Decline curve analysis (DCA) is the most common method to forecast future production and to estimate ultimate recovery and reserves/well. The traditional form of DCA proposed by Arps is however restricted to boundary-dominated flow regimes. In unconventional shale plays, it is however likely that the transient flow regime may occur for the first few months or years of production. Consequently, the applicability of the traditional forms of DCA to early-time production data may not be appropriate.
This work is divided in to two main sections. In the first section, we apply some of the more recent decline curve models proposed for shale wells to production data acquired from the Woodford shale in Oklahoma. A comparison is then made between the different decline curve models in terms of their ability to replicate production history in a forecast mode. In the second section, we extend earlier work performed by other authors to compare well performance across different shale plays and over different time periods. The DCA presented in the earlier work utilizes a simple Arps decline; in this work, we utilize a composite decline curve that works for the linear transient flow regime and subsequent boundary dominated flows. Our work indicates that while the Arps decline curve analysis approach may be erroneous, in comparison to a more rigorous DCA, the errors are less than 20% in predicting EUR.