Parent-child wells are horizontal wells drilled in close proximity to each other in unconventional basins. Simulation work in the technical literature demonstrates how depletion and fracture communication between parent and child wells can lead to child well underperformance. High-level, basin-wide data analysis of unconventional basins confirms this effect. However, as completion designs evolve and more state-of-the-art horizontal wells are completed in these basins, it is necessary to revisit this analysis and make adjustments and additions to the previous body of work. Specifically, initial production differences between parent and child wells need to be correlated to cumulative production differences, and more analysis regarding the effect of timing and spacing are needed. In this study, parent-child well pairs for wells completed within the last seven years in nine different unconventional basins are identified using a Python code applied to Enverus public data obtained in November 2020. These basins include the Bakken, Delaware, Eagle Ford, Haynesville, Marcellus/Utica, Midland, Niobrara, Powder River, and Scoop/Stack Basins. Our Python code also performs calculations to create the necessary comparative metrics for analysis. Four cumulative production proxies are created and First 12 Months BOE (barrel of oil equivalent) is chosen as the appropriate metric for analysis. Basin-to-basin comparisons are conducted, and the effects of well spacing and infill timing are investigated. The study finds that as stated in the technical literature, child well performance increases with spacing and decreases with infill timing. We show that parent cumulative production (BOE) at child well completion is a better indicator of child well performance. Overall, these assessments can help operators manage child well underperformance and can help them understand the effects of differing well spacing and infill timing on child well performance in different US unconventional basins.

You can access this article if you purchase or spend a download.