FDTD analysis of the laboratory measurement setup for measurements of frequency-dependent soil properties

Previously, it was shown through simulations and measurements that excessive sample length can lead to significant permittivity and resistivity errors in measurements of frequency -dependent soil properties. This paper presents recommendations for reducing these errors and possible corrections to existing measurement results. It has been demonstrated that the fringing capacitance makes a significant contribution to the permittivity error. However, additional unwanted capacitances (such as capacitance between connecting leads) can lead to even larger permittivity errors. Known expressions for fringing or stray capacitances (e.g., Kirchhoff formula) are useful for estimations of permittivity errors. However, FDTD simulations can be used for accurate calculations of the total excess capacitance of the measurement setup. The simulation results can then be used to determine the contribution of excess capacitances in the known soil models or future measurements. Voltage electrodes should not be very small to reduce possible resistivity errors. Either long rods or plates are preferred in measurements. The capacitance between the measurement circuits should be minimized (as even 10 pF can lead to appreciable errors). In addition, long wire lengths of the current and voltage measurement wires should be avoided.