protection of HV motor

7/29/2019 protection of HV motor

1/28

E/CT 165 first issed, May 1995

J.Y. Blanc

He graduated as an engineer fromSUPELEC in 1979. He joined MerlinGerin in 1981 as Medium Voltagedielectric engineer. Still in theMV Division, he was then in turn asproject manager for the developmentof:

- Fluair range primary substationsand equipment,- SF set built-in protection circuit-breakers.He is currently in charge ofdevelopment of MV primarydistribution and switchgear.

n 165

control/monitoring and

protection ofHV motors


2/28

Cahier Technique Merlin Gerin n 165 / p.2


3/28


control/monitoring and protection ofHV motors

contents

1. Reminder of the various types of Asynchronous cage motors p. 4

Asynchronous slipring p. 5rotor motors

Synchronous motors p. 6

Sizing tolerances p. 6

Dielectric withstand and tests p. 7

2. Classical HV starting processes Direct stator starting on p. 8full voltage

Stator starting on reduced p. 8voltage

Startor starting with capacitors p. 9

Rotor starting p. 10

Choosing the starting mode p. 11

3. Control and monitoring equipment Electromechanical solutions p. 13

Electronic solutions p. 15

4. Protection of HV motors Main fault types p. 18Protection principles p. 18

Technological evolution p. 22

Appendix 1: determining motor Calculation hypotheses p. 23

Overall approach p. 23

Appendix 2: coordination of protection devices p. 25

Appendix 3: bibliography p. 26

Unit power of rotating machinesfrequently exceeds 100 kW in industryand large tertiary. In such cases and/orif the length of the supply line is

particularly long (voltage drops,losses), use of high voltage motors isadvantageous.The purpose of this Cahier Techniqueis to analyse and compare thesemotors, their starting systems and thevarious protection devices which maybe used, in order to simplify technicalchoices.

AC motors

starting mode


4/28


current

torque

0

1

2

2

4

25 % 50 % 75 % 100 % (N/Ns)

1

25

4

6

3

2

1

Cn

In

current

torque

0 25 % 50 % 75 % 100 % (N/Ns)

Id

1. reminder of the various types of AC motors

Both high and low voltage AC motors,offer a large variety of electrical,dynamic and technologicalcharacteristics. However, except for asmall number of motors used for highlyspecific applications, they can bedivided into three families, namely:c asynchronous cage motors;c asynchronous slipring rotor motors;c synchronous motors.

They differ from each other in:c starting current and torque values;

c speed variation in normal operation;c power factor and efficiency values asa function of load.

HV motors are supplied with a voltagerarely exceeding 7.2 kV, their powerranges from 100 kW to over 10 MW,with an average of 800 kW.

asynchronous cage motorsThese HV motors fall into two maincategories according to their rotorcomposition which can be single ordouble cage.

This enables choice of starting currentand torque characteristics:c single cage rotors have:v a relatively low starting torque (0.6to 1 C

n),

v a maximum torque of around 2to 2.2 C

n,

v a starting current ranging from 4.5to 5.5 In,(Cn: rated torque, In: rated current);

c double cage or deep slot rotors have:v a slightly higher starting torque (0.8to 1.2 C

n),

v a maximum torque of around 2

to 2.2 Cn (slightly higher for deep slots),v a starting current ranging from 5to 6.5 In.

Figures 1 and 2 show the form of thesecurves as a function of speed (N/N

s).

Note that:c single cage motors have a minimumtorque (0.5 to 0.6 C

n), whereas the

torque curve, varying according to thespeed of the double cage or deep slotmotors, continues to increase up tomaximum torque.c these motors are ideal for intensive useand dangerous environments, due to:

fig. 2: curves C(N) andI(N) of an asynchronous double cage motor.

I

In

C

Cn

fig. 1: curves C (N) andI(N) of an asynchronous single cage motor.

C

Cn

I

In


5/28


fig. 4: curves C (g) of an asynchronous slipring rotor motor.

v the simplicity of rotor design in

short-circuit providing them with an

excellent mechanical and electrical

robustness,

v absence of brushes.

These two features allow maintenance

to be reduced to a minimum.

The torque characteristics of

asynchronous cage motors are

especially suitable for machines such

as centrifugal pumps, compressors,

converter sets, machine-tools and fans.

However, all these motors have the

drawback of a relatively low power

factor, around 0.8 to 0.9 on full load,

and even less when they are running

on low load (see fig. 3).

If asynchronous motor installed power

is high, reactive power compensation is

required. This may either be global, for

each set of motors or for each motor

(large units).

asynchronous slipringrotor motorsThe rotor winding of these motors

connected to sliprings means the

resistance of this circuit can be

modified by introducing external

resistances.In the motor stability zone,

corresponding to the positive slope of

curve C = f (g) (see fig. 4), the slippage"g" is proportional to the rotor

resistance:

g = 1

ARr C

where g % = Ns N

Ns100

where:

Ns

: synchronous speed,

N : operating speed.

A = 3V2 p

M

L1 = constant

where:

V: phase to neutral supply voltage,p: number of pole pairs,

: pulsation of supply currents,M: reciprocal stator-rotor inductance,

L1: total stator choke,

(L1

= M + Ls)

Rr: rotor resistance = rotor inherent

resistance + external resistances,

C: motor torque.

cos

0 2/4 3/4 4/4

0.5

0.6

0.7

0.8

0.9

cos

(P/Pn)

fig. 3: efficiency curves (P) and power factor curves cos(P) of an asynchronous double

cage motor.

CM

C

Rr R'r > Rr R''r > R'r

CnCd

C'd

C''d

1

0.50

0.5

1 2 g

operation

on generator

operation

on motor

operation in

regenerative

braking

rotor stability zone

in short-circuit


6/28


c the power factor which may be set bythe exciting current.Technologically, they are identical toAC generators.

In order to obtain asynchronous torqueand avoid oscillations, the rotors areequipped with a damping cage. Thiscage means synchronous motors canbe started with a low load torque insimilar fashion to asynchronous singlecage motors (they have practically thesame characteristic torque and currentcurves). To avoid surge voltages in theexciting circuit, this circuit is shuntedduring starting and on tripping by aresistance with a value chosenbetween 5 and 10 times the resistance

of the exciting circuit.In view of the fact that theasynchronous torque tends to zero onapproaching synchronous speed,coupling to the network when motorstarting is complete cannot take placeat synchronous speed as is the case forAC generators. The result is invariablya transient state varying according tothe speed acquired at the end ofstarting and motor power. To limit thistransient state:c either use a relay to monitor slippageby measuring frequency of the rotor

current passing through the startingresistance. This relay controls excitingcircuit supply when slippage is at itslowest.This device is practically indispensablewhen the synchronous motor accountsfor a large part of total installed power.c or apply the exciting current in twostages, automatically or manually.

The exciting sources can be either

separate:

c motor set, exciter;

c thyristor rectifier;

or placed at the end of the motor shaft:c reversed generator;

c rotating "diode rectifier and armature"reversed AC generator.

