Optimum Motor Protection en - LIBRO

8/12/2019 Optimum Motor Protection en - LIBRO

1/80


2/80


3/80

Contents Page

Introduction to the

Principles of Asynchronousand Synchronous Motors 3

Overview of Protection

Functions and Protection

Relays for Motors 11

Thermal Stress of

Motors and Necessary

Protection Functions 17

Motor Protection:

Requirements for

Current Transformers 25

Protection of

Low-Power Motors 31

Protection of

Medium-Power Motors 43

Protection of

High-Power Asynchronous Motors 51

Protection ofSynchronous Motors 63

Synchronization and Protection

of Synchronous Motors

with the 7UM62 69

Appendix 73

Optimum Motor ProtectionwithSIPROTECProtection Relays

Siemens AG 2006


4/80


5/80

Siemens PTDEA OptimumMotorProtectionwith SIPROTECProtection Relays 3

Introduction

This chapter is an introduction to three-phasemotors. On the basis of fundamental physicalrelationships and their structure, it is particularlythe characteristic operating modes of motorswhich are discussed. Essential mathematical rela-tionships, as are needed when deriving character-istic quantities of motors or for protection

settings, are also presented.

1. Basic principle of the three-phase machine

Electric machines convert electrical energy intomechanical energy. The energy conversion proc-ess is based on the interaction of magnetic fieldsand windings. In its electrically active part, a ro-tary electric machine consists of a stator with arotating field winding and the rotor. The rotor isconnected to the shaft, through which mechanicalenergy is dissipated (in the case of the motor). Arotating magnetic field (rotating field) is the fun-damental prerequisite for the operation of three-phase machines. This is generated by connecting a

three-phase winding to a symmetrical three-phasesystem. For a sinusoidal profile of the rotatingfield, the three phases must be placed at a 120offset and there must be a 120 offset between theindividual phase currents. The rotating fields ro-tation speed is called the synchronous speednsynand is defined by the machines number of polesand the frequency of the supplying power system.The following applies:

nsyn =f. 60/p [rpm]

where: p = number of pole pairs

Operating principle of the asynchronous machineA voltage is induced in the rotors conductors ifthere is a difference between the rotation speed ofthe rotating field and the rotor (induction law). Ifthe conductors are part of a closed winding, theinduced voltage produces a flow of current. Thisleads to tangential forces according to the forceeffect of a current-carrying conductor in the mag-netic field. The sum of all tangential forces gener-ates a resulting torque over the rotors radiusacting as a fulcrum.

According to Lenzs Rule, the force effect betweencurrents and the rotating field is oriented such

that it counteracts the cause of induction. If therotating field of the machine still at standstill runsbeyond the rotor, it begins to turn in the rotatingfields direction to reduce the relative speed be-

tween the rotating field and the rotor. The fre-quency (speed) of the rotor can never reach therotating frequency of the stator, because a voltageis then no longer induced in the rotor and so theforce effect becomes zero. The deviation betweenthe frequency of the rotating field and the fre-quency of the rotor is referred to as slips. Slip setsin so that a rotor current that is still just adequatefor the motors load arises.

s f f

f

n n

n= = =

rf r

rf

rf r

rf

syn r

nsyn

(1)

rf angular frequency of the stator's rotatingfield ( rf= S)

r angular frequency of the rotorf frequency

n speed

The resulting speed of the motor (rotor) resultsfrom the number of pole pairs p (p =1 signifies 2poles, as shown in Fig. 1) and the slip.

nr=nsyn(1 s) (2)

Introduction to the Principles

of Asynchronous andSynchronous Motors

Fig. 1 Principal components of the three-phase asynchronousmachine (U, V, W: conductor terminals UL1,UL2 and UL3)


6/80

Siemens PTDEA OptimumMotor Protection withSIPROTECProtection Relays4

Introduction

It was the imbalance between the frequencies ofthe rotating field and the rotor that led to thenameasynchronous machine. Due to the voltage

induced, the term induction machine is fre-quently used, too.

Structure of an asynchronous machineFig. 1 shows, in a longitudinal section and across-section, the principal components of athree-phase asynchronous machine with a slip-ring rotor. The number of pole pairs is p = 1.

StatorThe stator core secured on the housing accom-modates the isolated three-phase stator windingin mostly half-closed slots. The stator core con-

sists of mutually isolated electrical steel sheets.Lamination of the stator and rotor is necessary tominimize the eddy current losses arising due tothe alternating magnetic fields.

The three phases of the stator winding are con-nected in either star or delta form. Therefore it ispossible to operate the machines with differentrated voltages. Thus, machines in a star connec-tion operated at 400 V can also be operated at230 V with unchanged power output by changingover to a delta connection. Star-delta starting isalso possible if machines are operated at ratedvoltage in a delta connection.

RotorThe three-phase rotor core, which is also isolated,is housed in half or fully closed slots of the rotorcore. The ends of the rotor winding phases areconnected in a star point. The starts of the wind-ings are routed to sliprings so that the windingcan be shorted directly or with the aid of resistors.

Contrary to the slipring rotor, the winding of thesquirrel-cage motor consists of simple conductorbars without additional insulation. These conduc-tor bars are distributed concentrically around theshaft on the rotors circumference and are shortedby rings at the face ends of the rotor core. The

resulting winding cage led to the name of squirrel-cage rotor. The starting response of the motor(see also Section 3) can be influenced by specialcross-sectional shapes of the rotor bars (see Fig. 2)or by using two cages (double squirrel-cage rotor).

With a view to minimizing the magnetizationcurrent as far as possible, the air gap between thestator and rotor cores must be kept very small. Asall slot openings along the rotor or stator circum-ference act like an enlargement of the air gap andlead to distortions of the air gaps field (voltagedistortions and additional losses), the slots aregenerally half-closed, to some extent entirely

closed and only designed as open slots in the caseof high-voltage machines.

Operating principle of the synchronous machineIn the case of thesynchronous machine, therotating field and rotor frequencies are identical

and so the slip is equal to zero. Both rotatingfields run synchronously. As already explainedabove, the principle of the induction machinetherefore fails. The magnetic field required forenergy turnover is generated with the aid of anexcitation winding. In the case of electrically ex-cited synchronous machines, an adjustable directcurrent is needed for the excitation winding. Thisis supplied with the aid of an excitation system.In the case of brushless excitation, as takes placeespecially in large synchronous machines, theexcitation power is fed by means of an excitationmachine fitted on the motors shaft and with asynchronously rotating three-phase bridge circuit.This directly generates the excitation voltage.As an alternative, the excitation power can alsobe fed via sliprings.

One positive characteristic of the synchronousmachine is the possibility of synchronous com-pensator operation. A reactive power can begenerated in a required form, thus improving apower systems power factor (cos ).

Fig. 2 Bar shapes of a squirrel-cage rotora) Roundbar rotorb) Squirrel-cage rotor with parallel-flanked teeth

for cast cagesc) Deep-bar rotor with rectangular barsd) Deep-bar rotor with tapered barse) Double squirrel-cage rotor with roundbarsf) Double squirrel-cage rotor with parallel-flanked

teeth for cast cages


7/80


Introduction

2. Equivalent circuit diagram of theasynchronous machine and mathematicalrelationships

Essential characteristic quantities are to be de-rived with reference to the example of the equiva-lent circuit diagram. We can find analogies to thetransformer, and so the asynchronous motorsequivalent circuit diagram (see Fig. 3) can bederived from that of the transformer. The rotorvariables are converted to the variables of thestator via magnetic coupling. Similarly to thetransformer, this is done with the square of thenumber of turns per unit length of the rotor andstator windings (see equation (3)). To simplifymatters, the iron losses are ignored.

