Approaches to Subsurface Characterization of Unconventional Reservoirs – Marcellus Shale of the Appalachian Basin, USA
Critical factors to improve regional and local knowledge of variations in the productivity of unconventional shale reservoirs, such as the Middle Devonian Marcellus Shale of the Appalachian basin, are to improve our understanding the depositional dynamics of genetic units and map the distribution of lithofacies, which are amenable to both hydraulic fracture stimulation and have sufficient free and adsorbed gas content. Using an integrated core and log dataset from more than 1,000 wells from across the basin, a 3D shale model for the Marcellus Shale and related units is constructed using a sequence stratigraphic framework. The model is based on log derived estimates for mineral composition, rock geomechanical and total organic content (TOC). Core data were integrated with log data using multiple petrophysical and geostatistical approaches.
Total organic content was estimated and mapped from simple interpolation of gamma-ray logs and a modified “Passey” DlogR Method that was calibrated to core data. A set of eight derived parameters, commonly applied to unconventional reservoirs, was generated from abundant modern log suites and provided input parameters to an artificial neural network (ANN) model trained with the more limited data from core and advanced spectroscopy logs. The ANN model better define mineralogy and geomechanical properties in single wellbores and across large regional areas. The interpreted sequence stratigraphy and depositional environments constructed from a series of cross-sections were used to constrain 3D mineralogy, organic content and lithofacies models. The 3D model was compared to hydrocarbon production data to better understand regional variations in productivity in the Marcellus.
AAPG Search and Discovery Article #90154©2012 AAPG Eastern Section Meeting, Cleveland, Ohio, 22-26 September 2012