spastyle.doc - IEEE-SA - Working Groupgrouper.ieee.org/groups/railtransit/ocs/1627_wg/P1627D4.1... · Web view[B13] Roger C. Dugan, Mark F. McGranaghan, H. Wayne Beaty, “Electrical

August 28, 2009 IEEE P1627/D4.1

Draft Standard for Grounding Practices for DC Electrification Overhead Contact Systems, including Application of Lightning Arresters for Transit Systems

Prepared by Working Group 17 of the Overhead Contact System SubcommitteeSponsored by the Rail Vehicle Transit Interface Standards Committeeof the IEEE Vehicular Technology Society

Copyright © 2008 by the Institute of Electrical and Electronics Engineers, Inc.Three Park AvenueNew York, New York 10016-5997, USAAll rights reserved.

This document is an unapproved draft of a proposed IEEE Standard. As such, this document is subject to change. USE AT YOUR OWN RISK! Because this is an unapproved draft, this document must not be utilized for any conformance/compliance purposes. Permission is hereby granted for IEEE Standards Committee participants to reproduce this document for purposes of IEEE standardization activities only. Prior to submitting this document to another standards development organization for standardization activities, permission must first be obtained from the Manager, Standards Intellectual Property, and IEEE Standards Activities Department. Other entities seeking permission to reproduce this document, in whole or in part, must obtain permission from the Manager, Standards Intellectual Property, IEEE Standards Activities Department.

IEEE Standards Activities DepartmentManager, Standards Intellectual Property445 Hoes Lane, P.O. Box 1331Piscataway, NJ 08855-1331, USA

Abstract

This standard is a basis for the design and application of dc surge protection devices to protect Overhead Contact System (OCS) from transient overvoltages associated with lightning and switching surges. Switching surges are inherent characteristics of electrification system and are generally considered low energy transient overvoltages. Lightning surges are natural, caused by electrical discharge that occurs in the atmosphere between clouds or between clouds and ground. Lightning surges are known for causing very high energy transient overvoltages by direct or indirect coupling with OCS. This standard covers transient overvoltage protection of OCS used in dc transit rail electrification systems. Transient overvoltage protection of third rail dc and the running rails can be achieved by the application of same type of surge protection devices as are recommended for the OCS, although these rails are less likely to be affected by lightning transient overvoltages due to their proximity to earth where flash-over can occur to drain surge energy away from the rails without the application of surge overvoltage protection devices. The word surge arrester instead of lightning arrester has been used in this standard without affecting the technical contents of the standard.

Keywords

i

12

1

2

3

4

567

89

1011

12131415161718192021

22232425

26

2728293031323334353637383940

41

3

{INSERT DATE} P<designation>D<number>

Surge arrester, Lightning, Overvoltage, Switching surges.

Copyright © 2007 IEEE. All rights reserved.This is an unapproved IEEE Standards Draft, subject to change. ii

1

1

23

August 28, 2009 IEEE P1627/D1.3

IntroductionThe majority of the present operating dc electrified rail systems use OCS or third rail to supply power to the vehicles. The probability of lightning surge hitting an OCS, or third rail depends on its geometry, its own height, and length, and its relative location with respect to the presence of buildings, trees, towers etc. in its vicinity, and the lightning flash-to-ground density (number of lightning-to-ground strokes per square kilometer per year) of the area [B23][B24].

If after the lightning risk assessment, the expected frequency of direct lightning to the OCS (N D) to be protected exceeds its tolerable frequency of lightning (NC), as established by applicable standards and codes (NFPA 780, IEC 62305), lightning protection system (LPS) should be designed and installed. In this case OCS poles should be equipped with appropriate surge protection devices grounded by use of low resistance ground rods and OCS poles should be grounded by their separate ground rods.

Should the calculated average annual number of direct lightning strokes (ND) to the OCS be below the permissible strokes (NC), no LPS (with the exception of application of dc transient surge protection devices of low energy capability ) is necessary and the OCS is defined as self protected by IEC 62305 and NFPA780.

Arbitrary installation of direct stroke diverters such as lightning rods and ground wires above the OCS changes the geometry and may increase its chances of more lightning exposure. Based upon such reasoning it appears that even in the area of higher lightning strokes, the application of ground wires and ground rods above the OCS should not be used. Then this standard’s approach is to develop an OCS transient overvoltage protection that focuses on the basic approach of selection and application of dc surge arresters, their grounding configuration, and grounding of OCS metallic poles. This consists of combination of the following:

Selection of appropriately rated dc surge protection (lightning arrester) devices to protect OCS from transient overvoltages caused by switching and lightning phenomenon.

Grounding of surge arresters by individually dedicated shortest possible grounding conductor connected to ground rods.

Use of dedicated grounding of OCS poles by application of ground rods separate from surge arrester ground rods to keep the structure potential below insulation breakdown level of positive and negative feeder cables and to minimize damage to the vehicle.

Use of an additional surge arrester to protect the traction power substation equipment as necessary.

Use of OCS basic insulation levels to minimize outages due to switching and lightning surges.

Setting basic insulation levels for the OCS is in the scope of work of another subcommittee and thus reference [B22] on IEEE Draft P1626 standard and [B26] are included in this standard. DC electrification OCS and associated structure should not be compared to electric utility transmission lines as the physical configuration and location of the two are quite different whereby concept of lightning protection by use of ground/shield wire may not be applicable to dc electrification OCS. However, in case of single phase ac electrification OCS, an overhead ground wire which carries return traction ac current also performs the function of shield wire.

Prior to and during the development of this standard, there were reports of surge arrester failures on several transit systems resulting in service interruption and equipment failures. Questions were being raised regarding the grounding of OCS support poles, and the pros and cons of using

Copyright © 2007 IEEE. All rights reserved.This is an unapproved IEEE Standards Draft, subject to change. iii

1

123456

789

101112

13141516

17181920212223

2425

2627

282930

3132

3334

35363738394041

424344

23

August 28, 2009 IEEE P1627/D1.3

surge arresters in the OCS. Most importantly, there was no understanding of the cause of failures of surge arrester and their proper application to dc electrification OCS. Although, so far no personnel injuries have been reported involving failure of dc surge arresters at any of the operating transit properties in USA.

To establish lightning protection design measures, the derivation of lightning intensity and lightning stroke surge energy is established based upon the typical available lightning data. There are equal chances of a lightning strike hitting any of the system components due to their proximity, and thus flashover is certain if lightning strikes the OCS components since poles and OCS supporting members appear to be practically grounded for the surge voltage. Flashover appears to drain large amounts of lightning surge energy to ground, with the remainder of the surge energy relieved by the dc surge arresters placed at appropriate locations.

When considering lightning protection for OCS, numerous questions arise, such as:

a) Will dc surge arrester handle surge energy if a lightning flash (strike) directly hits the OCS wire near the arrester location?

b) If a lightning flash directly hits the OCS wire between two traction power substations, what will be the energy discharged through the surge arresters at feeder poles?

c) Is there a need to apply dc surge arresters at the mid point of two adjacent substations to enhance the lightning protection of the dc rail transit OCS?

d) Do we need to apply surge arresters at the connection points of underground dc supplementary feeder cable to OCS contact wire located approximately every 400 feet, where overhead messenger wire design is not possible in downtown areas due to aesthetics and other restrictions of height of the messenger wires?

e) Should there be shield/ground wire above the messenger to enhance lightning protection?

Responses to the above questions are discussed in this standard.

IEEE P1627 “Standard for Transient Overvoltage Protection of dc Electrification Overhead Contact System”, including application of dc surge protection devices was prepared by Working Group 17 of the Overhead Contact Systems Sub-Committee of the Rail Transit Vehicle Interface Standards Committee of the IEEE Vehicular Technology Society.

Origin and Development of IEEE P1627The Overhead Contact Systems Sub-Committee was formed in 2001 with the purpose of developing standards governing the design, construction, and maintenance of the OCS and current collection system.

Working group 17 was established to develop standards governing the transient overvoltage protection of OCS for the dc rail transit systems. The primary concern of the working group was a lack of uniform practices and a lack of understanding of the proper application of dc surge arresters and their grounding configuration to protect OCS and associated poles including dc traction power system and vehicles. Precise information of the transient environment, expected magnitude, duration and frequency of transient surges is very unpredictable and there has been no recorded data available to guide application engineers in the selection and application of dc surge protection devices. However, it is known that certain degree of threat of both the internal switching surges and the external lightning surges exist that could occur at random causing damage to OCS and dc system without proper application of surge overvoltage protection devices.

It appears that the lack of knowledge of the transient environment and understanding the parameters and rating of dc surge arresters lead to use of guesswork in the selection and

Copyright © 2007 IEEE. All rights reserved.This is an unapproved IEEE Standards Draft, subject to change. iv

1

1234

56789

1011

12

1314

1516

1718

19202122

2324

25

26272829

30

313233

3435363738394041424344

4546

23

August 28, 2009 IEEE P1627/D1.3

application of such devices causing failures at some installations. At some transit properties, failures of dc lighting arresters and equipment damage associated with those failures have been reported. Lack of understanding of characteristics of transient overvoltages, dc surge arrester ratings, their proper installation and grounding configurations, lack of their clear test and application data from the manufacturers has been a major concern in the industry that lead to the development of this standard. Proper application of dc surge protection devices to divert surge energy associated with the switching and lightning surges away from the OCS relates directly to their effect on equipment and personnel protection.

Patents

Attention is called to the possibility that implementation of this standard may require use of subject matter covered by patent rights. By publication of this standard, no position is taken with respect to the existence or validity of any patent rights in connection therewith. The IEEE shall not be responsible for identifying patents for which a license may be required by an IEEE standard or for conducting inquiries into the legal validity or scope of those patents that are brought to its attention.