R W

WR'r

S

rr=

2(3)

It becomes clear that the effective rotor resistanceis low during starting (s = 1). The high startingcurrent with a low cos is the result of this. Dur-ing normal operation, at s

0 the effective resist-ance is high and the cos approaches the value 1.

The necessary mathematical relationships areshown below. Equation (4) shows the stator cur-rent as the sum of the rotor and magnetizingcurrents (Im). If we neglect the idle current, thestator current is approximately equal to thetransformed rotor current.

IS m r mrf m

mr

rf r

S

Sr

S

= = +

+ + +

I I V

j L

V

R

sj L

V

R R

sj X X

' ''

'

( )'r(4)

The power factor results from the real portion ofthe rotor current in equation (4) and leads to therelationship in equation (5).

( )

cos

'

' '

r

r

rr

=

+

R

s

Rs

X

2

2

(5)

A further characteristic quantity is the machinestorque, which can be derived from the power bal-ance. The mechanical power given offPmechis the

result of the difference of the electrical effectivepowerPssupplied minus the losses in the statorPloss-sand rotorPloss-rof the machine.

P P P P mech s loss- s loss-r= (6)

The following applies if we neglect the statorlosses and introduce the air gap powerP:

P P Pmech loss-r= (7)

The mechanical power can be expressed with theaid of the torque M and the mechanical angular

velocitymechof the rotor.( )P M M

psmech mech

rf= =

1 (8)

By analogy, the torque can be calculated fromboth the air gap power and the rotors dissipatedpower using the following equation.

M P

p= rf

or M P p

s= loss-r

rf (9)

After neglecting the stator losses, we can calculatethe rotating fields power from the stator voltage(phase voltage) and the real portion of the statorcurrent which, for a three-phase machine, results

in the following:

( )P U e I = 3 S SRwhere

( )( )

Re SS

r

Sr

S r

IU

R

R R

X X

=

+

+ +

'

''

s

s

22

(10)

Inserted in equation (9), equation (10) leads tothe final relationship for the machines torque

( )M=

+

+ +

=

+

3

3 1

22

pU

R

s

R R

sX X

pU

R

S

2

rf

r

Sr

S r

S2

rf S2

'

''

( )X XR

s R

sR

S r

r

rS

++ +

'

'

'2

2

(11)

By derivation of the denominator of equation(11), which is set to zero, we can calculate the slipat which the torque reaches the extreme value.This slip is referred to as the pull-out slipspull.Consequently, the torque developing with this

slip is the pull-out torque.

Fig. 3 Equivalent circuit diagram of the asynchronousmotor


8/80


Introduction

( )s

R

R X Xpull

r

S2

S r

=+ +

'

' 2

(12)

( )M

pU

R X X Rpull

S2

rfS2

S r S

=+ + +

3

2

12 '

(13)

If we relate equations (11) and (13) and if we ne-glect the ohmic stator resistance, we arrive at theso-called Kloss relationship. This formula(equation (14)) describes the torque-slip charac-teristics for the entire slip range (-< s < +).The relationship for the pull-out slip is also sim-plified.

MM s

s

s

spull

pull

pull=

+2 wheres R

X Xpull r

S r'= + '

(14)

Thus, we have derived essential mathematical re-lationships for the operating variables of the asyn-chronous motor. These help when interpretingthe curves and data supplied by the manufacturerof the motor. In quality terms, Fig. 4 shows thecourse of the characteristic quantities. The typicalcharacteristics are also plotted.

Equation (4) shows that the starting current de-pends on the applied voltage and particularly onthe resistance of the rotor. The rotors equivalent

resistance is designed low so as to keep losses inthe rotor likewise low. This leads to relatively highstarting currents, which can certainly reach valuesof five to eight times the rated current. As demon-strated in equation (11), the low equivalent resist-ance of the rotor limits the starting torque.

There are limits to reducing the starting currentby decreasing the stator voltage. Equation (11)clearly shows the square dependence of the torqueon the applied stator voltage (M V2), whichleads to a drastic reduction in the starting torque(half the terminal voltage leads to a quarter of theoriginal torque).

Increasing the rotor resistance by means of start-ing resistors (slipring rotor motor as shown inFig. 1) leads to reduction of the starting current.As a result, the cos and thus the starting torqueincrease. During normal operation, the resistorsare shorted. The drawback of this solution is thegreater effort involved.

From the square voltage dependence of the torquewe can also derive that, as demonstrated in equa-tion (13), the pull-out torque is also clearly re-duced. Thus, when operating a motor with areduced voltage, we run the risk of the requiredmechanical torque exceeding the pull-out torque,

resulting in a possibility that the motor will liter-ally stand still.

Fig. 4 Operating quantities of the asynchronous motoras a function of the rotors speed or the slip(Mst starting torque, Mpull pull-out torque; MN ratedtorque, npull pull-out speed,nsyn = nrf speed of thestators rotating field,spull pull-out slip,sN ratedslip)


9/80


Introduction

3. Startup of asynchronous motors

From the point of view of the power system, ap-propriate proof must be provided of reliable start-

ing (from standstill) and restarting (from aparticular speed up to rated speed) in the event ofpossible residual magnetism of the magnetic field.The following impacts are possible if the startupconditions are not met:

Extreme delaying of startup and thus high ther-mal stressing of the motors and the upstreampower system elements

Standstill of the drive Tripping of the motor protection and possibly

tripping of the upstream protection Thermal stressing of all elements in the relevant

conducting path

Synchronism loss of motors in operation downto standstill

As an extension of Fig. 4, Fig. 5 shows the enve-lope curve of the starting current. Similarly toactivation of an inductance, on activation ofthe motor with the slips= 1 a transient inrushcurrent arises which then changes over to thequasi-stationary starting current. This current ini-tially drops very slowly and does not decrease fastto the normal operating current until near therated slip.

The starting factor (kst = 3 610) is often spec-ified to characterize starting of machines. It essen-tially depends on the design (see Section 2 andFig. 2). The restart factor (kr-st= 1310) de-

pends not only on the design, but considerablyalso on the speed and the still remaining magneticfield at the time of reactivation.

kststart

N

=I

Ior kr-st

restart

N

=I

I(15)

Fig. 6 shows the dependence of the restart factoron the slip by way of example with reference toasynchronous motors with different starting fac-

tors.

For every drive, the starting or restarting condi-tion is met if, at any time, the motors torqueMisgreater than the resistance torqueMr(opposingtorque of the machine, which is the sum of frictionmoments and other losses). A safety factor of 10 %is reckoned with. The required minimum voltagecan be determined by applying equation (16).

This lies in the order of magnitude ofVmin=(0.55 0.7)VN, M.

V Vn

n

M

Mmin N,M

pull

N

N

pull

= 11.

(16)

where:

VN,M rated motor voltageMN rated resistance torqueMpull pull-out torquenpull pull-out speednN rated speed

Restart factor kr-st asa functionof slip and fordifferent starting factors kst

Envelope curve of the starting current


10/80


Introduction

An equivalent power system circuit, an exampleof which is shown in Fig. 7, is needed to calculatethe terminal voltages.

The equivalent power system impedanceZNiscalculated on the basis of the short-circuit power,the transformer impedanceZTfrom the trans-formers rated data, the cable impedanceZcfrom the specific cable parameters and themotor impedanceZMfrom the starting factor kst(see equation 17).

( )Z ZM M A= +cos sin j(17)

whereZ V

SM

N,M2

st N, Mk=

where:VN,M rated motor voltageSN,M rated motor apparent power

For high-voltage motors, a phase angle of 90 canbe used for calculation to simplify matters, so that

ZMjXM is set.

