Rock thermal properties from well-logging data accounting for thermal anisotropy

•Technique for determining rock thermal properties from logging data is enhanced.•New technique allows determining thermal conductivity accounting for anisotropy.•Technique was applied for two unconventional reservoirs.•Uncertainty in thermal conductivity determination is less than 12 %.•Uncertainty in volumetric heat capacity determination is less than 5 %. The limitations of the existing techniques for in situ rock thermal property measurements and numerous cases with non-coring drilling determine the necessity for methods of rock thermal property determination based on well-logging data. Existing approaches for determining rock thermal properties from well-logging data are appropriate only for isotropic rocks. Since many rock types, especially organic-rich shales, exhibit a considerable degree of heterogeneity and anisotropy, advanced approaches for well log-based determination of rock thermal properties are highly desired. The implementation of the new thermal core logging technique, which provides continuous and high-precision measurements of the principal components of the thermal conductivity tensor and volumetric heat capacity from core samples, enabled the development of a new framework for the indirect determination of rock thermal properties. An enhanced technique for determining rock thermal conductivity and volumetric heat capacity from well-logging data accounting for thermal anisotropy and in situ thermobaric conditions is proposed and tested. This technique includes both the application of theoretical models and regression analysis of rock thermal properties, depending on the availability and quality of the input data. Three theoretical models involving a correction factor were compared to provide the best results. The experimental data of rock thermal properties inferred from the thermal core-logging and well-logging data from five wells (1630 samples) drilled through two highly anisotropic unconventional formations – the Bazhenov and the Domanic – were used as the basis of this newly developed approach. It is shown that rock thermal conductivity can be predicted from well-logging data accounting for thermal anisotropy with an uncertainty of less than 12 % and rock volumetric heat capacity with a total uncertainty of less than 5%.