Participants

At the time this standard was completed, the working group had the following membership:

Dev Paul, Chair

Ramesh Dhingra, Vice Chair

Alan BlatchfordButch CampbellRon ClarkIan HayesAlbert HoeGordon MacDonaldDev Paul

Chris PagniSteve MitanJay SenderJeffrey N. SissonVish MawleyEdward RoweSuresh Shrimavle

Carl WesselPaul WhiteKelvin ZanEthan KimRamesh DhingraStuart Kuritzky

The following members of the balloting committee voted on this standard. Balloters may have voted for approval, disapproval, or abstention. (To be provided by IEEE editor at time of publication.)_______________________________________________________________________________

Copyright © 2007 IEEE. All rights reserved.This is an unapproved IEEE Standards Draft, subject to change. v

1

12345678

9

101112131415

16

17

18

19

2021222324252627

28293031323334

353637383940

41424344

23

August 28, 2009 IEEE P1627/D1.3

Contents

1. Overview........................................................................................................................... 1

1.1 Scope................................................................................................................... 1

1.2 Purpose................................................................................................................1

1.3 Format of Standard...............................................................................................1

2. References....................................................................................................................... 2

3. Definitions, abbreviations, and acronyms...........................................................................2

3.1 Definitions.............................................................................................................2

3.2 Abbreviations and Acronyms.................................................................................4

4. Transient Surges...............................................................................................................4

4.1 Lightning and Switching Surges.............................................................................5

4.2 Surge Characteristics - Propagation......................................................................6

4.3 Magnetic Stored Energy of Surge..........................................................................7

5. Surge Environment – DC Electrification System.................................................................8

5.1 DC Surge Protection Device Requirements...........................................................8

6. Lightning Stroke Terminology..........................................................................................17

6.1 Lightning Intensity Estimation..............................................................................18

7. Lightning Stroke to OCS system......................................................................................23

7.1 Underground Supplementary Cable Connections to OCS....................................24

8. Grounding...................................................................................................................... 27

8.1 OCS Pole Grounding..........................................................................................27

9. DC Surge Arresters........................................................................................................28

9.1 Application..........................................................................................................28

9.2 Surge arrester Rating..........................................................................................30

10. DC Surge Arrester Service Requirements........................................................................31

11. DC Surge Arrester Testing...............................................................................................31

11.1 Design Tests.......................................................................................................31

11.2 In Service (Field) Tests.......................................................................................32

Copyright © 2007 IEEE. All rights reserved.This is an unapproved IEEE Standards Draft, subject to change. vi

1

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

23

August 28, 2009 IEEE P1627/D1.3

11.3 Bibliography........................................................................................................33

Copyright © 2007 IEEE. All rights reserved.This is an unapproved IEEE Standards Draft, subject to change. vii

1

1

2

23

Draft Standard for DC Electrification Overhead Contact Systems, including Application of Lightning Arresters for Transit Systems

1. Overview

1.1 Scope

The scope of this standard covers application of transient overvoltage surge protection devices and their grounding configurations to protect OCS used in dc traction electrification for heavy rail, light rail, and trolleybus systems.

1.2 Purpose

The purpose of this standard is to establish minimum design requirements for application of dc surge protection devices and their grounding configurations to protect OCS and associated traction power system components from transient overvoltages caused by switching and lightning surges. Such a design will provide a reasonable degree of protection to equipment by diverting surge energy and related hazards away from the OCS system. At the present, there are no uniform practices for application of properly rated dc surge arresters and their grounding configurations to protect OCS used in dc traction electrification. The use of this standard is intended to provide understanding of the transient overvoltage environment that exists at a particular dc electrification system and then a systematic approach to protect OCS and associated dc traction power system from transient overvoltages by use of appropriately rated dc surge protection devices.

1.3 Format of Standard

First the basic characteristics of transient surges, their origin, and energy and propagation behavior are described in Sections 4 and 7. This is followed by a brief description of the surge environment expected at a dc electrification OCS included in Sections 5 and 7. In Sections 6 and 7, this standard addresses analysis of lightning strike to the dc traction power system components.

This standard addresses OCS, the running rails, OCS supporting structure, metallic poles, messenger wire, underground supplementary conductors, traction power substations, and vehicles. This also includes lightning waveform parameters.

This standard also addresses the mismatch between the actual number of lightning strikes and specificity of the lightning parameters required in performing lightning protection analysis for an electric transit system. At present there is no data for the lightning effects and design experience available from the operating transit properties. Experience from operational lightning location systems [B18] [B19] [B20] [B24], ongoing research, and scientific ability to measure lightning parameters can avoid guesswork in the lightning protection design of electric transit systems.

Copyright © 2007 IEEE. All rights reserved.This is an unapproved IEEE Standards Draft, subject to change. 1

1

1

2

3

4

5

678

9

1011121314151617181920

21

2223242526

272829

303132333435

23

From a cause and effect standpoint, the maximum rate-of-rise, the peak current, and the wave front rise-time are associated with determining the maximum voltage that will be seen on the OCS subjected to unpredictable threat of a direct or nearby ground lightning discharge. The probability of a lightning strike is greater when the electric transit system is located in a high isokeraunic level area. [B2].

Lightning may be of concern when the dc system is in a relatively high isokeraunic area. Isokeraunic area map of the world can be seen in reference [B9]. In addition, dealing with a transit system involving the general public and, more importantly, the havoc due to interruption of system operations caused by lightning is a design concern for dc transit systems. Thus to minimize the effect of lightning and switching surge voltages to a typical dc traction power system, application of appropriately rated dc surge arresters should be included in the design.

2. References

This standard shall be used in conjunction with the following publications. If the following standards are superseded by an approved revision or new version, the latest revision shall apply. In case of conflict between this standard and the referenced documents, this standard shall take precedence. Those provisions of the referenced standard that are not in conflict with this standard shall apply as referenced. (all referenced standards need to be referred in the text of the standard and they need to be in alpha numerical order)

National Electrical Safety Code (NESC)

National Electrical Code (NEC)

Canadian Electrical Code (CEC)

British Standard BSN EN 50124-1:2001 Railway Applications – Insulation coordination

IEEE Standard Dictionary of Electrical and Electronics Terms, IEEE Std 100

IEEE Guide for the Application of Metal Oxide Surge Arresters for Alternating Current Systems, ANSI/IEEE Standard C62.22,

IEEE Guide for Application of Gapped Silicon Carbide Surge Arresters for Alternating Current Systems, ANSI/IEEE Standard C62.2 – 1987

EN 50526-1 Draft Feb. 2009, “Railway applications – Fixed installations – D.C. Surge arresters and voltage limiting devices – Part 1: Surge arresters”

EN 50526-2 Draft March 17, 2009, ““Railway applications – Fixed installations – D.C. Surge arresters and voltage limiting devices – Part 2: Voltage limiting devices”

3. Definitions, abbreviations, and acronyms

3.1 Definitions

Review all IEEE Dictionary terms For the purpose Insulation Standard IEEE P1626/D1

IEEE Glossary of OCS


1

12345

6789

1011

12

131415161718

19

20

21

22

23

2425

2627

2829

3031

32

33

34

35

23

3.1.1 Clearance: Shortest distance in air between two conductive materials.

3.1.2 Creepage Distance: Shortest distance along the surface of the insulating material between two conductive materials.

3.1.3 Electrical Section: Part of an electrical circuit having its own voltage rating for insulation coordination.

3.1.4 Grounded: Electrical section intentionally connected to earth that cannot be interrupted.

3.1.5 Insulated: All components isolated from the energized OCS conductors by at least one level of insulation. An insulated section may be under the influence of adjacent energized circuits. An insulated section may be considered as an electrical section.

3.1.6 Surge arrester: A device typically mounted on OCS poles and connected to the OCS, designed to protect insulated feeder cables against lightning, by providing a path to ground through a spark-gap, with or without variable resistance elements.

3.1.7 Nominal voltage: Value assigned to a circuit or system approximately equivalent to the working voltage for designating the voltage class.

3.1.8 Overvoltage: Voltage having a peak value exceeding the maximum steady state voltage at normal operating conditions.

3.1.9 Rated Voltage: Value of voltage assigned to a component, device or equipment.

3.1.10 Rated Impulse Voltage: Value of voltage assigned to the equipment referring to the specified withstand capability of the insulation against transient overvoltages.

3.1.11 Rated Insulation Voltage: RMS withstand voltage assigned to the equipment referring to the specified permanent (over five minutes) withstand capability of the insulation between energized components and earth.

3.1.12 Residual Discharge Voltage (VIR): Is the voltage across the surge arrester at the instant of surge peak current discharge of a specific surge current wave shape due to lightning or switching phenomenon. Same magnitude of surge peak current with different wave-shape may result in different value of VIR depending upon the type and manufacturer of the surge arrester.

3.1.13 Maximum Continuous Operating Voltage (MCOV): The maximum designated root-mean-square (rms) continuous operating voltage value in dc volts that many be applied continuously between the terminals of the arrester.

3.1.14 Surge Arrester: See Surge arrester.

3.1.15 Isokeraunic level: Number of thunderstorm days per year in an area is called isokeraunic level

3.1.16 Overhead Contact System (OCS): Contact wire and messenger wire electrically in parallel connected to traction power substation (TPSS) by underground feeders. Contact wire makes contact with vehicle pantograph and is acting as positive polarity contact point for the dc power supply delivered to the vehicles from TPSS.

3.1.17 Lightning flash: Very bright light stream seen in the sky between the cloud and the ground on a stormy day.

3.1.18 Lightning stroke: Same as lightning flash.


1

1

23

45

6

789

101112

1314

1516

17

1819

202122

23242526

272829

30

3132

33343536

3738

39

23

3.1.19 Lightning strike: Same as lightning flash.

3.1.20 Lightning Intensity: Number of lightning strikes/per square meter/year in an area is called its lightning intensity.

3.1.21 Temporary Overvoltage (ETOV): The maximum root-mean-square (rms) temporary overvoltage value in dc volts that may occur at the OCS due to vehicle regeneration or utility power supply voltage regulation.

3.1.22 Supplementary Cable: Cable connected in parallel with OCS is called positive supplementary cable. Cable connected in parallel with running rail is called negative supplementary cable.

3.1.23 Voltage Margin-of-Protection (VSA): The voltage seen at the OCS (voltage between the OCS and the local ground) when the lightning surge arrester conducts lightning surge peak current and associated energy to ground.

3.2 Abbreviations and AcronymsAC Alternating CurrentANSI American National Standards InstituteAREMA American Railway Engineering and Maintenance of way AssociationASTM American Society for Testing and MaterialsAWG American Wire GaugeDC Direct CurrentETB Electrified Trolley BusFRA Federal Railroad AdministrationICLP International Conference on Lightning ProtectionIEEE Institute of Electrical and Electronics EngineersISO International Organization for StandardsLRV Light Rail VehicleNEC National Electrical Code (NFPA-70)NEMA National Electrical Manufacturers AssociationNESC National Electrical Safety CodeNETA National Electrical Testing AssociationNFPA National Fire Protection AssociationOCS Overhead Contact SystemOSHA Occupational Safety and Health Administration ActRMS Root Mean SquareROW Right-of-wayTES Traction Electrification SystemUBC Uniform Building CodeUL Underwriters Laboratories, Inc.USASI United States of America Standards InstituteUSDOT United States Department of Transportation

4. Transient Surges

Transient overvoltage surges are caused by lightning, switching phenomenon within the system, induced and impressed voltages and become superimposed over the dc power system voltage. They are unpredictable and could be brief and quick. Their wave-shape, magnitude, and energy content may vary considerably depending upon their cause of origin, system configuration, and surge impedance parameters. Such transient overvoltages and associated energy shall be diverted away from the system by application of appropriate dc surge protection devices at specific locations within the overall electrification system. Lightning surges, called external surges


1

1

23

456

789

101112

131415161718192021222324252627282930313233343536373839

40

41424344454647

23

can impinge on the electrification system directly or indirectly, and can induce transient overvoltages with considerable energy to create hazards. Switching surges, called internal surges, are caused by sudden changes in the electrification system, such as dc breaker operation, and may not be as severe as lightning due to their relatively low energy.

