Advanced signal processing and modeling for partial discharge diagnosis on mixed power cable systems

Power cable grids have been growing enormously over the years; due to their aging, there is a strong need for Partial Discharge (PD) measurement techniques in order to identify incorrectly installed or damaged cable joints or cavities in cross-linked polyethylene (XLPE) insulated and paper-insulated lead-covered (PILC) cables. Mixed power cable systems consist of a mixture of interconnected cables of various types. The Time Domain Reflectometry (TDR) is a standard technique for the location of PD in power cables. However, the localization process on-the-field is impeded by various interferences and noise. Additionally, propagation velocities of PD pulses traveling on the line differ depending on the cable type (XLPE or PILC cables), which leads to significant location errors. This paper provides automatable signal processing tools for increasing the detection ratio as well as location accuracy in the field of mixed cable systems. A mixed cable system model provides the possibility of analyzing PD data of even complex cable interconnections.