March 2017
Shaleprofile.com has been able to collect Permian well data, allthoug partly only in aggregated form. The average well profiles shown there are aggregated profiles aligned on the first production month. Because the ramp up to the peak production month of a new well varies between one and several months, the obtained average peaks are not identical with average IP30 values. For example, for all Bakken wells with 2014 initial flows, the peak shown in shaleprofile.com is at 405 bo/d while the IP30 average for wells having their IP30 in 2014 is 559 bo/d.
But a Permian well curve, starting with the IP30 value, can be derived with the knowledge of the corresponding Bakken curve and by comparing cumulative production data of both plays.
The table below show cumulative oil production data for average Permian and Bakken wells for the years 2014 to 2016, drawn from shaleprofile.com in March 2017. The numbers represent the well averages for all lateral wells with initial production in those years.
| Barrels oil – cumulated production per average well (bo) |
|||||||
| Permian |
pre peak | 3 mo | 6 mo | 9 mo | 12 mo | 24 mo | y2/y1 |
| 2014 | 5278 | 24758 | 43438 | 56346 | 66801 | 95885 | 0.435 |
| 2015 | 6019 | 30652 | 55258 | 72692 | 86555 | (123700) | 0.429 |
| 2016 | 7824 | 38761 | 74080 | 98535 | |||
| Bakken |
|||||||
| 2014 | 8490 | 30989 | 56324 | 75537 | 90418 | 132003 | 0.460 |
| 2015 | 9052 | 32564 | 61917 | 84646 | 102425 | 148475 | 0.449 |
| 2016 | 9805 | 40362 | 75531 | 101389 | 118231 | ||
The data for 2016 are not final and the data for 2015 are preliminary for the 2nd year, with the Permian data probably subject to more important revisions than the Bakken data . Taking that into account, the average Bakken well has outperformed the average Permian well until 2016. For 2014 and 2015, the accumulation of production over time was relatively slower for Permian wells, suggesting a steeper well production decline rate. The last column, which indicates the ratio of y2/y1 production, indicates that the Permian well production decline was also steeper in the second year. The preliminary data for 2016 do not suggests a significant performance difference between Permian and Bakken wells.
The table also shows a column singling out the pre “peak” volumes (the production volumes before the peak month production shown in shaleprofile.com). The ratio of pre peak volume relative to the post peak cumulative volumes is lower in the Permian than in the Bakken, suggesting a faster production ramp up.
For a given year, the well population constituting the wells with initial flow of shaleprofile.com is slightly younger than the wells having their IP30 in that year, but the well population is between 85% to 90% identical.
The logic of using IP30 based curves
The well production profiles and the related cost model used here are based on IP30 values and decline curves starting with the IP30 month. The rationale is that the IP30 month is the first full production month and that the production level of this month and the production data of the following months determine the quality of a well. The ramp up phase to the IP30 month can last between 1 and 6+ months. That’s when the well gets tested and adjusted. Production during that phase is intermittent and erratic, not indicative of well quality.
Estimate of Permian IP30 and subsequent decline rates for the year 2016
The preliminary cumulative data show that the average 2016 Permian and Bakken wells have very similar shapes after the pre peak phase, with the Bakken well outperforming the Permian well by roughly 3%. That might be different when the final data are in. Proppant producers have reported an ongoing remarkable increase of proppant/well in the Permian throughout 2016, which could mean that the IP30 of the wells have continued to increase and closed the gap with the Bakken wells, possibly at the expense of the well decline rate.
Natural gas production
The evolution of the wet natural gas production of the wells has been analysed here.
Economic Model assumption
For 2016, IP30 oil of Permian and Bakken wells are considered to be similar. For the total pre IP30 volume of the Permian well 0.55 x IP30 is retained (vs 0.70 x IP30 observed for the Bakken well).
The second year y/y oil production decline rate of Permian wells is considered 2% worse than that of Bakken wells. It’s assumed that after the first two years, the y/y oil production decline rates in both plays become identical (see here)
Outlook
The Permian area seems to have more unused sweet spot capacity to maintain 2016 IP30 levels, compared to the Bakken area. On the other side, the Permian well performance increase has been partly the result of the use of very high levels of proppants, fracking fluids and power. Their expected cost increases could make the extreme use of these completion factors uneconomical. A reduction of these factors could have a negative impact on IP30 numbers.
