DP30D

DESIGN PRACTICESELECTRIC POWER FACILITIES

SYSTEM GROUNDING – LOW VOLTAGE

PROPRIETARY INFORMATION — For Authorized Company Use Only

SectionXXX-D

Page

DateEXXON

ENGINEERING

1 of 36

December, 1996

EXXON RESEARCH AND ENGINEERING COMPANY — FLORHAM PARK, N.J.

‘

CONTENTS

Section Page

SCOPE 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

REFERENCES 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

International Practices 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Other Literature 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

BACKGROUND 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

DEFINITIONS 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Charging Current 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Equipment Grounding 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Ground 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

High Resistance Grounded System 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Low Resistance Grounded System 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Low Voltage System 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Solidly Grounded System 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

System Grounding 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

System Neutral 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Ungrounded System 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

THE UNGROUNDED SYSTEM 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

THE GROUNDED SYSTEM 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Neutral Point Grounding 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The Solidly Grounded System 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The High Resistance Grounded System 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Other System Grounding Methods 9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

SOLIDLY GROUNDED SYSTEM DESIGN 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Fault Current Magnitude – General 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Maximum (Bolted) Line-To-Ground Faults 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Adequate Ground-Fault Protection per IP 16-4-1 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Maximum Distance Tables – IP 16-4-1 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Clarifications on Use of Maximum Distance Tables 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

When System Impedance Cannot be Neglected 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Impedance-Related Adjustments of Table Distances 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Examples of Percent Reductions in Maximum Distances 13. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Use of Tables 1A & 1B – Steel Conduit 13. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Use of Table 1A/1B Distance-Data – Without Interpolation or Extrapolation 14. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Interpolation of Table 1A/1B Distance Data 15. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Extrapolation of Table 1A/1B Distance Data 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Use of Tables 2A & 2B – Aluminum Conduit 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Use of Tables 3A, 3B, & 4 – Ground Return Cable or Wire 17. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Ground-Fault Calculations 19. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .


SYSTEM GROUNDING – LOW VOLTAGEPROPRIETARY INFORMATION — For Authorized Company Use Only

SectionXXX-D

Page

DateEXXON

ENGINEERING

2 of 36

December, 1996


CONTENTS (Cont)

Section Page

Methods for Estimating Reff, Xeff, or Zeff 20. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Xeff for Ground-Return Wire or Cable 20. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

GROUND FAULT PROTECTIVE DEVICES 22. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

GROUND FAULT PROTECTION DEVICE APPLICATION 23. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

GROUND FAULT PROTECTIVE DEVICE COORDINATION 24. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PROTECTIVE DEVICE APPLICATION FOR TYPICAL SYSTEM CONFIGURATIONS 24. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Use of Ground Sensor (50GS) Relays 27. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

SYSTEM AND COMPONENT DESIGN AND APPLIC ATION – HIGH RESISTANCE GROUNDED SYSTEM 29. . . . . . . . . .

Fault Current Magnitude 29. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Determining System Charging Current 31. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Ground Resistor Sizing 31. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Ground Fault Detection and Location Devices 32. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

APPENDIX 33. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Examples of Calculating Line-to-Ground Fault Currents 33. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Example 1 33. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Example 2 34. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Example 3 34. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Example 4 34. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Example 5 35. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Example 6 35. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Methods for Calculating Effective Impedance (Zeff) 36. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Zeff for Circuits Using Aluminum Conduit as Ground Return Conductor 36. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Zeff of Circuit in Non-Metallic Conduit or Having Ground Return Conductor within Cable Assembly 36. . . . . . . . . . . . . . . .

FIGURES

Figure 1 Ungrounded System Equivalent Circuit 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 2 Solidly-Grounded System Equivalent Circuit 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 3 High-Resistance Grounded System Equivalent Circuit 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 4 Low-Resistance Grounded System Equivalent Circuit 9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 5 Low-Reactance Grounded System Equivalent Circuit 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 6 Ground Sensor Protection 22. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 7 Residual Ground Fault Protection 23. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 8 Neutral Ground Circuit Protection 23. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 9 Secondary-Selective Substation Ground-Fault Protection 25. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 10 Radial Substation (Dedicated Primary Feeder) – Ground Relaying 26. . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 11 Radial Substation (Tapped Primary Feeder) – Ground Relaying 26. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 12 Use of 50GS Relay for Motors 27. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 13 Coordination of 50GS Relay with Contactor Interrupting Capacity 28. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 14 High Resistance Grounded System Under Normal Conditions 30. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 15 High Resistance Grounded System – Fault on Phase A 30. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 16 High Resistance Grounded System Using Distribution Transformer with Secondary Resistor 32. . . . .

Figure 17 Ground Fault Detection and Location 32. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .




SectionXXX-D

Page

DateEXXON

ENGINEERING

3 of 36

December, 1996


Revision Memo

12/96 Highlights of this revision are:1. Updated “References”.2. Revised “Background”.3. Revised most “Definitions”.4. Revised discussion of charging current limits in “The High Resistance Grounded System”.5. Completely replaced old section entitled “System and Component Design and Application Solidly Grounded Systems”

with new section entitled “Solidly Grounded System Design”, including explaining use of “Maximum Distance Tables”in IP 16-4-1.

6. Revised section on “Fault Current Magnitude” in “System and Component Design and Application – High ResistanceGrounded System”.

7. Completely revised “Appendix”, covering example calculations.



SectionXXX-D

Page

DateEXXON

ENGINEERING

4 of 36

December, 1996


SCOPE

This section covers the low-voltage (600 volts and lower) system ground fault protection practices normally used in Exxon plants.The section includes information on the methods normally used for grounding the low voltage system neutral and on the protectivesystem and component design and application.

REFERENCES

INTERNATIONAL PRACTICES

IP 16-2-1 Power System Design

IP 16-4-1 Grounding and Overvoltage Protection

IP 16-12-1 Switchgear, Control Centers and Bus Duct

OTHER LITERATURE

Beeman, D., “Industrial Power Systems Handbook,” McGraw Hill (1955).

IEEE (Institute of Electrical and Electronic Engineers) Standard 142-1991 “Recommended Practice for Grounding Of Industrial andCommercial Power Systems.”

The National Electrical Code, 1996 Edition (NFPA 70-1996). National Fire Protection Association.

IEEE Transactions on Industry Applications VOL. IA-13 No. 5 Sept/Oct. 1977 “The Reality of High Resistance Grounding,”J. R. Dunki-Jacobs.

IEEE Proceedings 1964 Industrial-Commercial Power Systems Technical Conference “Industrial Power Systems GroundingPractice,” L. W. Manning.

AIEE (American Institute of Electrical Engineers) Transactions Part II, (Applications and Industry) Volume 79, May 1960“Determination of Ground Fault Current on Common Alternating Current Grounded Neutral Systems in Standard Steel or AluminumConduit,” J. A. Geiger, O. C. Davidson and R. W. Brendel.

AIEE Transactions Part II (Applications and Industry) Vol. 73, July 1954, “Iron Conduit Impedance Effects in Ground CircuitSystems,” A. J. Bisson and E. A. Rochau.

Dunki-Jacobs, J. R., “The Impact of Arcing Ground Faults on Low Voltage Powers Systems” General Electric Co., publicationGET-6098 (1970).

Dunki-Jacobs, J. R. and Savoie, P. J., “A Guide to Ground Fault Protection” General Electric Co. periodical “Industrial PowerSystems” December, 1972, March, 1973, June, 1973.

Grissom, S. B., “Grounding of Power System Neutrals” Westinghouse Electric Co. publication Electrical Transmission andDistribution Reference Book, Fourth Edition (1964) Chapter 19.

Shields, F. J., “System Grounding for Low Voltage Power Systems” General Electric Co. publication GET3548.

“System and Equipment Grounding” and “Capacitance Constants” General Electric Co. publication Industrial Power Systems DataBook, Section 3 and Appendix C.

Wagner, C. L., “Effect of Grounding Impedance on the Magnitude of Transient Overvoltage Due to Arcing Ground Faults”Westinghouse Electric Corp. Report No. 60 – 166 (June, 1960).




SectionXXX-D

Page

DateEXXON

ENGINEERING

5 of 36

December, 1996


BACKGROUND

Per Exxon practices, low-voltage system neutrals are usually solidly grounded; but high-resistance grounding may be specifiedwhere permitted by regulations. Proposals for ungrounded systems and for systems grounded through Petersen coils need theapproval of Owner’s Engineer. The National Electric Code of the USA requires solid grounding of low-voltage systems that haveline-to-neutral loads. Ungrounded operation is not normally considered in the design of new low-voltage systems because of therisk of excessive overvoltages during arcing ground faults. Low-resistance grounding is not considered for low-voltage systemsbecause it would require sensitive ground-fault protective devices throughout the system, which would be costly.

Ground-fault protection for Exxon’s solidly grounded low-voltage systems is provided by phase overcurrent devices to the extentpermitted by Exxon practices and by regulatory requirements. Ground-fault relays are installed when phase overcurrent devicesdo not provide adequate ground-fault protection, and when local regulations require ground-fault relays on certain size circuits.

DEFINITIONS

CHARGING CURRENT

For an individual feeder, charging current is the steady-state per-phase current that flows in the feeder due to the inherentcapacitances of the feeder and of any electrical equipment connected to the feeder. For individually-shielded cable conductors,the charging current of the conductor is due only to the line-to-ground capacitive reactance, Xc0, of the conductors. For individualcables that are not shielded, and for bare-conductor overhead lines, both the line-to-ground and the line-to-line capacitances ofthe conductors give rise to charging current. For the purposes of this practice, only the line-to-ground charging current, dueto Xc0, is relevant.

For an electrical system, the system’s steady-state per-phase charging current (herein called Ic0) is the sum of all the chargingcurrents associated with that phase. The mathematical quantity that equals three times the system’s per-phase charging currentto ground is sometimes called the system’s total charging current to ground (1 to 2 amperes for most low voltage systems). Thisquantity (3 times Ic0) is equal to the capacitive component of the fault-point ground-fault current during a line-to-ground fault.

EQUIPMENT GROUNDING

An intentional connection to ground from non-current-carrying metal parts of a wiring system or from non-current-carrying metalparts of other electrical apparatus connected to the electrical system; e.g., metal enclosures, motor frames, etc.

GROUND

The term used for the earth and for a conducting connection to the earth – where “earth” is a voltage reference plane which, in theory,is large enough and conductive enough that currents passing through it do not cause a resistive voltage drop. A connection toground can be intentional as in system and equipment grounding, or it can be unintentional as in a ground fault. The purpose ofan intentional connection to ground is to establish and maintain the potential of earth at the connection point (for reference or safetypurposes), and to conduct ground-fault current into or out of the earth. References to “ground” or “earth” may be to the earth soil,or to a conducting body, in intimate contact with the earth, that has essentially the same potential as the earth; e.g., ground rods,buried lengths of bare wire, buried metallic water pipes, reinforcing bars in building footings, etc.

HIGH RESISTANCE GROUNDED SYSTEM

A system with a high-value resistor intentionally connected between the system neutral and ground. Normally the resistance valueis such that ground fault current is limited to 10 amperes or less.

LOW RESISTANCE GROUNDED SYSTEM

A system which has a low value resistor intentionally connected between the system neutral and ground. Normally the resistancevalue is such that ground fault current is within a range of 25 amperes to several thousand amperes.

LOW VOLTAGE SYSTEM

In this section, a system with line-to-line voltages of 1000 volts or less; e.g., 480 V in the U.S.A., 380 V in Europe.

SOLIDLY GROUNDED SYSTEM

A system grounded through an adequate ground connection in which no impedance has been inserted intentionally; e.g., a systemsupplied by a wye connected transformer whose neutral is connected directly to ground by a metallic conductor.



