The goal of this study is to explain the full spectrum of flow patterns observed before and after closure during diagnostic fracture injection tests (DFITs) by considering the dynamic nature of fracture geometry, variable leakoff rate and afterflow volume caused by wellbore storage.
A fit-for-purpose simulation model is used to simulate DFITs and generate pressure responses in low-permeability (tight) reservoirs. The cohesive zone model in Abaqus® is used to simulate hydraulic fracture propagation and closure. A customized leakoff model incorporated into the software accounts for variable leakoff rate as a function of reservoir properties, fracture pressure, fracture surface area and exposure time. The afterflow is modeled by including a wellbore volume and accounting for wellbore storage. Results are compared to field data to explain the full spectrum of flow patterns and fracture dynamics observed in pressure transient analysis of DFITs.
The overall falloff period is interpreted, using PTA diagnostic plots, for relative magnitudes of afterflow, leakoff rate and fluid flow in the formation. Initially, afterflow is high, resulting in fracture expansion, which is characterized by a unit slope on the Bourdet-derivative plot. The afterflow does not necessarily end after the unit slope terminates; the end of fracture expansion is signaled by a characteristic hump on the derivative plot. During fracture expansion, the afterflow decreases and the leakoff rate increases due to the larger fracture area. When the leakoff rate dominates over afterflow, fracture closure mechanics can be conceptualized as a moving hinge-closure, where the fracture volume reduces, and fracture tip extension occurs as fluid is pushed to the tip of fracture (indicated by fluctuations in the derivative). The transition from afterflow to leakoff dominance and the moving hinge-closure manifest as a semi-horizontal trend on the Bourdet-derivative. Subsequently, a progressive fracture closure occurs gradually along the fracture, identified by an increasing trend or a sharp decline on the primary pressure derivative, depending on the conductivity. Different estimates of closure pressure will be obtained early and late in this process. The pressure behavior immediately after full closure is observed to be affected by the residual leakoff and the continuing afterflow. Once all of these fracture, wellbore and leakoff processes are abated, the reservoir response is observed.
This study provides a clear understanding of the different mechanisms affecting pressure behavior during DFITs for tight reservoirs in order to arrive at more reliable estimates of fracturing parameters and reservoir properties. As an example, mechanisms leading to false before- and after-closure radial flow identification are explained.