Assessing the impact of quartz content on the prediction of soil thermal conductivity

Abstract

Accurate predictions of soil thermal conductivity k are strongly influenced by the volumetric fraction of quartz, Θq, data for which are very scarce. This paper reveals a new approach to estimate Θq from measured k records. First, an equation that relates the normalised k of soil (Ke) to the degree of saturation Sr is fitted to experimental k data, and the k at full saturation is assessed; then Θq is calculated from a geometric mean model. This modelling approach was applied to the k data of 10 Chinese soils obtained by Lu et al., also containing k measurements at full dryness, and to soils investigated by Kersten with measured quartz content data. The fitted Θq data for Chinese soils are noticeably different from the sand mass fraction, commonly assumed in the past as an equivalent of quartz content, consequently leading to irrational k estimates. Acceptably good agreement was obtained between fitted and measured quartz content for Kersten's soils. Five Ke(Sr) functions were tested against the experimental data for ten Chinese soils, supplemented with calculated k at full saturation. Overall, the normalised function by Lu et al. was the most suitable for the soils tested. The assumption that Ke(Sr) = 0, applied to Johansen's model extended to full dryness, worked well for fine soils, and was acceptable for coarse soils.