An essential aspect in the development of shale and tight (S&T) assets revolves around well-placement strategies to maximize economic value. Reservoir simulation is a key tool that provides physics-based estimation of inter-well interference under multiple scenarios, but the complexity and high computational demands hinder its widespread application. To address this issue, we propose a new method for predicting the inter-well interference factor (IF), which quantifies the production loss caused by interference from neighboring wells. Our approach leverages the concept of diffusive time of flight (DToF) coupled with volumetric calculations at a granular level.

In our analysis, we compute the DToF for each well under two distinct scenarios: Single Development (SD) and Co-Development (CD). We logarithmically segment the pore volume into bins based on the DToF values. Building on this segmentation, we introduce a novel concept of Active Depletion Volume (ADV), which we calculated for both SD and CD scenarios to estimate IF. A common engineering practice in S&T development involves comparing various configurations of well spacing and fracture geometry while maintaining the same dynamic properties, such as fluid properties and well controls. Our methodology simplifies this process by requiring only a single simulation run to capture the reservoir's dynamic behavior using a calibration coefficient. This allows for efficient scaling to evaluate multiple scenarios.

To verify our method, we generate multiple placement configurations of three horizontal wells in a realistic tight oil reservoir by varying the well spacings and fracture geometries. For each configuration, four simulations are run to calculate the 20-year production for individual wells in SD and CD scenarios, and the IF estimated by our proposed method is compared to that calculated by simulations. Results show a good agreement between the IF profiles as a function of time. We then applied the method to a six-well case study completed in the Haynesville shale, considering 15 fracture network configurations. The predicted IF aligns well with the actual IF calculated using full-physics simulations, with errors not exceeding 20%. For this case study, our method significantly reduced simulation requirements from 105 to 3 runs, achieving more than a 30-fold speedup while preserving the true ranking of fracture configurations. This considerable efficiency boost enables more rapid comparisons and expands the capacity to evaluate a greater number of development scenarios.

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