Correlation of the hydraulic conductivity of fine-grained soils with water content ratio using a database

The saturated hydraulic conductivity of fine-grained soils is of great importance, as it is directly linked to many fundamental calculations in geotechnical engineering. In this paper, various models for predicting saturated hydraulic conductivity using simple soil parameters are reviewed. A simplified semi-empirical prediction equation using water content ratio (w/w_L) as the predictor of saturated hydraulic conductivity is calibrated using a large database called FG/KSAT-1358 (n > 1300) of saturated hydraulic conductivity (k_sat) measurements on fine-grained materials. The regression equation can predict the saturated hydraulic conductivity of the measurements included in the database to within plus or minus an order of magnitude around 90% of the time. To study other factors which may affect the values of k_sat, the database is then subdivided according to liquid limit level, silt or clay classification, hydraulic conductivity test method and sample condition. Some variations in the regression equations for each of the aforementioned subsets are observed, but the effect on the value of the exponent in the derived power-law relationships is relatively minor.