The techniques most frequently used

are the thyristor rectifier and the"rotating diodes".

The latter does away with brushes,

removes the exciting cubicle and, in

addition, is often fitted with asynchronisation and recoupling

mechanism should synchronism fail.

These motors can supply reactiveenergy by increasing the excitingcurrent. This characteristic, whichenables compensation of network

reactive loads, is one of the mainreasons these motors are chosen.

The curves in figure 5 show statorcurrent variation as a function ofexciting current for a given constantload (Mordey curves). Use of this typeof motor for small powers is fairly rare.On the other hand, it is frequently usedabove 2,000 kW for its excellentefficiency and control of its powerfactor. In the case of highly regularmovements, synchronous motors are anecessity. However, the machines

being driven must have a relatively lowload torque during starting, and sizingof the dampening cage limits startingrate.

sizing tolerancesThe electromechanical characteristicsof motors are defined by standard

IEC 34-1. In the case of certain rated

characteristic values, the standard

defines the tolerances to be compliedwith by the manufacturer. It is useful tobe familiar with these tolerances since,

for certain characteristics, they directlyaffect the choice of motor and

equipment power and the setting of theprotection devices.

The table in figure 6 gives the

tolerances of the main characteristic

values.

By decreasing external resistance onstarting, the characteristic C (g) istranslated and the starting torqueadapted to the torque of the machine

being driven. Note that maximumtorque value does not depend on rotorresistance.

Moreover, for small slippages, rotorcurrent is inversely proportional to rotorresistance. Its amplitude is given by:

I2 = Bg

Rr

where

B = VM

L1.

Stator current follows the same law,except for the winding ratio and themagnetising current.

Consequently, the choice of initial rotorresistance solves virtually all theproblems concerning high startingtorque and inrush current on thenetwork and ensures these tworequirements are satisfied. The variouspossibilities for using asynchronousslipring rotor motors make themsuitable for driving machines with highstarting torque such as crushers,mixers, conveyors, etc.

Moreover, machines requiring highregenerative braking also use this typeof motor.

Just as for asynchronous cage motors,the power factor in normal operation isrelatively low: this characteristic and thepresence of sliprings and rotorresistances mean these motors areincreasingly replaced by double cage ordeep slot motors.

Figure 4 represents the characteristicC (g) curves, according to rotorresistance value and the stabilityzones. These curves show the

advantage of introducing a highresistance in the rotor circuit as ameans of obtaining efficientregenerative braking.

synchronous motorsThe main differences withasynchronous motors are:c their constant speed (synchronousspeed);c the rotor circuit supplied with DC;

fig. 5 synchronous motor: Mordey curves.

0 (i) excitation

(I) stator

stabilit

ylim

it full load1/2 load

1/4 load

cos AR cos AVcos = 1


7/28


dielectric withstand andtestsMotors, just like all electrical network

components, are subjected to a varietyof surge voltages. They are particularlysensitive to steep front surge voltagesor high frequency since they are"jammed" by the first turns of the statorwindings.

Switching surge voltages

These are the result of transientphenomena occurring during changesin status in the supply network.The following phenomena, specific toinductive circuits and thus to motors,must be taken into consideration:c current pinch-off on current breaking;

c multiple re-ignitions on breaking andprearcing on current making if thebreaking device is capable of breakingthe high frequency currentscorresponding to these phenomena.

Steep front surge voltages

These are the result of direct or indirectlightning strokes. They spread onto thenetwork, creating a dielectric stresswhich, even when limited by the use ofsurge arresters, can be considerable.Surge voltage is studied in detail in"Cahier Technique" n 151"Overvoltages and insulation

coordination in MV and HV" andspecial motor sensitivity in"Cahier Technique" n 143 "Behaviourof the SF6 MV circuit-breakers Fluarcfor switching motor starting currents".To check the motors capacity towithstand to these various surgevoltages, the motors undergostandardized tests performed asdefined by IEC 34-1 standard.

The test voltage is applied between thewinding being tested and the body ofthe machine to which the magneticcircuits and all the other stator and rotor

windings are connected.

Two types of tests are stipulated in the

standards: standard frequency tests

and impulse withstand tests.

Standard frequency testWithstand to switching surge voltages

is checked in compliance with standard

IEC 71, by the standard frequency

withstand test. Testing commences

with a voltage of less than U/2 which is

gradually increased up to 2 U + 1,000V,

at which level it is applied for one

minute.

For the stator, U is the specified supplyvoltage. For the rotor, U is the voltage

which appears, with the rotor circuit

open, when the specified statorsupply voltage is applied with the

rotor locked in rotation. If the motor

is reversible (change of rotation

direction of motor already started), the

test voltage applied to the rotor will be4 U + 1,000 V.

Impulse withstand test

This test consists in applying animpulse voltage representative of

lightning:

c buildup time: 1.2 s;

c dropdown time at Upeak/2: 50 s;

c test voltage:

Upeak

= 4 U + 5,000 V.

The windings are subjected to a

number of positive and negative waves.

Impulse withstand tests are not

currently mandatory in standards, as

they can lead to early ageing of

armature and winding insulation.

More generally, the dielectric testsmust not be repeated; if a second testis performed, it will be carried out at

80 % of the voltages indicated above.

value tolerance

asynchronous motors

current with locked rotor + 20 % of current

and short-circuited (no lower limit)

torque with locked rotor - 15 % to + 25 % of torque guaranteed

minimum torque during starting - 15 % of torque guaranteed for cage motors Cdu at

a third of rated torque andi

at half of torque withlocked rotor, at full voltage

maximum torque - 10 % of torque guaranteed provided that once thistolerance is applied, the torque remains u 1.6 timesrated torque

synchronous motors

current with locked rotor + 20 % of value guaranteed

torque with locked rotor - 15 % to + 25 % of value guaranteed

pull-out torque - 10 % of value guaranteed provided that once thistolerance is applied, the torque remains u 1.35 timesrated torque (1.5 for synchronous motors with salientpoles)

fig. 6: tolerances on the main characteristic values as in standard IEC 34-1.


8/28


0 0.2 0.4 0.6 0.8 1

0.2

0.4

0.6

0.8

1

2. classical HV starting processes

The main HV motor starting processesare as follows:c direct stator starting on full voltage;c stator starting on reduced voltage bystar-delta connection, by reactance orby autotransformer;c stator starting by capacitors;c rotor starting.

direct stator starting on fullvoltageThis starting mode is used forasynchronous motors with cage rotorand for synchronous motors.

Current peak on starting is around 4to 7 In, according to motorcharacteristics, and can last for roughly1 to 10 seconds depending on themoment of total inertia (motor +machine), motor torque and loadtorque.