4.1 Lightning and Switching Surges

The standards make a distinction between switching and lightning surges on the basis of the duration of the front or rise-time from zero-to-peak value. Surges with fronts of up to 20 s are defined as lightning surges, and those with longer fronts are defined as switching surges [B15].

Lightning surge intensity depends upon the isokeraunic level of the area. The probability of direct stroke can be estimated by the various factors listed in NFPA 780. To understand the characteristics and nature of such surges, considerable published data is available [B1] [B2].

For lightning surge application purposes, industry has standardized voltage impulse waveform 1.2/50 s indicating crest is reached in 1.2 s and it decays to half the crest in 50 s. Similarly, a current impulse wave of 8/20s is used where the crest is reached in 8 s and decays half the crest value in 20s.

A steep-fronted surge is one with a rise time of 0.1-0.5 s and a virtual time to half value of around 5 s. An impulse current wave shape of 10/1000 s (long wave) is more representative of the high energy surges usually experienced from the inductive elements [B4] [B5].

4.1.1 Causes of Lightning Surges

Lighting is an electrical discharge that occurs in the atmosphere between clouds or between clouds and ground. It is very high-energy phenomenon and can be a source of harm for transit system. Lightning flashes, in fact, can release many hundreds of mega-joules of energy.

A typical cloud-to-cloud lightning begins with a preliminary breakdown due to the intense electric field initiated in the lower part of the cloud, and is generally negatively charged. The polarity of the lightning, in fact, which is a function of the local territory, can be statistically assumed as negative in 90% of the cases.

The process is followed by a discharge, or leader, which creates a highly conductive channel, which advances, in a “zigzag” path, towards the earth and meets an upward advancing leader. This stepped discharge is caused by the non-uniformity of electric field. The medium between clouds and earth, in fact, is not uniform, as the air characteristics (density, pollution, humidity, etc.) continuously vary, hence the non-linearity of the lightning path.

As the discharge progresses towards the earth, the electric field existing between the advancing channels and the earth or objects on earth i.e. buildings, poles, and trees etc, increases. In fair weather conditions, the electric field at the ground is quite high, while in the presence of stormy clouds in the sky, the electric field ranges between 30 and 40 kV/m. At the point of strike, the field assumes the value of 400 kV/m for the duration of the lightning.

The dielectric strength of the air, during stormy weather, may be well below its strength of 3000 kV/m in dry weather conditions. Such circumstance can facilitate the cloud-to-earth electric charge to reach the surrounding air critical breakdown value and cause (due to corona effect) upward directed discharges (upward leaders), basically at earth potential.

An “attachment” process takes place between the two channels. The ground potential propagates upwards, circulating to earth the negative charge accumulated in the downward directed channel. This process, usually called a return stroke, causes the circulation of the high-intensity, impulse


1

1234

5

678

91011

12131415

161718

19

202122

23242526

2728293031

3233343536

37383940

414243

23

lightning current to ground. A very rapid rise to the peak within a few microseconds (s) and, then, a relatively slow decay, are the typical characteristics of a lightning surge current wave.

4.1.2 Causes of Switching Surges in the DC traction power system

Switching operations that cause overvoltages are mainly due to trapped energy in part of the circuit and subsequent release of that energy. The following system operations in dc transit system may cause switching surges of varying degrees.

a) DC Circuit Breaker Operation: Interruption of dc current generates an arc with transient overvoltage in the order of 2.5 times the dc system voltage [B21].

b) Pantograph Arcing: An uneven pantograph contact with the OCS wire causes arcing. The intensity of such an arc and associated voltage/current surge may depend upon various factors, such as load current, vehicle speed, air gap clearance, and weather conditions.

c) AC vacuum Breaker Operation: In rare circumstances, current chopping during interruption and pre-ignition during closing of the vacuum breaker may lead to transient overvoltage [B17]

d) Voltage Transient due to dv/dt across diodes: Surges are generated due to inherent characteristics of the circuit.

e) Induced switching surges from ac power system: Heavy direct lightning strike to an ac power line in close vicinity to the OCS system may induce surges in the dc system.

f) Current Limiting Fuse Blowing: Current-limiting fuses protecting the rectifier diodes may generate transient arc overvoltage.

g) Feeder Cable Arcing Fault: Loose cable connections arcing may lead to transient surges.

AC line making contact with the OCS wire can lead to voltage surge to dc system

4.2 Surge Characteristics - PropagationThe surge Impedance (Z) and the surge propagation velocity (v) are defined by the following expressions [B3].

Z = (1)

(2)

Where:

Z Surge impedance in ohms

L Inductance in henries per unit length

C Capacitance in farads per unit length

V Surge velocity will be in meter/sec if the if the units of L are in henries/meter and units of C are in farads/meter


1

12

3

456

78

9101112

131415

1617

1819

2021

2223

24

252627

28

29

30

31

32

33

3435

23

The surge current (I), the surge voltage (V) and the surge impedance (Z) are related by the expression below [B2].

V = IZ (3)

The rate-of-rise of surge voltage (dv/dt) and rate-of-rise of surge current (di/dt) are related to surge impedance (Z) as follows [B1]

di/dt = dv/dt (1/Z) (4)

Surge propagates at the speed of light, 1000ft/s, (304.8m/s) in the OCS wire and approximately half this speed 500ft/s (152.4m/s) in the dc feeder cables [B2] [B3]. Surge experiences a surge reflection and refraction at a junction point due to change in surge impedance values. Surge propagation theory is well-documented [B2] [B10].

The following expression [B15] [B21] may be used to derive the surge voltage that could propagate towards dc switchgear via feeder cables.

VFC = 2 VI (ZC/n)/ [ ZOCS + (ZC/n)] (5)

Where:

VFC :Surge voltage through feeder cables

VI : Incident surge voltage at the OCS feeder pole junction point, say 35 kV flash over value

ZOCS: OCS wire surge impedance, 400 [B3]

ZC: Each feeder cable surge impedance, 40 [B3]

n: Number of feeder cables in parallel, say 3

4.3 Magnetic Stored Energy of SurgeSwitching surge energy (w) is exchanged between system inductance L and capacitance C parameters [B2] [B5] during its propagation defined by the expression below.

W = (1/2) L I2 joules (watt-seconds) (6)

L Inductance in henries

I Current in rms. amperes

Estimation of the energy trapped in the OCS wire using expression (6) is as follows:

L= 2.0 mH/mile Ind. of the OCS wire [B2]

I = 2000 A (assumed surge current)

D= 2 miles assumed distance between adjacent traction power substations

It is reasonable to consider that 1/2 [B2] of the surge current will propagate towards each adjacent substation.

W = ½(2x10-3 )(1000)2 = 1000 joules


1

12

3

45

6

789

10

1112

13

14

15

16

17

18

19

202122

23

24

25

26

27

28

29

3031

32

23

It shall be noted that the above calculated energy will become 25 kilo-joules if the current surge is assumed as 10 kA.

5. Surge Environment – DC Electrification SystemIt is evident from rail electrification systems using OCS that there are two paths for the surges to impact the substation equipment, one from OCS and other from the ac primary power supply system. The expected upper limit of the surge without the dc surge arrester that could propagate through the OCS system will be equal to the dry flashover value of the OCS system, which is typically 35 kV peak for 800 V dc system [B27].

The incoming primary surge will see the doubling effect at the rectifier transformer due to change in surge impedance. The intensity of incoming surge will be related to distribution system parameters and the impinging surge. Internal switching surges and surges at the OCS may be related to any of the system operational causes listed under Section 4.1.2 earlier.

Due to proximity of the OCS positive wire and the running rails (or negative wire in case of ETB system) there are equal chances that the lightning may strike both of these components simultaneously. Any surge voltage impinging the OCS wire is free to propagate towards the dc switchgear, as well as to the vehicle. The associated surge current may divide according to the surge impedance paths and may experience some attenuation.

5.1 DC Surge Protection Device RequirementsFor dc application, ac surge arresters with nonlinear resistors are re-rated. The basic requirements of a dc surge protection device include the following:

a) At highest working voltage, the device shall be essentially non-conducting, with a minimal leakage current.

b) At an overvoltage moderately above the working voltage, the device while conducting shall permit only small increases in its own terminal voltage.

c) Device shall have an adequate energy absorption capability to handle the stored energy in the dc system.

d) Upon suppressing the system transient overvoltage, the device shall quickly interrupt the normal dc voltage follow current.

e) Device shall be suitable for outdoor and indoor application subjected to harsh weather without degradation in performance.

f) The residual voltage at the OCS at maximum expected surge current magnitude shall be less than the damaging voltage of the equipment to be protected. It is affected by the lead length inductance and rate of rise of surge current.

g) Device shall accommodate dc system temporary overvoltage condition as described in Section 5.1.3 without its failure due to repeated occurrence of ETOV at the OCS.

h) Device shall be capable of providing a close margin of protection without excessive maintenance and damage.

i) Device shall have indication of failure


1

12

345678

9101112

1314151617

181920

2122

2324

2526

2728

2930

313233

3435

3637

38

23

j) Device testing procedure is desirable

DC Surge arrester application configuration is shown in Fig. 1. The MOV dc surge arrester with characteristics shown in Fig. 2 (b) appears to meet above considerations and closely matches with the characteristics of an ideal device shown in Fig. 2 (a). However, its energy absorbing capabilities should be checked in dc system application.

The gapped MOV surge arrester shown in Fig. 2 (d) may not provide a close margin of voltage protection as the triggering of the gap will occur at relatively higher peak overvoltage condition. Spark gap type arrester with characteristics of Fig. 2(c) may not reseal causing system follow current flow to ground. Some manufacturers promote magnetically blown spark gaps with series connected non-linear resistors in a single stack. To improve the internal voltage distribution, a grading resistor and capacitor is provided for each spark gap. Such a device could provide adequate surge absorbing capability, as well as close margin of protection. This standard makes a recommendation to use properly rated MOV type surge arrester.

