CIRED

20th International Conference on Electricity Distribution

Prague, 8-11 June 2009 Paper 0652

EMTP SIMULATION OF LIGHTNING OVERVOLTAGE DISCHARGE TO MEDIUM VOLTAGE OVERHEAD LINES WITH COVERED CONDUCTORS
M.A. OMIDIORA, M. LEHTONEN, R.J. MILLAR Electrical Engineering Department, TKK Finland michael.omidiora@tkk.fi

ABSTRACT
This paper presents the evaluation and simulation of real lightning discharges that have occurred in medium voltage overhead lines equipped with covered conductors (CC). Lightning stroke current selections are based on real lightning data collected from the Finnish Meteorological Institute (FMI) from 1998 to 2008. The presented feeder is an overhead distribution line with a covered conductor subjected to direct lightning strokes to a line support. With the ATP/EMTP (Alternative Transient Program/ElectroMagnetic Transient Program), simulations of lightning strokes are conducted with different lightning current characteristics. All cases are considered with the modelling guidelines for fast transients, as specified in some lightning literature. Simulations are made to compare the resulting overvoltages and energy absorptions of surge arresters for all the analyzed cases.

ANALYZING LIGHTNING COVERED CONDUCTORS

EFFECT

ON

The lightning problem is associated with the incidence of thunderstorms in any given area. In Finland, lightning activities are monitored by a ground lightning location systems with eight sensors, whose detection covers almost the whole country [2]. Analysis of lightning data from 19982008 gives the number of average thunderstorm days as T = 12 days/year and a ground flash density of Ng = 0.786 strokes/ km2/year. The mean negative stroke is 15.2 kA (Fig. 1). The frequency of direct strokes to a conductor of 1 km length and average height h = 10 m is defined with

N = N g 0.028lh 0.6 [3] as 0.08761 strokes/km/year.

Assuming that: the value of the characteristic impedance of the CC overhead distribution line is 480 [1] the line-support earth resistance is 10 [4] the Basic Insulation Level (BIL) of the CC is 215 kV for 20 kV class CC (125 kV, support insulation and 90 kV, conductor insulation) the average lightning stroke per flash is 2 [4] the expected probability distribution of lightning currents is as in Fig. 1, then the conductor is prone to approximately 9 direct lightning strikes a year for a 100 km power line. Each stroke, to either a CC or line support, is expected to cause an interruption or outage, or damage to the CC.
Probability of exceeding ab scissa (p.u)
1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

INTRODUCTION
There has been an increase in the use of covered conductors (CC) in medium voltage (MV) distribution networks across the world over the last 30 years. The idea of developing covered conductor overhead lines is to minimize the deficiencies of traditional bare conductors (BC). In Nordic countries, covered conductors were initially introduced for easy power distribution through forest areas and for better operating safety [1]. The main benefits of these conductors are their ability to allow grounded objects to touch the line and the phases to touch each other without any tripping or outage. However, one common problem with covered conductors is lightning overvoltage. This overvoltage is in two forms: overvoltage produced from direct lightning strokes to a power line / line support and overvoltage induced from lightning strokes to ground. Unless effective protection systems are taken into use, a lightning surge can lead to flashover of a line insulator and then to multiple breakdowns of the solid conductor insulation. The assessment of the performance of a covered conductor from direct lightning strokes to a line support is the main purpose of this study. Lightning overvoltages were simulated with ATP/EMTP (Alternative Transient Program/Electro Magnetic Transient Program).

Lightning Peak Currents (kA )

Fig. 1. Cumulative distribution of lightning peak current in Finland according to lightning data collected from the Finnish Meteorological Institute (FMI). [4]

CIRED2009 Session 2

Paper No 0652

CIRED

20th International Conference on Electricity Distribution

Prague, 8-11 June 2009 Paper 0652

SIMULATING LIGHTNING COVERED CONDUCTOR Test system

EFFECT

ON

f (t ) = Amplitude[et e t ]
16 14 12 Max. time deriv ativ e = 3 kA/us Max. time deriv ativ e = 30 kA/us

(1)

Fig. 2 gives a schematic of the MV network simulated in this study. The network consists of a single phase infinite source (20/ 3 kV), a bare conductor of length 1000m, a covered conductor with length 200m, and a 20/0.4 kV substation. The ATP/EMTP simulation (single-phase) of the network is shown in Fig. 3. The network source was simulated with a single phase AC source and the bare conductor and covered conductor were simulated with an LCC model for both overhead line and covered conductor types respectively. The substation switch was replaced with a capacitance C = 4 nF [5]. The grounding resistance was taken as 10 [3].

Current (kA)

10 8 6 4 2 0 0

10

20

30

40 50 T ime (us.)

