Effect of humidity on partial discharge in a metal-dielectric air gap on machine insulation at trapezoidal testing voltages

Journal of Electrostatics 83 (2016) 88e96

Contents lists available at ScienceDirect

Journal of Electrostatics journal homepage: www.elsevier.com/locate/elstat

Review

Effect of humidity on partial discharge in a metal-dielectric air gap on machine insulation at trapezoidal testing voltages X. Wang*, N. Taylor, H. Edin KTH-Royal Institute of Technology, School of Electrical Engineering, Teknikringen 33, SE-100 44, Stockholm, Sweden

a r t i c l e i n f o

a b s t r a c t

Article history: Received 24 April 2016 Received in revised form 4 August 2016 Accepted 16 August 2016 Available online 2 September 2016

The atmospheric relative humidity (RH) has a great impact on the partial discharge (PD) process which can damage the insulation in operating machines. This work investigates how the relative humidity would affect the PD activities in a metal-dielectric air gap on machine insulation, which consists of mica, epoxy resin and glass-fiber, with the application of periodic alternating trapezoidal voltage waveforms. The results show that the PD characteristics, such as discharge amplitude, the average number of discharge pulses, can be varied greatly with the increasing humidity. This is mainly due to the increased surface conductivity in humid air. © 2016 Elsevier B.V. All rights reserved.

Keywords: Partial discharge Machine insulation Relative humidity Surface conductivity Surface charge decay

Contents 1. 2.

3.

4.

5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 2.1. PD measurement system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 2.2. Test voltages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 2.3. Test samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 2.4. Effect of relative humidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 3.1. PD at the relative humidity of 8% . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 3.2. PD at the relative humidity of 29% . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 3.3. PD at the relative humidity of 77% . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 4.1. PD behavior for the dry case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 4.2. Effect of relative humidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

1. Introduction

* Corresponding author. E-mail address: [email protected] (X. Wang). http://dx.doi.org/10.1016/j.elstat.2016.08.003 0304-3886/© 2016 Elsevier B.V. All rights reserved.

Stator insulation is a critical part of high-voltage rotating machines with respect to the efficiency and reliability of operation, manufacturing costs and maintenance. The stator insulation system is exposed to electrical, thermal, mechanical and ambient stresses

X. Wang et al. / Journal of Electrostatics 83 (2016) 88e96

simultaneously during its operation, resulting in the gradual deterioration of insulation properties which can reduce the machine's lifetime. Many failures of power generators failure are caused by insulation damage; the two main reasons leading to these damages are aging and partial discharge [1]. In general, several common PD sources exist in stator insulation, such as internal discharge, end-winding discharge, surface tracking, delamination and slot discharge [2,3]. Even though modern machines based on epoxy-mica insulation system have been designed to be able to withstand an appreciable level of discharges, for instance, internal discharge in small volumes, some other PD activities are even more detrimental to the insulation system, such as slot discharge. Slot discharge occurs in the air gap between the surface of the stator bar and the laminated magnetic core. It has been considered as the most severe damage to the stator winding groundwall insulation [4]. There are two main mechanisms that give rise to the slot discharge activity [5,6]. One is the mechanical slot discharge, which is developed from loose bars in the slot due to, for instance, the shrinking of insulation; this allows vibration of the bar in the slot leading to abrasion of the conductive coating of the bar. The other one is the electrical slot discharge, which is caused by poorly manufactured semiconductor coating with inappropriate surface conductivity or uniformity of the coating. If the surface conductivity is too low, a significant voltage could build up between the surface of the bar and the core in parts where the bar surface is not in direct contact with the core. The PD activity that follows may lead to the breakdown of the insulation between the bar and the core. The calculation in Ref. [7] gave a maximum acceptable surface resistivity of 25 kU per square for the surface coating. Besides, poor electrical connection of the conductive coating to ground also can initiate the slot discharge. It is well known that the typical phase resolved partial discharge (PRPD) pattern of slot discharge is characterized by a strongly asymmetric pattern, which is triangular with a sharp slope at the onset of the positive discharge appearing during the negative polarity of the applied voltage [2]. However, this pattern varies with other conditions, such as temperature and humidity in the air. One typical and systematic investigation of slot partial discharge has been presented in Refs. [8e10], including the influence of gap size, temperature and humidity, as well as the surface degradation caused by different stresses, such as electrical, thermal and mechanical stress, on the slot discharge activity. Moreover, it is apparent that the air humidity has a great impact on the PD activity in operating machines; one of the clear evidences from Ref. [11] is that by monthly on-line monitoring on the generator up to 9 years, the PD activities on generators in winter were higher than that in summer with a changing rate higher than 100 because of the seasonal changes in humidity. There exist a large number of studies about the effect of relative humidity on the PD activity [12e18], most of them focusing on the variation of PD inception and extinction voltage, PD intensity, breakdown strength, surface conductivity of the insulating material, space charge accumulation on the material and the statistical time lag of discharge, and so on. From those studies, it has been recognized that humidity has a negative correlation with PD activities: the discharge activity decreases as the humidity increases. For instance, it was observed in Ref. [15] that the external PD activities such as slot discharge are strongly reduced and even disappeared if the ambient air humidity goes above 50%. A similar behavior was reported in Ref. [16] for surface discharges on an epoxy bar disappeared at the eighth day of exposure to humidity of 80%. It has been understood this is due to the electronegative nature of water molecules, which can capture electrons and reduce the availability of free electrons to generate electron avalanches. However, lower PD activities at high-humidity

