Abstract

The development of unconventional plays has yielded valuable insights due to advancements in completion technologies and the identification of optimal drilling locations within plays. Among the notable techniques in the current commodity market, refracturing treatments have gained attention. Refracturing involves rejuvenating or creating new fracture networks in order to enhance flow rates and/or improve the Estimated Ultimate Recovery (EUR) of a well. Consequently, there is a growing interest in conducting performance analyses of refractured wells to identify key factors for candidate selection and post-production drivers that contribute to their success.

In this study, we propose a methodology that involves developing models trained on historical fractured and refractured wells to evaluate the impact of different parameters on the post-refrac performance of each well. These models are then used to screen and rank potential refrac candidates. For this purpose, we utilized the S&P Global Commodity insights public dataset to analyze the performance of fractured and refractured wells, leveraging data from more than 98,000 multi-fractured horizontal wells. We employed a machine learning approach (XGBoost + Factor Contribution Analysis) to identify potential candidates for refracturing. To evaluate the parameters affecting the post-refracture performance of wells, we collected data on 1127 refractured wells across 14 plays (including Bakken, Barnett, Haynesville, Eagle Ford, Austin Chalk, Woodbine, Wolfcamp Midland, Wolfcamp Delaware, Bone Spring, Scoop, Stack, Wattenberg, Woodford - Cana, and Woodford - Arkoma) with refractured vintages varying from 2006 up to 2022. We considered several potential features that could contribute to the post-refracture performance of each well.

Our analysis revealed that among the 11 potential features considered in this study, 3 showed a significant impact on the post-refrac performance of each well. These included rock quality, lateral length, and proppant per foot (refrac completion). Then, we proposed a workflow for screening and ranking refractured well candidates. Our methodology offers operators a rapid approach to identify potential refrac candidates, which can be further refined with more detailed information in subsequent steps of the process.

This content is only available via PDF.
You can access this article if you purchase or spend a download.