Using a field probe to study the mechanism of partial discharges in very small air gaps under direct voltage

Due to the growing use of high-voltage direct current (HVDC) for power transmission, it is of interest whether the diagnostic tools introduced in the past for insulation condition assessment of high-voltage alternating current (HVAC) apparatus are also applicable for quality assurance tests of HVDC equipment. This refers in particular to the measurement of partial discharges (PD) in compliance with the international standard IEC 60270 [1], [2]. As PD events in gaseous cavities embedded in the bulk dielectric of HV equipment are the most harmful ones, the mechanism of such internal discharges should be well known to judge the PD severity as well as to specify appropriate procedures and quantities for PD tests under direct voltage. According to Devins cavity discharges can be classified as "Townsendlike discharges" and "streamer-like discharges" [3]. Even if this classification seems to be reasonable for cavity discharges under both alternating and direct voltage [4]-[6], this has been rejected by Bartnikas. So, he argued in [7] that "the term 'streamer' by itself, when applied to PD in relatively short gaps or small cavity diameters, introduces unnecessarily a misleading term in the PD lexicon. PD, when occurring in short gaps, may assume different forms: rapid and slow rise time spark-type pulses." This is also consistent with the worldwide established PD model according to Kreuger [8], which is based on the hypothesis that the virtual cavity capacitance is discharged via a conductive spark channel. Despite simplicity, this concept was rejected by Pedersen and his coworkers [9], [10]. As an alternative, they promoted a more sophisticated field-theoretical approach based on a dipole model. As their concept is not easily understandable, this was presented in a more simplified manner in this Magazine [11], [12] as well as in [13].