89

do not directly mean that humidity reduces damage to the insulation surface; on the contrary, it was reported in Ref. [18] that degradation of epoxy resin exposed to PDs was more severe in moist air than in dry air. In this paper, the effect of relative humidity on partial discharges which take place in the air gap between a spherical metal electrode and the surface of epoxy-mica machine insulation is investigated. The PD analysis is performed with trapezoidal voltage waveforms as stimuli. The reason for this novel method is to investigate the character of the PD behavior with voltage stimuli that have two features, first a constant changing voltage dU/dt during rising as well as falling voltage period and a second feature of including a short period of constant voltage between rising and falling voltage periods, rather than having continuously changing voltage that is the case with alternating sinusoidal voltages. The derivative of the applied voltage dU/dt were varied from the maximum possible (approximately square-wave) to the minimum possible (trianglewave) for a given amplitude and period of the voltage. The influence of those voltage waveforms on the PD activity under different humidity levels was investigated, focusing on PD characteristics, such as PD repetition, average number of PD pulses per cycle, PD amplitude and delay time of PD appearance. 2. Experimental 2.1. PD measurement system The time-resolved PD measuring system consists of an Agilent 33120A function generator, a TREK 20/20 high-voltage amplifier, a detection resistance, an oscilloscope that was Yokogawa DL750 Scope Corder and a computer, as shown in Fig. 1. The test voltage was generated by a function generator and then amplified by a high-voltage amplifier (with amplification factor 2000). The high voltage was applied to a steel spherical electrode with a diameter of 22 mm, in order to concentrate the discharges at one spot on the insulating surface. A short length of commercial stator bar was placed below the spherical electrode; the minimum electrode-surface distance was about 1 mm. The machine insulation consisted of mica, epoxy resin and glass-fiber and its approximate thickness was 1 mm. The detection resistance R was 50 U, and the coupling capacitance Ck was 500 pF. The discharge pulses were acquired with the oscilloscope, with 12-bit A/D resolution, a sampling rate of 10 MS/s, and a deep memory of 250 MS, which made long-time pulse sequence analysis possible. The entire test cell was placed into a well-sealed plastic chamber where the relative humidity can be controlled by a saturated salt solution [19], and silica gel used as desiccant for the dry study. The relative humidity and the temperature inside the chamber were monitored by a Testo 625 thermo hygrometer. 2.2. Test voltages The testing voltage waveforms used were of a trapezoidal shape with different time derivative of the applied voltage (dU/dt), as shown in Fig. 2 (b). The duration of the linearly increasing period from zero till peak voltage Upeak is T1; the peak voltage keeps constant during the period of T2; then linearly falls back down to the zero with the duration of T3, the negative half-cycle is a mirror of the positive half-cycle. So the period of the applied voltage is T, where T ¼ 2 (T1 þ T2 þ T3). In the entire work, the up and down dU/ dt are equal, i.e. T1 ¼ T3. Moreover, there were two limit cases of this waveform. Fig. 2 (a) shows the triangular voltage at T2 ¼ 0 ms and the approximately square voltage could also be achieved in the case of T1 ¼ T3 z 0 ms, shown in Fig. 2 (c). In the entire work, the testing voltages shown in all the plots were recorded from the function

90

X. Wang et al. / Journal of Electrostatics 83 (2016) 88e96

Temp. and RH sensor

Salt solution

Stator bar

R H.V. amplifier

Function generator

GPIB

Ck

Plastic chamber

Scope corder

GPIB Fig. 1. Schematic diagram of time-resolved PD measuring system.