SectionXXX-D

Page

DateEXXON

ENGINEERING

6 of 36

December, 1996


DEFINITIONS (Cont)

SYSTEM GROUNDING

An intentional connection, either solid or through impedance, between ground and at least one point in the current-carrying partsof the system. The connection point in the system is usually at the neutral of a power transformer or generator. When systemgrounding through impedance is used to limit ground fault current in our plants, the impedance is normally resistive and is sizedto provide either a high-resistance or low-resistance grounded system.

SYSTEM NEUTRAL

A point in the current-carrying circuit of a system that is usually used for grounding purposes; e.g., the neutral of a wye connectedtransformer.

UNGROUNDED SYSTEM

A system without an intentional connection between ground and a point in its current-carrying circuits, except for potential indicatingor measuring devices, or other high impedance devices.

THE UNGROUNDED SYSTEM

An ungrounded system is one which has no intentional connection between the system and ground. The system is coupled toground via the distributed capacitance-to-ground of the system conductors, such as the windings of motors and transformers, andthe cables of the distribution system. Therefore, the so-called ungrounded system is actually a “capacitively” grounded system. Aground fault on one phase of the system will produce a small flow of current at the fault point supplied from the distributedcapacitance of the other two phases. Figure 1 illustrates the ungrounded system circuit.

FIGURE 1UNGROUNDED SYSTEM EQUIVALENT CIRCUIT

The chief advantage of the ungrounded system is that it permits continued operation of a circuit which has a single-line-to-ground fault(the most common fault type) without serious equipment damage or system upset. Ground fault sensing equipment can be used todetect ground faults and to sound an alarm. Orderly shutdowns can be scheduled in order to locate and repair faulted circuits.

There are serious disadvantages associated with the ungrounded system. These are:

� If a ground fault is not located and removed promptly, the possibility exists that the continuous current flow at the ground faultpoint, although small, will escalate the minimum damage fault into a double line-to-ground or three-phase fault which can causeserious damage. A typical maximum value of this small continuous current flow at the fault point for a 480-volt ungroundedsystem is 1 to 2 amperes.

� A ground fault on one line causes full line-to-line voltage to appear throughout the system between the two unfaulted lines andground. This voltage is 73% above the normal value, but conductor and equipment insulation is rated normally to withstandthis value. However, if this voltage is applied for long periods, it may result in failure of insulation which may have deteriorateddue to age or physical damage.




SectionXXX-D

Page

DateEXXON

ENGINEERING

7 of 36

December, 1996


THE UNGROUNDED SYSTEM (Cont)

� Ungrounded systems are vulnerable to line-to-ground transient overvoltages (as much as six times normal). The overvoltagescan place abnormally high stresses on conductor and equipment insulation. These stresses can cause aged insulation to failand new insulation to weaken prematurely. More frequent equipment and cable failures can result, and such failures may occursimultaneously in equipment on different circuits.

� While ground fault detection is relatively simple, ground fault location can be difficult and time consuming. Ground-fault locationtechniques include: de-energizing individual circuits one at a time until the faulted circuit is located; or using a tone or pulsegenerator and a coupling sensor. This methods are generally inconvenient, and location of grounded circuits may be postponedfor long periods.

The well-maintained ungrounded system, in which the first ground fault is promptly located and removed, has the capability forgreater service continuity than the grounded system. However, experience indicates that for most plants the grounded neutralsystem provides adequate service continuity and has other advantages including freedom from transient overvoltages.

THE GROUNDED SYSTEM

NEUTRAL POINT GROUNDING

The most desirable and common method of grounding a system is to make the connection to ground at the system’s neutral point.There are other methods such as line grounding and mid-phase grounding, but these methods are rarely used in new systems andwill not be covered in this practice.

The circuit points for system neutral grounding are obtained in different ways depending on winding configuration of the power sources.In a wye-connected system, the neutral points of the source transformer and/or generator windings are brought outside the equipmentenclosure and connected to ground. For a delta-connected power source or for wye-connected sources where the neutral point isnot available, zig-zag or wye-delta grounding transformers are used to provide the neutral points for system grounding.

Exxon normal practice is to use power transformers with wye-connected secondary windings, or wye-connected generators insystems where there is power generation at the low voltage level. The systems are grounded at the neutral points of thesewye-connected sources.

THE SOLIDLY GROUNDED SYSTEM

In solidly grounded systems, the neutral points of one or more power sources are connected to ground without intentional insertionof an impedance. The magnitude of ground fault current is about the same as three-phase fault current. Figure 2 illustrates thesolidly grounded system circuit.

FIGURE 2SOLIDLY-GROUNDED SYSTEM EQUIVALENT CIRCUIT



SectionXXX-D

Page

DateEXXON

ENGINEERING

8 of 36

December, 1996


THE GROUNDED SYSTEM (Cont)

The advantages of the solidly grounded system are:

� The magnitudes of transient and steady-state line-to-ground overvoltages are kept within safe limits. For systems havingground fault current approximately equal to three-phase fault current, line-to-ground voltages on unfaulted phases during faultsare close to the normal line-to-ground value.

� With properly applied protective devices, ground faults are quickly detected and removed. This fast fault clearing minimizessystem upsets and equipment damage.

� The magnitude of ground current is usually high enough so that most phase protective devices are sensitive enough to detectground faults. As a result, separate ground-fault protection is not always needed.

The disadvantages of solid grounding are:

� Although ground faults are quickly isolated, high magnitudes of ground fault current flow before the fault is cleared. Damageat the fault point can be severe, especially if first line protection fails.

� Ground faults cause service interruptions because faulted circuits are disconnected automatically by the protective devices.

� A severe flash hazard to personnel exists where arcing grounds occur. These arcing ground can be self-sustaining and verydamaging to equipment if protective devices are not sensitive enough to detect them.

Because the solid grounding method successfully controls overvoltages and provides enough fault current to allow protectivedevices to quickly and selectively remove faults, it is the most widely used grounding method for low voltage systems. InternationalPractice 16-2-1 requires that all low voltage systems be solidly grounded unless high-resistance grounding is specified, or unlessalternative methods (ungrounded system or Petersen coil) are approved by the Owner’s Engineer.

THE HIGH RESISTANCE GROUNDED SYSTEM

A high resistance grounded system is achieved by inserting a high ohmic value resistance between one or more system neutralpoints and ground. This resistance is sized to limit ground fault current through the grounding resistance to a value equal to orslightly greater than three times the per-phase system capacitive charging current to ground. The quantity equal to three timesthe per-phase charging current to ground is sometimes called the total charging current to ground (1 to 2 amperes for most lowvoltage systems), and is equal to the capacitive component of the ground-fault current during a line-to-ground fault. Protectivedevices are arranged for detection and alarm. Experience with high resistance grounded systems shows that the total systemcapacitive charging current to ground for the initial installation and including future expansions should not exceed 3.53 A(preferred) up to 7.07 A (the recommended maximum) where faulted circuits are not disconnected automatically by protectivedevices. Figur e 3 illustrates the high resistance grounded system circuit. The circuit labeled “Thevenin Equivalent” is anapproximate equivalent based on the assumption that the system and transformer reactances are very small compared to RNand XCO. High resistance grounding is discussed further later in this practice.

FIGURE 3HIGH-RESISTANCE GROUNDED SYSTEM EQUIVALENT CIRCUIT




SectionXXX-D

Page

DateEXXON

ENGINEERING

9 of 36

December, 1996



The advantages of high resistance grounding are:

� Unscheduled equipment shutdowns caused by line-to-ground faults are avoided because automatic tripping is not used.

� Since very little ground current is allowed to flow, fault point damage is usually small.

� Line-to-ground transient overvoltages, due to inductive-capacitive resonance and repetitive restriking of arcing faults, are limitedto safe values (250% of normal or lower). The effectiveness of the system in limiting resonant overvoltages is a function of theability of the neutral resistance to absorb the energy stored in the system capacitance-to-ground at the instant the fault occurs.

� Arcing grounds are usually self-extinguishing, with limited energy released by the arc.

� Flash hazard for line-to-ground faults is reduced greatly compared to solidly grounded systems.

� The disadvantages of high resistance grounding include:

� Ground fault location can be difficult and time-consuming. However, there are detection systems available which trainedpersonnel can use to shorten fault location time.

� If a second ground fault occurs on another phase before the first fault is located and cleared, a double line-to-ground fault resultswith the subsequent damage and automatic tripping of one or two circuits.

� System insulation is subjected to higher than normal voltage when a single ground fault occurs. System insulation on the twounfaulted phases will see line-to-line voltage as in the ungrounded system, this may result in failure of weak insulation if thefault remains for longer periods.

Although solid grounding is the most widely used low voltage grounding method, high resistance grounding is sometimes usedwhere an immediate service interruption, due to a ground fault, is undesirable.

OTHER SYSTEM GROUNDING METHODS

Low resistance and low reactance grounding of the system neutral are two other methods which can be used on low voltagesystems. However, these methods have not been used in Exxon plants and their use is generally uncommon.

Low resistance grounding is accomplished by inserting a low ohmic value resistance between the system neutral point and ground.The resistance value is selected to limit ground fault currents to 20% of three-phase fault current and lower, with a minimum of about400 amperes. Protective devices are arranged to trip when ground faults occur. Figure 4 illustrates the low-resistance groundedsystem circuit.

FIGURE 4LOW-RESISTANCE GROUNDED SYSTEM EQUIVALENT CIRCUIT

This method has the advantages that transient overvoltages are kept within safe limits and due to the lower fault current and theuse of ground fault protective devices, fault damage is reduced compared to damage on a solidly grounded system.



SectionXXX-D

Page

DateEXXON

ENGINEERING

10 of 36

December, 1996



The disadvantages of the low resistance grounded system are the additional cost of ground fault protective devices required toprovide selective protection, the more severe flash hazard compared with that of a high resistance grounded system, and theunscheduled equipment shutdowns caused by the automatic tripping of faulted circuits.

To date, low resistance grounding has not found wide application in low voltage systems because of the high cost of obtainingselective ground fault protection. This method may find wider acceptance for low voltage applications when small and inexpensiveground fault relays are developed. These relays must be sensitive and compatible with circuit protection used on motor controllers,small feeders, and branch circuits.

Low reactance grounding is accomplished by inserting a low ohmic value reactor between a system neutral point and ground. Thereactor is sized to limit ground-fault current levels to between 25% and 100% three-phase fault levels (usually closer to 100%).Figure 5 illustrates the low reactance grounded system circuit.

FIGURE 5LOW-REACTANCE GROUNDED SYSTEM EQUIVALENT CIRCUIT

This method is not often used in low voltage systems. It does have some applications, however, such as when low voltagegeneration is present. In a solidly grounded system with generators, ground fault current may be larger than three-phase faultcurrent (for close-in, bolted faults on the generator). To reduce fault current, reactance grounding is sometimes used.

Since ground-fault current levels are approximately equal to three-phase fault levels, this method is similar to the solid groundingmethod as far as advantages and disadvantages are concerned.

SOLIDLY GROUNDED SYSTEM DESIGN

FAULT CURRENT MAGNITUDE – GENERAL

In order to design an effective protective scheme for a solidly grounded system, the designer must know the range of possibleground fault currents. This range is generally considered to lie between the maximum bolted line-to-ground fault and aline-to-ground fault with arc resistance at the end of an outgoing feeder. The maximum ground fault current must be checked againstthe line-to-ground fault-interrupting capability of interrupting devices. The minimum fault current due to an arcing ground fault atthe end of a feeder is used to determine the adequacy of the feeder’s ground fault protection.

The general equation for calculating ground-fault current is:

IL � G �

3EL � NZ1 � Z2 � Z0

Eq. (1)

where: IL–G = Single-line-to-ground fault currentEL–N = System line-to-neutral voltageZ1 = System equivalent positive sequence impedanceZ2 = System equivalent negative sequence impedanceZ0 = System equivalent zero sequence impedance




SectionXXX-D

Page

DateEXXON

ENGINEERING

11 of 36

December, 1996


SOLIDLY GROUNDED SYSTEM DESIGN (Cont)

Note: All sequence impedances are in complex form (Z1 = R1 + jX1), and can be in either ohms or per unit, depending on thecalculation. In the general case, Z0 includes the R0 + jX0 of phase conductors and windings between the relevant neutraland the fault point, plus three times any impedances that exist only in the ground return path between the fault pointand the relevant neutral; e.g., 3ZN (neutral grounding impedance), 3ZF (fault impedance), 3RGR (ground return pathresistance). X0 of the phase conductors takes account of the inductive reactance between the phase conductors and theground return path.

