ELECTRICAL GENERATION PLANT DESIGN PRACTICE

INTERN EXPERIENCE AT

POWER SYSTEMS ENGINEERING, INC.

An Internship Report

by

TING-ZERN JOE LEE

Submitted to the College of Engineering of

Texas A&M University
in partial fulfillment of the requirement for the degree of

DOCTOR OF ENGINEERING

December 1981

Major Subject: Electrical Engineering

 ELECTRICAL GENERATION PLANT DESIGN PRACTICE

INTERN EXPERIENCE AT

POWER SYSTEMS ENGINEERING, INC.

An Internship Report

by

TING-ZERN JOE LEE

Approved as to style and content by:

c g r-
(Member) (Member)

l
| \ X ^ f v v x - -- j V

(Member) (Member)

/ ^ .

(Member (Head oiTDepartment)

December 1981
 ABSTRACT

A survey of the author’s internship experience with Power Systems

Engineering, Inc. during the period September 1980 through August, 1981

is presented. During this one year internship, the author was assigned

to two engineering projects. One involved design of a 480 MW power

plant. The other was the design of a 8.2 MW induction generator for

cogeneration.

The author’s activities during this period can be categorized into

two major areas. First, technically oriented, he designed protective

relaying and SCADA systems for the projects. Secondly, he assisted the

Project Manager in project management activities such as project pro­

gress and cost control.

The intent of this report is to prepare a training manual for PSE

young engineers. It covers both technical guidelines for power plant

design and nonacademic professional codes. Although this report is

primarily written for young engineers, it can also be used as a refer­

ence by older and experienced engineers.

I wish to acknowledge with sincere thanks the following individuals:

Professor A. K. Ayoub, Electrical Engineering Department, Texas A&M

University, for his invaluable guidance throughout the course of my

work;

All my committee members, Dr. C. W. Brice, Dr. Biman Das, and Dr. G.

Schugart for their valuable suggestions;

Mr. E. E. Hogwood, Electrical Engineering Manager, Power Systems

Engineering, Inc., for his supervision and guidance throughout my

internship;

My colleagues, Mr. R. E. Henry and Mr. A. K. Munjal, who gave freely of

their time to review the manuscript and made suggestive comments;

Mrs. Judy Humphrey who graciously provided her competent and profes­

sional assistance with the typing.

Finally, I would like to thank Electrical Power Institute of Texas A&M

University for supporting me as a Research Assistant.

 TO

My Father who gave me so much

CHAPTER 1 INTRODUCTIN 1

1.1 History of Power Systems Engineering, Inc. 2

1.2 Project Management 4

1.3 Scope of Internship Assignments and Report 7

CHAPTER 2 SOME FUNDAMENTALS OF ELECTRICAL POWER SYSTEM DESIGN 9

2.1 Fault Current Calculation 10

2.2 Steady-State Performance of Electrical Conductors 22

2.3- Transient Performance of Electrical Conductors and

Supports 27

2.4 Applications of Instrument Transformers 43

2.5 The R-X Diagram 47

CHAPTER 3 GENERATOR PROTECTION 52

3.1 Philosophy of Generator Protection 53

3.2 System Connection 56

3.3 Fault Protection 57

3.4 Abnormal Operation Protection 64

3.5 Protection of Synchronous Generators 76

3.6 Protection of Induction Generators 81

CHAPTER 4 TRANSFORMER PROTECTION 86

4.1 Philosophy of Transformer Protection 88

4.2 Effects of Magnetizing Inrush on Differential

Protection 89

4.3 Fault Protection 92

4.4 Abnormal Operation Protection 106

4.5 Protection of Power Transformer 109

 PAGE

CHAPTER 5 MISCELLANEOUS ITEMS OF POWER SYSTEM SUPERVISION,

PROTECTION AND CONTROL 109

5.1 Transmission Substation Configurations 124

5.2 Breaker Failure Protection 125

5.3 Typical Schemes for Breaker Failure Protection 127

5.4 Digital Computer Applications on System Supervision

and Control 131

CHAPTER 6 - BUSINESS ASPECTS OF ENGINEERING PROJECTS 144

6.1 The Life Cycle of Engineering Projects 144

6.2 Legal Aspects of Engineering 146

CHAPTER 7 SOME MANAGERIAL CONSIDERATIONS FOR YOUNG ENGINEERS 148

7.1 Character and Personality Considerations 149

7.2 In Relation to Your Work 150

7.3 In Relation to Your Boss 152

7.4 In Relation with Your Associates and Clients 153

CHAPTER 8 SUMMARY AND CONCLUSIONS 154

REFERENCES 157

APPENDIX I 159

VITA 163
 L IS T OF F IG U R E S

PAGE

Figure 1.1 Power Systems Engineering Organization Chart 5

Figure 1.2 Power Systems Engineering, Inc. Engineering Division 6

Figure 2.1 Multiplying Factors for Total Current Rating Basis

Devices 19

Figure 2 .2 Multiplying Factors for Symmetrical Current Rating

Basis Devices 21

Figure 2.3 Shape Correction Factors for Strap Buses 35

Figure 2.4 Ice and Wind Loading on Electrical Conductors 37

Figure 2.5 System Conditions on the R-X Diagram 50

Figure 3.1 Differential Protection for a Wye-Connected

Generator 60

Figure 3.2 Typical Differential Protection Schemes 62

Figure 3.3 Loss of Excitation Protection 67

Figure 3.4 Recommended protection for Synchronous Generators,

Direct-Connected 78

Figure 3.5 Unit Generator Transformer Arrangement 80

Figure 3.6 Overall Protection Scheme of Cross-Compound

Generators 82

Figure 3.7 Protection of an Induction Generator 84

Figure 4.1 Procedures for Transformer Differential Protection

Design 100

Figure 4.2 A Three-Winding Transformer Bank and CTS’

Connection for Differential Protection 101

Figure 4.3 The Time-Current Curves for the Transformer

Differential Relay 102

Figure 4.4 Protection Arrangement for a Generator Transformer 111

Figure 4.5 Protection Arrangement for a Wye/Wye Transformer

with Delta Tertiary (Buried) 112

Figure 4.6 Protection Arrangement for a Wye/Delta Transformer 116

Figure 4.7 Protection Arrangement for a Delta/Wye Transformer 117

Figure 4.8 Protection Arrangement for a Delta/Delta Transformer 118

Figure 4.9 Protection Arrangement for a Wye/Delta Transformer

with a Grounding Transformer 119

Figure 4.10 Protection Arrangement for a Delta/Delta Transformer

with a Grounding Transformer 120

Figure 4.11 Protection Arrangement for a Delta/Wye Transformer

with a Circuit Switcher 121

Figure 4.12 Protection Arrangement for a Delta/Wye/Wye

Transformer 122

Figure 5.1 Simplified Breaker Failure Schemes 126

Figure 5.2 Single Bus Breaker Failure Protection Using One

Timer Each Bus 129

Figure 5.3 Single Bus Breaker Failure Protection Using One

Timer per Breaker 130

Figure 5.4 Breaker-and-A-Half Breaker Failure Relaying Using

One Timer per Breaker 132

Figure 5.5 Breaker Failure Relaying for a Ring Bus 133

Figure 5.6 Generator Pair Protective Scheme 134

Figure 5.7 Substation Protection Scheme 135

Figure 5.8 Functional Block Diagram-RTU System 141

Figure 5.9 Common Control Module 142

Figure A.l Effect of Instant Fault Occurrence on the Fault

Current if 160
 L IS T OF T A B LE S

PAGE

Table 2.1 Rotating Machine Reactance Multipliers 18

Table 2.2 Atmospherical Data for Ampacity Calculation 26

Table 2.3 Short Circuit Forces on Conductors 34

Table 2.4 Bus Conductor Deflection and Stresses Formulas 40

 CHAPTER 1

INTRODUCTION

It is the intent of the Doctor of Engineering degree program to

prepare the student either as a professional and qualified engineer

with strong business and administrative background or as a manager with

enough technical background. The internship, as required in the program,

is designed to expose the student to the real industry environment and

to bridge the gap, if any, between the academic and the industrial

environments.

The objectives of the internship are as follows:

(1) To become familiar with the structure of the organization, or­

ganizational goals and objectives, departmental responsibilities and

functions, project management procedures, technology assessments and

engineering decisionmaking procedures, and employee's performance eval­

uation and motivation.

(2) To make an identifiable contribution in the Electrical Engi­

neering area of projects undertaken by the company.

(3) To gain experience in the non-academic technical and business

activities of the company and to be aware of the industry standards,

ethical practices, and the interactions between the company and the in­

dustrial environments.

In addition, the following considerations have also been taken into

account in selecting a successful engineering firm to intern with:

(1) In Taiwan, it is very important either to improve the effi­

ciency of energy usage or to find new energy sources other than petro­

leum since there are very few resources.

 (2) In Taiwan, being an experienced engineer is almost the only

way to become an engineering manager. Consequently, ever since I en­

rolled in the Doctor of Engineering program, I have been preparing my­

self to be a qualified engineer with sufficient business and administra­

tive background to be an engineering manager; not a manager with an

engineering background. Considering all these factors, I chose to have

my internship with Power Systems Engineering, Inc., a Houston based

professinal engineering and construction services organization.

1.1 History of Power Systems Engineering, Inc.

Power Systems Engineering, Inc., was founded early in 1969 by

Mr. Albert J. Smith and Mr. Thomas C. McMichael, who at that time had

over 50 years of combined experience in the design and installation of

power generation equipment and industrial power systems. Since its

inception, the firm has stamped itself as an innovator and a pioneer of

high efficiency power generation in the power and process industries.

Its first project was the design, construction, and the operation of a

300 MW combined cycle power generation plant, the largest base load

combined cycle gas turbine power plant ever built at that time.

Today the firm provides engineering and project management services

wherever required. Through one of its subsidiaries, Power Operating

Company, the firm also provides services of operating and maintaining

most types of power generating, steam or heating and cooling facilities

for clients. In response to the critical demands imposed on industry

by the shift in fuel cost and availability today, Power Systems Engi­

neering is providing coal, lignite, and nuclear plant technology logis­

tics and other energy-related services.

 The firm's goals are innovation in high efficiency power genera­

tion, being owner-operator of cogeneration and heat recovery systems,

and to maintain a sound and stable financial structure commensurate to

the requirements of the firm's objectives.

Today, the firm's objectives include:

(1) To serve a great variety of energy-related industrial clients

and meet their specific requirements by the most cost effective methods.

(2) To build up a highly qualified business team to develop co­

generation projects.

(3) To support energy intensive industries with energy resources

such as coal, lignite and others.

(4) To continue development of a highly qualified engineering

staff in all disciplines capable of implementing the design, engineering

and construction of energy-related projects.

(5) To become an internationally recognized professional firm.

The firm's philosophy is based on innovation backed by thorough

feasibility studies, conservative design and the use of proven compo­

nents and systems best suited for a particular project. As an engi­

neering consultant, the firm always tries its best to employ procedures

and techniques which are cost effective and result in overall savings

for its clients. Power Systems Engineering has designed and built a

wide range of industrial power and utility projects from small to very

large, complex systems, and has constructed numerous chemical plants,

compressor stations and other chemical and petroleum industry oriented

installations.

Figures 1.1 and 1.2 show the overall structure of the firm, with
deparmental functions. It is clear that the organization form is of

modified matrix type. Generally, one of the department managers is

assigned as the Project Manager for a particular project, depending

upon the nature of the project. This type of organizational structure

has the following advantages for this particular firm:

(1) The project team can respond more quickly and move effective­

ly to the client’s needs.

(2) It is a flexible structure such that each employee can take

more than one responsibility at a time, if required.

(3) Each individual employee takes wider and higher diversity of

assignments and gains broader experience in all phases of project de­

velopment than might be feasible in a larger firm.

(4) A closer personal relationship at the business level is pos­

sible between employees, clients, and the employer.

1.2 Project Management

For any project undertaken, Power Systems Engineering offers ser­

vices covering all phases of project development: definition of the

needs, conceptual development, feasibility studies, environmental per­

mitting and licensing strategy and applications, preliminary engineer­

ing, design engineering, specification of equipment, procurement, cost

control, complete working schedule, financing and leasing, construc­

tion management, start-up, operation and maintenance and others.

Engineering and corporate assignments are directed and oriented

in a project manner. A Project Manager is assigned to act as liaison

between the client representative and all organizational functions in

the firm. This provides a single source of responsibility to the

POWER SYSTEMS ENGINEERING, INC
AND SUBSIDIARIES

Figure 1.1 POWER SYSTEMS ENGINEERING

ORGMIZATION CHART
 POWER SYSTEMS ENGINEERING, INC.
ENGINEERING DIVISION

Figure 1.2 POWER SYSTEMS ENGINEERING, INC. ENGINEERING DIVISION

client such that the client’s project can receive the maximum support

from all corporate services. A responsible project engineer or engi­

neers from each discipline as needed are assigned to handle all engi­

neering requirements and are responsible to the Project Manager. Under

the project engineer are design, drafting, and logistic teams. Each

individual project engineer also acts as liaison with the client for

his discipline through the Project Manager.

1.3 Scope of Internship Assignments and the Report

During a one year internship, I was assigned to two project teams

for the following projects:

(1) Engineering and design of a 480 MW combustion gas turbine gen­

eration plant in Saudi Arabia.

(2) Engineering and design of a 8.2 MW induction generator for co­

generation in southern Texas.

My contributions to the firm included project electrical engineer­

ing design as well as engineering services. As an electrical engineer,

I conceptually and detail designed protective relaying for the 480 MW

power plant, wrote specifications of equipment, designed microprocessor-

based supervision control and data acquisition (SCADA) and sequence of

events (SOE) systems. I also implemented relay coordination and ven­

dor's bid evaluations, both technically and economically. As an assis­

tant to the engineering service manager, my accomplishments included

designing a complete engineering progress control chart for electrical

engineering discipline, helping the project engineers to process and

expedite most of the procurement, and assiting the project manager to

prepare cost control charts and monthly progress reports.

 Another assignment during my internship was to standardize both

technical and managerial guidelines for young engineers. This report

summarizes and highlights the results of this assignment. It consists

of two parts. Part one, including Chapter 2 through Chapter 5, covers

technical fundamentals of electrical generation plant design. Part two,

including Chapters 6 and 7, discusses business aspects of engineering

projects and some engineering managerial considerations for young

engineers.
 CHAPTER 2

SOME FUNDAMENTALS OF ELECTRICAL POWER SYSTEM DESIGN

As stated in Chapter 1, the scope of engineering business includes

marketing, planning, management, scheduling, control, and all technical

activities. In this chapter only technical design tools such as short

circuit current calculation, ampacity of electrical conductors, tran­

sient performance of conductors, instrument transformers, and the R-X

diagram, will be covered.

Among.these, the short circuit current calculation is the most im­

portant. It includes maximum and minimum current calculations under

fault conditions. Both of these calculations will be discussed in Sec­

tion 2.1 in detail. The selection of electrical conductors may gener­

ally be made on the basis of voltage drop, power loss, maximum allowable

temperature rise, or mechanical strength. Each of these characteris­

tics will be covered under Sections 2.2 and 2.3 separately. Instrument

transformer performance and the R-X diagram are two key topics in power

system relaying. They will be elaborated on under Sections 2.4 and

2.5 separately. Although this chapter and the following three chapters

emphasize only technical aspects of engineering practices, by no means

is it implied that other aspects are not important and can be neglected.

In fact, it should be pointed out that economic considerations are

usually at least as important as technical requirements. For example,

in designing the substation buses to resist short circuit stresses, an

economic balance should be reached by providing means for limiting

possible short circuit forces by increasing bus spacing, or by changing

bus arrangements. Frequently even landscaping becomes a deciding

factor. Similarly, the application of protective relays to the system

components is essentially a compromise between the cost of protection

and the risk of no protection at all.

2.1 Fault Current Calculation

It is not feasible to design an electrical system completely free

of faults. Even the best designed electrical system occassionally e x ­

periences short circuits resulting in high fault currents. Therefore,

protective devices such as circuit breakers and fuses should be applied

to isolate faults before any severe damage or long service interruption

occurs. On the other hand, other parts of the system should be capable

of withstanding the mechanical and the thermal stresses resulting from

maxi m u m flow of fault currents through them. The magnitude of fault

currents are usually established by calculation. Most of the system

components are specified and selected using the calculation results.

2.1.1 Two Simplifying Assumptions

The basic relationship used in determining fault current is O h m ’s

Law. There are several methods or algorithms for fault current calcu­

lation. However, certain simplifying assumptions are customarily made

for engineering design purposes [1]. An important assumption is that

the fault is bolted, that is, it has zero impedance. It should be

recognized, however, that actual faults often involve arcing, which

reduces the fault current magnitude due to arc resistance or impedance.

Furthermore, a three-phase fault is generally assumed because this

type of fault generally results in the m a x imum short circuit current

available in a circuit. In high voltage systems the three-phase fault

is frequently the only one calculated. In a low voltage system, ground

currents are often less than the rated load currents, but can be very

destructive. This condition leads to the requirement for line-to-ground

fault calculation.

Impacts of these two assumptions on engineering design are not only

simplifications of the calculation but also increases of safety margins

on all the equipments selected on the basis of the calculation results.

2.1.2 Fault Current Calculation Results and Their Applications

It is well known that an electrical fault current is usually of an

asymmetrical wave form whose magnitude varies with time after the

fault. Then a question like "When should the fault current be calcu­

lated?" arises naturally.

According to IEEE Standard 141-1976, short circuit currents may be

calculated at the following recommended times:

(1) The first cycle: First cycle maximum symmetrical values are

always required. They are often the only values needed for low voltage

systems and for fuses in general. These values are usually related to

circuit breaker momentary (closing and latching) capabilities.

(2) From 1.5 to 4 cycles: Maximum values are required for high

voltage circuit breaker applications. These values are generally re­

lated to contact interrupting capacities of high voltage circuit

breakers.

(3) About 30 cycles: The reduced magnitude of the fault current

gives the minimum available fault current supplied by the system.

These minimum fault currents are important in protective device coor­

dination in the system.

Generally speaking, the maximum fault current values are used to

select adequate interrupting ratings and to check the ability of system

components to withstand both mechanical and thermal stresses. The

minimum values are needed to check the sensitivity requirements of pro­

tective relays. For overall coordination of system protective devices,

both the maximum and the minimum values are often required. Sometimes,

the minimum values are estimated as fractions of the maximum values for

the sake of simplicity. In this case, only the maximum magnitude of

fault current should be calculated.

2.1.3 Sources of Fault Current

Power frequency currents which flow during a fault come from elec­

trical rotating machines. Rotating machines involved in fault calcula­

tions may be divided into the following categories:

(1) Synchronous generators

(2) Synchronous motors and condensers

(3) Induction machines

(4) Electric utility system

The fault current from each rotating machine source is limited by

the impedance of the machine and the impedance between the machine and

the fault. The impedance of a rotating machine is not a simple value,

but is complex and varying with time. For purposes of fault current

calculations, industry standard practice [1 ] has established the follow­

ing guidelines for the application of this variable reactance.

For synchronous machines, the subtransient reactance (X^j) should

be used to determine fault current during first cycle after fault

occurs; the transient reactance (X£), to determine fault current

after several cycles (about 5 cycles at 60 Hz); the synchronous

reactance (X^), to determine the current flow after a steady-state con­

dition is reached. As most fault protective devices, such as circuit

breakers and fuses, operate well before steady-state conditions, is

seldom used in calculating fault currents for applications of these de­

vices. For induction machines, only the subtransient reactance (X^)

is used. This value is about equal to the locked-rotor reactance.

Therefore, the fault current from an induction machine is approximately

equal to the full voltage in-rush current of the machine, if shorted at

the terminals.

Generally speaking, the current contributed to a fault from the

utility system appears to be merely a small increase in load current

to the remote and relatively large central station generators, and this

current contribution tends to remain constant. Consequently, the uti­

lity system is usually represented by a constant voltage source behind

a constant impedance whose value is determined as follows:

cvi_i ) 2
Z syg (ohm)
Available short circuit level in MVA from
the utility sytem at the point of connection

Where Vi_]i is the system line-to-line operating voltage in KV.

2.1.4 Fault Curent Calculation Procedure [1]

The procedures for calculating system short circuit currents con­

sist of the following steps:

(Step 1) Prepare system one-line diagram.

This diagram should show all sources of short circuit currents and all

significant circuit elements.

(Step 2) Collect and convert impedance data to a common base for

for all significant circuit elements.

A significant part of the preparation for a short circuit current study

is establishing the impedance of each circuit element, and converting

them to the same base. Sources of impedance values are nameplates,

handbooks [3], standards [1 ] [4], manufacturers' catalogs, direct

contact with the manufacturer and others.

The major elements of impedance include transformers, busways,

cables, conductors, and rotating machines. There are other circuit

impedances such as those associated with circuit breakers, instrument

transformers, bus connections and structures, etc., which are usually

negligibly small in high voltage (above 1 KV) systems. However, in

low voltage systems, particularly at 208V, omission of these impedances

may result in over-conservative results. All the impedances, except

those for rotating machines, are assumed to be constants throughout

the fault current calculation. The impedances of rotating machines

depend upon the purpose of the study as discussed in further detail

under Step 5.

Also, it is normal practice to disregard all static loads such as

lighting, electric heating and so on in the network despite the fact

that they are in parallel with other network branches. This is so be­

cause system load currents are usually negligibly small compared with

the high fault current. This assumption implies that from a system

operating point of view, the fault is essentially a disturbance around

the prefault operating point. In other words, the prefault system

voltage (Eo) can be taken as the system nominal voltage at the moment

fault occurs. Therefore, when using the per-unit system, the driving

voltage is equal to 1.0 per unit if the voltage bases are equal to
system nominal voltages. This will greatly simplify the calculation

in Step 5.

(Step 3) Prepare the system impedance diagram.

Impedance information gathered in Step 2 should be entered on the one

line diagram obtained from Step 1. For many short circuit studies at

medium or high voltages, and for a few at low voltage, it is sufficient­

ly accurate to ignore resistances and use reactances only. However, for

many low voltage calculations resistance should not be ignored because

the calculated currents would be unnecessarily over-conservative.

Resistance data are definitely required for calculations of X/R

ratios when applying high and medium voltage circuit breakers, but they

should be kept on a separate drawing called resistance diagram [1 ].

(Step 4) Combine impedances from sources to fault location.

This step is the most straightforward one, but it may be most time con­

suming too.

(Step 5) Calculate short circuit currents to anticipated fault

locations.

This step uses different impedances for the rotating machines to calcu­

late the short circuit currents depending upon the purpose of the

study. Industrial applications of short circuit calculation results

can be classified into the following four general categories:

(a) First cycle duties for fuses and low voltage (below 1 KV)

circuit breaker.

(b) First cycle duties for high voltage circuit breakers (1 KV

and above).

(c) Interrupting duties for high voltage circuit breakers.

 (d) Short circuit currents for time delayed relays.

Each of these categories will be discussed subsequently in further

detail.

For first cycle duties of fuses, switches, and low voltage circuit

breakers, the short circuit current should be calculated in accordance

with ANSI C37.13, ANSI C37.41, NEMA AB, and NEMA SG latest standards.

With low voltage fault calculations, all motors should be considered

regardless of horsepower size. However, the standards allow using an

equivalent reactance to represent a group of low voltage induction or

synchronous motors fed from a low voltage substation. The equivalent

reactance is 0.25 p.u. based on the lump sum KVA rating of the machines.

For other machines, whether generators or motors, the subtransient re­

actance should be used in the equivalent network.

