New Log And Core Analysis Methods For An Enhanced Shaly Sand Saturation, Using A ''M*'' Varying Mth The Clay Content

The Juhasz (Normalized Qv) method is a very attractive way of applying the Waxman & Smith theory, as it provides a full visual control of the clay parameters over the shaly water bearing intervals. Unfortunately, it does not work directly at each time: sometimes, the theoretical water straight line of the master cross-plot (F*Ct Vs Qvn) may appear as a curve. Even worse, the theoretical increasing trend of this line may appear as a decreasing curve, thus suggesting that the Waxman & Smith theory may not apply. This paper demonstrates that this is actually an effect of m* variation with increasing clay content. It is shown through both log and core analysis. First, core data including CEC and clay content measurements were used for the determination of the CEC of clay, which drives the clay end point of the Juhasz master cross-plot (F*Ct Vs Qvn). By using the log data within water bearing intervals, it was then possible to define a relationship between m* and the clay content. The saturation computed with this new variable m* showed a good fit with the core derived ones (capillary pressures and Dean-Stark). The absolute demonstration was still to be performed: these variations of m* had to be confirmed by direct core sample measurements. Unfortunately, the classical way of determining m*, which relies on four points Co Vs Cw measurements, cannot be applied to very shaly formations: due to their very low permeability, such samples cannot be swept by different brines as required by this method. A new technique was thus developed in order to achieve these measurements. Its accuracy was checked through a comparison with "four-points" m* when available on the same sample. The low cost of this method allowed an increasing number of measurements, always useful for assessing possible statistical variations. Then, by considering samples with varying clay content, a core derived relationship between m* and the clay content was established. This core derived relationship is remarkably close to the log derived one, thus validating both the variable m* concept, and its computation methodology from logs. As practical results, the Juhasz (normalized Qv) method can now be applied to a wide range of field cases, thus providing enhanced saturation values; the log analysis alone may be able to provide good quality results; the maximum confidence on these results is reached when log analysis is completed by core analysis, now possible through a new low cost methodology.