Fig. 2. Three types of testing voltage waveforms: (a) triangular; (b) trapezoidal; (c) approximately square.

generator. In reality the output of the square wave is a little bit later than the input; that's appropriate for a square shape. The square waveform here has a rise time of several V/ms in these measurements. In this work, the influence of time derivative of the applied voltage (dU/dt) was studied, which was implemented by varying the duration of T1 at a specific applied voltage and a frequency of 50 Hz (T ¼ 20 ms). 2.3. Test samples The test sample was a short length of commercial stator bar. It has two stacks of 14 separated strands. The strands were all connected by silver-based conductive paint at one end. A hole was drilled between the stacks to provide a connection point to the strands. The insulation between the strands and the outer surface had a thickness of 1 mm after removal of the slot semiconductor layer and several layers of main insulation, as shown in Fig. 3. The main insulation of the stator bar consists of small pieces of micapaper tapes with a backing material of glass-fiber and epoxy resin as a binder. The insulation between the strands is not considered to be stressed during the measurement in this work [5].

Fig. 3. A short length of stator bar used in the study.

would affect the dielectric properties of the insulation only if it is subjected to water for very long time [17]. However, the moisture can strongly affect the surface conductivity of insulation by forming a conductive layer over the insulation surface. A saturated salt solution was used to control the relative humidity of the ambient air inside the chamber for the wet studies, and a silica gel desiccant was used for the dry study. The test sample was kept in the chamber at room temperature (22e23  C) for at least 15 h in the presence of the saturated salt solution in order to reach the steady state of the whole system before the measurement. After 5 or 6 h' continuous measurement, the chamber was left open at least 2 h to release the ozone from the chamber and let it recover back to the room humidity before introducing another saturated salt. PD measurements based on different voltage waveforms shown in Fig. 2 have been carried out at three different relative humidity levels of 8% (dry condition given by silica gel), 29% (room condition given by magnesium chloride) and 77% (humid condition given by sodium chloride), respectively. The timesequential PD pulses were captured after each voltage was applied to the test sample for one up to three minutes. 3. Results

2.4. Effect of relative humidity

3.1. PD at the relative humidity of 8%

For modern insulation materials of large rotating machines, it seems impossible to have significant moisture absorption from the ambient air into the insulation bulk, and water contamination

In this study, the test sample was exposed to PD activities in dry air of RH ¼ 8%. The applied voltage was Uapp ¼ 7 kV (z1.5e1.6 Uinc) with the varying values of dU/dt, that is, at the same voltage level

X. Wang et al. / Journal of Electrostatics 83 (2016) 88e96

but with different values of T1: 0 ms, 1 ms, 2 ms, 3 ms, 4 ms and 5 ms. Generally, the PD inception voltage Uinc was a little different for each waveform; it decreases slightly with the increasing dU/dt [20]. The time-sequential PD pulses were captured during consecutive 24 cycles of the applied voltage, with one example of 10 consecutive cycles at T1 ¼ 1 ms shown in Fig. 4. To have a more aggregated view of discharge behavior, a phase-resolved PD pattern was made by converting all the discharge pulses during the recorded cycles into one single reference cycle of the applied voltage, as shown in Fig. 5. The relative voltage position of PD occurrence has been marked with star on the waveforms in the following figures. PD activities in dry condition show some common features: Firstly, PD number and amplitude: there are only one or two discharge pulses per half cycle, but with quite big amplitude, which could reach to 10 V in the case of T1 ¼ 0 ms, as shown in Fig. 5 (a); however, the PD amplitude becomes smaller when the voltage increases more slowly in the period of T1, for instance, Fig. 5 (e) shows the PD magnitude can get to around 6 V. Secondly, phase of PD occurrence: it can be clearly observed from Fig. 5 that most of the PDs occur at the voltage rising period of T1, but only very few PD events appear during the constant-voltage period of T2, and No PD pulses were observed to appear after dU/dt was changed from zero to negative or to positive. It must be mentioned that the appearance of the ‘pulses’ with opposite polarity with their applied voltage, such as the negative ‘pulses’ in the positive half-cycle of the applied voltage in Fig. 5 (a) and (b), is due to the oscillation of the big positive discharge pulses. Thirdly, PD time shift tshift: there is a significant time shift for PD occurrence from the ‘edge’ of the ‘square voltage’ in Fig. 5 (a); it has been presented in Ref. [21] that the detected PD events appear long time after the voltage switching on at semi-square voltage waveforms, in the order of milliseconds. With the increasing of T1, the PD pattern shows an increasing delay time after the zero-crossing voltage, such as, Fig. 5 (f) shows a big time shift tshift ¼ 1.22 ms at the triangle voltage waveform. It is clear that the inception voltage Uinc plays an important role in this phenomenon. The time to reach Uinc is different for different voltage waveforms. When T1 is increased, the time needed for reaching Uinc increases. 3.2. PD at the relative humidity of 29% In room condition of RH ¼ 29%, the PD measurement has been performed at the same applied voltage waveforms as the last study but with a lower applied voltage of Uapp ¼ 6 kV (z1.5e1.6 Uinc). This

