Modeling the Development of Low Current Arcs and Arc Resistance Simulation

Xin Zhang and Alasdair BruceNational Grid

Electricity National Control CentreWokingham, RG41 5BN, United Kingdom

Simon Rowland and Vladimir Terzija The University of Manchester

School of Electrical and Electronic EngineeringManchester, M60 1QD, United Kingdom

Siqi BuThe Hong Kong Polytechnic UniversityDepartment of Electrical Engineering

Kowloon, Hong Kong

ABSTRACTLow current arcs in the range 0.5 ~5 mA occur in power networks in situations such as on overhead line insulators and cable terminations. These arcs are important because of their potential contribution to surface ageing, asset failure and potential flashover. In this paper, the development of low current arcs is classified in three stages: a formative leakage current phase (~µA), a stage where discharges occur but are unstable with each half power cycle (<1 mA) and a period of stable discharges (>1 mA). Arc resistance is a key element in controlling arc behavior in each stage, and is modeled as the combination of a stable arc resistance, an oscillating resistance and a surface resistance. The resulting arc model has been developed in PSCAD/EMTDC, to simulate an arc/discharge in each development stage. Simulations compare well with experimental data. The simulation reveals that peak arc current plays a key role in the transition from an unstable to stable arc. Analysis shows a significant increase in discharge energy as a result of its stabilization. These models explain the conditions required for accelerated ageing of polymeric insulators and can be used to design and interpret testing regimes, and for polymeric insulator asset management.

Index Terms — Arc discharges, insulators, modeling, simulation, high voltage testing, arc resistance, leakage currents, energy, PSCAD/EMTDC, condition monitoring

1 INTRODUCTIONLOW amplitude surface currents and resulting discharges and

arcs are a concern in polymeric outdoor insulation design and management. Components such as overhead line insulators, cable and transformer bushings, and all-dielectric, self-supporting (ADSS) cables are subject to electric fields, surface leakage currents, discharges and arcs [1-4]. Electrical arcs can lead to surface erosion and tracking [5-6], thereby reducing the reliability of insulation. This may eventually lead to mechanical failure of insulators, flashover or the puncture of bushings [7-11].

A widely used low-current arc model is based on Mayr’s equation, which describes a dynamic arc resistance with a non-linear dependence on arc current [12]. Application of Mayr’s equation illustrates that arc resistance plays an important role in

low-current arc behavior and associated models [13]. More recent models are based on differential equation systems [14], multiple-arc electrical equivalent circuits [15], partial-arc and spark models

[16], direct current / alternating current / impulse voltage sources [17-18], and various configurations of insulation media such as insulator profiles [19] and ice coated insulator surfaces [20]. Such models have had various degrees of success in modeling the electrical behavior of arcs, with a few of them having capacity to simulate dynamic arc resistance.

However, most arc discharge models are based on high current flashover [16-20], over long distances [21-22], or short arcs between high current contacts [23-24]. This is to meet the need to Manuscript received on 14 October 2017, in final form 15 April 2018,

accepted 16 April 2018. Corresponding author: X. Zhang.

understand system reliability and control. Modeling of low current arcs is less well developed, and previous low-current arc models are applicable only to arc discharges in equilibrium with repeatable dynamic characteristics [14-15]. No model has yet been proposed to simulate the dynamic behavior of low current arcs from initial formation to full development. Such a model is required to better understand the processes behind long-term insulator ageing.

Experimental work has identified two forms of discharge during low current arc development: unstable discharges with peak currents typically less than ~1 mA, and stable arcs with peak currents typically above ~1 mA. These are distinguished depending on whether a discharge is stable within half a power cycle. The transition between unstable and stable discharges has been confirmed in laboratory conditions [25].

Based on these experimental results, this paper identifies three stages of discharge development and these are simulated in PSCAD/EMTDC. The stages are: a formative leakage current phase (~µA), a stage where discharges occur but are unstable within each half power cycle (<1 mA) and a period of stable discharges (>1 mA). Arc resistance is simulated as three sub-units named ‘stable arc resistance’, ‘oscillating resistance’ and ‘surface resistance’. The simulations of voltage and current in each of the three stages, allow associated discharge/arc energies to be analyzed.

