Parent-Child Well Dynamics: Insights for Mitigating Frac Hits Using a Data Science Approach

Abstract During the development of unconventional reservoirs, interactions between parent and child wells, known as Fracture Driven Interactions (FDIs), are frequently observed. These interactions can either enhance, diminish, or have no effect on the performance of parent wells. This study addresses the industry's challenge of predicting the response of parent well to neighboring completion activities, known as frac-hit, by correlating these responses with controllable factors such as well spacing, completion design and operational strategy during offset completion activities. The analysis employs a metadata including 15,000+ wells from the Marcellus play. After a comprehensive quality control of the drilling, completion and production data, wells were grouped based on landing proximity, represented as 3D line segments using well coordinates. Parent and child wells were identified using spatial and temporal criteria, including drilling and completion dates. Frac-hit events were detected by analyzing the mean slope and variability of production profiles. In addition, frac-hit signatures were confirmed with child well completion dates. The assessment of the impact of a frac-hit on parent well involves evaluating production trends over ten months period, including five months before, the exact month, and five months after the detected frac-hit event. The clustered wells with that met the frac-hit criteria resulted in identification of over 1,000 parent wells, each with at least one significant frac-hit event. Through a data quality procedure, approximately 450 events with high confidence were selected. Predictive modeling and interpretability techniques were employed to analyze key independent variables and define thresholds. These thresholds were then used to construct a two-dimensional quadrant map to visualize the relationships between impact categories and influencing factors. Quadrant of impact vs well spacing and completion design demonstrates the minimum required spacing and fracturing criteria to avoid negative impact on parent wells. This study bridges multiple disciplines, combining geospatial analysis, statistical modeling and data science for identifying and interpreting frac-hits to optimize well performance. The methodology and outcome provide operators with a practical screening tool to mitigate detrimental impacts on parent wells as suboptimal drilling and/or completion designs often lead to excessive communication, reduced productivity, and financial inefficiencies.