If this starting mode is used, thenetwork must be able to withstand theabove current overload without

disturbing the other loads, and themachine being driven must be able towithstand the mechanical impact due tothe motor torque. The simplicity of bothequipment and motor and the resultingsavings mean that this mode is verypopular and even recommendedprovided that voltage drop on thenetwork on start up is acceptable. Thedecisive factor lies in the motor power/short-circuit power ratio.

stator starting on reducedvoltage

Star-delta startingThis starting mode is used to reduce:c current in a ratio of e;c starting torque by a third.

It is used in LV and for low powers, butrarely in HV due to the high currentpeaks when moving to delta. In thiscase it is replaced by reactancestarting.

Voltage reduced by resistanceCommonly used in LV, it is rarely usedin HV due to the Joules to be dissipatedand resistance insulation problems.

Voltage reduced by reactance

This starting mode (refer to powerdiagram, figure n 8) is the one whichreduces current inrush on the networkin the simplest manner. Since motorstarting torque is low, the machinesbeing driven must have a relatively lowload torque during start up:compressors, centrifugal pumps,converter sets, etc.In point of fact, asynchronous motortorque varies according to square of the

supply voltage , whereas the absorbedcurrent remains proportional to thisvoltage.

C'd = Cd UdUn

2

where :C'

d: starting torque with reduced

voltage,C

d: starting torque with full voltage,

Ud: starting voltage,

Un: rated operating voltage.

I

'd = Id

Ud

Un

where:I'd: starting current with reducedvoltage,Id: starting current with full voltage.These relationships can also beexpressed using the ratedcharacteristics as follows:

I'dIn

= Id

In

UdUn

.

The curves in figure 7 give the ratiovariations as a function of

the ratio UdUn

.

Voltage at the motor terminalsincreases gradually during starting. Theresulting start up is smooth.c operation and schematic diagram:v first stage:operation on reduced voltage byclosing the line contactor C

L,

v second stage:normal operation by closing the short-circuit contactor C

C;

c determining a starting reactance(see fig. 9)

Starting voltage is determined by the

maximum current inrush l'd authorised

on the network:

Ud = Un I'dId

.

The phase-to-phase voltage drop in the

reactance has the following value:

Un

Ud

= j 3 L I'd

.

The diagram in figure 9 shows that this

relationship can be expressed in anarithmetic form for asynchronousmotors, since the power factor at theinitial moment start up moment is

initiated is virtually the same as the

power factor of the starting inductance.

Therefore:

L = Un Ud3 I'd

.

Knowledge of start up time and

operation rate is required to determinereactive power.

Voltage reduced by autotransformer

This starting mode sometimes makes it

possible to reconcile reduction of

current inrush on the network and

motor torque value. In point of fact, it

has the advantage of reducing current

inrush according to the square of the

winding ratio:

fig. 7: graphs showing start-up on reduced

voltage (by reactance or star-delta).

UdUn

I'dId

andC'dCd

C'd

Cd

I'dId


9/28


I'd

d

Un

Ud

rating of the neutral point contactor tobe reduced;c there is an alternative version of thisconfiguration in which the neutral point

contactor is removed. This version isnot recommended since the move fromreduced to full voltage necessites themotor-network link be cut. In view of therelatively short switching time, and theresidual voltage at the motor terminals,a current inrush exceeding startingcurrent can ocur on transfer to normaloperation. This is unacceptable for boththe network and the motor and can leadto the tripping of the protection devices.

stator starting withcapacitorsThis process enables motor full voltage

starting characteristics to bemaintained. It is mainly used forkeeping the starting torque ofsynchronous motors constant duringstart up, for example in cement worksand crushing plants.

Capacitors, as well as motor, supplypart of the reactive energy during thestarting phase: the motor power factoris small at this stage. Power inrush on

I'dIn

= Id

In

UdUn

2

C'dCn

= CdCn

UdUn

2

where:

I'd: starting current on network side with

reduced voltage.

These relationships are used to

determine the value of reduced voltage

authorised on the network, as a function

of the ratio'dn

or of the ratioC'dCn

authorised by the machine being driven.

The curve in figure 10 gives the

variation ofUdUn

as a function

of'dn

orC'dCn

.

c operation and schematic diagram

(see fig. 11)C

L: line contactor,

CC: short-circuit contactor,

CPN

: HV neutral point formationcontactor,

AT: autotransformer

v first stageoperation on reduced voltage by

closing CPN

which causes CL

to close,

v second stageoperation in inductance by opening CPN,

v third stage

operation on full voltage by closing CC.

Remarks:

c in theory the second stage is short(around one second) since in mostcases it is a slowing-down time;Use of an autotransformer with air gapsconsiderably reduces this fault, butrequires knowledge of the value of thecurrent absorbed by the motor at theend of the first stage;

c the move to full voltage invariablyresults in a transient state whoseduration varies according to the speedacquired at the end of the first stageand the value of the current absorbed;

c the current flowing through theneutral point on start up is equal to thedifference between the motor currentand the line current, excluding themagnetising current of theautotransformer. This enables the

fig. 9: vectorial diagram to determine L.fig. 11: power diagram: starting by

autotransformer.

fig. 10: graphs showing start up byautotransformer on reduced voltage.

0 0.2 0.4 0.6 0.8 1

0.2

0.4

0.6

0.8

1

I'dId

andC'dCd

Ud

Un

fig. 8: power diagram: starting by reactance.

J L I'd

3

M

neutral point

AT

3L

CL CC CPN

3L

CC

CL

M

3L

L

j


10/28


the network is reduced accordingly(see fig. 12).

This technique is tricky to implementand calls for a study of the motor and

capacitors to avoid resonance andsurge voltage (due to motor self-excitation), as well as mechanicaloscillations on the transmissionsystem.

Moreover, control equipment must bechosen especially for capacitorswitching.

rotor startingThis starting mode solves virtually allthe problems which may occur onstarting, namely:c reduction of current inrush on thenetwork with increase in motor torque;c adaptation of motor torque to loadtorque;c long, progressive starting (e.g. forloads with high inertia).

This mode can only be used forasynchronous slipring rotor motors andfor synchronous induction motors (its

use is increasingly rare in industry).It is particularly used for load starts.

Example of a rotor start in n time

This start up is illustrated in figures 13and 14. Motor torque varies betweentwo values at each notch. The lowervalue is taken as being equal to therated torque. At each notch, rotorresistance changes value and thetorque-speed characteristic evolves. Inthe last stage, rotor resistance is simplyequal to the rotor internal resistance.c first stagestator supply and starting on full rotorresistance by closing C

L;

c second stageshort-circuiting of the first section of therotor resistance by closing C

1;

c third stageshort-circuiting of the second section ofthe rotor resistance by closing C

2;

c Nth stageshort-circuiting of the n-1 section of therotor resistance by closing C

n-1.

The number of stages, or notches, n isalways greater by 1 than the number ofsections or contactors.

This number n is determinedapproximately by the formula:

n = log gn

logCnCp

or byCnCp

= gnn

where:C

p= peak torque,

gn

= rated slippage.

According to each case, n can bededuced from Cp, and vice-versa.