The use of this equation or modifications of this equation are discussed below.

Example 1 in the Appendix is a calculation of ground fault current using this equation for a bolted fault at the end of a feeder ona solidly grounded system. Example 1 makes the simplifying assumption that all impedances upstream of the faulted cable aresmall enough that they can be assumed to be zero.

MAXIMUM (BOLTED) LINE-TO-GROUND FAULTS

Although bolted line-to-ground faults rarely occur, they are important because they cause maximum ground current flow. Themagnitude of this current depends on the line-to-neutral source voltage and the impedances of the sequence networks, perEquation 1. For the maximum line-to-ground fault current at a main source bus of a solidly grounded system, the impedances ofthe ground return path are all assumed to be zero. Typically, the maximum ground fault current in our solidly grounded systemsis somewhat greater than the maximum three-phase fault current.

ADEQUATE GROUND-FAULT PROTECTION PER IP 16-4-1

IP 16-4-1 specifies the basis for determining if circuit ground fault protection is adequate, per the following statement:

“The combined impedance of the ground return path and the supply circuit line conductors shall be low enough to insure operationof the circuit overcurrent protective device in less than two seconds on a single line-to-ground fault at the load end of the circuit.An arc voltage of 40 volts in phase with the line-to-neutral source voltage shall be assumed at the point of fault.”

Our general practice is not to provide ground-fault relays on low-voltage circuits if the above basis is satisfied by the phaseovercurrent devices. When ground fault relays are applied, the basis in the statement is almost always satisfied by high-sensitivity50GS relays. The adequacy of ground fault protection should be verified if a 51N relay protects a low-voltage circuit. If localregulations regarding ground fault protection are more restrictive than IP 16-4-1, the local requirements shall be followed. Forexample, some codes may require ground-fault relays on all motors of a certain size or larger.

The adequacy of ground-fault protection in low-voltage solidly-grounded systems should be determined, to the extent possible, viathe use of the “Maximum Distance” tables in IP 16-4-1. These tables are discussed below.

MAXIMUM DISTANCE TABLES – IP 16-4-1

IP 16-4-1 provides tables for use in determining the maximum circuit distances allowed by our practices for various cableinstallations in solidly grounded low-voltage systems. The maximum distances in the tables are based on the “two second, 40-voltarc” criterion discussed above, and are presented as a function of the rating of the overcurrent-device, device operating speed,and the sizes of conduits, cables, and ground-return conductors.

The notes associated with the tables describe how to use the tables; however, clarification is provided below regarding the tablesand their use. (Because the “A” tables deal only in feet and the “B” tables deal only in meters, references to distances in these tablesare presented in only the units used in the specific table.)

Clarifications on Use of Maximum Distance Tables

Clarifications on the use of the “Maximum Distance” tables in IP 16-4-1 are provided as follows:

� The tables assume that the impedance upstream of the cable circuit is negligibly small compared to the cableimpedance. This assumption is usually valid, but if the upstream impedance is not negligible compared to the cable-circuitimpedance, the tables are not directly applicable, and calculations must be performed to take account of upstream impedances.See “When System Impedance Cannot Be Neglected” below.

� The word “Rating” in the heading of each table means “Relay Pickup” (expressed in line-side amperes) when a relay isinvolved. The word “Rating” is correct for fuses and molded case circuit breakers.



SectionXXX-D

Page

DateEXXON

ENGINEERING

12 of 36

December, 1996



� The notes in the tables refer to protective device operation at a nomina l settin g of 10 times their rating. The words “at anomina l setting of” mean “at a curren t of”. For example, a fuse characterized by Facto r B in Table 1 must operate in lessthan 2 seconds at a curren t of � 6 times its rating. In Tables 2 and 3, the words “divide d by devic e setting” at the endof the next to last sentence of Note 1 mean “divide d by the multipl e of devic e ratin g at whic h devic e operate s in justunder 2 seconds.”

� The actua l maximum-total-clearin g tim e characteristic of a protective device must be used to determine the maximumcircuit length allowed under the 2-seconds rule of IP 16-4-1. Note 1 of Tables 2 and 3 indicates how to adjust the tabledistances upward for protective devices that totally clear a fault within 2 seconds at multiples of rating less than 10. The tabledistances would have to be adjusted downwar d if there were an MCCB protecting a feeder (without contactor switching),and the MCCB had a 2-second maximum-total-clearing time at a multiple of rating greater than 10. The distances in Tables1A and 1B canno t be adjusted by the current-ratio method of Note 1 in Tables 2 and 3. Distance data in Table 1A or 1Bmust be interpolated or extrapolated when it becomes necessary to adjust distance data in these tables. Use of the distancetables is discussed below.

WHEN SYSTEM IMPEDANCE CANNOT BE NEGLECTED

Distance data in the “Maximum Distances” tables is based on the assumption that the impedance upstream of the cable is negligible.A ground fault relay must be added if the actual circuit distance is greater than the maximum allowable distance per the tables.However, if the actual circuit distance is less than the distance determined using the tables, the effect of upstream impedances ontable distance should be evaluated if either of the following conditions apply:

� The actual circuit distance is between 90% and 100% of the maximum distance determined using the tables, or,

� The load on a feeder is greater than 10% of the local transformer’s OA capacity. Before doing this comparison, adjust thetransformer OA capacity downward. If the total impedance upstream of the local transformer is more than 5% of the localtransformer impedance; i.e., adjust the transformer capacity downward by the ratio of the transformer’s impedance divided bythe sum of the transformer impedance plus the upstream impedance.

IMPEDANCE-RELATED ADJUSTMENTS OF TABLE DIS TANCES

When the effect of impedance upstream of the cable circuit should be evaluated (per the above), either a short circuit calculationshould be performed (as discussed later), or the following method should be used to adjust distances in any of the tables to accountfor impedance upstream of the cable.

For each maximum distance (Lmax) in any of the tables, there is an inherent total-circuit maximum impedance (Zmax), which is equalto 237 V divided by the 2-second-clearing current used in the table. The 2-second-clearing current used in Tables 2 and 3 is equalto the Device Rating times 10. The 2-second-clearing current used in Tables 1A and 1B is equal to the Device Rating times either10 or 6 or 4, depending on the Trip Setting Factor used to determine Lmax. To the extent source impedance. ZS, upstream of thecable, uses up some of the total-circuit maximum impedance, the maximum cable-circuit distance in the table must be adjusteddownward. The adjusted maximum distance can be closely approximated via the following equation:

L�max � Lmax �1 –ZS

Zmax�

where : Lmax is the maximum distance for a cable circuit found by using the tableL’max is the adjusted maximum distanceZS is the magnitude of the impedance upstream of the cable circuit, expressed in ohms and referenced to

the low-voltage side of the transformerZmax is the magnitude of the maximum allowable impedance upon which Lmax is based. Zmax is equal to 237V

divided by the 2-second-clearing current used in the table (as explained in the paragraph above).




SectionXXX-D

Page

DateEXXON

ENGINEERING

13 of 36

December, 1996



The upstream impedance, ZS, for a line-to-ground fault can be determined via symmetrical components (Z1 plus Z2 plus Z0, alldivided by 3). However, when the impedances upstream of the local transformer are negligible, ZS can be assumed to be equalto ZT (the impedance of the local transformer).

Note: The biggest impedance-related reductions in table distances occur with the smaller transformers and the higher-ratedprotective devices. The following table shows example percent reductions in table distances for two local transformersizes (5% impedance) and two protective device ratings (2-second operation at 10 times rating). The local transformersare assumed to be the only significant upstream impedance.

EXAMPLES OF PERCENT REDUCTIONS IN MAXIMUM DISTANCES

ÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

500 kVA TRANSFORMER ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

1000 kVA TRANSFORMER


125 A Device ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

12% ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

6%


70 A DeviceÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

7%ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

3.5%

The following example illustrates the need for evaluating the effect of upstream impedance on distances found in the MaximumDistance tables:

A 500 kVA, 480 V transformer, with ZT = 5%, has an impedance of 0.023 ohms reflected to the 480 V secondary side. Assume a480V, 100 HP motor, a 2-inch steel conduit ground return, a 400-foot circuit length, and a 150 A protective device that totally clearsa fault in 2 seconds at 1500 A (10 times rating). Assume all other upstream impedances are negligible. The maximum allowablecircuit distance in Table 1A is 450 feet (150 A device, 2-inch conduit, TSF = A); however, the load is more than 10% of thetransformer rating and the effect of the transformer impedance should be checked. The inherent impedance, Zmax, in the table is0.158 ohms (237 V divided by 1500 A). Per the calculation below, the adjusted maximum distance is 384 feet, which is about a15% reduction in allowable distance. Since the actual circuit length exceeds this adjusted limit, a ground fault relay should be added.

L�max � Lmax �1 –ZS

Zmax� � 450 �1 – 0.023

0.158� � 450 (0.854) � 384 feet

The adjustment factor of 0.854 applies only to the distance of 450 feet because Zmax is different for each different distance in anyof the tables.

USE OF TABLES 1A & 1B – STEEL CONDUIT

Tables 1A and 1B are to be used when rigid steel conduit is the ground return conductor. The distances in these tables are a functionof the conduit size, and the protective device’s “Trip Setting Factor” (TSF) and nominal rating – which together define the currenta given device requires to operate within 2 seconds.

The size of the line conductor is not a factor in the steel-conduit distance tables. Note 2 of these tables gives percent increasesin maximum distances when there is a ground-return conductor in the conduit. Distances in Tables 1A and 1B cannot be adjustedper the ratio method described in Note 1 of Tables 2 and 3, because the impedance of steel conduit varies with current.

Depending on actual circuit and protective-device parameters, one of the following uses of the data in Tables 1A and 1B will apply:

� The distance data in the tables may be used without interpolation or extrapolation when it is obvious that a ground faultrelay is either needed or not needed, per the guidelines in the next section, entitled “Use of Table 1A/1B Distance-Data –Without Interpolation or Extrapolation”.

� If it is not obvious from the table whether or not a ground fault relay should be added per the guidelines in the next section, thefollowing procedures should be used:

+ The distance data in the tables should be interpolated if the protective device’s actual 2-second clearing current is withinthe range of 2-second-clearing currents inherent in the table (see below).

+ The distance data in the tables should be extrapolated if the actual 2-second clearing current is outside the range of2-second-clearing currents inherent in the table.



SectionXXX-D

Page

DateEXXON

ENGINEERING

14 of 36

December, 1996



+ In all cases, the distance data in the tables may have to be adjusted if the impedance upstream of the cable is not negligible(as explained above).

USE OF TABLE 1A/1B DISTANCE-DATA – WITHOUT INTERPOLATION OR EXTRAPOLATION

To use the distance data in Tables 1A and 1B without interpolation or extrapolation, use the following procedures and guidelines(substitute “pickup” for “rating” when the protective device is a relay):

1. Determine the actual multiple of device rating at which the circuit’s protective device totally clears a fault in 2 seconds.

2. Given the circuit’s conduit size and phase protective-device rating, compare the actual circuit distance against the distancesfound in Table 1A or 1B, per the following guidelines:

� When there is no data in the table for the device rating and conduit size, interpolation or extrapolation of distance data willbe required, unless otherwise noted.

� If the actual multiple of device rating (per 1. above) is greater than 10:

+ The circuit does need a separate ground fault relay if the actual circuit distance is greater than the distance forTSF = A.

+ If the actual circuit distance is less than the distance for TSF = A, interpolation or extrapolation of the table data willbe required.