60

70

80

90

Fig. 4. Waveforms of the lightning stroke currents

Surge arrester characteristics

The V-I characteristics of metal oxide surge arresters are a function of the incoming surge steepness [4]. In this study, MOV arresters were employed for the protection of lightning surges at various locations on the test system. Obtained from the manufacturer's datasheet [6], the V-I characteristic curve of the arrester used for the simulation is shown in Fig. 5. The energy rating of this 30kV class of MOV arrester is 74.8 kJ, 3.4 kJ/kV, where Uc= 22 kV.
91.0 U [kV]

83.3

75.5

Fig. 2. A schematic of part of a Finnish 20kV MV network having a bare conductor, a covered conductor and a substation in series
Source
LCC LCC LCC

67.8

Bare conductor
LCC LCC LCC LCC LCC LCC LCC

Covered Conductor
LCC
V

Substation Switch
V

60.1 1.5 11.1 20.7 30.4

I [kA] 40.0

Csb Rg

Fig. 5. V-I characteristic curve of the 30kV MOV arrester (8/20 s) (V: residual voltage and I: discharge current)

Rg

RESULTS AND DISCUSSIONS

Fig. 3. A single-phase ATPDraw model of a 20kV MV network, consisting of a bare conductor in series with a covered conductor and substation. Rg is the grounding resistance and Csb is a substation switch represented with a capacitance.

Lightning stroke
In this study, the resulting overvoltages were simulated for a mean lightning stroke current of peak value 15.2 kA (Fig. 1) with a maximum rate of rise of current of (i) 30 kA/s and (ii) 3 kA/s. Fig. 4 gives the waveforms of the lightning currents. In the ATP/ EMTP simulation program, an ideal source was used for the lightning stroke modeling. It was modeled with the Heidler Type 15 ideal source in ATPDraw characterized by two exponentials with amplitude and constant and , in accordance with the ATP rule book, with the function:

Induced voltages were computed for a 20 kV network of Fig. 2, consisting of a bare conductor in series with the covered conductor and the substation. The lightning stroke was terminated on the line support where the bare conductor and the covered conductor meet. The induced voltages were produced on the bare conductor, the covered conductor and the substation terminals. Figs. 6 and 7 give the induced voltages on the conductor and substation with two different rates of rise of the lightning current (i) 30 kA/s and (ii) 3 kA/s respectively. It can be observed from the two figures that lightning to a power line support may produce an induced voltage capable of damaging the insulation of the covered conductor in several places (e.g. at the cable heads where the characteristic impedance of the bared conductor and the covered conductor changes) and the transformer windings. The resulting voltages are far higher than the Basic

CIRED2009 Session 2

Paper No 0652

CIRED

20th International Conference on Electricity Distribution

Prague, 8-11 June 2009 Paper 0652

Insulation Levels (BILs) of the conductor insulating layer (90 kV) and the substation transformer (150 kV). It was also seen from the figures that for an increase in the maximum rate of rise by a factor of 10, induced voltages in the conductors and substation produce more distortion, which can be damaging to any protective device along the feeder. For the same factor, the increase in the maximum induced voltage at the substation is marginal. Since the bare conductor and covered conductor have different characteristic impedances, an attempt was made to study its influence on the induced voltage on the feeder. Fig. 8 gives the induced voltage on the bare conductor and the covered conductor 100 m away from the lightning strike point (line support). The lightning stroke was simulated with the maximum time derivative, 30 kA/us. From the figure, it can be seen that the induced voltage on the BC undergoes higher distortion than the induced voltage on the CC. There is also a negative pre-pulse and further negative pulses in the induced voltage waveform on the BC. The difference in the maximum induced voltage on the bare conductor (240 kV) and the covered conductor (200 kV) is marginal for the kind of lightning strike simulated in this study. Thus, any power line with a covered conductor is prone to breakdown from a close-range lightning strike, unless it is well protected with surge protective devices, such as surge arresters, arc gaps or other devices.
250 Vc (30 kA/us) Vc (3 kA/us) 200

250 Induced v oltage on the CC Induced v oltage on the BC 200

150 Voltage (kV)

100

50

-50 0

10

20

30

40 50 T ime (us)

60

70

80

90

Fig. 8. Induced voltage on the covered conductor (CC) and the bare conductor (BC) 100m from the strike point (maximum time derivative, 30 kA/us)

C onductor Voltage (kV)

150

100

50

In order to assess the protective capability of an MOV for a lightning current simulated in this study, an MOV arrester was added on every pole of the bare conductor (the line span was taken as 100m for 20 kV). Fig. 9 gives the ATPDraw model of the line with surge arresters. Shown in Figs. 10 and 11 are the remaining voltage at the conductor head and the terminal of the substation respectively, for the two different rates of rise of the lightning currents. All simulation results show that effective reduction in the surge voltage is possible with adequate protection, i.e. installation of arresters at every pole leading to a covered conductor head will prevent multiple breakdowns of the covered conductor and other equipment along the power line. From Figs. 10 and 11, it can be observed that if there is an increase in the rate of rise of lightning currents, the peak value of the remaining induced voltages increases marginally with little distortion, and hence it does not have a great influence on the protective capability of surge arresters.