Complete determination of rotor startingequipment calls for knowledge ofoperation (hourly rate and startingtime). Lack of standards for HV motorsmeans this equipment is determined ineach individual case by specialists.

RemarkA linear start is sometimes required.This calls for power electronics

enabling monitoring of rotor energy: forexample using a Gratz bridge and achopper to make a continually variableresistance (see fig. 15).

fig. 13: power diagram: rotor starting.

fig. 12: vectorial diagram for starting by

capacitor.

fig. 14: symmetrical rotor starting diagram.

power suppliedby the network

active power

power suppliedby capacitors

apparentmotor poweron starting

inductivereactivepower

M

neutral point

3L

CL

3L

C1

C2

Cn - 1

4th 3rd 2nd stage 1st stage

0 1 g

Cn

Cp

CM

C

gn g4 g3 g2 g1


11/28


C

C

3

3

C

C

3

C

3 n

3 n

3 n

n

n

characteristics examples

constant torque piston compressor

parabolic torque pump:

valve opened

- - - - valve closed

high breakaway crusher

torque

Starting conditionsIn view of the required startingcharacteristics, it is necessary to checkthat starting can take place in correctly

as far as motor torque, current inrushand starting time are concerned, for thestarting techniques underconsideration:

c motor torque is always greater thanload torque (see fig. 17);c current inrush on the network and thecorresponding voltage drop are

acceptable to the network;c starting time is compatible with theequipment used.

Approximative calculation of startingtimeOperation of the motor-driven machineassembly is governed by the followingmechanical equation:

Cm Cr = Jddt

where:C

m: motor torque with Un voltage,

Cr: load torque,

J: inertia of rotating frames (motor andmachine driven),

ddt

: angular acceleration.

Angular velocity varies from 0 to nthroughout starting time t. Moreover, amean accelerating torque C

a, equal to

the mean difference between Cm

andC

rcan be defined:

This results in Ca

= (Cm

- Cr) mean

= J(n 0)

thence:

t = J nCa

.

Bearing in mind that real motor torquevaries according to the square of itssupply voltage:

C'mCm

= UrealUn

2

with real voltage.

Cm

= motor torque with real tension.Reducing this voltage will thus reduce

Ca, hence increasing starting time.

fig. 16: rreminder of load torque curves of machines to be driven (loads). fig. 17: case of non-starting (C'm < Cr).

choosing the starting modeChoice of starting mode is conditioned

by ensuring that the motor torque and

the load torque of the load are properlymatched.

Knowledge of the load torque is

required (see fig. 16).

torque

1.8

0.9

0speed

Cm

C'm

C'r

C

Cn

NI

I0

R1

I

N0

R2

H

L

--

+ +N

M

3L

tachometer dynamo

fig. 15: setting speed by rotor chopper.


12/28


Starting mode selection table

The table in figure 18 summarises the

advantages and disadvantages of the

main starting modes for the various

applications.

For a given torque, the current

absorbed on the network is establishedin the following increasing order:

c rotor starting;

c starting by autotransformer;

c starting by stator impedance;c direct starting.

Choice of starting mode calls for good

communication between the electricalenergy supplier and the manufacturerof the motor and machine being driven.

The vital characteristics for making thischoice are:c supply network power and maximumauthorised current inrush;

c motor torque and current at fullvoltage as a function of rotation speed;c load torque of the machine driven(see fig. 16);

c moment of inertia of the rotatingframes.

If the supply network power/motor powerratio is less than 5, particular care mustbe paid to choice of starting mode andchoice of coordination of the protectiondevices (see appendices 1 and 2).

fig. 18: starting mode selection table for the most common applications.

application application starting mode controlled by advantagesneeds characteristics disadvantages

permanent or machines requiring direct 1 or 1 simplicity,quasi-permanent high start torque less investment.

process i 1/jour on starting:frequent starts > 1/day motors with low current direct chigh torque;

inrush or low power 1 chigh current inrush,;chigh mechanical stresses.

pumps, fans, machines starting stator 2 reduction of current inrushcompressors, with low torque by reactance on startingfrequent starts (possible adjustment).

optimisation of when starting current stator by 3 optimisation of torquestarting characteristics must be reduced, but autotransformer (reduced) and of current inrush

the necessary starting on startingtorque maintained (possible adjustment).

optimisation of the most difficult starts rotor generally small current inrush andhigh torque 3 high starting torque.starting characteristics


13/28


3. control and monitoring equipment

The function of this equipment is

threefold:

c energising and stopping (control);

c disconnecting the motor should a

fault occur (protection);

c monitoring the motor (monitoring).

When we talk about monitoring, thisimplies the equipment is capable (or

not) of:

c initializing the start (starting sequenceautomation);

cacting on motor speed;c supplying information on motor

electrical status and contributing to

protection.

The monitoring function chiefly relies on

power electronics and low currents

(digital technology); it is currently being

fully developed. Protection of

HV motors will be dealt with in the next

chapter.

electromechanical

solutionsThe choice between the various

devices (switch, circuit-breaker or

contactor) depends on:

c operation rate;

c electrical endurance;

c motor power.

The main breaking device

characteristics are summarised in the

table in figure 19.

Fuse-switches

By their very design, the breaking

capacity, mechanical and electrical

endurance of switches is low. Thislimits their use to small powers

(In

= approx. 50 A - 5,500 V) andto rates of two to three operations a

day.

Moreover, the low breaking capacity ofthese devices makes choice ofprotection devices tricky.

Circuit-breakersCircuit-breakers are generally used forhigh motor powers of more than 300 A,with a small operation rate, and foroperating voltages of more than 6.6 kV.

Their use can naturally be extended tolower powers, operable by switch orcontactor.

Fuse-contactorsc operation rateTheir simple control mechanismcombined with the robustness and

simplicity of their contacts meancontactors have a high operation rate.This rate cannot be withstood by circuit-breakers, even special ones, and evenless so by switches.

Some installations use contactors withmechanical latching to do away withpermanent consumption of the closingelectromagnet. This may reduceendurance as a result of the greatercomplexity of the kinematic chain.

c network short-circuit powerThis factor does not really affectcontactors thanks to the presence of

fuses placed immediately after theisolating switch or next to thedisconnecting contacts on the busbarside. These fuses with their highbreaking capacity, limit the short-circuitcurrent.

This special feature means that,if network power is increased, themotor feeder cubicles can bemaintained. The busbar supports arestrengthened if required.

FusesFuse rating is determined according to:c rated current In;c the I

d/I

nratio (I

d= starting current);

c starting time determined using thechart in figure 20 page 14.

Finally, it should be pointed out thatfuses protect the motor againstovercurrents roughly five times greaterthan motor rated current and that theymust be combined with additionalprotection devices (thermal relays, etc.

see "protection device" chapter).Current transformersThe increasing use of digital protectiondevices means unconventional currentsensors can be used (e.g. Rogowskitoroids). These sensors have theadvantage of being linear and thusdelivering an accurate signalthroughout the useful current range.

