T2 Relaxation Time Versus Mercury Injection Capillary Pressure: Implications for NMR Logging and Reservoir Characterisation

Abstract

Abstract A link between distributions of T2 relaxation time and capillary pressure curves has been demonstrated in several articles published over the last few years. Efforts to establish a link have mostly concentrated on high pressure air/mercury experiments (MICP). Establishing a link between T2 and MICP is important because it adds confidence to estimates of bound fluid, free fluid, and permeability from NMR logs. The link also provides another route to saturation height functions. However, experience working with NMR logs has shown that estimates of bound fluid are frequently under or overestimated and the default parameters used to determine bound fluid (e.g. T2 cutoffs) have to be adjusted to compensate for this. Often, little is known about how these adjustments relate to the actual pore system. Analysis of 176 core samples representing a broad spectrum of sandstone reservoirs demonstrates the link between T2 and MlCP is unpredictable. This result challenges use of simplistic pore models for predicting bound fluid and explains why bound fluid estimates from NMR logs are often wrong. However, there is sufficient similarity between data sets for T2/Pc functions to be developed locally from core on a well-by-well basis. These functions can be applied to NMR logs using a process of pattern recognition which identifies similar NMR response type's within logs and core samples. NMR response types and their associated T2/Pc function provide information about upscaling and flow units. P. 323