2 LABORATORY TEST AND RESULTSDetailed test procedures have been reported previously [25-

26], and are summarized in Figure 1. A silicone-sheathed pultruded rod of outer diameter 22 mm is tested with electrodes directly attached and separated by 200 mm. Water of conductivity 16,000 µS/cm is sprayed onto the surface and a variable 50 Hz voltage of up to 42 kV (peak) is applied across the sample and a 6 MΩ resistor (modelling the resistance of a large wet insulator body). The data acquisition system (PC and LabVIEW) includes a voltage divider with the ratio of 1:10,000 to obtain the voltage at the HV end of the rod (approximately the discharge voltage), and a 2 kΩ resistor to enable measurement of the leakage current.

0-42 kV

Voltage transformer

Current limiting resistor

Test sample

Current signalia(t)

Vol

tage

div

ider

Voltage signalua(t)

spray

2 kΩ 6 MΩ

Figure 1. The experimental set-up.

2.1 FORMATIVE LOW CURRENT PHASEIn the early stage of moisture deposition on the hydrophobic

sample surface, the voltage and current traces are initially 90 degrees out of phase (capacitive), and these gradually become in phase (resistive) as the test progresses and the surface becomes more hydrophilic, as shown in Figure 2. These

results were generated from tests in fog conditions to allow a gradual and uniform wetting process [27]. The leakage current increases due to the moisture being able to form more continuous, and so conductive, paths on the silicone surface, which reduces the surface resistance. This is widely reported behavior on silicone surfaces along with methods to detect and predict leakage current [28]. No arcing activities are observed at this stage.

(a) (b)

(c) (d)Figure 2. Voltage and current curves show, as the current increases: (a) time 0, voltage lags current by 90 degrees; (b) at 90 seconds from test start, voltage

lags current by 60 degrees; (c) at 150 seconds from test start, voltage lags current by 30 degrees; (d) at 220 seconds, voltage and current are in phase

[27].

2.2 UNSTABLE DISCHARGESUnstable discharges are observed for currents over 0.6 mA.

In these cases the applied voltage is high enough to strike an arc across a dry band on the insulator, but the current is insufficient to sustain an arc, so that the discharge is ‘unstable’. Following the discharge development, typically after an extended 2 minutes of wetting time, the peak discharge current gradually increases. This could be due to the increase in surface conductivity of the insulator, or a reduction in the natural discharge length [25-26]. Increased peak current is accompanied by discharges gradually becoming stable arcs, with less oscillation of arcing voltage and current, as seen in Figure 3.

(a) (b)

(c) (d)

Figure 3. Unstable discharges become stable arcs: (a) Initial unstable discharges appear; (b) Reduced instability and discharge oscillation frequency

after 2.5 minutes; (c) Stable arcs appear but are accompanied by some instability after 7 minutes; (d) Stable arcs finally dominate after 8.5 minutes

[25].

The development time from an unstable discharge to a stable arc varies depending on the spray rate. In this test, a spray rate of 0.12 g/cm2/hr is used throughout the test. It is observed that a higher spray rate would accelerate the arc development process, due to more effective moisture deposition on the insulation surface.

2.3 STABLE ARCSWhen an equilibrium is achieved between an electrical arc

and the surrounding moisture conditions, stable arcs appear with smooth and repeatable voltage and current characteristics, typically when the arc current is above 1.0 mA. In Figure 4, the impact of increasing the source voltage can be observed to generate 1.0, 1.5, 2.0, 2.5, 3.0, 3.5 and 4.0 mA arcs. Each test is conducted for at least 30 minutes to allow equilibrium to be established. These current and voltage traces are of the classical ‘dry-band arc’ form, but of lower current magnitude than is normally studied.

(a) (b)

(c) (d)

(e) (f)

(g)Figure 4. Voltage and current curves show stable arcs with: (a) 1.0 mA; (b)

1.5 mA; (c) 2.0 mA; (d) 2.5 mA; (e) 3.0 mA; (f) 3.5 mA; (g) 4.0 mA.

3 ARC RESISTANCE MODELINGThe stability of an arc is determined by two things. Firstly,

sufficient voltage is required to strike an arc across a high resistance gap. Secondly, once the arc is established, it can only be maintained if sufficient voltage remains available across the lower resistance arc to maintain the current flow.

The available voltage is dependent on the supply voltage and ratio of the resistance of the arc to the resistance of the supply circuit, which is normally the resistance of the rest of the insulator surface. In our experiment this is the 6 MΩ resistor. In turn, the resistance of the arc is a strong function of current [11], which is controlled by the supply circuit which in this situation has a higher resistance than the arc itself. Arc resistance is a thus a key parameter in the arc development process.

