Contents lists available at ScienceDirect Electrical Power and Energy Systems journal homepage: www.elsevier.com/locate/ijepes A new method for estimation of time parameters of standard and non- standard switching impulse voltages Božidar Filipović-Grčić a, ⁎ , Dalibor Filipović-Grčić b a University of Zagreb, Faculty of Electrical Engineering and Computing, Unska 3, 10000 Zagreb, Croatia b Končar – Electrical Engineering Institute, Fallerovo šetalište 22, 10000 Zagreb, Croatia ARTICLE INFO Keywords: Switching impulse voltages Time and amplitude parameters IEC 60060-1 High-voltage test techniques ABSTRACT In this paper, a new method for estimation of time parameters of standard and non-standard switching impulse voltages is presented. Method is based on estimation of time difference between true and virtual origin of the switching impulse waveform. An analytical expression was derived for calculation of time to peak, which is more accurate than the expression given in IEC 60060-1. The presented method was verified with mathematically generated double-exponential waveforms and waveforms given in IEC 61083-2 with time to peak values within the range 20–300 μs and time to half values within the range 1000–4000 μs. An experimental verification of the proposed method was successfully demonstrated by comparison with an approved impulse voltage measuring system. 1. Introduction Switching overvoltages (SOVs) in high voltage networks, which are caused by circuit breaker operations, stress the insulation of the high voltage equipment [1]. Therefore, most of high voltage equipment designed for operating voltages above 245 kV should be tested under laboratory simulated switching-impulse voltages [2]. Test requirements along with definitions of standard switching-im- pulse voltage parameters are given in [3]. The time parameters of standard switching-impulse voltage are shown in Fig. 1. True origin O is an instant where the recorded curve begins a monotonic increase (or decrease for waveforms of negative polarity). Virtual origin O 1 is an intersection of the time axis with a straight line drawn through the reference points A and B in the front. Time to peak T p is a time interval from the O to the time of maximum value of a switching-impulse vol- tage, while time to half value T 2 is a time interval between the O and the instant when the voltage has first decreased to half the maximum value. Standard switching-impulse voltage has a T p of 250 μs and a T 2 of 2500 μs. Acceptable tolerances between specified values and those re- corded in laboratory conditions are: ± 20% for T p , ± 60% for T 2 and ± 3% for value of test voltage. T p for standard switching-impulse voltages is defined as follows [3]: = T KT · , p AB (1) where K is a dimensionless value given by: = − + − − K T T 2.42 3.08·10 1.51·10 , AB 3 4 2 (2) and T AB is given by: = − T t t . AB 90 30 (3) There are several issues when estimating time parameters of switching impulse voltages. In practice, it is not easy to determine an instant O and time at which maximum of recorded waveform occurs. The reason are oscillations caused by operation of impulse generator around O and noise present in the recorded signal. In the peak area of the impulse there is a problem related to analogue-to-digital conversion which is used for sampling the recorded analogue signal and quantizing its amplitude. Due to this process, the signal values are only available at discrete time intervals. Therefore, the signal amplitude cannot be clearly determined even with high resolution recorder because there can be several discrete points in the peak area often with the same value. Noise could be reduced by averaging the recorded signal, but this affects the parameters of the recorded waveform. Literature survey showed that only a few papers have been pub- lished regarding the issues mentioned above. In [3] it is stated that for non-standard impulses, T p can be determined by various methods of digital curve fitting dependant on the actual shape. The problem is that there is no guidance on how to determine T p for non-standard impulses. The requirements for software used for evaluation of impulse parameters from recorded impulse voltages are given in [4]. It provides test waveforms and reference values for the software required to meet the measuring uncertainties and procedures specified in [3,5–7]. Some http://dx.doi.org/10.1016/j.ijepes.2017.10.001 Received 11 August 2017; Received in revised form 15 September 2017; Accepted 1 October 2017 ⁎ Corresponding author. E-mail addresses: bozidar.filipovic-grcic@fer.hr (B. Filipović-Grčić), dfilipovic@koncar-institut.hr (D. Filipović-Grčić). Electrical Power and Energy Systems 96 (2018) 126–131 0142-0615/ © 2017 Elsevier Ltd. All rights reserved. MARK