� If the actual multiple of device rating is less than 10 and greater than 6, compare the actual circuit distance to thedistances for Trip Setting Factors (TSFs) A and B:

+ The circuit does not need a separate ground-fault relay if the actual circuit distance is less than 90% of the smallerdistance (TSF = A). Note: If there is no distance data for the device rating, the circuit does not need a separateground-fault relay if the actual circuit distance is less than 90% of the “TSF = A” distance for the next smaller conduitsize that has distance data for the given device rating.

+ The circuit does need a separate ground-fault relay if the actual circuit distance is greater than the larger distance(TSF = B).

+ If the actual distance is between the larger distance and 90% of the smaller distance, interpolation is required andthe effect of upstream impedances should be checked.

� If the actual multiple of device rating is less than 6 and greater than 4, compare the actual circuit distance to thedistances in the table for TSF = B and TSF = C:

+ The circuit does not need a separate ground-fault relay if the actual circuit distance is less than 90% of the smallerdistance (for TSF = B). Note: If there is no distance data for the device rating, the circuit does not need a separateground-fault relay if the actual circuit distance is less than 90% of the “TSF = B” distance for the next smaller conduitsize that has distance data for the given device rating.

+ The circuit does need a separate ground-fault relay if the actual circuit distance is greater than the larger distance(for TSF = C).

If the actual distance is between the larger distance and 90% of the smaller distance, interpolation is required and the effectof upstream impedances should be checked.

� If the actual multiple of device rating is less than 4:

+ The circuit does not need a separate ground fault relay if the actual circuit distance is less than 90% of the distancefor TSF = C. Note: If there is no distance data for the device rating, the circuit does not need a separate ground-faultrelay if the actual circuit distance is less than 90% of the “TSF = C” distance for the next smaller conduit size that hasdistance data for the given device rating.

+ If the actual circuit distance is greater than 90% of the distance for TSF = C, interpolation or extrapolation of the tabledata will be required, and the effect of upstream impedances should be checked.

3. For a circuit which already has a ground fault relay, with an operating time of 2 seconds, or less, at a current that is 4 timespickup or less, perform the following check. Check that the actual circuit distance does not exceed the table distance in thefollowing column and row:

� In the column of the device rating that equals (or next exceeds) the relay’s line-side pickup current; and

� In the row for “TSF = C” for the circuit’s conduit size (or the next smaller conduit size with distance data).




SectionXXX-D

Page

DateEXXON

ENGINEERING

15 of 36

December, 1996



It is unlikely that the actual circuit distance will exceed the table distance, but if it does, calculate the ground-fault current (fromimpedance data), and use it to determine if the relay clears the fault in less than 2 seconds. In the absence of any otherground-return impedance data for the fault calculation, use 0.0006 ohms per foot for cable-circuits in 2-inch and smaller steelconduit, and use 0.0005 ohms per foot for circuits in 2.5-inch and larger steel conduit. If the ground fault protection is found to beinadequate per this analysis, a more sensitive ground fault relay should be used if available.

INTERPOLATION OF TABLE 1A/1B DISTANCE DATA

Interpolation or extrapolation of Table 1A and 1B distance data is required when the limiting distance cannot be determined perthe guidelines in the preceding section. When interpolation is not possible, the data should be extrapolated if accuracy is possible(see the “Extrapolation” section below). If extrapolation is not considered accurate enough, a ground fault calculation must bedone.

The key points regarding interpolation of distance data in Tables 1A and 1B are the following:

� Interpolation or extrapolation of data in Tables 1A and 1B is possible because all of the distances in the tables for a given conduitsize are points on a curve of distance versus 2-second-clearing current, regardless of protective device rating.

� The 2-second clearing current inherent in Tables 1A and 1B for any given distance is equal to the Device Rating (above thedistance) multiplied by the factor (10, 6, or 4) associated the Trip Setting Factor (left of the distance).

� Interpolation of the distance data inherent in Tables 1A and 1B for a given conduit size is possible when the actual2-second-clearing current of a protective device is within the range of 2-second-clearing currents inherent in the table for thatsize conduit.

� The range of 2-second-clearing currents inherent in the tables for a given conduit size is from 4 times the smallest device ratingwith distance data, up to 10 times the highest device rating with distance data.

� Given the actual protective device’s 2-second-clearing current, it is important to interpolate between the two nearest data points;i.e., interpolate between the distances whose 2-second-clearing currents are nearest to the device’s actual 2-second-clearingcurrent. The two nearest 2-second-clearing currents in the table may not be on the same TSF line. Prior to interpolating betweentwo distance-versus-current points, check that the two nearest data points have been found by ensuring that there is no distancein the table (for the given conduit size) that is between the two distances chosen.

For example, for a 3-inch steel conduit and a device with a 2-second-clearing current of 1450-A, the nearest currents in the tablefor a 3-inch conduit are 1400 A (350 A rating, TSF = C, 690 feet) and 1500 A (250 A rating, TSF = B, 660 feet). Since there is noother distance between 690 and 660 feet for 3-inch conduit, it is certain that the two nearest data points have been selected forinterpolation. With the actual 2-second clearing current of 1450 A being midway between 1400 A and 1500 A, interpolation yieldsthe maximum distance of 685 feet, midway between 690 feet and 660 feet.

Even when there is no distance data in the table for a given device rating and conduit size, interpolation is possible when theactual 2-second-clearing current of the device falls within the range of 2-second-clearing currents inherent in the table for the conduitsize. For example, a 300 A device protecting a circuit that is in a 4-inch steel conduit has no explicit distance data in Tables 1A(or 1B). However, there is inherent distance data for 4-inch conduits in Tables 1A and 1B for any device that has a 2-second-clearingcurrent between 1600 A (4 times 400 A) and 6000 A (10 times 600 A).

The two scenarios which follow illustrate use of the distance versus current data inherent in Tables 1A and 1B:

� Assume a 175 A device totally clears a fault in 2 seconds at 1800 A, and that the associated circuit is 565 feet long in a 4-inchsteel conduit. There is no distance data in Table 1A for this scenario. Since 1800 A is exactly 4 times a 450 A rating (for whichthere is distance data), the maximum distance of 630 feet can be found at the intersection of the 450 A device-rating andTSF = C. Since the actual circuit distance is less than 90% of the table distance, the effect of upstream impedance does nothave to be checked, and no ground fault relay is needed.

� Assume the 175 A device in the above scenario totally clears a fault in 2 seconds at 1900 A. Since 1900 A is not a 4, 6 or 10times multiple of any device rating in the table, the maximum distance must be found by interpolation between the distancesassociated with the two operating currents nearest to 1900 A. In this case, the interpolation must be done between 630 feetfor 1800 A (4 times 450 A), and 590 feet for 2000 A (4 times 500 A). These two points are the nearest to the devise operatingcurrent of 1900 A because there is no distance in the table for 4-inch conduit that is between 630 feet and 590 feet. Since 1900A is midway between 1800 A and 2000 A, interpolation yields the limiting distance of 610 feet, which is midway between 630 feetand 590 feet. Because the actual circuit distance of 565 feet is greater than 90% of the distance found using the tables, theeffect of upstream impedances should be checked:



SectionXXX-D

Page

DateEXXON

ENGINEERING

16 of 36

December, 1996



+ Assume a 1000 kVA transformer with an impedance of 0.0115 reactive ohms, and a heater with impedance of 0.002resistive ohms. The resultant upstream impedance magnitude is 0.0117 ohms. The maximum impedance for a clearingcurrent of 1900 A is 0.125 ohms (= 237V/1900A). The distance adjustment factor is 0.9064 (= 1 – 0.0117/0.125) per theequation in the section (above) entitled “When System Impedance Cannot Be Neglected”. Therefore the adjustedmaximum allowable distance is 553 feet (= 0.9064 times 610 feet). Since the actual circuit distance is 570 feet, a groundfault relay must be added.

+ If the transformer in this example had a 2000 kVA rating, with a resultant upstream impedance of 0.007 ohms, theadjustment factor would be 0.944 (= 1 – 0.007/0.125), and the adjusted maximum distance would be 576 feet. Since theactual circuit distance is 565 feet, there is no need for a ground fault relay.

EXTRAPOLATION OF TABLE 1A/1B DISTANCE DATA

When a device’s 2-second-clearing current is outside the relevant data range in Tables 1A and 1B, graphical extrapolation of thedistance versus current data can provide reasonable results.

USE OF TABLES 2A & 2B – ALUMINUM CONDUIT

Tables 2A and 2B indicate maximum allowable circuit distances when rigid aluminum conduit is used as the ground return. Thedistances in these tables are a function of the conduit size, the protective device’s nominal rating, and the size of the line conductor.The distances are for protective devices that operate in two seconds at a current of 10 times the device rating.

Where the distance is blank for a given aluminum-conduit size, line-conductor size, and device rating, multiply any one of thepublished distances (for the given conduit and line size) by the ratio of the “device rating” for that published distance, dividedby the actual device rating or actual relay pickup (in line-side amperes). For example, if a 300 A device protects a 95 square mmline conductor in a 2-inch conduit, there is no distance entry for this situation. Therefore, multiply the Table 2B distance of 285 meters(for a 250 A device, 2-inch conduit, and 95 square mm line) by 250/300.

If the protective device operates in 2 seconds at a higher multiple of rating than 10 times rating, the distance per the table shouldbe decreased via multiplication by the ratio of 10 divided by the higher multiple of rating.

Protection is adequate if the actual circuit distance is less than the applicable distance in the tables, or is less than an adjusteddistance derived from the tables, as discussed in the following paragraphs. Distances in the tables can be adjusted upward whenthe protective device operates in less than 2 seconds at a current ten times its rating or pickup or setting.

If a phase overcurrent device does not provide adequate protection based on the actual or adjusted table distance, a ground faultrelay must be added to the circuit. The adequacy of protection provided by a ground fault relay must be checked in the same wayas for a phase fault device, using the ground-fault relay’s line-side pickup (for 51N) or line-side setting (50 GS) in place of the “DeviceNominal Rating” in the tables.

If the circuit distance is greater than the distance per the table and the phase protective device operates within 2 seconds at a currentless than 10 times its rating, the distance in the table can be increased via the following adjustment: multiply the table distanceby the ratio of 10 divided by the multiple of device rating at which the device operates within 2 seconds. For example, if a 250 Afuse operates in 2 seconds at 7.33 times its rating, multiply the distance in the table by 1.364 (= 10 � 7.33).

If a smaller line conductor is used than is given in the table for a given size aluminum conduit:

� For the conduit size and device rating, check if the actual circuit distance is greater than the shortest distance (or shortestadjusted distance per previous paragraph) – in which case the protection is not adequate and a ground-fault relay is required.

� If the above check does not establish the need for a ground fault relay, check if the actual circuit distance is less than thepublished (or adjusted) distance for the same size line conductor in the next smaller size conduit (for the same device rating)– in which case the protection is adequate.




SectionXXX-D

Page

DateEXXON

ENGINEERING

17 of 36

December, 1996



� If the actual circuit distance is between the above distances, the designer should perform a calculation to determine the circuitimpedance taking account of the smaller line conductor, and adjust the maximum permissible length accordingly, usingEquation 8 below. Since the combined line and conduit impedance is essentially all resistance, this impedance can be adjustedusing line resistances from manufacturer’s data, or using line resistance data that can be back-calculated from the distancetables by using Equation 7 below.

If the circuit is protected by a ground fault rela y, Note 1 of these tables calls for a distance calculation using actual impedances.Protection is adequate if the actual circuit distance is less than the distance in the table for the actual conduit and line size at thesmallest device rating that equals or exceeds the ground fault relay pickup. If the actual circuit distance exceeds this table distance,perform the following calculation, which uses the impedances inherent in the distance tables. This calculation assumes that theground-fault relay is either instantaneous (50GS) or has a time dial setting (51N) such that the relay operates in 2 seconds or lessat 4 times pickup.

