Summary

The Granite Wash tight sand play represents a set of stacked Pennsylvanian-age formations in the Anadarko Basin spanning across the Texas Panhandle into western Oklahoma. The highly radioactive sands were eroded from granites, re-deposited, and later diagenetically altered to become heterolithic reservoir rocks. Large uncertainties in lithology, grain density, water resistivity, and other key rock and fluid properties have resulted in poor correlation to well performance. This is further complicated by the inability to predict fluid type and well deliverability, leading to additional scrutiny of logging programs.

Nuclear magnetic resonance (NMR) technologies have been employed in the reservoir characterization with varying degrees of success. Their application in Granite Wash characterization has faced challenges because of the low porosity of the formation and the current limit of tool capability for direct gas detection.

This study offers a practical method for integrating high-pressure mercury injection capillary pressure (HPMICP) data and NMR data from legacy whole cores, along with extensive conventional logging suites and borehole NMR. Using standard pore geometrical relationships between pore size from NMR data and pore throat size from HPMICP data, we generate pseudo-capillary pressure curves as a function of depth and calculate pseudo-PC based relative permeability/ fractional flow approximations. Furthermore, we use the effective gas and water permeability predictions (constrained by the dual water saturation model and NMR T2 relaxation times) to predict fluid phase and flow rates. This study reviews the critical assumptions required by this practical method and present the results as a guide for well productivity analysis.

This content is only available via PDF.
You can access this article if you purchase or spend a download.