They present no saturation or thermalproblems, as is sometimes the casewith classical Current Transformers(see "Cahier Technique" n 112 "Thebreaking process with a Fluarc SF6

puffer-type circuit-breaker").

Special features due to use of fusesor circuit-breakers

c operation in single-phase due to fusemelt, with the striker not working.

Today striker reliability is such that therisk is slight. Their dependability maybe increased still further by use of anadditional protection device(undervoltage or unbalance relay).

fig. 19: breaking device application field.

device mean rate endurance acceptable motor

nb operations powerfuse-switch low: 2-3/day 2,000 low i 50 A

circuit-breaker low: 10/day 10,000 highu 7.2 kVA> 300 A

fuse-contactor high > 10/h > 100,000 average i 300 A


14/28


fig. 20: fuse determination chart.

0.01

0.02

0.03

0.04

0.05

0.1

0.5

1

5

10

50

100

t seconds

100A 125A 160A 200A 250A

rating of 7.2 kV fuses

example:

motor In = 100 A

Id = 6 . Intd = 5 s

fuse rating = 200 A

motor

Amp In

10

20

30

40

50

60708090

100

150

200

I

Id=4

In Id=

5In

Id=6

In Id=7

In


15/28


c discrimination with upstreamequipment.This can be hard to achieve when both:v the fuses used have a high rating

(200 A or 250 A),v and the feeder protected by thesefuses accounts for a large fraction ofthe power supplied by the main circuit-breaker (see fig. 21 and 22).

However, the high breaking capacity ofthe Rollarc contactors combined withthese fuses allows the use of slightlytime delayed overcurrent relays, thusensuring discrimination. Discriminationis easier to achieve if the motor feederis protected by circuit-breaker butsince, for high short-circuit currents thecurrent is not limited, there is an

increase in thermal stresses.c surge voltagesSome types of device, in particularvacuum breaking devices, generatesurge voltages on motor energising andstopping (due to their capacity to breakhigh frequency currents, resulting forexample from the current pinch-offphenomenon - see "Cahier Technique"n 143).In order to prevent these surgevoltages progressively damaging motorinsulation, manufacturers place ZnOtype surge voltage limiters in the

equipment if required.It can be concluded that today'selectromagnetic solutions are reliable,robust, economic and entirely suitablefor most applications.

electronic solutionsThese provide users with additionalpossibilities and advantages such as:c variable speed;c possibility of speed regulation;c high operation rate;c energy savings.

The electronic solution is rarely usedjust for starting.

Before dealing with standard cases ofelectronic devices, it should be pointedout that their use calls for a certainnumber of constructive precautions atmotor level:

fig. 21: protection diagram showing a high current motor feeder.

fig. 22: discrimination diagram for a high current motor feeder.

c safety margin in temperature rise dueto harmonics: a 15 % margin on currentis generally sufficient;c forced ventilation is recommended(motors can run at low speed);

c reinforced insulation between turnsdue to the high voltage gradientsgenerated by thyristor switching (whichmay reach the magnitude of impulsewithstand test ones).

t

I

contactor

breaking

time

contactor

overcurrent

relay

contactor

breaking

capacity

circuit-breaking

melting curve

remain circuit-breaker

overcurrent relay

discrimination

loss

M

overcurrent

relay

circuit-breaking

contactor

main

circuit breaker

overcurrent

relay


16/28


The devices described below are theones most frequently used in mediumvoltage. Only the general principles arepresented.

The table in figure 23 gives an idea ofthe adequations between the type ofspeed controller, the type of motor andthe type of load driven.

Autonomous rectifiers/inverters

With their capacity to deliver variablevoltage and frequency, they guaranteecomplete control of motor speed andtorque.

A reminder for asynchronous motors:c torque is proportional to U2 if f and Nare constant;c torque is inversely proportional to f fora given voltage and speed.

There are three types of autonomousrectifiers/inverters.

c voltage rectifier/inverterv the thyristor rectifier sets voltage,v the thyristor inverter suppliesAC current at variable frequency.

This configuration is also used byUninterruptible Power Supplies (UPS)used in LV to supply computers. Theonly difference is that frequency andvoltage are fixed.

c rectifier/inverter with Pulse WidthModulation (PWM)

v the diode rectifier supplies theinverter,v the inverter generates voltage pulsesenabling reproduction of a sine wave

with variable period and amplitude (seefig. 24).c current rectifier/inverter (switch)v the thyristor rectifier combined with asmoothing choke acts like aDC generator,v the inverter switches current in turn in

the motor windings using capacitors.Motor frequency and thus speeddepend on switching speed.A few elements of comparisonbetween these three types ofrectifiers/inverters:c voltage inverterv suitable for high reactance motors,v often requires a filter between inverterand motor,

v enables regenerative braking if therectifier is reversible.c the PWM inverterv allows a wide range of speeds,v maximum speed limited by themaximum switching frequency

authorised by the inverter thyristors.Use of power transistors (IGBT) makesit possible to work at far higherfrequencies but at lower powers,v reversible operation possible (tworotation directions).c the current switchv suitable for low reactance motors,v enables operation in the fourquadrants.

motor loads speed variations power overall speed controllerefficiency type

asynchronous pumps, fans, 0 % to over 100 % a few 10 kW 0.85 to 0.90 autonomousor synchronous compressors, extruders to a few 100 kW rectifier/inverter

asynchronous same 60 % to 100 %* a few 100 kW 0.90 to 0.95 subsynchronousslipring to a few MW cascade

synchronous same 0 % 100 kW to a 0.90 to 0.95 selfcontrolledcentrifugal machines to several times 100 % few 10 MW rectifier/inverterTGV bogies(high speed)

asynchronous crushers, rolling 0 % to 33 % 100 kW 0.85 to 0.90 cyclo converteror synchronous mills, cement kilns to a few 10 MW

(low speed)

*: 100 % corresponds to relative speed at 50 Hz

fig. 23: application areas of electronic speed controllers for AC motors.

fig. 24: generation of variable voltages and frequency with PWM speed controllers.

U

U

0

0

t

t


17/28


Subsynchronous cascadeAsynchronous slipring rotor motors arenormally supplied by the network.Speed is adjusted using the rotor

current by means of a rectifier/inverterset. The rectifier draws off energy from

the rotor circuit, thus increasing

slippage. The energy drawn off

depends on the conduction setting of

the thyristors of the inverter which

reinjects energy into the network (see

fig. 25). The subsynchronous cascade

enables continuous speed variation

with a maximum slippage of

around 40 %.Converter set power is low compared

with motor power, and the energy

regenerated results in outstandingoverall efficiency.

Note that the converter is put into

operation after starting by rotor

resistance.

This assembly can operate at speeds

higher than synchronous speed

(supersynchronous) in the case of

driving loads.

Selfcontrolled rectifier/inverter

The motor stator phases (in this case

synchronous) are supplied in turn just

as for the current switching

autonomous rectifier/inverter.