For the actual aluminum conduit size and line size, read any Lmax and its corresponding device rating (IR1) in Table 2A or 2B. ThisLmax is based on an operating current of 10 times the device rating, IR1. If we call IGFR the ground-fault relay pickup or instantaneoussetting, and if 4xIGFR assures relay operation in 2 seconds or less, the calculated maximum length, LGFR, for the ground fault relayapplication is:

LGFR � Lmax x10 � IR14 � IGFR

Eq. (2)

where : LGFR is the calculated maximum distance allowed with a ground fault relay.Lmax is any maximum distance per Table 2A or 2B for the size conduit and line conductor in the circuit.IR1 is the device rating (in primary side amperes) corresponding to Lmax in the table.“10” is the multiple of device rating on which the table is based.IGFR is the ground fault relay pickup (or instantaneous setting) in line-side amperes.“4” is the multiple of IGFR that ensures relay operation (accounting for CT error, etc.).

For example, for a 1-inch conduit, a 70 A MCCB, and a 10 Awg line conductor, the maximum distance from Table 2A is 310 feetfor a 70 A device (which operates within 2 seconds at 10 times its rating). If the actual circuit distance is 1600 feet, a ground faultrelay, 50GS, must be added. Assume the 50GS relay picks up at 10 A (1 A relay setting with a 50/5 CT).

The highest published distance for 1-inch conduit and a 10 Awg line is 1070 feet for a 20 ampere device. Since the circuit distanceis 1600 feet, we need to do a calculation of the maximum allowable distance using Equation 2. Using Lmax = 1070 feet andIR1 = 20 A from Table 2A, and IGFR = 10 A for the 50 GS:

1070 x 10 x 204 x 10

� 5350 feet

Therefore, the ground fault relay provides adequate protection.

USE OF TABLES 3A, 3B, & 4 – GROUND RETURN CABLE OR WIRE

Tables 3A and 3B are for circuits not in steel or aluminum conduit, with the ground return conductor within the cable assembly. Thedistances in these tables are a function of the protective device’s nominal rating, and the sizes of the line and ground returnconductors. The distances are for phase protective devices that operate in two seconds at a current of 10 times the device rating.Table 4 provides adjustment factors to decrease the Table 3A and B distances if the ground return conductor is outside of the cableassembly. These adjustment factors are a function of the spacing between the ground return conductor and the phase conductor.

In the rest of this section, references to Table 3A and 3B distances mean the distances adjusted by Table 4 factors ifapplicable.

If the protective device operates in 2 seconds at a higher multiple of rating than 10 times rating, the distance in the table shouldbe decreased via multiplication by the ratio of 10 divided by the higher multiple of rating. For example, if a device operates in2 seconds at 11 times its rating, multiply the table distance by 10/11.



SectionXXX-D

Page

DateEXXON

ENGINEERING

18 of 36

December, 1996



Where the distance is blank for a given line-conductor size, ground-return conductor size, and device rating, multiply any one ofthe published distances (for the line size and ground-return conductor size) by the ratio of the “device rating” for that publisheddistance, divided by the actual device rating or actual relay pickup (in line-side amperes). For example, a 50 A device protects a25 square mm line conductor with a 25 square mm ground-return conductor, but there is no distance entry for this situation. Thereis a distance of 225 meters published for a 70 A device, and 25 square mm line and ground-return conductors. Therefore, multiply225 meters by 70A/50A to determine that the maximum allowable distance for 25 square mm line and ground-return conductorsis 315 meters when the circuit is protected by a 50 A device that operates in 2 seconds at 10 times its rating.

Protection is adequate if the actual circuit distance is less than the applicable distance in the tables, or is less than an adjusteddistance derived from the tables, as discussed in the following paragraphs. Distances in the tables can be adjusted upward whenthe protective device operates in less than 2 seconds at a current ten times its rating or pickup or setting.

If a phase overcurrent device does not provide adequate protection based on the actual or adjusted table distance, a ground faultrelay must be added to the circuit. The adequacy of protection provided by a ground fault relay must be checked in the same wayas for a phase fault device, using the ground-fault relay’s line-side pickup (for 51N) or line-side setting (50 GS) in place of the “DeviceNominal Rating” in the tables.

If the circuit distance is greater than the distance in the table and the phase protective device operates within 2 seconds at a currentless than 10 times its rating, the distance in the table can be increased via the following adjustment: multiply the table distanceby the ratio of 10 divided by the multiple of device rating at which the device operates within 2 seconds. For example, if a 250 Afuse operates in 2 seconds at 7.33 times its rating, multiply the distance in the table by 1.364 (= 10 � 7.33).

If the ground return conductor is larger than any in the table for a given size line conductor:

� First check if the actual circuit distance is less than the longest published distance (or longest adjusted distance per theprevious paragraph) for the line conductor size and device rating – in which case the protection is adequate and a ground-faultrelay is not required.

� If the check above does not determine that the protection is adequate, the designer should perform a calculation to determinethe impedance due to the larger ground-return conductor, and adjust the maximum permissible length accordingly, usingEquation 8 below. Since the combined line and ground-return conductor impedance is essentially all resistance, thisimpedance can be adjusted using line resistances from manufacturer’s data, or using line resistance data that can beback-calculated from the distance tables by using Equation 7 below.

If the circuit is protected by a ground fault rela y, Note 1 of these tables calls for a distance calculation using actual impedances.Protection is adequate if the actual circuit distance is less than the distance in the table for the actual line and ground-returnconductor sizes at the smallest device rating that equals or exceeds the ground fault relay pickup. If the actual circuit distanceexceeds this table distance, perform the following calculation, which uses the impedances inherent in the distance tables. Thiscalculation assumes that the ground-fault relay is either instantaneous (50GS) or has a time dial setting (51N) such that the relayoperates in 2 seconds or less at 4 times pickup.

For the actual line and ground-return conductor sizes, read any Lmax and its corresponding device rating (IR1) in Table 3A or 3B.This Lmax is based on an operating current of 10 times the device rating, IR1. If we call IGFR the ground-fault relay pickup (orinstantaneous setting), and if 4xIGFR assures relay operation in 2 seconds or less, the calculated maximum length, LGFR, for theground fault relay application is obtained using Equation 2, repeated below:

LGFR � Lmax x10 � IR14 � IGFR

Eq. (2)

where : LGFR is the calculated maximum distance allowed with a ground fault relay Lmax is any maximum distance per Table 3A or 3B for the line and ground-return conductor sizes in the circuit.IR1 is the device rating (in primary side amperes) corresponding to Lmax in the table.“10” is the multiple of device rating on which the table is based.IGFR is the ground fault relay pickup (or instantaneous setting) in line-side amperes.“4” is the multiple of IGFR that ensures relay operation (accounting for CT error, etc.)




SectionXXX-D

Page

DateEXXON

ENGINEERING

19 of 36

December, 1996



For example, for a 70 A MCCB, and 10 Awg line and ground-return conductors, the maximum distance from Table 3A is 160 feetfor a 70 A device (that operates within 2 seconds at 10 times its rating). If the actual circuit distance is 1600 feet, a ground fault relay,50GS, must be added. Assume the 50GS relay picks up at 10 A (1 A relay setting with a 50/5 CT).

The highest published distance for 10 Awg line and ground-return conductors is 560 feet for a 20 ampere device. Since the circuitdistance is 1600 feet, we need to do a calculation of the maximum allowable distance using Equation 2 :

560 x 10 x 204 x 10

� 2800 feet

Therefore, the ground fault relay provides adequate protection since the circuit distance is 1600 feet.

GROUND-FAULT CALCUL ATIONS

IP 16-4-1 indicates that the calculation of ground fault current, when required, should involve the use of specific circuit impedances,and that the fault clearing time should be determined for the calculated current and the characteristic of the actual protective device.If specific impedance data is not available, this design practice provides a method for estimating the circuit impedance.

When calculation is required for an end-of-circuit ground fault with a 40-volt arc voltage drop, Equation 1 would be modified asfollows for a rigorous solution:

IL – G �

3 x (EL – N – 40)

Z1 � Z2 � Z�0

Eq. (3)

where: The variables are in amperes, volts and ohms.The 40-volts arc drop at the fault point is subtracted from the line-to-neutral voltage, with bothat the same phase angle of zero-degrees.Z1 and Z2 are the system positive and negative sequence impedances. (In the general case, thesystem includes the cable and the upstream source impedances.)Z0’ = Z0 – 3ZF,The components of Z0’ are as follows:+ Per above, Z0’ does not include 3ZF, the fault impedance term, the effect of which has

been accounted for by the 40-volts drop in the numerator.+ 3ZN is zero in this case because the system is solidly grounded.+ Z0’ includes three times the circuit’s ground-return-path cold resistance, RGR.+ Z0’ includes the hot resistance of the phase conductor, RPC, (as does Z1 and Z2).+ Z0’ includes the zero-sequence inductive reactance of circuit components, including (in

the general case) the X0 of relevant source generators and transformers, and the X0Cbetween the cable phase conductors and the ground return path. See Equation 4.

Equation 3 can be used when impedance data for the installation is available or can be reasonably estimated, and should be usedwhen the source impedance upstream of the cable is not negligible.

When the sum of the sequence impedances of the source system (upstream of a cable) is very small compared to the sum of thesequence impedances of the cable (say 1:10), the source impedance can be assumed to be zero with reasonable accuracy, andonly the cable sequence impedances need to be used in Equation 3. This would result in the following equation:

IL – G �

3 x (EL – N – 40)

3 RPC � 3 RGR � j (X1C � X1C � X0C)Eq. (4)

where: The denominator of Equation 4 represents the sum of the sequence impedances of the cable andits ground return path (the source impedances having been assumed to be negligibly small).

RPC is the phase conductor hot a-c resistance, from the source to the fault.RGR is the ground return conductor cold a-c resistance, from the fault back to the relevant source neutral.X1C is the positive sequence reactance of the cable, which is equal to X2C, the negative sequence

reactance of the cable.X0C is the zero sequence inductive reactance of the cable/ground-return circuit.



SectionXXX-D

Page

DateEXXON

ENGINEERING

20 of 36

December, 1996



Example 2 in the Appendix uses Equation 4 .

If RGR and/or X0C for the cable installation is not available, the following calculation techniques may be employed to estimate thecable/ground-return impedance for use in the denominator of either Equation 3 or Equation 4 .

Rewriting Equation 4 :

IL–G �

EL – N – 40

L �(R2eff

) � (X2eff

)�1�2

�

EL – N – 40

L (Zeff)Eq. (5)

where : L = Length of conductor from source to fault point, meters.Reff = Effective resistance (hot) of one phase conductor from source to fault point plus the resistance

(cold) of the ground return path, ohms per meter.Xeff = Effective circuit inductive reactance, accounting for effect of both the phase conductor and the

ground return path, ohms per meter.Zeff = Effective impedance magnitude of the cable phase conductor and ground return circuit, ohms per

meter.

The fault current, IL–G, calculated using Equations 3 , 4 or 5 would be used to check that the device providing ground-fault protectionoperates quickly enough to satisfy IP 16-4-1 or local regulations.

METHODS FOR ESTIMATING Reff, Xeff , OR Zeff

Methods for estimating Reff, Xeff, or Zeff in Equation 5 (when RGR and/or X0C data are not available) are presented in the followingparagraphs:

Xeff For Ground-Return Wire or Cable

For a ground-return wire within or external to a cable assembly, and not in a metallic conduit/pipe

Reff = RPC + RGR

where : RPC is the a-c resistance (hot) of one phase conductor from source to fault, and RGR is theresistence (cold) of ground return wire from source to fault point, all in ohms per meter.

Xeff = 4�f x 10–7 (0.5 + ln 4 S2

d1 d2) ohms per meter Eq. (6)

where : Xeff = Reactance between two round conductors, with diameters d1 and d2, separated by distance S.This approximates the average of the three sequence reactances of the cable/ground-return circuit.

