Practical Application of NMR Logging in Carbonate Reservoirs

Abstract

Abstract Nuclear magnetic resonance (NMR) data acquisition and interpretation in carbonate reservoirs is much more challenging than in sandstones, where it is a well-established technology. Heterogeneous porosity distribution, a broad range of pore sizes, a wide variety of complex textures, and low surface relaxivity combine to complicate the picture considerably. The successful practical application of NMR in these reservoirs requires the development of acquisition and interpretation techniques specifically suited to the task. In carbonate reservoirs dominated by intercrystalline or intergranular porosity, NMR can deliver accurate estimates of porosity, permeability, bound-fluid volume, and residual oil saturation. In vuggy, heterogeneous carbonates more complex interpretation models, based on the integration of whole-core and log data, are required for reliable answers. NMR answer products, based on these new techniques, are presented and validated with core data and by comparison to other logs. Introduction In many clastic reservoirs the CMR Combinable Magnetic Resonance tool has proven its ability to easily and accurately provide a number of answers not possible with conventional logging tools. From a single measurement of signal amplitude and transverse relaxation time (T2), it is possible to determine porosity, permeability, and bound- and free-fluid volumes and to estimate residual oil volumes. Extending this success to the carbonate reservoirs of West Texas is the focus of this study. CMR interpretation in these formations is not always straightforward and normal acquisition parameters are not necessarily sufficient to produce data relevant to the task at hand. For instance, under normal reservoir conditions, the oil signal and the water signal cannot be differentiated in most carbonates. Also, permeability estimated using the same simple treatment given to sandstones does not always match up well with core permeability. Despite these hurdles, quality answers are still attainable. The CMR* tools' accurate lithology-independent porosity is often critical in these complex carbonate reservoirs. Correct bound-fluid volumes are easily obtained using the right cutoff. Good permeability estimates are possible in carbonates, although this may initially require calibration versus core data and other logs in each field. And finally, a simple mud- doping procedure will allow the correct determination of residual oil saturation (ROS). NMR Petrophysics of Carbonates Petrophysically speaking, the most obvious difference between carbonates and sands lies in the heterogeneity of porosity distribution. In general, carbonates can be said to possess a wider range of pore sizes and geometries than sandstones, which are homogeneous and predictable by comparison. This gives rise to a number of physical properties in carbonates that directly affect NMR measurements. First, there are the properties that affect the T1 and T2 distributions of the formation. Because a wider range of pore sizes occurs in carbonates, the T2 distribution will generally be more dispersed than in sands. The largest of these pores will result in very long relaxation times; we show that this directly impacts logging speed and interferes with residual oil measurement. Additionally, an inherent matrix property of carbonates, low surface relaxivity, makes for longer relaxation times (Timur, 1972). Sandstone reservoirs consistently contain about 1% iron by weight. This results in a surface relaxivity of about 15 microns/s. By contrast, a typical carbonate matrix contains less impurities and has surface relaxivity in the range of 5 microns/s (Chang et al, 1994). P. 217^