Switching from one stator phase to thenext is selfcontrolled by motor speed bymeans of a "notched disk" sensor.

There is thus a correspondence

between the excitation flux and the

armature flux, as for DC machines, and

the pull-out risk is zero.

Since switching presents problems both

on starting and at low speed, the con-

verter control system must be modified.

This solution is ideal for synchronous

motors.

Cyclo converter

Each motor phase is "supplied" by athree-phase double bridge.The first bridge is used to draw off

current throughout the positive cycle oneach of the network phases accordingto the required frequency.

The second bridge is used for reversecurrent during negative alternation toone of the phases.

The cyclo converter produces a three-

phase pseudo mains requiring filteringand with a frequency varying between0 % and a third of network frequency.

fig. 25: subsynchronous cascade power diagram.

M

P

Pr

3L

3L

3L


18/28


"Motor protection" groups all theprotection devices used to preventserious damage due to abnormaloperating conditions at supply, motor orprocess level.

The protection devices to be installedare chosen according to the followingcriteria:c operating conditions;c importance of the operationperformed by the motor;c the degree of dependability required;

c the relative cost of the protectiondevice with respect to the motor;c the likelihood of the faults consideredoccurring.

As well as:c the type of load driven;c disturbances which could occur onthe network;c the type of motor protected.

The faults listed below may thereforerequire use of a protection device.

main fault types

Asynchronous motors

c overloads;c short-circuits;c phase breaking, reversal andunbalance;c insulation fault between turns;c stator frame;c under and overvoltage;c incomplete starting.

Synchronous motors

All the above faults, plus:c loss of synchronism;c loss of excitation;

c rotor frame;c prolonged operation in asynchronousmode on starting;c overloads and short-circuits in theexciting winding;c reverse power (operation onAC generator).

Other faults linked to the process orload

c over frequent starts;c locked rotor;c underpower or undercurrent.

The detection/protection processes forthe main fault types are studied in theparagraph below.

protection principles

Overloads

Overloads can be detected by reverse

time overcurrent relays, thermal imagerelays or heat sensors.

The relays process the information

"current absorbed by the motor" whichis generally detected by current

transformers.

The heat sensors are inserted in the

live parts of the motor.

c reverse time overcurrent relaysTheir use requires:

v either an operating curve I(t) allowing

starting, or a device blocking the relay

during starting,v an operating threshold I0 close to therated current In of the motor

I0 1.10 In .

These relays do not memorise theoverloads.

c thermal image relays

These relays are certainly the mostsuitable since they ensure the greatest

possible advantage is derived frommotor overload possibilities without

damage.

The operating curve I(t) of the relay

must enable the starting current to flowwithout tripping and be approved by the

motor manufacturer.

c heat sensors

These are resistors, the ohmic value ofwhich varies with temperature.

In theory these devices are not used bythemselves, but rather back up therelays using the current absorbed as ameasuring means.

Overloads due to temperature rise of abearing are, in theory, insufficient fordetection by the overload relays.

Bearings must therefore be protectedby thermostats or heat sensors.

Short-circuits

Short-circuits on circuit-breakerequipment are detected byinstantaneous operation overcurrentrelays, set above the starting current.

On fuse-contactors, short-circuits arecleared by the fuses.

However, an useful solution is to addslightly time delayed overcurrent relaysto the fuses. This means that thecontactor can be used right up to itsbreaking capacity.

Phase breaking, reversal andunbalance

These faults are detected by a filterwhich highlights the negative phasesequence components.

It is vital to monitor phase breaking andunbalance since these faults give riseto:c increased current, in the stator,;c additional temperature rise by Jouleeffect, due to the fact that all out-of-balance states result in the appearanceof reverse currents flowing through the

rotor at twice supply frequency in therotor.

Phase reversal is detected either bycurrents or voltages:c by currents: this reversal is detectedafter contactor closing: the drivenmachine receives the fault;c by voltages: this means contactorclosing can be prohibited, if necessary,if phase order is not the normal networkone.

Insulation fault in winding

Stator windings may have faultsbetween turns on the same phase or

between windings of different phases.As a result of its electrical position, thefault may not be detected quicklyenough by the overload protectiondevice, thus causing serious damage.These faults are normally detected bycurrent comparison.

c longitudinal earth leakage protectionProvides protection against faultsbetween windings of different phases.For this, the ends of the motor windingsmust be accessible on the neutral side.

4. protection of HV motors


19/28


Faults are shown up by comparing theinput and output currents of the samephase (see fig. 26).If there are no faults, these currents are

identical and the protection relay is nottripped. It trips when the differencebetween these currents reaches avalue set by the relay setting.

c transverse earth leakage protectionProvides protection against faultsbetween turns of the same phase.It applies to machines with dividedphases, i.e. with two windings perphase.The operating principle is the same asabove, that is the currents of eachwinding are compared (see fig. 27).

Stator frame

This protection device is vital to complywith the statement of the 14.11.1988 onworkers' protection. It is chosenaccording to the earthing system of thenetwork supplying the motor.

c motor protection supplied by networkwith earthed or impedance-earthedneutral.The fault is detected by measuring thezero sequence current formed betweenthe faulty phase and the network frame.Low threshold overcurrent relays areused for measuring.The zero sequence current is deliveredby three parallel-connected currenttransformers or, better still, by a toroid(see fig. 28).Use of a toroid prevents theappearance of a false zero sequencecomponent due to unequal saturation ofcurrent transformers on motor starting,and allows a relatively low operatingthreshold.

These relays must operate for a faultcurrent value such that frame potentialcompared with earth is never raised tomore than 24 V in conductiveenvironments, with frames

interconnected, or 50 V in otherinstallation cases.

Knowledge of earth connection valueand of the frame interconnectionconfiguration is therefore necessary todetermine this setting point.If the frames are not interconnected,the value of the operating threshold isgiven by:

IF i 24 or 50 V

RTM.

fig. 28: diagram showing a stator frame zero sequence protection with toroid sensor or CT +

earthed or impedance-earthed neutral.

NM

protection

relay

fig. 26: diagram showing a longitudinal earth leakage protection.

M N

protection

relay

zero

sequence

current relay

toroid

CT

zero

sequence

current relay

eventually toroid

fig. 27: diagram showing a transverse earth leakage protection.


20/28


RTM

is the value of the earth resistance

of the frame in question.

Note that the lower the threshold, the

earlier the detection, and the smaller

the risk of damage to the magnetic

circuits.

c unearthed neutral

The fault is detected by continuous

measurement of global network

insulation compared to earth by means

of DC injection devices such as

continuous insulation monitors (see

diagram 1 in figure 29) or by zero

sequence overvoltage relays where this

overvoltage is delivered by three

voltage transformers with the

secondary winding in open delta (see

diagram 2 in figure 29).