S = Distance between centers of phase and ground return conductor, meters.d1 = Diameter of phase conductor, meters.f = System frequencyd2 = Diameter of ground return conductor, meters.ln = Natural logarithm (base e)

If the data for Equation 6 is not available, a calculation can be made to extract the impedance used in the IP 16-4-1 distance tables,which impedance can be used for Zeff. Solving Equation 5 for Zeff, and substituting Lmax for L, and “trip current” for ILG:

Zeff �

EL � N � 40

Lmax x trip currentohms per meter Eq. (7)

where : Lmax is the distance found using the distance tables in IP 16-4-1.“Trip current” is the fault current at which the overcurrent-device operates in less than 2 seconds; i.e., the trip

current upon which the distance Lmax is based. For example; for a 50 A device rating, the trip currentis 500 A for “Factor A” in Tables 1A and 1B.




SectionXXX-D

Page

DateEXXON

ENGINEERING

21 of 36

December, 1996



Excluding steel conduit installations, this calculated Zeff (in ohms per unit distance) can then be used in Equation 5 , where itis multiplied by the actual circuit distance, L, to find ILG. ILG found by any of Equations 3 through 6 would then be used to checkthat the circuit’s ground fault relay operates reliably in less than 2 seconds.

If the Zeff per Equation 7 is adjusted to Zeff’ for an alternative circuit condition (such as a different wire size than what is in the table),then a new maximum distance Lmax’ can be calculated by solving Equation 7 for Lmax:

Lmax� �

EL � N � 40

Zeff� x trip currentEq. (8)

Example 4 in the Appendix uses Equation 7 to calculate a Zeff, which is very nearly equal to the Zeff found in Example 3 (whichuses Xeff from Equation 6 ).

For steel conduit systems using the conduit as the ground return conductor.

Reff = Ra–c of the steel conduit only, ohms per meter.

Xeff = Xa–c of the steel conduit only, ohms per meter.

These values may be obtained from manufacturer’s data or the equivalent effective impedance (Zeff) may be calculated by themethod covered in Example 5 in the Appendix.

The use of the steel conduit a-c resistance and reactance as the circuit resistance and reactance has been verified by test. Thisis due to the effect of the encircling magnetic conduit inductive reactance which is greatest when the fault current returns on a remotepath, substantial when it returns over the conduit itself, and still appreciable when a ground return conductor inside the conduit isused.

For steel conduit system with a separate ground wire run inside the conduit:

Reduce the calculated effective impedance (Zeff) as follows:

ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

CONDUIT SIZE*ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

Zeff ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ20 mm and 25 mm

ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ0.9 Zeff of conduit onlyÁÁÁÁÁÁÁÁÁÁ


32 mm, 40 mm and 50 mmÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

0.7 Zeff of conduit only


65 mm, 80 mm and 100 mm ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

0.4 Zeff of conduit only


* inside diameter ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

The factors are derived from AIEE paper “Determination of Ground Fault Current on Common Alternating Current Grounded NeutralSystems in Standard Steel or Aluminum Conduit” listed in References.

For aluminum conduit system using the conduit as the ground return conducto r.

Reff = Ra–c (hot) of one phase conductor from source to fault plus the resistance of the conduit from thesource to fault, ohms per meter.

Xeff = The reactance of one phase conductor from source to fault plus the reactance of the conduit, ohmsper meter.

Where data is not available for Aluminum conduit resistance and reactance, the effective reactance (Zeff) may be calculated by themethod covered in Example 4 in the Appendix , but using Tables 2A and 2B.

Usually, the system using a separate ground wire inside an aluminum conduit is not used since its effect in reducing circuitimpedance is small. If this system is used, the values calculated for the aluminum conduit system should be used. The resultingcalculated ground fault current will be a small amount lower than if the ground return wire effect was included.

With the calculated ground fault current value, circuit protective devices can be applied to meet the 2 second maximum requirement.Also, the circuit impedance can be checked to ensure that feeder and ground return conductors are adequately sized to permitsufficient ground fault current to flow for fault clearing.



SectionXXX-D

Page

DateEXXON

ENGINEERING

22 of 36

December, 1996


GROUND FAULT PROTECTIVE DEVICES

There are two types of ground fault protective devices. The first type is the dual function protective device which protects for phasefaults as well as ground faults. The fused switch or molded case circuit breaker used alone or as part of a combination magneticmotor starter and low voltage power circuit breakers are typical of this type device. It should also be noted that power circuit breakersare available with solid state selective trip elements which provide phase and ground fault protection as part of a single solid statepackage. The second type are protective relays such as the ground sensor (50 GS device) or the residually connected ground faultrelay (51N or 50N/51N device). These devices are used specifically for ground fault protection. Their use is usually limited to lowvoltage power circuit breaker applications such as motor controllers, transformer secondary breakers and feeder breakers.However, use of a protective ground relay is required in other circuits if the phase device will not provide a clearing time of 2 secondsor less.

The following describes various types of ground fault protective relays:

� Ground sensor protection (50 GS) – the most common type is an instantaneous ground overcurrent relay supplied by a zerosequence type current transformer. As shown in Figure 6, the current transformer surrounds all phase conductors (and neutralconductor, if used). Under normal operating conditions, the currents add vectorially to zero, and no signal is sent to the relay.However, a ground fault causes an unbalance, which is sensed by this over-current relay. Ground sensors can also be providedwith time delayed tripping where required.

FIGURE 6GROUND SENSOR PROTECTION

� Residually connected ground fault protection – a time overcurrent relay (51N) is connected in the current transformer circuitas shown in Figure 7 so that only the residual current flows through the relay operating coil. Under normal conditions, currenttransformer currents are balanced and the residual current is zero. When a ground fault occurs, an unbalance results, and therelay operates from the flow of residual current. The residual connection is seldom used to supply an instantaneous groundrelay (50N) since transient unbalanced currents can cause undesired relay operation. The high current inrush conditions typicalof motor starting produce such transient unbalances. One special application for the residually connected 50N relay is an aground fault blocking relay in the automatic transfer circuit of secondary selective substations. Sometimes, a 50N element isused in combination with a residually connected 51N relay and set higher than the expected error currents.




SectionXXX-D

Page

DateEXXON

ENGINEERING

23 of 36

December, 1996


GROUND FAULT PROTECTIVE DEVICES (Cont)

FIGURE 7RESIDUAL GROUND FAULT PROTECTION

� Neutral ground circuit protection – a time overcurrent relay (51G) supplied from a current transformer located in the neutral toground connection of a transformer or generator (see Figure 8 ). The relay measures current as it returns to the power source.Normally, this relay provides first line protection for the zone between the transformer secondary (or generator winding) andthe transformer secondary (or generator) circuit breaker. It also provides backup protection for breakers connected to the buswhich the transformer or generator supplies.

FIGURE 8NEUTRAL GROUND CIRCUIT PROTECTION

GROUND FAULT PROTECTION DEVICE APPLICATION

International Practice 16-2-1 , “Power System Design,” and International Practice 16-12-1 , “Switchgea r, Control Centers,and Bus Duct,” specify applications where separate ground fault protection is mandatory or preferred.

1. Transformer relaying: Neutral backup relaying (51G) should be applied to all substations 500 kVA or larger. For substationshaving low resistance or solidly grounded neutrals, the relay should trip the transformer primary breaker or controller; andshould alarm only at substations with high resistance grounding. (IP 16-2-1, par. 5.27).

2. Feeder relaying: Ground fault protection, energized residually or from a zero sequence current transformer is required orpreferred as stated below for feeder breakers requiring relaying (IP 16-2-1, par. 5.35).

� Required in substations with low resistance grounded neutrals.

� Preferred in all substations with solidly grounded neutrals.

3. Feeder relaying: Ground fault relays should be instantaneous if energized from a zero sequence current transformer and if nofuses or relays exist downstream which can sense the same ground fault. Otherwise time delay relays should be used, definitetime type or the least inverse characteristic available (IP 16-2-1, par. 5.36).



SectionXXX-D

Page

DateEXXON

ENGINEERING

24 of 36

December, 1996


GROUND FAULT PROTECTION DEVICE APPLICATION (Cont)

4. Backup phase and ground fault protection is required for all fault locations, except for ground faults in transformer mainsecondary connections with impedance grounding of the secondary. Backup protection for ground faults in transformer mainsecondary connections need not be sensitive to faults with large arc voltage components (IP 16-2-1, par. 5.41).

5. Motor relaying: Ground fault relaying is required for low voltage breaker-controlled motors on low-resistance or solidly groundedsystems. The relay should be instantaneous and supplied from a zero sequence current transformer (IP 16-12-1, par. 8.23).

International Practice 16-4-1 – “Grounding and Overvoltage Protection” – in par. 7.6 states that “the combined impedance of theground return path and the supply circuit line conductors shall be low enough to insure operation of the circuit overcurrent protectivedevice in less than 2 seconds on a single line-to-ground fault at the end of the circuit. An arc voltage of 40 volts in phase with theline-to-ground source voltage shall be assumed at the fault point” If phase devices cannot meet this requirement, separate groundfault protection is required.

The National Electrical Code (NEC) 1996 Edition in Section 230-95 covering “Ground-Fault Protection of Equipment” states:“Ground fault protection of equipment shall be provided for solidly grounded wye electrical services of more than 150 volts to groundbut not exceeding 600 volts phase-to-phase for each service disconnecting means rated 1000 amperes or more.

1. Setting. The ground fault protection shall operate to cause the service disconnect to open all ungrounded conductors of thefaulted circuit. The maximum setting of the ground fault protection shall be 1200 amperes and the maximum time delay shallbe one second for ground fault currents equal to or greater than 3000 amperes. Exception No. 1: The ground-fault protectionprovisions of this section shall not apply to a service disconnect for a continuous industrial process where a non orderlyshutdown will introduce additional or increased hazards. Exception No. 2: The ground-fault protection provisions of this sectionshall not apply to fire pumps.”

In the usual Exxon plant system which has a medium voltage distribution system supplying low-voltage substations, there is no“service” or “service disconnecting” means which fits the National Electrical Code definitions. The transformer secondary circuitand circuit breaker are closest to meeting the Code definitions. Sometimes ground relay settings exceeding the 1200 amperemaximum setting and the one second maximum time at 3000 amperes or higher fault values are required to achieve selectivity withprotective devices closer to the fault. These settings meet NEC requirements per the “Exception” to Section 230-95(a).

GROUND FAULT PROTECTIVE DEVICE COORDINATION

Ground fault relay coordination is accomplished mainly on the basis of time. The ground device located furthest downstream is seton a low pickup setting with an instantaneous response. Upstream devices are also set on low pickup but with time delay to allowthe downstream device to clear the fault first.

For coordination purposes, the solidly grounded low voltage system can usually be treated separately from its medium voltage supplysystem because most transformers have delta-connected primaries, and wye-connected secondaries. A line-to-ground fault on thesecondary side of the transformer appears as a line-to-line fault on the primary side of the transformer. Therefore, ground relay settingsfor the low voltage system do not have to be selective with ground relays in the medium voltage system, because the medium voltagesystem’s ground relays do not sense ground faults on the secondary side of the delta-wye-connected transformer.

Depending on the neutral grounding, this may not be the case for wye-wye connected transformers and coordination betweenprimary and secondary ground relays may be required.

International Practice 16-2-1 , par. 5.39, specifies feeder relay selectivity requirements between phase and ground devices:“Selectivity between upstream ground fault relaying and downstream phase fault relaying is unnecessary if there is ground faultrelaying at the downstream location. If there is none, selectivity is required up to at least 0.3 seconds and should be sacrificed above0.3 seconds only if necessary to achieve reasonable sensitivity of upstream ground fault protection.” Figures 12A and 12B illustratethese conditions.

PROTECTIVE DEVICE APPLICATION FOR TYPICAL SYSTEM CONFIGURATIONS

The substation configurations most encountered in refinery electrical systems are:

� Secondary selective substation

� Radial substation (dedicated primary feeder)




SectionXXX-D

Page

DateEXXON

ENGINEERING

25 of 36

December, 1996


PROTECTIVE DEVICE APPLICATION FOR TYPICAL SYSTEM CONFIGURATIONS (Cont)

� Tapped radial substation (tapped primary feeder)

Ground relaying schemes for these three configurations follow, with relay descriptions and coordination requires for each:

� Secondary Selective Substation – Figure 9 shows the ground relaying used for this type of substation.

