Evaluation of diffusion-driven material property profiles using three-wavelength interdigital sensor

The response of a three-wavelength interdigital dielectrometry sensor to th e moisture diffusion process in oil-impregnated transformer pressboard has been simulated using an empirical relationship between the moisture concent ration and the dielectric properties of the pressboard. A benchmark test of the moisture diffusion process has been developed with the purpose of comp aring alternative parameter estimation algorithms used in w-k (frequency wa venumber) dielectrometry. The results of simulations highlight characterist ic features of the multi-wavelength sensor response, such as the sensor res ponse delay from the start of the transient moisture diffusion process as a function of sensor wavelength, the influence of moisture boundary conditio ns, and the relation between the signal magnitude and variations of dielect ric properties. One of the parameter estimation algorithms, linear calibrat ed admittance-based estimation (LCABE), has been applied to both simulated and measured data. Adequate performance of the LCABE approach in the absenc e of strong discontinuities of dielectric properties in the electric field penetration region is demonstrated and contrasted to an electrically shield ed region case, in which the signal response becomes nonlinear. The propose d approach offers significant potential for the measurement of diffusion pr ocesses in various dielectrics, especially for cases with highly irregular geometry or material structure. Measurement results from moisture diffusion process monitoring are included. Parameter estimation of measurement resul ts with LCABE confirm its applicability to the monitoring of moisture dynam ics in transformer pressboard.