5.1.1 DC Surge Arrester Test Data and Energy Capability

Without a standard on the manufacture of dc surge arresters, the testing and rating method among the various suppliers may vary. The energy absorbing capability varies, depending upon the quality and quantity of basic material (zinc oxide) used in the development of surge arrester. It appears that the manufacturers rely on the test data provided by the suppliers of the basic surge arrester material. For example, the data shown in Table 2 by the surge arrester manufacturer is the same as shown by Harris Semiconductor Corporation [B5] for the Type CA Metal-Oxide Varistors. However, there is no indication how the individual units were tested after their assembly in their housing. The design of connection terminals may change the test data. The unit shall be tested after assembly. It shall be noted that the terminology used in Table 4 is in accordance with ANSI/IEEE Std. C62.33 [B11], however, it differs from the terminology used in Tables 1, 2 and 3.

Manufacturer's technical data shown in Table 1 does not list all required test data including the energy capability for the surge arrester as seen in Tables 2 and 3. Fig.3 from another manufacturer does not indicate the test surge wave shape. Fig. 4 shows the generic relationship of surge arrester MCOV rating and energy capability. Surge arrester’s energy capability can be increased by using a series parallel combination of the basic MOV discs without increasing its MCOV rating which may be required for a dc surge arrester [B1].

Two different nomenclatures, discharge voltage (V IR) and clamping voltage (VC), have been used by the dc surge arrester manufacturers as shown in Tables 1, 2, 3 and 4. This may create confusion to the application engineer.

The lack of test data and test procedures for dc surge arresters has created a challenge for the application engineer to evaluate their protection capabilities.

Table 1: DC Surge Arrester Parameters from Published Catalogue

MCOV VoltsDC

0.5s, 10kAMaximum Discharge

Voltage - VIR

(kV peak)*

500A Switching SurgeMaximum Discharge

Voltage - VIR

(kV peak)**

8/20s Impulse Wave peak current maximum discharge voltage - VIR

(kV peak)

1.5 kA 3 kA 5 kA 10 kA 20 kA

900 3.4 2.2 2.5 2.6 2.8 3.0 3.5

1800 5.8 4.4 5.0 5.1 5.5 6.0 7.0


1

1

2345

6789

10111213

14

1516171819202122232425

262728293031

323334

3536

37

23

* Maximum discharge voltage for a 10 kA impulse current wave, which produces a current wave cresting in 0.5s.

**Based upon a current surge of 45s time to crest, 500A peak.


NominalVoltage

RatedVoltage MCOV

ThermalEnergy

absorbing capability

Max. values of the residual voltages in kV (discharge Voltage-VIR) at peak discharge currents of impulses

30/60s 8/20s 1/2s0.5kA 1kA 5kA 10kA 20kA 20kA

kV kV kV kJ kV kV kV kV kV kV0.75 1 1 10 1.9 2 2.3 2.4 2.7 2.61.50 2 2 20 3.8 4 4.5 4.8 5.3 5.13.00 4 4 40 7.6 8 9.0 9.6 10.6 10.2


Nom.Volt.kV

(Type)

MCOV

kV

Max. Values of the residual voltage in kV (discharge Voltage-VIR) at peak discharge currents of impulses. Energy capability, 2 impulses – 10.5 kJ/kV

MCOV30/60s 8/20s 1/2s

250A 500A 1 kA2 kA 1kA 1.5k

A5kA 10kA 20kA 10

kA20kA

1.0 1.0 1.96 2.01 2.06 2.13 2.10 2.16 2.31 2.40 2.64 2.67 3.001.5 1.5 2.92 2.99 3.06 3.19 3.15 3.22 3.46 3.60 3.96 4.04 4.472.0 2.0 3.89 3.99 4.08 4.25 4.20 4.29 4.61 4.80 5.28 5.38 5.962.5 2.5 4.95 5.07 5.19 5.41 5.34 5.45 5.86 6.10 6.71 6.84 7.573.0 3.0 5.84 5.98 6.12 6.38 6.30 6.43 6.92 7.20 7.92 8.07 8.934.2 4.2 8.10 8.30 8.50 8.85 8.75 8.92 9.60 10.0 11.0 11.2 12.4

Table 4: Metal-Oxide Disc Varistors (CA Series) from basic MOV material [B5]

Maximum Ratings (85 C ) Characteristics (25 C)Cont.DC

Volt.

TransientEnergy with Wave

10/1000s

Peak Amp Wave8/20s

Varistor Voltage at1 ma dc test current

Max. Peak ClampingVoltage (VC) with

surge wave200A peak, 8/20s

TypicalCap. @ 1MHz

Vm(dc)

(Volts)Wtm

(Joules)Itm

(kA)Min.(Volt)

VN(dc)

(Volt)Max(Volt)

VC

(Volt)Pico-farads

(pf)970 2600 70 1080 1200 1320 1880 35001150 3200 70 1290 1500 1650 2340 27001400 3200 70 1620 1800 2060 2940 22001750 5000 70 2020 2200 2550 3600 18002150 6000 70 2500 2700 3030 4300 15002500 7500 70 2970 3300 3630 5200 12003000 8600 70 3510 3900 4290 6200 1000


1

12

3

45

6

7

8

23

3500 10000 70 4230 4700 5170 7400 800

Wtm : Rated Single Pulse Transient Energy, VN(dc): Nominal varistor dc voltage

Itm : Single Pulse Transient Peak Current, Vm(dc): Max. Cont. Operating Voltage (MCOV)


1

1

2

23


1

1

2

23

Fig. 2: Characteristics of Various Surge Arresters

Fig. 3: Clamping Voltage v/s Peak current (check copy right)

Fig. 4: Surge Arrester Energy Capability (copy right check)


1

1

2

3

4

5

23

5.1.2 DC Surge Arrester Application: Surge arresters configuration and surge impedance parameters are shown in Fig. 1. There is a need for research for surge arrester application.1

A systematic approach shall be applied based upon the system configuration, surge parameters, and the expected intensity of surges and careful review of the manufacturer’s data on arresters. The selection of an arrester will require establishing its MCOV rating, energy handling capability, and the arrester discharge voltage.

5.1.3 DC Surge Arrester MCOV Rating

DC surge arrester maximum continuous operating voltage (MCOV) will definably be greater than the system operating voltage, but should also be greater than the temporary overvoltage value (ETOV) determined by the following expression [B21].

ETOV = (F) (VR)(RG)E (7)

Where:

E: System No Load Voltage, 800V

VR: Specified Voltage Regulation of Transformer Rectifier Unit in Per Unit, which will be (1.06) for 6% voltage regulation

F: Upper limit of utility primary power supplies voltage regulation factor, suggest this value not to be less than 1.10.

RG: Vehicle Regeneration Factor, use 1.15

Using these values, ETOV will be 1073 volts, thus dc surge arrester MCOV should be greater than 1073 V in this example. This statement is based upon the consideration that in a dc transit system temporary overvoltage may appear at the OCS as many times as the operating trains go through the regeneration mode especially if there are no other trains nearby to absorb the regenerated power. Such ETOV may also occur during night time when utility voltage may go up and dc load (trains) is relatively less. To assure that selected surge arrester MCOV should be greater than ETOV , an engineering analysis of the actual dc power system performance parameters listed in equation (7) above and adequate voltage-margin of protection is required. This standard recommends MCOV ≥ ETOV to avoid surge arrester premature failure due to ETOV

especially when indication of the failed (degraded without enclosure rupture) surge arrester is not easily available unless field testing of the installed surge arresters is conducted. Degraded surge arrester may be conducting harmful dc stray current to earth.

5.1.4 DC Surge Arrester Voltage – Margin-of-Protection

Surge arrester voltage-margin-of-protection (Vsa) above remote earth is defined as the peak voltage seen at the OCS by conduction of a surge current that results in maximum front of the wave internal discharge voltage (VIR), and maximum voltage drop across its grounding leads on both sides. For correction application of a dc surge arrester to OCS, it’s this peak voltage that should be as close to OCS dc voltage as possible such that no damage should occur to the dc traction power system components, including dc feeder cables, vehicles and substations. This voltage can be calculated by the following equation:

Vsa = VIR + L. di/dt + IZSG (8)

1 There is a room for research to establish similar concept of High-Low or the Low-High cascaded arresters [B14] at the OCS pole and the vehicle


1

12

3456

7

89

10

11

12

13

1415

1617

18

192021222324252627282930

31

32333435363738

39

23

45

Where:

ZSG Surge Arrester ground rod (electrode) impedance measured in ohms at 60 Hz, usually less than 5 ohms

VIR Front of the wave maximum IR discharge, voltage drop of arrester in kV peak

L Inductance in henries of surge arrester leads

I Peak surge current in amperes

di/dt Rate-of-rise of the surge current in kA/sec

It shall be noted that the voltage drop across the ground electrode impedance (IZSA) does not affect the dc equipment protection margin as the system negative, which acts as a reference point is grounded via leakage resistance to ground of the running rails insulators. In addition, all metallic components, including the OCS pole, are grounded to the same earth near to surge arrester ground electrode. Thus, for the surge arrester Vsa calculation, as shown in the equation (9) below the factor IZSG should not be used.

Vsa = VIR + L. di/dt (9)

Without complete knowledge of the surge environment and test data on the dc surge arresters, one application approach is to apply a lower voltage surge arrester, such as 970 MCOV dc for the 800V dc system, knowing it has relatively lower energy capability, but better voltage margin of protection. In case of its failure in actual application, it will provide the measure of surge environment. To increase surge energy handling capability of the surge arrester without increasing its MCOV rating, another approach may be to apply two surge arresters of low MCOV rating in parallel with individual ground leads and grounding electrode connections. However, the concern of increased leakage current under normal system operation with parallel arrester approach should be checked based upon this analysis.

Based upon above described analysis, this standard recommends MCOV ≥ ETOV to avoid surge arrester premature failure due to ETOV especially when indication of the failed (degraded without enclosure rupture) surge arrester is not easily available unless field testing of the installed surge arresters is conducted. Degraded surge arrester at the OCS system may be conducting harmful dc stray current to earth until it is replaced with a new surge arrester.

5.1.5 DC Surge Arrester Energy Discharge Capability

Surge arrester surge energy (W) in joules can be calculated by the following expression below:

W= (10)

Where:

V: Surge arrester front of the wave protective level in volts

I: Peak discharge current in amperes

t: Time in seconds the surge reaches voltage (V)


1

1

23

4

5

6

7

89

10111213

14

151617181920212223

2425262728

29

30

31

32

33

34

35

36

23

If the surge wave shape is known, then another easier expression for the energy discharged through an arrester may be calculated by using the equation (11) below [B5].