Under and overvoltage

(see fig. 30)

c undervoltage

This relatively common protection

prevents the motor from working on

overload and waiting for tripping by the

overload protection device. Moreover, if

the contactor coil is supplied by an

auxiliary LV source not from the

network, the latch undervoltage

protection device is vital to prevent

untimely startup on restoration of

power.

The "voltage" information is provided by

a voltage transformer and processed by

a time-delayable threshold device.

c overvoltage

This protection is necessary if there is a

risk of strong fluctuations occurring on

the supply network. It means it is no

longer necessary to wait for the

overload relays to trip, since

overvoltage results in motor overcurrent

and increased motor torque which

could be detrimental for the machine

being driven.

The fault is detected by time delayed

overvoltage measuring relays.

Incomplete or overlong starting

This protection is justified for starting in

several stages.

It is provided by a time delay relay put

into operation at the beginning of

starting and removed at the end. The

value monitored may be speed or

current.

Prolonged use of the starting system,calculated to run for a set time, is thusavoided.

Loss of synchronismThis protection is vital for synchronousmotors.

In point of fact, the dampening cageof a synchronous motor is relativelyfragile. If the motor jams, inducedcurrents will flow through this cage andcould destroy it if the motor is notdisconnected.Jamming may occur following amechanical overload, undervoltage, aloss of or drop in excitation.

This fault is detected by underimpedance or power factor relayssupplied by voltage transformersand current transformers (see fig. 31).

fig. 29: insulation fault monitoring diagram with continuous insulation monitor or zero sequence

voltage relay - unearthed neutral

fig. 30: diagram showing under and

overvoltage protection. In most cases, two

VTs supply the relay phase-to-phase

voltages.

voltage

relay

M

zero sequence

voltage relay

DC injection

insulation

monitorsurge

voltage

limiterC

R = load resistance

R R R

diagram 1

diagram 2


21/28


directional power

relay

M

under

impedance

or cos relay

M

Loss of excitationThis loss due, for example, to a break inrotor winding, causes the motor to jam.It can be detected by:

c the "loss of synchronism" protectiondescribed above;c or an exciting undercurrent orundervoltage relay.

Rotor frame of a synchronous motor

This protection is determined accordingto the supply configuration and theDC production mode.

If the entire DC exciting circuit isunearthed, an insulation fault will notaffect motor operation.However, should a second fault occur,it may give rise to an overload or short-

circuit with all its consequences. Therelays used to detect this type of faultare generally devices injecting lowfrequency 10 Hz or 20 Hz AC current(see fig. 32).

The 50 Hz frequency is also used,which requires that no 50 Hzcomponents must be present in theexciting circuit.

Prolonged operation in theasynchronous mode on starting

Overlong starting on synchronousmotors causes excessive temperaturerise of the damping cage.

Either the "incomplete starting"protection described above is used or athermal device adapted to the rotorthermal time constant, placed in series

with the inductance during starting (seefig. 32 element b).

Overload and short-circuit in theexciting windingThese protection devices preventdamage due to temperature rise of theexciting winding and its power supply.The fault is detected by an excitingovercurrent relay. Moreover an excitingundervoltage relay is normally used,tripping on voltage drop, caused forexample by a short-circuit.

On starting, it is used as an excitingvoltage presence relay and authorisesclosing of the exciting contactor atthe end of starting in the

asynchronous mode (see fig. 32element g).

Reverse power

This protection particularly applies tosynchronous motors.

When the supply circuit-breaker trips, itprevents energy from being sent backto the loads connected on the samebusbar. It also prevents a fault on thisbusbar from being supplied by themotor.

The protection device has to detect areversal in current or power direction.

Its function is therefore performed by adirectional power relay (see fig. 33).

Frequent starts

Too many starts over a given period oftime may cause motor damage if themotor is not designed for such anoperation.

This protection is provided by a relayperforming metering and time delayfunctions, which automatically limits:c either the number of starts over agiven period of time;c or the time interval between start upsover a given period of time.

fig. 31: protection against loss of

synchronism.

fig. 32: rotor protection devices on startup

and in operation.

a. exciting windingb. thermal protection: prolonged operation inasynchronous modec. starting resistanced. exciting contactore. rotor frame protectionf. overcurrent protectiong. undervoltage protectionh. to tripping of motor feeder contactors. continuous source

fig. 33: protection against reverse power.

ab

c

d

e

f

g

h

s

Z

U


22/28


Locked rotor

Jamming of a motor for mechanicalreasons causes an overcurrent roughlyequivalent to starting current. The

resulting temperature rise isconsiderably greater since rotor lossesare maintained at their highest valuethroughout jamming and ventilation isno longer present if connected to rotorrotation. As a result, when there is arisk of jamming, the "locked rotor"protection is necessary since overloadrelays sometimes have an excessivelylong response time.

This fault is detected by an overcurrentrelay set at a value less than thestarting current. This value is validated

after a time delay initiated when themotor is energised. This time delay isset at a value greater than or equal to

normal starting time.

Undercurrent or underpower

Pumps may be damaged when

unpriming occurs. Unpriming could also

result in a drop in the active power

absorbed by the motor. Protection

against this fault is provided by an

undercurrent relay.

technological evolutionThe term "relay" is often used in the

above description of the various

protection devices. This is a thetraditionally used term (relay = a type ofprotection device) which goes back tothe time when motor protection

required the use of separate "relays"each with a single protection function.

In the seventies, manufacturers broughtout RACKS able to house a number ofdifferent protection devices in order tomeet needs more closely.

In the eighties digital technologyincreased adaptation possibilities stillfurther. Thus a single programmabledevice performs the various protectionand control/monitoring functionsrequired in each specific case.


23/28


overall approachThe approach chosen by the designeris to look for the best technico-economic choice. For this, first try andvalidate the easiest and mosteconomical solution. Then, if this doesnot work, follow the order shown in thetable in figure 34.

Direct starting

Motor apparent power at the beginningof starting:

Sm = Pn

cos Id

In

Sm = 1,500

0.845 = 8,925kVA

where, with power factor on starting:cos

d= 0.15, i.e. :

d= 81.

This power is vectorially added to thepower delivered by the transformer onthe other feeders (see fig. 35).

The total value of the apparent powerrequired by the transformer isgraphically deduced from the following:

S 9,580 kVA.The maximum inrush authorised is6,000 kVA: direct starting is thus notpossible.

Starting by reactance

Introduction of a reactance reduces theapparent power absorbed by the motor.The available starting power isgraphically determined (see fig. 35).

When the motor starts, the presence ofthe reactance means that the powerfactor approaches zero. Thus

d 90.

appendix 1: determining motor starting mode

The purpose of the example below isnot to fully deal with a problem butrather to illustrate, through an example,an approach wich helps to choose theappropriate starting mode for a givenapplication.

calculation hypothesesAsynchronous motor with:c rated power P

n= 1,500 kW;

c rated voltage Un

= 5,500 V;

c efficiency x power factor: x cos = 0.84;c starting torque/rated torque ratio onfull voltage:

CdCn

= 0.8

c starting current/rated current ratio onfull voltage:

Id

In

= 5 ;

c breakaway torque of the drivenmachine: 0.2 C

n;

c main supply transformer power:Pt = 3 MVA;c maximum apparent power inrushauthorised by the transformer network:S

t= 6 MVA.

