The research activity described in this report is aimed at both increasing the resilience and improving the reliability of the electrical system, with particular reference to the cable Transmission and Distribution lines.

Today the transmission and distribution of electrical energy are ensured mainly via two distinct types of electrical connections, namely overhead lines and cable lines. The former, which are much more widespread (especially in suburban areas), entail the disadvantage of a significant environmental impact. On the other hand, while allowing energy to be transmitted without substantially modifying the territory, the latter entail further critical issues, above all cable vulnerability (due to the mechanical stresses they must endure). One of the most critical elements within a cable connection system is the “joint”, namely the connection between the cables themselves. This configuration is necessary because each cable has a finite length both for construction reasons and for installation issues.

Joint construction is a very delicate operation and is carried out on site by highly qualified workers. Often having to operate in narrow environments subject to powdery material, in many cases the construction of such joints features impurities and imperfections that, once the line is energized, lead to a more or less rapid deterioration of the insulating material.

Defects prompt micro-discharges called Partial Discharges (SP), which over time discharges erode the insulating material placed between the conductive core and the external metal shield. This erosion is called “ Treeing ”. Once the dielectric strength of the material is compromised, the cable undergoes a destructive discharge that inevitably leads to the line being put out of service.

The use of current diagnostic techniques, which include both sensors and algorithms, does not enable the localization of defects nor, consequently, the prediction of their evolution.

The aim of this work is therefore to better and more accurately localize defects inside a joint and to recreate their evolution over time. This aspect is key to estimate the “health status” of joints with greater accuracy, as this allows preventing failures and consequently line disruptions.

Initially, a simplified and miniaturized mock-up of a joint was produced, in which a prototype of a sensor specifically designed and developed, called Nabla, was implemented. A known defect was added inside the mock-up, in order to generate partial discharges. These discharges were acquired and analyzed using the localization algorithms previously developed in order to identify their source.