W = KVCIτ (11)

K = Constant, 0.5 for triangular wave, 1.0 for rectangular wave and 1.4 for exponential decaying wave

W = Energy in joules

VC = Clamping voltage in volts

I = Impulse current in amperes

τ = Impulse duration in seconds

5.1.6 DC Surge Arrester Application Analysis Calculation

An engineering analysis calculation for application of dc surge arrester to OCS operating at 800 V nominal dc voltage is discussed. Calculations are based upon using surge arresters with MCOV rating of 900 V and 1800 V and other parameters published by their manufacturer.

a) Using rate-of-rise of incoming surge voltage (dv/dt) at the OCS surge arrester location of 11 kV/s per kV MCOV [B1], the calculated values of dv/dt will be 9.90 kV/s and 19.8 kV/s for the 900 V and 1800 V arresters respectively.

b) Surge arrester lead lengths inductance using 0.4 H/ft with 25 feet length will be in the order of 10 H for each arrester.

c) Rate-of-rise of the surge current at the arrester location by using equation (4) and surge impedance of arrester leads of 400 will be 0.025 kA/s and 0.050 kA/s respectively for each arrester.

d) Using the published data shown in Table 1 for surge arrester front of wave protective level at 0.5 s, 10 kA peak discharge current, discharge voltage (VIR) will be 3.4 kV and 5.8 kV respectively for each arrester.

e) Margin of protection voltage (Vsa) by using equation (9) without considering (IZSG) factor will be:

900V : Vsa = 3.4+0.25 =3.65 kV

1800V: Vsa = 5.8+0.50 =6.30 kV

f) Energy (W) in Joules discharged by the surge arrester for a switching surge may conservatively be estimated by the equation (11). The values indicated in Table 4 are in the order of 2600 and 5000 joules respectively for 900V and 1800V arresters for the long wave 10/1000s and may be compared with the calculations in this step for 45/1000s wave:

900V DC MCOV Rated Surge Arrester:

W = 0.5 * 2.2 * 0.5 * 45 +1.4 * 2.2* 0.5* 0.5*1 000

W = 24.75 + 770.00 = 794.75 Joules


1

12

3

45

6

7

8

9

10

111213

141516

1718

192021

222324

2526

27

28

2930313233

34

35

36

23

1800V DC MCOV Rated Surge Arrester:

W=0.5 *4.4.2*0.5*45+1.4*4.4*0.5*0.5*1000 =1589.5 Joules

It should be noted that the calculated values of energy for each surge arrester using 45/1000s wave parameters are lower than the published data of energy discharge capability of the arrester using 10/1000s wave. Since the characteristics of the two wave shapes are different, therefore manufacturer should be consulted to provide discharge energy capability for (45/1000s) wave.

g) Voltage surge through feeder cable (VFC) in kV at the dc switchgear can be calculated by use of equation (5):

900V MCOV Surge Arrester: VFC=2x3.65x(40/3)/[400+40/3] = 0.24 kV

1800V MCOV Surge Arrester: VFC=2x6.30x(40/3)/ [400+40/3] = 0.41kV

It appears that without the effect of surge impedances, the conservative voltage impressed at the equipment may be 3.65 kV and 6.30 kV, respectively as indicated in item e) above.

h) Current surge (IS) associated with voltage surge derived in item g) above can be calculated by use of expression (3)

900V MCOV Surge Arrester: IS = ES/ZS = 0.24x(3/40) = 0.018 kA

1800V MCOV Surge Arrester: IS = ES/ZS = 0.41x(3/40) = 0.0.031 kA

Although the energy capability of 1800V arrester is higher than the energy capability of 900V arrester, however, the 900V arrester provides better voltage protection if it can withstand system energy capability. It shall be noted that 900V MCOV rating appears to be lower than the system temporary overvoltage (ETOV) value of 1073 volts derived by using equation (7) and thus may lead to premature failures in actual installation. Thus the surge arrester with MCOV rating of 1150 V may be the proper choice for 800 V dc system.

6. Lightning Stroke TerminologyPerhaps it is best to clarify the terminology; reference [B20] makes a distinction between the traditionally used term “stroke” and a more precise reference to the term, “flash”. A flash describes the entire electrical discharge to the stricken object. Stroke, on the other hand, describes only the high-current components of a flash. Because of the observed multiplicity of strokes, the relationship between the terms “flash” and “stroke” is that there can be many strokes in a single flash. Research into flash characteristics indicates that 55 percent of all flashes contain multiple strokes, with an average value of three strokes [B20]. This information is important because of the differences in wave shape of the successive strokes. The term “flashover” is described as an electrical discharge completed from an energized conductor to a grounded support structure, which will be OCS poles in case of an LRT system.


1

1

2

3456

78

9

10

1112

1314

15

16

171819202122

2324252627282930313233

23

Fig.5 Surge Arresters Configuration – Surge Impedance Diagram2

6.1 Lightning Intensity Estimation Lightning intensity within a specific area is generally based upon the ground flash density, Ng, in flashes per km2/year. At present, this data is not available in the United States and thus, lightning intensity must be based upon the isokeraunic level, or the number of thunderstorms per year, T d.

The value of Ng may be approximated by using the following empirical expression [B4] [B19]. With more research data available in the future such an expression may change.

2 Future research may provide a software program to provide surge current distribution


1

1

2

3456789

2

34

Ng = 0.04 Td1.25 (12)

For example consider the area of a light rail transit (LRT) where T d is in the order of 40-60 [B3][B4]. Using expression (12) and Td of 60, the calculated value of Ng will be near 6.68. It is noted that the exponent value of 1.25 in expression (12) is somewhat uncertain, for some published literature indicates this value to be 1.35. However, 1.25 has been accepted by the committee responsible for the development of the standard [B19] and thus, for this example, OCS lightning protection analysis will be based upon the value of Ng to be around 6.68.

This calculated number for Ng provides some measure of likelihood of lightning strike to ground in the area. The actual number of lightning flashes/year, NOCS, that may strike the light rail OCS or nearby ground inducing direct or indirect lightning surge waves, may be calculated by using the following expression:

NOCS = wLNg (13)

Where:

L = length of LRT system in kilometers

w = Width of area covering LRT tracks in kilometers

Assuming a double track dc system with width near 0.015 kilometer and Ng of 6.68, by expression (13), calculated value of NOCS will be: L/10.

NOCS = L/10 (14)

Assuming the probability of direct hit of lightning strike to OCS (ND) is 20% of the value calculated for the actual number of lightning flashes/year, NOCS in that case, for the above example following relationship applies.

ND = 1/5 (NOCS) = L/50 (15)

Where: L is the dc electrification system length in kilo-meters

Assumed low probability of 20% of direct lightning hit to OCS indicated above is based upon the reasoning that there are equal chances that lightning may hit any of the OCS support structures, nearby buildings, trees, substation structures, communication and control cabinets including running rails.

Thus in the above example for a dc system with 10 km length, the calculated number of lightning flashes per year (NOCS) that may strike the OCS system is one (1) and perhaps 50 km length is needed for direct hit to OCS. For 10 km length of OCS system, the expected single lightning flash per year may not be a direct hit to the OCS system. In addition, the expected single lightning flash may or may not be of concern, depending upon the severity and energy associated with the lightning stroke (surge) contained in the flash.

6.1.1 Lightning Stroke – OCS Flash Over

This discussion is intended to establish the lightning overvoltage intensity to the OCS components, especially the contact wire, which is generally protected by dc surge arresters. The various components of the OCS, including messenger wire, contact wire, feeder cables and supporting structure (which consists of metallic poles, cross-arms, and running rails), are relatively close to each other. There are equal chances that the lightning strike may hit any of the above-described OCS components.


1

1

234567

89

1011

12

13

14

15

1617

18

192021

22

23

24252627

282930313233

34

353637383940

23

The messenger wire, cross-arms, and grounded metallic poles may provide some measure of shielding of direct lightning strike to the OCS contact wire. In rare circumstances, if the lightning strikes directly to the OCS wire, flashover is almost certain since the insulated air gaps and clearances from the grounded metallic components including the poles is relatively low with wet and dry flashover values near 20 kV to 35 kV peak respectively for 750 V nominal dc. Lightning strike energy after the flashover at the OCS pole will go to ground via grounding path of the poles.

After the flashover, the maximum voltage expected at the OCS contact wire would not be more than the actual dry flash over value of the insulator. The time to flashover from stroke, the energy contained in the remaining surge wave at the OCS, and its propagation away from the point of strike will depend upon the rate-of-rise of the incoming surge waves of the lightning flash strokes. As indicated earlier, 60 percent of the strokes may strike the OCS poles and the remainder at mid-span of poles.

The maximum distance that a lightning surge will need to travel before hitting the grounded pole for flashover phenomenon is ½ of the pole spacing distance, which in terms of the surge wave propagation time is relatively small. Without the application of dc surge arrester at each OCS pole, the metallic grounded OCS poles will provide adequate path to the lightning strokes with peak voltages exceeding dry flash over voltage of the insulators. This OCS poles flashover to ground will cease automatically once the OCS surge voltage falls below the insulator’s actual flash over voltage. The flashover may occur again if there are repeated lightning strokes in a particular flash.

If the flashover occurs near the dc feeder poles with dc surge arresters, the dc surge arrester may also start discharging during the pole flashing. It is also apparent that as the propagation time of the surge to adjacent feeder pole towards the next substation is small, the surge arrester on adjacent substation will also start conducting. In addition, the surge wave will also propagate via an underground feeder cable to the dc switchgear with a reduced surge magnitude indicated by equation (5) in section 4.2 earlier. Thus, surge arresters applied at the dc feeder breakers will reduce the effect of surge propagation on feeder cables and the substation equipment.

For a LRT system in a high isokeraunic area, if the flashover occurs to OCS poles or rails, then the induced surge voltage will get into the running rails, or if the surge strikes directly to the running rails then the surge may propagate to the substation negative bus via the negative underground dc feeders. Therefore in such areas surge arresters should also be applied at the dc negative bus or the running rails.

In case of high isokeraunic lightning stroke, a concern of damage to the surge arrester rises due to its limited surge energy handling capability. However, it appears that for such a severe lightning stroke, flashover across the outer surface of the surge arrester may occur due to its short length. Such flashovers will drain the surge energy to ground leaving lesser surge current and energy to be discharged through the surge arrester. If there is still some concern of the dc surge arrester to be inadequate in handling the surge energy, then properly rated surge arrester with higher surge energy handling capability should be applied.

The dc surge arresters applied at the dc feeder poles or other locations should be adequate to handle the discharge current of the lightning surge wave deposited by the lightning flash strokes after the OCS flash over occurs. In addition, the dc surge arrester discharge voltage should be such that it provides adequate voltage margin of protection to the operating Light Rail Vehicle (LRV) and the traction power substations. Since these surge arresters at the OCS contact wire are first lines of defense to trap the lightning/switching surge voltage below the protection level of the connected equipment, an engineering analysis of surge arrester voltage ratings should be performed for proper selection of the surge arresters [B21].


