Temperature effect on NMR surface relaxation
an abstract by Jiansheng

The reservoir carbonate sample studied has a porosity of 24% and a water permeability of 1.5 mD. Electronic microscopy of this rock exhibits the presence of crystals of CaCO3 varying in size for 1 to 20 mm. From mercury porosimetry, the pore throat radius distribution is narrow and peak at 0.5 mm. Electron Spin Resonance (ESR) experiments have shown the presence of Mn2+ paramagnetic ions. The sample has been cleaned according to IFP’s standard procedure. The USBM wettability test performed after cleaning indicates a wettability index of +0.50. During the wettability test, a spontaneous imbibition of water was observed but the cleaning procedure was not sufficient to remove some absorbed components from the surface, resulting in a non-strongly water-wet state.


Discussion and application to log calibration
From the various experiments described above, we can observe general trends describing the temperature dependence of the surface relaxivity (Table 4). For water- or oil-wet surfaces composed mainly of silica, (SiC grain packing, Clashash and reservoir sandstones), the water surface relaxivity parameters increase with temperature according to Eq. 2 to 4 with an observed effective activation energy in the range [-2; -1.7 kcal/mol], and the oil surface relaxivity parameters decrease with temperature with an observed effective activation energy in the range [1.5; 2 kcal/mol]. For water-wet or weakly water-wet surfaces composed mainly of CaCO3 (calcite grain packing, Lavoux limestone and reservoir carbonate), the water or oil surface relaxivity decreases with temperature with an effective activation energy in the range [1.9; 2.5 kcal/mol]. Note that we used a refined oil (dodecane) as the oil phase to determine the oil surface relaxivity. More complex fluids (mud oil filtrate or native crude oil) might complicate the analysis but the general trends should still be valid.


When laboratory and log data are compared in similar saturation conditions (i.e. at 100% water saturation for a water based drilling mud, or at oil-water irreducible water saturation for an oil based drilling mud), the relaxation times might be shifted by a factor equal to the right exponential term in Eq. 3. For instance, from 25∞ C up to 120∞ C, a shift factor of 2.1 is possible (taking DE = 2 kcal/mol), either in reducing or increasing the relaxation times depending on the type of surface. In practice, the temperature effect might be hindered by pore coupling when considering bimodal pore structures, or by mixed surface composition (quartz and calcite). There are also further complications related to the existence of static magnetic filed gradient present in the logging tools that prevent the measurements of long relaxation times.


On the basis of temperature dependence measurements of NMR relaxation times and a theoretical description of the solid-liquid interactions occurring at the pore surface, we have evidenced two different kinds of temperature dependence of the relaxation times of liquid in pores. The temperature dependence of the relaxation times of water confined in pores with silica surface is anomalous, that is to say, decrease with increasing temperature. On the other hand, relaxation times of oil in pores with silica surface increase with temperature. Relaxation times of water and oil in pores with calcium carbonate surface both increase with increasing temperature. The consideration of these characteristic temperature behaviors can be used as guidelines to compare NMR well logging data acquired at elevated temperature and laboratory experiments performed at room temperature.