The multi-stage hydraulically fracturing process of horizontal wells in tight reservoirs may form complex networks of fractures. The hydraulic behavior of such networks as revealed by numerical simulations may not correspond to linear flow. Nonetheless, methods for rate transient analysis that rely on the existence of linear flow are widely applied in the industry. It is possible that many unconventional wells are being incorrectly characterized. This paper shows a single analytic solution that can be applied to the characterization of complex networks of fractures. It encompasses different flow regimes (sub-linear, linear and sub-radial) to interpret pressure and flow rate behavior. We demonstrate physical mechanisms that may lead to sub-linear and sub-radial flow in complex networks of fractures and apply a solution originally derived for fractional dimension flow to fracture networks in tight porous media. In this new approach the diagnostic log-log plot commonly used in pressure transient analysis becomes the primary diagnostic tool to identify flow regime as well as some basic parameters. Those parameters are then used to formulate an analytic solution applicable to reservoir characterization, estimation of reservoir volume and decline analysis. Validation with numerical modeling and field cases are presented.
The last decade has seen an accelerated development of unconventional oil and gas reservoirs. Typical completions in shale gas wells consist of long horizontal wells with multiple stages of hydraulic fracturing separated by a relatively small distance. These multiple fracturing events by themselves, or in the presence of previously existing natural fractures, often lead to networks of fractures of much higher complexity than previously anticipated (Cipolla et al, 2011). The reservoir modeling tools now available make it possible to simulate the pressure response to fluid extraction in complex networks of fractures. We have observed that diagnostic log-log plots of pressure and pressure derivative for simulated drawdowns and buildups do not show linear flow behavior characterized by a derivative slope of 0.5. Standard plots such as pressure versus squared root of time however, may show a clear straight line for part of the data sometimes combined with a negative intercept that has been explained as the effect of a large area open to flow due to fracture complexity (Jiang et al, 2015). We have observed, as well as some other authors (Cipolla and Wallace, 2014), that the application of analysis techniques based on linear flow to complex fracture network simulation results may give parameters that are different from those used as input for the simulations.