The use of nuclear magnetic resonance T(2) relaxation measurements in the characterization of the porous microstructure and reservoir properties of sandstones

NMR T{dollar}sb2{dollar} relaxation distributions were generated from the raw decay curve using two Pennsylvanian sandstone suites from Oklahoma. The T{dollar}sb2{dollar} distributions were calibrated with pore size distributions obtained using petrographic image analysis (PIA) procedures in order to determine the petrophysical relevance of T{dollar}sb2{dollar} as a measure of the pore system. Properties derived from the T{dollar}sb2{dollar} distribution, namely log mean T{dollar}sb2{dollar} and free fluid index (FFI), show a strong relationship with the portion of porosity controlling fluid flow, indicating that T{dollar}sb2{dollar} is capable of resolving the effective porosity. Predictive models relating pore size distributions derived from image analysis to the T{dollar}sb2{dollar} distribution reveal that different parts of the T{dollar}sb2{dollar} distribution are related to specific portions of the pore size distribution. These models have shown that the T{dollar}sb2{dollar} distribution is capable of predicting the amount of porosity contributing to flow to a high degree of accuracy and that the predictive ability is stronger for the pores which compose this portion of porosity as a whole rather than individually.; This indicates that T{dollar}sb2{dollar} is measuring clusters of porosity acting as a whole rather than the individual pores of which these clusters are composed. Analysis of the relationships between NMR endmembers and PIA pore type frequency distributions reveals that T{dollar}sb2{dollar} is primarily a function of the spatial arrangement of porosity and not individual pore size. NMR endmembers with large mean T{dollar}sb2{dollar} values are associated with the pore types which compose the clusters of porosity contributing to flow while the NMR endmembers with smaller mean T{dollar}sb2{dollar} values are related to the isolated pore types which compose the ineffective porosity, regardless of size. This is evidence that pores associated with each other in these clusters are having a combined or coupling effect on the NMR signal. Lastly, the weighted mean T{dollar}sb2{dollar} of these clusters of porosity show a strong linear relationship to the mean pore size, allowing for the empirical determination of the surface relaxivity.