In this paper, arc resistance is simulated in three parts corresponding to ‘Stable Arc Resistance’ in series with ‘Oscillating Resistance’, and ‘Surface Resistance’, as shown in Figure 5. ‘Stable Arc Resistance’ represents the well-established stable arc characteristic, ‘Oscillating Resistance’ simulates unstable discharge characteristics, and the ‘Surface Resistance’ is used to simulate the low current arc formative stage.

Oscillating Resistance

Rosci(t)

Stable Arc Resistance

Rstab(t)

(a) (b)Surface Resistance

Rsurf(t)Figure 5. The simulation of arc resistance consists of ‘Stable Arc

Resistance’, ‘Oscillating Resistance’ and ‘Surface Resistance’ for different stages in arc development.

23

3.1 STABLE ARC RESISTANCEFor stable arcs, three periods are identified within each half

cycle in Figure 6 as pre-arcing, arcing and post-arcing periods. A detailed mathematical model to simulate voltage and current traces for each period has been reported and validated previously [29].

t1

Pre-arcing

Ut1

Post-arcing

Volta

ge s

imul

atio

n

Ut2

t2

Cur

rent

sim

ulat

ion

0

Arcing

ua(t)

ia(t)

ua(t) ia(t)

40 ms

kV

Figure 6. Double sinusoidal model to simulate I-V curves for stable arcs.

In this paper, the accuracy of that model is improved by simulating the arcing voltage ua(t) with a sloping value from Ut1 (arc ignition voltage) to Ut2 (arc extinction voltage). The model is also extended to simulate arcs for different current levels, Ia (peak current). Between times t1 and t2, the new model gives equations (1) and (2):

ua (t )=U t 1+

U t 1−U t 2

t 1−t 2(t−t 1) (1)

ia (t )=I a sin ωi [t−( πωu

−t 2)] (2)

where Ia is the peak value (mA) of the sinusoidal current, ωi

and ωu are the angular frequency (rad/ms) of the sinusoidal arcing current and source voltage, t1 is the arc ignition time (ms), t2 is the arc extinction time (ms), Ut1 is the arc ignition voltage (kV), and Ut2 is the arc extinction voltage (kV).

The stable arc resistance between times t1 < t < t2 is then given by equation (3), R stab ( t )=∞ in the pre-arcing and post-arcing periods.

R stab (t )=ua (t )ia (t )

=U t 1+

U t1−U t2

t1−t2( t−t1 )

I a sin ωi[ t−( πωu

−t2)](3)Arc length is reported as an important parameter for arc

resistance [14, 26]. The relevant experimental data is summarized in Table 1. Arc peak current (Ia), arc ignition voltage (Ut1), and arc extinction voltage (Ut2) are linked to arc length (La) by the curve fitting method of least squares estimator (LSE) as shown in Figure 7:

(a) (b)

(c) (d)

(e)Figure 7. (a) The relationship between arc peak current Ia and arc length La;

(b) The relationship between arc ignition voltage Ut1 and arc length La; (c) The relationship between arc extinction voltage Ut2 and arc length La; (d) The arc

ignition time t1; (e) The arc extinction time t2.

Table 1. Summary of the test data from low current stable arcs. The final row gives the model derived from the data in each case.

ArcLength La (cm)

Arc PeakCurrent Ia (mA)

Arc Ignition VoltageUt1 (kV)

Arc Extinction VoltageUt2 (kV)

t1

(ms)t2

(ms)1.0 1.0 5.80 3.78 3.60 8.691.4 1.5 7.91 5.57 2.26 8.602.1 2 12.11 8.21 2.81 8.402.3 2.5 12.50 7.72 2.39 8.74

2.9 3 15.92 11.23 2.95 8.403.3 3.5 17.48 12.31 2.91 8.433.6 4 16.51 11.04 3.46 8.63La 1.12La-0.15 4.52La+1.88 3.16La+1.06 2.91 8.55

Substituting Ut1, Ut2, t1, t2, the stable arc resistance for an arc of length La is given by,

R stab ( t , La )=3.84 La+1.47

(1.12 La−0.15 ) sin 0.44 (t−1.45 )(4)

where 2.91 ms < t < 8.55 ms. Finally, the arc resistance for a given arc length can be calculated from equation (4), and is shown to be a good fit to the corresponding measured arcs in Figure 8.

Figure 8. Simulated and measured arc resistances for current levels from 1.5mA to 4.0mA through a complete half cycle.

3.2 UNSTABLE DISCHARGE RESISTANCEUnstable discharges repeatedly switch the arc on and off

due to insufficient current being available to sustain an arc. This phenomenon is modelled as a rapid increase in discharge resistance when an arc is switched off. The Dirac delta function δ (t ) is used to simulate the ‘impulse’ feature of unstable discharge resistance as shown in Equation (5) and Figure 9a,

δ (t )= ϵπ (t2+ϵ 2 )

(5)where ε is a parameter to control the impulse peak.