4. Operating modes of asynchronous motors

The mechanical and thermal stresses depend onthe load state of motors. Duringcontinuous opera-tion, the temperature rise must not exceed thepermissible overtemperature limit for the insula-tion system used. It is obvious that a motor maybe subjected to a higher load for a certain periodof time without the winding overtemperatureexceeding its permissible value. After that, it isimperative to take a break for cooling downbefore the motor is loaded again. In IEC 60034,such operating cases have been standardized inan idealized form as so-called duty typesS1 toS10.

S1 Continuous running dutyS2 Short-time duty (e.g. S2 60 min)S3 Intermittent periodic duty

(operation with similar loading cycles)S4 Intermittent periodic duty with startingS5 Intermittent periodic duty with electric

brakingS6 Continuous operation periodic duty

S7 Continuous operation periodic duty withelectric braking

S8 Continuous operation periodic duty with

related load/speed changesS9 Duty with non-periodic load/speedvariations

S10 Duty with discrete constant loads and speeds

The predominant duty type of motors is continu-ous duty (approximately 90 % of all applications).

Fig. 8 shows a selection of duty types. Besides theovertemperature curves, it also shows the lossesand the speed, with the result that the characteris-tics of the individual duty types can be derivedfrom the depictions alone.

5.

Equation (18) describes the power for stable oper-ation of a synchronous three-phase motor.

P PEV

XV

X X

X Xmech syn d

d q

d q

= = +

3 3

222sin sin

where:

Pmech mechanical power outputPsyn active power consumed electrically

E rotor voltageV terminal voltage (phase voltage)

Xd synchronous direct-axis reactanceXq synchronous quadrature-axis reactance rotor angle

Equivalent power system circuit on activation of a motor

LossesPloss, speedn and the overtemperature forthe dutytypes S1, S2, S3 andS6

(TB loading duration, TP pause duration, Tdcduty cycle,Ti idle duration)


11/80


Introduction

If, in accordance with equation (18), the terminalvoltage or the excitation voltage is reduced at agiven constant mechanical power, the rotor angle

increases until the motor falls out of step whenthe limit angle ( > 90atXd=Xq) is exceeded.The motor falls out of step, no longer runs atsynchronous speed.

If the power system is undisturbed, this case isencountered whenever the excitation is too lowfor the required mechanical power or fails com-pletely. In this case we speak ofloss of synchro-nism due to underexcitation. However, loss ofsynchronism can also occur as a result of a powersystem fault in which the terminal voltage dropsfor a short time or disappears entirely and thenreturns. When the voltage returns despite full ex-

citation, the synchronizing torque cannot sufficeto re-establish synchronism of the motor. In thiscase we speak ofloss of synchronism due to a

power system fault.

We can distinguish between two possibilitieswhen a motor is not running synchronously:

1. Motor operation is stable

with the excitation circuit closed like an asyn-chronous motor with two rotor windings (exci-tation and damper circuit) in the longitudinalaxis (double-cage characteristic), but only onerotor winding (damper circuit) in the trans-

verse axis (single-cage characteristic) with the excitation circuit open like anasynchronous motor with one rotor winding(damper circuit) with a single-cagecharacteristic

2. The pull-out point for stable asynchronousrunning is exceeded and the motor is braked de-pending on the torque characteristic of the loadmachine. It slips and generally comes to a stand-still.

Equation (18) no longer applies to mathematicaltreatment of non-synchronous operation.Instead, the complete equation of motion must be

applied for an oscillatory machine.

Below, the response of motors to the two failuremechanisms is discussed on the basis of a dy-namic power system calculation program:

a) Loss of synchronismdue to underexcitationIf the excitation voltage of a fully loaded synchro-nous motor is deactivated, synchronous running

is no longer guaranteed and the motor loses syn-chronism.

Fig. 9 shows the measured values in the admit-tance diagram. Due to the adequately high asyn-chronous pull-out torque, the motor is capable oflargely stable output of the required mechanicalpower as asynchronous power with low slip. Bycontrast, the differing reactance values in the lon-gitudinal and transverse axes result in consider-able oscillations of the conductance andsusceptance or the active and reactive power. In

this specific case, the speed of the motor oscillatesaround an average of 0.99nsynwith an amplitudeof 0.02nsyn.

Oscillation is repeated cyclically and the pull-outlimit of the asynchronous motor is not ex-ceeded. In all cases, the susceptance remainshigher than 1/Xd.

Excitation failure (shorted excitation circuit) wasalso simulated on a second motor with other data(see Fig. 10).

Admittance in the event of excitation failure(2.22 MVA motor)

Fig. 10 Admittance in the event of excitation failure(9.88 MVA motor)


12/80


Introduction

In this case the asynchronous pull-out torque isexceeded and the motor comes to a standstill at aspeed of about dn/dt= 7 %/s as the slip rapidly

increases. The rotor angle change rises from600/s to ever increasing values.

Due to differing input reactance values in the lon-gitudinal and transverse axes, the admittancemoves almost in circles. The circle diameters be-come increasingly smaller because, in terms oftheir synchronous values, these reactance valuesdecrease towards their sub-transient values with in-creasing slip. The final point of this phenomenon isreached at a conductance of almost zero and asusceptance of approximately 2/(Xd" +Xq").

b) Loss of synchronism due to short-time

power system voltage failureIn the event of a three-phase power systemshort-circuit and a related breakdown of thepower system voltage, a synchronous motor canno longer obtain the mechanical power requiredby the load as electrical power from the powersystem. The lacking power is covered by the(negative) acceleration power. As this is propor-tional to the change in the rotor angles rate ofrise (d2/dt2), a fast rotor angle change is en-forced. In this case, the rotor angle is understoodto be the angle between the rotor voltage and thepower system voltage at an unchanging fre-quency. After short-circuit tripping and the re-

lated return of voltage, either resynchronizationor a loss of synchronism can occur depending onthe motor constant, the mechanical load and theshort-circuit duration.

Fig. 11 shows the circle diagram, with the motorjust not having lost synchronism. Although thespeed drops to 5 % and the rotor angle has almostreached 180 in the time without voltage, the syn-chronizing torque after the voltage returns is highenough for the motor to resynchronize.

Extension of the short-circuit duration by 50 msleads to a greater speed drop (8 %) and the motorloses synchronism (see Fig. 12). Similarly to the

underexcitation shown in Fig. 10, the circle dia-gram shows circular motions around the point1/Xd". As the full excitation is effective, the circlediameters are accordingly larger. Here also, thesusceptance always remains higher than 1/Xd.

c) Conclusion

The examples clearly show that loss of synchro-nism of synchronous motors can be coped withvery well by means of underexcitation protectionwith an admittance characteristic (see 7UM61 or

7UM62 manuals). Only one protection character-istic is needed, which must be set to 1/Xd. An ade-quate delay should be set to give the motors anopportunity to resynchronize. During simulationruns, this resynchronization took no more than0.5 s in the worst case of motor stressing at ratedload. Therefore, a tripping delay of 1 s is recom-mended for loss of synchronism.

As the motor terminal voltage changes onlyslightly due to the mostly low input reactance,reactive power monitoring can also be used as analternative.

References:[1] Wenigk, K.-D.: Kraftwerkselektrotechnik.

