IEEE Web Hosting - pres n rC.A. Nucci Dept. of Electrical Engineering University of Bologna 40136...

C.A. Nucci

Dept. of Electrical Engineering

University of Bologna

40136 Bologna, [email protected]

F. Rachidi

Swiss Federal Inst. of Technology

Power Systems Laboratory

1015 Lausanne, [email protected]

IEEE ‘99 T&D ConferencePanel session "Distribution Lightning Protection".

New Orleans, April 14, 1999

Lightning-Induced Voltages

1. Aim of Presentation

To give some answers to the questions that are

most commonly raised by engineers and

scientific researchers dealing with the problem of

protection against lightning-induced voltages.

2. What causes induced voltages ?

Electromagnetic coupling between the field radiatedby a lightning stroke and the line conductors

Incident fieldScattered field

TOTAL FIELD

2. What causes induced voltages ? Cont.

Essentially the return-stroke phase is responsible ofthe induced voltage

However, when lightning strikes the ground nearbythe line at close distance from the line, also thepreceding leader phase can results in a significantinduced voltage

Return-Stroke Current

i (0,t) RSC i (z,t)

Lightning ElectroMagnetic Pulse (appr. expr. for E horiz.)

i (z,t) LEMP E, B

ElectroMagnetic Coupling

E, B EMC V, I

3. How to evaluate them?

A review of the various return-stroke models has beenrecently made by Rakov and Uman onIEEE EMC Transactions, Special Issue on Lightning,1998 where they have discussed, among others, thefollowing ‘engineering’ models

• Bruce-Golde (BG)

• Transmission Line (TL) Uman, McLain, Krider

• Traveling Current Source (TCS) Heidler

• Modified Transm. Line - Linear (MTLL) Rakov and Dulzon

• Modified Transm. Line - Exponential (MTLE) Nucci et al.

• Diendorfer-Uman (DU)

Return-stroke current models

3. How to evaluate them? Cont.

Experimental validation

Given a channel-base current ==>the RSC model must reproduce the corresponding Electromagnetic field

For Natural lightning:

PROBLEM: practically no existing data sets ofsimultaneously measured current and fields

Data of this kind have been collected using

the Triggered lightning technique



nn TRIGGERED LIGHTNING:TRIGGERED LIGHTNING:Lightning is artificiallyLightning is artificiallyinitiated firing small rocketsinitiated firing small rocketstrailing grounded wirestrailing grounded wiresupwardupward a a few hundred few hundredmetersmeters under underthunderstormsthunderstorms..



Triggered lightning: A sequence of frames



TCS

a) b)

microseconds microseconds

V/m TCS MTL

Validation by means of triggered lightning



Basically, three coupling models have been used:

• Rusck [1958]

• Chowdhuri [1969]

• Agrawal et al. [1980]

Of the three models only the Agrawal one is‘rigorous’ for a general external field excitation

However, for a lightning channel perpendicular tothe ground plane ===> Rusck = Agrawal

Electromagnetic coupling


Reduced scale model at the University Of São Paulo - Brazil

The Agrawal model: Experimental validation


0 2 4 6 8 10 12

Time in µµs

0

20

40

60

80

100

120

140

CALCULATEDMEASURED

Using reduced-scale line model

Experimental data: by A. Piantini, Univ. Of São Paulo

70 m

130 m 40 m 100 m

OBSERVATION POINT



0 0.5 1 1.5 2 2.5 3 3.5 4Time in us

-60

-40

-20

0

20

40

60

80

100

120

140

160Calculated dV/dt

Measured dV/dtMeasured dV/dt (filtered)


Using reduced-scale line model

Experimental data: by A. Piantini, Univ. Of São Paulo




Using NEMP simulators

Which component of the LEMP does affectmost the induced voltages?

Vertical E component?

Horizontal E component?

Other components?

Let us assume, for simplicity, a lossless line


0

u (x)

i(x) L'dx

x x+dx

+-

i(x+dx)

+

-

+

-

L

us (x+dx)-u i (0)

R 0

-u i (L)

R L

u i(x)

iE x dx

s C'dx

Agrawal et al.