1

1234567

89

10111213

1415161718192021

22232425262728

2930313233

34353637383940

4142434445464748

23

6.1.2 Lightning Stroke MagnitudeResearch on the stroke current peak amplitude reported that the mean value of first stroke is near 31 kA, with a 95 percent probability of the stroke magnitude being between 10 and 100 kA [B20]. The first stroke wave shape mean value just before the current peak has been reported to be near 24.3 kA/s, which is helpful in understanding the impulse voltages that can occur for discharges through inductances. It is necessary to indicate that although the average value of the peak magnitude of the subsequent stroke(s) is generally less than the first stroke, the wave front(s) of the subsequent stroke(s) are typically faster. The average value is near 39.9 kA/s, although values in excess of 70 kA/s have been reported. The above mentioned stroke parameters relate to the flash itself and much of the data was obtained from mountaintop observatories [B28]. It is also reported that 60 percent of the direct flashes hit the tower where they would flashover to the ground and the remainder hit on the spans between the towers.

The above listed current lightning waves develop very high corresponding voltage waves based upon their relationship provided by the equations (3) and (4) described under Section 4.2 earlier.

Consideration must be given to some modification of the flash characteristics striking OCS, especially when tracks may be surrounded by urban development. Any high-rise buildings including the trees and street light poles that are taller than the OCS poles will provide some degree of lightning flash shielding to the OCS. However, since there is no measured research data specifically for the dc transit OCS, the conservative approach is to use the data available for the transmission towers for the OCS.

6.1.3 Lightning Stroke Induced Over-voltage

Lightning overvoltages are also possible due to electric and magnetic fields induced from nearby lightning, often referred to as indirect or induced surges. For transmission lines, peak overvoltages induced by first strokes varied between +150 kV and –40 kV, the mean being 23 kV. The mean rise time for these voltage surges was 6 s. This provides rate of rise of the voltage wave to be approximately 4 kV/s. The study further revealed that induced overvoltges caused by subsequent lightning strokes had 11 kV peaks, with a mean rise time of 4 s. This provides a rate of rise of the voltage wave to be approximately 3 kV/s, which is much lower than the values reported for the direct lightning flash hitting the transmission lines. Such lightning wave parameters may be used for LRT system design and engineering analysis of lightning protection, which is the purpose of this standard.

6.1.4 Lightning Stroke Surge Energy

Surge energy (J) may be calculated by the expression [B1]:

J = Joules (16)

Assume lightning stroke with the following parameters:

kV/ s (2x1011 V/ sec)

Assume surge impedance for surge voltage to be near 40 ohms, parallel combination of OCS with supplementary feeder cable. Thus surge current wave by use of expression (18) will be as follows


1

1

23456789

101112

1314

151617181920

21

22232425262728293031

32

33

34

35

36

373839

23

s (5x109 A/sec)

The maximum flashover kV peak for OCS is near 35 kV (dry weather condition), thus within 35/200 s (0.18s), OCS poles will flashover to ground with or without the application of dc surge arresters.

Thus the lightning stroke energy that may pose threat of OCS damage or the dc surge arresters will be for flashover time of 0.18 s with calculated stroke energy value indicated below.

t = 0.18 s (18x10-8 sec)

J = 2x1011 x5x109.x joules

J = 1021x 10-24 x183 x 1/3 joules = 5.83/3 kJ (17)

The OCS system appears to get self-relief from the heavy lightning stroke energy (responsible for damage to dc surge arresters and other OCS equipment) due to flashover near 35 kV peak surge magnitude without the help of surge arresters. However, 35 kV peak voltages are quite damaging to the system components, such as dc switchgear and also LRV components. Thus dc surge arresters of proper rating should be applied. These surge arresters will discharge current and will handle energy as indicated in expressions (20) and (21) below.

6.1.5 Arrester Discharge Energy

Arrester discharge current is a function of many interrelated parameters, including: Surge impedance of the OCS Stroke current characteristics, wave shape, peak current magnitude, and its rate-of-rise Distance of the surge arrester from the point of stroke Ground resistance at the location of stroke Number and locations of flashoversFlashover characteristics of the OCS insulators Arrester discharge voltage

The following expression [B4] is used for power distribution overhead lines and may be used for the OCS system:

IA = (ES - EA)/ Z (18)

Where:

EA = Arrester impulse discharge voltage (kV) for current IA (kA)

ES = Prospective surge voltage (kV)

Z = Surge impedance of the conducting path of the surge arrester ()

IA = Impulse current (kA) associated with impulse voltage

Energy discharged by the arrester, J, in kilojoules (kJ), may be conservatively estimated by the following expression [B18]:

J = 2 DL EA IA /v (19)


1

1

234

56

7

8

9

101112131415

16

1718192021222324

2526

27

28

29

30

31

32

3334

35

23

Where:

EA = Arrester discharge voltage (kV)

IA = Switching impulse current (kA)

DL = Line length (miles) or (km)

v = the speed of light (190 miles/ms) or (300 km/ms)

The expression assumes that the entire line is charged to a prospective switching surge voltage and is discharged through the arrester during twice the travel time of the line.

If the surge wave shape is known, then another easier expression for the energy discharged through an arrester may be calculated by using the equation (11) described under section 5.1.5.

7. Lightning Stroke to OCS systemFor analysis purposes, assume maximum distance between the substations to be near 1 1/2 miles. Assume that there are surge arresters installed only at the feeder poles adjacent to each traction power substation, and there are no other poles between the substations that are equipped with surge arresters as shown in Fig. 5. A lightning strike hitting the OCS wire midway between the two substations will propagate equally with 1/2 the impinging surge current magnitude to each substation [B2]. Thus, in this example the surge will travel maximum distance of 3/4 mile before reaching a pole with dc surge arresters.

For a 750V dc nominal LRT system voltage, consider a dc surge arrester rated at 2 kV duty cycle with MCOV rating near 1800 V dc with discharge voltage rated at 7.0 kV. This discharge voltage is the surge arrester test voltage, which is based upon 20 kA peak current of a standard 8/20s wave.

Time in milliseconds to travel 3/4 mile will be 3/4x1/190 (4s). The energy discharged through the surge arrester in kJ using equation (19) may be calculated as shown in equation (21).

Assume surge wave is magnified to twice its magnitude (2 times 35 kV dry flashover value of OCS) due to open circuit condition of a sectionalizing dc disconnect switch. Using OCS surge impedance of 40 ohms (surge impedance of OCS wire in parallel with underground supplementary cable), and assuming surge arrester discharge voltage (7 kV) to be the test voltage at 20 kA peak, the discharge current IA and surge energy discharge will be as follows:

IA = (2x35 – 7.0)/40 = 1.575 kA (20)

J = 2 x ¾x(1/190) x 7.0 x 1.575 kJ = 0.088 kJ (21)

It should be noted that the time to travel 3/4 mile distance by the lightning stroke is very small and it is possible that the lightning stroke time may be longer than two times the travel time for 3/4 mile distance. Under such circumstances, the maximum estimated time for the lightning stroke should be used for estimating the energy discharge through the surge arrester. The time 3/4x2/190 ms (8 s) used in calculating energy J in kilojoules should be increased to a reasonable value, say 300 s, the maximum estimated time the lightning flash containing more than an average of three strokes may exist. This will lead to calculated energy of 3.30 kJ, which will still be below 4.4 kJ (2.2 kJ/kV) value for a 2000V dc surge arrester. This estimation of surge energy is very conservative as the assumed value of surge time and arrester discharge voltage seems to be on the high side. However, dc surge arrester selection based upon such high energy discharge requirements will assure that the arrester will not be damaged by high isokeraunic lightning flash hitting very close to its location.


1

1

2

3

4

5

67

89

1011121314151617

18192021

2223

2425262728

29

30

313233343536373839404142

23

In the above calculation it has been assumed that lightning surge impinging the OCS wire will flash over to grounded metallic OCS poles once the surge wave voltage reaches 35 kV peak. Time to reach 35 kV peak will depend upon the rate of rise of lightning surge wave. If the rate of rise for example is near 200 kV/µs, then the time to reach 35 kV peak will be 35/200 µs, which is far less than the travel time of 4µs for 3/4 mile distance. Thus, OCS wire will not charge more than 35 kV peak voltage unless the surge comes across an open circuit caused by open position of a disconnect switch. However, as the surge propagation time to reach open circuit location is quite higher than the time to develop 35 kV peak voltages at lightning flash striking location, the OCS pole flash will occur before double peak voltage (2x35 kV) is impressed upon the OCS contact wire. Thus, flashover phenomena will reduce the surge energy that will be discharged through the dc surge arrester.

To provide assurance that OCS wire flashing over occurs in case of a direct lightning strike impinging the system in high isokeraunic areas horn air gap type arresters should be considered midway between adjacent traction power substations.3

7.1 Underground Supplementary Cable Connections to OCSIn certain sections of the OCS, it is assumed that there will be underground supplementary cable. For discussion purposes, it is assumed that the average distance between the OCS contact wire and the underground supplementary cable tap connections are in the order of approximately 400 feet. For calculation purposes, assume 4000 feet length of such supplementary cable which will then require a total of eleven (11) OCS contact wire-to-supplementary cable tap connections. This configuration of OCS and supplementary cable connections will have total of nine (9) underground cable splices. Such underground supplementary cable installation, electrically in parallel to OCS will require cable splice connections to be located in the underground manholes. Lightning surge withstand capability of such underground cable splices in the manholes and the tap point connection of cable to overhead OCS contact wire are of concern.

Analysis of the cable splices and cable connections to the OCS contact wire would require a derivation of the peak value of the lightning surge voltage expected at these connection points and cable splices. Then this surge voltage peak value will be compared to the tolerable values of basic surge withstand impulse voltages (BSL) of the cable splices and cable-to-OCS connections.

The basic switching surge level (BSL) of the 2 kV cable is near 75 kV peak. Underground cable splice BSL levels to match with the cable BSL level are also available.

Assume the following:

ZOCS = 400 ohms (OCS contact wire surge impedance) [B2]

ZC = 40 ohms (cable surge impedance) [B2]

Vi = Voltage magnitude of the incident lightning wave at the impedance junction point (connection of cable to OCS contact wire or at the underground cable splice)

Ii = Current magnitude of the incident lightning wave at the impedance junction point

3 Such air gap horn type surge arresters do not pose any threat of dc leakage current or uncertainty of their damage due to ambient temperature. They shall be bonded to the OCS poles that are grounded with its own grounding electrode with low resistance-grounding impedance by appropriately sized wire (not less than #6 AWG) 600V insulated cable to avoid jeopardizing the OCS double insulators criteria. Another alternate to the horn gap is to consider surge arrester with higher duty cycle voltage.


