The study estimates initial effective fracture pore-volume (Vfi) and fracture volume loss (dVef) for 20 wells completed in the Montney and Eagle Ford Formations. Also, it evaluates the relationship between dVef and choke-size strategy. We conduct this study through the following 3 key steps. First, we apply rate-decline analysis on water flowback data from these Montney and Eagle Ford wells to estimate the ultimate water recovery volume, approximated as Vfi; Second, we estimate dVef using a fracture compressibility relationship to evaluate the fracture volume loss of Eagle Ford wells; Third, we investigate the effect of choke size on dVef for the Eagle Ford wells with fastback and slowback strategies.
Semi-log plots of flowback water rate versus cumulative water volume for the Montney and Eagle Ford wells show straight-line trends which represent harmonic decline. The estimated Vfi accounts for about 84% and 26% of the total injected water volume in the Montney and Eagleford wells, respectively. The results show that about 10% of the fracture volume can be lost during flowback. This loss in fracture volume mainly happens during early flowback period and becomes minimal during late flowback period. Comparative analysis shows a relatively higher dVef for fastback wells compared with that for slowback well, indicating that slowback may slow down the loss in fracture volume. This paper proposes a method to estimate initial fracture volume and investigate the loss in fracture volume during flowback processes. Field data analyses lead to an improved understanding of the factors controlling water flowback and effective fracture volume.
The analysis of flowback rate and pressure data is used to characterize fracture networks in unconventional reservoirs. As such, several flowback models have been proposed and applied on early flowback data to estimate effective fracture pore-volume (Vef) for shale and tight gas/oil wells (Abbasi et al. 2012, 2014; Xu et al. 2015, 2016, 2017; Ezulike et al. 2016; Fu et al. 2017). However, input parameters such as fracture compressibility are generally unknown or hard to measure. This can lead to high uncertainty in estimates of output parameters like Vef.
