**1. Introduction**

Gas-insulated systems (GIS) have been the state-of-the-art in high voltage alternating current (HVAC) transmission grids since the 1960s [1]. These systems have major advantages compared to air-insulated systems (AIS) like their space-saving design, their independence from environmental conditions, and their higher reliability. These advantages are beneficial for high voltage direct current (HVDC) transmission systems as well. Especially due to the growing importance of renewable energy sources and their integration into the existing power grid, longer transmission lines have to be built. This is economically feasible only by using HVDC technologies. The advantageous space-saving design of GIS can be used, for example, to reduce the size of offshore converter-platforms, to ensure a reliable power transmission in densely populated areas, and to allow building high voltage infrastructure close to protected landscapes due to their low visual impact. Therefore, gas-insulated HVDC systems are a space-saving solution for HVDC substations.

Due to the transition of the electric field, the directed movement of charge carriers, and the accumulation of charge carriers on gas-solid interfaces, the development of gas-insulated HVDC systems is challenging [2,3]. Reliable operation can be ensured, using partial discharge (PD) measurements during type tests, factory acceptance tests, on-site tests, and operation (monitoring). For AC applications, electrical PD measurements using the phase-resolved partial discharge pattern (PRPD) have been well known since the 1960s for various types of defects [4]. Due to the missing phase relation, the procedures and analysis tools used under AC voltage stress cannot be applied directly for DC equipment. Additionally, the directed movement of charge carriers, space-charges in the gas, and accumulated charges at the gas-solid interfaces lead to a significantly different partial discharge behavior under DC voltage stress [5–12]. One major challenge, reported in the literature, is the occurrence of pulseless discharges, which cannot be detected using the conventional methods according to IEC 60270 [13] and ultra-high frequency (UHF) measurements [5,10,11]. Summarizing these challenges, the PD experts stated in the literature that a major problem during PD measurements and analysis under DC voltage stress is the lack of experience [14]. In addition, the physical PD behavior under DC voltage stress is not as well understood as under AC voltage stress, since space charges can accumulate over a long time period. However, a reliable PD diagnosis requires the knowledge of the discharge behavior. Hence, the aim of this investigation is to bridge the gap of knowledge for one typical defect in gas-insulated systems: a protrusion. Therefore, the partial discharge current of a fixed needle in a weakly inhomogeneous electric field is measured. The measurements reveal four different discharge types that may occur depending on the electric field strength, gas pressure, and voltage polarity.

The insulating gas sulfur-hexafluoride (SF6) has been used as a dielectric in gas-insulated systems since the early 1960s. Due to the high global warming potential (GWP) of this gas and the related political will to reduce the emission of fluorinated gases, alternative gases with a low GWP come to the fore of manufacturers and customers [15–18]. These gases can be pure natural gases or mixtures of natural gases, like synthetic air. Furthermore, these alternatives can be gas-mixtures with highly electron affine components in order to improve their dielectric behavior. In this contribution, the discharge behavior of SF6 is compared with the behavior of pressurized synthetic air as one example for an alternative gas.