After the first four steps above, a single driving impedance Zf fo r

each fault location can be obtained. Then the symmetrical short circuit

current can be calculated as follows:

^■sym = So p.u.
Z f

For those low voltage circuit breakers whose short circuit ratings

are expressed in symmetrical rms currents, this is the minimum required

capacity for satisfactory performance. For those whose ratings are

given in terms of total rms currents, the symmetrical short circuit cur­

rent obtained above should be multiplied by a multiplication factor ac-

y
cording to ANSI Standard C37.13. This factor depends upon the f/R f
y
ratio at fault location. If f/R^ ratio is equal to or less than 6 .6 ,
y
the factor is 1. If f/Rf is greater than 6 .6 , a straight factor of

1.15 can generally be used for purposes of selecting the breakers.

For fuse application, the multiplication factor should be obtained from

ANSI Standard C37.41.

For first cycle duties of high voltage circuit breakers, the cal­

culation shall be implemented in accordance with the latest ANSI C37.5

and C37.010 standards. With the calculations, only motors with 50 HP

or more ratings should be considered. Table 2.1 lists the reactances

for rotating machines that should be used in the calculation.

After the first four steps stated above, a single impedance Xf for

each fault location can be obtained. Then the first cycle asymmetrical

short circuit duties can be calculated as follows:

Itot = — x 1 - (>
Xf

This is the short circuit duties for comparison with high voltage (above

1 KV) circuit breaker momentary ratings (total current rating basis) or

closing and latching capabilities (symmetrical current rating basis).

There are significant differences between total current and symmet­

rical current calculations of high voltage circuit breaker interrupting

duties. However, in both calculations, only motors rated 50 HP or more

should be considered. The reactances used for the rotating machines

is shown in Table 2.1. For these calculations, the resistance network

is also necessary. In the resistance network each rotating machine

resistance must be multiplied by the same factor as its corresponding

reactance multiplier from Table 2.1. Consider first the duties for

the high voltage circuit breaker rated on the total current basis.

Reduce the reactance network to a single equivalent reactance X f . De-

Y T7
termine f/Rf and o/Xf ratios. Then select the multiplying factor for

f/Rf correction from curves shown in figure 2.1 [5]. To use the curves
 Power S ystem s
ROTATING m a c h i n e reactance multipliers
E ngineering, Inc.
P.O. Box 19398
Houston. Texas 77024
JL JO B NO. N/A DATE

T Y P E OF R O T A T I N G
MACHINE MOMENTARY INTERRUPTING

All turbine generators: all

hydrogenerators with amortisseur
windings, all condensers 1.0 X"d 1.0X"d

Hydrogenerators without amortisseur

windings 0.75 X"d 0.7 5 X"d

All synchronous motors 1.0 X Md 1.5 X"d

Induction Motors ( = 1800 RPM)

Above 10 00 HP 1.0 X"d 1.5 X"d

250 < HP 1000 HP 1.2 X"d 3.0 X"d

50 < HP 250 HP 1.2 X"d 3.0 X"d

HP < 50 HP neglect neglect

Induction Motors ( = 3600 RPM)

Above 10 00 HP 1.0 X"d 1.5 X"d

250 < HP iz. 1000 HP 1.0 X"d 1.5 X"d

50 < HP 250 HP 1.5 X"d 3.0 X"d

HP < 5 0 HP neglect neglect

See Tablel8 IEEE Std. 141-1976, page 197 for typical

reactance values of induction and
synchronous machines, in per unit of
machine KVA ratings

Table 2.1 ROTATING MACHINE REACTANCE MULTIPLIERS

 MULTIPLYING FACTORS FOR Power System s
TOTAL CURRENT RATING BASIS DEVICES E ngineering, Inc.
P.O. Box 19398
Houston. Texas 77024
y JL JO B NO. N/A DATE Sh 1 of 1

M u ltip ly in g Factors (Total Current R ating

M u ltip ly in g Factors (Total Current R ating
Basis) for Three-Phase Faults Fed Basis) for Three-Phase and Line-to-Ground
P redom inantly from Generators Through Faults Fed P redom inantly T hrough Two or
No More Than One T ransform ation (Local)
More T ransform ations (R e m o te )
[From A N S I C37.5-1969 (R 1974)] [From A N S I C37.5-1969 (R 1974)]

Figure 2.1 MULTIPLYING FACTORS FOR TOTAL

CURRENT RATING BASIS DEVICES
the following information must be known:

(a) Three-phase or single-phase short circuit study?

(b) Rating interrupting time of the circuit breaker used.

Minimum contact parting times usually used in 60 Hz systems

are defined following:

Rated interrupting time (cycles) 8 5 3 2

Contact parting time (cycles 4 3 2 1-1/2

(c) Fault point ^f/Rf ratio

(d) Proximity of generators. The proximity of generators deter­

mines the choice between graphs for fault fed predominantly from local

or remote sources. Unfortunately, for hybrid type sources, that is,

both significant local and remote sources exist, no single multiplying

factor is available from related standards. Different methods have

been suggested for industrial applications. One of the most usual

methods is given in Huening’s paper [6 ]. Then multiply o/Xf by the

multiplying factor selected. This is the calculated total rms current

interrupting duty to be compared with the circuit breaker interrupting

capability.

Next consider the high voltage circuit breaker rated on the symmet-

F Y
rical current basis. Calculate o/Xf and f/Rf ratios and select the

multiplying factor from Figure 2.2 as in the total current basis case.

Then multiply o/Xf by the multiplying factor selected. The result is

an interrupting duty only if the multiplying factor for o/X^ is 1 .0 .

The rated symmetrical interrupting capablity of the circuit breaker is

calculated as,

minimum required symmetrical = ^o x multiplying factor x rated max. E

interrupting capability Xf EQ
 MULTIPLYING FACTORS FOR SYMMETRICAL
Power S ystem s
CURRENT RATING BASIS DEVICES
Engineering, Inc.
P.O. Box 19398
Houston, Texas 77024
BY JL JO B NO. N/A DATE

3 CY CL E
'C IR C U I T
BREAKER

i.O 1.1 1.3

M ULTIPLYING FACTORS

M u ltip ly in g Factors for Three-Phase Faults Fed Predom inantly from Generators
T hrough No More Than One Transform ation (From IE E E Std 320-1972)

multiplying f a c t o r s

M u ltip ly in g Factors for Three-Phase and Line-tcGround Faults Fed Predom inantly
T hrough Two or More Transformations (From IEEE Std 320-1972)

Figure 2.2 MULTIPLYING FACTORS FOR SYMMETRICAL CURRENT RATING BASIS DEVICES
 The short circuit study for time delayed relays, known generally as

30 cycle study, is intended to investigate whether there are sufficient

fault currents to operate relays. Consequently, the network used is

often a minimum source representation and includes only generators that

contribute fault current through the relay under consideration to the

fault point. The generators should be represented by their transient

impedance. All induction motors should be omitted in the calculation.

It becomes obvious that the complete procedure of short circuit

study is tedious and long-hand calculation will soon become impractical

and impossible as the size and complexity of the system increase. Digi­

tal computers are generally used for major fault studies in the industry.

2.2 Steady-State Performance of Electrical Conductors

Under normal operating conditions, the prime considerations in the

applications of electrical conductors are current carrying capacity

(defined as ampacity), and maximum allowable voltage drop, and energy

losses. The cost of the energy losses through resistance is also im­

portant in the case of conductors carrying heavy currents at low volt­

ages such as, for instance, in the electrochemical industry. In such

cases, it may pay to use the conductors at less than their ampacity.

If the selection of conductor size is based on energy loss and conduc­

tor cost, then the most economic choice of conductor according to

classical economic theory is the one that makes the annual cost of the

resistive losses equal to the annual interest on capital cost of the

installed conductor material plus depreciation [7].

For substation buses, ampacity and energy loss are two factors

which normally govern the selection of the conductors. For overhead

transmission lines, ampacity and voltage drop determine the size of

the conductor, and the voltage is considered in choosing the spacing

and insulation of the line. Because of the extra high voltage generally

used in transmission systems, corona plays a major role in the design

of transmission systems. Besides these two factors, the overall system

operation requirements may supersede all the considerations discussed

so far and place a dominant limitation on the selection of the trans­

mission conductor. For example, system stability may require a minimum

conductor size. For overhead distribution in urban area, close voltage

regulation and continuous service are two important factors to be con­

sidered. In addition, because of the heavy load involved, operating

costs per unit of current must be kept as low as possible. Therefore,

in the design of urban distribution systems, much study should be given

to the most economical arrangement of the line, conductor sizes, and

the methods of operationg, etc. For overhead distribution in rural

area, it is extremely important that the line be low in initial con­

struction cost, and still give satisfactory and reliable service. As

a result, long spans are usually used. For building wires and cables,

the determining factor is the ability of the insulation of the conduc­

tors to withstand heat without being deteriorated or damaged. On the

other hand, the National Electric Code specifies that the size of

feeder conductors should be such that the voltage drop will not be

more than 3% for power loads and 2% for lighting loads or combined

lighting and power loads.

In this section, calculation of ampacity for bare conductors will

be discussed. Since the ampacity of an insulated conductor is limited

by the maximum allowable temperature rise of the insulation used, it

cannot be represented by a simple formula. In addition, ampacities can

be easily obtained from NFPA Standard No. 70 [8 ]. Therefore, it will

not be further considered thereafter. Neither voltage drop nor power

loss calculation will be covered since they are very straightforward.

2.2.1 Ampacity of Bare Conductors

Electrical current carrying capacity of a bare conductor is deter­

mined by the tempeature rise resulting from the power loss of conduction

and any other heat gained by the conductor. Since most equipment to

which connections must be made from the conductor is designed to operate

at a temperature rise of 30°C above 40°C ambient in accordance with NEMA

and IEEE standards, it is undesirable to operate the conductor at a

temperature that would cause heat to flow into the apparatus. Such heat

may raise the temperature of that apparatus to a damaging point. In

addition, the conductor may begin to oxidize at an excessively rapid

rate at temperatures higher than 80°C. It is general practice to limit

the maximum operating temperature below 80°C for copper conductors, be­

low 100°C for aluminum conductors, below 95°C for single layer ACSR

(aluminum conductor, steel-reinforced) and 105°C for multi-layer ACSR.

The heat caused by electrical losses in a conductor is dissipated

primarily by convection and radiation. Under steady-state conditions,

the rate of heat dissipatipn is equal to the rate of heat supply.

12R + Q, = Qr + Qc

where I = ampacity in amperes

Qc = convected. heat loss in watts/ft.

0

0r = radiated heat loss in watts/ft.

 Q s = heat gain from sun in watts/ft.

R = nominal AC resistance at operating temperature tc in ohms/ft.

Table 2.2 summarizes atmospheric data required for ampacity calculation.

Each of these items is discussed as follows:

9
(A) Convected heat loss Qc :

For natural convention cases, that is, no wind:

Q = 0.072 D0 *7 5 (t - O 1,25 at sea level

C a

Qc = 0.283 d ^ ’5 at altitudes
(1 0 , 0 0 0 feet and above)

For forced convection cases:

CLL. = [1.01 + 0.371 (DdV)0

■— *5 2 ] K (t C- td) if 0.1 1 D u___
dV < 1,000
u u

*Q„
>-
= 0.1695 [— DdV ] 0 # 6 K (t C - ta) if 1,000 < DndV
U ___
< 18,000
u u
where D = conductor diameter in inches

D0 = conductor diameter in feet

tc = conductor operating temperature in degrees C

ta = ambient temperature in degrees C

d = air density at tc in lb./cu. ft.

V = air velocity in feet per hour

u = absolute viscosity of air at (tc+ t a ) / 2 in lb./(hr.-ft.)

K = thermal conductivity of air at (tc+t a)/2

Among these parameters, d, u, and K can be read from Table 2.2.

#
(B) Radiated heat loss Qr :

Q_ = 0.138 D e [(K )4 - (K )4]

100 100

where e = coefficient of emissivity; 0.91 for black conductors,

0.23 for new conductors and 0.5 for average oxidized

conductor.
CMOSPHERICAL DATA FOR AMPACITY CALCULATION Power S ystem s
Engineering, Inc.
P.O. Box 19398
JL N/A Houston, Texas 77024
JO B NO. DATE

TABLE A VISCOSITY, DENSITY AT SEA 1 E V E L T O 15,000 FT, AND THERMAL CONDUCTIVITY OF AIR
M l - ---

THERMAL
T E M PER A T U R E / K \‘ A ESO IU T E DENSITY, d CONDUC-
I _ 1 VISCOSITY, TIVITY,
V io o y SEA LE V tt 5,000 FT 10,000 FT 15,000 FT
K u
■' K !
32 C j ./ 3 55.55 C.C415 0.0807 a ni 71 0.0554 0.0455 0.00739
41 5 273 59.73 0.C42! 0.0793 0.0660 0.0545 0.04.47 0.007 50
‘ 50 ■C 2 63 *4 14 0.04 27 0.0779 0.0648 0.0535 0.0439 0.00762
59 15 2S8 68.80 0.0433 0.0765 0.0636 0.0526 0.0431 0.CC773
68 in 292 7370 0.0439 0.0752 0.0626 0.0517 0.04 24 0 00784
77 25 296 7c. 86 0.0444 0.0740 0.061 6 0.0508 0.04 T7 C.C0795
86 30 303 E*.29 0.0450 0.0728 0.0606 0.0500 0.041 1 0.00807
95 35 305 89.99 0.0456 0.0716 0.0596 0.0492 0.0404 0.0051 e
104 40 313 95.98 0.0461 0.0704 0.0586 0.0484 0.0397 0.00830
113 45 3l 6 102.26 0.0467 0.0693 0.0577 0.0476 0.0391 0.00841
122 50 323 1OS.85 0.0473 0.0683 0.0568 0.0469 0.0385 0.00852
131 55 328 1 15.74 0.0478 0.0672 0.0559 0.0462 0.0379 0.00864
140 60 333 122.96 0.0484 0.0661 0.0550 0.0454 0.0373 0.00875
149 65' 338 130.52 0.0489 0.0652 0.0542 0.0448 0.0367 0.00886
158 70 34 3 138.41 0.0494 0.0643 0.0535 0.0442 0.0363 0.00898
167 75 348 146.66 0.0500 0.0634 0.0527 0.0436 0.0358 0.00909
176 80 353 155.27 0.0505 0.0627 0.0522 0.0431 0.0354 0.00921
185 85 358 164.26 0.0510 0.061 6 0.051 3 0.0423 0.0347 0.00932
194 90 363 173.63 0.0515 0.0608 0.0506 0.041 8 0.0343 0.00943
203 95 368 183.40 0.0521 0.0599 0.0498 0.0412 0.0338 0.00952
212 100 373 193.57 0.0526 0.0591 0.0492 0.0406 0.0333 0.00966

Hf — cb so lu te vijc oiify , Ib /'h rj^ft). .

P/ = density, lb of csir/ftJ.
kj = thermal conductivity of oir, watts/(iq fi'i'C) c! // = [le + )»}/2.
to = ombient temperature C.
It — conductor temperoture °C .

TABLE B HEAT-TR ANSMISSION FACTOR FOR TABLE D ALTITUDE AND AZIMUTH IN

ALTITUDES A B O V E SEA LEVEL DEGREES OF SUN AT V A R IO U S LATITUDES
AT DECLINATION OF 23.0 DEGREES,
ELEVA TIO N A B O V E MULTIPLIER FOR NORTHERN HEMISPHERE, JU NE 10 AND JULY 3
SEA LEVEL, FT V A LU E S IN TABLE C
LO C A L SUN TIME
0 1.00
5,000 1.15 DEGREES 10:00 A .M . 12 NO O N 2:00 P.M.
10,000 1.25
15,000 1.30 .ATITUDE He Z. H. z. H. Z.

20 62 78 87 0 62 282
25 62 88 88 180 62 272
30 62 98 83 180 62 262
TABLE C TOTAL HEAT RECEIVED BY SURFACE 35 61 107 78 180 61 253
AT SEA LEVEL N O RM A L TO SUN’S RAYS 40 60 1 15 73 180 60 245
45 57 122 68 180 57 238
50 54 128 63 180 54 232
q.. WATTS/SO FT 60 47 137 53 180 22 3
i ^
SO LA R 70 40 143 43 180 | 40 217
ALTITUDE, CLEAR INDUSTRIAL
Hc, DEGREES A TM O SPH ER E A TM O SPH ERE

5 21 J 12.6
10 4 0.2 22.3
15 U .2 30_5
20 64.4 39.2
25 71.5 46.6
30 . 77.0 53.0
35 £ 1_5 57.5
40 £4.5 61.5
45 £7.4 64.5
50 90.0 67.5
60 92.9 71.6
70 9 r-0 75.2
80 95.8 77.4
90 96.4 78.9

Table 2.2 ATMOSPHERICAL DATA FOR AMPACITY CALCULATION

 K c = conductor operating temperature in degrees Kelvin

K a = ambient temperature in degrees Kelvin

(C) Solar heat gain Qs :

#
Q s = a q A sin 8 and

6 = cos -1 [(cos H c) cos (Zc-Z e)]

where a = coefficient of solar absorption; 0.95 for black conduc­

tor, 0.23 for new conductor and 0.5 for average oxidized

conductor.

q = total solar and sky radiated heat in watts/ft.

6 = effective angle of incidence of sun's rays

A = projected area of conductor in sq. ft./linear ft. = D/12

H c = altitude of sun in degrees

Zc = azimuth of sun in degrees

Ze = aximuth of conductor in degrees; 270° for E-W line and

180° for N-S line.

Among these, H c and Zc are available from Table 2.2.

It is important to notice that this equation is derived on the bases of

the following two assumptions:

(1) Heat loss carried off by conduction through supports and con­

nections is negligibly small. This is usually true in most cases.

(2) The conductor is isolated, that is, no proximity effect. Un­

fortunately, this is generally not true. However, the equation can

still be used to calculate conductor ampacity if an effective resistance

value including proximity and skin effects is used for R.

2.3 Transient Performance of Electrical Conductors and Supports

Under fault conditions, mechanical considerations rather than

electrical requirements usually govern the selection and design of

electrical installations, such as substation buses and supports, over­

head lines, and cables. For medium voltage substations, the size of

buses is generally dictated by mechanical stresses caused by short cir­

cuit currents. For high voltage buses, where currents are relatively

lower and the phase spacing is larger, the deflection under short cir­

cuit of the conductor usually determines the bus size. A good practice

is to limit the maximum deflection of the bus to ■^ q- of the span

where only two supports are used and qq ~ where three or more supports

are used. For overhead lines, fault currents are more likely to cause

thermal damage, which may be followed by loss of mechanical strength of

the conductor, than direct mechanical damage. However, the high elec­

tromagnetic forces of fault currents play an important role in the de­

sign and selection of line supports, spacers, dampers, and other equip­

ment. For all insulated wires and cables, heating caused by short

circuit current is a deciding factor in their applications. On average,

the life of cable and wire insulation is about halved and the failure

rate is doubled for each 10°C increase in operating temperature. It

is the intent of this section to cover various stresses, including

mechanical stresses, thermal stresses, and other stresses under abnormal

operating conditions.

2.3.1 Short Circuit Forces and Other Mechanical Stresses

Conductors carrying electric currents will exert forces on each

other according to Ampere’s Law. The basic formula for force between

two infinite parallel conductors is: Fs = 5 . 4 1^ I 2

D (107)
Where Ij and 1 2 are current in each wire, D is center-to-center spacing

of two conductors in inches, and Fs is the force acting on the conduc­

tors in pounds per foot. This equation is directly applicable to

direct current systems.

Calculation of forces set up by alternating currents is somewhat

complicated by the fact that fault currents are usually asymmetrical

and are decreasing with time. There are various methods to calculate

short circuit forces in industry. Some advocate the theory of natural

frequency of vibration of the bus itself and possibility of resonance

with the alternating current frequency. If resonance should occur,

very great forces might result. This method involves data on buses,

bus clamps, insulators, bases and structure which would be practically

endless for all possible combinations. In fact, for reasons of clear­

ance and economy, high voltage substations are generally designed with

relatively long spans and relatively flexible structures. As a result,

the natural frequencies of such high voltage bus systems are normally

much lower than the f r e q u e n c y of the current in the bus. Even for

those applications where resonance might occur, the problem can be eas­

ily solved by proper utilization of dampers or loose scrap conductors.

Consequently, the resonance theory is unnecessarily complicated for

general applications. Others advocate the use of the average force

developed over a half cycle instead of that delivered by the wave peak.

Since the deflection of the conductor subjected to successive impulses

is proportional to the average magnetic force rather than to the maximum

force [9]. This theory is true only if the average force is maintained

indefinitely and the forces are a series of impulses. Even worse, there
is no clearly defined safety margin for all design based on this

theory.

The use of the first half-cycle peak value of the current wave is

a sound and sensible basis for the worst case design. This gives the

maximum forces as successive peak currents are smaller due to various

decrements. Since part of this instantaneous force is absorbed by in­

ertia of the conductor system, a safety factor is inherent. Generally,

either a line-to-line or a three-phase fault is assumed because these

types of fault will result in the maximum short circuit forces available

in a single phase or a three phase system.

In the subsequent paragraphs, various short circuit force formulas

are derived on the base of peak current value during the first half­

cycle assuming that no decrement will be present during this period.

A fully offset wave is assumed for asymmetrical calculations, which

provides a ratio of between the rms value of the composite current

wave and the rms value of the corresponding symmetrical current wave

if no offset is present.

(A) Line-to-Line Faults:

In the case of a symmetrical fault the fault current does not

have any offset, the peak value of the wave is equal to yfl times the

rms value.

Fs - 5.4 (7? Is ) 2

D(107 )

= 10.8 Ig2

D(107 )

Where Is is the rms symmetrical current value.

 For an asymmetrical fault that is fully offset, if the rms symmet­

rical value is given, the peak current equals 2 Is

Fg = 5.4 (2/T Ig)2

D(107 )

= 43.2 Is2

D(107 )

If the rms asymmetrical value over the first half-cycle is given, the

peak value equals 1.63 Ia as shown in Appendix 1.

Fs =5.4 (1.63 Ia)2

D(107)

“ 1 4 -4 ! a2

D(107)

where Ia is the rms asymmetrical current over the first half-cycle.

(B) Three-Phase Fault:

The usual three-phase conductor arrangement is a flat configura­

tion with equally spaced conductors side by side vertically or horizon­

tally. It is proven in Appendix 1 that the maximum possible force will

be impressed on the center conductor, phase B in our case, and the

short circuit must occur 45° after zero current value in phase B. Any

other combination of circumstances gives a smaller force than this con­

dition. At the instant of short circuit, phase B current will have the

value of 0.7071 (that is, sin 45°) times the prefault current peak

value; phase A current will have the value of 0.2588 (sin 15°) times

and phase C current will have - 0.9659 (sin 255°) times the prefault

current peak value. The instantaneous values of their transient di­

rect current components will be the negatives of the above amounts.

The maximum force will then occur a half-cycle after the instant of

short circuit [10] . Neglecting the decrement during the first half­

cycle and multiplying phase current A by phase current B and B by C,

the total force acting on conductor B is

F» ■ 1 8 -7 v 2

D(107)

where Ip is the peak value of the symmetrical alternating current com­

ponent of the fault current.