-50 0

10

20

30

40 50 Time (us)

60

70

80

90

Source
LCC LCC LCC

Bare conductor
LCC LCC LCC LCC LCC LCC LCC

Covered Conductor
LCC
V

Substation Switch
V

Fig. 6. Induced voltage on covered conductor head for different maximum rates of rise of lightning current
250 Vs (30kA/us) Vs (3kA/us) 200
Rg

PE

Substation Voltage (kV)

150

Fig. 9. ATPDraw model of the power line configuration with arrester at every pole.

MOV

MOV

MOV

MOV
H

MOV

MOV

MOV

MOV

MOV

MOV

Csb Rg
I
H

Rs

Lightning source

Rs

100

50

-50 0

10

20

30

40 50 Time (us)

60

70

80

90

Fig. 7. Induced voltage on substation for different maximum rates of rise of lightning current

Fig. 12 gives the energy absorbed by the arrester at the closest range to the lightning current magnitude. It can be seen that the simulated lightning stroke is never a threat to MOV arresters if they are installed on every pole. However, an arrester can respond differently to a lightning magnitude with different rates of rise of the lightning current, as shown in the figure. Surprisingly, a lightning stroke current of 15.2 kA with 3 kA/ us is more stressful for the surge arrester than with 30 kA/ us. Thus decrease in the rate of rise of lightning

CIRED2009 Session 2

Paper No 0652

CIRED

20th International Conference on Electricity Distribution

Prague, 8-11 June 2009 Paper 0652

current by a factor of 10 increases the MOV energy absorption by 1.15.

70 60 Remaining Conductor Voltage (kV) 50 40 30 20 10 0 0 10 20 30 40 50 Time (us) 60 70 80 90 Vc (30kA/us) Vc (3 kA/us)

CONCLUSION
The induced voltages due to lightning strokes the line support of a power line consisting of a bare conductor, covered conductor and a substation have been simulated using a digital program. The following are the main remarks from the study: Change in lightning characteristic (rate of rise of lightning current) does not lead to a significant change in the maximum induced voltage for a lightning stroke to a line. But the induced voltage distortion increases with an increase in the rate of rise of the lightning current. For the lightning stroke that was simulated, the induced voltage on the bare conductor is higher than the induced voltage on the covered conductor. Negative pulses emerge in the induced voltage waveform on the bare conductor. This can be very stressful for surge protective devices. The arresters energy dissipation increases as the rate of rise of lightning current decreases. Effective protection of a power line against induced voltage from close-range lightning strokes can be achieved by installing arresters on every pole, for a 100 m line span. This will offer good protection against any slow or fast rising surges around the power line.

Fig. 10. Remaining induced voltage on covered conductor for different maximum rates of rise of lightning current

80 70 Remaining Substation Voltage (kV) 60 50 40 30 20 10 0 0 10 20 30 40 50 Time (us) 60 70 80 90

REFERENCES
Vs (30 kA/us) Vs (3 kA/us)

Fig. 11. Remaining induced voltage on substation for different maximum rates of rise of lightning current
4.5 4 Arrester Energy Dissipated (kJ) 3.5 3 2.5 2 1.5 1 0.5 0 0 10 20 30 40 50 60 Time (us) 70 80 90 100 E (30 kA/us) E (3 kA/us)

Fig. 12. Energy absorbed by MOV arrester installed on the connecting point of the bare conductor and covered conductor for different maximum rates of rise of lightning current.

[1] G. M. Hashmi, 2008, Partial discharge detection for condition monitoring of covered-conductor overhead distribution network using rogowski coil, Published doctoral thesis dissertation, TKK, Finland [2] J. Tuomi and M. Antti,. 2006, Lightning observations in Finland. Report of Finnish Metrological Institute, Helsinki. No. 6. [3] IEEE working group on lightning performance of Electric Power Overhead Line. 2004. IEEE Guide for improving the Lightning Performance of Electric Power Overhead Lines. IEEE Std 1410tm- 2004, NY [4] M.A. Omidiora, M. Lehtonen, 2008. An Approach to the Lightning Overvoltage Protection of Medium Voltage Lines in Severe Lightning Areas, Current Themes in Engineering Technologies, American Institute of Physics, CP1007, USA, pp. 140-151. [5] W. Skomudek, 2004, computer analysis of overvoltage hazard due to lightning discharges in medium voltage overhead lines with covered conductors, Journal of Electrical Engineering, Vol. 55, No. 5-6, 2004, 161164. [6] Ultrasil housed varigap arresters, April. 2003, Cooper Power Systems Electrical Apparatus. I235-39.

CIRED2009 Session 2

Paper No 0652