1

123456789

1011

121314

1516171819202122232425

2627282930

3132

33

34

35

3637

38

23456

78

Vr = Voltage magnitude of the reflected lightning wave at the impedance junction point

Ir = Current magnitude of the reflected lightning wave at the impedance junction point

V = Total voltage magnitude (refracted voltage) at the impedance junction point

I = Total current magnitude (refracted current) at the impedance junction point

The following expressions are well documented [B1].

V =Vi + Vr (22)

I = Ii + Ir (23)

At the interface of two surge impedances Z1 and Z2, the expressions for the above indicated surge voltage and current are related by the following expressions:

V = [2x Z2 /(Z1 + Z2)] Vi (24)

I = [2x Z1/(Z1 + Z2)] Ii (25)

For the sake of completeness, expressions for the surge current as well as the surge voltage have been described. However, analysis of the surge wave voltage is more critical for the cable insulation protection when compared to surge current. It is well understood that the cable can tolerate excessive magnitude of surge current for short duration without appreciable heat rise to create damage to cable insulation. Hence, only the surge voltage analysis is presented under Sections 7.1.1 and 7.1.2 below. Single Line diagram indicated in Fig.1 shows basic elements of the power system.

7.1.1 Lightning Hits OCS ahead of the Supplementary Cable Connections

The initial and final surge voltages at the junction points of cable to OCS or splice point of supplementary cable may be calculated by using expression (24) which requires knowledge incident voltage magnitude (Vi) through surge impedance Z1 before hitting the junction point of surge impedances Z1 and Z2. In the case of OCS and supplementary cable, the surges impedances are as follows.

Z1 = 400 Ώ, for OCS and Z2 = 40 Ώ for the supplementary cable.

If all cable to OCS taps is spaced equally and the installation is uniform, then, for practical purposes neglecting the effect surge impedances of cable splices and the OCS/cable connection tap points, the combined surge impedance (Z) of underground supplementary cable and OCS contact wire may be represented by expression (26) shown below.

Z = Z1 x Z2 / (Z1 + Z2) ≈ Z2 (since Z1 >> Z2) (26)

If voltages (V1), (V2), (V3) and (V11) are successively represented as the surge voltages at the first, second, third and last (eleventh) junction points when the surge voltage travels along the OCS section with underground supplementary positive cable, the expressions for these surge voltages will be as follows:

V1 = [2x Z2/(Z1 + Z2)] Vi (27)

V2 = [2x Z2/(Z1 + Z2)][2x Z2/(Z2 + Z2)] Vi (28)

V3 = [2x Z2/(Z1 + Z2)][2x Z2/(Z2 + Z2)]2 Vi (29)


1

1

2

3

4

5

6

7

89

10

11

12131415161718

19

2021222324

25

26272829

30

31323334

35

36

37

23

V10 = [2x Z2/(Z1 + Z2)][2x Z2/(Z2 + Z2)]9 Vi (30)

Using values indicated for the surge impedances in ohms, surge voltages represented by expressions (27) through (30) will be practically equal in magnitude, approximately 18% of the incident stroke surge voltage magnitude.

The final (eleventh) point will be end of the supplementary cable where the surge impedance will become again Z1 and the surge voltage will be escalated as follows:

V11 = [2x40/(400+40)].[2x400/(400+40)]Vi (31)

This final voltage appears to be approximately 33% of the initial surge voltage that appeared when the surge entered parallel combination of OCS and supplementary cable.

If installation of the underground supplementary cable riser feeders and OCS connections is uniform, then the surge impedance will be practically the same, slightly less than 40 ohms. The above calculations indicate that voltage will never be more than the striking voltage unless there is a switch that may be in an open position to make this voltage two times the initial surge voltage. This twice the initial surge voltage can be derived by using the expression (24) as shown below.

V = [2x Z2 /(Z1 + Z2)] Vi = [2/(Z1/Z2+ 1)] Vi (32)

Since Z2 at open switch will be infinite, thus Z1/Z2 will become zero in expression (32) making surge voltage V as two times the initial voltage V i. These surge voltage calculations do not take into account the effect of surge attenuation due to cable inductance and capacitance effects.

If we assume that the striking voltage is limited to 35 kV by the flashover phenomenon, then it appears that the underground supplementary cable splices and the OCS-to-supplementary feeder cable tap connections may not require surge protection, except at the first and the last connection points at the OCS.

7.1.2 Lightning Hits OCS within Supplementary Cable Connections Zone

The surge voltage analysis for this case will be identical to the analysis presented under Section 7.1.1 above, with the exception that the incident surge in the air at the OCS will propagate to each side traversing the OCS/cable tap connections and underground supplementary cable splices. From a theoretical point of view, the current surge that propagates in each direction will practically be half the magnitude of the incident surge stroke current. The final maximum surge voltage will be at the outermost cable-to-OCS connection tap points, and it will practically become double the traveling surge voltage as indicated by the following calculation.

V = [2x 400 /(400 + 40)] Vi = 1.82x Vi (33)

This voltage V will be equal to the initial lightning stroke surge voltage, which initially split into half the magnitude at the strike locations. All intermediate tap points will see a lesser surge voltage in the order of 9% (from analysis in 7.1.1) magnitude of the initial lightning stroke before it splits into half the magnitude.

Thus, if the OCS flashover voltage is near 35 kV without the application of the dc surge arresters, then the maximum surge voltage will be near 35 kV peak or 70 kV if the design installation involves dc disconnect switch which is open.


1

1

234

56

7

89

101112131415

16

171819

20212223

24

25262728293031

32

33343536

373839

23

7.1.3 Lightning Surge Propagation Discussion for Supplementary Cable

From a theoretical point of view, more surge current may tend to propagate through the underground supplementary cable as compared to the OCS wire due to difference in the values of their respective surge impedances. However, the speed of surge propagation through OCS wire is two times the speed of surge through supplementary cable and feeder taps [B2]. This may lead to balancing out the surge energy propagation through OCS and underground supplementary cable and feeder taps.

It should also be mentioned that the underground supplementary cable switching surge peak impulse voltage withstand level far exceeds 35 kV peak surge wave that can be expected without considering the doubling effect. Thus it is not necessary that dc surge arresters be applied at every 400 feet at supplementary cable-to-OCS connection locations if there is no dc disconnect connection.

However, doubling peak voltage effect cannot be avoided at the dc disconnect switches when they are in open position. Thus at disconnect switch locations, dc surge arresters are required on each side of the switch.

In addition, it is prudent to perform surge analysis based upon actual configuration of the OCS/supplementary cable to optimize the design of surge arrester applications to the dc system.

In specific configurations of OCS under the high voltage utility overhead lines, consideration shall be made to prevent live ac wire contacting the OCS by use of guard wire or some other means. Additional surge arrester at the OCS in the vicinity of OCS/high voltage utility lines should not be used due to its misapplication for such a hazard. If the high voltage line touches the OCS, then the dc system will be exposed to a sustained over voltage condition equal to dc voltage plus the peak ac system voltage leading to damage of the surge arresters and other dc system components until the utility fault is cleared

It is a general opinion that guesswork and overconcern of lightning protection without performing surge analysis indicated in this standard has led to a design of applying surge arresters at each OCS-to-supplementary cable taps. 4

8. Grounding

8.1 OCS Pole Grounding8.1.1 OCS support poles with surge arresters shall have dedicated ground rod (ground electrode) for connecting the surge arrester to ground by use of exothermic weld. Measured grounding resistance of surge arrester ground electrode should not exceed 5 ohms to avoid excessive surge arrester residual voltage at OCS when a heavy lightning surge current discharge current takes place either through flash over across the arrester or by current discharge through the arrester.

Surge arrester ground rod should not be bonded to OCS pole foundation re-bars to avoid the uncertainly of damage to the footing due to energy dissipation of a lightning surge at a rapid speed. Bonding the pole with surge arrester grounding system may also transfer surge hazard to the pole which is not desirable. In addition, separate dedicated grounding rod is needed to ground OCS pole. All OCS poles with surge arresters shall be grounded individually by a maximum of 25

4 Such a design should be avoided based upon the surge voltage analysis presented in this standard. Addition of such excessive number of dc surge arresters to the OCS is an application concern, especially when such surge arresters do not have any indication that the arrester is in a degraded mode and may be injecting undesirable dc stray current to ground.


1

1

234567

89

101112

131415

1617

18192021222324

252627

28

29303132333435

3637383940

2345

67

ohms grounding system. For those poles that are accessible to public the grounding resistance shall be 5 ohms or less. [B27] Following the installation of the pole mounted surge arrester its grounding system resistance shall be measured and recorded for future reference.

9. DC Surge Arresters

9.1 Application

9.1.1 Surge arresters shall be provided and shall be connected to the OCS as a minimum at the feeder poles. Two (2) parallel surge arresters with separate ground conductor and separate ground rods may be needed at feeder poles for high intensity lightning area. [B27]

a) Considering the low profile of OCS, proximity of all components, inherently grounded poles, and major portion of the dc rail transit system close to high-rise structures and trees, the probability of lightning striking the OCS is very low. With this configuration, application of the ground shield wire above the messenger and contact wire does not appear to provide any greater degree of protection, especially when the lightning strike tends to flashover the grounded structures.

MOV surge arresters are sensitive to ambient temperature. In the summer when ambient temperature is high, metallic tip of the MOV dc surge arrester may become hot leading to transfer of heat to the surge arrester material. This may cause premature surge arrester failures. In addition, the surge arrester provides leakage current under TOV conditions leading to heat dissipation as well degradation of its internal MOV elements. Thus the installation should consider excessive temperature effect on performance and selection of MOV surge arresters.

a) Wire lead from arrestor to positive (or negative OCS) shall not exceed 610 mm where possible.

MOV dc surge arresters should be installed at the following locations:

At positive (and negative for electric trolley bus) feeder poles, close to pole-mounted or pad-mounted dc disconnect switches on load side of the switches.

At positive (and negative for electric trolley bus) pole-mounted or pad-mounted OCS sectioning switches. Arresters shall be installed on both sides of the switch.

The application of dc surge arrester for dc switchgear is not covered in this standard; however, review of surge arrester at the OCS and dc switchgear should be coordinated.

The application of dc surge arrester for vehicle is not covered in this standard; however, review of surge arrester at the OCS and vehicle should be coordinated.

For OCS system within high isokeraunic areas, consider installing surge arresters at the negative bus to protect the equipment under rare circumstances of lightning surge reaching the negative bus via running rails and dc negative feeder cables.5

Install dc surge arrester at positive (and negative for electric trolley bus) underground feeder tap location. As a minimum at the first and last OCS to underground positive (negative) supplementary feeder cable tap location.

