Lithology-specific influence of particle size distribution and mineralogical composition on thermal conductivity measurements of rock fragments

Abstract

Abstract Thermal conductivity data from reservoir rocks are key parameters for the analysis of heat transfer processes and heat flow patterns in geothermal exploration. Laboratory measurements of regional sample sets still provide the most reliable database. Since well cores are rarely available, rock fragments (e.g., drill cuttings) are the most common sample type for laboratory thermal conductivity measurements. However, the effects of the particle size of the rock fragments on the determination of thermal conductivity have not yet been discussed in a systematic study for different rock types. In this study, we used a line source device to investigate the influence of the particle size distribution of rock fragments and powder samples from different rock types on thermal conductivity measurements. The study focuses on three main objectives: 1. Does the particle size distribution of rock fragments impact thermal conductivity measurements of sample–water mixtures with a line source? 2. Are there lithology-specific differences in suitable particle sizes for measurements of thermal conductivity? 3. Is there an ‘ideal’ or an inappropriate particle size for measurements of rock fragments? Our study demonstrates that particle size effects on thermal conductivity determination of rock fragments and powders are unequivocal and that the accuracy of thermal conductivity determination can be improved by consideration of suitable, lithology-specific particle size ranges. The accuracy of measurement is related to the particle size distribution, mineralogical composition, competence under mechanical stress, and porosity of the sample–water mixtures during the measurement. For almost monomineralic limestones without significant clay volume fractions, powder samples with particle sizes ≤0.063\u202fmm are most appropriate. In the case of argillaceous sandstone, powder samples are an inappropriate sample type, and instead, coarse rock fragment samples with mean particle sizes of 0.9–1.3\u202fmm are recommended.