The Middle Devonian Marcellus Formation as a Depositional System, a Stratigraphic Unit and a Reservoir System: Perspectives From Northern Pennsylvania
The Marcellus Formation is a succession of organic rich mudrock and mudstones forming part of the prograding fluvial/ deltaic Catskill Clastic Wedge during the second tectophase of the Acadian Orogeny (Ver Straeten, 2010). A series of hypotheses have evolved to better understand the Marcellus in terms of a reservoir body, each centered on the idea that if a detailed facies model was developed, created through a series of lithofacies (from core) and petrofacies (from wireline log) measurements, predictions of the distribution of reservoir properties and ultimately well performance would be possible. Proximal to the source of clastic sedimentation, the Union Springs Member is dominated by a succession of carbonate free black shales with limited facies cyclicity. This litho-stacking pattern occurs in an area where a salt basin formed in the Silurian which likely influenced the formation of a foredeep during the deposition of the Marcellus, creating a dysaerobic environment favorable to organic matter deposition. To the south and north of this paleo-deep there is conversely, a high degree of cyclicity in stacking patterns, where organic-rich shales give way to carbonate and gray clastic shales, which suggests that these areas may have been long-lived bathymetric highs, decreasing the paleo-water depth. The increase in cyclicity to the north correlates with a large salt-cored (Towanda) anticline, which suggests that salt migration in the Appalachian basin had begun prior to Alleghanian deformation. Lithofacies distributions in cored wells correlate well with a petrofacies model derived from electric logs and show a strong correlation with hydrocarbon-filled porosity and matrix permeability. From this integrated facies model, a more detailed sequence stratigraphic correlation, based on facies stacking patterns, has the potential to redefine the limits of reservoir bodies through the Appalachian Basin. Geologic attributes (lithologic, petrophysical, structural), for 425 producing wells, along with the engineering design (e.g. lateral length) were used in a multi-variable regression analysis to determine the controls on well productivity. Geology controls ∼40 of the 60% variance in productivity the model is able to constrain, with a significant portion of that explained by facies distributions. The results of this regression model have enabled us to become predictive of well performance and better understand what drives production in unconventional reservoirs.
AAPG Datapages/Search and Discovery Article #90189 © 2014 AAPG Annual Convention and Exhibition, Houston, Texas, USA, April 6–9, 2014