-20

0

20

40

60

80

0 2 4 6 8

Time (µs)

E contribution

E contribution

Total

X

Z

0

u (x)

i(x) L'dx

x x+dx

+-i(x+dx)

+

-

+

-

L

us(x+dx)-ui(0,t)R0

-u i(L)RL

ui(x)

iEx dx

s C'dx


∫∫ +−=h

iz

ix

hiy dztzxE

xthxEdztzxB

t00

),,(),,(),,(∂∂

∂∂

Nucci and Rachidi, IEEE Trans. on EMC,Vol. 37, No. 4, November 1995.

z

xy

E i, B i


0

u (x)

i(x) L'dx

x x+dx

+-

i(x+dx)

L

u (x+dx)R 0

R L

i(B y (x,z) dz)dx

C'dx

⌠⌡

ddt

i(E z (x,z) dz)dx⌠ddt⌡

-C’

Taylor et al.


-60

-40

-20

0

20

40

60

80

B contribution

E contribution

Total

z

y

0

u (x)

i(x) L'dx

x x+dx

+-i(x+dx)

L

u(x+dx)R0

RL

i(B y(x,z) dz)dx

C'dx

⌠⌡

ddt

i(E z(x,z) dz)dx⌠ddt⌡

-C’


0

iB y (x,L)⌠⌡

1L’

-

h

dz

0

u (x)

i(x) L'dx

0R L

C'dxiB x(x,z)⌠⌡

∂∂ y

1L’

0

-

h

dz[ ]

0

iB y (x,0)⌠⌡

1L’

-

h

dz

Rachidi


-20

0

20

40

60

80

0 2 4 6 8Time (µs)

B contribution

B contribution

Total

X

y

0

iB y(x,0)⌠

⌡1L’

-

h

dz

0

u (x)

i(x) L'dx

x x+dx

i(x+dx)

L

u (x+dx)R0 RL

C'dx

iB y(x,0)⌠

⌡1L’

-

h

dz

iBx (x,z)⌠

⌡∂

∂ y1L’

0

-

h

dz[ ] dx

3. How to evaluate them? Cont

The contribution of a given electromagnetic fieldcomponent in the coupling mechanism dependsstrongly on the used model.

Thus, when speaking about the contribution of agiven electromagnetic field component to theinduced voltages, one has to specify the coupling

.

3. ? Cont.

Induced voltage magnitude and shape significantlydepend on

• lightning return stroke parameters (channel-basecurrent parameters, return stroke velocity),

• distance and relative position with respect to thetransmission line,

• line configuration and terminations.

Induced overvoltages can reach magnitudes up to fewhundreds of kV and can therefore cause line flashover.

4. are for lightning induced

1 2 3 4 5 6 7

12I in kA

Time in us

10

8

6

4

2

0

-20-10

10

203040

5060

0 1 2 3 4 5 6 7Time in us

UA in kV

1 km

Shape

4. What magnitudes and shape are typicalfor lightning-induced overvoltages ? Cont.

However ...

1 2 3 4 5 6 7

12I in kA

Time in us

10

8

6

4

2

0

-20-10

102030405060

0 1 2 3 4 5 6 7Time in us

UA in kV


Magnitude

1 km


However ...

For perfectly conducting grounds and for an

infinitely long wire


the simplified Rusck formula allows for asatisfactory estimation

d

hIZU max

max 0=

Ω== 30/4/1 00 oZ εµπwhere

HOWEVER …

5.How far away can lightning strokes be thatcause an induced voltage flashover ?

Generally within 200 m

However it depends on many parameters(=> computer code)

Lightning strokes occurring beyond a fewhundred meters from the line can cause a lineflashover for poor conducting soils (De LaRosa et al., IEEE Trans. on PWDR, 1988)

6.How does the induced voltage drop as afunction of distance from the line ?

-20

-10

10

20

30

40

50

60

70

80

90

0 1 2 3 4 5 6 7

Time in us

U A in kV

30m

60m

90m

Perfectly conducting ground and stroke location equidistant to theline termination => nearly proportionally to 1/dHOWEVER ...

1 km

d

7.What BIL is needed to prevent inducedflashovers ?

50 100 150 200 250 300Basic Insulation Level [kV]

0.01

0.10

1.00

10.00

100.00

1000.00

No.

of e

vent

s ha

ving

am

pli tu

des

exce

edin

gth

e B

IL /

( 100

km

· ye

a r)

perfect ground

ground conductivity = 0.01 S/m

ground conductivity = 0.001 S/m

8.What arrester spacing is needed to preventflashover?

