Random Uncertainty in the Measurement of Matric Potential Using the Miniature Tensiometer

Abstract

Abstract The miniature (high-capacity) tensiometer is increasingly being used both in practice and in research to determine the matric potential in unsaturated soils. Although the potential which can be recorded with this type of device tends to be limited to a maximum of approximately 1500 kPa (a function of the air entry value of the porous element), simplicity of use, together with the ability to measure potentials under atmospheric conditions, makes this a unique piece of equipment for investigating and characterizing the hydro-mechanical response of unsaturated soils in the laboratory and in the field. The use of the miniature tensiometer is now well established and documented, and the particulars of a number of such devices, together with modifications to standard laboratory equipment to accommodate them, have appeared in the literature. The measurement of potential with the tensiometer, like any other measurement of a physical quantity, is not exempt from a certain degree of uncertainty. Additionally, the device is usually calibrated by applying a positive water pressure; however, when in use, the fluid filling the internal reservoir experiences a tensile stress. Although extrapolation of the calibration curve to negative values of water pressure is justified, it is nevertheless interesting to assess the level of uncertainty associated with a single measurement of potential in a soil. To this effect, data on gravimetric water content and matric potential from a statistically representative set of clay samples, carefully prepared and compacted in a similar manner to the same initial conditions, were analysed. This paper presents details and results from this exercise, together with the statistical treatment of the data.