VDE-Verlag GmbH, Berlin, Offenbach, 1993[2] Brendler, W.; Roseburg, D.: Elektrische

Maschinen. 4. Lehrbrief frs Hochschul-fernstudium in der ehemaligen DDR, 1988

[3] Mller, G.: Grundlagen elektrischerMaschinen. VCH VerlagsgesellschaftmbH, Weinheim, 1994

[4] Clemens, H.; Rothe, K.: Schutztechnik inElektroenergiesystemen. Verlag TechnikGmbH, Berlin, 1991

[5] IEC 60034-1 Rotating electrical machines.Part 1: Rating and performance, Release 2004

[6] Fischer, A.: Auertrittfall und Auer-trittfallschutz von Synchronmotoren.Etz Bd. 111 (1990) H. 6, S. 306 309

Fig. 11 Circlediagram of the admittance dueto a three-pole short-circuit lasting 0.15s(2.22 MVA motor)1 Initial value

2 Value during the short-circuit3 Value after short-circuit tripping

Circlediagram of the admittance dueto athree-pole short-circuit lasting 0.2 s (9.88 MVA motor)a) Power circle diagramsb) Admittance circle diagrams1 Initial valus2 Value at the startof the short-circuit3 Value on short-circuit tripping4 Value after short-circuit tripping


13/80


Protection Functions and Protection Relays

In this chapter, suitable protection functions arediscussed on the basis of possible faults in asyn-chronous and synchronous motors. An overviewfollows, showing how protection functions aredistributed over individual protection relays. Toconclude, the basic use of protection facilities forthe various motor types and power classes is

discussed.

1. Faults and protection functions

With motors, various causes of faults are possible.Table 1 shows an overview of essential causes offaults and their influences. For both the stator andthe rotor, thermal overloading plays a centralrole. Such inadmissible stressing has a general in-fluence on the useful life of a motor and leads to

premature aging. This, in turn, has a detrimentalinfluence on insulation capacity and simulta-neously increases the probability of follow-upfaults which can consist of earth faults and shortcircuits.

Cause Fault type

Thermaloverload

Interwindingfault(2-pole, 3-pole)

Earth fault Interturnfault

Mechanicaldestruction

Insulation aging

Stationary and transient

overvoltages

Undervoltage

Asymmetrical voltage

Conductor discontinuity,phase failure

Asynchronous changeover

Bearing damage

Mechanical overload duringcontinuous operation

Inadmissibly long starting time

Rotor locking during starting

Inadmissibly short pause time,too frequent reclosing

Soiling

Inadmissibly high ambienttemperature

Cooling failure

Overview of causes of faults and types of faults in asynchronous motors

Overview of

Protection Functions andProtection Relays for Motors


14/80



Fault Protection functions ANSI

Thermal overloading of the stator due toovercurrent, cooling problems, soiling, etc.

Thermal overload protection with completememory

Measurement of stator temperature bytemperature sensor (e.g. PT100)

49

Thermal overloading of the rotor during starting Too long, rotor locked Too frequent

Overload protection by two principles Starting time supervision Restart inhibit

4866, 49R

Voltage asymmetry, phase failure Negative-sequence protection(definite-time; thermal stage)

46

Overloading and thus overtemperature of thebearings

Measurement of bearing temperature bytemperature sensor (e.g. PT100)

38

Overstressing of drives running idle(pumps or compressors)

Minimum power consumption UndercurrentI< Active powerP

46

Thermal unbalanced-loadprotection (I2

2 t)46

Temperature monitoring

(via RTD-box)

38

1)

Undercurrent protection 37

Active power protection (P


16/80



Scope of protection functions provided by theprotection relays

In particular, numerical protection is noted for its

multiple functionality. Depending on the applica-tion, various protection functions can run on oneitem of standard hardware. Regardless of whethera cable, an overhead line, a generator or a motoris being protected, we are always dealing with thesame device type. This simplifies engineering bymeans of an identical amount of hardware. Appli-cations can be standardized and the need forspare parts can also be reduced.

In addition to choosing the hardware, users mustorder the appropriate combination of protectionfunctions.

Table 3 provides an overview of the functions formotor protection applications integrated in theprotection units and a selection of additionalfunctions is listed.

3. Use of the protection relays

The scope of protection to be applied dependsessentially on the rated power of the motor, itsmode of operation and its role and importance interms of the technological process linked to it. Itgoes without saying that costs also have to be con-sidered. Total costs should be considered because,to some extent, the costs of repairing a motor arerelatively low in comparison with the repair costs

of other electrical equipment. In case of a failure,however, the follow-up costs for the productionsystem often amount to a multiple of the repaircosts.

The overview in Fig. 1 shows a breakdown ofprotection units according to power classes ofmotors and applies to asynchronous motors.The preferred types of protection relays are listedin the left-hand column.

If additional functions are required, for examplefreely programmable logic, more scope for serialcommunication, a wide frequency operating

range and much more, refer to the units in theright-hand column. As the majority of protectionunits are installed directly in the medium-voltagecubicle, increased use is being made of controlfunction. Here, the relays with graphics displayoffer advantages in the form of illustration of afeeder with the switching devices. The conven-tional mimic diagram can be dispensed with.

Table 4 describes the advantages of the optionalrelays from Fig. 1.

Classification Options

Low

(100 500 kW)

7SJ61:more functions

(e.g. PLC/CFC), more communica-tion scope, and more binary I/Os

Medium(500 kW 2 MW)

7SJ63/64:graphics display andthus better local control, scalablebinary I/O quantities; 7SJ64 hashigher PLC/CFC performance andone additional interface.

7UM61: larger volume of binaryI/Os, wider frequency range (reli-able protection functionality evenduring running down of a motor)

Explanation of the options


17/80



Fig. 1 Allocation of protection relaysto motor power classes

The power of synchronous motors is generallyclearly higher than 2 MW. This is why it is imper-ative to provide differential protection.

A SIPROTEC 7UM62 with the Generator Basicoption is recommended as the protection unit.This option features underexcitation protectionto detect excitation problems (failure or regula-tion problems) and out-of-step cases (loss ofsynchronism).

You will find further notes on application, con-nection and setting of the protection relays in thefollowing chapters.

LSP2001.eps

LSP2299.e

ps

LSP217

1.eps

LSP2361.e

ps

LSP2299.eps

LSP2176.e

ps

Preferred Options7SJ60 7SJ61

7SJ62 7SJ63, 64 7UM61

7UM62


18/80




19/80

Siemens PTDEA Optimum MotorProtectionwith SIPROTECProtection Relays 17

Thermal Stressof Motors

Thermal stress of motors is discussed in this chap-ter. The various operating states are consideredand the thermal influence is derived from them.The second part contains discussions of suitableprotection principles and a basic explanation oftheir implementation in the numerical protectionunit. Taken from motor data sheets, notes on set-

ting the protection functions are provided. 1. Heating phenomena

In the process of conversion from electrical tomechanical energy, losses occur which leads toequipment heating up, i.e. a motor. Their designpermits certain thermal stress.

Heating sets in when the heat sources become ef-fective. The fed energy is first of all stored in eachvolume element as heat. The initial curve of thetemperature rise =f(t) is independent of thethermal resistors through which the heat is laterdissipated. Moreover, it is in no way identicaleverywhere because different loss density values

prevail in the various windings and also in theindividual segments of the magnetic circuit. Themotor can therefore generally not be looked uponas a homogeneous-body system.

Temperature differences develop between theparts of the motor and with respect to the sur-rounding coolant, and heat begins to flow in thedirection of the coolant. The temperature differ-ence between the coolant (which can also be air)and the machine part is calledtemperature rise(orovertemperature). After an adequately long time,the heat flows reach a state of equilibrium. Then,heat is no longer stored. The temperature has

reached itstemperature-rise limit value. Fig. 1shows how thetemperature rise(or overtempera-ture) develops for characteristic points within thestator of a motor.

Thepermissible temperature riseof an electricmachine is primarily limited by the relatively lowthermal resistance of the insulators for the wind-ings. The insulation systems are subdivided intotemperature classes A to Hin relation to their ther-mal resistance. The permissible temperature riseof a winding at its hottest point differs from the

permissible temperature of the insulation systemused by the agreed maximum temperature of thecoolant (ambient temperature), which is definedas 40 C. A safety factor of 5 to 15 K is generallyalso added.