91

is because the PD inception voltage becomes lower with the increasing relative humidity level. Fig. 6 shows the phase-resolved PD patterns and the time-sequential PD pulses during 10 consecutive recorded cycles of the applied voltage at T1 ¼ 1 ms are shown in Fig. 7. Compared with the PD characteristics in dry air, PD activities at a relative humidity of 29% show that discharge amplitude becomes smaller; all the PD pulses are less than 2 V, however, the number of PD events obviously increases, about 15e20 PD pulses per cycle. Regarding to the phase of PD appearance, the same behavior of those PD patterns from Figs. 5 and 6 is that the PD activities during voltage rising period of T1 are still dominant and no PD events turn up in voltage falling period of T3. However, the major difference is that except the PD pattern in the triangular waveform in Fig. 6 (f), all the other patterns have PD occurrence in the constant-voltage period of T2, even at approximately squarewave voltage of T1 z 0 in Fig. 6 (a). The similar PD pattern under square voltage waveform has been presented in Ref. [22]. 3.3. PD at the relative humidity of 77% In order to understand how the relative humidity would affect the PD process and to obtain the significant different features from PD patterns compared with the last two studies, a high relative humidity was chosen, that is, RH ¼ 77%. The applied voltage was even lower Uapp ¼ 5 kV (z1.5e1.6 Uinc) due to its lower PD inception voltage at the high humidity level. Fig. 8 and Fig. 9 show one example of time-sequential PD pulses at T1 ¼ 1 ms and the phase-resolved PD patterns at the applied voltage waveforms, respectively. It can be observed that PD activities in humid condition contain some distinctive features: all the PD amplitude dramatically decreases, dropping to less than 500 mV, on the contrary, the number of PD pulses increases significantly, roughly 100e200 discharge pulses per cycle. There are a large number of PD events dispersing not only during the voltage rising period of T1, but also covering the whole constant-voltage period of T2 for both polarities, even extending a little to voltage falling period of T3, for instance, PD patterns at T1 ¼ 4 ms and T1 ¼ 5 ms shown in Fig. 9 (e) and (f). It seems that all the PD pulses distribute quite uniformly during their appearance phase. Generally, the time shift tshift for the PD appearance in humid condition seems longer compared with that in the lower relative humidity levels, for instance, tshift ¼ 1.22 ms in Fig. 5(f), tshift ¼ 1.37 ms in Fig. 7(f) and tshift ¼ 2.70 ms in Fig. 9(f). 4. Discussion 4.1. PD behavior for the dry case

PD Voltage [V]

10 5 0 −5 −10 0

0.05

0.1 Time [s]

0.15

0.2

Fig. 4. Time-sequential PD pulses during 10 consecutive recorded cycles of the applied voltage at T1 ¼ 1 ms in dry condition of RH ¼ 8%.

It should be mentioned that the electric field from the applied voltage and from space charges will drive the initiation and evolution of the discharges in the air gap. Therefore, the benefit of the non-sinusoidal testing voltage waveforms is that the local conditions in the gap, such as electric field distribution, surface charge decay, vary with the applied non-sinusoidal voltage waveform. As a result, the PD process will also vary with the applied voltage waveforms. The application of trapezoidal voltage waveforms for PD study can show more particular PD behavior and provide more useful information about the discharge physics than traditional sinusoidal voltage waveform. The PD behavior in dry condition acts as a reference of studying the relative humidity effect on the PD activities. Fig. 10 shows PD characteristics at different values of T1 in dry condition. Generally, with the increasing dU/dt in the period of T1, the average number of PDs per cycle Navg and the total PD charges per cycle Qtot is increased. After one discharge event, the electric field in the air gap

X. Wang et al. / Journal of Electrostatics 83 (2016) 88e96

10

5

5

0 −5 −10

(a) 0.005

0.01 0.015 Time [s]

0 −5 −10 0

0.02

(b) 0.005

0.01 0.015 Time [s]

5

0

−5 0

0.02

10

10

5

5

5

0 −5 −10 0

(d) 0.005

0.01 0.015 Time [s]

PD Voltage [V]

10 PD Voltage [V]

PD Voltage [V]

0

10 PD Voltage [V]

10

PD Voltage [V]

PD Voltage [V]

92

0 −5 −10 0

0.02

(e) 0.005

0.01 0.015 Time [s]

0.02

(c) 0.005

0.01 Time [s]

0.015

0.005

0.01 0.015 Time [s]

0.02

0 −5 −10 0

(f) 0.02

4

2

2

2

0 −2

(a) 0.005

0.01 Time [s]

0.015

0 −2 −4 0

0.02

(b) 0.005

0.01 Time [s]

0.015

0 −2 −4 0

0.02

4

4

4

2

2

2

0 −2 −4 0

(d) 0.005

0.01 Time [s]

0.015

0.02

PD Voltage [V]

PD Voltage [V]

−4 0

PD Voltage [V]

4 PD Voltage [V]

4

PD Voltage [V]

PD Voltage [V]

Fig. 5. Phase-resolved PD patterns of the time-sequential PD pulses during 24 consecutive recorded cycles of the applied voltage at different values of T1: (a) 0 ms; (b) 1 ms; (c) 2 ms; (d) 3 ms; (e) 4 ms; (f) 5 m in dry condition of RH ¼ 8%.

