Technique of rock thermal conductivity evaluation on core cuttings and non-consolidated rocks

The paper focuses on the development of a new technique to evaluate thermal conductivity of rock core cuttings and non-consolidated rocks. Very often consolidated cores cannot be recovered even during the well drilling in high porous and/or fractured reservoirs. At the same time reliable data on rock thermal properties are required for basin and petroleum system modeling, hydrodynamic modeling of hydrocarbon producing with thermal methods of enhanced oil recovery, interpretation of temperature logging data, heat flow density determination, etc. The proposed methodology is based on preparation of the synthetic composite specimens consisting of rock particles mixed with material-fillers (wax, water, air), measurements of the thermal conductivity of these specimens and evaluation of the thermal properties of the particles solving inverse problem of homogenization. The thermal conductivity is measured using the optical scanning technique which is a rapid, non-destructive, contactless methodology providing high accuracy (providing evaluation of thermal conductivity with uncertainty not more than ± 6% at confidence level of 0.95). The results of measurements are used as the effective properties in various homogenization schemes to evaluate properties of the inhomogeneities. To validate the approach, we calculated effective thermal properties using Mori-Tanaka-Benveniste and Maxwell micromechanical schemes and compared analytical predictions against experimental data. Our results show a good correspondence between micromechanical approximations and experimental measurements with average absolute error no more than 4%. Inverse homogenization problem was formulated to reconstruct thermal properties of considered rock cuttings from the known properties of remaining constituents and measured effective properties. Approximations based on inverse Mori-Tanaka-Beneveniste and Maxwell schemes are compared against experimental data measured on solid rock specimens. Satisfactory agreement between inverse solutions and experimental data were observed for both MTB and Mawell schemes with with average absolute error no more than 6%. The sensitivity analysis of the results to the shape of inhomogeneities corresponding to the entrapped air was performed.