An oscillating frequency f is introduced to control the instability of low current discharges. The oscillating resistance is proposed based on δ (t ) and f in Equation (6). The graphic presentation of ‘oscillation resistance’ based on the Dirac delta function is shown in Figure 9b.

Rosci ( t )=∑n=1

N

δ (t−nf ) (6)

where 1f

represents the impulse interval, N is the total

number of impulses during the experimental period. N=f for a period of 1 s.

(a) (b)

Figure 9. (a) Dirac delta function δ (t ) with impulse peak 1πε

; (b) The

modeling of oscillating resistance with frequency f.

By analyzing features of the unstable discharges from test results such as those shown in Figure 3, each oscillation width is found to take approximately 0.01 ms. In Equation (5), ε is set to 0.00006 corresponding to an impulse peak value of 2000 MΩ, which is 1000 times higher than the measured stable arc resistance. The experimental results also show the oscillating frequency f takes an initial value of approximately 2000 Hz for a 0.6 mA unstable discharge, and gradually reduces to 0 Hz for a 1.0 mA stable arc. This phenomenon may be explained in that higher current discharges contain higher energies, thereby sustaining more stable discharges. Thus, the oscillating frequency f is a function of discharge peak current Ia. A simple linear relationship is assumed between f and Ia for 0.6 mA < Ia < 1.0 mA so that

f =5000 (1−I a ) (7)

Substituting f and ε , the oscillating resistance for unstable discharges is given by,

Rosci (t , I a )=0.006π ∑

n=1

N 1

(t− n5000 (1−I a ) )

2

+0.0062

(8)where 0.6 mA< I a<1.0 mA .

The entire unstable discharge resistance is modelled as the oscillating resistance Rosci (t , I a ) in series with the stable arc

resistance R stab ( t , La ) with arc current < 1 mA, as shown in Figure 10.

Figure 10. Simulated unstable discharge resistance for current levels from 0.6 mA to 0.95 mA through a complete half cycle.

3.3 SURFACE RESISTANCEIn the formative leakage current stage before arc discharges

appear, the material surface resistance is modelled as a variable resistor in parallel with a capacitor as shown in Figure 11.

1µF (b)(a)

Capacitor

Water-layer resistor

(wet)10 Ω

(dry)∞

Figure 11. Electrical model of surface resistance in both dry and wet conditions.

By analyzing the shift change between voltage and current traces in Figure 2, when the sample surface is still dry, the variable resistor taps at point ‘a’ with an open circuit. The leakage current flows in the capacitor. An initial dry current of 20 μA corresponds to a capacitance of around 1 pF. The current leads the voltage at this stage. When the sample becomes completely wet due to moisture deposition, the variable resistor drops to a value of ~10 Ω (equivalent to a moisture conductivity of 16,000 µs/cm), which is negligible compared to the MΩ levels of arc resistance. The wet surface will shunt the capacitor and make the surface purely resistive, and the current flows in phase with the voltage.

The simulated surface impedance is therefore,

R surf ( t )=

Rwater (t)ωC

Rwater(t)+1

ωC(9)

where rwater is a variable water-layer resistance from 10 Ω to ∞, ω is angular frequency at 50 Hz, and C is fixed at 1 µF.

4 ARC SIMULATION IN PSCAD/EMTDCThe PSCAD simulation is built as shown in Figure 12. The

source voltage is produced by a 50 Hz generator (output of 5-42 kV), with a current limiting resistor of 6 MΩ. The arc resistance is simulated as stable arc resistance (R stab) and oscillating resistance (Rosci) as proposed in section 3.1 and 3.2. The surface resistance (R surf) is simulated as a variable water-layer resistor in parallel with a capacitor as illustrated previously in section 3.3. This simulation replicates the arrangement of the experimental setup.

Current limiting resistor

Source voltage5 kV~42 kV

Surface resistance Arc resistance

Figure 12. Simulation of low current arc discharge in PSCAD using stable arc resistance, oscillating resistance and surface resistance.

234

4.1 SIMULATION OF ARC RESISTANCEThe simulation of arc resistance consists of two blocks in

PSCAD: stable arc resistance (R stab) and oscillating resistance (Rosci) as shown in Figure 13. The stable arc resistance block is constructed based on equation (4), and the oscillating resistance block is based on equation (8). The coordination control between two blocks and surface resistance is given in Table 2.

