A Physically Guided Porosity-Compensated Model for Shear-Wave Velocity Prediction in Sandstone Reservoirs

Accurate estimation of shear-wave velocity (Vs) is fundamental for reservoir geomechanics, as it directly influences the calculation of elastic properties used in Mechanical Earth Models (MEMs). However, shear-wave sonic logs are frequently unavailable in legacy or data-limited wells due to high operational costs and technical constraints. Therefore, reliable prediction of Vs has become essential. This study proposes a physically guided porosity-compensated compressional-wave predictor, Vp (1 − PHIT), derived from the Wyllie time-average equation, to mitigate porosity-induced variability and enhance sensitivity to rock-frame stiffness. The proposed model was evaluated using a multi-well sandstone and shaly sand dataset comprising 29,426 data points from 19 wells in the Sirte Basin, Libya. Its performance was benchmarked against five widely used global correlations using statistical metrics including R2, RMSE, MAE, and MAPE. The results demonstrate that the proposed model achieves superior predictive performance: R2 = 0.908, root-mean-square error (RMSE) = 0.00047 ft/µs, mean absolute error (MAE) = 0.00037 ft/µs, and mean absolute percentage error (MAPE) = 4.05%, outperforming conventional empirical correlations. The developed correlation provides a simple, physically interpretable, and field-applicable solution for predicting Vs in sandstone reservoirs and similar formations where shear-wave measurements are unavailable.