In unconventional reservoirs, the presence of natural fractures coupled with high pore pressures leads to the creation of complex fracture networks. During drawdown, the fracture network experiences large changes in the stresses which can affect the fracture conductivity, and hence the production rate. We present a workflow to find an optimum drawdown strategy in which the fractures can remain conductive while maintaining a high enough drawdown to maximize production.

A fully coupled geomechanical reservoir simulator is developed to simulate production from complex fracture networks. Flow in the fracture and reservoir domains is solved in two separate conforming meshes which are coupled through matrix-fracture transfer indices. The complex fracture network is represented as an explicit discontinuity in the reservoir domain which is essential to capture the stress variations in the vicinity of the fractures due to reservoir depletion and fracture closure. The fracture closure process is modeled dynamically using the Barton-Bandis contact relationship, and the fracture conductivity is determined using the fracture width and proppant concentration. This model is used to study the impact of drawdown strategy on fracture conductivity and well productivity.

It is observed that the estimated ultimate recovery (EUR) from complex fracture networks depends upon the connected fracture conductivity and the applied drawdown. A conservative drawdown strategy maintains the fracture conductivity for a longer period but results in a lower initial production rate.  As the drawdown is increased, the unpropped fractures close and can cause a large portion of the fracture network (the part behind the closed segment) to get disconnected from the wellbore. This reduces the available fracture area for production. Although an aggressive drawdown strategy results in higher initial production rates, it can lead to faster fracture closure, in turn resulting in a lower EUR. Impact of drawdown strategy on productivity is analyzed at different fracture closure rates.

We show that the optimum choke management strategy depends on the sensitivity of the fracture conductivity to stress. A coupled geomechanical reservoir model is presented that can simulate production with dynamic fracture closure in complex fracture networks to quantify this effect.