In order to prevent direct-stroke flashover, arresterspacings of 300-400 m is generally recommended.

For the case of induced flashovers, a givenconfiguration of line arresters can result in differentperformances depending on the location of lightningstrike [29].

Further studies are needed in this respect.


X0 X1 X2 X3 X4 X50

20

40

60

80

100

120

Transf.Transf. + Surge arr.Matched

Stroke location: B1

a)


X0 X1 X2 X3 X4 X50

50

100

150

200

250

300

350

400


Stroke location: E1

b)


X0 X1 X2 X3 X4 X5

Observation Point Along the Line

0

50

100

150

200

250

300

350

400


Stroke location: H1

c)

a)

1

2

3

10 m

3.7 m

3.7 m

3.7 m

4

1 2 3

3.7 m 3.7 m

10 m

4 5

3.7 m

3.7 m

b)

9.Will a shield wire help ?

9.Will a shield wire help ? Cont.

Shield wires help in reducing the magnitude ofinduced voltages by a factor of about 20 to 40 %.

This implies about the same reduction of the faultfrequency.

wireProtective

RatioVerticalConfig.

HorizontalConfig.

1

2

3

PR1PR2PR3

0.81

0.78

0.72

0.67

0.60

0.67

10. Is horizontal or vertical construction best ?

a)

1

2

3

10 m

3.7 m

3.7 m

3.7 m

4

1 2 3

3.7 m 3.7 m

10 m

4 5

3.7 m

3.7 m

b)

10. Is horizontal or vertical construction best ? Cont.

a)

VoltageRatio

VerticalConfiguration

HorizontalConfiguration

V1/V(h1)

V2/V(h2)

V3/V(h3)

0.75

0.79

0.89

0.85

0.81

0.85

Ratio between peak values of the induced voltages on a lineconductor Vi and those corresponding

to a single-conductor line of the same height V(hi).

The induced voltage magnitude for typical distributionlines is virtually proportional to the line height.

As a consequence, an important factor determining themagnitude of lightning-induced voltage is the lineheight above ground, rather than the type ofconstruction.

In general, a construction allowing a shorter height forthe conductors is epected to experience lower inducedovervoltages.

10. Is horizontal or vertical construction best ? Cont.

11. What effect does pole grounding andground resistivity have ?

Pole grounding affects the performance of the groundwire in reducing the induced overvoltages.

In general, lower the pole ground impedance, betterthe performance of the ground wire.

Influence of pole grounding

The ground resistivity affects:

1. electromagnetic field

2. propagation of the surges

Influence of ground resistivity

11.What effect does pole grounding andground resistivity have ? Cont.

Lossy ground

Perfect ground

microseconds 0.5 km 0.5 km

Lossy ground:

σ = 0.001 S/m

Influence of ground resistivity Cont.


b)

Ideal field

Ideal line

Lossy ground

Lossy ground(σ = 0.001 S/m)

Ideal field: ground resisitivity only in the expression of the ground impedance

Ideal line: ground resistivity only in the expression of the incident field



Lossy ground

Ideal line

Ideal field

microseconds 2.5 km 2.5 km

Lossy ground(σ = 0.001 S/m)

Ideal field: ground resisitivity only in the expression of the ground impedance

Ideal line: ground resistivity only in the expression of the incident field



How does the ground resistivity affectmagnitude and shape of the inducedvoltages?

Influence of ground resistivity


0.5 km 0.5 km


0 200 400 600 800 1,000

Obs. point along the line in m

30

50

70

90

110

σ infinite

σ =0.01 S/m


1 km


0 200 400 600 800 1000Obs. point along the line in m

-40

-20

0

20

40

σ infinite

σ = 0.01 S/m


The ground resistivity can increase or

decrease the magnitude of the induced

voltages depending on the stroke location

and the observation point along the line

=> the calculation is not trivial



- Lightning electromagnetic field characterizationusing natural and artificially-initiated lightning

- Experimental validation of field-to-transmission linecoupling models

- Development of engineering tools for the protectionof power networks against lightning-inducedovervoltages

- Leader-induction effect

- Effect of ground conductivity on lightning-inducedovervoltages

12.Some of the current researches which are being done on induced voltages

The LIOV-EMTP code

The LIOVLIOV code calculates:

• LEMP (using the MTL model and Cooray-Rubinstein expr.)