Other data necessary for calculation:c torque-speed C (N) characteristic ofthe motor;c load torque-speed C

r(N);

characteristic of the driven machine

c transformer delivery on feeders otherthan the motor one: 1,200 kVA undercos ' = 0.87.

OA

= 1,200 kVA : transformer power

used on other feeders.

OB

= St = 6,000kVA: maximumapparent power authorised. Theapparent power available for starting(motor + reactance) is graphicallydeduced.

AB

= Sd = 5,300kVA.Power reduction to be caused by the

reactance:SdSm

= 5,3008,925

0.6 .

fig. 34: decisive criteria for choosing motor starting mode.

starting solutions main acceptance criteria

direct power inrush compatible with network

reactance c starting torque greater than breakawaytorquec current peak (for full voltageresumption) acceptable by network

autotransformer the same

fig. 35: starting power diagram.

S = 9,580*

B

d= 81

' = 29.51,200*

St= 6,000*

Sd

= 5,300*

Sm

= 8,925*

A

active power*: in kVA


24/28


(a)

(c)

(d)

(b)

(e)

0 0.1 0.2 0.5 0.8 1

0.2

0.8

5

1

1

Cn

In

point of balance

at reduced

voltage

current peak

on move

to full voltage

(N/Ns)

I'd is the new starting current value.

Sd = Un I'd 3

Sm

= Un

Id

3

thus: I'd = 0.6 Id.

Moroever:I'd

Id

= UdUn

= 0.6 .

The value of voltage at the motorterminals is U

d= 0.6 U

n. Once the

electrical problem has been solved, itremains to be seen if this solution isvalid from the mechanical standpoint. Inthe case of direct starting, the startingtorque equals : C

d = 8 Cn (see fig. 36).

In the case of starting by reactance, thestarting torque C'

dequals:

C'd = 0.8 Cn UdUn

2

= 0.288 Cn

This value is compatible with thebreakaway torque of the machine beingdriven.

One last point remains to be checked:if the point of mechanical balanceC

m= C

ris placed at an excessively low

speed, a current peak may occur on themove to full voltage. If this peak is toohigh for the network, the starting modewill have to be reconsidered andstarting by autotransformer, for

example, chosen (see fig. 36).Remark 1Let us assume that the value of thebreakaway torque of the drivenmachine is 0.35 C

ninstead of 0.2 C

n.

Starting by reactance is thenincompatible with the breakawaytorque.The solution of starting byautotransformer must then beenvisaged.

The apparent power available remainsS

d= 5,300 kVA. From this value the

magnetising force of the auto-

transformer Smg must be deducedwhich, at the initial starting moment, isarithmetically added to motor apparentpower. S

mgis around 0.2 to 0.4 times

motor apparent rated power.

i.e. with the factor 0.4:

Smg = 0.4Pn

cos

Smg = 0.41,500

0.84 = 720 kVA .

The power reduction factor is then:

Sd SmgSm

= 5,300 7208,925

= 0.513

At constant voltage Un, the current

inrush on the network side is therefore:0.513 Id.

Determining the reduced startingvoltage U

d.

Equality of primary and secondaryautotransformer powers results in thefollowing:

0.513 Id Un = I"d UdI"

dis the reduced voltage starting

current on the motor side

0.513 Id Un = Id UdUn

Ud

hence:UdUn

2

= 0.513

i.e. Ud = 0.718 Un

New starting torque:

0.8 Cn UdUn

2

= 0.41 Cn .

This value is sufficient for starting totake place.

Remark 2

For a breakaway torque greater than0.41 C

n, stator starting with this motor

is no longer possible.

Either a slipring rotor motor with rotor

starting must be used, or a motor with aspecial cage possessing a high startingtorque.

fig. 36: torque and current curves for starting by reactance.

a: curve C (N) on full voltage

b: curve C (N) on reduced voltage (0.6 Un)

c: curve I (N) on full voltage

d: curve I (N) on reduced voltage (0.6 Un)

e: curve Cr (N)

C

Cn

I

In


25/28


appendix 2: coordination of protection devices

Once the protection devices have beenchosen according to operatingrequirements, they must be coordinatedto obtain the maximum advantage fromtheir possibilities. A balance must besought between untimely tripping anddelay on fault removal. The problem ofprotection device coordination is solvedby studying the curves t (I) of therelays, of the circuit-breaker and of thebreaking capacity of the contactor. Thecharacteristics of the motor shown inthe figure are as follows:

c Pn

= 550 kW;c U

n= 3150 V;

cIn

= 130 A;cId = 5 In.

The contactor is of the Rollarc fuse type.

Type of protection devices

c thermal relay with indirect tripping,set at In = 130 A for overloads;c positive phase sequence componentrelay set at 6 I

n, time delayed at 0.05 s

for balanced faults;c negative phase sequence componentrelay set at 0.3 or 0.4 In, time delayed

at 0.6 s.In the case of networks with quasi-permanent unbalances, a two-thresholdrelay is used:c a lower time delayed threshold, setjust above the permanently acceptednegative phase sequence componentratios;c an upper instantaneous threshold forphase break protection.

Instantaneous tripping by the positivephase sequence component relayenables best use to be made of thecontactors breaking capacity and

avoids fuse melting. Analysis of thecurves in figure 37 shows that themotor and network are protectedagainst:c unbalances of approximately 0.3 I

nto 10 I

n,

c balanced faults from 6 In to 28 In.

The fuses only take action beyond 15 Infor unbalanced faults and 25 I

nfor

balanced faults.

The maximum current the contactormay have to break is 28 In = 3,640 A.This value is considerably lower than itsbreaking capacity of 10 kA.

fig. 37: minutes thermal relay operating zone.

1 2 5 10 20 50 100

130 260 650 1,300 2,600 6,500

0.01

0.03

0.05

0.1

1

5

10

20

40

1

2

34

68

10

20

30

40

60

cold

hot

fuse

phase sequence

component:

negative

positive

contactor

breaking

time

minutes

seconds

Rollarc contactor

breaking capacity

thermal relayoperating zone

3,640 A

k InIn (A)


26/28


appendix 3: bibliography

Standards

c IEC 34-1: Rotating electricalmachines.

c IEC 71: Insulation coordination.

Merlin Gerin's Cahier Technique

c The breaking process with a FluarcSF6 puffer-type circuit-breaker,Cahier Technique n 112 -J. HENNEBERT

c Behaviour of the SF6 MV circuit-breakers Fluarc for switching motor

starting currents,Cahier Technique n 143 -J. HENNEBERT and D. GIBBS

c Overvoltages and insulationcoordination in MV and HV,Cahier Technique n 151 -D. FULCHIRON


27/28



28/28

Documents

protection of HV motor