Abstract
In order to more accurately characterize reservoir and hydraulic fracture properties from well performance, a workflow has been developed that effectively integrates variable quality data from a variety of sources. This workflow applies analytical techniques designed specifically for shale gas wells followed by as-needed numerical modeling. The analytical techniques can be applied to multiple wells through time to: a) identify groupings of like-performing wells, b) detect wells with anomalous behaviors, c) develop hypotheses about production mechanisms, and d) choose specific wells for more detailed analysis and numerical modeling. Numerical modeling provides the functionality needed for complex mechanism forensics, performance forecasting, and completion optimization studies.
Conventional numerical models typically use finite-difference grids, but these are neither sufficiently complex nor sufficiently flexible for shale gas reservoirs. For this reason, a finite-element modeling technology has been applied that places a large number of closely-spaced nodes near hydraulic fractures, "where all the action takes place" in the early life of a well. The finite-element technique also allows complex fracture geometries to be modeled.
This workflow, incorporating analytical and numerical solutions, has been applied to multiple shale gas projects, including industry consortia in the Haynesville (US) and Montney (Canada) shales and individual operator projects in the Woodford (US), Horn River (Canada), and Marcellus (US) shales. Through the application of these techniques, fracture and reservoir properties have been characterized and uncertainty associated with forecasted well performance has been reduced. This work has profound implications for quantifying gas reserves, understanding those factors responsible for variations in well performance, and for optimizing well spacing, lateral lengths, and completion techniques.
