Вы находитесь на странице: 1из 11

CHAPTER

14

Distance

protection:

optimum

dynamic

 

design

of static

relay

comparators

  design of static relay comparators BYL.JACKSON,J.B.PATRICKSON,ANDL.M. WEDEPOHL. SYNOPSIS V,, V,
  design of static relay comparators BYL.JACKSON,J.B.PATRICKSON,ANDL.M. WEDEPOHL. SYNOPSIS V,, V,

BYL.JACKSON,J.B.PATRICKSON,ANDL.M.

WEDEPOHL.

SYNOPSIS

V,,

V,

= particular

levels

of voltage

in

a level

The

ease

with which

transistor

comparators

for distance

 

detector,

corresponding

to time

relays

can be designed

for high-speed

operation

must

be

V,

intervals

= polarising

T,

and

voltage

T,,

respectively

 

tempered

with consideration

of overall

performance

and

 

integrity

of operation.

Operating

speed

must

be defined

 

T

d

= system

periodic

time

 

over

the

whole

of

the

working

range

of the

relay,

and

Z,_ =

= phase-comparator

VL/IL =

impedance

angular setting

of a section

of

neither

the

speed

nor

the measuring

accuracy

should

be

protected

 

unduly

affected

by

the

severe

transients

generated

by

Za

power

system

setting

 

modern

e.h.v.

networks.

The

comparator,

as

well

as

= relaying-system

impedance

being

proof

against

damaging

voltage

surges,

must

oper-

 

ate

accurately

in

the

presence

of

long-duration

offset

transients

accentuated

by low-loss

modern-plant

para-

meters

and

the

random

point-on-wave

inception

of

INTRODUCTION

 

faults

due

to natural

hazards

and

closure

or reclosure

of

modern

pressure-head

circuit

breakers.

The

attainment

Until

a decade

ago,

relay

design

was

dominated

by the

of high-speed

operation

under

these

practical

conditions

use of electromechanical

elements.

Such

an element,

of

precludes

the

adoption

of

many,

apparently

practical

whatever

basic characteristic;

e.g.

square-law

induction

transistor-comparator

circuits

and

favours

the adoption

 

element,

has

a dynamic

behaviour

 

special

to

that

ele-

of circuits

with

well

defined

dynamic

 

performances.

ment,

and

design

freedom

is consequently

restricted

by

Extensive

laboratory

investigation

has shown

that

the

factors

such as the conflicting

requirements

of sensitivity

block-average

cise

design

comparison in all respects,

principle

and

practical

is amenable

to pre-

and

mechanical

robustness.

The

same

period

saw

the

 

fast-operating

 

emergence

of a method

for

assessing

the

dynamic

per-

relays

can

be designed

with

good

transient-free

charac-

fomance

of different

relaying

systems

by displaying

tim-

teristics.

are presented

The

results

obtained

on for a phase

such

a practical

comparator

in the paper

relay

with

a

ing

The

contours

relative

under

practical

conditions

deficiencies

of several

types

of

of relay

switching.5

could

polarised-mho

characteristic.

It

is

shown

that

a

thus

be exposed,

and

criteria

for

dynamic

performance

minimum

inherent

operating

time

of

one

halfcycle

of

be

established.

Furthermore,

it

became

possible

to

the

power

frequency

can be defined

for this comparator

establish correlation

between

operating

time,

switched

arrangement

and

that

both

the

static

and

dynamic

measuring

accuracy

and

overall

integrity

under practical

operating

characteristics

are predictable

over

the whole

operating

conditions.

Over

the

last

ten

years,

relays

working

range.

Equivalent

performance

for

the

using

transistors

have

been

shown

to be practical

alter-

amplitude-comparator

 

counterpart

is

justified

in

an

natives

to relays

using

conventional

 

componentsi

 

and

appendix,

and

underlines

earlier

work.