In the case where the rms symmetrical value is given, the peak

value equals^ 2 ~ Is

Fs - 18.7 ( /2 Is)2

D(107)

- 3 7 ~* !s2

D(107)

From Appendix 1, it is known that the maximum force obtainable

from a three phase fault is v/3/2 times that from a single phase fault,

assuming that the current per phase of the two types of circuits are

the same. However, on the same circuit, the three phase short circuit

currents are usually greater than those of line-to-line faults and the

short circuit forces are accordingly larger. Therefore, if given the

half-cycle rms value of the fault current, the forces on phase B con­

ductor can be calculated as following:

Fg = 14.4 (0.866) Ia 2

D(107 )

" 1 2 ‘5 Xa 2

D(107 )
 By a similar analysis, the maximum force on either outside conduc­

tor can be derived. Table 2.3 summarizes the maximum forces due to

various currents, faults, and conductor configurations. It may be

noticed that a symmetrical three phase fault is not covered in this

table. This is so because a symmetrical three phase fault is impossible

since the currents are 120° out of phase and no time can be selected so

that all three are simultaneously zero. It should be pointed out that

all formulas are derived on the basis of the following two assumptions:

(i) The largest dimension of the conductor is negligibly small

compared with the center-to-center distance between the conductors.

If this assumption is not true, then a shape correction factor

should be applied to the force obtained from these formulas as

shown in Figure 2.3.

(ii) The conductors are assumed to be long, straight, and parallel

to one another. When bends, crossovers, short lines, etc., must

be considered, no simple and generally applicable formulas are

available. Each application should be studied and analyzed

individually.

Of extreme importance in calculating short circuit forces is that

the short circuit current information obtained from short circuit

studies should not be misused. Remember that in computing short cir­

cuit currents, the subtransient reactance of rotating machinery, ex­

cept hydraulic generators, is used for the first cycle duty. Therefore,

the amperes obtained are rms symmetrical containing no direct current

component. This component must be added to the calculated value before

short circuit force analysis. On the other hand, if the results from a
SHORT-CIRCUIT FORCE ON BUS CONDUCTORS Power S ystem s
Engineering, Inc.
P O. Box 19398
Houston, Texas 77024
JL JO B NO. N/A DATE

Table 2.3 SHORT-CIRCUIT FORCE ON BUS CONDUCTORS

M A X IM U M f,
SHORT-CIRCUIT
FORCE ON IN P O U N D S
CURRENT, /, TYPE OF FAULT C O N FIG U R A T IO N
CO N D U C TO R PER FT.
G IV EN AS
(D in IN C H ES)

A B
o o F_ 5.4 \-
D-C A or B
r— D—H D (107)

Single- A B
o o F _ 10.8 /2
RAAS symmetrical Phase A or B
Symmetrical ~ D (107)

Single- A B
RMS symmetrical Pnase o o F _ 43.2/2
A or B
Asymmetrical h-D -H ~ D (107)

Single- A B
RMS asymmetrical Phase o o A or B /: _ u -4 I s ■
Asymmetrical H— d —-! D (107)

Three- A B C
o. o o F 37.4 /2
RMS symmetrical Phase B
Asymmetrical hD-r-DH D (1 0 7)

Three- A B C
RMS asymmetrical Phase o o o B F_ /2
Asymmetrical H-d -t -d H ~ D (107)

Three- A B C
RMS symmetrical o o o F_ 34.9 I 2
Phase A or C
Asymmetrical k-D-n*-DH ~ D (107)

Three- A B C
RMS asymmetrical o o o F_ n .6 / 2
Phase A or C
Asymmetrical h-d -4*-d H D( I O-)

Three-
V D D / F 37.4 r-
RMS symmetrical Phase A, B or C
Asymmetrical
XO O X ~ D (1 0 7)
A k- d —; c

Three- D ° D F_ 12.5 I 2
RMiS asymmetrical Phase A, B or C
Asymmetrical ~ D (1 0 7)
A —1 C

NOTE: Single-phase notation indicates either a single-phase system or a

line-to-line fault on a three-phase system.
 SHAPE CORRECTION FACTORS FOR STRAP BUSES Power System s
E ngineering, Inc.
P.O. Box 19398
jo b no . N/A DATE Houston. Texas 77024
BY

S-3 _ AIR SPACE BETWEEN STRAPS

a+b SUM OF SIDES

F=force between two equal long parallel bars before F can be determined.
Fl=force between two long parallel straight round
The above curves have been drawn neglecting skin
conductors (see section on short circuit forces)
effect and assuming that the current density is uniform
To use the curves, F 1 must be computed according to over the section of conductors.
the discussion on forces due to short circuit currents

Figure 2.3 SHAPE CORRECTION FACTORS FOR STRAP BUSES

short circuit study are expressed in terms of rms asymmetrical values,

it is necessary to ensure that these values are those immediately after

short circuit and before the decrement sets in.

In addition to the short circuit forces described above, the con­

ductor is also subjected to the loads from its own weight, ice, and

wind. The weight loading is quite straightforward. Ice and wind

loading are determined by geographical location. This country is di­

vided into three districts for which standard loading conditions are

specified in National Electric Safety Code [11]. The loading for the

various districts is shown in Figure 2.4. The wind load on the con­

ductor is

H = P(D + 2t)
12

where D = diameter of the conductor without ice in inches

t = thickness of ice in inches from Figure 2.4

P = horizontal wind pressure in pounds per square foot from

Figure 2.4

H = wind force in pounds per foot.

The weight of ice on the conductor is

I = 1.24t (D + t)

in which I = weight of ice loading in pounds per foot.

Then the loaded weight on the conductor is the resultant of the

vertical and horizontal weights plus the proper constant specified in

Figure 2.4. The wind load is generally in the horizontal direction;

the ice weight and the conductor weight are in the vertical direction;

the short circuit forces can be either horizontal or vertical depending

upon the physical layout of the conductor. Expressed in mathematical

 ICE AND WIND LOADING ON ELECTRICAL CONDUCTORS Power S ystem s
Engineering, Inc.
P.O. Box 19398
BY
JO B NO. N/A DATE Houston, Texas 77024

| Fig 250-1
I General L oading Map o f United States w ith Respect
I to Loading o f Overhead Lines

Table 250-1. Ice, W ind and Temperature

Extreme W ind
L oading Districts L oading
(For use w ith (For use w ith
Rule 250B) Rule 250C)

Heavy M edium Light

R ad ial thickness of
ice (in) 0.50 0.25 0 0
H orizon tal wind
pressure in lbs
per sq ft 4 4 9 See Fig 250-2
T em perature (°F ) 0 +15 +30 +60

C onstant to be
added to the re­
sultant in pounds
per foot:
All conductors 0.30 0.20 0.05 0.0

From NESC Std. C2-1981, page 227

Figure 2.4 ICE AND WIND LOADING ON ELECTRICAL CONDUCTORS

form, the resultant load on the conductor is:

F h 2 + Fv 2 + K

where Ft = resultant load on the conductor in pounds per foot

= horizontal load on the conductor in pounds . per foot

= sum of H and Fs , if applicable.

Fv = vertical load on the conductor in pounds per foot

= sum of I, the conductor weight, and Fs , if applicable

K = constant specified in Figure 2.4

During short circuits, the stresses imposed upon the conductor may

be lateral, longitudinal, or torsional. These stresses are transferred

to the insulator and other conductor support hardware. Of these three

stresses, the lateral stress is probably the most destructive as it

applies a strong cantilever force on the support system. In designing

the conductor support systems, the following considerations can be

taken as general guidelines.

(1) Try to keep the insulator in compression, if possible. This

is because porcelain is roughly ten times stronger in compression than

in either tension or cantilever.

(2) Good substation bus design, especially rigid metal bus, re­

quires the use of flexible connectors at certain points in the bus.

In general, they should be used at all terminals of machines, transform­

ers, circuit breakers, disconnect switches and at intervals in long

straight bus runs as required for thermal expansion.

Before leaving these topics, it is proper at this moment to discuss

the calculation of the deflection of metallic buses and the mechanical

stresses on the bus support systems. Because the loads to which they
are subjected, that is, bus weight, short circuit forces, ice and

wind, are uniformly distributed along their lengths, bus conductors

are generally considered as uniformly loaded beams. For conductors

requiring only two supports, deflection and stress calculations can be

made by using the formulas for a simple beam. If more than two supports

are required for a single length of bus, the formulas for a continuous

beam apply. Formulas for these two cases are summarized in Table 2.4.

2.3.2 Transient Thermal Stresses

Electrical conductors are occasionally subjected to current over­

loads due to either short circuits or emergency conditions.

Under short circuit conditions, the ultimate conductor temperature

depends upon (1 ) the magnitude of fault current, (2 ) the duration of

fault current flow, (3) the material and cross sectional area of the

conductor, and (4) the •conductor temperature before short circuit

occurs. Generally speaking, the duration of fault current flow is so

short that for most practical purposes it can be assumed that no heat

loss occurs by convection and radiation. Consequently, the temperature

rise is determined by only the specific heat of the metal, the size of

the conductor and the heat input. The specific heat varies as the

temperature increases. Therefore, it is difficult to determine accu­

rately the temperature rise of a conductor under short circuit condi­

tions. For a particular type of conductors, manufacturers usually

provide the short-time current temperature equation. For bare conduc­

tors, the maximum safe short circuit currents can be easily calculated

from this equation, if given maximum allowable conductor temperature.

Standard industrial practice recognizes a temperature of 300°C as the

 BUS CONDUCTOR DEFLECTION Power S ystem s
AND STRESSES FORMELAS hJUi1E ngineering, Inc.
P O. Box 19398
Houston, Texas 77024
iy JL JO B NO. N/A DATE

deflcction in inches /= moment of inertia, inches ,<J

h-= loadin lb /in. of Si * bending moment in pound-inches
S*section modulus, inches®
k»,.h /'•fiber stress in Ib/sq in.
total uniform load in pounds {wl) / —maximum allowable fiber stress in lb/sq in.
I « span in inches (The value of minimum yield strength is com­
E —modulus of elasticity, lb /sq in. monly used.)
©Maximum deflection occurs in the end spans and is only slightly less than that for a continuous
beam of 2 spans.
5,Maximum moment and fiber stress for simple beams occur at the center of the span.
©Maximum moment and fiber stress for beams fixed at both ends occur at the points of support.
©Maximum moment and fiber stress for continuous beams occur at the second support from each end.

Table 2.4 BUS CONDUCTOR DEFLECTION AND STRESSES FORMULAS

maximum allowable for determining short time rating of buses; 340°C

for all-aluminum conductors and 645°C for ACSR conductors. For insu­

lated conductors, the IPCEA has established maximum allowable short

circuit temperatures for various types of insulation.

Emergency loading normally occurs in transmission or distribution

systems due to coincidence of peak loads with high summer ambients,

shifting of additional loads to an already loaded conductor, use of

high loadings to prevent icing and others. The procedure of determin­

ing maximum overload currents under emergency loading conditions is

the same as that under short circuit conditions. In an emergency load­

ing application, caution should be given to the fact that the effect of

heating is cumulative.

2.3.3 Low Voltage Conductors: Special Considerations

As stated in Section 2.1, predictions of possible short circuit

currents for high voltage circuits, where arcs are of a sustained

character, can be made fairly accurately. However, in low voltage (600

volts or less), the effect of fault impedance is such that the actual

current resulting from a fault is usually much smaller than that calcu­

lated. If this difference is not recognized, unnecessarily high costs

for bus structures may be incurred.

Caution should also be exercised in applying relay current-time

characteristics for the settings of protective relaying on low voltage

lines that may be subjected to arcing burndown. The settings should

be such that the relay will operate before any arcing burndown occurs.

2.3.4 Vibration of Overhead Conductors

Vibration is a frequent cause of failure or serious damage on

overhead distribution and transmission lines. Because of the importance

of such lines, even occassional failures must be avoided. Vibration

of overhead conductors may be divided into two general classes:

(i) Resonant vibrations of high frequency and low amplitude. This

type of vibration is most destructive and may be caused by either

fatigue due to successive reflection of vibration or chafing. In

either event, the failure usually takes place at the point of ver­

tical support, and only occasionally elsewhere, such as at a dead­

end, and even less frequently at a joint.

(ii) Vibrations of low frequency and high amplitude, usually known

as dancing or galloping. This type of vibration is caused by

passage of a current of air over the conductors perpendicular to

the major axis. Damage caused by dancing conductors is usually

due to flashover caused by conductors whipping about at high

amplitudes and reducing the clearances.

To date, the most popular methods of minimizing the damage due to

vibration involve either improvements in design of supporting clamps,

or addition of accessories to the line. These methods come under three

general categories as follows:

(i) Suspension clamps designed to permit vibration to pass through

them without damaging reflection.

(ii) Reinforcements or armor rods at the suspension clamps which

will increase the effective section modulus at the point of support.

(iii) Dampers which function to absorb vibration energy and pre­

vent destructive vibration from building up.

2.4 Applications of Instrument Transformers

Instrument transformers are used both to protect personnel and

apparatus from high voltage and to allow reasonable insulation levels

and current carrying capacity in relays, meters, and instruments.

Instrument transformer performance is critical in protective relaying,

since the relays are only as accurate as the instrument transformers.

Generally speaking, instrument transformers for metering and instrumen­

tation require higher accuracy than those for protective relaying. In

this section, only those instrument transformers for protective relaying

will be covered. Special emphasis will be given to performance estima­

tion of current transformer for relaying applications.

2.4.1 Applications of Current Transformers

The major criteria for selecting a current transformer ratio are

the continuous current ratings of the connected relays and of the sec­

ondary windings of the current transformer itself. In practice, with

load current normally flowing through phase relays, the ratio is usually

selected so that the secondary output is around 5 amperes at maximum

primary load current. In general, the performance of current trans­

formers is not so critical for transmission and distribution line pro­

tection. However, it is most critical for differential protective

schemes, particularly under transient conditions.

Since most relays commonly operate at currents of several times

the maximum normal current of the current transformer, performance up

to 20 times normal current is the basis for accuracy specification.

At the same time there is usually little need to limit the ratio error

to less than 10% and hardly any need to limit the phase angle for
relaying. Consequently, the basis for classification of performance

for relaying is an error of 10% at any current from 1.0 to 20 times

normal current. Instead of a description of accuracy, ANSI accuracy

classification is really a description of how much voltage the trans­

former can supply to the external circuit or burden. The performance

of current transformers is then specified in terms of the secondary

voltage which can be supplied by the transformers at 20 times normal

current without exceeding the error limit of 10%. In addition to volt­

age ratings, the ANSI standard [12] also uses two letters, C and T,

for class designation. Letter C means that the effective internal

impedance is only secondary resistance. This implies that the exciting

current by-passed through the exciting branch can be rather easily cal­

culated if one has the saturation curve at hand. The C classification

covers bushing current transformers with uniformly distributed windings,

and any other transformers whose core leakage flux has a negligible

effect on the ratio within the defined limits. On the other hand,

letter T means that even if the internal reactance of the transformer

is known, the calculation of the exciting current by-passed in the

exciting branches is most tedious or even impractical. It is usual

that performance of the transformer under this class should be deter­

mined only by test. The T classification covers most wound-type trans­

formers and any others whose core leakage flux affects the ratio

appreciably.

For relaying service application, the following current transformer

data should be obtained from the manufacturer.

(1) Relaying accuracy classification.

 (2) Mechanical and thermal short time (1 second) ratings. Both

ratings define rms values that the transformer is capable of

withstanding and should be checked against system fault current

at the location of installation.

(3) Resistance of the secondary winding between the winding

terminals.

(4) For Class C transformers, typical excitation curves on log-log

coordinate paper are required. For Class T transformers, the manu­

facturers should supply typical overcurrent ratio curves on rec­

tangular coordinate paper.

It is important to note that the above ANSI ratings are applicable

to the full winding only. Where there is a tapped secondary, a propor­

tionately lower voltage rating should be used for each tap.

From the preceding discussion, it is clear that the errors in ratio

of a current transformer depend upon the total impedance connected to

the secondary of the transformer. Unfortunately, many relays have var­

iable impedance. For these applications, analysis of the transformer

performance is usually based on an equivalent value at normal current.

This can be justified on the basis that the burden at higher current

is usually less and thus the current transformer will perform better

than expected from the equivalent burden.

A current transformer’s performance is measured by its ability to

reproduce the primary current in terms of the secondary without satura­

tion and large errors. Two common methods for performance estimation

are excitation curve method and ANSI transformer relaying accuracy

class method.
 (A) Excitation curve method:

This method consists of the following steps.

(Step 1) Determine total burden on the transformer using the

following expressions:

(1 ) with wye-connected current transformers:

Zf- = 1.13 R]_ + Z + R for three-phase faults

Zt = 1.13 (2R^) + Z + R for single phase-to-ground faults

(2 ) with delta-connected current transformers:

Z-f- = 3 (1.13 Ri + Z + R) for three-phase faults

Z t = Z(1.13 Ri + Z) + R for single phase-to-ground faults

where R^ = one-way lead resistance from the transformer to

relays in ohms at 75°C

Zt = total burden in ohms on the transformer

1.13 = a multiplier used to accommodate temperature

rising during fault

Z = burden impedance of any devices connected or

reflected to the current transformer in ohms

R = internal resistance of the transformer in ohms

(Step 2) Determine secondary CT voltage required, E s , to

deliver a given amount of secondary current, say Is .

(Step 3) From excitation curve of particular tap of the

transformer being used, determmine excitation current Ie

corresponding to the required secondary voltage, E s .

(Step 4) Determine the percentage error for the transformer.

Applications of this method will be illustrated in Section 4.3, under

the topics of differential protection.

 (B) ANSI Transformer Relaying Accuracy Class Method:

In contrast with excitation curve method, this method gives only

a qualitative appraisal for current transformer applications. To en­

sure the ratio error of a Class C transformer being less than 10% with

maximum secondary fault current Is flowing, the following criterion

should be met.

NV + 100 R > Is (Z + R)

where N = proportion of total current transformer turns in use

V = voltage rating class, that is, Class C voltage

2.4.2 Applications of Potential Transformers

Two major criteria for selecting a potential transformer are the

system voltage level and the basic impulse insulation level as required

by the system on which they are to be used. ANSI Standard [12] recog­

nizes three groups of potential transformers for different capabilities

and connections.

Standard accuracy classification of potential transformers range

from 0.3 to 1.2, representing percent ratio corrections to obtain a

true ratio. These accuracies are normally high enough so that any

standard transformer will not create any problem for relaying applica­

tion as long as it is adequately used. Potential transformers are

usually connected in open delta for balanced system; in wye-wye when

line-to-neutral loading or metering is required; in wye-open delta for

ground detection in an ungrounded system; and any specific connection

as required by a particular relay.

2.5 The R-X Diagram

As the system complexity increases and higher operation reliability

is required, it is an increasing trend to replace single operating

quantity relays with multiple operating quantity relays. Surely this

will increse performance exactness of protective relaying systems.

But it raises the question, "How can we relate the characteristics of

these more complicated relays to different system operation conditions?"

The answer to this question will be found in the R-X diagram. Before

examining this powerful tool, a brief review of ohmic relay character­

istics is first given.

An ohmic relay is defined as a relay which responds to the follow^*

ing three variables: voltage, current, and phase angle. In general

the relay will respond to at least three of the following four torque-

producing components:

(1) Voltage component (torque proportional to V 2 )

(2) Current component (torque proportional to I2)

(3) Product component (torque proportional to VI F(0))

(4) Control spring torque

Thus a general torque equation for the relay will look like:

Torque = ± Kj V 2 t K£ I2 t K 3 V I f (r, 6 ) +

where Torque is positive for a contact-closing torque

K 4 is a spring torque assumed to be constant

K_i, K 2 , and K 3 are independent design constants of either sign and

variable magnitudes.

r is the design angle of maximum torque.

Since a good example can serve hundreds of words, the following example

will be used to illustrate the value of the R-X diagram. Considering

a directional relay with voltage restraint, set K 2 = K 4 = 0 and select

a minus sign for K^, positive sign for K3 and F(r. 6 ) = sin

(90° + r-e).

Torque = K 3 V I sin (90° + r-e) - Kx V 2

Consequently, the relay characteristic is

K 3 V I sin (90° + r-6 ) = V2

K 3 I sin (90° + r-6 ) = K 2 V, if V i 0

It is obvious that the relay characteristics are a function of system

operation parameters V and I which are usually unknown under transient

conditions. Even more complicated, prefault system operating conditions

affect post-fault conditions, and different types of faults result in

different system transient conditions for the same initial conditions.

One key to this problem is the R-X diagram.

Remember V/l = Z.

Z = K3 sin (90° + r-e)

Kl

It is obvious that system impedance seen by the relay should be less

than K 3 /K 1 sin (90° r-6 ) in order to operate the relay. From this

example, it is clear that the value of the R-X diagram lies in the

following two facts: relay characteristics can be simply shown and

system conditions can also be shown on the same diagram as proven in

the subsequent paragraph.

To illustrate how system conditions can be represented on the R-X

diagram, a two machine system is shown on Figure 2.5(a). On the same

figure is also shown the impedance seen by the relay located at substa­

tion A for different fault locations and one loading condition. By

similar analysis, other system operation conditions can be shown on

the diagram. Figure 2.5(b) illustrates loci of loss-of-excitation in

 P ow er S y ste m s
SYSTEM CONDITIONS ON THE R-X DIAGRAM E ngineering, Inc.
P .O . Box 1 9 398
Houston, Tex as 7702*1
JL JOB N O . N/A DATE

c «
■u
t: u
3
. o a X'O
°
J w 2Z
w z ? ®
o o •» O)
< -
*
O u.
o
—* ©
_ > ■
- - tr o ■<
u
W o U s ft-
vT o

I ©©©©©
a synchronous generator.

After both the relay characteristics and system conditions have

been plotted on the same diagram, the proper protective scheme for the

system can be decided from the R-X diagram. This will be illustrated

later in Chapters 3 and 4.

 CHAPTER 3

GENERATOR PROTECTION

Surely the frequency of generator failure is low with modern design

practices and improved materials. Yet failure can occur and may result

in severe damage and long outages if not properly protected. The de­

gree of protection for each generator is mainly a result of economic

compromise. It should be carefully weighed against possible risk if

no protection were applied, and acccording to the size and the impor­

tance of each individual machine in the system. Overprotection is as

costly as underprotection. Optimum generator protection can only be

obtained by individual engineering analysis of the particular installa­

tion. Since there are very few [4] provisions for generator protection,

the design of generator protection will heavily rely upon individual

engineering experience. This chapter will cover some recommended mini­

mum protective features applicable to generators of various types and

sizes together with some additional considerations in designing genera­

tor relaying schemes. Since the connection of a generator to an exist­

ing power system will dictate the type of grounding and protective

system employed, system connection will be discussed briefly, too.

Generator faults can be either phase-to-ground or phase-to-phase

faults, either in the stator winding (internal faults) or in the leads

external to the machine terminals. Most generator faults will involve

ground, and the amount of fault current will depend upon the method

used to connect the generator to the system. Internal faults generally

develop as a ground in one of the phase windings because of insulation

failures and may occasionally involve more than one phase. Insulation
failures occur due to overvoltage or reduced insulation strength, or

more frequently a combination of both. Overvoltage caused by lightning

and switching transients should be protected against by surge arrestors

and will not be covered in this chapter. Reduced insulation strength

is usually due to abuse and aging. Both result in drying out and bak­

ing the insulation until it loses its elasticity and approaches the

carbonized stage. The most frequent types of abuse are overloading,

lack of attention to the cooling system, and allowing dust to accumu­

late. All short circuits in the stator require the repair or replace­

ment of the damaged stator coils. Consequently, it is extremely impor­

tant to limit the duration of a stator fault for the following reasons:

(1) To prevent igniting the insulation material which may result

in a fire.

