Estimation of shear-wave velocities in unconventional shale reservoirs

Analysis of compressional and shear-wave sonic logs in seven organic-rich shale reservoirs exhibiting a wide range of velocities, total organic content, thermal maturity, fluid compressibility and mineral composition indicates that compressional-wave velocity is a strong predictor of shear-wave velocity as is the case for conventional reservoirs. Excluding data with high water saturations, a simple linear relationship is found between the compressional and shear-wave velocities in shale reservoirs in accordance with laboratory measurements. The relationship can be further refined for prediction purposes by local calibration or with a linear correction for total organic content. Correcting for compositional variation and fluid effects requires more complex treatment. An empirical rock physics model that explicitly incorporates these effects results in high accuracy (0.2% average mean error) in shear-wave velocity estimation for the entire data set. The model also exhibits good precision with mean absolute error of 2%. This is significantly better than what is achieved by rock physics models that do not explicitly consider fluid properties. These results are obtained without local calibration and without requiring correction for thermal maturity, degree of lithification or the shape or distribution of the solid organic matter. Explicit consideration of mineral composition, fluid substitution effects and amount of solid organic matter was necessary in this formulation to achieve this high average prediction accuracy with near zero bias as well as good precision across all formations and without local calibration. The successful use of Gassmann's equations in this algorithm suggests that errors in shear-wave velocity prediction in shale reservoirs due to violation of the assumptions underlying these equations are either small or self-compensating.