 

both

phase

and

amplitude

comparators

have

been

used

Relays

using

the

block-average

comparison

principle

with

double

and

multiple

inputs.

Although

the relation-

have

been

used

successfully

in field

trials

since

1957,

and

ships

of

geometrical

duality

between

phase-

and

this principle

now forms

the basis

for various

production

 

amplitude-comparators

in

the

steady

state

are

well

designs.

Sufficient

field

experience

is now

available

to

established,7

the respective

dynamic

performances

have

justify

the theoretical

analysis

and

laboratory

test results

not

been

adequately

rationalised.

 

given

in the

paper.

The

transistor

comparator

affords

great

freedom

of

 

design

for specific

laws

of operation

and/or

characteris-

tics;

this

has

no

counterpart

in

the

electromechanical

LIST

OF PRINCIPAL

SYMBOLS

 

relay,

where

the

basic

characteristics

are

prescribed

by

the

behaviour

of

the

element

itself.

This

freedom

of

V1,

V2 =

input

signals

to

a 2-input

relay

comparator

 

4

= phase

displacement

between

Vi and

 

v2

VL = phase-to-phase

voltage

 

I,

=

line

current

Vsl,

Vs2 =

level-detector

voltage

settings

 

Si

=

p.u.

input

1 related

to setting

S2 =

p.u.

input

2 related

to setting

design

embraces

both

static

characteristics

and

dynamic

performance.

Rationalisation

beyond

elementary

reproduction

of

conventional

dynamic

performance

becomes

possible,

design

procedure

is clarified,

and

the

relative

assessment

of comparators

operating

to differ-

ent

principles

is faciliated.

 

The

transformation

of input

quantities

defining

iden-

tical

steady-state

operating

characteristics

using

phase

or

amplitude

comparators

is

shown

below

to

have

204

operating characteristics using phase or amplitude comparators is shown below to have 204
extended significance in rationalising comparator per- formance, especially in the case of

extended

significance

in rationalising

comparator

per-

formance,

especially

in the

case

of circuits

using

transis-

tors.

DESIGN

PRINCIPLES

OF

TRANSISTOR

DIS-

TANCE

RELAYS

Useful characteristics

obtained

using transistor

comparators

There

concerning

ill-considered

mance

parators

teristics,

is little

doubt

that

some

of the

past

relays

have

the

element

uncertainties

resulted

from

perfor-

com-

charac-

compared

electromechanical

attempts

of

phase

with

e.g.

to compare

inherent

amplitude

output

comparators

and

different

fundamentally

a

linear

moving-coil

Basis for design

A transistor

of

close

hand,

square-law

relay

may

be designed

These

to

as

different

a

characteristics.

comparator6

approximation

such

to have

include,

a wide

on

the

range

one

an

the

electromagnetic

induction

cup

or

with

a square-law

induction-cup

element.

With

transis-

 

tor

relays,

similar

misconceptions

can

arise,

and

it

is

important

to recognise

that

the

number

of basically

dif-

ferent

comparator

methods

of obtaining

useful

to

circuit

is confined

characteristics

the

following:

from

a

balanced-beam

type

or,

on the

other

hand,

a characteris-

(4

Block

instantaneous

comparison’

in

which

the

tic

not normally

obtainable

 

by conventional

means,

 

such

duration

of

polarity

coincidence

determines

the

as

an

inverse

relationship

between

operating

time

and

output.

The

tripping

criterion

is that

the duration

 

comparator

output.

 

Equally,

nonlinear

timing

relation-

 

of the

first

coincidence

should

exceed

a specified

ships

can

be

obtained.

It

is,

in

any

case,

illogical

to

time,

usually

one

quarter

 

of the power-frequency

 

emphasise

the

reproduction

of

conventional-relay

period.

 

characteristics

without

reconsideration

based

on

the

Cc)

Block

average

comparison,

a development

of (a),

newly

available

design

freedom

 

using

transistors.

