Estimation of thermal conductivity of plutonic drill cuttings from their mineralogy: A case study for the FORGE Well 58–32, Milford, Utah

Abstract

The thermal conductivity (TC) of rocks plays a significant role in geothermal-related research. The matrix TC (MTC) could be indirectly inferred from modal mineralogy as a parameter incorporated into mixing models instead of the time-consuming direct measurement of cuttings. This approach lacks robust references due to limited study. In this study, we analyzed 44 plutonic drill cuttings data covering from diorite to granite collected from the FORGE Well 58–32, Milford, Utah, including XRD, MTC, and porosity. We employed eight mixing models commonly used in thermal studies to estimate MTC based on mineral composition (determined by XRD). The MTC depends on the quartz content, and there is a positive linear correlation between them, with R2=0.988. For the mean TC of individual mineral phases, the estimation of MTC with outliers removed shows that the geometric mean model provides a better statistical fit to the measured MTC with a mean absolute relative error (MARE) and a root mean squared error (RMSE) of 6.0 ± 4.4% (2σ) and 0.21 W/(m•K), respectively. By selecting the best mixing model considering the cutting's rock types, the MARE, and RMSE between the estimated and measured MTCs are significantly minimized. Therefore, we suggest that in case of the volume of plutonic cuttings is inadequate to support direct measurement of TC, the MTC of plutonic cuttings could be estimated with an acceptable magnitude error by assessing their mineralogy and using the appropriate mixing model according to the rock types.