Stable arc resistance

Oscillating resistance

Figure 13. The simulation circuit for arc resistance consists of stable arc resistance and oscillating resistance.

Table 2. The coordination between Stable Arc Resistance, Oscillating Resistance and Surface Resistance.

StageArc

Current (mA)

Stable Arc Resistance

Oscillating Resistance

Surface Resistance

Phase Shifting <0.6 0 0 Equation (9)

Unstable Discharge 0.6 – 1.0 Equation (4) Equations (8) 0

Stable Arc 1.0 – 5.0 Equation (4) 0 0

4.2 SIMULATION OF UNSTABLE DISCHARGESThe simulation results for unstable arcs including the

transition process from unstable to stable conditions are shown in Figure 14. The correlation coefficients between the simulation and experiment results (in Figure 3) are above 0.75 for all cases. The PSCAD simulation here reflects the main electrical features of unstable discharges. It can also simulate the reduced oscillating frequency when arc peak current rises from 0.6 mA to 1.0 mA. This is a key development feature from unstable discharges to stable low current arcs.

(a) Simulated unstable arcs of 0.6mA with oscillating frequency at 2000 Hz

(b) Simulated unstable arcs of 0.8mA with oscillating frequency at 1000 Hz

(c) Simulated unstable arcs of 0.9mA with oscillating frequency at 500 Hz

(c) Simulated unstable arcs of 0.95mA with oscillating frequency at 250 Hz

Figure 14. Simulation result for unstable arcs as current increases.

4.3 SIMULATION OF STABLE ARCSThe PSCAD simulation for arcs with 1.0 mA, 2.0 mA, 3.0

mA and 4.0 mA peak current are shown in Figure 15. Correlation coefficients above 0.95 between simulation and experiment results (illustrated in Figure 4) confirm the validity of the PSCAD simulation for simulating stable arcs.

(a) Simulated stable arc of 1.0 mA

(b) Simulated stable arc of 2.0 mA

(c) Simulated stable arc of 3.0 mA

(d) Simulated stable arc of 4.0 mA

Figure 15. Simulation of stable arcs at current levels from 1.0 mA to 4.0 mA.

4.4 SIMULATION OF SURFACE RESISTANCEThe PSCAD simulation results for the early stage of leakage

current formation are shown in Figure 16. Results show that a reduction in the water-layer resistance results in leakage current growth, and gradual reduction in phase difference with voltage. This agrees with experimental results illustrated previously in Figure 2.

(a) Simulated leakage current of 0.015 mA and 90° phase lag

(b) Simulated leakage current of 0.055 mA and 30° phase lag

(c) Simulated leakage current of 0.08 mA and 15° phase lag

(d) Simulated leakage current of 0.15 mA in phase with the source voltage

Figure 16. Simulation of leakage current formation on wet insulation surface.

5 DISCUSSIONThe stability of low current discharges has been previously

explored using experimental methods [25]. The low current arc has an oscillating feature in voltage and current in the range between 0.6 mA to 1.0 mA. The modeling of unstable discharges here reveals a relationship between oscillation frequency f and discharge peak current Ia. Equation (6) also quantifies this relationship. When the discharge peak current increases from 0.6 mA to 0.95 mA, the oscillation frequency reduces from 2000 Hz to 250 Hz, representing a more stable arc.

The transition from unstable discharges to stable arcs is a critical process that increases arc energy and so presents an increased threat to the surface of polymeric insulation as a result of increased arc temperature and enhanced energy transfer to the polymer surface [30-31]. Experiment work reveals that arc energy is increased by a factor of up to 3, from

an initial unstable state to a well-developed stable state. Based on the simulated arc voltage and current from PSCAD/ EMTDC, the arc energy per half a cycle is calculated as:

Ea=∫0

T

|ua (t ) ia ( t )|d t (10)

where Ea is arc energy (joules) per cycle, ia(t) is the simulated arc current curve, ua(t) is the simulated voltage curve, and T is the period of the source voltage (40 ms).

The energy trend as the discharge activity transitions from unstable discharges to stable arcs is shown in Figure 17 for both PSCAD/EMTDC simulations and experiment results. Increased arc energy has been found associated with the development from unstable discharges to stable arcs. This is due to the increase in discharge current and reduction in oscillations which prolongs the total arcing period per power cycle. In addition, 0.01 joule/cycle is identified as a threshold between unstable discharges and stable arcs.

Figure 17. Arc energy trends from unstable discharges to stable arcs for both PSCAD simulation and experiment results.