 

Both

in which

the

duration

of

polarity

coincidence

 

is

inverse

and

inverse-square

 

timing

characteristics

arise

in

measured

on both

halfcycles

of the

input

signals,

conventional

relays,

because

they

are

inherent

to

the

and

the

average

value

is determined

in an integ-

electro-mechanical

 

elements

used

 

and

not

because

they

rating

circuit,

a

trip

signal

being

produced

if

a

are

necessarily

desirable

in a functional

sense.

 

specified

value

 

is maintained

for

more

duction

of existing

characteristics

 

in this

Repro- way can lead

to

than

average

a prescribed

duration.

The

principles

of this

unwarranted

circuit

complexity,

without

leading

to tim-

form

of

comparison

 

have

already

been

 

ing or other

characteristics

 

which

are specially

suited

to

described.2,

3, I”

 

power-system

protection

requirements.

From

 

broad

fc)

Pulse

comparison,’

in which

the

polarity

of one

consideration

 

of

the

protective

 

requirements,

it

is

the

signal

is measured

during

a short

interval

in the

authors’

opinion

(and

a view

which

appears

to

have

 

cycle

of the

second

signal,

usually,

but

not

neces-

majority

support

among

those

engineers

intimately

 

con-

sarily,

at the

latter’s

peak.

 

cerned

with

protection

design

and

application),

that

a

 

a practical

is based

 

comparator

or

amplitude

in categories

comparison

 

definite

time

characteristic

 

is the

 

most

desirable

one.

Whether

(a)

or

(b)

transistor

on

phase

 

Finally,

however,

it

is

necessary

to

assess

a

relay

has

been

shown

to

be

immaterial.

To

date,

practical

design

beyond

 

the

philosophical

factors

discussed

comparators

comparison

falling

into

category

(c)

are

of the

phase-

above.

Purely

technological

design

aspects,

such

as

long-term

circuit

stability,

susceptibility

to damaging

type

only,

even

though

equivalent

amp-

       

litude

versions

can

be

conceived.

Thus,

the

relative

 

transient

under

surges,

nonideal

economic

feasibility

system

conditions

have

and

performance

influences

which

merits

conveniently

of

practical

comparators

compared

by

of

considering

each

category

are

phase-angle

often

redirect

design

thinking

away

from

the

narrow

comparators

only,

in

detail.

This

choice

has

practical

 

requirements

 

of the

laboratory

prototype.

significance

in that

the

inherent

characteristics

of trans-

 

istors

lend

themselves

most

readily

to phase-comparator

 

Phase-comparison

and amplitude-comparison

principles.

Notwithstanding

the fact that

Ellis

‘x8established

that

there

were

no

fundamental

differences

between

these

Fundamental

principles

of operation

of transistor

 

comparators

two

principles,

unfounded

comparisons

have

been

Considering

 

phase

 

comparators

 

in

the

 

three

made.

For

example,

Mathews

and

Nellist’

presented

an

categories

of

Section

2.3,

the

operating

 

criterion

is

analysis

of

the

differential

rectifier-bridge

comparator

expressed

in the

equation:

   

and

mentioned

its inferior

transient

response

relative

to

(1)

the

transistor

phase

comparator

described

by Adamson

 

-

d

5

4

2

+

d

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

and

Wedepohl.’

In

order

to clarify

this point,

it is estab-

where

$I is the

phase

difference

 

between

 

the

two

input

lished

in Appendix

9 that

with both

a basic

phase

and an

signals

and

a is the phase-angle

 

setting,

 

usually

7r/2. For

amplitude

comparator,

each

with

specified

operating

 

the

block

comparators

 

of

categories

(a)

and

(b),

the

criteria,

the

output

signals

are

identical,

instant

by

operating

limit

0~may

be preset

between

0 and

7r to

give

instant,

provided

that

the

correct

input

relationships

overall

characteristics

 

comprising

sectors

of circles

and

specified

by Ellis’

are

observed.

