Search

This study compares different polymer-nanofiller blends concerning their suitability for application as insulating thermoplastic composites for High Voltage Direct Current (HVDC) cable application. Two polymer blends, PP/EOC (polypropylene/ethylene-octene copolymer) and PP/PP-HI (polypropylene/ propylene - ethylene copolymer) and their nanocomposites filled with 2 wt.% of fumed silica modified with 3-aminopropyltriethoxysilane were studied. Morphology, thermal stability, crystallization behavior dynamic relaxation, conductivity, charge trap distribution and space charge behavior were studied respectively. The results showed that the comprehensive performance of the PP/PP-HI composite is better than the one of the PP/EOC composite due to better polymer miscibility and flexibility, as well as lower charging current density and space charge accumulation. Nanosilica addition improves the thermal stability and dielectric properties of both polymer blends. The filler acts as nucleating agent increasing the crystallization temperature, but decreasing the degree of crystallinity. Dynamic mechanical analysis results revealed three polymer relaxation transitions: PP glass transition ( $\beta$ ), weak crystal reorientation ( $\alpha 1$ ) and melting ( $\alpha 2$ ). The nanosilica introduced deep traps in the polymer blends and suppressed space charge accumulation, but slightly increased the conductivity. A hypothesis for the correlation of charge trap distribution and polymer chain transition peaks is developed: In unfilled PP/EOC and PP/PP-HI matrices, charges are mostly located at the crystalline-amorphous interface, whereas in the filled PP/EOC/silica and PP/PP-HI /silica composites, charges are mostly located at the nanosilica-polymer interface. Overall, the PP/PP-HI (55/45) nanocomposite with 2 wt.% modified silica and 0.3 wt.% of antioxidants making it $a$ promising material for PP based HVDC cable insulation application with $a$ reduced space charge accumulation and good mechanical properties.

Starting from measurements performed on insulation models, estimation of partial discharge inception voltage, PDIV, and inference of the different behavior of PD for a MV cable models are presented and discussed. The estimations are validated through laboratory testing on a cable with an artificial defect. It is shown that PDIV under DC can become considerably higher or lower than that under AC, depending on temperature and material properties, which affect PD repetition rate as well. However, resorting to the model presented in this paper, measuring PD under AC will help to identify and validate the results of the PD DC measurements.

This paper has the purpose to derive an analytical model ("continuum" model) able to describe the behavior of partial discharge (PD) repetition rate and amplitude, occurring in a cavity embedded in polymeric insulation, as a function of the frequency of the supply voltage, going from AC power supply frequency, 50–60 Hz, to DC. In the range between DC and 50–60 Hz focus is made on data coming from tests under AC sinewave with very low frequency (VLF) such as 0.1 Hz and 0.01 Hz, which are commonly used for cable testing. It is shown that the proposed "continuum" model can provide reasonably good fit to the experimental results obtained in the range DC to 60 Hz, regarding PD repetition rate and amplitude. To reach such result, the equivalent circuit is modified from that commonly used and made by fully-capacitive or resistive components, in order to take into account the change of polarization mechanisms which, depending on dielectric material, may play a non-negligible role to establish the repetition rate from low frequency to DC power supply. In addition, the residual voltage after a PD event has to vary with frequency to reach good fitting. Also, it is shown that PD amplitude under DC and VLF can be lower than under AC 50–60 Hz due to the delay time of the firing electron, thus experimental PD amplitude varies with frequency depending on material and defect typology and location.

New nanostructured insulating materials are investigated, within the European project GRIDABLE, to be used for DC cable and capacitors. Besides electrical, thermal and mechanical properties, and life behavior, work is being done to evaluate their capability to endure highly stressing conditions as those cause by the inception of partial discharges, PD. This paper, in particular, analyzes what happens during energization of a DC cable or capacitors, when voltage goes from zero to the nominal value in a few seconds, while the internal electrical field takes longer time to reach the DC steady state configuration. During an energization transient, indeed, electrical field in insulation, and insulation defects (as cavities), is driven by permittivity, not conductivity as in steady state. Hence, PD might occur with high repetition rate, which would not occur, or at much lower repetition rate, in steady state. The way to evaluate the time constant of the transient, that is, through charging current measurement, is described and successfully fitted to the results of PD measurements performed on two types of polypropylene, PP: neat and nanostructured, and a cross-linked polyethylene, XLPE, having significantly different electrical characteristics.

In this paper the AC and DC partial discharge inception voltages, PDIV, are modeled and validated through measurements performed on multi-layer flat specimens with artificial defects. The results highlight that, depending on temperature and material conductivity, there can be a large difference between PDIV under DC and AC, which depends on insulation temperature, thus, e.g., cable loading. Indeed PDIV-DC can be much larger than PDIV-AC at room temperature, but the two values can be very close (with PDIV -DC being even smaller than PDIV-AC) at maximum operating temperature, i.e. full load.