The modeling approach here allows a deeper understanding of energy change during the discharge/arc transition process, and provides a tool by which the energy in a discharge can be calculated and monitored. This can now be developed into an online monitoring system that measures real discharge current and interprets such measurements in terms of arc energy through the PSCAD/EMTDC simulation program.

6 CONCLUSIONThe development of low current arc discharges between

10 μA and 5 mA has been classified into three stages: pre- formative leakage current, unstable discharges and stable arcs. Arc resistance is modelled in three states: surface resistance, oscillating resistance and stable resistance. A circuit using arc resistance models was built in PSCAD/EMTDC to simulate arc discharges in corresponding development stages. The simulation results of arc voltage and current are found to have good correlation with experimental results.

It has been found that the observed increase of discharge stability with peak current at around 1 mA, in the form of reduced oscillation in arc voltage and current, is reproduced well in the model. Similarly an increase in thermal energy is reproduced in the model due to the change in arc stability, providing an indication of enhanced power in the arc, and so

identifies situations in which more aggressive/accelerated ageing of a polymer insulator surface may occur.

Future work will introduce a stochastic process into arc modeling, to simulate the random features during the arc development. The modeling of multiple-arcs with individual-controlled growth and extinction is also to be studied.

REFERENCES[1] D. Zhang, Z. Zhang, X.L. Jiang, Z. Yang, Y. Liu and M. Bi, “Study on

the Flashover Performance of Various Types of Insulators Polluted by Nitrates”, IEEE Trans. Dielectr. Electr. Insul, vol. 24, No. 1, pp. 167-174, 2017.

[2] Z. Jia, C. Chen, X. Wang and H. Lu, “Leakage Current Analysis on RTV Coated Porcelain Insulators During Long Term Fog Experiments”, IEEE Trans. Dielectr. Electr. Insul., vol. 21, No. 4, pp. 1547-1553, 2014.

[3] M. Albano, R.T. Waters, P. Charalampidis, H. Griffiths and A. Haddad, “Infrared Analysis of Dry-band Flashover of Silicone Rubber Insulators”, IEEE Trans. Dielectr. Electr. Insul., vol. 23, No. 1, pp. 304-310, 2016.

[4] B. Moula, A. Mekhaldi, M. Teguar and A. Haddad, “Characterization of Discharges on non-Uniformly Polluted Glass Surfaces Using a Wavelet Transform Approach”, IEEE Trans. Dielectr. Electr. Insul., vol. 20, No. 4, pp. 1457-1466, 2013.

[5] R. A. Ghunem, S. H. Jayaram and E. A. Cherney, “Erosion of Silicone Rubber Composites in the AC and DC Inclined Plane Tests”, IEEE Trans. Dielectr. Electr. Insul., vol. 20, pp. 229-236, 2013.

[6] Y. Xiong, S. M. Rowland, J. Robertson and R. J. Day, “Surface Analysis of Asymmetrically Aged 400 kV Silicone Rubber Composite Insulators”, IEEE Trans. Dielectr. Electr. Insul., vol. 15, No. 3, pp. 763-770, 2008.

[7] S. M. Rowland, “Prevention of dry-band arc damage on ADSS cables”, IEEE Trans. Dielectr. Electr. Insul., vol. 13, No. 4, pp. 765-772, 2006.

[8] R. S. Gorur, E. A. Cherney and J. T. Burnham, Outdoor Insulators, ISBN 0-9677611-0-7, Phoenix, Arizona, USA, 1999.

[9] S. M. Rowland, K. Kopsidas and X. Zhang, “Ageing of Polyethylene ADSS Sheath By Low Currents”, IEEE Trans. Power Del., vol. 25, No. 2, pp. 947-952, 2010.

[10] S. E. Kim, E. A. Chemey and R. Hackam, “Effect of Dry Band Arcing on the Surface of RTV Silicone Rubber Coatings”, in Proc. IEEE Int. Symp. on Electr. Inul., 1992, pp. 237-240.

[11] M. Kumosa, L. Kumosa and D. Armentrout, “Failure Analyses of Nonceramic Insulators Part 1: Brittle Fracture Characteristics”, IEEE Electr. Insul. Mag., vol. 21, May/June, pp. 14-27, 2005.

[12] O. Mayr, “Beitrag zur théorie der statischen und der dynamischen lichtbogens”, Archiv fur Elektrotechnik, vol. 37, pp. 588–608, 1943 (in German).