(2) To prevent burning of the iron, possible costly replacement,

and longtime service interruption.

3.1 Philosophy of Generator Protection

In order to minimize damage, when a short circuit develops in

stator windings or in the connections up to and including the breakers,

which connect the generator to the rest of the system, these breakers

must be tripped immediately. This will prevent any outside source

from supplying current to the short circuit. To stop a synchronous

machine from feeding its own short circuit, it is also necessary to

disconnect and collapse the field, and to de-energize the prime mover

of the machine from its source of power. If the neutral of the genera­

tor is grounded through a neutral circuit breaker, the simultaneous

tripping of this breaker along with the main breaker and the field
circuit will help to limit the duration of arcing at a ground fault.

It is general practice to trip all these breakers through a hand-reset

a u x i l i a r y or lockout relay. This same relay should also shut down the

prime mover and energize fire extinguishing equipment if required.

Overcurrent relays, which protect against ground faults only, may

have very sensitive current settings. However, selectivity require­

ments may prevent the use of low current or fast time settings. It

is also necessary to protect against those faults which start as phase-

to-phase faults before the ground relays can operate. Unfortunately,

phase-to-phase overcurrent protection has inherent limitations from

the standpoint of sensitivity, selectivity, and speed of operation

since overcurrent relays must be set above maximum load current and

must also have time settings which should be coordinated with other

relays in the system. For these reasons ordinary overcurrent relaying

is seldom satisfactory for generator protection. Because of its selec­

tivity and sensitivity, differential protection is generally accepted

as standard for the protection of all generators above 1 MVA. The

minimum size of machine for which it is desirable to provide differen­

tial protection is determined by balancing the cost of the protection

against the importance of the machine. Differential protection is

strongly recommended for smaller machines under the following condi­

tions [4] .

(1) Any KVA ratings, 5KV and higher

(2) 501 KVA and larger, 2.2 KV and higher

(3) Machines paralleled with larger machines or a system, which

are differentially protected.

 (4) Machines, regardless of size, which are important to the op­

eration and reliability of the system.

In the instances where the cost of differential protection is not

economically justifiable, current balance relaying can provide an ac­

ceptable substitute. In addition to winding faults, this type of re­

laying will also afford protection against single-phase operation.

Most field circuits of modern generators are usually operated un­

grounded so that a single ground will not damage the machine at all.

However, since a potential second ground may short part of the field

winding and cause severe mechanical vibration, a ground indicating

equipment should be used to detect the occurrence of the first ground

and initiate an alarm.

The principles of external fault backup in generator protective

relaying is to minimize damage from a feeder or bus fault in case the

first line relays normally expected to protect these circuits fail to

trip the proper circuit breakers. This backup relaying should be

chosen to provide coordination with those first line relays.

For a turbine generator, a preheating schedule during startup may

call for running the unit at reduced speed for several hours with the

field energized. An induced voltage at reduced frequency will exist

on the stator winding during this period. Adequate relays should be

selected to detect possible faults during this reduced frequency

operation.

For a hydro-generator, a sudden drop of full load may result in

an overspeed of 14% or more of the rated value. The terminal voltage

may reach as high as 200% or more if AC overvoltage protection is not

provided. Therefore, overvoltage protection is essential and strongly

recommended for all hydro-generators.

3.2 System Connection [13] [14]

The method of connecting a generator to an electrical system dic­

tates the type of neutral grounding and protective schemes employed.

In spite of various connection methods, all of these ramifications can

be classified under two categories; direct-connected system and unit

system, insofar as system protection is concerned.

(A) Direct-Connected System:

This system applies where at least some of the power generated

will be distributed at the generator terminal voltage. It can be

either solidly grounded or low resistance and reactance grounded. In

its application, it has to be assured that sufficient ground fault

current is available to operate system protective devices.

Where the ground fault capacity of the machine is approximately

equal to the level required for the system protection, the machine

could be solidly grounded although this is not generally recommended

[13]. For installations where the possible ground fault level exceeds

the values required by the system protection devices, the maximum

ground fault current should be reduced to a level of 100 ampere minimum

and 150% maximum of full load current by the use of grounding resistance

or reactance. Usually, reactance grounding is applied where the gener­

ator supplies four-wire feeders at generated voltage, and resistance

grounding, where three-wire feeders are supplied at generated voltage.

(B) Unit System:

In this system, each generator is directly connected to its

individual transformer bank, the low-voltage side being delta, and the

high-voltage size, wye. So far as ground fault currents are concerned,

each machine is isolated from the others and the system. Hence, the

generator ground fault currents can be limited to a very low value,

but not less than the minimum necessary to actuate a protective relay

[14]. A variation of this arrangement is sometimes used where two gen­

erators supply a common transformer bank.

The optimum method of limiting the ground fault current of a unit

system is to insert the primary of a distribution transformer with a

resistor across its secondary between the star point and the ground

for a wye-connected generator. The resistor and the transformer are

usually sized to limit the ground fault currents to a level of 5 to 15

amperes. For delta-connected generators [13] and unit generator system

installations where there is no switching at generated voltage, poten­

tial transformer neutral grounding has also been applied successfully.

In spite of these variations, both are known as high resistance ground­

ing. Another alternative for limiting the ground fault current of a

unit system is low reactance grounding. The neutral reactor will gen­

erally be selected so that ground fault current will not exceed the

three-phase fault level.

3.3 Fault Protection

Generator fault protection functions to detect any faults either

internally in stator windings or externally in terminal leads but with­

in the generator relaying zone. It usually includes differential pro­

tection and stator ground fault protection. It is general practice

that all fault protection devices should be arranged to trip the

generator main circuit breaker, the field circuit and shut down the

prime mover.

(A) Differential Protection (Device No. 87G)

In differential protection, the currents in each phase, on both

sides of the machine, are compared in a differential circuit. The

term "differential" is properly applied only to the connection of cur­

rent transformers so that the net input to the relay is the difference

between input and output currents. There are various modifications of

the differential protection schemes, depending upon the type of relay

used and the connection of the machine itself. However, in all these

schemes, it is always good practice to use current transformers with

the same characteristics wherever possible, and to avoid connecting any

any other equipments such as meters and other relays to these circuits.

It should also be noted that differential relays do not protect any

open circuits, or turn-to-turn short circuits in most of these schemes.

Where perfectly matched current transformers at opposite ends of

the windings are available, overcurrent relays in differential connec­

tion can provide excellent selectivity as well as very fast tripping.

Actually, very few current transformers, even though they are commer­

cially identical, have the same characteristics, especially under ex­

ternal fault conditions. Furthermore, the rapid increase in the com­

plexity and loading of integrated power systems requires a more dis­

criminative line of differential relays on each generator to all exter­

nal faults without sacrificing selectivity and sensitivity.

The percentage differential relay meets all these requirements.

This relay has two or more restraint windings which produce opening
torques and desensitize the relay to high external fault currents.

The percentage indicates the unbalanced current required to operate

the relay as a percentage of the restraint currents. There are two

types of percentage differential relays generally used for generator

protection: constant percentage and variable percentage differential

relays. A variable percentage differential relay is designed to be

more sensitive and faster on light internal faults, but less sensitive

on heavy external faults. However, it costs about 25% more than a con­

stant percentage relay. It is generally recommended for the following

applications:

(i) The situation where heavy saturation currents in current

transformers may be encountered.

(ii) Full-voltage transformer energization during black start or

following out-of-step synchronizing. This is so because these condi­

tions will produce DC current flow in the machine. Since the time

constant of rotating machinery is large, it may produce current trans­

former saturation, probably occurring at different times for the two

current transformers in a given phase, that will persist for many

cycles after the initial disturbance.

Most generators have wye-connected windings. Figure 3.1 shows a

differential protective scheme for these generators, including DC trip­

ping circuit. This scheme can protect against phase-to-phase fault

whether generator neutral is grounded or not. With a solidly grounded

neutral, the differential relay will operate on phase-to-ground short

circuits in any part of the winding except a small portion very close

to the neutral, depending upon relay sensitivity and setting. If

 DWG.NO. REV,
DIFFERENTIAL PROTECTION FOR P ow er S y ste m s
WYE-CONNECTED GENERATOR E ngineering, Inc.
Houston, Tr*»« 7703K
Y JL jo e no. N/A date Sh 1 of I

ct
Ul
o.
UJ
< Q_
CC
cc
I—
CL
CD
CL
UJ Qc _J CTL
00
Z <
21
(X CD
CH
i- a.
o:
O
5 Z> 00 Gj c c X
< iT) L lI

O O O o
I-

o g

HH
O
H H
O
HH I T
HP
8 2 cr
<
i£>
CD
iS>
CD oo CO O UJ
in 2 :
fcja
Ll I ^
b o
o c r iljJ
CL Ih-
o
o Ll I
Q
> -1 z :
o < z:
r\J p o
z: V
UJ , ,
QC y
UJ > 7

Q <
current-limiting devices are inserted in the neutral, the effective­

ness of differential relaying against phase-to-ground fault will be

diminished slightly and there will be a somewhat greater portion of

the winding next to the neutral for which no protection can be provided

by the differential scheme. If the machine neutral is ungrounded, the

scheme can still protect against phase-to-ground fault if the system

is grounded elsewhere and the fault current is greater than the relay

setting. It is quite obvious that the differential scheme will not

respond to line-to-ground fault in a unit system. Figure 3.2(a) shows

a differential scheme for a wye-connected generator having only one

neutral lead brought out of the machine. In this scheme, residual

line current, that is, the phasor sum of the line currents, is balanced

against the neutral current. Consequently, it will protect against

phase-to-ground faults for grounded neutral and ungrounded neutral pro­

viding the system is grounded elsewhere. However, phase-to-phase

faults will not operate the relay and separate phase relays are re­

quired in this scheme. Figure 3.2(b) illustrates a similar protective

scheme for delta connected generators. It will protect against phase-

to-phase faults and phase-to-ground faults if the system is grounded

elsewhere. In the instances where two sets of current transformers

are not economically justifiable, a flux balancing differential scheme

provides a substitute. Figure 3.2(c) shows an example of this scheme.

This scheme requires that attention be given to the machine primary

terminal connections since the generator protection zone will vary in

accordance with the physical location of the current transformers.

(B) Stator Ground-Fault Protection (Device No. 51GN or 64GN)

TYPICAL DIFFERENTIAL PROTECTION SCHEMES
P ow er S y stem s
E ngineering, Inc.
P O. Box 19 398
Houston. T exas 77024
JL J 0 8 NO. N/A DATE Sh 1 of 1

P rotection o f a G e n e ra to r H avin g O n ly Four Leads

B rought O u t. — Differential Protection o f a .Defto CcmmwcW Cwwrwlar

(a) Protection of a Generator (b) Differential Protection of a

Having Only Four Leads Delta Connected Generator
Broucrht Out

Figure 3.2 TYPICAL DIFFERENTIAL PROTECTION SCHEMES

 The method of grounding affects the degree of protection afforded

by differential relays as described in the preceeding paragraph. Con­

sequently, it is general practice to insert a separate ground fault

relay in the neutral as a part of fault protection. This relay can be

either a current sensitive type (Device No. 51GN or 50GN) or a voltage

sensitive type (Device No. 64GN).

Typical applications of the current type ground fault relay are

in direct-connected systems, that is, a generator connected directly

to a bus with other circuits. Since a fault on one of these circuits

might also misoperate the relay, a time delay is usually required to

coordinate with other overlapping relays. Although this type of relay

can provide sensitive protection to some extent, it becomes less appli­

cable as the grounding impedance of the generator is increased. This

is so because the lower the relay pickup, the higher its burden on the

current transformer and the more difficult it is to discriminate ground

fault current and the third harmonic current flowing in the neutral.

This third harmonic current may reach 15% of the maximum generator

output with solid grounding.

Applications of voltage sensitive type relays for ground fault

protection can be best illustrated in a unit generator-transformer

system. The relay used must be insensitive to the third harmonic volt­

age which may normally exist between neutral and ground, and yet still

respond to the voltage shift of the generator neutral with respect to

ground which accompanies a fault. Since the power transformer and the

station service transformers are all connected in delta on the generator

side, coordination with other protection is virtually unnecessary and

the ground fault relay used can be very sensitive. However, some time

delay in relay operation is usually provided to prevent tripping on

system transient disturbances or faults in the potential transformers,

if used. The general practice is to use the relay to trip the generator

main circuit breaker and shut down the machine.

A ground fault relay is also economically justifiable in an un­

grounded generator. Though no damage is likely to occur from a single

ground, a second ground will result in phase-to-phase short circuit

and requires a service interruption. The relay commonly used for such

detection is a voltage-sensitive relay connected across the secondary

of a potential transformer (bank) or an indicating voltmeter. The re­

lay is generally used to initiate an alarm.

3.4 Abnormal Operation Protection

A generator in abnormal operation modes may be subjected to severe

damage. The most frequent abnormal operation conditions include loss

of excitation, motoring, negative sequence current, uncleared system

fault, overloading, overtemperature, overvoltage, overspeed, over and

under frequency, grounded rotor and so on. In addition to the potential

damage to the generator, some of these conditions may also be disastrous

to the system operation depending upon the relative size of the genera­

tor. In this section, protection schemes for these conditions will be

discussed.

(A) Loss of Excitation (Device No. 40)

Partial or complete loss of excitation on a synchronous generator

will result in a drop of terminal voltage, causing it to draw reactive

power from the system. If the system is large enough to supply the
deficiency in excitation through the armature, the machine will operate

as an induction generator, running at 1 to 3 percent above the synchro­

nous speed. Since synchrous generators are not designed for asynchro­

nous operation, the machine output will oscillate as the rotor rotates

in an attempt to lock into synchronism and severely damage the rotor,

especially a non-salient one, by overheating. Furthermore, an excessive

voltage drop represents a potential disaster. Consequently, it is a

safety practice that loss of excitation protection should be provided

for all synchronous generators. The decision remains as to whether the

relay should trip the unit or sound an alarm. From the standpoint of

generator protection, most modern generators can run two or three min­

utes at full load with zero excitation without endangering the genera­

tor. Consequently, the generator can stand a delay in tripping. From

the standpoint of protecting the system against instability, the system

cannot withstand the large reactive power displacement following loss

of excitation accident. Therefore, it is a proper approach to plan

to trip the generator on loss of excitation unless there is definite

justification that the accident will not result in a major system

disturbance.

Two types of relays are available for use as loss of excitation

protection [1]. The simplest type is an undercurrent relay connected

in series with the field circuit. However, this type of relay would

not necessarily detect loss of excitation resulting from short circuit

in the field winding. Where more complete protection is required, an

impedance relay should be used. The basic element of loss of excitation

relays is an impedance unit. When a generator loses excitation, the

impedance seen by the relay looking into it will depend upon the

machine characteristics, the load flow to the loss of excitation, and

the type of excitation failure. Therefore, it may be necessary to use

a timing unit, directional unit, undervoltage relay and others to pre­

vent possible malfunctions of the impedance unit. As an example, Fig­

ure 3.3(a) illustrates application of an offset mho distance relay

for a unit system protection [15]. Figure 3.3(b) shows system behav­

ior under stable swings resulting from system disturbance. It is

obvious from the R-X diagram that a time delay unit or another mho

relay is required to prevent undesired tripping. This requirement

depends upon the possible modes of generator operation, generator

characteristics and the particular system concerned. Each application

shall be individually evaluated and the generator manufacturer shall

be consulted for protection requirements and relay settings.

An important consideration in the application of loss of excitation

relays relates to the potential transformer secondary practice. In

general, the potential transformers that provide potential to the relay

will have other burdens connected between phases and possibly between

phase and neutral. If a secondary fuse or circuit breaker on one phase

is open, the potential on that phase will not necessarily go to zero.

It will generally assume some potential depending on the impedance of

the total burdens connected between that phase and the other phase and

ground. Thus, when potential transformer secondary fuses are used, it

is recommended that the relay be fused separately from all other

devices. The blowing of a potential transformer primary fuse, when

used, presents a similar problem that cannot generally be solved as

 DWb. NU„ ftEV.
P ow er S y ste m s
LOSS CF EXCITATION PROTECTION \ 0 | J E ngineering, Inc.
I Houston. Trxes. 77024
iv JL JOB NO. N/A DATE Sh 1

* - * » © » ■ :-£ i t c # 17 £ i f i e ld i t f u t i l c i b
■ - S W 0 tT - 'U » O ilT E 3 r i f i r I T MODERATE LOAD

C - S n e t T - C lt c u n T t* n t L O »T * 0 1 0 * 5 0 * O F f l -
Cl«£li"!t fin e II KO 1,310.

(a) Loss of Excitation for a Unit System

by use of an Offset Mho Relay

IT S . CQ IIIf.
/ITT_____ | m r m iT r
BUS

f 34 f a u lt

l «* U » |T* * »f :e « I. c t i t i n 11 1 0 * 9 i* > f S » * c { .

1*. I ,!:[,(.» *f j {» f * y i l ,15 ( U»MS

* * V c W .'' f ®* C C «C .*I0«5 «f U i i i t C C .S i f t » ( » iFlitTO* LOAD. k*t>
*U ' v ' - u * ,|* 5 $ t C f tlt iC M . S » iH C * li6 T iih E , «■ » f O t liW l( C M ,» I W M T Of X M I C L ( M M * liiM lfS il)
! • -' C.9i L f J S lv g » c * t i M C TW 1 ,0 i « l l - L flijt i« * fC « * C (
S(. im.(s,ii:u* «»Tr u fwif 15(■) mu.

(b) System Behavior Under Loss of Excitation Conditions

Figure 3,3 LOSS OF EXCITATION PROTECTION

easily as the secondary fuse situation. In order to prevent malfunc­

tion on a blown primary fuse, a voltage balance relay (Device No. 60)

can be used to supervise the loss-of-excitation relay tripping contacts.

The basic considerations above apply to all types of prime movers.

However, special caution must be observed in the case of hydro-genera­

tors and under-excited synchronous condensers. Hydroelectric generators

may be operated severely underexcited during light load to compensate

for system distributed capacitance. Therefore, the operating condition

during a light load period can approach the loss of excitation condition

in terms of impedance seen by the relay and may result in undesired

tripping.

(B) Motoring Protection (Device No. 32)

If the turbine throttle or stop valves are closed while the genera­

tor remains connected to the system, the generator will run as a motor,

receiving sufficient power from the system to supply friction and wind­

age losses in the unit. The approximate power required to motor a gen­

erator depends upon the individual design of the generator. Listed

below is the approximate power expressed as a percentage of nameplate

rating from General Electric and Westinghouse catalogs.

General Electric Westinghouse

Condensing Steam Turbine 0.5 - 1.0% 3%

Noncondensing Steam Turbine 1-4% >3%
Hydraulic Turbine 2 - 100% 0.2 - >2 Z
Gas Turbine • ---------------- 50%
Single-shaft Gas Turbine 100% ---
Double-shaft Gas Turbine 10 - 15% ---
Diesel Engine - 2 Cycle 25% 25%
Diesel Engine - 4 Cycle 15% ---

It should be noted that motoring will not endanger the generator

itself at all. In fact, generator motoring protection is intended for

the prime movers and the system as backup protection, rather than for

the generator. For example, steam turbines will be overheated because

of the lack of adequate steam flow to cool the blades and shroud, and

generally will be protected by steam temperature devices. Similarly,

hydraulic flow indicators will protect against blade cavitation on low

water flow.

There are considerable variations in industry in the methods for

protecting against damage from motoring conditions. There is so much

divergence in practice that it is difficult to say which one is the most

widely accepted. Since safe motoring times will usually be in the order

of one minute or less, some companies, particularly in older or smaller

stations, depend upon operators to notice the condition and initiate

corrective measures. Other companies use directional power relays

(Device No. 32) to detect the condition, to initiate an alarm, or to

trip the generator main circuit breaker.

Regardless of these variations, a trend is developing to use an

electrical reverse power relay to trip the unit from the system when

motoring because of the increasing complexity of modern stations and the

huge investment involved in larger generators. This is the protective

scheme strongly recommended because it affords maximum protection to the

prime mover. In case of diesel-engine generating units, reverse power

relaying is imperative because of inherent danger of explosion and fire

from unburned fuel.

(C) Unbalanced Current Protection: Negative Sequence Current

(Device No. 46)

Negative sequence currents, inherent in unbalanced phase currents

due to external phase faults or unbalanced loads, will induce twice-

rated frequency rotor currents. These currents tend to flow in the

rotor iron, slot wedges, and retaining rings, resulting in rotor heat­

ing. The capability of machines to withstand this heating can be best

expressed as a function of the negative sequence current and of time

[16]. This heating will not be excessive if the following condition

is satisfied:

a 2 )2t < k,

where l£ is the negative sequence current in p.u. stator current,

t is the duration of the fault in seconds, and

k is a constant for the machine.

Typical k values are tabulated below for design reference:

Type of Machine k

Indirectly cooled Turbine Generators 30

Conductor-cooled Turbine Generators (up to 800 MVA) 10
Conductor-cooled Turbine Generators (800 MVA to 1600 MVA) 5
Hydraulic Turbine Generators 40
Engine Driven Generators 40
Synchronous Condensers 30
Frequency Changer Sets 30

Since the capability of machines to withstand unbalanced faults

can be expressed as a function of (I2 ) t, the phase-balance current

relay selected for generator protection shall have a time characteristic

matching the (I 2 ) t characteristics of the machine.

There has been controversy on whether or not negative-sequence

relaying is a primary protection. A fault which results in excessive

negative-sequence heating will usually be detected by one of the first-

line relays, either on the generator or on the system. In this respect,

negative-sequence relaying may be regarded as backup relaying. On the

other hand, none of the other relays are designed to protect expressly

against this condition, nor can the negative-sequence relay serve as

backup for balanced faults. Then, it seems proper to consider it as

primary protection. Regardless of the above arguments and in view of

possible major damage from overheating, it is recommended practice

that all major generators be equipped with negative-sequence relays

and that the relay trip the generator instead of initiating an alarm.

In case the negative-sequence relaying is intended as backup, it

is essential that the relay time characteristic be coordinated with

other relays in the system. If required, additional time delay units

should be used to prevent undesired tripping due to system temporary

disturbances. In the application of this relay, it is also important

to check current transformer performance to assure that a false unbal­

anced current caused by current transformer errors will not trip the

unit.

(D) System Fault Backup Protection (Device No. 51^/21)

Generator backup relaying is applied to protect either the machine

itself or the external system in the event of failure of primary pro­

tection on the machine or system. The application and setting deter­

mination of backup relays are usually more difficult than for first-line

relays. Not only is it necessary to foresee the system emergency con­

ditions asking for the relay to operate, it is further necessary to

know the operation of other protective devices under these conditions.

Because the application of backup relays depends so directly upon sys­

tem conditions, no specific protective scheme can be recommended which

will be suitable for all cases. However, it is general practice that

some form of backup relaying be provided to open the generator main

circuit breaker in case of system uncleared faults.

