet has steadily increased through 2017 due to higher production per well and higher efficiency frac crews. The reason for the production plateau since then is that frac crew efficiencies have been balanced by lower US well productivity from infill drilling and a slow but gradual move to lower reservoir quality. 

For our frac crew forecast model for 2020 and 2021, we used the 2017-2019 plateaued yearly production per fleet and its associated legacy decline data, assuming further frac crew efficiency gains would be equally offset by infill drilling and lower reservoir quality.

To maintain about 5.5 MMBOPD US shale production (slightly below the EIA estimate, so our low-case scenario) would require an average of ~165 liquid-rich frac fleets for 2021. Together with about 30 – 40 gas-rich frac fleets, mostly in the Marcellus, Utica and Haynesville, this would mean about 200 frac fleets are required in the Lower 48.

Most of these crews will have to work hard to offset production decline from older wells. To increase oil production by another 1MM BOPD per year (slightly above the EIA estimate for 2021 – our high-case scenario), an additional 85 liquid-rich frac fleets are needed to bring that total of liquid-rich frac fleets to about 250.

Will fleet efficiencies remain at 12,000 BOPD/fleet? There is certainly some room for improvement through the current relaxation of stage intensity goals and the industry’s focus on cluster intensity and efficiency. This should allow us to do fewer stages in a well, while keeping proppant and fluid intensities the same, allowing us as an industry to complete more lateral feet per day. The industry is also transferring fast to solutions like monoline that can reduce wear and tear on auxiliary equipment and improve pumping efficiencies by tens of minutes every day.

The question is if this will outweigh the slightly lower production levels per lateral feet we see from wells that are completed in slightly lower quality rock and are starting to show some interference from parental neighboring wells.

But it remains difficult to make predictions, particularly about the future. I will review the quality of my 2021 frac crew prediction in about 15 months. In the meantime I will take your feedback.