FIGURE 9SECONDARY-SELECTIVE SUBSTATION GROUND-FAULT PROTECTION

+ 50N/51N: Induction overcurrent relay, inverse, 0.5 – 4 or 0.5 – 2.5 amp time, 0.5 – 4. or 1 – 4 amp instantaneous, 2 – 16or 2 – 6 amp time, 2 – 16 or 2 – 8 amp instantaneous.

Typical relays include GE IAC51 and Westinghouse CO-6 definite minimum time. The instantaneous unit is used to blockautomatic transfer on bus faults or uncleared feeder faults. It should be set above the maximum non-fault current expectedbefore a transfer, and also above motor contribution to a fault on the incoming line or transformer.

The time unit is set to coordinate with the highest set downstream relay that provides ground fault protection (somenon-coordination is allowed per IP 16-2-1, par. 5.39).

+ 51G: Induction overcurrent relay, inverse, 0.5 – 4 or 0.5 – 2.5 amp time or 2 – 16 or 2 – 6 amp time (no instantaneous). Typicalrelays include GE IAC51 and Westinghouse CO-6. This relay provides ground fault protection for the transformer secondarywinding and connections to the switchgear and also backs up the 51N relay for ground faults on or downstream of theswitchgear main bus. It should be selective with the 51N. The 51G trips the primary breaker through a lockout relay 86T.

� Radial Substation (Dedicated Primary Feeder): Figure 10 shows the typical ground relaying scheme for a radial substation.



SectionXXX-D

Page

DateEXXON

ENGINEERING

26 of 36

December, 1996



FIGURE 10RADIAL SUBSTATION (DEDICATED PRIMARY FEEDER) – GROUND RELAYING

The 51N and 51G relays are the same type as used for the secondary selective substation. In this case, they do not have to beselective because substation service will be interrupted if either relay or both sense a ground fault. Note this differs from thesecondary selective substation practice where selectivity is required between 51N and 51G. The 51N can be omitted unless:(a) future expansion to secondary selective is expected, or (b) primary feeder will become a tapped feeder in the future.

� Radial Substation (Tapped Primary Feeder): Figure 11 shows this substation and feeder arrangement.

FIGURE 11RADIAL SUBSTATION (TAPPED PRIMARY FEEDER) – GROUND RELAYING




SectionXXX-D

Page

DateEXXON

ENGINEERING

27 of 36

December, 1996



The 51G should be selective with the 51N because the 51G trips the primary breaker interrupting service to all substations on the feeder.The 51N trips the secondary breaker for downstream faults on the substation and does not interrupt service to the other substations onthe feeder.

USE OF GROUND SENSOR (50GS) RELAYS

Ground sensor relays are often used to supplement the usual phase protection on feeders for the following reasons:

1. To meet the International Practice requirement of tripping for ground faults in 2 seconds or less.

2. To provide more sensitive protection for large, or critical motors to limit damage.

3. It may be desirable to use a 50GS relay on large motors to make the upstream protection more sensitive. As shown inFigure 12(A) , the 51N relay is set to meet the selectivity requirements of IP 16-2-1, par. 5.39. Figure 12B shows that by usinga 50GS for motor ground fault protection, the upstream 51N no longer has to be selective with the fuse. The 51N setting islowered which means it provides faster, more sensitive protection for the main bus and can still be set to be selective with 50GS.

FIGURE 12USE OF 50GS RELAY FOR MOTORS

Figure 12(B) shows how addition of 50GS allows full selectivity for ground faults on a motor circuit.

When 50GS ground sensor relays are used with combination starters having fused contactors, the contactor’s interruptingcapability must be considered. The fuse characteristic and 50GS setting may be such that the 50GS will trip the contactor whenfault currents are in excess of the contactor’s maximum interrupting rating. Figures 13(A) and 13(B) show an example of this.



SectionXXX-D

Page

DateEXXON

ENGINEERING

28 of 36

December, 1996



FIGURE 13COORDINATION OF 50GS RELAY WITH CONTACTOR INTERRUPTING CAPACITY

IC is the maximum interrupting capacity of the contactor. The crosshatched area of Figure 13(A) shows the “dander range,” wherethe 50GS will attempt to trip the contactor before the fuse has a chance to operate. The currents in this range are beyond thecontactor’s interrupting capacity. Figure 13(B) shows that the problem becomes less severe if time delay is added to the groundsensor relay.

This application can be handled in one of the following ways:

� Time delay the ground sensor relay in order to minimize the danger range. However, some of the benefits of fast ground faultprotection are lost.

� Use a contactor with adequate interrupting capability.

� Use a combination starter with molded case circuit breaker. The circuit breaker, equipped with a shunt trip operated by 50GS,would interrupt the ground fault current instead of the contactor.

� Reduce the ground return impedance, if possible, so that the fuses operate in 2 seconds or less making the use of 50GSunnecessary.




SectionXXX-D

Page

DateEXXON

ENGINEERING

29 of 36

December, 1996


SYSTEM AND COMPONENT DESIGN AND APPLIC ATION – HIGH RESISTANCE GROUNDED SYSTEM

FAULT CURRENT MAGNITUDE

In high-resistance grounded systems, the neutral grounding-resistance component of ground-fault current is designed to beslightly higher than the capacitance-to-ground component of ground-fault current. The basis for this design is to limit themagnitude of overvoltages due to ground faults. The neutral grounding resistance and the system capacitive-reactance-to-groundare normally high enough that the other system impedances are negligible in the calculation of ground fault current forhigh-resistance grounding.

Recommended practice for high resistance grounded systems is that this system should not be used if its application results in aline-to-ground fault current, IG, exceeding 10 amperes; however, a maximum fault current level of 5 amperes is preferred, and anyapplication where the total ground fault current exceeds 7.8 A should be carefully considered. Since the neutral resistor is sizedso that the current through it during a line-to-ground fault is slightly higher than the system’s total charging current, the line-to-ground

fault current value is about 2� times the resistor current. As used herein, the term “system total charging current” means three timesthe system’s per-phase charging current to ground.

With the resistive and capacitive currents 90 degrees out of phase and approximately equal in magnitude (per Figure 15B ), theresistive and capacitive components of total ground-fault current (IG) must each be limited to 3.53 A to achieve IG = 5 A. Likewise,they must each be limited to 5.5 A to achieve IG = 7.8 A, and to 7.06 A to achieve IG = 10 A.

ICO is defined herein as the per-phase capacitive charging current to ground in an unfaulted system. The magnitude of each ofthe currents flowing through the capacitors in Figure 14 is ICO. During a line-to-ground fault, the capacitive component of the totalground-fault current is equal in magnitude to three times ICO. The vector sum of IcB + IcC in Figure 15B is equal in magnitude tothree times ICO.

The future maximum per-phase charging current to ground for a system cannot exceed 1.18 A if the total ground fault current(IG) is to be limited to 5 A. In this case the system total charging current (three times ICO), which equals the capacitive componentof fault current, would be limited to 3.53 A (3 times 1.18) and the grounding resistance fault current would also be 3.53 A, resultingin a 5 A total fault current. Likewise, the future maximum per-phase charging current to ground for a system cannot exceed 2.35A if the total fault current is to be limited to 10 A.

In low voltage systems, the range of the system total capacitive charging current is 0.1 to 2.0 amperes per 1000 kVA of systemcapacity. The total capacitive charging current for most low voltage systems would be expected to fall in the 1 to 2 ampere range,well below the 3.53 A limit for a 5 A total ground fault current. Therefore, most high resistance grounded low-voltage systems willhave line-to-ground fault current levels well within the 5 ampere preferred limit.

Figure 14A shows a simple high resistance grounded system. XcA, XcB, and XcC are the system capacitive reactances to groundof phases A, B, and C, respectively. Under normal conditions, IcA, IcB and IcC are equal in magnitude and 120� out of phase witheach other as shown in Figure 14B and IR = 0. The magnitude of total system charging current is:

Ic � | IcA | � | IcB | � | IcC | � 3 IcA �

3 VL � NXcA

Figure 15A shows the system with a ground fault on phase A. IcA is zero (shorted by the fault). The system charging current, Ic,is now the sum of IcB + IcC. Ground fault current, IG, is equal to the vector sum of the neutral resistor current, IR, and system chargingcurrent Ic, as shown in Figure 15B .

If the system total charging current is 1A and R is sized so that:

IR = 1A, IG would therefore be 1.41A.



SectionXXX-D

Page

DateEXXON

ENGINEERING

30 of 36

December, 1996


SYSTEM AND COMPONENT DESIGN AND APPLICATION – HIGH RESISTANCE GROUNDED SYSTEM (Cont)

FIGURE 14HIGH RESISTANCE GROUNDED SYSTEM UNDER NORMAL CONDITIONS

FIGURE 15HIGH RESISTANCE GROUNDED SYSTEM – FAULT ON PHASE A




SectionXXX-D

Page

DateEXXON

ENGINEERING

31 of 36

December, 1996


SYSTEM AND COMPONENT DESIGN AND APPLIC ATION – HIGH RESISTANCE GROUNDED SYSTEM (Cont)

DETERMINING SYSTEM CHARGING CURRENT

The magnitude of the system charging current depends on the size, type, length and number of system elements which areinherently capacitively coupled to ground. The elements are insulated cables, overhead lines, transformer and rotating machinewindings, and surge protection and power factor correction capacitors. In the usual industrial system, cable capacitors and motorwindings are the major elements.

The designer can calculate the charging current values for specific system elements either from manufacturer data, from publishedtypical values and formulas for calculating typical values, or from data obtained by performing tests on the system. An exampleof one such test can be found in Section 32, “System Capacitance Data,” General Electric Co. publication, Industrial PowerSystems Data Book, listed in REFERENCES. Capacitance constants data for motors, cable and systems, and other componentsare covered in Appendix C of the same publication.

GROUND RESISTOR SIZING

The resistance of the neutral grounding resistor is selected so that the current through it during a line-to-ground fault is slightlyhigher, about 5%, than the system’s total charging current including any allowance for the system’s future growth. Usually, a tappedresistor with a resistance range of about 3:1 is provided to permit final selection of the resistance value closest to that determinedby after installation field measurement of a system’s actual charging current.

The nominal resistance of the neutral grounding resistor can be determined from the following:

R � 0.95 x(EL � N)

Ic

where : R = Resistance, hot, in ohmsEL–N = System line-to-neutral voltageIc = Total system charging current, in amperes (three times ICO).

Since the neutral resistor will carry current until the ground fault is located and removed, the resistor must be rated to carry theneutral current continuously without excessive temperature rise. The normal power rating of the resistor can be determined fromthe following:

W �

(EL � N)2

R

where : W = Power in wattsR = Resistance, hot, in ohmsEL–N = System line-to-neutral voltage

Figure 16 illustrates an alternative method for high resistance grounding which sometimes is less costly than the method shown inFigure 15 , where the grounding resistor is directly connected to the system’s neutral point. A single-phase distribution transformeris used with its primary winding connected to the system’s neutral point and the grounding resistor connected across the distributiontransformer’s secondary winding. The resistor used would have the same watt rating as the resistor for Figure 15 , but is reducedin ohmic value by the square of the transformer’s turns ratio. Details of this method are found in Design Practice XXX-B .



SectionXXX-D

Page

DateEXXON

ENGINEERING

32 of 36

December, 1996


SYSTEM AND COMPONENT DESIGN AND APPLICATION – HIGH RESISTANCE GROUNDED SYSTEM (Cont)

FIGURE 16HIGH RESISTANCE GROUNDED SYSTEM

USING DISTRIBUTION TRANSFORMER WITH SECONDARY RESISTOR

GROUND FAULT DETECTION AND LOCATION DEVICES

Ground faults are detected by a sensitive voltage relay (64) connected across the neutral resistor. When ground current flows, avoltage drop across the resistor is produced. The relay senses this and sounds an alarm. A contactor is used to short intermittentlypart of the resistor in order to produce a pulsing current which makes it possible to trace the fault with an ammeter type detector.The resistor and detection equipment is available as a manufacturer’s package unit.