Following are two schemes which can satisfy requirements of most

generator protection. One scheme uses three voltage sensitive time

overcurrent relays, the other uses three single phase distance relays

in conjunction with a timer. The choice between these two schemes is

mainly governed by the type of system protective relaying with which

these relays are to be selective. In general, the former scheme shall

be used when the generator is directly connected to a bus at generator

voltage, because such relaying is most selective with inverse time

overcurrent characteristics generally used for lines connected to such

a bus. The latter scheme is used with unit generator-transformer

arrangement because the relay is most selective with distance or pilot

wire relaying generally used for high-tension lines. In both schemes,

the current source should be current transformers at the neutral ends

of the generator windings if such CT ’s are available. With such con­

nection, the relays will provide generator fault backup protection as

well as system fault backup protection even if the generator main cir­

cuit breaker is open or there are no other generation sources in the

system. If the neutral CT ’s are not available, then line-side CT ’s

should be used. With such connection, the relays will backup the

system and generator only when the generator main breaker is closed

and there are no other sources of generation in the system. One point
which deserves special caution in the design of external fault backup

protection is that the phase-to-phase voltage shall be obtained from

generator potential transformers. Loss of potential to the relays due

to fuse blowing may cause misoperation of the relays. This can be re­

solved by the use of voltage balance relays (Device No. 60).

For protection of uncleared ground faults in the station bus, and

the system, an inverse time overcurrent relay shall be used to provide

backup protection. For directly connected machines, this backup pro­

tection can be obtained by connecting the relay across an additional

CT on the neutral of the machine. For unit connected machines, a third

harmonic restraint overcurrent relay in the generator neutral can de­

tect ground faults only up to and within unit transformer low voltage

windings. In order to obtain backup protection for uncleared faults

on high voltage buses, transformer windings, or the system, another

ground fault relay fed from a CT in the transformer high voltage neutral

should be used.

In any scheme, backup protective relays should be arranged to trip

the main circuit breaker, the field breaker and also shut down the

prime mover.

(E) Overload and Thermal Protection

For most synchronous generators, overload protection is commonly

supplied by temperature measuring elements embedded in the stator wind­

ings which will initiate an alarm. Thus, the condition can be corrected

either by reducing the load, or by increasing any cooling means. Simi­

larly, overheating of bearings (Device No. 38) shall also be alarmed by

the same means. However, it is not general practice to provide relaying

against rotor overheating since this heating is not measured directly.

Usually a negative-sequence current relay is accepted as a substitute

to indicate rotor heating.

(F) Voltage Relaying (Device No. 27, 59, or 60)

Voltage relays can be used for protection against alternating cur­

rent over or under voltage, for permissive control or tripping of other

devices, and for ground fault detection on equipments and feeders as

discussed in the preceeding paragraphs.

Overvoltage relaying is recommended for prime mover generator in­

stallations, especially for hydro-electric generators, which may be sub­

jected to overspeed and consequent overvoltage upon loss of load. In

this application, the relay shall be arranged to open the generator and

field circuit breakers. It may also be used as backup relaying for a

voltage regulator and for its potential transformer in any installation

since most overvoltage, that is overexcitation, cases are caused by

leaving the voltage regulator on while generator speed is slowing down.

In this application, the relay will either sound an alarm or initiate

other equipment to reduce the generator excitation.

It is recommended practice to provide undervoltage relaying for

synchronous converters and all alternating current motors. The relay

will disconnect the armature circuit from the system upon loss of volt­

age or upon the occurrence of low voltage. Usually this protection

should function with time delay dropout to prevent shutdown on transient

voltage dips.

Voltage balance relaying is used to block relays or other devices

that will operate incorrectly when a potential transformer fuse blows.

Some of these relays and devices are voltage restraint overcurrent re­

lays, voltage controlled overcurrent relays, loss of excitation relays,

synchronizing relays, automatic voltage regulators and some impedance

relays. In this application, the voltage balance relay shall initiate

an alarm as well as block the misoperation of other devices.

Another typical application of voltage relaying is capacitor

switching control and automatic transfer of power supplies to critical

circuits. In cogeneration applications, an induction generator may be

operated in parallel with a capacitor bank. Overvoltage relaying is

recommended for the capacitor protection. This application will be

discussed in more detail later.

(G) Overspeed and Over Frequency Relaying (Device Ho. 81)

Since most modern generators are equipped with some type of mechan­

ical or electromagnetic speed regulator, overspeed protection is not

always necessary. In the absence of a mechanical overspeed protection,

frequency relaying can be used to detect an over speed condition indi­

rectly. However, protection by mechanical devices is preferable because

operation is more direct and does not depend on the presence of alter­

nating current voltage.

(H) Rotor Ground Detection (Device No. 64F)

Although modern generator excitation systems are almost invariably

operated ungrounded so that an accidental ground will not cause major

damage, some means of continuous supervision of the field circuit to

detect the first ground is generally applied. The detector will be

used to initiate an alarm rather than trip the generator.

 (I) Synchronism-Check and Synchronizing Relays [1]

The synchronism-check relay is used to verify when two alternating

current circuits are within the desired limits of frequency and voltage

phase angle to manually switch them to operate in parallel. These re­

lays can also be employed on switching applications on systems known

to be normally paralleled at some other location. In this case, they

are only checking these two sources to insure that they have not become

electrically separated or displaced by an unacceptable phase angle.

The synchronizing relay, on the other hand, monitors two separate

systems that are to be paralleled and automatically initiates a switch­

ing action when the beat frequency, voltage deviation and phase angle

difference are within the acceptable limits.

(J) Out-of-Step Protection [16]

As generator per-unit reactances have steadily increased over the

years, and inertia constants have decreased with increased machine

ratings, these conditions intensify the need for out-of-step detection

as part of generator relaying complement. Ideally, generator out-of-

step relaying should be designed to open circuit breakers at a few pre­

determined locations so that generation and load in each subsystem are

reasonably balanced.

The basic element of an out-of-step relay is an impedance unit in

conjunction with a timer, a directional unit, or another impedance

unit. The relay differs from the loss-of-excitation relay in that it

is usually equipped with undervoltage supervision.

3.4 Protection of Synchronous Generators

All power systems, whether utility, industrial, commercial, or

residential, have a common purpose of providing electric energy to the

utilization equipment as safely and reliably as is economically feasible.

At present, synchronous generators play a core role in supplying elec­

tric energy in all these systems. This section will cover the minimum

protective features for synchronous generators.

In industrial and commercial power systems, local generators are

quite often included to either supply all or part of the total load or

provide emergency power in the event of a failure of the normal source

of energy. Generators used to supply all or part of the required sys­

tem energy are operated almost continuously. They are usually turbine

generators. Generation is usually at the system’s higher voltage

level. Those used to supply emergency power are normally shut down

and are briefly operated when the normal power source fails or during

maintenance, testing and inspection. They are usually diesel engine

or gas turbine driven. Generation is usually at the voltage level of

the largest load to be served. Generally, generators for industrial

and commercial power systems are connected directly to the systems.

Figure 3.4 shows the recommended protection for generators that are

direct-connected to the system [4]. This figure is self-explanatory

except 87G Relay. This relay is used to supplement the differential

relay (87) where an impedance is applied to limit the maximum ground-

fault current to less than the generator full load current. A current-

polarized directional relay may be used for this application, its op­

erating coil being connected to the differential circuit and its polar­

izing coil from a neutral current transformer.

Before leaving this topic, it is proper to clarify the following

 D W G 1.. iNI 0 1 IR E V

RECOMMENDED PROTECTION FOR SYNCHRONOUS iP o ^ M i^ d r S v s t g in n s

GENERATORS, DIRECT-CONNECTED E ngineering, Inc.

H ouun si o n .. T f » j : % 7 7 © '2 4
BV JL JO ® N O '. N/A .D A T E Sh 1 orf X
TJ
'Qj
4J.
QJi >1 c
e +J id ■fd
-HI 11^ •H 1— 1 4 -» w >1

ffl
£
0
T3
'ffl 'C
&
ci
ltd
1— !! c
m > , , (0
IQ rH '
tji' +J 1— t © H. rH ffll
ffl' >11 0 ©■ 4J a ffli a ;
G iffli i—m X: rH ’ ■H C ffli (X;
■HI •H I »— t u ffl ‘h ©I 1— t irp!
*o
ffl ffll' if) C 0 iHI Amdi. rH td'
IP 0) a ■"H: irfll *H1 fH 4 -J' P (fl £
o f
irO
o
;|. 1.
t-l'
OQ 0 0 rH ■HI ■0 irH
m © j p
£=
V_-‘ m U ffl •.HI
■HI
u jj' •HI nj
JZ
— m 4 -J1
Re

,rC c fd -H i fd. ff l ■fd. c 4 U>' *,H

IH
m © [Lii 0 O +J
>4
c ffl' tfl >
>■ +jP 113 C 0 O S ffll IM ffii
u
c | j'
k in» B--II IP' k ffl'
e
(ti
O' u O c ffl'
—1 ■
m M p 0 © -HI 1© ri£! ffll k
a; 4J 3
0 1 O'
e T3
O

8 IE-*
© t r O' c c * w Q © K
05 iK n 0 © U' tT1 0 ;m C © ■ -iH -MM
© k c £ Q *0' 'MM!
r—fl ltd 1 +J
ro
"-H. >
© 1— ! ii***
.P ■HI CD
©
W'
0 0 (0
tw C[ -iHI: *HI
(k CJ O'
to ■>— I B +J H Q
O ' id
rH rH C O1 4 -» c ip ; Xi O' c iP
1—1! .m c
0 c C ■HI
ss:
q O (d 0 ffll II
ip ft. k 1© '» k ■
■ ■H O p nj >1 e ra + J ■ffl >1: f f l i ffl' IMi + J
J p k P © -P > - i p 1 ,rC , l* rt ir f ll ■H u e fli to O' ( d 11— II
§o c P 4J 0 c O' *H 1© ir— H X I IH ■B-1 ; k JM •H i
O ' c O ,P © 0 ffl -P e ffl' O' © a!' ffli 0 |RJ,
1 2
H u ffll ■
iIM' 'ffli c O' ■--- V lH! a ms 1 " e i l^ll iH I C l > pij. Ipb x x
I, u © C (0 r fij •H ia © H ffl' ■HI iHj C II
P. ffl k > -P' '<4-11 ffll O +j 4 Jt + J' ;> O -IP' H 00 4J 131

S£
,r-J
y
0 3 O u c -H I :rC 0 M U c •HI 4J :P c c:
4 J,
(X rt O 0
4 J.
td
4 JI
c k 5-41 0 © ,rC ’ C b *T3 I 0 ffl' ,p
4J
k ffll 1© P © ffli a; ffl in JZ
0
u CU rC +J IMl tfd. iH I ■0 0
w ■rH © Ei IQ. W ■ to O' £ ,+ ji (C lMl t T ' .0 ' +J ■O' X c: shi

9 :p, k O' Ip j 4J1 p ■ CL) Q *H1 mi p 0

0
m > •H I k c •H I O *
Q > O IE-: ■ffl' H t j. P hi ■HI £ CO 0 2 Cz-,1 ft »3 1
'4 ■+J1 B

I) 11 I i 1
I
! III III I 1 CP'
I III III i il *
t It ! i I ■ • IP
k
If III I H I
3
III I I
III I:
O ' t7>
III
I > “iHf
1
1
I
II.
II
I ||Z4
II 0 } II
I
!
$
o o 1 > III 1 O '
O' (M O ' H0 O n,C ir -
til m H£« \0 ffl '0 0 CO'

iHII

■**°
2 Z
u k ...
0 0 '<M>: HUH
■H •HI *ir+
.HA
3 irH S =
& m 0 H
© c ETt ■CP iHI
IP O 'U > 10 ufl :>
-"iH' ■ffl X ■*)■ IE
10 41 iiH ; irH1 m
ji—ii^ O z. ,^_1 'Tjr
iflj O' > •I > IP*;
.iHI1 '13 K r«ic
ffll. 1® ■® > . 'ffl
Ik liM' rHf um 0 c irH1 i i i
fflt iHI > m r-r b
HIIIM •cl ® Q ■w (0 .HI
O' * 0 O' > M 0 "0
g «r < H^JI-
*4 xz ® > ffl' ffl
0 «
r» fijl jQ < K II-PI nX-Jl <
c: " t l 1 ifHII > 1—4 :>
zs irH1, < X O O ifHI < X
c P : -HI > 0 O .HI i> 0
5 % 0 ■ ifHI1 3 ' *4 0
m 0 0 0
m s z ■0® 0 ITT+ ■ffll ■IB O ■—i-
Oil k 0 1 > IM O 1
0 O' lull O O uri O '
c lixi. 4J 1. 1 O 4 .a O'
0 ffll ffl! O ' Ufl us 0 C i
Hji 0 1 — Hi
*r+ ■ffll ©,
J> C' ,»— >
!■ C
llU ,J3i D f fl itf '"^1
"W DS?- a©' ""l— ■

■r,i «rj.

05
04
two points. First, where suitable current transformers are available,

the flux balancing differential scheme shown in Figure 3.2(c) may be

used. The relay for Device 87 should be of instantaneous type in this

instance. In general, the sensitivity of this scheme should be selected

as such not to require the use of 87G. Second, it is necessary that

the static exciter, if energized from the stator circuit of the genera­

tor, shall include some means of maintaining excitation should the

stator become short circuited. This is so because this condition may

result in insufficient flow of fault current to ensure proper relay

operation.

Contrasted with the direct connected system, the unit system is

commonly used in generating stations. Figure 3.5 shows an overall pro­

tection scheme for a tandem compound unit connected generator. When a

station service transformer is included in the zone of the generator

differential protection, it may require that 87G utilize harmonic re­

straint. Generally, the type of generator prime mover will determine

whether it is necessary to do so. In addition to the differential re­

lay 87G protecting the generator, another percentage differential relay,

87T should be installed as shown in the figure. This relay serves as

primary protection for faults in the generator leads and in the trans­

former as well as backup for the generator differential relay 87G.

Current transformers for this relay should be installed on the bus

side of the breaker in order to provide overlapping protection between

the bus and the power transformer. Although the station service trans­

former is included in the power transformer protective zone, the dif­

ferential relay 87T cannot be expected to protect the station service

 DW G. NO, REV „

A Pow er S y ste m s
UNIT GENERATOR TRANSFORMER ARRANGEMENT
E ngineering, Inc.
p.o. Box 19 39 8
JL JOB N O . N/A DATE 1 H ou st on , Tex as 7 7 0 2 4
Sh of

Figure 3.5 UNIT GENERATOR TRANSFORMER ARRANGEMENT

transformer due to the relatively small size of this transformer. A

separate transformer or an overcurrent relay supplied from C T ’s in the

station service transformer HV bushings will probably be required.

Another overcurrent relay is also recommended to trip the station ser­

vice LV circuit breaker only as a backup for the circuits on the sta­

tion service bus.

Figure 3.6 illustrates an overall protection scheme for a cross­

compound unit connected generator. It is self-explanatory except the

following point. On cross-compound units, field is usually applied

while the set is on turning gear. Application of field to such machines

normally excites the generator step-up transformer as well as the gen­

erator so that protection is required for all of the unit system. The

filter circuit of some harmonic-restraint relays may cause failure to

operate if a fault occurs before the machine is up to synchronous

speed. Since relay designs vary, the relay performance at reduced fre­

quency should be checked.

3.5 Protection of Induction Generators

Recently, the cogeneration concept has gained the intense interest

of electric utilities, industry, government, and most facilities with

substantial energy requirements because of the increasing cost of fuel.

Cogeneration is defined as the production of electric power and some

other form of useful energy such as heat or process steam in the same

or adjoining facilities.

Either the induction generator or the synchronous generator can be

used in cogeneration applications. The induction generator has the ad­

vantage of simplicity in design and operation. The construction is

 OVERALL PROTECTION SCHEME P ow er S y ste m s
OF CROSS-COMPOUND GENERATORS % J y E ngineering, Inc.
P .O . Box 19 3 9 8
JL JOB N O . N/A DATE
H ou st on . Tex as 7 7 0 2 4

21 Distance Relay
NOTE: ( ) shows number of 40 Loss-of-Excitation Relay
relays required 41 Field Circuit Breaker
46 Negative-Sequence Current Relay
50 Instantaneous Overcurrent Relay
51v/27 Voltage Sensitive Time Overcurrent
Relay
51N Time Overcurrent Relay
64N Voltage Sensitive Ground Fault Relay
64NT Voltage Sensitive Ground Fault Relay
63 Sudden Pressure Relay
87G Generator Differential Relay
87T Transformer Differential Relay

Figure 3.6 OVERALL PROTECTION SCHEME OF CROSS-COMPOUND GENERATORS [16]

that of a squirrel cage induction motor. The field windings, provisions

for rotating excitation supply to the field or separate field excitation

equipment and slip rings as well as voltage regulation equipment are not

required, but would be for a synchronous generator. The induction gen­

erator always assumes the same output voltage and frequency as that of

the external system, so synchronizing equipment is not necessary. The

protective relaying is essentially the same as that which would be pro­

vided for an induction motor of comparable size. However, the induction

generator cannot be operated as an independent source when the external

source is down because of its need for an external magnetizing power

supply. Consequently, in an industrial plant where an uninterrupted

supply is vital, the synchronous generator should be used.

Figure 3.7 illustrates a protection scheme for an induction gen­

erator. Unlike the protection scheme for a synchronous generator, the

loss-of-excitation relaying is not required. Also, the flux self-

balancing differential scheme is generally used for induction generator

protection. The differential relay used can be the inverse-time over­

current type, not the percentage type. The ground fault protection re­

lay (50G) is generally fed from the residual of the primary side current

transformers. Where a shunt capacitor bank is used to improve the over­

all power factor, the voltage relaying is required to protect the gen­

erator against over and under voltage. There are considerable varia­

tions in the industry in the applications of the anti-motoring protec­

tion to the induction generator. Where the machine is designed to

operate both as a generator and a motor, the relay should be arranged

to initiate an alarm but not to trip the unit circuit breaker. Where
 DWG. NO. REV.

PROTECTION OF AN INDUCTION GENERATOR

P ow er S y ste m s
*E ngineering, Inc.
1 Houston, Tex as 7 7 0 2 4
JL JOB N O . N/A DATE Sh 1 of

n w esr c v -r u o D e p ^ L f-o ^ e p h l a ’v j*|

32. v .e s t! C \J ■DiR££.'f/nkJA'- T O ^ e R ‘P ttL A V

5 9 )5 .1 5 v j£ t)'f. C A Ex*=?. ikW £R5 £ CORRECT f?cLA Y

V , £ 3 l 5 (t A .lit. r e la y

H I* vJ=5T, c m P M A 6E B A L A tO C t R E L A Y
S o G r j v j£ 5 1 ”. it H i w i T ^ B O J R R e M f F .E lA Y
Z \G c V J tft D V c f i^ O L t A e E F ? £ L A Y
67 | W S 5 1 If H < j£ > i. D t r F E R E U l l A L RELAY
EX> i ^ N 'T C M P c r? A fU 'R £ M O M lfO f?
S o O / c l u J E j T .u J U UXKO l ^£LO Y
k o iC l ' * j E 5 l I'M \w y T O V 6 R C O R R E U T -R E L A Y
< W £ R : \t f L l4 < j£ R iU k y |
{ { iC P , I i* r U R i3 ilo £ S H O f C O w O O R £ L A .y j

Figure 3.7 PROTECTION OF AN INDUCTION GENERATOR

the machine is designated to operate only as a generator, the relay

should trip the unit circuit breaker and shut down the generator. In

some instances, reverse power relaying is provided to prevent power

flow into the utility's incoming line. The relay for this purpose is

usually arranged to trip the main incoming breaker. However, this

type of application should be carefully studied for the particular

installation since the tripping of incoming main breaker might be far

more harmful than a reverse power flow to the utility.

 CHAPTER 4

TRANSFORMER PROTECTION

At first thought, a power transformer might seem to be a highly

reliable device, and this is correct as far as the mean failure rate

is concerned. Yet, if we consider the mean down time per transformer-

year, it is obvious that transformer protection is no less important

than generator protection [4]. Contrasted with the several types of

abnormal operation to which generators may be subjected, a transformer

presents a much more limited protective problem.

A transformer may be subjected to short circuit, open circuit, and

overload. Transformer failures can be grouped in the descending order

of occurrence as follows [17]:

(1) Winding failures are the most frequent. Reasons for this in­

clude insulation failure and voltage surge. Voltage surge resulting

from lightning and switching transients should be protected by voltage

surge protective devices such as rod gaps and arresters. Two principal

causes of insulation failure are overheating and overvoltage. Over­

heating because of overloading or failure of the cooling system will

deteriorate the insulation. Insulation thus weakened may fail when it

is stressed by only a moderate overvoltage. Insulation may also fail

when a transformer is returned to service without first making sure

that the insulation is free from moisture. Ground faults in ungrounded

systems and sudden load shedding can subject transformer windings to

severe voltage stress which may also cause winding failure.

(2) Tap changer and terminal board failures are attributed to

improper design, assembly, handling in transportation, or mechanical

contact problems.

(3) Bushing failures due to aging, contamination, cracking, etc.

(4) Core failures involving clamping brackets or short circuited

laminations.

(5) Miscellaneous failures including current transformer failures,

liquid leakage, etc.

Transformer protection is generally achieved by current sensing

devices, mechanical sensing devices, or more frequently combinations

of the two. Current sensing devices including fuses, overcurrent

relays, differential relays and pressure relays are used to protect

transformers against short circuit. Mechanical sensing devices, in­

cluding thermal relays, gas relays, etc., are designed to detect ab­

normal operating conditions.

Compared with other equipments in power systems, power transformer

protection is different from others in the following aspects:

(1) Different voltage levels, including taps. This implies dif­

ferent primary current ratings in the connecting circuits, and conse­

quently different current transformer types, ratios, characteristics

with possible mismatch of ratios, etc. This might create problems in

relaying.

(2) Magnetizing inrush current which appears to differential re­

lays as an internal fault.

(3) Phase shift between the primary and the secondary.

Since published application guides covering transformers are few in

number [17], protective relaying engineers must rely on their sound

engineering judgment to achieve an adequate or optimal protection

system. This chapter will outline some important points in designing

power transformer protection systems. Special emphasis is put on power

transformer relaying. Since magnetizing inrush has definite impact on

relay function, it will also be briefly covered.

4.1 Philosophy of Power Transformer Protection

Among the current sensing devices for small transformer protection,

fuses are widely used for this purpose. Although fuses have certain

limitations, they have the merits of being economical, simple to main­

tain, and generally failsafe. This is true for many transformer sizes;

however, transformers exceeding 2 MVA in rating are usually provided

with more sensitive protection. Fuses can be reliably used to protect

against external short circuit currents under most conditions, but can­

not be relied upon to protect the transformer from internal faults or

long time overload.

If considering fusing of a transformer, the following factors shall

be taken into consideration in the selection of fuse links.

(1) Time current curves published by manufacturers should be cor­

rected for ambient temperature conditions and prefault load currents.

(2) Maximum permissible load currents and the transformer heating

curve recommended by standards.

(3) Transformer connections.

(4) Available primary system fault current .and the transformer

impedances.

(5) Coordination with low-side protection equipment.

(6 ) Maximum allowable fault time on the low-side bus conductors.

(7) Maximum degree of sensitivity for protection from high imped­

ance faults.
 Where more sensitive protection is required, it can be achieved by

proper selection and combination of different types of relaying. Power

transformer relaying in general includes abnormal operation sensing de­

vices to protect the transformer and fault detecting devices to protect

the system and limit damage in event of a fault. Abnormal operation

sensing devices include thermal relays, pressure relief devices, gas

detector, liquid level indicator, combustible gas relay, etc. Fault

detecting devices include sudden pressure relays, overcurrent relays,

differential relays and system backup relays. Again, differential re­

laying is the principal form of transformer protection. However, this

type of protection cannot be as sensitive as that of generator protec­

tion due to the factors stated above although it does protect turn-to-

turn and phaseto-ground faults. Pressure sensitive relays are generally

used to detect either the rate of rise of gas pressure (sudden pressure),

a slow accumulation of gas (gas detector), or more commonly a combina­

tion of both. Such relays are supplemental to differential and ground

fault relaying for fault protection, and to overload relaying for ab­

normal operation protection. System backup protection and other forms

of abnormal operation protection are sometimes used depending upon the

requirements of the individual system design. In most cases, fault cur­

rent sensing devices and combustible gas relays should be arranged to

trip the transformer; on the other hand, abnormal operation relaying

should be arranged either to initiate an alarm only or automatically

shed parts of transformer loads.

