Improved Characterization and Performance Assessment of Shale Gas Wells by Integrating Stimulated Reservoir Volume and Production Data

Abstract

Gas flow in shale gas reservoirs occurs primarily from ultra low permeability shale rocks through a complex network of natural and induced hydraulic fractures. Consequently, fracture parameters (conductivity and half length), fracture location and distribution are the dominant factors influencing well drainage volumes and shale gas well performance. Stimulated reservoir volume or SRV, estimated from microseismic event clouds or rate/pressure transient analysis, describes a measurement of overall reservoir volume impacted by fracture treatments. With SRV as well as the dynamic production/pressure response, reservoir simulation models can be calibrated to actual well performance in shale gas reservoirs leading to improved understanding, forecasting and future well placement.In this paper, we first introduce a novel approach for computing well drainage volume for shale gas wells with multistage fractures and fracture clusters. Next, we calibrate the shale gas reservoir model by matching the drainage volume with the SRV within specified confidence limits. The matching of the SRV is done in addition to the traditional history matching of production/pressure response and further constrains the estimation of fracture parameters. An evolutionary algorithm with design of experiments is used for the assisted history matching. Sensitivities to various parameters such as fracture conductivity, fracture half lengths and rock compaction have also been investigated. The proposed approach has been applied to a generic shale gas well designed after a real field case. The results clearly indicate the benefits of including SRV during history matching, leading to improved fracture/matrix parameter estimation and performance forecasting. Our proposed approach provides an important tool that can be used to optimize well placement, fracture treatments and improve the economics of shale gas plays.

Temperature-Prediction Model for a Horizontal Well With Multiple Fractures in a Shale Reservoir

“…These methods can reduce computational efforts comparing to the discrete fracture network model, but typically the analytical solution used in these models cannot capture the very long transient behavior in the matrix blocks exhibited by shale gas reservoirs (Cipolla et al, 2010a). In addition, Yin et al (2011) used an enhanced permeability area to approximate the enhancement effect by the network fractures as higher permeable matrix zone in the vicinity of the fractures. In this method, the hydraulic fractures were expressed as the higher permeable grids, and the enhanced permeability area exists along the fracture grid to consider the stimulated reservoir volume.…”

“…For the expression of induced or reactive complex fracture networks around the created hydraulic factures, we took the same approach presented by Yin et al (2011). This is one of the single porosity models with the simple approximation of the enhancement effect by network fractures.…”

“…In this work, we took the same approach with Yin et al (2011) who used an enhanced permeability area to approximate the enhancement effect by the network fracture as the higher permeable matrix zone in the vicinity of the fractures. In this method, the hydraulic fractures were expressed as the higher permeable grids, and the enhanced permeability area exists along the fracture grid to consider the stimulated reservoir volume (Fig.…”

“…However, these models require consideration of additional parameters such as matrix-fracture shape facture. In this work, in order to reduce the computational efforts for simulation and reduce the number of parameters to be considered, we take the approach of enhanced permeability area used by Yin et al (2011) in which the enhancement effect by the network fracture is considered as the change of parameter values in the vicinity of planar primary fractures. The network fractures are expressed as higher permeable matrix region comparing to the unstimulated matrix region.…”

Fast-Marching Methods for Complex Grids and Anisotropic Permeabilities: Application to Unconventional Reservoirs

“…We generate the "observed" bottom-hole pressure data using the finite difference reservoir simulation coupled with geomechanical modeling for the first 5 years of the production. History matching is performed using the genetic algorithm (Yin et al 2011;Xie et al 2012b) with model parameters and their ranges shown in Table 2. The genetic algorithm is carried out for 20 generations and each generation has a population consisting of 80 realizations.…”