oa Nuclear Magnetic Resonance (NMR) Properties of Unconsolidated and Unsorted Materials Under Low-Field and Mid-Field Conditions

Failing dams and levees experience changes in porosity, permeability, and water content. Surface nuclear magnetic resonance (sNMR) may allow direct and noninvasive detection of water distribution in these dams and levees. in particular, the T2 response is dependent upon such hydrologic parameters as porosity and permeability that in turn affect water content. the properties of sNMR make it a potentially good technique for monitoring dams and levees. However, a better understanding of the NMR response to the subsurface under the conditions associated with leaky dams and levees is required to fully develop this approach. Though NMR measurements are routinely performed on unsaturated materials in the petroleum industry, these measurements are typically used to detect the locations of water or hydrocarbons in well sorted consolidated rock. Also, most laboratory measurements performed to date utilize a magnetic field that is much higher than the earth’s field, which is used for sNMR. Thus, there is a need to characterize the NMR response of materials typically used in dam and levee construction under the low-field condition and variable water saturations. A typical earthen dam is constructed of an impermeable clay core covered with a less impermeable layer of gravel, sand, and clay sized particles. Due to a lack of sorting and the wide distribution of grain sizes these sediments are typically unsaturated. We have carried out a series of laboratory sample measurements using a low-field and a mid-field performed on unsaturated, unsorted, and unconsolidated samples that are analogous to the permeable layer of dams so that we can better understand the sNMR responses under the conditions of a leaky dam. using spherical glass beads we prepared samples of varying porosities and permeabilities. We then performed both T2 and T2* lab measurements on the saturated samples using the low-field lab equipment. We then performed T2 measurements using mid-field laboratory equipment. these measurements were inverted to estimate the T2 time constantdistribution. Parallel to the saturated measurements, we repeated the same set of measurements on the gravity drained samples thereby simulating a naturally occurring condition of partial saturation. in this talk, we will present preliminary results obtained from these measurements.