4.2 Effects of Magnetizing Inrush on Differential Protection

The small component of current consumed in the power transformer

for producing the necessary flux is negligible for differential relaying

application in steadystate. This is true over the full range of trans­

former operation and even for several times full load current when a

short circuit first occurs, because the unbalance due to the magnetizing

current is small compared with the full load current and far below the

relay pickup.

However, when a transformer is first connected to a power source

after the transformer has been completely deenergized, there is a mag­

netizing inrush from 8 to 30 times that of full load current [16].

There are two other types of inrush, that is, recovery inrush and sym­

pathetic inrush, but of lesser magnitude. There are several conditions

that cause particularly severe magnetizing inrush phenomena. One in­

volves the energizing of a transformer at a station at which at least

one other transformer on the same bus is already energized. The inrush

involves the already energized transformer as well as the one being

energized, and the inrush transient is of particularly long duration.

It is important to realize that the inrush into the transformer being

energized occurs during the opposite half-cycle to that of the already

energized transformer. Thus, the net inrush into all transformers on

the same bus may be nearly a sine wave of fundamental frequency. This

phenomenon makes necessary a separate harmonic suppression relay unit

for each bank. Another inrush phenomenon involves the energizing of a

transformer by means of an air switch. Arcing of the switch can re­

sult in successive half cycles of arc of the same polarity. Thus, if

the first half cycle results in substantial residual magnetism in the

transformer core, succeeding half cycles can cause a cumulative increase

in residual magnetism, each time resulting in a more severe inrush.

Important differences between a magnetizing inrush current and a

short circuit current are:

(1) A magnetizing inrush current has a dominant harmonic component

generally either the second or the third order which is generally not

present in a short circuit current.

(2) Accompanying a short circuit current, there is a voltage dip

which is not present in a magnetizing inrush current.

To differential relays, the inrush current looks like an internal

fault. Some methods commonly used to distinguish magnetizing inrush

from fault include:

(1) Reduced sensitivity of percentage differential relays to the

inrush current, that is, either a higher pickup setting or an intended

time delay in operation.

(2) Restraints by harmonic components, especially the second order,

of inrush current.

(3) Desensitization of the differential relays during energization.

In applying these methods, it is important that precaution shall

be observed for the following considerations:

(1) While it is simple to prevent misoperation of a differential

relay during inrush, care shall be taken that means should not delay

tripping if a short circuit should occur during this period.

(2) When two or more transformers are connected to a common bus

and switched separately, harmonic ■components generally existing in.

inrush currents may not be available for preventing mistripping depend­

ing upon the specific system configuration under consideration. This

is so because a transformer suddenly energized tends to take the har­

monic components of the inrush current from the parallel banks while

the fundamental component comes from the system [17].

(3) In the event of load injection, the transformer may be subject

to substantial overvoltage and saturation. If saturation occurs, sub­

stantial exciting current will flow. In this case there are principally

odd harmonics, particularly the third order. This may misoperate sec­

ond-order harmonics constraint differential relays [17].

4.3 Fault Protection

Transformer fault protection should function to detect any faults

either inside the transformer or outside in the terminal leads but with­

in the transformer zone. It generally includes differential protection,

ground fault protection, and sudden pressure relaying.

4.3.1 Differential Protection (Device No. 87)

(1) Types of differential relays:

ANSI Standard C37.91 [17] lists four general classes of transformer

differential relays as follows:

(1) Differentially connected time-overcurrent relays

(2) Percentage differential relays

(3) Harmonic restraint percentage differential relays

(4) Desensitized percentage differential relays.

Among these, time-overcurrent relays connected in differential

scheme are seldom used in transformer protection due to their suscep­

tibility to misoperation from the magnetizing inrush.

The transformer percentage differential relay differs from that

for generator protection in having higher percentage, that is, larger

characteristic slope so that it has reduced sensitivity to the inrush

current. The required percentage may vary between 15% and 50% depending

on the type of the transformer protected and the system configuration.

Generally speaking, 15% slope is for standard transformers; 25% to 50%

for load tap-changing transformers, depending upon the range of tapping;

40% for special application [4]. Drawbacks of its application are:

(a) Its setting should be above the maximum possible inrush cur­

rent. Since most pickups of differential relays on the market cannot

be set in the field, replacement may be required as the system expands.

(b) Higher percentage means reduced sensitivity not only to inrush

currents but also to fault currents.

This type of relay is particularly applicable to power transformers of

moderate sizes located at some distance from the major source of gen­

eration [17].

The harmonic restraint relay makes use of harmonics, particularly

the second order, present in the inrush current to discriminate mag­

netizing inrush from short circuit. It functions either to block false

tripping or increase the pickup during the inrush period. Consequently,

a lower normal pickup setting and a faster operation is possible without

risk of false tripping. Unfortunately, there are still some problems

as follows:

(a) During inrush, it will either not trip or delay tripping if

a short circuit should occur.

(b) Sufficient harmonics may exist in high internal fault currents

or in the system to inhibit relay operation.

(c) In the event of load rejection, it may respond to overvoltage

magnetizing inrush.

The relay is most suitable for large power transformers located near

large sources of generation where either severe inrush is highly likely

or recovery and sympathetic inrush may create problems.

To overcome the inrush problem, the desensitized percentage differ­

ential relay utilizes the fact that a voltage dip always accompanies a

short circuit. An undervoltage relay is used to control the desensi­

tizing circuit of the differential relay. Under short circuit condi­

tions, the undervoltage relay will remove the desensitizing circuit

from the differential relay. Consequently, it also provides protection

against short circuits happening during the inrush period.

(2) General Guidelines for Power Transformer Differential Relaying

[16] As stated in prior sections, transformer protection design heavily

relies upon engineeering experience and sound judgment. The following

guidelines are designed to assist in selecting and applying differential

relays for power transformer protection. These guidelines do not neces­

sarily apply to special-purpose transformers such as those for power

rectifiers, converters, inverters and induction furnaces

(a) There is no clear-cut answer to the questions of where a

differential relay should be used, and which type should be selected if

a differential relay is required. As a general rule, however, it is

general practice to use percentage differential relays for all trans­

former banks rated 1 MVA and above and with 2400 V and above secondary

voltage [4]. Percentage differential relaying is also recommended for

transformer banks rated below 1 MVA that operate in parallel with

differentially protected banks and have circuit breakers for all

parallel-connected windings. This is so because it may prove just as

damaging to the service as a similar fault occurring in a large bank

if a fault occurring in a small parallel-operated bank is not promptly

removed. For transformers from 1 MVA to 2 MVA and below 15 KV, desen­

sitized percentage differential relays are recommended. For all trans­

formers rated 2 MVA and above, or having windings rated 15 KV and high­

er, high-speed harmonic restraint percentage differential relays are

recommended.

(b) Regarding the proper type of current transformers for differ­

ential relaying, it is general practice that bushing-type current

transformers are used on the high tension side and are balanced either

against other bushing type current transformers on the low tension side

or against instrument type current transformers. Unfortunately, no

general answer can be given to this question, but the procedures for

current transformer accuracy calculation outlined in section 2.4 can

be used to determine the answer to such a question.

In order that the relay will get as much current as possible, and

hence be most certain to operate when transformer short circuits occur,

the ratios of currrent transformers used should be chosen so that they

will supply the relay with current approximately equal to its tap rat­

ing at maximum load. In case the multi-ratio current transformer is

used, a tap which will give approximately 5 amperes at maximum load

shall be used. This arrangement will generally provide good sensitivity

without introducing thermal problems in the current transformer, the

leads, or the relay itself.

In general, the use of low ratio bushing type current transformers

should be avoided. Bushing type current transformers used for differ­

ential protection should have a m i n i m u m of 40 turns [17]. This is be­

cause these transformers show a marked departure from the turn ratio

over the range from very low current to that limited by the power trans­

former on through faults, especially at 345 KV and higher.

For power transformers having three or more windings, all windings

should be treated as having the same capacity as the K V A rating of the

highest rated winding when determining the proper current transformer

rating to use. This is because it is the voltage ratio, not the indi­

vidual K V A 'rating of each winding which determines the current that

will flow to an external fault. For tap changers, it is customary to

choose current transformer ratios and relay adjustments on the basis of

using a transformer tap at the middle of its range. Thus, with a plus

or minus 10% range, the minimum, current transformer unbalance will be

plus or minus 10%.

In applications where a power transformer is connected to the

rest of the system through two breakers, for example a ring bus or

breaker-and-a-half arrangements, the CT ratios must be selected so

that the secondary windings will not be thermally overloaded on load

current flowing around the ring in addition to the transformer load

current.

(c) A useful rule for determining current transformer connections

in differential protection is as follows: The current transformers in

the leads to a wye connected winding should be connected in delta; cur­

rent transformers in the leads to a delta connected winding should be

connected in wye. The connections should be so arranged as to block

the zero sequence current from the differential circuit on external

ground faults, and to compensate for the 30° phase shift introduced by

the wye-delta connected power transformer bank.

For current transformer connections in a power transformer having

three or more windings, consider any two windings first and connect the

current transformers for those two windings as if they were the only

windings; then consider a pair consisting of one of those windings and

another, and so on till finishing all the connections.

Before leaving this topic, it is well to observe precaution never

to ground interconnected current transformers such as those used for

differential protection, at different points on a ground bus or to two

different station grounds. A difference of potential betwen two points

may occur and cause misoperation of the relay when ground current flows.

(d) Differential relays should be connected to receive "in" and

"out" currents that are in phase for a balanced load condition. Where

there are more than two windings, all combinations must be considered,

two at a time.

(e) After the current transformer ratios have been determined,

the relay taps should be so selected that the relay ratio is as close

as possible to the ratio of the secondary current in the current trans­

formers. Then the continuous rating of relay windings should be checked

for compatibility with the power transformer load. If the currents

supplied by the current transformers exceed the relay continuous rating,

a higher current transformer ratio or relay tap should be used. How*-

ever, it is desirable to increase CT ratio in preference to the relay

tap. Since the relay burden is likely to be relatively small compared

 to the lead burden; increasing the CT ratio tends to improve the rela­

tive performance of the CT's as a result of reducing the maximum sec­

ondary fault current and increasing the accuracy of the C T ’s.

(f) The percentage of current mismatch should always be checked

to ensure that the relay taps selected have an adequate safety margin.

Percentage mismatch M can be determined as follows [17]:

in which II, Ir are currents supplied by current transformers on low

and high voltage sides, respectively. Tl , Tr are relay tap settings

for low and high voltage sides respectively. S is the smaller of

(IL /IR ) or (TL/TR ).

For power transformers having three or more windings, all combina­

tions should be calculated, two at a time. It is important that the

total mismatch should include not only M but also load tap changing

values (LTC).

(g) To ensure correct operation of the relaying scheme, the cur­

rent transformer ratio error must not exceed a specific limit depending

upon the specific relay characteristics used. Again, the procedures

outlined in section 2.4 shall be applied to determine the percentage

errors of the current transformers. In the next paragraph, an example

will be used to illustrate checking the current transformer ratio.

(h) The total error of the differential scheme equals M plus LTC

and CT ratio error. This value should be smaller than the relay slope

used by a safety margin, usually at least 5%, to ensure correct

operation of the scheme.

(i) It is not desirable to protect two or more power transformers

connected in parallel with one set of differential protection as pointed

out in section 4.2

(j) On a wye-delta transformer, which has a grounding transformer

connected on the delta size, external ground faults on the delta side

will cause zero sequence currents to flow in the wye connected CTs.

These currents are not balanced by corresponding currents from the

delta connected CTs and may cause spurious operation. To eliminate

this misoperation, a zero sequence current trap can be inserted in

the wye connected CTs circuit to correct the inbalance.

(3) Checks for Applying Transformer Differential Relays

The complete procedure for applying transformer differential relays

is summarized in Figure 4.1 and can be best illustrated by the following

example.

Considering the protection of a three-winding bank as shown in

Figure 4.2, a harmonic restraint percentage differential relay with

the following characteristics have been selected for the protection.

Tap Operating Circuit Restraint Circuit Burden

Setting (A) Burden (VA)____ Burden (VA) Impedance (Ohms)

2.9 3.2 1.3 0 180

3.2 2.7 1.2 0 156
3.5 2.4 1.1 0 140
3.8 2.0 1.0 0 120
4.2 1.9 0.9 0 112
4.6 1.6 0.8 0 0'96
5.0 1.5 0.7 0 088
8.7 0.7 0.5 0 048

The relay has a 25% slope and a time current characteristic curve as

shown in Figure 4.3. The continuous and short time current ratings of
Yes!

Figure 4
A THREE-WINDING TRANSFORMER BANK AND CTS' P ow er S y ste m s
CONNECTION FOR DIFFERENTIAL PROTECTION E ngineering, Inc.
P O. Box 19 398
Hous to n. Tex as 7 7 0 2 4
IV JL JO B N O . N/A DATE Sh 1 of 1

110k v /44 k v /13.8 k v

30 MVA OS/37.5 MVA FOA

4 A

U J ^ 44KV
110 K V A
— m -

i
P A 1 3 .8kv

tr
= 5

L-L Voltage Transforiner Connections CT Connections

la

110 KV

IB

44 KV

*la

13.8 KV

klB

Figure 4.2 A THREE-WINDING TRANSFORMER BANK AND CTS'

CONNECTION FOR DIFFERENTIAL PROTECTION
 D W G .N O . REV.
THE TIME CURRENT CURVES FOR THE P ow er S y ste m s
TRANSFORMER DIFFERENTIAL RELAY E ngineering, Inc.
P . O . Box 1 9 3 9 8
BY JL JOB N O . N /A DATE Sh of 1 Houst on, Te x a s 7 7 0 2 4

Figure 4 .3 THE TIME-CURRENT CURVES FOR THE TRANSFORMER DIFFER EN TIAL RELAY
the relay are 10 A and 220 A for one second, respectively. Further­

more, assume the maximum available fault level for a three phase fault

is 1.6 GVA. Determine the differential protecting scheme for this

transformer.

The designing procedure of the scheme is illustrated step by step

as follows:
TRANSFORMER WINDINGS
110 KV (Y) 44 KV (A) 13.8 KV (Y)

(i) Select CT ratio:

ip = k v a / ( y r v ^ )

= 37,500/( Vi-i) 197 A 492 A 1,569 A

CT connections delta wye delta
relay current factor ( ) ✓3 1 JT
CT ratio IpKj 400/5 500/5 3000/5

(ii) Select relay taps:

Relay current (A) 4.27 4.92 4.53
tried relay taps (A) 4.2 5 .0 4.6

(iii) Check mismatch errors:

4.27 _ 4.2
Mismatch error between 4.92 5.0 '
110KV & 44 KV 4.2

3.0 %
5.01 /
4.27 _ 4.6
Mismatch error between 4.53 5.0' X 1002
110 KV & 13.8 KV 4.6
5.0
•3w
2 r '/o

4.92 _ 5.0
Mismatch error between 4.53 4.6 X 1002
44 KV & 13.8 KV 4.92
4.53
0.082
All are less than 5%, therefore, mismatch error is not excessive

(iv) Check relay ratings:

continuous relay ratings (A) 10 10 10
maximum fault current (A) 8.398 20,995 66,941

= 1.6 GVA/( J T V ^ )
relay fault currents (A) 181 210 193
Select desired operating time according to the system requirement. From

Figure 4.3, assume the required operating time is less than 3 cycles at

8 times relay tap value. This means that the current transformers

should be able to supply the relay with 8 times rated relay tap current

with an acceptable error. On the other hand, the relay short time cur­

rent rating should be at least greater than the fault current supplied

by the C T s . Given short time current rating of 220 A for one second,

the short time rating for 3 cycles can be calculated as following:

( 2 2 0 ) 2 x 1 = ( i ) 2

i = 983 A for 3 cycles

Therefore, both continuous and short time ratings are not exceeded,

(v) Check CT ratio errors:

Assume one way resistance is 0.25 ohms. The internal resistances

of the CT's are 0.15-ohms for 400/5, 0.12 ohms for 500/5, and 0.09

ohms for 3000/5. Furthermore, assume that there are no other de­

vices connected to these C T ’s. Then the total impedance burden

on each CT can be easily calculated according to section 2.4.

The results are:

Impedance burden, Ohms (Z) 1.33 0.49 1.23

8 times relay taps, amps (Ig) 33.6 40.0 36.8

Secondary voltage required (ZIg) 44.7 19.6 45.3

Excitation current from

excitation curve (Ig) 2.5 1.2 2.0

% ratio error (Ig/Ig) 7.4% 3.0% 5.4%

(vi) Check total errors:

Maximum of M + CT ratio errors = 10.4%

 (vii) Safety margin available = 25% - 10'.4% = 14.6% minimum.

Therefore, all CT taps selected are satisfactory.

4.3.2 Sudden Pressure and Buchholz Gas Protection (Device No. 63)

Buchholz relays can be applied to conservator type transformers and

sudden pressure relays, to sealed tank construction transformers. These

relays are more sensitive to light internal faults than the differential

relay and generally recommended for all units of 5 MVA or more [16].

With the application of this type of pressure relay, many transformers

can be simply protected by a plain differential relay with higher set­

ting instead of harmonic-restraint type. Generally speaking, the relay

should trip a separate lock-out relay from that tripped by the differen­

tial relay. In case the transformer is protected against short circuits

by fuses, the relay should initiate an alarm.

4.3.3 Ground Fault Protection (Device No. 50/51N, or 67N)

It is general practice to apply ground overcurrent relays to each

transformer winding which is supplied from a grounded system, to detect

possible ground faults within that winding.

Where the transformer primary winding is not a zero sequence source

and a single donut type CT can be fitted around all three phases, a

sensitive instantaneous ground overcurrent relay should be used. If

a donut type CT cannot be fitted, an inverse time/instantaneous over­

current relay fed from the residual circuit of the phase CTs should be

used.

Where a transformer primary winding is a zero sequence source, a

sensitive instantaneous relay generally cannot be used since it will

operate for primary system faults. An inverse time/instantaneous

 relay can be employed only if it can provide coordination with the sys­

tem protection for both transformer and system faults. If the coordin­

ation required is difficult or even impossible to achieve, a product

type directional ground overcurrent relay, current or potential polar­

ized, should be used. If the current polarized relay is used, the

phase and neutral CTs shall be relatively close together. If they are

remote from each other, such as those in a transformer feeder, the po­

tential polarized directional ground relay can be used. This relay

should be fed from the phase CTs and polarized from an open delta

potential transformer.

In general, an inverse time overcurrent relay fed from a CT in the

grounded neutrals of all secondary windings and grounding transformers

is used to provide transformer protection as well as system back-up

protection.

In applying ground fault relays, precaution should be observed that

only those circuit breakers connected between all ground fault current

sources and the transformer will be tripped. This should be carefully

studied and planned for the particular application. If the ground fault

current can flow in either direction, the directional overcurrent relay

should be used.

4 .4 Abnormal Operation Protection

Transformer abnormal operation protection is designed to protect

the transformer against any abnormal operation such as over-temperature,

uncleared system faults, gas accumulation, and high pressure. It gen­

erally includes over-temperature protection, system backup protection,

gas sensing and pressure devices. In this section, all mechanical

 devices, including gas sensing and pressure devices will not be elab­

orated since their applications are quite straightforward.

4.4.1 Over-Temperature Protection (Device No. 49)

Over-temperature relays, whose operation depends on the temperature

of the transformer windings or the oil, are inherent selective and are

recommended for all transformers 10 MVA and above [4].

These relays can be depended upon to trip the transformer main cir­

cuit breaker only when it becomes necessary, especially in unattended

stations. In attended stations, they are generally arranged to initiate

an alarm when the transformer overheats.

4.4.2 System Backup Protection (Device No. 51, 67, or 21)

A fault external to the transformer zone will result in an overload

which can cause transformer failure if the fault is not cleared promptly.

The transformer can be isolated from the fault before damage occurs by

using either phase overcurrent relays or distance relays. As in gener­

ator protection schemes, the choice between these two types of relays

is mainly determined by the type of system relaying with which the

transformer relaying should be coordinated. As a rule of thumb, the

same type of relaying should be used in both relaying systems, if

possi b l e .

Compared with generator protection, phase overcurrent relaying is

far more complicated because of the wide range of load current and more

coordination requirements in transformer protection. Generally speaking,

system backup overcurrent protection should be fitted to all phases on

the source side of the transformer. Overcurrent protection fitted to the

load side of the transformer is considered as protecting the secondary

system. For the transformer protected by separate differential scheme,

phase overcurrent relaying also backs up the differential protection.

For the transformer without differential relaying, it is considered as

main protection for the transformer.

There are two types of overcurrent relays generally used for trans­

former protection. One is the inverse time overcurrent relay. This

relay should have two elements, that is, instantaneous trip and inverse­

time characteristic. The required degree of inverseness depends upon

the particular application. To allow transformer overloading when

necessary, the pickup value of the inverse unit should be set above

this overload current. Settings of 200% to 300% of the transformer self­

cooled rating are common, although higher values are sometimes used.

Fast operation of the inverse time unit is usually impossible, since

this relay must coordinate with all other relays it overreaches. Con­

sequently, the applications of the overcurrent relay is limited by the

insensitive setting and the delayed operation for coordination. Before

advancing to the other type of overcurrent backup, it is proper to

point out an application which requires special caution. Where an over­

current is used to protect a grounding transformer, it should be fed

from CTs connected in delta to prevent misoperation due to system

ground faults.

The other type of overcurent relays generally used is the direc­

tional overcurrent relay. This relay provides best protection where

fault current can flow in either direction. Generally, the relay

should be connected to operate when the current flow is into the trans­

former. The directional element should be fed from the same CTs as the
overcurrent element and, in addition, a polarizing voltage derived

from a PT. In the applications of the directional overcurrent relay

on a transformer primary which is a zero sequence source, it is general

practice to use an inverse time overcurrent relay, fed from a CT in the

primary winding neutral, to provide primary system backup protection.

Of course, the relay should coordinate with other system relays. More

discussions on the overcurrent backup protection will be covered in

the next section.

The application of distance relaying on transformer backup protec­

tion is quite similar to that on generator backup protection. The dis­

tance relay used should be of the directional type. It should be con­

nected to operate when the fault current flows toward the protected

transformer. The setting should reach into, but not beyond, the trans­

former .

4.5 Protection of Power Transformer

Industrial applications of power transformers fall into the follow­

ing four general categories; generator transformers, system intercon­

nection transformers, power utilization transformers, and special pur­

pose transformers such as voltage regulator, phase regulator, and so

on. In this section, protective relaying for each category, except

special purpose transformers, will be discussed. Some protection

schemes are also given to illustrate the considerations covered in the

first four sections. Since protective relaying design is really an art,

the schemes given are for illustration only and should not be taken as

standards. For any particular application, individual engineering study

is required.
 4.5.1 Protection of Generator Transformers

Since transformer protection of a unit connected generator is cov­

ered in section 3.4, no duplication will be given here. A protective

scheme for a direct connected generator transformer is shown in Figure

4.4. Generally speaking, the transformer is a step-up transformer

with its delta winding connected to the generator bus. The secondary

winding can be either delta or wye connected depending upon the exter­

nal system connection. The ground overcurrent relay (51G) in the

grounded neutral of the wye winding should be of inverse time type and

be coordinated with the system ground fault protection for a system

ground fault. Similarly, the ground fault overcurrent relay (51G) in

the grounded neutral of the grounding transformer is designed to backup

generator bus ground fault protection and should be coordinated with

generator bus ground fault relaying. In this scheme, a zero sequence

trap is required on the primary side since a grounding transformer is

shunt across the delta winding.