Two modes of operation must be distinguishedwhen considering a motor from the point of viewof thermal influences.

a) Normal operation under load

Depending on the torque to be applied, the corre-sponding current is picked up from the powersystem which, depending on the design, leads tothe heat development described above. We canrefer mainly to the stator. The motor manufac-turer defines a machine thermally for the permis-sible temperature class. Operation under ratedconditions leads to a temperature rise that cangenerally be assigned to a lower temperatureclass. For example, if you find the letters F/B (de-signed/operated) in the motor data, the motor isdesigned for the temperature class F, but is usedin accordance with temperature class B. Thesethermal reserves results in a continuous overloadcapacity of about 10 %. A higher overload is pos-sible for a brief time, but the manufacturer doesnot specify any data because of the various influ-

encing parameters. The thermal withstand curve(see Fig. 2) is to be referred to.

Thermal Stress of

Motors and NecessaryProtection Functions

Fig. 1 Temperature rise at three characteristicpointson the stator


20/80

Siemens PTDEA Optimum Motor Protection withSIPROTECProtection Relays18


The curves are specified for when the motor isboth cold and warm. Warm means that the motorhas been operated with rated quantities and anoverload then occurred. The graphic above showsthe worst case, the curve of the copper time con-stant for short-time duty. The time constant for

continuous duty must be used for the protectionsetting. The continuously permissible current canalso be found in the characteristic. According toFig. 2 this is approximately 1.15I/IN,M.

In the American (NEMA) region, the overload isdescribed by the service factor (SF), which speci-fies the permissible overload for the rated powerin the S1 mode. This overload refers to the overallsystem (electrical and mechanical parts). Accord-ing to NEMA MG1, the permissible limitovertemperatures of the relevant temperatureclass may be 10 K higher. The values for the ser-vice factor lie, for example, at 1.1, 1.15, etc. How-

ever, if SF = 1 is specified, this means that themotor may only be operated at the rated power.Thermal overload protection refers only to themotor winding. If SF = 1 is specified, a continu-ously permissible winding overcurrent of 10 %can be assumed because the insulation is generallydesigned in accordance with temperature class Fand the motor is mainly operated in accordancewith temperature class B. If SF = 1.1, this may bemore than 10 % (e.g. quite possibly 15 %). Referto the thermal characteristic for the exact value ofthe continuously permissible overload current(see Fig. 2).

For detection of temperature rise, two tempera-ture sensors (preferable PT100) are usually in-stalled for each phase in the stator winding.

Motor manufacturers have placed them at thethermally critical points. The tripping tempera-ture can be derived from the temperature class.F signifies a maximum permissible temperatureof 155 C and for B the value lies at 130 C. Thelatter temperature ought to set in approximatelywhen the motor is loaded with rated quantities.Minus a safety factor of 10 K, the tripping value isapproximately (155 C -10 K = 145 C).

The rotor additionally heats up when the motor isoperated on anasymmetrical voltage. The mostcritical case is discontinuity in a phase. Voltageasymmetry means that there is a negative-

sequence voltage system that is driving a negative-sequence current. This negative-sequence rotatingfield leads to an AC current in the rotor (100 Hzrelative velocity) in the opposite direction of ro-tation. Due to the skin effect, a larger rotor re-sistance takes effect which, in turn, leads toincreased heating up. It can be assumed that1 % negative-sequence system voltage leads toabout 5 % to 6 % negative-sequence systemcurrent.

Fig. 2 Maximum permissible thermal overload of a motor

(Thermal copper time constant (short-time duty): 2.7 minThermal time constant (continuous duty (S1)): 15 minCooling t ime c onstant at s tandstill: 105 m in

Temperature class (designed/operated): F/B


21/80



The stator current can be estimated by way of theequivalent circuit. It must be split into thepositivephase-sequence and negative phase-sequence sys-

tems. As the rotatingfield of thenegative-sequencesystemruns in the negative direction at thesynchro-nous speed, the slip forthenegative-sequence sys-temis 2-s. The influence of the different rotorresistance values is very easily recognizable in Fig. 3.

Unless otherwise specified, it can be approxi-mately assumed that about 2 % negative-sequence system voltage does not lead to anyadditional heating of motors. This correspondsto a continuously permissible negative-sequencesystem current of approximately 10 %. If thecurrent rises above this value, tripping must takeplace after defined times. Unfortunately, motormanufacturers seldom specify any data in thisrespect.

b) Starting

During starting, thermal stress occurs in the rotordue to the starting currents. Typical startingphenomena are dealt with in the first chapterentitled Introduction to the principles of asyn-chronous and synchronous motors.

The cases of the locked rotor and the acceleratingrotor must basically be considered. In the case ofmotors with critical rotors (the majority of them),the locked rotor case constitutes the higher stressfrom the thermal point of view. This is why motormanufacturers frequently specify the thermal lim-its of the locked motor in the thermal limit curve.

These curves are shown on the right in Fig. 2. Ifthe motor accelerates, the thermal limits are gen-erally above them.

The motors starting time, for both the rated volt-age and the 80 % case, is also found in the techni-cal data. These times apply on connection of themachine.

For example, the motors starting time is 24.0 swith a starting current of 5.6I/INunder ratedvoltage conditions (VN) and 52 s at 4.17I/INand0.8VN. In Fig. 2, we read a time of about 40 s forthe warm condition at 5.6I/IN, with the resultthat the characteristic of starting time monitoring

can lie below the locked rotor characteristic.A speed check is not imperative here.

A further thermal stress results from the fre-quency of successivestartupsof the motor. Here,manufacturers distinguish between warm andcold. We frequently find the typical value of threestartups from the cold state and two startups fromthe warm state. This statement also applies tostarting with a reduced voltage (80 %).

A further important characteristic quantity is thenumber of permissible startups per year. Thisamounts to 1000, for example, and must not be

exceeded.

2. Implemented protection configurations

2.1 Stator thermal overload protection(normal duty)

In Section 1, we worked out that temperatureresponse is complex and really ought to be de-scribed by a thermal network. This descriptionleads to differential equations of a higher order,for which the input parameters are missing, how-ever. Due to the limited availability of data frommotor manufacturers, the simplified descriptionbased on a homogeneous-body model has estab-lished itself. Fig. 4 shows the model with the heatsource, realized as the current sourceI2, heat stor-age in the capacitorCth, heat dissipation throughthe resistorRthand the ambient temperature a.The model is based on a reference temperature 0of 40 C. The temperature describes the motorsovertemperature.

Fig. 3 Equivalent circuit diagram of a motor as positive and negative-sequence systems(left: positive-sequence system, right: negative-sequence system)

Fig. 4 Equivalent circuit diagram of the thermal model


22/80



The following differential equation (1) can be de-rived from the equivalent circuit diagram, whichis simplified further when we assume a constant

reference temperature 0.

I C2 0 0= +tha

th

d

d

( ) ( ) t R

I R R C t

20th th th a

d

d= +

( ) (1)

When a constant ambient temperature is as-sumed, the maximum permissible current willlead to the maximum permissible temperature(I2maxRth max). If we scale equation (1) withthe maximum permissible quantities, we arrive atthe differential equation to be implemented in theprogram. The differential equation must be calcu-

lated separately for each phase current. Trippingoccurs if the value 1 is reached in at least onephase (reason: calculation is based on scaledquantities).

It

2p.u. th a

d

d= +

(2)

with the following scaling

I I

I

I

Ip.u.

Nk= =

max

= =max k

2N

th th th=R C a a 0 a 02

Nk= =

max

The solving of the differential equation leads tothe well-known exponential expression. If thecurrent is increased abruptly, the temperatureexponentially approaches the stationary value.

