Time-Dependent Fracture-Interference Effects in Pad Wells

Abstract

Summary Hydraulic fracturing in shale formations induces microseismic events in a region we refer to as the microseismic volume. Many of these microseismic events are signatures of failure in the formation that are believed to be a result of induced unpropped (IU) fractures beyond the primary propped fracture. Areally extensive microseismicity may be evidence that these IU fractures occur and extend spatially beyond the propped fracture during pumping in many unconventional reservoirs. We present evidence that these fractures close over time after pumping is stopped and that this closure of IU fractures can have a significant impact on stress interference between fractures. To illustrate these effects, microseismic and radioactive-tracer data are presented for four laterals drilled and fractured from a single pad. Two wells on this pad were fractured with the consecutive-fracturing sequence, and the other two wells were fractured with the zipper-fracturing sequence. Geomechanical simulations were performed to model the pad scenario and explain the microseismic and tracer observations, with emphasis on the two different fracturing sequences. Our simulations show that the opening of the IU fractures results in significant temporary changes to the stress field in the rock. One consequence of this is that later fracture stages tend to propagate into the open-fracture networks of IU fractures created earlier because of stress reorientation. This can lead to inefficient usage of fluid, proppant, and capital because the region that is being stimulated has already been stimulated by the previous stage. By analyzing the net pressure, radioactive-tracer data, and microseismic data from the four-well pad, we show that these IU fractures close over time because the fracture fluid leaks off. This reduces the stress shadow, and subsequent induced fractures are no longer subjected to the significantly altered stresses, allowing for more-efficient fracture-network coverage by subsequent fractures in a horizontal well. On the basis of the data presented and computer simulations, we propose the idea of maximizing the time between fracturing in the microseismic volume of a recently fractured region (within operational constraints). The time required for the IU fractures to close can be estimated from our models and varies on the basis of the reservoir and fluid properties from several hours to days. One example of how this is accomplished in practice is zipper fractures. However, our work suggests that there also may be other fracture-sequencing strategies for accomplishing this.