Thus,

the

only

way

in

straight

lines

in

the

complex

plane.3,

4 The

case

of

an

which

differences

in dynamic

performance

in

the

two

operating

limit

of 7r/2 yields

characteristics

which

com-

cases

can

occur

is if there

are

differences

in the

passive

prise

either

straight

lines

or circles.’

Thus,

in the

steady

networks

processing

the

input

signals,

or

in the

circuits

state,

there

are

no

basic

differences

between

the

three

connected

to

the

comparator

output.

 

comparators,

but

it

can

be

observed

that

the

block-

205

comparison

principle

allows

for

greater

versatility.

passive

circuits

before

the

signals

are

compared.

Consideration

of

relative

dynamic

performance

dis-

(b)

Timing

characteristic:

The

timing

characteristic

 

closes

sharper

contrasts,

however.

Comparators

in

the

should

be

of

the

definite-minimum

type

for

all

faults

first

and

last

category

are

inherently

susceptible

to sys-

within

the

protected

zone,

allowance

being

made

for

tem

transients

and

other

spurious

signals

by

virtue

of

controlled

performance

in the

immediate

vicinity

of the

their

near

instantaneous

operation.‘,

2

Unless

all

operating

boundary.

An operating

time

of the order

of

1

unwanted

surges

and

transients

are

effectively

removed

 

cycle

of power

frequency

is considered

desirable

over

from

the

signal

inputs,

their

measuring

accuracy

cannot

the

majority

of the

practical

working

range.

be

maintained

under

dynamic

conditions

without

sac-

(c)

Stability: The

comparator

should

have

inherent

rifice

in

operating

speed.

A

compromise

solution

has

resistance

to

high-amplitude

 

short-duration

system-

been

proposed*

wherein

two

identical

comparators

are

generated

surges,

both

with

regard

to maloperation

and

arranged

to compare

signals

on alternate

halfcycles,

and

to surge

damage.

 

their

outputs

are gated

so that

transient

overreach

in one

 

element

is blocked

by the other.

Other

arrangements

use

It

is of

fundamental

importance

in developing

a sound

‘first-block-rejection

or

‘gap-timing’

circuits.

relaying

philosophy

 

to consider

in close

detail

the inher-

All

such

attempts

circuits’” to preserve

dynamic

measuring

accu-

ent

performance

as

set

out

above,

bearing

in mind

that

racy

sacrifice

speed

of operation,

because

the

form

of

modern

high-speed

relays

are

required

to operate

cor-

comparator

with

which

they

are

associated

have

no

rectly

in the

presence

offset

d.c. trans-

inherent

transient-free

characteristics.

 

ients.

The

major

part

of long-duration of any discussion

on performance

 

The

block-average

comparator,

however,

has

useful

must

thus

centre

on

the

dynamic

response

of the

relay

inherent

transient-free

characteristics,

and

it is shown

in

and

on its operating

mechanism

in the presence

of offset

the

following

Sections

that

a relay

can be designed

to

a

d.c.

components,

with

the

clear

understanding

that

the

theoretical

 

minimum

operating

 

time

of one

half

of the

 

steady-state

response

is merely

a particular

case

of the

power-frequency

 

period

without

incurring

transient

dynamic

response.

 

It

is on

this

basis

that

relays

using

the

overreach

 

and

without

resorting

to special

filtering

cir-

principle

of block-average

comparison

are

cuits

in the

input

signal

paths.

The

operating

time

is not

 

tage

over

simpler

and/or

faster

arrangements

at an advan-

significantly

affected

by the

instant

of fault

initiation

or

design

is based

 

on

steady-state

considerations

in which only.

degree

of the d.c. offset

transient

in the input

signals,

and

can

be

precisely

defined

 

following

the

procedure

 

Basic considerations

 

described

 

in

Section

3.

Timing

over

the

full

working

Fig.