( ) (t) 1 ep.u. a t =0 t / t 0th= + + =( )I2 (3)

As all values are scaled, the tripping time when(t) =1 is set can be calculated on the basis ofequation (3). If we also use the scaling quantities

of equation (2) and assume a measured ambienttemperature, the tripping time can be calculatedwith the following equation. If no ambient tem-perature is measured, a, measured = 40 C must beinserted and this leads to further simplification ofequation (4).

t

I

I

I

I=

th2

N2

previous load

Nn

1

k

1

k

1l

2 2

k

1

k2 N2

a,measured

N

I

I

+

2

0 1

A quasi-stationary state (constant over 5th) ispresupposed for the previous load and the meas-ured ambient temperature.

Therefore, for overload protection, two character-istic quantities of the motor are needed, thefactork, which describes the continuously maxi-

mum permissible current referred to the ratedcurrent, and the thermal time constantth. If theambient/coolant temperature is measured, thetemperature that sets in under the rated condi-tions is also needed as a scaling quantity. If this isnot available, it can be measured via the tempera-ture sensors.

During operation with a different load, the heat-ing time constant is equal to the cooling timeconstant. Longer cooling (no forced cooling bythe fan) must be assumed if the motor is shutdown, however. If, in this process, the currentfalls below a minimum value, standstill can be

assumed and it is possible to switch over to thelonger cooling time constant (in the exampleshown in Fig. 2, this is 105 min).

The behavior of the thermal model on activationof the motor must now also be considered. Here,however, the thermal load acts on the rotor. Thissignifies restricted validity of the stator model orof the setting values. Two strategies are currentlypursued:

Freezing the thermal memory during startingThis is intended to avoid overfunction of theoverload protection during starting.

Internal limiting of the starting current

The current fed to the thermal model is limitedduring starting. This slows down heat devel-opment. Current limiting should lie at approxi-mately 2.5I/IN,M(see also Fig. 2). Furtherreduction of limiting is recommended ifstarting times are long and successive warmstarts are possible. In the example, a value of2I/IN,Mmakes sense for a starting time of 52 s(at 0.8VN).

Different operating states were assumed to pro-vide an insight into the thermal behavior of thethermal model. Fig. 5 encompasses the followingsequence: starting with rated voltage (starting

time = 24 s), continuous duty under rated condi-tions, deactivation and direct reactivation at areduced voltage (starting time = 52 s), brief dutywith a reduced load (0.9I/IN,M) and then shut-down of the motor. After shutdown we recognizethe effect of the longer cooling time constant verywell. During starting, the starting current for thethermal model was limited to 2I/IN,M.


23/80



The tripping threshold for the thermal model is1 (corresponding to 100 %). The higher, but al-ready reduced, current leads to a visible tempera-ture increase during starting. However, thetripping threshold is not reached despite the longstarting time. During operation under rated con-ditions (I=IN,M= 1), the thermal memory has afilling level of about 76 % (rule of proportion:1.152 / 100 % like 12 / x %).

2.2 Thermal protection with an asymmetricalvoltage (unbalanced-load protection)

Voltage asymmetry leads to a current asymmetry,which can be described by the negative-sequencesystem current. On the basis of the phase cur-rents, the protection algorithm calculates (corre-sponding to the definition equation of thesymmetrical components) the negative-sequencesystem current that is to be assessed in protectionterms. In the simplest case, thresholds are queriedand tripping is initiated after a set delay.

To take the heat development process into ac-count, anI22tcharacteristic must be simulated,which is a well-known characteristic (e.g.I22 t=10 s) in synchronous machines. The characteristic

is released when the permissible negative-sequence (unbalanced-load) threshold (e.g. 10 %)is exceeded. Below the threshold, the motor mustbe allowed to cool down. An inverse curve resultsas the tripping characteristic (see also Fig. 9).

2.3 Starting time supervision

The rotor is thermally overloaded if starting is toolong. The thermal limit (see Fig. 2) is described byanI2tcharacteristic. This characteristic must besimulated. The following equilibrium condition(see equation (5)) can be set up and the permissi-ble starting or blocked-rotor time can be deter-mined on the basis of it.

I2 t=I2starttstart =t I

Itstart start

2(5)

where:

Istart maximum permissible starting currenttstart maximum permissible starting timeI measured starting current

Tripping occurs if the actual time is longer. Thetripping characteristic has an inverse characterand adapts very well to the starting conditions(with rated voltage and reduced voltage).

The calculation in accordance with equation (5) isnot released until starting is detected. For typicalmotors, the necessary current threshold amountsto approximately 2.5IN,M. The threshold must belowered accordingly for motors with reducedstarting currents.

By way of example, Fig. 6 shows starting cur-

rents/times measured for a motor. Adaptation ofthe protection characteristic to the starting condi-tions is easily recognizable.

2.4 Restart inhibit

Inadmissible heating of the rotor occurs when themotor is started too many times in succession. Inthe simplest case, the permissible limits can be

monitored by counter and the specified pausetimes can then be kept to.

Another approach is to model heating of the rotorto determine the thermal limit for the permissiblerestarts. In this way, we get closer to the physicalconditions and can generally load the motorbetter.

A homogeneous-body model is also expedient formodeling. The two necessary parameters consist-ing of the k factor for the rotor (kr) and the rotortime constantrmust be found. Both values aregenerally not given. However, they can be derived

from data of the manufacturers such as the num-ber of cold starts (nc) and warm starts (nw), thestarting time and the associated starting current.We create an equation system that describes the

Fig. 5 Behavior of the model in different load cases(data:k = 1.15; heating = 15min; cooling = 105 min)

Fig. 6 Measured starting currents and tripping characteristic


24/80



cold and warm states with the known quantitiesand we determine from it the two unknown val-ues (krand r). Equation (6) shows the approxi-

mated solution.

krc

c w

n

n nr c w ( )n n I2starttstart (6)

These parameters calculated internally in thefunction are incorporated into the thermal model(by analogy to equation (2)). Restart inhibit be-comes active if the thermal memory has reached acorresponding filling level. This threshold lies at(nc 1)/nc.

At standstill, the slower cooling is taken into ac-

count by extending the time constant. Beforethen, the thermal replica is frozen for an adjust-able time to take internal transient phenomenainto consideration. Only then does exponentiallydecaying cooling take place. Renewed restart isprevented for this time to allow the motor to rundown.

To meet demands for a minimum standstill time,an additional delay time is started after blockingby the thermal restart inhibit function. This timetakes effect only if the thermal model has enabledit beforehand.

Fig. 7 shows the thermal behavior for a possible

operating case. The motor is started from the coldstate and runs under rated conditions for a certaintime. After that, it is restarted twice from thewarm state and operated at 90 % of the ratedcurrent. The number 1 describes the thermallimit for the rotor and 0.67 the restart inhibitthreshold.

The first restart from the cold state leads to rotorheating, which will also set in during rated opera-tion. Two restarts are permitted from this warm

state. When the warm motor restarts a secondtime, the restart inhibit threshold is exceeded andthe temperature approaches limit value 1. Thisvalue represents the maximum permissible rotortemperature. Subsequent continued operationunder load conditions leads to correspondingcooling of the motor. If the motor were now de-activated, it could be restarted immediately againbecause there is a sufficient thermal reserve.

If, by contrast, it were deactivated immediatelyafter the second start from the warm state, therestart inhibit function would immediately takeeffect and would prevent renewed restarting.

From the thermal point of view, restart is onlypermitted when the temperature falls below the67 % threshold.

Fig. 7 Behavior under different loads according to thethermal rotor model(tstart =52s;Istart = 4.17 I/IN,M;nc =3and nW = 2)


25/80



3. Protection setting

To summarize, the setting parameters for thethermal protection functions are gathered to-

gether in a table and are commented on briefly.a) Stator overload protection

b) Negative-sequence protection

c) Starting time supervision

Setting parameters Value Comment

Heating time constant 15 min Take this from the motor data sheet ( see Fig. 2) or calculate iton the basis of a specifiedcharacteristic (equation (4) can beusedto dothis).