5 Surge arrester at appropriate locations on running rail may be applied in elevated guideways in areas of high isokeraunic levels


1

123

4

5

678

91011121314

151617181920

2122

23

2425

2627

2829

3031

323334

353637

234

56

Surge arresters shall be considered at locations where surge impedance changes due to OCS configuration, such as bridges and end points of the tunnels.

Surge arresters shall be considered for high isokeraunic areas, at the midway between the adjacent substations.

Surge arrestors shall be considered in the above applications for negative contact wires and feeder cables in electric trolley bus systems.

b) An engineering analysis should be performed to determine appropriate voltage and energy capability rating of the dc surge arresters. The analysis should take into consideration the OCS location, ambient environment and operating voltage characteristics.

c) MOV dc surge arresters continuously conduct milli-amperes level of current to ground. This current may increase if the internal material becomes defective. 6

9.1.2 Name Plate Information

The following information shall be provided:

DC Surge Arrester

Rated voltages, duty cycle voltage and MCOV

Nominal peak discharge current with current waves of 8/20 s, and 100/1000 s

Surge discharge capability kA peak

Energy discharge capability (klo-joules)

Manufacturer’s name

Year of manufacture

Serial number

9.2 Surge arrester Rating

9.2.1 Surge arresters discharge voltage shall be no more than 80% of the BIL of the equipment that is being protected.

9.2.2 MCOV voltage rating of surge arresters shall be greater than Temporary Overvoltage TOV of the OCS system. For dc surge arrester ratings refer to Tables 1, 2 and 3 under Section 5 and Table 5 below.

6 Future development of dc surge arresters should provide visual indication when the surge arrester becomes defective or fails so that it can be removed to avoid the uncertainty of draining continuous dc stray current to ground.


1

12

34

56

789

10

1112

13

14

15

16

17

18

19

20

21

22

23

2425

262728

2345

67

Table 5: OCS System Surge Arrester Voltage Levels

System Nominal Voltage (Volts) dc

System Max. Continuous Voltage (Volts) dc

Temporary overvoltage voltage (ETOV )

(Volts) dc

600 to 850 1,020 1,150

1,500 1,800 2,000

3,000 3,600 4,000

9.2.3 DC Traction Power System Components’ Voltage Withstand Capability. DC Traction Power System components’ including OCS and dc switchgear voltage withstand capability listed in Table 6 is the per IEEE draft Standard P1626 [B22] and IEEE STD C37.14 [B7].

Table 6: DC Traction Power System Components’ Voltage Withstand Capability

DC Switchgear OCS minimum dc Spark over Voltage

Rated Voltage

(Volts dc)*BIL

(kV Peak)*Dry Weather

(kV Peak)**Wet Weather

(kV Peak)**

1,300 65 45 25

2,100 65 45 25

4,200 65 45 25

* Per IEEE STD C37.14 [B7]

** Per IEEE Draft STD P1626 [B22]

10. DC Surge Arrester Service Requirements

Operating ambient temperature shall be within the range of the operating environment between -40 degrees Fahrenheit and + 104 degrees Fahrenheit. System OCS to ground voltage shall be within the rating of the arrester under all system operating conditions. For OCS systems above 1,800 feet consult the surge arrester manufacturer for revised ratings.

11. DC Surge Arrester Testing

11.1 Design Tests

Surge arresters shall be subjected to the following design tests:


1

1

2

345

6

7

8

9

10111213

14

15

16

23

Insulation withstand test between the terminals

DC dry and wet spark over test

Discharge voltage test

Impulse protective level voltage time characteristics

Accelerated aging procedure

Pressure relief test

Leakage current at MCOV and 5% voltage in access of MCOV, leakage current v/s voltage

Effect of higher temperature on leakage current, current v/s temperature graph

Peak discharge voltage under 8/20 µs lightning surge, with peak current of 200 A, 500 A, 5 kA and 10 kA

Peak discharge voltages under 45/1000 µs switching surge, with peak currents of 100 A, 200 A and 500 A.

Energy absorbing capability for the 45/1000 µs switching surge with 100A, 200A and 500A peak current

Energy absorbing capability for the 8/20 µs lightning surge with 200A, 500A , 5 kA and 10 kA peak

New and clean arresters shall be used for each design test. For additional test requirements, see IEC Draft Standard NE 50562-1.

The arrester shall be mounted in the position(s) in which it is designed to be used.

Ambient temperature for test shall be -40 to +104 degrees Fahrenheit.

11.2 In Service (Field) Tests

All surge arresters shall be visually inspected per IEEE Std. P1628 recommendations.

Surge arresters shall be subjected to electrical tests per manufacturer’s recommendations.

Apply dc test voltage in very slow increments of 20V dc each till the voltage reaches MCOV rating of the surge arrester. At each voltage step, measure leakage current through the arrester as excessive leakage current will result if arrester is defective. Do not apply voltage above MCOV rating for longer than manufacture’s recommended time to avoid damaging the surge arrester during testing. Voltage above TOV rating of OCS system shall only be surge waves and not the dc voltage. Surge wave shall be 8/20s with peak current magnitudes of 0.5kA, 1.0kA and 1.5 kA. Such surge waves if applied should be for a very short duration less than 25s. Review such testing requirement with the manufacturer of the surge arrester under test before application of the suggested surge current for testing.


1

1

2

3

4

5

6

78

9

1011

1213

1415

1617

1819

20

21

22

23

24

252627282930313233

23

11.3 Bibliography

[B1] Dr. Allen Greenwood, Electrical Transients in Power Systems, New York, Wiley 1971 Edition

[B2] IEEE Standard 141 – 1986, (Red Book), IEEE recommended Practice for Electric Power Distribution for Industrial Plants

[B3] Dev Paul and S.I Venugopalan, " Power Distribution System Equipment Overvoltage Protection", IEEE Trans. Ind. Appl., vol. 30, pp. 1290-1297, Sept./Oct. 1994

[B4] ANSI/IEEE Standard C62.22-1997, IEEE Guide for the Application of Metal -Oxide Surge Arresters for Alternating Current Systems

[B5] Transient Voltage Suppression Devices Data Book .1992 Edition by Harris Semiconductor Corporation

[B6] ANSI/IEEE Standard C37.20.1- 1987, IEEE Standard for Metal - Enclosed Low - Voltage Power Circuit-Breaker Switchgear.

[B7] ANSI/IEEE Standard C37.14 - 1999, IEEE Standard for Low-Voltage DC Power Circuit Breakers Used in Enclosures

[B8] NEMA Standard, Publication No. RI9-1968 (1978), Silicon Rectifier Units for Transportation Power Supplies.

[B9] Tseng WU Liao and Thomas H. Lee “Surge Suppressors for the Protection of Solid-State Devices," IEEE Trans. Ind. Appl., vol. IGA-2, pp.44-52, Jan. /Feb. 1966

[B10] Luke Yu, “ Quick Evaluation of Voltage Surge in Electrical Power Systems,” IEEE Trans. Ind. Appl., vol 31, pp. 379-383 Mar./Apr. 1995

[B11] ANSI/IEEE Standard C62.33-1982, IEEE Standard Test Specifications for Varistor Surge-Protective Devices

[B12] ANSI/IEEE Standard C62.35-1987, IEEE Standard Test Specifications for Avalanche Junction Semiconductor Surge Protective devices.

[B13] Roger C. Dugan, Mark F. McGranaghan, H. Wayne Beaty, “Electrical Power Systems Quality” Chapter 4, McGraw-Hill Edition.

[B14] Jih-sheng Lai and Francois D. MartZloff, " Coordinating Cascaded Surge Protection Devices: High-Low versus Low-High", IEEE Trans. Ind. App. vol. 29, pp. 680-687, July/Aug. 1993

[B15] Nirmal K. Ghai, “Design and Application Considerations for Motors in Steep- Fronted Surge Environments”, IEEE Trans. Ind. App. Vol. 33, pp. 177-187, Jan./Feb. 1997

[B16] Keith W. Eilers, Mark Wingate, “Application and Safety Issues for Transient Voltage Surge Suppressors” IEEE Trans. Ind. App. vol. 36, pp. 1734-1740, Nov/Dec. 2000


1

1

2

34

56

78

910

1112

1314

1516

1718

1920

2122

2324

2526

2728

2930

3132

23

[B17] Paul G. Shade, “Vacuum Interrupters: The New Technology for Switching and Protection Distribution Circuits” IEEE Trans. Ind. App. vol. 33, pp. 1501-1511, Nov/Dec. 1997

[B18] Edward A. Bardo, Kenneth L. Cummins, William A. Brooks, “Lightning Current Parameters Derived From Lightning Location Systems: What Can We Measure?” 18th International Lightning Detection Conference June 6-8, 2004, Helsinki, Finland www.vaisala.com/ILDC2004

[B19] IEEE Committee Report, “A Simplified Method of Estimating Lightning Performance of Transmission Lines”, IEEE Transactions on Power Apparatus and Systems, PAS – 104, pp 919 – 932, 1985.

[B20] Transmission Line Reference Book, 345 kV and Above, Second Edition, Electric Power Research Institute, pp .545-552.

[B21] Dev Paul, “Light Rail Transit DC Traction Power System Surge Overvoltage Protection” IEEE Trans. Industry Application, Vol. 38, pp 21- 28, Jan. /Feb. 2002

[B22] IEEE Draft Standard P1626, “Standard for DC Overhead Contact System Insulation Requirements for Transit Systems”.

[B23] NFPA 780 “Standard for the Installation of Lightning Protection Systems”

[B24] IEC 62305, “Protection against Lightning”, 2006-01

[B25] IEEE Recommended Practice on Surge Voltages in Low-Voltage AC Power Circuits, IEEE Std C.62.41 – 1991, pp 34.

[B26] British Standard BSN EN 50124-1:2001 Railway Applications – Insulation coordination

[B27] Dev Paul “DC Rapid Transit System: OCS Pole Grounding Technical Analysis and Safety, APTA 2004”

[B28] Dev Paul “Lightning Protection Analysis of Light Rail Transit DC Overhead Contact System”, IEEE I&CPS Conference 2005, pp.160-169


1

12

345

678

910

1112

1314

15

1617

1819

20

2122

2324

23

http://www.vaisala.com/ILDC2004

Documents

spastyle.doc - IEEE-SA - Working Groupgrouper.ieee.org/groups/railtransit/ocs/1627_wg/P1627D4.1... · Web view[B13] Roger C. Dugan, Mark F. McGranaghan, H. Wayne Beaty, “Electrical