• Coupling using the Agrawal model.

The EMTPEMTP :

• calculates the boundary conditions• makes available a large library of power components

Γ0

LIOV-line

n-port

OverheadDistribution Line

The LIOV-EMTP code Cont’d

LIOV has been developped within the framework of an international collaboration involving

• University of Bologna• Swiss federal Institute of Technology (EPFL, Lausanne)• University of Roma ‘La Sapienza’

Its link with EMTP has been realized in collaboration with ENEL-CESI (Univ. Bologna)

Other methods have been proposed EdF (EPFL )

0

u (x,t)

i(x,t) L'dx

C'dx

x x+dx

+-i(x+dx,t)

u (x+dx,t)

u i (x,t)

E ix

(x,h,t)dx

ss

+

-

Γ0

i0 u 0

-u i(0,t)

u u u u E x t dzt s i szi

h= + = − ∫ ( , )

0

i2i1Γ

i0

u 1 u 2

linea Bergeron linea LIOV

+

-

0

u 1'i0 '

-u i(0,t)

Bergeronline

LIOV line

i 0Z

+

-

u 1

V

+

-

+

-

Γi2i1

u 2

V11'

0 Z cc

u1 '

'i0

-u i (0,t)

Link between LIOV and EMTP


∂∂

∂∂

u x t

xL

i x t

tE x h t

s

xi( , ) ( , )( , , )+ ′ =

∂∂

∂∂

i x t

xC

u x t

t

s( , ) ( , )+ ′ = 0

( )u t u t i t E z t dzso z

ih

( , ) ( ) ( , ) ( , , )0 0 010

= = − + ∫Γ

[ ]u t Z i t u t t Z i t t

Z i t V t tc c

c

1 0 1 0

0 1

( ) ( ) ( ) ( )

( ) ( )' '= − + − + −

= − + −

∆ ∆

∆

[ ]u t Z i t u t t Z i t t

Z i t V t tc c

c

1 0 1 0

0 1

' '

' '

( ) ( ) ( ) ( )

( ) ( )

= + − − −

= + −

∆ ∆

∆


Validation using data from a more complex systemData: courtesy of Dr. A. Piantini, Univ. Saõ Paulo

12.Some of the current researches which are being done on induced voltages Cont.

0 1 2 3 4 5 6 7 8

Time in us

-100

-50

0

50

100

150

200 Calculated voltageMeasured voltage

Validation using data from a more complex systemData: courtesy of Dr. A. Piantini, Univ. Of Saõ Paulo


0 1 2 3 4 5 6 7 8

Time in us

0

10

20

30

40

50

60 calculated voltageMeasured voltage

Validation using data from a more complex systemData: courtesy of Dr. A. Piantini, Univ. Of Saõ Paulo


- Return stroke modeling and influence of elevatedstrike objects on lightning current and radiated fields

- Adequacy of the available lightning return strokecurrent statistical data

- Effect of corona on lightning-induced voltages

- Lightning detection and location systems

- Lightning channel tortuosity and inclination


13.What is the CIGRE working group doing oninduced voltages ?

Within the framework ofCIGRE working group WG 33.01 "Lightning",

Task Force 33.01.01 "Lightning induced voltages"established some years ago.

C.A. Nucci (responsible member), P. Chowdhuri, G. V.Cooray, M.T. Correia de Barros, M. Darveniza, F. De la Rosa,G. Diendorfer, F. Heidler, M. Ishii, W. Janischewskyj, T.Kawamura, C. Mazzetti, P. Pettersson, F. Rachidi, V. Rakov,M. Rubinstein, T. Short, J.V. Shostak, M.A. Uman, S.Yokoyama

13.What is the CIGRE working group doing oninduced voltages ?

TF 33.01.01 has already produced two paperspublished in Electra dealing respectively withlightning return stroke models (August 95) andlightning electromagnetic field-to-transmission linecoupling models (October 95).

A third paper, dealing with a sensitivity analysis andaimed at providing ranges of overvoltage values tobe expected in the different typical lineconfigurations, is in preparation.

IEEE Web Hosting - pres n rC.A. Nucci Dept. of Electrical Engineering University of Bologna 40136...

Documents

Transcript of IEEE Web Hosting - pres n rC.A. Nucci Dept. of Electrical Engineering University of Bologna 40136...