Cooling time constant 105 min Take this from the motor data sheet ( see Fig. 2) or set anexperience-based value, e.g. 7 times the heating time constant.

k factor 1.15 Take this from the thermal characteristic; if this is not available,a k factorof 1.1 can beused(see Section 1 temperature class F/B).

Thermal alarm stage 90 % During operation under rated conditions, the thermal warning stageis always below 90%. With the default setting of1.1, this is 83% andinthecaseof k = 1.15, only76 %.

Current limiting 2 I/IN,M

A slightly lower valuehas been chosen due to the long starting time,in particular with a reduced voltage. Thevalue2.5 IN,M can be usedfora short starting time.

Current alarm stage 1.15 IN,M

This should be set to equal the k factor.


Continuously permissibleunbalanced-load or alarmstage

0.1 I2/I

N,MUnless otherwise specified, this valueshould be set.

If this stage shouldalso trip, it must be delayed appropriately(approximately 15 - 30 s); suggestion: 20 s.

Tripping stage 0.4 I2/IN,M Two-phase operation under rated conditions leads to a negativephase-sequence system current of about66 %. Dueto thepowertobe produced,it will rise andwill certainly reach valuesof 100 %.

A tripping delay of about 3 s is recommended to allow fortransientphenomena.

Thermal unbalanced-load time((I2/ IN,M)

2 t = K )(Internal designation:FACTOR K)

2 s The conservative setting of 2 s is chosenif the motor manufacturerdoesnot specify any details. The timereally ought tobe longerif aneffect comparable to synchronous machines is presupposed.

Cooling time of thethermal model

200 s Usethe following relationship:

tcooling2,permiss.

K ss=

= =I

IN, M

2 2

2

01200

.


Starting c urrent ( pickup c urrent) 5.6 I /IN,M

Take this from the motor data sheet (starting curves).

Maximum permissible startingtime

35s Tothisend, the specified startingtimes must becompared with thecharacteristic curve times. Go below the thermal characteristicifthere is an adequate safetyclearance. According to Fig. 2, a value of40s is permissible in the case of5.6 I/IN,M in the warmstate.For thiscurrent, the motors starting time wasspecified as 24 s.

Chosen permissible starting time: 35 s

Start detection 2 I/IN,M

With thevoltage reduced, the starting current amounts to 4.17 I/IN,M

and current limiting for unbalanced-load protection was alsodefined at 2 I/IN,M. The typical value of 2.5 I/IN,M can beusedforstandard motors.


26/80



d) Restart inhibit

e) Characteristics resulting from theprotection setting values

By analogy to Fig. 2, Fig. 8 shows the trippingtime of the thermal characteristics. The thermaloverload protection characteristic can be seen on

the left and the starting time supervision charac-teristic can be seen on the right. The locked rotorcharacteristic of the manufacturer is also shown(curve just above it).

Fig. 9 shows the tripping response to a negative-sequence. In addition to the thermal characteris-tic, the two definite-time characteristics are alsoshown.

Fig. 7 adequately explains the response of therestart inhibit function for the chosen settingparameters.

References[1] IEC60034-1 Rotating electrical machines.

Part 1: Rating and performance,Release 2004

[2] IEC62114 Electrical insulation systems(ISM). Thermal classification, Release 2001


Startingcurrent (pickupcurrent) 4.17I/IN,M The current in the case of the longest starting time is taken from themotor data sheet. This is thevalueat thereduced voltage.

Starting time 52 s Choose the time that belongs to the current; if a motor starts clearlyfaster, e.g. in 2.6 s, you can be a little moregenerouswiththe setting.Here, wechoose the lowest setting of 3 s.

Cold starts 3 Take this from the motor data; if values are missing, assume thevalue 3.

Warm starts 2 Take this from the motor data; if values are missing, assume thevalue 2.

Rotor temperatureequalization time

1 min This time is considered practicable andalso takes running down tostandstill into account.

Note: Restarting of themotoris not possible duringthis time. Thetime must beset to zeroif only blocking is to takeplace when thethermal limit is reached.

Extension of time constantat stop

5 This valueis recommended forlong starting times. It leads to releaseof the restart inhibit function after approximately 30 minutes. Atleast thevalue10 must be chosen if the starting times areclearlyslower.

Minimum r estart i nhibit t ime 30 m in A t ime b etween 15 m in a nd 3 0 min i s r ecommended i f n o data i savailable. The longertime must be chosen forlonger starts.

Fig. 8 Thermal characteristics according tothe setting Fig. 9 Negative-sequence protection characteristics


27/80


Current TransformerRequirements

This chapter discusses the design of current trans-formers for motor protection applications. Vari-ous operating cases and typical protectionprinciples are considered. Design is explainedwith reference to examples, based on a choice oftwo approaches, checking of existing currenttransformers and new design. These consider-

ations are based, among other things, on the rele-vant IEC standards (IEC 60044-1, IEC 60044-6).

Formula symbols and definitions used

Kssc= factor of the symmetrical rated short-circuitcurrent(example: transformer CI. 5P20) Kssc= 20)

K'ssc= effective factor of the asymmetricalshort-circuit current

Ktd= transient rated dimensioning factor

Ipn= primary rated transformer current

Isn= secondary rated transformer current

Rct= secondary winding resistance inat75 C (or another specified temperature)

Rb= rated resistive burden in

R'b= Rlead+Rrelay= connected burden in

Rrelay= relay burden in

R l

leadA

= 2

where:

l= single conductor length between currenttransformer and device in m

q= specific resistance = 0.0175mm2/m(copper) at 20 C (or another specifiedtemperature)

A = conductor cross-section mm2

Istart= motor starting currentIstart trans= transient starting currentktrans factor for the transient starting current

For current transformers defined via the symmet-rical rated short-circuit current factor Kssc and therated resistive burdenRb (e.g.: 5P,10P), theeffective factor of the symmetrical short-circuitcurrent K'ssc can be calculated in accordance withthe following formula:

K Ksscct b

ct bssc' '=

+

+

R R

R R

Starting is crucial to stability when current flowsthrough and with regard to the design of thetransformers.

Requirements arisingout of motor startingThe starting currentIstartis in the range from fourto seven timesINwith a DC time constantTAofabout 40 ms (1 MW). This issuperimposed with a rush current of approxi-mately the same order of magnitude which, how-ever, decays in two to three cycles in accordancewith a time constant

Trushof about 20 ms.

The starting current is therefore:

Istart trans= ktransIstart, where ktrans= 2 - 2.5

The minimum required factor of the symmetricalshort-circuit current K'ssccan be calculated inaccordance with the following formula:

K' Kssc tdstart trans

pn

I

I

the following condition must be met:K'ssc(required) K'ssc(effective)

Motor Protection:

Requirements forCurrent Transformers


28/80


Current Transformer Requirements

Example1 Definite-time overcurrent-time protection, check of the existing current transformer

I P

VN

N

N

kW

kVA=

=

=

3

1000

3 6 3 0 851078

cos . ..

I Istart N A A= = =5 5 107 8 537 5. .

I Istart trans start A A= = =2 2 537 5 1075.

Setting valueI >>: 1.3 Istart trans= 1.3 1075 A = 1397 A

Note:Fort>> = 0 ms.With a time setting oft>> = 50 ms, the setting valuecan be reduced.(Refer to Chapter Protection of medium-power motors)

Requirement: K A

Assc

pn

' . >>

= =I

I

1397

1509 3, but at least 20

The CT's rated burden in amounts to:

R S

Ib

n

sn2 2

VA

A= = =

5

15

The actually connected burden (cable + device) amounts to:

R R R l

' ..

.b lead relay

A

mm

mm

m= + =

+ =

2

012 0 0175 10

2 5

2

m 2 01 0 24+ =. .

This results in the effective factor:

K' Kssc sscct b

ct b

= +

+ =

++

=R R

R R'

.

