Search

Summary: Providing people with fresh water is one of the greatest challenges of the century. Since most of the world’s liquid fresh water is groundwater, the knowledge of aquifer storage properties is essential. Moreover, there is a need to focus research on poor aquifers (i.e. capable of providing about 100m3/day) which will play an increasing role for supplying many human communities. This paper concerns a study carried out in a clayey sandstones aquifer in Northern Cambodia. Conventional hydraulic methods used to characterize aquifers are costly, time-consuming and thus they are usually not used in most of the water projects in developing countries. Therefore, geophysical methods can be useful if they improve aquifer characterization. As compared to other non-invasive geophysical methods, magnetic resonance sounding (MRS) is selective to groundwater. MRS results are the distribution of both water content and pore-size related-parameters as a function of depth. However, relationships between the field scale MRS results and hydrogeological storage-related properties have not been well established yet. We present in this paper a comparison of MRS results with both specific yield calculated from pumping tests and effective porosity calculated from tracer tests. We found that the MRS water content is equal or higher than the specific yield and the effective porosity, thus indicating that MRS also measures capillary water in unsaturated zone and part of the bound groundwater attached to the aquifer solid matrix. We also found that the MRS pore-size parameter is linearly correlated with both the effective porosity and the specific yield, thus suggesting that the hydrogeological storage properties are mainly controlled by the size of the pores of the aquifer. Consequently, we adapted an approach used in the oil industry for differencing gravitational water from capillary water and from bound water, based on the MRS pore-size parameter. In the clayey sandstones of Cambodia, our approach named MRS apparent cutoff time approach, allowed calculating specific yield with an average error of 23% (which is far less than the previous published results), and for the first time it allowed calculating effective porosity (with an average error of 11%). We conclude that the MRS apparent cutoff time approach is useful for estimating aquifer storage properties down to 50–80m deep, in a single day and at an affordable cost. [Copyright &y& Elsevier]