Managing pumping schedule during hydraulic fracturing operation is paramount to initiation and propagation of fracture systems. At the end of treatment when pumps are turned off, specialists have reported based on field datasets (e.g., microseismic and fiber-optics) that fractures might continue to grow, increasing reservoir contact area. However, no concrete answer has been presented to demonstrate this physical question yet. The objective of this work is to characterize strain fields when pumping stops using fiber synthetic data and consequently determine if fracture systems keep propagating in such conditions.
We simulate Distributed Acoustic Sensing (DAS) with an in-house code based on the Displacement Discontinuity Method (DDM). Simulation setup consists of a single hydraulic fracture and cross-well fiber deployment with operation and monitor wells distancing 300 ft. Fracturing process lasts about 60 min and when pumping stops, fiber monitors 30 min strain fields of 6 pressure drop gradients representing distinct fluid leak off scenarios. We perform two case studies to determine the pair extension-compression (i.e., polarity changes) in DAS time history plots when pumping is over: in case 1 fracture propagation terminates and in case 2 it does not.
Strain rate results of case 1 indicate polarity changes immediately at pump off instant for all fluid leak off scenarios: (i) fracture corridor sign switches from extension (i.e., opening) to compression (i.e., closing) and (ii) fracture neighboring sign shifts from compression to extension. On the other hand, DAS dataset of case 2 reveals that fracture may not close at monitor well location when pumping stops. The smaller the pressure drop gradient, the later fracture will start to close. Therefore, we infer that after treatment phase, hydraulic fracture continues propagating if no compression signature is identified in the fracture corridor present on DAS waterfall plot.
To the best of our knowledge this is the first work that addresses with synthetic strain data factors controlling fracture extension when pump is turned off. If such phenomenon is properly considered in the field, fracture geometric properties may be estimated with greater precision leading to the development of models that better characterize stimulated reservoir rock. Moreover, this project demonstrates that numerical simulation has significant value when used to assist the interpretation process of fiber-optics field data, which in turn is complex and holds a full potential not explored yet.