4.5.2 Protection of System Interconnection Transformers

A system interconnection transformer is used to interconnect two

systems such as from transmission system to subtransmission system, or

between two transmission systems of the same voltage, and so on. It is

normally a wye-wye transformer or an autotransformer with a delta

tertiary. The power flow can be in either direction. Figure 4.5

shows a protective scheme for a wye-wye transformer with delta tertiary.

It is self-explanatory and straightforward except for the following

considerations.

The tertiary winding is usually of much smaller rating than the

 RE'

P o w e r S y stem s
E n gin eerin g, Inc.
P.O. Box 19398
JL JO B NO. N/A DATE Sh
Houston. Texas 77024

fe? KV
m ain windings and overcurrent relays set to protect the main windings

offer very little protection to the tertiary. If the rating of the

tertiary winding can be exceeded by the heavy fault current circulating

in it on any external ground fault, a single inverse time overcurrent

relay (51) , fed from a current transformer in one leg of the delta

should be fitted as shown. It is also noticed that two neutral current

transformers and the current transformer on delta winding provide cur­

rent polarizing sources for directional units, if used in the adjacent

protection zones.

4.5.3 Protection of Utilization Transformers

Power utilization transformers are used at the point of industrial

or other power consumption. They are normally step-down transformers

connected between the high voltage (above 34.5 KV) power source and

the distribution system or load. The power flow is normally from high

voltage to low voltage. Since motors represent a typical load which

requires special considerations in relaying design, they are the only

load used for the purpose of illustration.

Various types of transformer installations and their protective

schemes are shown in Figures 4.6 through 4.12, The characteristics of

the phase overcurrent relays (50/51) should he of the inverse or very

inverse type such that their settings can be co-ordinated with the sec­

ondary phase overcurrent relays (51), shown in dotted lines on the

drawings. Along the same reasoning, 51 relays should coordinate with

the downstream protection devices.■ When the motor is comparable in size

with transformer rating, it may be difficult to coordinate the 51 relay

with the motor protection while obtaining fast clearance tines for high
level faults and a low enough current setting to comply with the NEC,,

provision 340. Where the transformer supplies a single bus, this

problem can be resolved by using a second set of phase overcurrent re­

lays. Qme set should have a long time inverse characteristic and the

second set normally has a standard inverse or very inverse characteris­

tic. On the other hand, the motor must be protected! by a differential

scheme for sensitive fault detection. This arrangement can achieve

good sensitivity as well as good selectivity, both of which may be

unobtainable with only one set of overcurrent relays. Where the trans­

former is one of two or more feeding a multi-section bus, a partial

differential scheme with a second set of phase overcurrent relays can

provide the solution to the problem stated above.

If the motor on the low voltage bus is synchronous rather than in­

duction type, it will continue to supply current to a fault until the

motor comes to rest. For this protection, it is necessary to ensure

that 50/51 relays will not operate for a high voltage system, fault*

If the misoperation may occur, directional overcurrent relays (67)

should be used as shown in alternative 1 in Figures 4.6 through 4*12.

For the same reason, directional overcurrent relays should also* replace

50/51 relays if the low voltage bus has another infeed and fault cur­

rent can flow through the transformer in either direction.

As shown in the figures, power transformers with a delta primary

should be fitted with an instantaneous ground fault relay (50GS) sup­

plied from a "donut" type current transformer. If a donut CT is not

available, a relay with an inverse time/instantaneous characteristic

(50/51N) supplied from the residual of the phase current transformers

should be used. In case of a wye primary, the transformer contributes

current to a high voltage system ground fault. Instantaneous ground

fault relay (50GS) cannot therefore be used. A residual inverse time

instantaneous overcurrent relay (50/51N) or a directional overcurrent

relay (67N1) with another ground fault overcurrent relay (51G) should

be used. The 51G relay should be fed from a transformer neutral cur­

rent transformer and be coordinated with the high voltage system ground

fault protection.

Power utilization transformers are sometimes provided with circuit

switchers instead of circuit breakers. In the design of the protection

scheme, it is necessary to block the circuit switcher from tripping

when the fault current is above its interrupting capacity. This usually

requires that the remote source relay operates instantaneously for

faults above the circuit switcher interrupting capacity.

In designing protective schemes for parallel transformer banks, a

separate differential scheme for each transformer is recommended. This

is especially important when the transformers are of different sizes.

If considering using only one differential relay, extra phase and

ground overcurrent should be provided for the smaller transformer.

 D W G .IN 0 .

P o w e r S y stem s
E n g in eerin g , Inc.
P.O. Box 19398
Houston, Texas 77024
BY JL JO B NO. N/A DATE Sh 1

F I G U R E 4 ,6 PROTECTION ARRANGEMENT f o r
A W Y E / DELTA T R A N S F O R M E R
 d w g .n o . REV .

PROTECTION A R R A N G E M E N T FOR P o w e r S y ste m s

A D E L T A /DELTA T R A N S F O R M E R E n gin eerin g , Inc.
Houston, Texas 77024
Y JL JO B NO. N/A DATE Sh 1 of 1

i
v._ _ _ m
NOTE

1. Relays marked -> are alternative. Relay 50GS is used with a donut CT ■
Relay 50N/51N is used with phase CT's.
 P o w e r S y stem s
E n g in e e rin g , Inc.
P O. Box 19398
Houston. Texas 77024
JL JO B NO. N/A DATE

■ ___ ______________ ___

F IG U R E 4 ..9 p r o t ec t io n a r r a n g e m e n t for a w y e / d e l t a
TRANSFORMER WITH A GROUNDING TRANS.
 d w g .n o . REV.

P o w e r S y ste m s
.. /'.SI E n gin eerin g , Inc.
P.O. Box 19398
JL JO B NO. N/A date Sh
Houston, Texas 77024

1. Relays m a r k e d * are a l t e r n a t i v e . R e l a y 50GS is u s e d w i t h a d o n u t C T , R e l a y

5 ON/5IN is used w i t h p h a s e C T 1s .

F I G U R E 4 , 1 0 PROTECTION ARRANGEMENT fora d e l t a / delta

TRANSFORMER WITH A GROUNDING TRANSFORMER
 P o w e r S y stem s
E n g in e e rin g , Inc.
P.O. Box 19398
JL JO B NO. N/A DATE
Houston. Texas 77024

( 1)

PREFERRED a l t e r n a t iv e
CT CT. f

CIRCUIT A

MOTOR
LOAD

F IG U R E 4.11 PROTECTION ARRANGEMENT FORA D ELTA /W YE

tra n sfo rm er w it h a c ir c u it s w it c h e r
 CHAPTER 5

MISCELLANEOUS ITEMS OF POWER SYSTEM SUPERVISION,

PROTECTION AND CONTROL

One deciding factor in modern electric power system operation is

service continuity. Service interruption means not only loss of rev­

enue but, more importantly, possible social disasters. Various ap­

proaches for improving service continuity such as more lines, more

generators, power pooling, etc., are widely used. Among these, a proper

relaying system is the most important one. A well designed relaying

scheme should trip all necessary circuits, not more or not less. Con­

sequently, selectivity and sensitivity are two major considerations in

the design of protective relaying. These two requirements are usually

not hard to meet in primary protection through proper selection of dif­

ferent types of relays. However, this is not necessarily true for

backup protection. For example, remote backup protection for a trans­

mission line m a y trip circuit breakers on both ends and interrupt all

loads tapped along the line, which may not be necessary. Breaker

failure relaying is a local backup protection designed to solve this

problem.

Applications of electronic computers for electric power system

protection, operation control and supervision are not new at all.

Electromagnetic protective relays, widely used since sixty years ago,

are nothing more than analog computers designated for individual speci­

fic function. As an example, an impedance relay is an analog device

to calculate the driving point impedance and to initiate other devices

such as circuit breakers, if necessary. Since digital systems have

definite advantages over analog systems, it is a natural trend to

replace conventional devices by their digital equivalents wherever

possible. In this chapter, microprocessor applications on SCADA

(Supervision _Control And .Data Acquisition) systems will be discussed

in detail.

5 .1 Transmission Substation Configurations

Generally speaking, transmission switchyards can make use of the

following four arrangements: the single bus, the double bus, the ring

bus and the breaker-and-a-half arrangements.

The single straight bus configuration is the simplest arrangement

and is usually applied where circuit outage for equipment maintenance

is acceptable. In the event that the loss of all circuits due to any

single bus fault or breaker failure is intolerable, bus sectionalization

should be provided.

The double bus configuration can provide a higher degree of reli­

ability than the single bus. It permits equipment maintenance without

circuit outages. Its design consists, essentially, of the compromises

between cost of transferring circuit protection and cost of service

interruption. As with the single bus arangement, sectionalization

may be necessary for the same consideration.

The ring bus and the breaker and a-half configuration provide both

the m a x i m u m degree of reliability and the ability to maintain equipment

without more than two circuit outages. The use of the ring bus arrange­

ment is limited to no more than four circuits. For applications of

more than four circuits, the breaker-and-a-half scheme should be used.

5.2 Breaker Failure Protection [16]

In the past, it was standard practice to backup transmission, line

protection simply by overreaching primary protection to line sections

beyond the remote bus. Selectivity of this remote backup depends main­

ly upon the time settings of the relays involved. Frequently, unneces­

sary tripping and service interrupting occurred. Remote backup* in

addition to not being selective, may also not be sensitive enough

because of the relatively small proportion of the total fault current

flowing from the remote bus. This is true especially when very long

transmission lines are involved.

Wi t h the advent of extra high voltage systems and increasing con­

cern about service continuity and possible breaker failures,, a trend

to use breaker failure protection is developing. Unlike remote backup

protection, this local backup is applied at a local station. If the pri­

mary relays fail, local secondary relays will trip the local breakers.

In case the local breaker fails, either the primary or the secondary

relays will initiate the breaker failure protection to trip all local

breakers adjacent to the failed breaker. It should be noticed that

although breaker failure protection has advantages and is widely used

now, remote backup protection can not be completely eliminated.

Breaker failure protection is usually applied to equipment such

as generators, transformers, and substation buses. In this section,

basic schemes and operation theory of breaker failure protection will

be discussed. Typical schemes for' various equipment will be given in

the subsequent section.

Figure 5.1 illustrates a basic simplified breaker failure scheme.

 DWG. N 0. REV,,
SIMPLIFIED BREAKER FAILURE P o w e r S y ste m s
PROTECTION SCHEMES
Q
>E n g in e e rin g , Inc.
P.O. Bo* 19398
Houston. 7 7 7 0 7 4
BY JL J O B NO. N/A DATE Sh 1 of 1

PR!, RELAY
SYSTEM
r --- 1

TO TRIP ALL ADJACENT CB.

125 VDC c+; 5V l fr)

I25YDC

i SEC. RELAY

NOTE'S: 1) 94 F u n c t i o n is Auxiliary Tripping Relay

W h i c h m a y n o t be Required Wi t h Dual
Tr i p Coils.
21 M u l t i - C o n t a c t s to Trip Local Circuit
B r e a k e r s a r o u n d Failed Circuit and
In i t i a t e Remote Trip as Required to 125VDC ( - )
Is olate F a u l t , Block Reclosing, etc.

Figure 5.1 S I MPLIFIED BR E A K E R FAILURE P R OTECTION SCHEMES

Operation of either the primary or the secondary relays will trip the

circuit breaker and start the timer 62 through auxiliary relays 62X

and 62Y. If the circuit did not open before 62 times out, this indi­

cates a failure to clear the fault. Then failure detector (device No.

50BF shown in the figure) will energize a hand-reset multi-contact

auxiliary relay 86BF, which in turn will trip all the adjacent breakers.

Usually, the 86BF relay is also arranged to block reclosing, to initiate

transfer trip, and to stop carrier with blocking-type pilot relaying.

Normally, timer 62 should be supervised by instantaneous overcur­

rent relay 50BF rather than the breaker "a" contact. This is because

the "a" contact may be open while the main contacts in a damaged breaker

are actually closed. Three overcurrent units used can be set below

mini m u m fault current for better sensitivity. Where transformer faults

do not provide sufficient-current to operate 50BF relays, a transformer

differential relay contact can be used instead, supervised by a breaker

"a" contact.

It is recommended practice that, (i) one breaker-failure circuit

per breaker be applied, regardless of substation bus configurations,

(ii) all adjacent breakers be tripped, regardless of fault location,

and (iii) one timer per breaker be used, rather than one timer per bus.

This practice will provide maximum flexibility in design and, more

important, maximum selectivity.

5.3 Typical Schemes for Breaker Failure Protection

As stated in the last section, current flow should cease shortly

after the trip circuit is energized in a properly functioning breaker.

The time interval between trip energization and current flow cessation
is the breaker interrupting time. If interruption does not occur in

about this much time, the breaker is assumed to have failed and the

breaker failure relaying should initiate tripping of adjacent and/or

remote breakers to isolate the protected breaker. Today, breaker fail­

ure protection is mostly used in high voltage substation bus protection.

In this section, breaker failure protection for various bus configura­

tions will be discussed [16].

Figure 5.2 shows the simplest breaker failure scheme for the single

bus configuration using one timer per bus. In transmission line protec­

tion, it is general practice to connect primary and secondary line re­

lays to separate current transformers and separately fused direct cur­

rent supplies as shown in the figure. Two zener diodes TRB-2 are used

to prevent false indication of supervision lights, R and W. The zener

diode inside each overcurrent detector isolates the multiple trip cir­

cuits of all overcurrent detectors on the same bus so that only the

desired target will drop.

Figure 5.3 shows breaker failure relaying for the same bus arrange­

ment using one timer per breaker. Although using a separate timer for

each breaker is more costly than a timer for each bus, the following

advantages can pay for the price:

(1) If the breakers on the bus have different interrupting times,

the common timer must be set to accommodate the slowest breaker. Sep-

a rate timers provide faster backup clearing for faster breakers.

(2) If a fault begins on one line and subsequently spreads to

another, the common timer might misoperate and strip the bus. As the

number of lines increase, it might become impossible to set the timer

properly.
 P o w e r S y ste m s
Jkjy E n g in eerin g , Inc.
P.O. Bo* 19398
Houston, Texas 77024
JL JOB NO. DATE N/A

to
D
CQ

1
W
cc
1
—® n
£w■i~s
i2 m
2 5I o2 £o
2, i e>
*X ~ c
•
z
^ C t
o5
£■ u
> ©
to ^ —
S.- 4* 0

D
J
a
fa
Pi
h

53
<<
ZC
3 0>
m
■=r£y O- ? E
S £

'91J* 3 ® »
5 “<-r=5-£ s
CQ
oat?£ ~ " -U. iy. ia <
,
*” 3 3 I I
~ “ V J jiS a
: ; H J u
n O O v = £S S £ ffi ,i
,H
W

CM
-in
a
u
a
* o- ;^ x > Qn,
Q Z • | fcO$£>i£> '<£>
fa
 P o w e r S y ste m s
E n g in e e rin g , Inc.
P.O. Bo* 19398
Houston, 77024
JL JO B NO. N / A DATE
 (3) With separate timers, the bus can be reconfigured without

rewiring the breaker failure circuits.

The telephone relay TX as shown in the figure should be used when

breaker failure initiate auxiliary relays 62X and 62 Y are energized

by potential polarized relays. This is because of the possible col­

lapse of polarizing potential during bus f a u l t s .

Figures 5.4 and 5.5 show breaker failure relaying for the breaker-

and-a-half bus and the ring bus, respectively. The basic functioning

of these schemes is the same as for the single bus, and requires no

further explanation. To conclude this section, an overall protection

scheme for a generator pair is shown in Figures 5.6 and 5.7. The

scheme includes both generator and bus breaker failure relaying.

5.4 Digital Computer Applications on System Supervision and Control

5.4.1 SCADA Systems ■

SCAD A systems for power system applications are designed to provide

remote control and monitoring of substation and power plant equipment

f rom the system control center. The equipment should be controlled

and monitored by a local remote terminal unit (known as RTXJ) which in

turn is under the direct control of the S'CADA central processor* Cowr

municationS' link between RTU’

s and the SCADA Master station can be

▼O'ice grade,, telephone,, Microwave,, or power line carrier circuit, de­

pending u p o n particular application requirements'.

Information flow between the master station and BLIP1 s is M - d i r e c -

tiociai. Brills transmit status, analog, and digital inf o r a t i o n to the

station; the master station, on the other hand, issues control cooniaads

to RTUs. Status information includes dry contact open/closed messages

 P o w e r S y ste m s
E n g in e e rin g , Inc.
P .0. Box 19398
Houston. Texas 77024
JL JO B NO. N/A DATE

it « 3 C
Q ED Q j
^ ^^
-S -c 5 o r ij
^ i <V ® rc *J &S cvn
a- 5 fxJ5 -2=1
h < -£S2
a S
3| ut SC <
-©
"Z c
w5 3“fI Sx
O a * <£
o > I « •- U.- u.' < u.' * ©» O
P ' NJ r
IS) O a. co cd cd cb ,- i ^ 5
*- » ?
1 >8

iUULJLJUULJ
;S 03 O '•® COCD ® CD COQ
i t£ o
r 5 ,v *
► lo m c
££«?
„ a
X S oo
fU"*><VCN tD
* l£>< s n a h
C C N -V
oo od co a

o
2
H
<
J

8 £
D 2
*1 <
s s
CQ
& OS
W W
\4 Oh

CQ §
H
Cn H
w
s
o
o
2
M
w
D

CQ

in
a
3
Cn
•H
tn
 DWG .NO. REV.
BREAKER F A I L U R E R E L A Y I N G FOR A RING BUS P o w e r S y ste m s
' E n g in e e rin g , Inc.
P.O. Box 19398
Houston. Texas 77024
BY JL JO B NO. N/A DATE Sh of 1

co
D
«
o
S
a
o
Pn
u
2H
s

in
in
0)
u
p
CP
•H
Cn
Thij drawing <s the proo«ny of P o w e r S y lt e m * Engineering, tnc. Neither this d e w in g ncr reor<xiiJClion<,
of i?, nor information derived from it shad tx* gtv*n to others without the t?*or**ssed written .. im sm -w jt
S y ite m a . No us® is to De made of m which >s. or mav De injurious Pow m S y t t e m i
 D W G. NO. REV.

SUBSTATION P R O T E C T I O N SCHEME P o w e r S y ste m s

E n g in e e rin g , Inc.
P.O . Box 19398
Houston. Texas 77024
V JL JO B N O . N/A D ATE Sh ^ of 2

Figure 5.7 S U BSTATION PROTECTION SCHEME

and is used for such functions as alarms, system topology update, etc.

Analog information includes variable voltage and current levels and is

used for transmitting such quantities as line voltages, currents, watts,

vars, frequency, turbine and/or generator temperature, wind speed and

transformer tap position. Digital words represent information such as

total number of contact transitions since last polling, etc. Control

information is usually of a digital nature and contains the address of

the equipment under control and control signal itself. It usually ac­

complishes a relay closure which in turn controls an appropriate item

of substation or power plant equipment such as breaker close/open,

transformer tap raise/lower and generator governor raise/lower.

In addition to data acquisition and control discussed above, a

m o d e r n SCADA system should also have the following functions:

(1) Data Base:

The effectiveness value of a SCADA system is primarily a function

of the data processing efficiency of the digital computer subsystem as

measured by data thruput. To enhance the data thruput, the SCADA sys­

tem requires an efficient and effective data base subsystem. Therefore,

from the data thruput point of view, the data base is the heart of the

SCADA system.

(2) Operating System (also known as the Executive):

The operating system is required to provide a real-time priority

oriented processing environment and to allocate computer resources

accordingly.

(3) Man-Machine Interface:

This sybsystem is designed to provide capabilities such as CRT dis­

plays, alarm generation, data entry, data logging and picture compiler.
 (A) Performance Monitoring:

This subsystem functions to enhance the operational availability

of the SCADA or central control system. Generally, every primary hard­

ware subsystem should have an identical unit for backup.

Each of these functions is a highly specialized topic and the design of

a complete SCADA system may require huge engineering effort. Conse­

quently, no further discussion on SCADA systems will be given. In the

subsequent paragraphs, emphasis will be placed exclusively on RTU de­

sign. An example is given to illustrate the organization and operation

of a RTU system.

A remote terminal unit (RTU) is a part of a process controller

and/or SCADA system in which automatic functions are implemented within

the R T U s . An intelligent RTU is one in which much of the logic asso­

ciated with each function- has been programmed and is executed locally

in a digital processor (usually microprocessor) instead of transmitting

data to and processing in the central processor of the SCADA system.

The first point of consideration in RTU design is the content of

data required. Usually it is the responsibility of the operations and

application groups to determine what data is needed for power system

operation, control and for data input to SCADA programs. They should

also define the required data characteristics of accuracy, resolution,

sampling period, range, format, etc. Generally speaking, an intelligent

RTU should have the following basic functions and capabilities.

(1) Data Acquisition:

It includes (a) discrete inputs such as status monitoring, pulse

counting and pulse timing; (b) digital inputs such as byte codes,

binary coded decimal and others; (c) analog inputs to maintain a fresh

internal data base.

(2) Control Outputs:

Included are (a) supervisory and manually initiated control such

as check back-before-operate interlocks, etc.; and/or (b) fully automa­

tic and software or firmware initiated control.

(3) Communications, including communications with SCADA master

station and service communications interface to prevent the loss of

data.

(4) Operational Security:

It includes (a) minimization of false commands and data due to

communication errors, component failure, electromagnetic interface,

etc.; (b) internal monitoring of analog reference inputs, power failure

and restart, etc.

(5) Functional and Field Expandability:

Expandability includes processor time, memory, input and output

resources, wiring and space spare for the future addition of new points

or functions. As a rule, future expansion of points or functions should

not require any modifications to the RTU controlling logic circuits.

In addition to these basic functions and capabilities, optional

functions such as data concentration, sequence of events reporting,

digital metering, digital relaying, control sequences and many others

can also be implemented in an intelligent RTU as necessary.

After these functional requirements are defined, the operating en­

vironment for RTUs should be designed accordingly. The environment for

RTU operation must provide not only the data and instructions associated

with a particular function but, more importantly, must provide these

inputs in a predefined priority order. The controlling logic in an RTU

processor must decide how, when and to whom the processing resources

are to be allocated according to a predefined priority order. Methods

generally used by the controlling logic to manage processing resources

include event interrupts (a hardware contact is closed), time interrupt,

timer interrupts (a predefined time period has elapsed) and flag (a

predefined condition is recognized).

The next step is to determine whether hardware or firmware should

be used for a particular function defined above. As a rule of thumb,

those general purpose functions making up the microprocessor should be

implemented in firmware; those special functions such as operational

security in hardware. However, the trade-offs between hardware and

firmware are actually an art and should be carefully studied for any

particular application.

Possibly the most important design consideration related to RTU

systems is how to manage the resources of the microprocessor such that

data is not lost and is processed correctly. A communications interface

is the key point in the consideration. As the number of points in­

creases, the microprocessor executive overhead may be overloaded by

huge communication requirements so incurred. Generally used communica­

tions interface techniques include polling and interrupt approaches.