[13] T. E. Browne, “An Approach to Mathematical Analysis of a-c Arc Extinction in Circuit Breakers”, AIEE Trans., vol. 77, part III, pp. 1508-1517, 1959.

[14] Q. Huang, G. G. Karady, B. Shi and M. Tuominen, “Numerical Simulation of Dry-band Arcing on the Surface of ADSS Fiber-optic Cable”, IEEE Trans. Dielectr. Electr. Insul., vol. 12, No. 3, pp. 496-503, 2005.

[15] N. Dhahbi-Megriche and A. Beroual, “Self-Consistent Multi-arcs Dynamic Model for High Voltage Polluted Insulators”, IEEE Trans. Dielectr. Electr. Insul., vol. 23, No. 5, pp. 2899-2907, 2016.

[16] R. T. Waters, A. Haddad, H. Griffiths, N. Harid and P. Sarkar, “Partial-arc and Spark Models of the Flashover of Lightly Polluted Insulators”, IEEE Trans. Dielectr. Electr. Insul., vol. 17, No. 2, pp. 417-424, 2010.

[17] S. Taheri, M. Farzaneh and I. Fofana, “Improved Dynamic Model of DC Arc Discharge on Ice-covered Post Insulator Surfaces”, IEEE Trans. Dielectr. Electr. Insul., vol. 21, No. 2, pp. 729-739, 2014.

[18] N. Dhahbi-Megriche and A. Beroual, “Flashover Dynamic Model of Polluted Insulators under ac Voltage”, IEEE Trans. Dielectr. Electr. Insul, vol. 7, No. 2, pp. 283-289, 2000.

[19] R. Sundararajan and R. S. Gorur, “Dynamic Arc Modeling of Pollution Flashover of Insulators under DC Voltage”, IEEE Trans. Electr. Insul., vol. 28, no. 2, pp. 209-218, 1993.

[20] L. Shu, S. Wang, X. L. Jiang, Q. Hu, J. Liang, P. Yin and J. Chen, “Modeling of AC Flashover on Ice-covered Composite Insulators with Different Shed Configurations”, IEEE Trans. Dielectr. Electr. Insul., vol. 21, No. 6, pp. 2642-2651, 2014.

[21] V. V. Terzija and H. J. Koglin, “On the Modeling of Long Arc in Still Air and Arc Resistance Calculation”, IEEE Trans. on Power Del., vol. 19, No. 3, pp. 1012-1017, 2004.

[22] V. V. Terzija, G. Rreston, M. Popov and N. Terzija, “New Static ‘AirArc’ EMTP Model of Long Arc in Free Air”, IEEE Trans. Power Del., vol. 26, No. 3, pp. 1344-1353, 2011.

[23] P. Borkowski and E. Walczuk, “Thermal Models of Short Arc between High Current Contacts”, in Proc. 47th IEEE Holm Conf. Electr. Contacts, 2001, pp. 259-264.

[24] S. Nitu, C. Nitu, and P. Anghelita, “Electric Arc Model, for High Power Interrupters”, in Proc. Int. Conf. on Compr. as a Tool, Eurocon, 2005, pp. 1442-1445.

[25] X. Zhang and S. M. Rowland, “Stability and Energy of Low Current Surface Discharges on Wet Surfaces”, IEEE Trans. Dielectr. Electr. Insul., vol. 19, No. 6, pp. 2055-2062, 2012.

[26] X. Zhang, S. M. Rowland, and V. Terzija, “Increased Energy in Stable Dry-band Arcs due to Length Compression”, IEEE Trans. Dielectr. Electr. Insul., vol. 17, No. 2, pp. 473-480, 2010.

[27] X. Zhang, J.B. Zhu, S.Q. Bu, Q. Li, V. Terzija, and S.M. Rowland, “The Development of Low-Current Surface Arcs under Clean and Salt-Fog Conditions in Electricity Distribution Networks”, IEEE Access, In press, 2018.

[28] A. N. Jahromi, A. H. El-Hag, S. H. Jayaram, E. A. Cherney, M. Sanaye-Pasand and H. Mohseni, “A Neural Network Based Method for Leakage Current Prediction of Polymeric Insulators”, IEEE Trans. Power Del., 21, No. 1, pp. 506-507, 2006.

[29] X. Zhang and S. M. Rowland, “Modeling of Dry-band Discharge Events on Insulation Surfaces”, in Proc. Int. Symp. Electr. Insul., 2010, pp. 1-5.