0 −2 −4 0

(e) 0.005

0.01 Time [s]

0.015

0.02

(c) 0.005

0.01 Time [s]

0.015

0.02

0.005

0.01 Time [s]

0.015

0.02

0 −2 −4 0

(f)

Fig. 6. Phase-resolved PD patterns of the time-sequential PD pulses during 24 consecutive recorded cycles of the applied voltage at different values of T1: (a) 0 ms; (b) 1 ms; (c) 2 ms; (d) 3 ms; (f) 4 ms; (d) 5 ms in room condition of RH ¼ 29%.

drops to residual electric field quickly. If the applied voltage increases more quickly to the peak value, the field across the air gap could quickly reach to the inception field again, giving more PD events. On the other hand, the increase of dU/dt gives rise to enhancement of the charge accumulation affect, leading to a strong reversed superposed field and longer stay of the space charges on the insulating surface, thus the total discharge charge can be increased. With respect to the phase of discharge, almost all the PDs concentrate on the voltage rising period of T1. This is because many PD activities have happened due to enough free electrons in dry air and faster increasing of the applied voltage before reaching its peak value, and the electric field across the air gap has reached its equilibrium states. In addition, the slower surface charge decay due to a lower surface conductivity in dry air will lead to a higher surface potential on the insulating material; it can decelerate or extinguish the discharge [23]. Therefore, in dry air the discharge shows the similar behavior to the PD activity at DC voltage

waveform, that is, very few discharge events.

4.2. Effect of relative humidity Humidity has a complex influence on the discharge mechanisms, not only by reducing the effective free electrons in the air to initiate and sustain the PD activity, but also by modifying the surface condition of the insulating material. The electronegative nature of water molecules can increase the effective attachment coefficient, thus reduce the effective ionization efficiency in the low field region, which leads to the reduction of locally available free electrons [24]. Once water molecules capture free electrons, they form more stable ‘water ions’, with a very low probability to release their electrons during molecule collisions [10,25]. The consequence of water vapor presence is also mentioned in another concept as cluster formation in the air [25e27], which leads to a decrease of charge mobility due to their increased size and mass, and an

X. Wang et al. / Journal of Electrostatics 83 (2016) 88e96

PD Voltage [V]

4 2 0 −2 −4 0

0.05

0.1 Time [s]

0.15

0.2

Fig. 7. Time-sequential PD pulses during 10 consecutive recorded cycles of the applied voltage at T1 ¼ 1 ms in room condition of RH ¼ 29%.

3

PD Voltage [V]

2 1 0 −1 −2 −3 0

0.05

0.1 Time [s]

0.15

0.2

Fig. 8. Time-sequential PD pulses during 10 consecutive recorded cycles of the applied voltage at T1 ¼ 1 ms in humid condition of RH ¼ 77%.

increase in their lifetime because the clustering molecules can be considered to form a protective shield around the ions, thus increase their stability. Moreover, water molecules can also limit photoionization by absorbing the irradiated photons [25]. All of these consequences will weaken the ionization during discharge process and reduce the PD activities in the air. As mentioned above, the PD inception voltage decreases with the increasing relative humidity, as shown in Fig. 11. It may be easy to understand that the lack of free electrons due to the presence of water vapor will increase the breakdown strength of air. However, the condensation of water vapor on the insulating surface may change the electric field distribution on it and the electric field will be enhanced around the local condensation. It is possible that the decrease of PD inception voltage is also due to the effect of insulation surface degradation under humid condition. This would lead to the reduction of the inception voltage, as presented in Refs. [12,13].

