Deriving Capillary Pressure and Water Saturation from NMR Transversal Relaxation Times

Abstract

Abstract Determination of water saturation (Sw) from conventional resistivity and porosity logs has proved difficult in several gas fields within Bolivia. To address this challenge, a method was developed to determine Sw profiles based on pseudo-capillary pressure curves (Pc) derived from transversal relaxation times (T2 distribution) measured by nuclear magnetic resonance (NMR) logging tools. Some of the Bolivian sand reservoirs, producing gas and condensate with a high gas/oil ratio, are characterized by a lack of resistivity contrast above and below the oil-water contact. This problem is attributed to complex mineralogy, including thin shale laminations within sand bodies of variable petrophysical quality, and mostly to very fresh formation waters that average a salinity of 5000 ppm of total dissolved solids. Also, the lack of lateral extension of each reservoir makes cross-well correlation very difficult. The core hypothesis of this original method is to assume that the relation between capillary pressure and pore throat sizes is similar to that between T2 values and pore-size distribution. A consistent scaling factor is used to derive a pseudo-capillary pressure from an NMR T2 distribution. After the pseudo-pressure is calibrated with capillary pressure measurements from laboratory-derived core data, it is combined with density (difference between the produced hydrocarbon and the formation water) and free-water level (FWL) information to compute Sw along the wellbore trajectory. Even so, in general terms, resistivity-based models are still the most consistent and documented foundation in order to compute water saturation; Examples demonstrate the success of this new standalone method of uncovering unforeseen hydrocarbon in place and obtaining accurate Sw as an alternative to the standard, but dubious or inadequate, resistivity method in Bolivia where cretaceous and carboniferous rocks with intergranular porosity are found within several sedimentary basins. Introduction When the conventional methods for the calculation of the water saturation (Sw) profile in the reservoir are not reliable for different reasons, a viable alternative is the calculation based on capillary pressure (Pc) curves. Using the Pc curve, the water saturation (Sw) in the reservoir can be calculated for any height above the free water level (FWL). By knowing the capillary pressure in a point, Sw can be easily calculated without the need of any standard resistivity model, whenever FWL and the fluid density of hydrocarbon can be determined with certain precision. As the capillary pressure is measured directly from cores or samples taken from the formation, normally it is difficult to find continuous information of Pc that allows creating a calculation throughout the whole formation thickness. In such sense, if it is realistic to acquire the value of Pc indirectly, for example by means of electrical logs, then one could virtually compute a continuous Sw. Some authors, Volokitin1 et al., have studied the option of obtaining a continuous curve of Pc from the T2 distribution of transverse relaxation times of a Nuclear Magnetic Resonance (NMR) tool. In our case, we have experimented with the processing of the data acquired from a CMR* tool that includes an application to calculate a continuous curve of Pseudo-Capillary Pressure (P-Pc) and the Sw, from the T2 distribution of transverse relaxation times, the density difference between water and hydrocarbon enclosed in the reservoir and the FWL. The method is essentially based on that the size of the poral throat has a relation with the size of the pore itself. Then as the capillary pressure, by means of calibration, can be derived from the size of the poral throat and likewise from the T2 time distribution of the CMR, we are subsequently able to get the pore size distribution. In conclusion, from a T2 time distribution curve one can, by means of calibration, obtain a Pseudo-Pc curve so that it is used in the calculation of water saturation Sw with the knowledge of the free water level FWL.