[30] A. Xiao, S. M. Rowland, X. Tu and J. C. Whitehead, “Thermal Features of Low Current Discharges and Energy Transfer to Insulation Surfaces”, IEEE Trans. Dielectr. Electr. Insul., vol. 21, No. 6, pp. 2466-2475, 2014.

[31] L. H. Meyer, S. H. Jayaram and E. A. Cherney, “Correlation of Damage, Dry band Arcing Energy, and Temperature in Inclined Plane Testing of Silicone Rubber for Outdoor Insulation”, IEEE Trans. Dielectr. Electr. Insul., vol. 11, No. 3, pp. 424-432, 2004.

Xin Zhang (M’17) was born in Shandong, China. He received the B.Eng. degree in automation from Shandong University, China, in 2007; the M.Sc. and Ph.D. degrees in electrical power engineering from The University of Manchester, U.K., in 2007 and 2010 respectively.He is a Senior Power System Engineer in the Electricity National Control Centre at National Grid, U.K.. He is currently responsible for generation scheduling and technical analysis of the GB power

system, and previously worked in the areas of electricity network planning and market operation. He gained valuable industrial experience in the technical and commercial aspects of power transmission and distribution systems. His research interests include electrical arc phenomenon and condition monitoring techniques for high voltage equipment; intelligent methods in power system planning and operation; and smart distribution networks with the integration of renewable and multi-vector energy sources.Dr. Zhang is a Chartered Engineer with the U.K. Engineering Council, a member of CIGRE SC B5 Protection and Automation working group, a Professional Registration Interviewer (PRI) and Professional Registration Advisor (PRA) to interview and advise Chartered Engineer (CEng) and Incorporated Engineer (IEng) applicants for professional registration within the U.K.

Siqi Bu (S’11, M’12, SM’17) received the Ph.D. degree from the electric power and energy research cluster, The Queen’s University of Belfast, Belfast, U.K., in 2012, where he continued his postdoctoral research work before entering industry.He joined National Grid UK as a Power System Engineer and then became an experienced UK National Transmission System Planner and Operator. He is an Assistant Professor with The Hong Kong Polytechnic University, Kowloon, Hong Kong, and

also a Chartered Engineer with UK Engineering Council, London, U.K.. He has received various prizes due to excellent performances and outstanding contributions in operational and commissioning projects during the employment with National Grid UK. He is also the recipient of Outstanding Reviewer Award from IEEE Transactions on Power Systems and IEEE Transactions on Sustainable Energy. His research interests are power system stability analysis and operation control, including wind power generation, PEV, HVDC, FACTS, ESS and VSG.

Alasdair R. W. Bruce received the B.Sc. degree in physics from the University of Warwick, U.K. in 2010, and both the M.Sc. degree in sustainable energy systems and the Ph.D. degree in power systems engineering from the University of Edinburgh, U.K. in 2011 and 2016, respectively. Dr. Bruce is currently a power system engineer and data scientist at the Electricity National Control Centre at National Grid. His research interests include power system resilience and optimization, flexibility requirements, and

machine learning.

Vladimir Terzija (M’95, SM’00, F’16) was born in Donji Baraci (former Yugoslavia). He received the Dipl-Ing., M.Sc., and Ph.D. degrees in electrical engineering from the University of Belgrade, Belgrade, Serbia, in 1988, 1993, and 1997, respectively.He is the Engineering and Physical Science Research Council Chair Professor in Power System Engineering with the School of Electrical and Electronic Engineering, The University of Manchester,

Manchester, U.K., where he has been since 2006. From 1997 to 1999, he was an Assistant Professor at the University of Belgrade, Belgrade, Serbia. From 2000 to 2006, he was a Senior Specialist for switchgear and distribution automation with ABB AG Inc., Ratingen, Germany. His current research interests include smart grid application of intelligent methods to power system monitoring, control, and protection; wide-area monitoring, protection, and control; switchgear and fast transient processes; and digital signal processing applications in power systems.Prof. Terzija is Editor in Chief of the International Journal of Electrical Power and Energy Systems, the China One Thousand Talents Plan Scholar, an Alexander von Humboldt Fellow, as well as a DAAD and Taishan Scholar.

Simon M Rowland (F’14) was born in London, England. He completed the B.Sc. degree in physics at The University of East Anglia, and the PhD degree at London University, UK. He has worked for many years on dielectrics and their applications and has also been Technical Director within multinational companies. He joined The School of Electrical and Electronic Engineering in The University of Manchester in 2003, and was appointed Professor of Electrical Materials in 2009, and Head of School in

2015. Prof. Rowland was President of the IEEE Dielectric and Electrical Insulation Society in 2011and 2012.