93

The occurrence of a discharge depends not only on the applied voltage, but also on the charge distribution on the insulating material. In a short air gap, surface condition of the insulating material plays an important role in the propagation of a discharge, because the charges deposited on the insulating surface from the previous PDs have a strong relation with the surface condition, mainly referring to surface conductivity. The surface conductivity has a great impact on the superposition of the electric field during the PD process. It has been recognized that the surface conductivity increases significantly at higher relative humidity, thus surface charge decay is dominated by electric conduction along the insulation surface [28]. With the increasing conductivity of the insulating surface, the mobility of charges deposited on the insulating surface will be enhanced, so the charges would spread over larger area on the insulating surface. The surface area where charges deposited on the insulating material plays an important role in variation of the discharge magnitude [29]. Thus the surface potential on many spots will be built up due to the deposited charges. Therefore, the lower overvoltage on larger area on the insulating surface will generate lots of small-amplitude discharge pulses. However, on the insulating material with a lower conductivity, the charges could remain on very limited area of the insulating surface, leading to a large overvoltage on other areas which are not affected by the surface charges, and thus causing several large-amplitude discharge pulses [30]. Fig. 12 shows one typical period of time-sequential PD pulses under the same applied stress at T1 ¼ 1 ms with three different humidity conditions. In summary, a few big discharge pulses in dry air turn into a large number of small pulses in humid air mainly due to the increasing surface conductivity in higher humidity condition. It is possible that the rapid degradation of insulation surface could also contribute to the change of PD activity behavior in humid condition. It could be observed from Fig. 12 (c) that PD activities show a higher repetition rate and larger amplitude at the beginning of voltage applied if the voltage increases quickly in the period of T1, and then the repetition rate is decreased, and the amplitude also becomes smaller and uniform during the constant-voltage period. This decreasing of PD repetition rate is the result of surface potential on the insulating material caused by the deposited charges, leading to a reduction of the discharge activities [23]. Moreover, the appearance of PD in the constant-voltage period at higher humidity levels reflect that the electric field in the gap is not constant even during the short period of T2, but is affected by surface charge decay or lateral spread. The charges deposited by the previous PD events will be faster decaying due to a higher surface conductivity, causing a lower surface potential built up on a larger area of the insulating surface. So the voltage difference across the air gap would be higher than the inception voltage during the most of the applied voltage period. Also, there are still available free electrons left at that period which are not completely consumed during the period of T1. Therefore, the PD occurrence during the constant-voltage and the voltage falling period depends on the availability of free electrons and the spread of the space charges on the insulating surface. The variation of PD characteristics with the changing of relative humidity can be seen in Fig. 13, it can be observed that the decreasing behavior of Qtot and Navg with the increasing of T1 is enhanced in the humid condition, and the same to the increasing behavior of time shift of PD occurrence tshift. Fig. 13 (c) shows that with the increasing period of T1, the time shift is clearly increased at three humidity conditions; it is clear because the applied voltage increases slowly, more time will be needed to have the PD appeared even though the free electron is available. When the relative humidity is raised, the PD events occur significantly later, mainly due to the lack of free electrons caused by the existence of water molecules in the air. In general, the reduction of free electrons and the

94

X. Wang et al. / Journal of Electrostatics 83 (2016) 88e96

Fig. 9. Phase-resolved PD patterns of the time-sequential PD pulses during 24 consecutive recorded cycles of the applied voltage at different values of T1: (a) 0 ms; (b) 1 ms; (c) 2 ms; (d) 3 ms; (e) 4 ms; (f) 5 ms in humid condition of RH ¼ 77%.

(a)

increase of surface conductivity in humid air both contribute to the changing of PD behavior.

3 2.5

5. Conclusions

N

avg

2 1.5 1 0.5 0

0

2 T [ms]

4

1

(b)

The relative humidity has a great effect on the PD activities, not only by reducing the effective free electrons in the air, but also by modifying the surface condition of the insulating material. This work presents the effect of relative humidity on the PD activities on machine insulation at trapezoidal testing voltage waveforms, with the derivative of the applied voltage dU/dt ranging from the maximum possible (approximately square-wave) to the minimum possible (triangle-wave). The PD measurements have been performed in a metal-dielectric air gap on a short length of stator bar, whose groundwall insulation consists of mica, epoxy resin and glass-fiber. The variation of PD characteristics with the changing of relative humidity has been studied, for instance, average number of PD pulses, PD amplitude, the phase of PD appearance and the discharge time shift.

20

6

V0 [kV]

5

10

Q

tot

[nC]

15

5 0

4 3

0

2 T [ms]

4

1

Fig. 10. PD characteristics at different values of T1 in dry condition: (a) average number of PDs per cycle Navg; (b) total PD charges per cycle Qtot.

2 0

20

40 RH [%]

60

Fig. 11. PD inception voltage varies with the relative humidity in air.