FIGURE 17GROUND FAULT DETECTION AND LOCATION

The following sensing methods are used to locate ground faults:

� Permanent ammeters provided as part of the distribution switchgear metering for main feeder circuits.

� Portable clip-on ammeters for individual branch circuits.

The number of permanent sensing meters versus portable meters is an installed cost versus maintenance cost tradeoff. The morepermanent meters installed, the quicker fault location will be. The owner’s requirements should be obtained before the design is fixed.




SectionXXX-D

Page

DateEXXON

ENGINEERING

33 of 36

December, 1996


APPENDIX

EXAMPLES OF CALCUL ATING LINE-TO-GROUND FAULT CURRENTS

Note: The following examples make the simplifying assumption that the impedances upstream of the cable circuit to be analyzedcan be assumed to be zero because they are very small compared to the cable impedances. This is an acceptablyaccurate assumption when the sum of the resistances of the cable’s phase and ground-return conductors is aboutten times the impedance upstream of the faulted cable.

A direct buried 3-phase feeder circuit from the turnaround power center in a 480 V, 60 Hertz substation has the followingcharacteristics, which are used in Examples 1 through 4:



PHASE CONDUCTORÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

GROUND RETURN CONDUCTOR


Size ÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

4/0 Awg ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

No. 4 Awg



ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

ÁÁÁÁÁÁÁÁDiameter ÁÁÁÁÁÁÁ

ÁÁÁÁÁÁÁ16.4 mm ÁÁÁÁÁÁÁÁÁÁÁ

ÁÁÁÁÁÁÁÁÁÁÁ9.1 mm

ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

Resistance, RGR,

A.C. @ 25�C’

ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

— ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

.000850 �/m


Resistance,RPC,

A.C. @ 75�C’ÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

.000209 �/m ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ

—

ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁReactance X1C

ÁÁÁÁÁÁÁÁÁÁÁÁÁÁ.000081 �/m

ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ—ÁÁÁÁÁÁÁÁ

ÁÁÁÁÁÁÁÁReactance X0C

ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ.000147 �/m

where : RGR is the resistance of the ground return pathRPC is the resistance of the resistance of the phase conductorsX1C is the positive sequence reactance of the phase conductorsX0C is the zero sequence reactance between the phase and ground-return conductors

The data above was obtained from the Westinghouse “Transmission and Distribution Reference Book” page 70 and “The SimplexManual” pages 5 – 16, 5 – 20, and 5 – 22. (See REFERENCES.)

The spacing between the phase and ground return conductor is .01159 meters and the circuit length is 150 meters.

Example 1

Calculate the fault-current magnitude for a bolted line-to-ground fault at the load end of the cable using positive, negative and zerosequence impedances Equation 1 . The source impedance upstream of the cable is considered so small that it is taken as zeroin this example.

Z1 = Z2 = R1 + jX1 = RPC + jX1C = .000209 + j .000081 �/m

Z0 = R0 + jX0 = RPC + 3RGR + j X0C = .000209 +3(.000850)+ j .000147 �/m

IL – G =3EL – N

�Z1 � Z2 � Z0 � 3RGR� � L

�

3EL – N

�(3RPC � 3RGR) � j (X1 � X2 � X0)� � 150

=3EL – N

�(3RPC � 3RGR) � j (X1C � X1C � X0C)� � 150

=

3 x 4803�

[3(.000209) � 3(.00085) � j(.000081 � .00081 � .000147)] � 150

= 831.38(.003177 � j.000309) � 150

�831.38

.4765 � j.04635

IL–G = 1736 amperes



SectionXXX-D

Page

DateEXXON

ENGINEERING

34 of 36

December, 1996


APPENDIX (Cont)

Example 2

Calculate the fault-current magnitude for an arcing line-to-ground fault at the load end of the cable using positive, negative and zerosequence impedances Equation 4 (source impedance assumed to be zero).

IL – G =3(EL – N – 40)

3RPC � 3RGR � j(X1C � X1C � X0C)

=

3 x (4803�

– 40)

[.000209 � .000209 � .0000209 � 3(.00085) � j(.000081 � .000081 � .000147)] � 150

= 711.38(.003177 � j.000309) � 150

�711.38

.4765 � j.04635

IL–G = 1485 amperes

Example 3

Calculate the magnitude of arcing ground fault current at the load end of the cable using Equations 5 and 6:

Xeff = 4�f x 10–7 (0.5 � ln 4S2

d1d2) Ohms�meter

Xeff = (12.5664)(60) x 10–7(.5 � ln4 x 106(.01159)2

16.4(9.1)) ��m

Xeff = .000134 �/m

ILG =EL – N – 40

L�(R2eff

) � (X2eff

)�1�2

�

EL � N � 40

L�(RPC � RGR)2 � (Xeff)2�

1�2�

EL � N � 40

Zeff

IL – G =

4803�

– 40

150�(.000209 � .000850)2 � (.000134)2�1�2

IL – G = 2370.1601

� 1480 amperes

Note that by back-calculation, the equivalent X0 is 0.00024 �/m per the above, versus the published value of 0.000147 per the datafor these calculations. The answers in Examples 2 and 3 are almost equal because the cable resistance predominates.

Example 4

Use the method of Equation 7 to back out Zeff from Table 3 of IP 16-4-1, and compare it the 0.1601 ohms calculated above:

Using Table 3A, the maximum distance for a 4/0 line conductor and a #4 Awg ground-return conductor is 490 feet for a 150 A deviceoperating at 10 times rating (i.e., trip current = 1500 A).

Per Equation 7 :

Zeff �

EL – N – 40

Lmax x trip current�

277 – 40490 x 1500

� .000323 ��ft � .001065 ��m

Zeff = .001065 �/m X 150 m = 0.16 ohms, which is essentially the same answer as in Example 3.




SectionXXX-D

Page

DateEXXON

ENGINEERING

35 of 36

December, 1996


APPENDIX (Cont)

Example 5

The following example illustrates a calculation using Table 1A or 1B to determine the adequacy of a phase-fault device to provideground-fault protection for a circuit that uses steel conduit as the ground return conductor.

Assume three circuits having three different circuit lengths of 400, 500, and 600 feet. Each circuit is in its own 3-inch steel conduit,and each circuit is protected by a 300 A phase overcurrent device that operates in 2 seconds at 2200 A (7.333 times rating).

The distance data for 300 A devices which protect circuits in 3-inch steel conduit, are based on three different 300 A devices whichoperate in 2 seconds at currents of 3000 A (Factor A), 1800 A (Factor B), and 1200 A (Factor C). Since the impedance of steelconduit varies with current, and the limiting current of 2200 A in our example does not match any of these currents, we cannot directlyuse the impedances inherent in the 3-inch-conduit/300-A-device data.

However, for two of the circuits in this example, data for the 300 A device can be used to determine the adequacy of protection:

� A slower device than ours allows a circuit distance of 480 feet (Factor A ), therefore the faster speed of our device will protectthe 400-foot circuit, and a ground-fault relay is not needed.

� A faster device than ours provides protection only up to 580 feet, therefore neither it nor our slower speed device can protectthe 600-foot circuit, and a ground-fault relay is required.

Rule: When a protective device trips in 2 seconds at a multiple of current between two of the multiples in the table (e.g., the multiple7.33 falls between Factor A and Factor B ), a ground-fault relay is not needed if the actual circuit distance is less than the smallerdistance (Factor A ), and a ground fault relay is required if the actual circuit distance is greater than the larger distance (Factor B ).

Since the 500-foot circuit falls between the Factor A and Factor B distances, further analysis is required for the 500 circuit.

One potentially easy check is to find a position in the table with operating amperes equal to or slightly higher than the operatingamperes of the actual device (at 2 seconds). If the distance corresponding to this position is greater than the actual circuit distance,the protection is adequate and a ground-fault relay is not required. In this example, the circuit device operates in 2 seconds at2200 A. Since 2200 A divided by 4 (Factor C) is 550 A, and there is distance data for a 550 A device protecting a 3-inch conduit,we have an exact match. The distance at this position in the table is 510 feet. Since the 500 foot-circuit is less than the 510-footmaximum, a ground fault relay is not required.

If the 550 A data had not been in the table, the next higher amperes in the table would be 2400 A at the Factor B (6 times) positionin the 400 A device column. Since the corresponding allowable distance in the table is only 480 feet (versus the actual circuitdistance of 500 feet), and the operating amperes are not an exact match, no firm conclusion could be drawn, and further analysiswould have been necessary.

Such further analysis for the 500 foot circuit in this example would involve using Equation 7 to calculate the impedancecorresponding to currents either side of 2200 A (the device operating current), and interpolating to find an estimated impedanceper foot for 2200 A. In this case, 2400 A ( = 6 x 400 A), and 2000 A ( = 4 x 500 A), equally bracket the 2200 A target. The impedancesfound using Equation 7 are .000206 ohms per foot and .000219 ohms per foot. Using .000212 ohms per foot as the interpolatedvalue, the 500-foot circuit impedance is 0.106 ohms, which produces 2236 amperes of fault current per Equation 5 . Since this ismore current than the minimum of 2200 amperes needed for device operation in 2 seconds, the protection is adequate.

Example 6

For the 600 foot circuit in the previous example, for which a ground-fault relay is required, the following check could be made:

For the given conduit size of 3 inches, the longest distance is 850 feet, corresponding to 1000 amperes of operating current.Assuming a 50GS relay with an operating current of 40 A (4 x 10A pickup), the allowable distance would be significantly higher than850 feet, therefore the 50GS relay provides adequate protection for the 600 foot circuit.

If for some reason this rough check had not been conclusive, the distance for a smaller conduit with a higher operating current couldbe checked. The lowest operating current in Table 1A is 60 A, which is higher than the 40 A operating current assumed for the 50GS.The corresponding distance for this smaller conduit and higher operating current is 1270 feet, therefore the 600 foot circuit isadequately protected by the 50GS.

For systems with a ground return conductor run inside the steel conduit reduce the Zeff value by the factors specified in the textsection “Ground Fault Current and Circuit Impedance Calculations.”



SectionXXX-D

Page

DateEXXON

ENGINEERING

36 of 36

December, 1996


APPENDIX (Cont)

METHODS FOR CALCUL ATING EFFECTIVE IMPEDANCE (Zeff)

Zeff for Circuits Using Aluminum Conduit as Ground Return Conductor

The calculated value is the effective impedance of the phase and ground return conductors circuit.

Table 1B and 2B values are derived from the AICC paper “Determination of Ground Fault Current on Common Alternating CurrentGrounded Neutral Systems in Standard Steel or Aluminum Conduit.”

1. Refer to Table 2B (metric units) IP 16-4-1 which provide maximum permissible lengths for copper conductors run in aluminumconduit as a function of conduit size and protective device trip rating.

2. Follow the procedures specified for calculation of Zeff of steel conduit circuits. Note that Table 2B distances are based on atrip setting factor of 10.

The inclusion of a separate ground return wire inside the aluminum conduit has a small effect on the effective impedance and thevalue calculated for conduit alone may be used. The resulting calculated ground fault current will be a small amount lower than ifthe ground return wire effect was included.

Zeff of Circuit in Non-metallic Conduit or Having Ground Return Conductor within Cable Assembly

The calculated value is the effective impedance of the phase and ground return conductor circuit.

1. Refer to Table 3B (metric units) IP 16-4-1 which provides maximum permissible lengths for copper conductors as a functionof line conductor size, ground conductor size and protective device trip rating.

2. Follow the procedure specified for calculation of Zeff of steel conduit circuits. Note that Table 3B distances are based on a tripsetting factor of 10.

DP30D

Documents

electric power facilities

solidly grounded neutrals

national electrical code

induction overcurrent relay

design practicessection xxx

ensures relay operation

steel conduit varies

maximum permissible length