Secondary data value for geostatistical estimation of flow and transport parameters in low-permeability media

This study investigates the relative data value of different types of secondaryinformation to estimate the flow and transport parameters in low-permeability media. Data from four boreholes in two different clay formations are analyzed. Cross-validation is used to estimate the relevance of the different types of secondary data. The flow and transport parameters (hydraulic conductivity, diffusion coefficient and diffusion accessible porosity) are first estimated by kriging using primary information only and then estimated by co-kriging using primary information and one type of secondary information (grain size, gamma ray and resistivity). The kriging residuals, i.e. the difference between the estimated and the measuredvalues, of the different kriging and co-kriging variants are compared to evaluate the relevance of each type of secondary information. This analysis shows that in the different clays and the different boreholes, different types of secondary information result in the largest improvement of the estimates of the flow and transport parameters. It is not possible to determine the "best" type of secondary information for improving the estimation of the flow and transport parameters of low-permeability media. This probably depends on local factors such as the quality of the loggings and the amount of nearby primary and secondary data.