Implementing a communications interface by interrupt technique requires

that the microprocessor be able to handle external hardware interrupts

in addition to the real time clock interrupts. The advantage of the

interrupting interface over the polling approach is a reduction of

microprocessor executive overhead. In a well designed system, communi­

cations functions should not use more than 30% of microprocessor re­

source time.

As in the SCADA system, it is desirable that every primary hardware

subsystem or module have an identical unit for backup. It is also pre­

ferable to have automatic switchover (known as failover), and recovery

and restart characteristics to achieve high reliability and availability

as usually required.

Figure 5.8 shows a functional block diagram of a RTU system. This

system has capabilities ranging from a few status or analog points up

to m a n y hundreds of point types including SOE, set points, raise-lower,

etc. It has a dual-redundant backup. It is divided into the following

modules: common control, control and indication, accumulator, analog

to digital, raise-lower and power supply module.

The common control module performs message buffering between the

various point modules and communication subsystems. The central element

is a microprocessor with communication to and from all other modules.

The interrogation of the point modules is under direct control of the

microprocessor and the remote scan program stored in ROM. Figure 5.9

shows internal organization of the common control module. The watchdog

timer is designed to provide a programmable timing signal which is fed

into an interrupt generation logic circuit. If the timer is timed out,

a signal is sent to initiate an interrupt to the CPU.

The control and indication module continuously scans its indication

points and maintains a queue of all changes. On request from the SCADA
 P o w e r S y ste m s
FUNCTIONAL B L O C K diagram - RTU SYSTEM
E n g in e e rin g , Inc.
P O. Box 19398
JL JO B NO. DATE Sh
Sh 1 of, 1 Houston. Texas 77024

s
13
W
w
D
Eh
OS

U
<
H
Q
b
o
CQ
a

EH

00
in
Q)
u
P
Cn
A •H
 DWG. NO. REV
P o w e r S y ste m s
C O M M O N C O N T R O L MODULE
s ^M E n g in e e rin g , Inc.
P.O. Bo* 19398
v JL JO B NO. N /A DATE Sh ^ of 1 Houston. Texas 77024

a
o
s
9os
H
£O
o
2O

C\
LO
©
s
Cn
•H
£n
master station, the module can report the number of state changes

accumulated, point address, and the present state of each point.

Remote control of a point requires the reception of two sets of point

address. The module compares these two sets of point address; if they

agree, the control operation will take place.

The analog to digital module digitizes each analog input in se­

quence and produce a serial bit pattern for serial transmission* The

module in cooperation with a precision direct voltage source anid resis­

tive tap change potentiometer can also provide remote indication of

transformer tap position in the SCADA master station.

The raise-lower module provides remote control of transformer tmp

from the SCADA master station. As in the control module, two sets of

point address are required 'before the operation occurs*

 CHAPTER 6

BUSINESS ASPECTS OF ENGINEERING PROJECTS

During my internship, I was under the direct guidance of project

managers. Consequently, I also gained broad experience in engineering

project management. My activities in this area included market analy­

sis , feasibility study, vendors 1 bid evaluation, and project supervi­

sion and control. This chapter will highlight some of my experience.

It is divided into two parts; part one discusses life cycle of engineer­

ing projects, part two concentrates on legal aspects of engineering.

6 .1 The Life Cycle of Engineering Projects

Generally speaking, the life cycle of an engineering project b e ­

gins with identification of needs (or market a n a l y s i s ) , followed by

feasibility study, preliminary design, detailed design, purchasing*

formal design review, and ended with hand over and comp let Ion,. Inter­

weaving through these phases of project life is project supervision and

control. In this section all these different stages will be briefly

discussed.

The purpose of market analysis is to view the project over its

life cycle on the basis of economic considerations. A market can exist

due to either current deficiencies or anticipated future deflci.ei.ncie?s.

The data collected at this stage is usually quite rough and a decision

is usually made on the considerations of the market potential and s i z e ,

competition, possible obsolescence, and so on.

The purpose of a feasibility study is to extend the market analy­

sis with the intent of arriving at a proposed system configuration in

response to an identified need. In general, feasibility studies in­

clude the following aspects:

(1) Technical study: choice of technologies, design require­

ments, manpower requirements, etc.

(2) Economic and financial study: cost-benefit analysis, cash

flow study, adequacy of funds, supply-demand analysis, and marketing

program.

(3) Social and political study: community resistance, legal con­

straints, environmental suitability, and so on.

The results of the feasibility study are generally presented in the

form of a proposal to management or clients covering the aspects stated

above.

Preliminary design constitutes the starting point of a design pro­

ject and directs toward defining a set of useful solutions or system

configuration to the problem being addressed. It starts with the tech-

.nical baseline for the system defined in the feasibility study and pro­

ceeds to translate the established system level requirements into de­

tailed qualitative and quantitative design requirements. Usually it

includes functional analysis and allocation, trade-offs and optimiza­

tion, and synthesis. In this stage, the following specifications

should be prepared by design engineers: system specification, develop­

ment specification, and procurement specifications.

The detailed design phase begins with the concept and configura­

tion derived through preliminary system design, and. includes system

component design and interface, preparation of design documentation*

In this stage specifications of process, material, and construction

should be p r e p a r e d .
 Purchasing is one of the key engineering functions. It consists

of purchase requisitions, sources of supply, request for quotation,

source selection and quotation evaluation, and placement of purchase

order.

The functions of formal design review are:

(a) It provides a formalized check of the project. Major problem

areas are discussesd and resolved.

(b) It provides a common baseline for all project personnel*

(c) It provides a means for solving interface problems and pro­

motes the assurance that all system elements will be compatible.

The completion and hand over phase of a project prepares the pro­

ject for phasing out and hand over to another fora of administration*

Project completion consists of scaling down and dismantling the pro­

ject organization. It also involves the transfer of project personnel,

assets or other facilities.

Project supervision and control is the backbone to an engineering

project. Generally speaking, it consists of accomplishments evaluation,

expenditure control, and investment control. The tools generally used

are simple bar c h a r t , milestone c h a r t , program evaluation and review

technique (PERT), critical path method (CPM), PERT/COST, Gantt chart,

and line of balance (LOB).

6*2 Legal Aspects of Engineering

Never before in its history has the engineering profession faced

the legal exposure that it does today. The rapid expansion of technol­

ogy has stimulated equally rapid developments in the law. Merely keep­

ing abreast of developments in the laws that have a direct effect on

an engineer would be a full-time task. Lack of time will prevent engi­

neers from developing familiarity with all the laws that may affect

the engineering profession. However, ignorance of the law is no excuse

is still a valid maxim.

The engineer must develop a sixth sense that will alert him to

possible legal implications and permit him to seek competent advice

before undertaking any project. This sixth sense can be developed by

acquiring an awareness of major legal concepts that affect the engineer­

ing profession. Of current significance to all engineers is the concept

of product liability. "Product liability" is a collective term describ­

ing three legal theories. These theories, which form the basis for re­

covery against the manufacturer and seller of a defective product, are:

(1) breaches of express and implied warranty; ( 2 ) negligence; and (3)

strict or absolute liability. Because of his relationship to the de­

sign and manufacture of products, there is a strong likelihood that the

engineer will become involved in a product claim or litigation as an

"expert." The engineer called upon to render technical assistance as

an expert should ensure that he has the requisite educational back­

ground and practical knowledge regarding the subject of the contro­

versy. The engineer can also prove instrumental in avoiding product

liability exposure by instigating and pursuing preventive management

techniques.
 CHAPTER 7

SOME MANAGERIAL CONSIDERATIONS FOR YOUNG ENGINEERS

C. R. Gow reported in his book [19] that the chief obstacles to

the success of each individual or of the group comprising a unit were

of a personnel and administrative rather than a technical nature. What

Gow said in 1932 still holds true. This is largely due to the fact that

99% of the emphasis in the training of engineers is placed upon purely

technical or formal education.

In this chapter a set of "house rules" or professional code for

young engineers will be discussed. They are results of my observations

or experience during my internship. Although they are not exhaustive,

they may be of worth to young engineers just starting their careers,

and to "older" engineers as well who know these things perfectly well

but who all to often fail to apply them in practice.

This chapter is divided into two parts; part one discusses purely

personality and character considerations for e n g i n e e r s , part two con­

centrates on what a young engineer really needs to learn at once and

consists of three sections - in relation to your work, in relation to

your boss, and in relation to your associates and clients [20 ].

Just one point should be clarified before entering into the con­

text: none of these rules is theoretical or imaginary, and however

obvious and bromidic they may appear, their repeated violation, in my

experience, is a main source of frustration and embarrassment in an

engineering career.
7.1 Character and Personality Considerations

The subject of personality and character is, of course, very broad

and much has been taught in schools from the social, ethical, and reli­

gious points of view. In this section, the following considerations are

given and discussed from a purely practical point of view based upon

principles of general engineering practice.

One of the most important personal traits of engineers is the abi­

lity to get along with all kinds of people. This is rather a compre­

hensive quality but it defines the prime requisite of personality in

any type of industrial organization. Of course, it is a mistake to try

too hard to get along with everybody merely by being agreeable and

friendly on all occassions. You must earn the respect of your asso­

ciates by demonstrating your will to stand up for your belief. This is

extremely important, especially when you know you are in the right. If

you can be pushed around easily, the chances are that you will be pushed

around. Just remember do not be too a f f a b l e . There will be times when

you do well to start a fight yourself, when your objectives are worth

fighting for.

Self-respect is another important character for engineers. One of

the most striking phenomena of an engineering office is the transparency

of character among the members of any group who have been associated for

any length of time. In fact, it frequently happens that a man is much

better known and understood by his associates, collectively, than he

knows and understands himself. It is also usually true that people

respect only those who respect themselves. Consequently, regard your

personal integrity as one of your most important assets and maintain

the highest standard of ethics of which you are capable.

It is also important to note that engineering is essentially a

gentleman's profession, and a little profanity may go a long way.

T h e r e f o r e , watch your p's and q's and be careful of your personal

appearance.

Dependability is another important personal trait for any success­

ful engineer. In fact, a reputation of dependability and reliability

can be one of your most valuable assets.

7.2 In Relation to Your Work

Rule 1 : No matter how menial and trivial your early assignments

m a y appear, give them your best efforts.

M any young engineers feel that the minor chores of a technical

project are beneath their dignity and unworthy of their professional

training. They are anxious to prove their extraordinary capability in

some major enterprise. Actually, the spirit and effectiveness with

which you tackle your first humble tasks will usually be carefully

watched and m a y affect your entire career. Remember, there is always

a premium upon the ability to get things d o n e . Do your present job

well and the future will take care of i t self.

Rule 2 : In carrying out a project, do not wait for others to hand

you results; go after your results and keep everlastingly after them.

This is one of the first things a young engineer has to learn in

entering engineering business. It is not sufficient to place the order

and sit back waiting for the result. In fact, most jobs move in direct

proportion to the amount of follow-up and expediting applied. Look

immediately for some way around each obstacle encountered and keep the

job rolling without losing momentum.

 Rule 3: When sent out on any field assignment, stick with it and

see it through to a successful finish.

Rule 4: Before asking for approval of any major action, have a

definite plan and program worked out to support it.

It always pays for the time to prepare a well considered proposal

rather than a half-baked one. OR you will not get credit for your

great i d e a s .

Rule 5 : Strive for conciseness and clarity in oral or written

reports.

There is a wisespread tendency among engineers to surround the

answer to a simple question with so many preliminaries and commentaries

that the answer itself can hardly be understood. It is so difficult to

get a clear answer out of the answer that its usefulness is thereby

greatly diminished. Therefore, try to convey the maximum of significant

information in the minimum time and by the minimum w o r d s .

Rule 6 : Don’
t be timid, speak up and express yourself and promote

your i d e a s .

M any young engineers tend to think that their job is simply to do

what they are told to do. Of course, there are times when it is very

wise and prudent to keep your mouth shut, but, as a rule, it pays to

express your point of view whenever you can contribute something.

Rule 7 : Be extremely careful of the accuracy of your statements.

This sounds almost as trite as one plus one equals two, but many

engineers lose the confidence of their superiors and associates by

habitually guessing when they do not know the answer to a direct ques­

tion. Remember next time, if you do not know, say so, but always say,

"I will find out as soon as possible."

7.3 In Relation to Your Boss

To get along with your boss, the basic principle is that he must

know what is going on in his territory. From this idea are derived

the following rules.

Rule 1 : Do not overlook the fact that you are working for your

boss.

It is not uncommon for young engineers, in their impatient zeal

to get things done, to ignore the boss, or attempt to go around him.

Sometimes they move a little bit faster that way for a while, but even­

tually they find that such tatics do not work at all.

Rule 2 : Whatever the boss wants done takes top priority.

Unless you obtain permission it is unwise to put any other project

ahead of a specific assignment from your own boss. As a rule, always

take for granted that your boss has good reasons for wanting his job

done now. Also, if you are instructed to do something and you subse­

quently decide it is not worth doing because of new data or events,

don’
t forget to inform your boss of your intentions and reasons.

Rule 3 : Try your best to inform your boss of all significant

developments.

Since your boss is constantly called upon to account for, defend,

and explain your activities to the higher-ups as well as to coordinate

these activities with other departments, he needs to know all informa­

tion promptly and accurately.

Rule 4 : Do not be too anxious to follow the boss’

s leads.

In general, the program laid down by the department head is tenta­

tive, and is intended to serve only until a better program is proposed

and approved. Therefore, do not just follow the lines laid down by

your boss. Try to sell your ideas to him and make contributions.

Rule 5 : Be as particular as possible in the selection of your

boss.

It is well known that the influence of the senior engineer, or

even the boss, is a major factor in molding professional character of

younger engineers. Consequently, if your boss turns out to be somewhat

less than half the man he ought to be, transfer to some other company

at the first opportunity. Remember, a person who constantly plays golf

with a dub becomes a d u b .

7.4 In Relation to Your Associates and Clients

Rule 1 : Never invade the territory of any other department without

the knowledge and consent of the executive in charge.

Rule 2 : In all transactions be careful to deal-in everyone who

has a right to be in.

Be especially careful about whom you mark for copies of letters,

memos, etc., when the interests of other departments are involved.

Rule 3 : When you are dissatisfied with the services of another

department, make your complaint to the individual most directly respon­

sible for the function involved.

Complaints made to a m a n ’
s superior should be resorted to only

when direct appeal fails. Give the man a fair chance to correct the

grievance before filing a complaint with his s u p e r i o r .

Rule 4 : In dealing with customers and outsiders, remember that

you represent the company, ostensibly with full responsibility and

authority.
 CHAPTER 8

SUMMARY AND CONCLUSIONS

During a one year internship, I made the following contributions

to the firm:

(1) Technical Activities - I designed a protective relaying scheme

for the 480 MW power plant, microprocessor-based SCADA and SOE systems.

I also designed the power plant overhead and underground distribution

circuits as well as substations.

(2) Engineering Services - Designing a complete engineering pro­

gress control chart and helping the Project Manager to prepare cost

control charts for a project which cost over $1.4 million merely for

engineering.

(3) Standardization of the managerial procedures for young engi­

neers in the firm. This report summarizes and highlights the results

of this standardization process. It is prepared to be a training manual

for young engineers just starting their engineering careers.

This report is divided into two parts. Part one emphasizes the

following technical design guidelines:

(1) Fundamentals of Electrical Power System Design

Technical design tools such as short circuit current calculation,

ampacity of electrical conductors, instrument transformers, and the

R-X diagram.

The short circuit current calculation includes both maximum and

m i n i m u m current calculations under fault conditions. The maximum fault

current values are used to select adequate interrupting ratings and to

check the ability of power system components to withstand both mechani­

cal and thermal stresses. The minimum values are needed to check the

sensitivity requirements of protective relays. For overall coordina­

tion of system protective devices, both the maximum and the minimum

values are often required.

Under normal operating conditions, the prime considerations in the

applications of electrical conductors are ampacity, maximum allowable

voltage drop, and energy losses.

Under fault conditions, mechanical considerations such as minimum

required mechanical strength and maximum allowable deflection usually

govern the selection and design of electrical installations.

Instrument transformers are used to protect personnel and appara­

tus from high voltage and to allow reasonable insulation levels in

relays, meters, and other- instruments. Their performance is critical

in protective relaying. Consequently, performance estimation of in­

strument transformers is elaborated in the report.

The R-X diagram is another important tool for protective relaying

design. It integrates both system operation conditions and relay char­

acteristics in the same diagram. Its applications are also illustrated

in Chapter 2.

(2) Generator Protection

Generator fault protection functions to detect any faults either

internally in stator windings or externally in terminal leads feet srftMin

the generator relaying zone. It includes differential and stator ground

fault protection. On the other hand, generator abnormal protection is

designed to protect the following abnormal operation conditions: loss

of excitation, motoring, negative sequence current, uncleared system

fault, overloading, overvoltage, overspeed, grounded rotor and loss of

synchronism.

(3) Power Transformer Protection

Power transformer relaying includes fault detecting devices and

abnormal operation sensing devices. Fault detecting devices function

to protect the system and limit damage in the event of a fault. Abnor­

mal operation relaying is designed to protect the transformer and in­

cludes overtemperature and system backup protection.

(4) Power System Supervision, Protection and Control

Miscellaneous items such as transmission substation configurations,

breaker failure protection and SCADA system design are discussed.

Part two gives business aspects of engineering projects and in-

house guidelines to young -engineers. These guidelines are given to help

young engineers who are starting their professional careers to adjust

themselves to the industrial environment. The guidelines fit into the

following four topics: character and personality consideration, in

relation to your work, in relation to your boss,, and in relation to

your associates and clients.

From a personal point of view, my internship has been one of the

best possible and the three internship objectives have been met. The

positive attitude of m y internship supervisor, Mr* E. E* H 0*gw©'0«dl,, and

all of m y colleagues was a significant factor in the success of w y

internship.
 REFERENCES

IEEE Recommended Practice for Electrical Power Distribution

for Industrial Plants
IEEE Std. 141-1976, pp.31

[2] PSE Electrical System Evaluation and Short Circuit Study

Report No. ET-4065-02, September, 1980

[3] D. G. Fink e t . a l .
Standard Handbook for Electrical Engineers
10th edition, McGraw-Hill Co., 1969

[4] IEEE Recommended Practice for Protection and Coordination of

Industrial and Commercial Power Systems
IEEE Std. 242-1975

[5] ANSI Test Procedures for Low Voltage AC Power Circuit Breakers
ANSI Std. C37.50, 1973

[6 ] W. C. H e u n i n g , Jr.
"Interpretation of New American National Standards for
Power Circuit Breaker Applications"
IEEE Trans. Industry & General Applications,
Vol. IGA No. 5, 1969

[7] H. R. Everitt
"Formulas for Conductor Size According to Cost of
Resistance Loss”

[8 ] NFPA National Electrical C o d e , 1981

[9] 0. R. Shurig and M. F. Sayre

"Mechanical Stresses in Busbar Supports During Short
Circuits" pp. 217-237, AIEE 44, 1925

[10] IEEE Recommended Practice for Electric Power Distribution for

Industrial Plants
IEEE Std. 141-1976, Chapter 5

[11] NFPA National Electric Safety C o d e , 1980

[ 12 ] ANSI Requirements for Instrument Transformers

ANSI Std. C57.13, 1978

[13] IEEE Application Guide for Grounding of Synchronous Generator

S yst e m s , IEEE Std. 143-1954

[14] Westinghouse
Electrical Transmission & Distribution Reference Book
Westinghouse, 1964
[15] General Electric
"Loss of Excitation Relay" GEK-27887C, 1977

[16] Westinghouse
Applied Protective Relaying, Westinghouse, 1977

[17] IEEE IEEE Guide for Protective Relay Applications to

Power Transformers, IEEE Std. 273-1976

[18] Westinghouse
"Type KC-4 Overcurrent Relay",
Westinghouse I.L. 41-776.1J, 1978

[19] C. R. Gow
Elements of Human Engineering
Macmillan Co., New York, NY, 1932

[20] R. E. Shannon
Engineering Management
John Wiley & Sons, Co., New York, NY, 1980
 APPENDIX I

Figure 7.1 shows a simplified circuit for fault current calcula­

tion. Since a fault may occur at any instant defined by 9 shown in

the figure, when 0 is the angle between the instant of positive volt­

age crest and the instant of closing switch. Expressed in mathematical

form, the fault current if for a single line to ground fault is

L if = ig(t) = E (coswt + 0)
dt

— ,. if = J L (coswt + 0)
dt L

if(t) = JL. sin (wt + 0 ) + constant

WL

Initial condition at t=o,

if (o) = o

constant - _ E
WL

if(t) = J L [sin (wt + 0) - sin 0]

WL

To find 0 values such that if(t) will have maximum value

_A_ if = JL [cos (wt + 0) - cos0]

dq WL

= 0

cos wt cos0 - sin wt sin 0 - cos0 = 0

(coswt-1) cos0 - sin wt sin0 = 0

and the peak of if occurs at wt = *180°, that is, a half-cycle after

the fault occurrence.

E F F E C T OF I N S T A N T F A U L T OCCURRENCE P o w e r S y ste m s
ON THE F A U L T C U R R E N T if E ngineering, Inc.
P.O. Box Il'KWi
Hoiuiiori, T«"xa« 77024
JL JO B NO. N/A date

Single Frequency Circuit for

Fault Current Calculation

F ig u r e A„1 EFFECT O'F IIISTABT FAULT OCCUP.r£-iCE

OH THE FAULT CUEREBPT i f
Consequently, the maximum fault current can be shown as

if (t) = J L [-coswt * 1]
WL

peak value of if(t) is 2 sfl times the rms value of its

symmetrical component (-J L coswt)*

WL

:l Ip = peak of if (t )

= 2y T is

The rms value of if (t) over the first half-cycle after fault occturrMC*

JL 1-2 coswt + (cos dut

WL

Ip :s= 2 I a * 1*63 l a

yiTi
OlnriiBwiisly,
ia = y ? ! ®

'fin prove that the fladam short circuit force «• 1

conductor, the short circuit mist occur at 0 * 4.5% eoMider #

three phsfie sjraraetrical fault#

ifk = J L [ sin (wt + 0) - sin 0 ]
WL

kfc = ^ [sin (wt + 0 - 120 °) - sin (0- 120 °)]

WL

F s = short circuit force on phase R conductor

= 5.7 [sin (wt + 0 + 120 °) - sin (wt + 0 120°)

-sin (0 + 120 °) + sin (0 - 120 °)]
D(WL) 2 (1 0 ) 7 [sin (wt + 0) - sin 0 ]

- 2 2 .8 sin (wt + 20) sin 2 _wt

2
D(WL) 2 ( 1 0 ) 7

- 45.6 sin 2 wt cos (wt + 20)

d@ ' 2
D(WL) 2 (10 ) 7

cos (wt + 28) = 0

 VITA

Candidate’
s Full Name: Ting-Zern Joe Lee

Place and Date of Birth: T a i w a n , China

February 3, 1950

Permanent Address: #7, LN 40, Pei-Ta Road

Hsinchu, Taiwan
R •0 •C •

Education: University of New Brunswick

Fredricton, N. B.
Canada
M. Sc. (Electrical Engineering, with
emphasis on power system operation,
protection and control), 1978

National Cheng-Kung University

Tainan, Taiwan
R.O.C.
M. Sc. (Electrical Engineering, with
emphasis on VLSI design), 1976

National Cheng-Kung University

Tainan, Taiwan
R.O.C.
B. Sc. Eng. (Electrical Engineering), 1972