*4.1. Partial Discharges*

PD has been established to evaluate different polymeric materials for application in HV apparatus, be it cable, electrical motors, transformers, etc. PD is defined as a localized electrical discharge that partially bridges the conductors' insulation and maybe closer or far away from a conductor [100,101]. In another context, the PD is regarded as an incomplete breakdown of the HV insulation system. PD activity mostly relies on the electric field strength applied to a specific area and its non-uniformity. For short gaps, PD possesses fast and slow-rise-time pulses that generate small electrical sparks or arcs or pseudo glow discharges, or pulseless glows [102,103]. Four main types of PD occur Corona, Surface discharges, cavities, and electrical trees (Figure 8).

**Figure 8.** Types of Partial Discharge.

Internal discharge is a common phenomenon in voids within solid or liquid dielectrics [104]. These voids can embed in the insulation material during the manufacturing process of the apparatus. This form of discharge is extremely important and has a significant effect on polymeric insulation. A surface discharge, on the other hand, occurs at the HV insulating with a high tangential electric field [105]. Internal discharge is a common phenomenon in voids within solid or liquid dielectrics [104]. These voids can embed in the insulation material during the manufacturing process of the apparatus. This type of discharge, on the other hand, is less concentrated and hazardous than internal discharge. Corona discharge, on the other hand, occurs in air insulation and is usually harmless. Treeing is a form of discharge that is triggered by a sequence of internal discharges. Since the advent of polymer nanocomposites, various studies have been conducted on PD properties of such materials. Kozako et al. [106], for example, reported that only two wt% nanofiller is needed to increase the PD resistance of polyamide/layered silicate nanocomposites. Another investigated by Tanaka et al. [7] looked at PD resistance in epoxy/layered silicate nanocomposites. The authors concluded that by incorporating a limited amount of nanofillers into epoxy resins, PD tolerance of polymer nanocomposites could be greatly enhanced. Many experimental findings on the enhancement of partial discharge property of polymer nanocomposites have been reported. Henk et al. [107] studied the effects of nanoparticle amorphous silica dispersion in epoxy and polyethylene. Nano-silica was found to have a significant impact on thermoset partial discharge durability (epoxy and cross-link polyethylene). However, no effect was found with thermoplastics (low density polyethylene and medium density polyethylene).

#### *4.2. High-Current Arc Resistance to Ignition (HAI)*

The High-Ampere Arc Ignition (HAI) testing is a method that studies and assesses the electrical insulation flammability. This method is described in detail at the standard for safety UL 764A [108–110]. The HAI test subjects three specimens of the studied electrical insulation to electric arcs, recording the average number required to produce ignition in it, with a maximum of 200 (over this number is considered no ignition of the material). The test generates the electric arcs by using two round electrodes in contact with the studied electrical insulation; these electrodes have 3.2 mm of diameter; the difficulty is described in Figure 9. One of the electrodes is fixed and made of copper, where the other is stainless steel (303 alloys) movable rod; through the separation of these rods, an electric arc is generated. The electrodes are placed in a 45◦ plane of the studied electrical insulation specimen, and the fixed rod is sharpened to a 30◦ chisel point and the movable rod to a 60◦ conical point. The stainless steel rod movement is done using an air piston controlled by an electrical relay [108,109].

**Figure 9.** Electrode positioning and operating mechanism in HAI. Adapted from [109]**.**

The testing method proposed by the standard UL 764A starts by applying a voltage of 240 V AC at a frequency of 60 Hz. An air core impedance is connected in series to the electrodes for yielding a short circuit current of 32.5 A, with a power factor of 0.5. The rate of electric arcs generated is 40 arcs/min, and the separation of the electrodes is an average of 254 ± 2.54 mm/s. There is no need to synchronize the electrodes' separation with a particular value of the sinusoidal current variation [108,109].

Several issues have been established for the replicability of this test, such as the identification of the ignition in the studied electrical insulation and the placement of the electrodes, that by not being described in detail, could lead to different kinds of arcs (random intensity or white flashes) [108,109]. To fix the latter issue, it was identified that the electrodes should return to their original position between cycles [108]. The identification problem was, in the beginning, attenuated by the use of a dark glass shield that allowed for a better distinction of the candle-like flame, which indicated the origin of the ignition in the studied specimen. However, this was improved by the use of a liquid crystal light attenuator [109]. The liquid crystal light attenuator allows for a better distinction of a low or high-energy arc, improving the test by providing the operator the crucial information regarding the test's continuity. The continuous series of soft energy arcs indicate that the electrodes should be moved to a different testing area of the studied specimen. Additionally, to enhance the distinction of a high from a low energy arc, the use of a digital arc energy meter is advised. This device records the magnitude of current and voltage for discrete intervals to calculate the electric arc's energy magnitude. These recommendations allow for a more descriptive signature between different materials taking into account its resistance to ignition [109].

The initial interest in making an HAI test in DC conditions originated from the automotive industry, which changed the traditional 12 V DC battery supply to a 36 V DC battery supply with a 42 V DC charging unit. This originated an interest in determining the ignition resistance of the automotive polymeric materials [111]. Initially, a primary circuit for studying the glass-fiber-reinforced plastics (GRP) for a DC supply was proposed, as shown in Figure 10. This test was based on the HAI test described by the standard UL 764A. This circuit uses a 200 V AC source; the capacitors generate a DC discharge between electrodes A and B, the release takes 0.5 to 12.8 ms depending on the capacitance values. The rate of the discharges is 20–40 arcs/min until the arc ignites the specimen. The use of a high-speed camera (4800 frames/s) allowed recording the length of the whole process, registering the arc discharge, arc ignition, and the combustion duration [112]. Similar studies have been realized in common polymers for the automotive industry [113–115].

**Figure 10.** Basic circuit diagram of high current DC arc ignition testing apparatus formed by power switch (S1), regulator (Re), AC voltmeter (V1), rectifier (G), DC voltmeter (V2), DC ammeter (Am), charging resistance (r), capacitor (C), discharge resistance (R), fixed electrode (A), mobile electrode (B) and Inductance for current stabilization (L). Adapted from [112].

The interest in the DC testing method has increased in recent years due to the rise of renewable technologies like solar photovoltaics and wind power and DC microgrids' implementation [116]. This lead to proposing new DC-HAI testing, based on the original HAI testing described by the standard UL 764A, with the inclusion of some improvements, such as an automated control via a LabView-based program. The proposed testing setup is shown in Figure 11, where the movable electrode has three different positions: start, stable and safe. For this test, there are nine configuration parameters [116]:


**Figure 11.** Resistance to high-ampere arc ignition (DC-HAI). Adapted from [117].

The safe position is the most considerable distance between the electrodes, allowing for a better judgment regarding the specimen's ignition. On the other hand, the stable position's purpose is to establish and arc stability [116]. The test realized by this proposed method resulted in a standard deviation ratio of 5% for DC arc that took 200 ms and 8% for an arc that has currents of 10 A; this proved to be consistent and repeatable [116].