80

X. Wang et al. / Journal of Electrostatics 83 (2016) 88e96

(a)

(a)

6

250

2 0

150 100

−2 50

−4 −6 0

(b)

RH = 8 % RH = 29 % RH = 77 %

200 Navg

PD Voltage [V]

4

95

0.005

0.01 Time [s]

0.015

0 0

0.02

(b)

4

1

2 3 T1 [ms]

70

Qtot [nC]

PD Voltage [V]

−2

30 10 0 0

0.005

0.01 Time [s]

0.015

0.02

(c)

0.6

3

[ms]

2.5

shift

0.4 0.2

t

PD Voltage [V]

40 20

−4 0

(c)

50

0

5

RH = 8% RH = 29 % RH = 77 %

60

2

4

2

1

2 3 T1 [ms]

4

5

4

5

RH = 8 % RH = 29 % RH = 77 %

1.5 1 0.5

0 −0.2 0

0 0

1

2 3 T [ms] 1

0.005

0.01 0.015 Time [s]

0.02

Fig. 12. One typical period of time-sequential PD pulses at T1 ¼ 1 ms with different humidity conditions: (a) RH ¼ 8%; (b) RH ¼ 29%; (c) RH ¼ 77%.

The results show that discharge amplitude can be reduced greatly with the increasing humidity; on the other hand, the average number of discharge pulses is increased dramatically with the increasing humidity. This is mainly due to the increased surface

Fig. 13. PD characteristics at different values of T1 and different relative humidity levels: (a) average number of PDs per cycle Navg; (b) total PD charges per cycle Qtot; (c) time shift of PD occurrence tshift.

conductivity in humid air, leading to the transition from few bigamplitude discharges in dry air to plenty of small-amplitude discharges in humid air. The phase of PD appearance indicates that in the higher humidity level PD activities have a wider range based on the applied testing voltage waveforms. In the dry condition

96

X. Wang et al. / Journal of Electrostatics 83 (2016) 88e96

(RH ¼ 8%) most of PD events concentrate during the voltage increasing period of T1. With increasing humidity to RH ¼ 29%, PD will continue to occur even during the constant-voltage period of T2 in the trapezoidal waveform, this thus reflect that field in the gap is not constant even during the short period of T2, but is affected by surface charge decay or lateral charge spread on the surface. Moreover, in the humid condition (RH ¼ 77%), discharges even continue to occur during the voltage falling period of T3. Moreover, the PD occurs significantly later in humid air. The time shift of PD occurrence is increased with the raised relative humidity mainly due to the lack of free electrons caused by the presence of water molecules in the air. Acknowledgment The project was funded by the Swedish Energy Agency, Elforsk AB, ABB AB and Swedish Railway Company via the ELEKTRA program (project No. 36161), which is gratefully acknowledged. References € hlich, T. Weiers, R. Vogelsang, Insulation failure [1] R. Brütsch, M. Tari, K. Fro mechanisms of power generators, IEEE Electr. Insul. Mag. 24 (No. 4) (2008) 17e25. lec, Partial discharge signal interpretation for generator di[2] C. Hudon, M. Be agnostics, IEEE Trans. Dielectr. Electr. Insul. 12 (2005) 297e319. [3] M. Farahani, H. Borsi, E. Gockenbach, M. Kaufhold, Partial discharge and dissipation factor behavior of model insulating systems for high voltage rotating machines under different stessses, IEEE Electr. Insul. Mag. 21 (No. 5) (2005) 5e19. [4] D.L. Evans, IEEE working group report of problems with hydrogenerator thermoset stator windings Part I-Analysis of survey, IEEE Trans. Power Apparatus Syst. PSA-100 (No. 7) (1981) 3284e3291. [5] N. Taylor, Dielectric Response and Partial Discharge Measurements on Stator Insulation at Varied Low Freqency, KTH, Stockholm, Sweden, 2010. PhD thesis. [6] G.C. Stone, C.V. Maughan, D. Nelson, R.P. Schultz, Impact of slot discharges and vibration sparking on stator winding life in large generators, IEEE Electr. Insul. Mag. 24 (No. 5) (2008) 14e21. [7] R.J. Jackson, A. Wilson, Slot-discharge activity in air-cooled motors and generators, IEE. Proc. 129 (1982) 159e167. Pt. B. lec, M. Le vesque, Study of slot partial discharges in air-cooled [8] C. Hudon, M. Be generators, IEEE Trans. Dielectr. Electr. Insul. 15 (2008) 1675e1690.  vesque, E. David, C. Hudon, M. Be lec, Effect of surface degradation on slot [9] M. Le partial discharge activity, IEEE Trans. Dielectr. Electr. Insul. 17 (2010) 1428e1440.  David, C. Hudon, M. Be vesque, E. lec, Contribution of humidity to the [10] M. Le evolution of slot partial discharges, IEEE Trans. Dielectr. Electr. Insul. 19 (2012) 61e75. [11] K. Younsi, D. Snopek, J. Hayward, P. Menard, J.C. Pellerin, Seasonal changes in partial discharge activity on hydraulic generators, in: IEEE Electr. Insul. Conf. And Electrical Manufacturing & Coil Winding, Cincinnati, Ohio, USA, 2001, pp. 423e428. [12] M. Fenger, G.C. Stone, Investigations into the effect of humidity on stator winding partial discharge, IEEE Trans. Dielectr. Electr. Insul. 12 (2005)