1 shows

a schematic

diagram

of a basic

relay

using

range

approaches

 

the

ideal

definite-time

characteristic,

the

phase-comparator

principle;

the definitive

equations

but

as

the

critical

phase

angle

d

(usually

n/2)

is

approached,

the

timing

tends

to infinity

at the boundary

 

“1

of operation.

This

controlled

timing-characteristic,

and

“L_

measuring

and

m,x,ng

<

colncldence

 

hxlr

Ievel

 
 

clrcult

-

mtegrator

-

detector

&

the

use

of both

halfcycles

for

measurement,

contribute

IL

-

circuits

_

 

most

significantly

to the dynamic

accuracy,

and contrasts

 

“2

with

the

other

two types

of comparator

with

their

varia-

 

FIG.

1.

BASK

BLOCK

AVERAGE

COMPARISON

RELAY

tion

in timing

depending

on the point-on-wave

instant

of

fault

initiation

and

the

uncontrolled

timing

characteris-

 

can

be derived

and

performance

will be identical

for

an

tic

at

or

near

the

boundary

of

operation.

This

latter

equivalent

amplitude

comparator.

Two

input

quantities

property

largely

accounts

 

for

their

poor

dynamic

Vi

and

Vz, derived

from

input

voltage

and

current

VL

measuring

properties

and

for

their

susceptibility

to

and

I,_ in a measuring

and

mixing

circuit,

are

compared

maloperation

ient

signals;

resulting

from

the comparators

spurious

surges

and

in categories

(a) and

trans-

(c) are

in

a

coincidence

 

circuit

producing

standard

output

probably

of lowest

merit

in these

respects.

pulses,

 

which

are

positive

when

Vi

and

V2 are

of the

           

same

polarity

and

negative

when

they

are

of opposite

 

polarity.

The pulses are applied

to an integrating

circuit

BLOCK

AVERAGE

COMPARISON

 

whose

output

increases

linearly

during

the

time

when

Specification

of design requirements

 

the

pulse

is positive

and

falls

at the

same

when

the

 

polarity

reverses.

The final element

 

rate

is a level

The

following

are

the

factors

 

of

most

significance

in

detector

which

switches

when

the

in the relay integrator

output

relating

inherent

comparator

performance

 

to

the

con-

exceeds

some

preset

value,

and

resets

when

the

output

junctive

performance

 

of

any

practical

distance-

 

falls

below

some

second

value.

measuring

 

relay

and,

as such,

they

are

used

as the

basis

Figs

 

2

and

3

show

the

relevant

waveforms

in

the

for

specifying

the

block

average

system:

steady

state

for

phase

displacements

4

>

r/2

and

r#~<

(a)

Measuring

accuracy:

The

specified

accuracy

7r/2, respectively.

It is evident

that

the output

signal from

should

be maintained

over

the

full working

range

when

the

integrator

is sawtooth

in nature,

and

that

there

is an

measured

under

realistic

dynamic

conditions

with offset

effective

gain

in output

only

for

the

condition

$J <

n/2.

d.c. transients

and

other

spurious

signals

superimposed

 

The

rise

and

fall

rates

in

the

integrator

are

at

the

on the input

quantities.

Long-term

stability

of measuring

designer’s

disposal,

so that

the

critical

phase

angle

may

accuracy requires

that

the

comparator

design

levels

be

be

set

to any

desired

value.

Both

the

level-detector

 

set

such that

all vectorial

signal

mixing

should

be

done

in

and

reset

levels

are

critical

in relation

to the

total

excur-

206

(1)

-

----

 

----

--_--

----

 

(ii)

 

-----

-

----

_---------

 

(iii)

-----

-----

_-

----_

 

FIG.

2.

RELAY WA~ORWS($

3 42)

 

a Inputsignals to coincidence circuit

 

b

Output

from

coincidence

circuit:

(i)

Upper limit

(ii)

Set

level

(iii)

Rest

level

('Integrator output

sion limits

of integrator

linearity

and

also

to the

slope

of

the