341e346. [13] Z. Nawawi, Y. Murakami, N. Hozumi, M. Nagao, Effect of humidity on time lag of partial discharge in insulation-gap-insulation system, in: 8th International Conference on Properties and Applications of Dielectric Materials, 2006, pp. 199e203. Bali, Indonesia. [14] C. Kane, A. Golubev, C. Patterson, R. Astasiewicz, The importance of correlating dynamics when performing partial discharge measurement and analysis, in: Conf. Record of the 2004 IEEE International Symposium on Electrical Insulation, Indianapolis, USA, 2004, pp. 117e122. [15] R. Soltani, E. David, L. Lamarre, Study on the effect of humidity on dielectric response and partial discharge activity of machine insulation materials, in: IEEE Electrical Insulation Conference, Montreal, QC, Canada, 2009, pp. 343e347. [16] E. Binder, A. Draxler, M. Muhr, S. Pack, R. Schwarz, H. Egger, A. Hummer, Effects of air humidity and temperature to the activities of external partial discharges of stator winding, in: 11th Intern. Symp. High Volt. Eng, ISH, London, UK, 1999, 5.264e5.257. [17] R. Soltani, E. David, L. Lamarre, L. Lafortune, Effect of humidity on charge and discharge current of large rotating machines bar insulation, in: Conf. Record of the 2004 IEEE International Symposium on Electrical Insulation, Vancouver, BC, 2008, pp. 412e415. [18] D.M. Hepburn, I.J. Kemp, A.J. Shields, J. Cooper, Degradation of epoxy resin by partial discharges, IEE Proc.-Sci. Meas. Technol. 147 (No. 3) (2000) 97e104. [19] L.B. Rockland, Saturated salt solutions for static control of relative humidity between 5 ºC and 40 ºC, Anal. Chem. J. 32 (No. 10) (1960) 1375e1376. [20] B. Florkowska, P. Zydron, Analysis of conditions of partial discharges inception and development at non-sinusoidal testing voltages, in: Conf. On Electrical Insulation and Dielectric Phenomena (CEIDP), 2006, pp. 648e651. [21] E. Lindell, T. Bengtsson, J. Blennow, S.M. Gubanski, Influence of rise time on partial discharge extinction voltage at semi-square voltage waveforms, IEEE Trans. Electr. Insul. 17 (2010) 141e148. [22] D. Fabiani, G.C. Montanari, A. Cavallini, G. Mazzanti, Relation between space charge accumulation and partial discharge activity in enameled wires under PWM-like voltage waveform, IEEE Trans. Electr. Insul. 11 (2004) 393e405. [23] X. Wang, M. Ghaffarian Niasar, R. Clemence, H. Edin, Partial discharge analysis in a needle-plane gap with repetitive step voltage, in: Int. Conf. On Electrical Insulation and Dielectric Phenomena (CEIDP), Montreal, Canada, 2012, pp. 92e95. [24] E. Kuffel, Electron attachment coefficients in oxygen, dry air, humid air and water vapour, Proc. Phys. Soc. 74 (1959) 287e308. [25] R. Messaoudi, A. Younsi, F. Massines, B. Despax, C. Mayoux, Influence of humidity on current waveform and light emission of a low-frequency discharge controlled by a dielectric barrier, IEEE Trans. Dielectr. Electr. Insul. 3 (1996) 537e543. [26] J.M.K. MacAlpine, C.H. Zhang, The effect of humidity on the charge/phaseangle patterns of ac corona pulses in air, IEEE Trans. Dielectr. Electr. Insulation 10 (2003) 506e513. [27] G. Berger, E. Marode, O. Belabed, B. Senouci, I. Gallimberti, A. Osgualdo, Effect of water vapour on the discharge regimes and the deviations from similarity law in compressed SF6 for positive polarity, J. Phys. D. Appl. Phys. 24 (1991) 1551e1556. [28] B. Lutz, J. Kindersberger, Influence of relative humidity on surface charge decay on epoxy resin insulators, in: Int. Conf. On Properties and Applications of Dielectric Materials, Harbin, China, 2009, pp. 883e886. [29] K. Wu, Y. Suzuoki, L.A. Dissado, The contribution of discharge area variation to partial discharge patterns in disc-voids, J. Phys. D. Appl. Phys. 37 (2004) 1815e1823. [30] C. Hudon, R. Bartnikas, M.R. Wertheimer, Spark-to-glow discharge transition due to increased surface conductivity on epoxy resin specimens, IEEE Trans. Electr. Insul. 28 (1993) 1e8.