. ..10

1 3 5

1 3 0 2440 9

K'sscrequired = 20, K'ssceffective = 40,9 The dimensioningof the CT is correct.

Fig. 1


29/80



Example 2 Differential protection, design and checkingof a current transformer

2.1 Design of the current transformer T1

I P

UN

N

N

kW

kVA=

=

=

3

11165

3 10 0 85758

cos .

A CT withIpn=1000 A is chosen.

I Istart N A A= = =5 5 758 3790

I Istart t rans start A A= = =2 2 3790 7580

Requirement: K' K A

Assc td

start trans

pn

= =I

I5

7580

100037 9.

where Ktd= 5 for the generator/motor differential protection(Catalog SIP 2006)

The CT's class should be 10P with 20 VA rated burdenand an internal resistanceR ct 4

The CT's rated burden in is:

R S

Ib

n

sn

VA

A= = =

2 2

20

120

The actually connected burden (cable + device) is:

R R R l

' ..

b lead relay A

mm

mm

6= + =

+ =

2

012 0 0175 200

2

mm 2 + =01 1 266. .

This results in the rated factor:

K K'sscct b

ct b

ssc +

+ =

++

=R R

R R

' .. .

4 1 226

4 2037 9 8 3

Kssc= 10 is chosen

With this, all necessary CT data are available:

1000 A/1A, 10P10, 20 VA,R ct 4

Fig. 2


30/80



2.2 Checking the existing transformer T2

K'ssc= 37.9 (from design of transformer T1)


R S

Ib

n

sn2

VA

A= = =

15

115

2


R R R l

' ..

.b lead relay

A

mm

mm

m= + =

+ =

2

012 0 0175 10

2 5

2

m 2 01 0 24+ =. .


K Kssc sscct b

ct b

'' .

.= +

+ =

++

=R R

R R10

4 15

4 0 2444 8

K'ssc required = 37.9, K'ssc effective = 44.8 The dimensioningof the CT is correct.

Note: Contrary to the star point CT (-T1),the required apparent power can be slightly lowerbecause the distance from the protection is clearlyshorter. If the final distance is not clear, the worstcase must be reckoned with and identical currenttransformers must be provided.

Example3 Differential protection with bushing-type transformers

3.1 Current transformer T1

The classic motor protection functions (overload, starting supervision,negative-sequence (unbalanced-load), etc. operate with this current transformer).

I P

VN

N

N

kW

kVA=

=

=

3

11165

3 10 0 85758

cos .

Fig. 3


31/80



A CT withIpn= 1000 A is chosen.

I Istart N A A= = =5 5 758 3790

I Istart trans start A A= = =2 2 3790 7580

Setting valueI I>> = =: . .1 3 1 3 7580 9854start trans A A

Note:Fort>> = 0 ms.With a time setting oft>> = 50 ms, the setting valuecan be reduced.(Refer to Chapter Protection of medium-power motors)

Requirement: K' A

Assc

pn

>>

= =I

I

9854

10009 85. , but at least 20

The CT's class should be 10P with 5 VA

rated burden and an internal resistanceRct 4.The CT's rated burden in amounts to:

R S

Ib

n

sn2

VA

A= = =

5

15

2

The actually connected burden(cable + device) amounts to:

R R R l

' ..

.b lead relay

A

mm

mm

m= + =

+ =

2

012 0 0175 10

2 5

2

m 2 01 0 24+ =. .


K Ksscct b

ct b ssc

+

+ =

+

+ =

R R

R R

'

'

.

.

4 0 24

4 5 20 942

Kssc= 10 is chosen

With this, all necessary CT data are available:

1000 A/1 A, 10P10, 20 VA,Rct 4

3.2 Bushing-type transformer T2

For machines connected via cables, reliable differentialprotection with high response sensitivity can be realizedwith this method. The prerequisite is that the threephases are returned separated from the star point sideand are routed through the bushing-type transformer

in the opposite direction. If the machine winding is fault-free, the currents in the transformer cancel each otherout and no current flows into the (differential) currentrelay. The current comparison is highly precise withoutsaturation problems, taking place as a magnetic(self-balancing) operation.No stabilization is necessary, i.e. simple current relayscan be used. The pickup value can be adjusted to 2 to 5 %of the machines rated current.One 7UM62 can also be used instead of two 7SJ6 units,in which case the overcurrent function must be assignedto the bushing-type transformer.


32/80



CT requirement for-T2The (differential) current relaymustrespond to high-current internal faults.

As the setting value is very low and the total currentis zero during fault-free operation,a CT with 500 A/1 A, 10P10, 5 VA,Rct=2 is chosen, with the requirement:K'sscW 20 (analogous to the definite-timeovercurrent-time protection)


R S

Ib

n

sn

VA

A= = =

2 2

5

15


R R R l' . .b lead relay A

mm

m mmm

= + = + = 2 01 2 0 0175 2006

2

2 01 1 266+ =. .


K' Kssc sscct b

ct b'

= +

+ =

++

=R R

R R10

2 5

2 1 226214

.

.

K'ssc required = 20, K'ssceffective = 21.4 The dimensioningof the CT is correct.

Summary

The requirements for current transformers are definednot only by the starting operation, but also by therequirements of the protection principle. Further

important defining variables are the burden through thefeeder and also the current transformers internal burden.The current transformer types chosen in the examplescan certainly be used for orientation. However, a generalcheck is nevertheless recommended.


33/80


Low-PowerMotors

Protection of

Low-Power Motors

Principles and settings for the protection of asmall (asynchronous) motor are discussed below.Particular attention is devoted to design of theearth-fault protection for earthed, isolated and

1. Protection functions

The following protection functions are available:

7SJ600:

Time-overcurrent protection phase/earth,thermal overload protection, starting timesupervision, restart inhibit and negative-sequence(unbalanced-load) protection

7SJ602 (from V3.5) additionally:

Sensitive earth-fault protectionDirectional earth-fault protection for isolatedand compensated systems

Time-undercurrent protectionTemperature monitoring by RTD-box

resonant-earthed power systems, representative ofall kinds and sizes of motors. The 7SJ600 or the7SJ602 is considered a preferred protective relayfor small motors.

* Required depending onstar point connectionandsystemtopology

** With 7SJ602only

Besides the SIPROTEC relays 7SJ600 and 7SJ602presented here, it goes without saying that the7SJ61,62,63,64 relays in the SIPROTEC 4 familywith a wider scope of functions are also suitable.Refer to the chapters on protection of medium-and high-power motors.

Protection functions Abbreviations ANSI

Time-overcurrent relay, phase I>>,I>,Ip 50, 51

Non-directional/directionalearth-fault protection*,**

IEE dir.>>,IEEdir.>,VE> 51N, 51Ns,67Ns, 59N

Displacement voltage protection 59N

Thermal overload protection I2 t> 49

Starting time supervision Istart2 t 48

Restart inhibit I2 t 66, 49R

Negative-sequence(unbalanced-load) protection

I2>,t=f(I2) 46

Temperature monitoring ** (Thermobox) 38


34/80


Low-PowerMotors

3. General power system data

The current and voltage transformer data and thetype of connection are entered in the unit under

1100 Power System Data.

Here are the DIGSI parameters of a 7SJ602:

1118: Via this setting, the entries for the startingtime supervision and restart inhibit motorprotection functions are referred to rated motorcurrent. 1119: Motor starting or locked-rotorcurrent. 1120:Tthe rotor locking time for 6.5IN,Mis 18 s, and the startup time for the 422 kW pumpused in the example is 2.1 s. An interme- diatevalue of 8 s was chosen (see also comments on

starting time supervision).

No. Parameter Value

1100 POWER SYSTEMDATA

1101 Rated system frequency fN 50 Hz

1105 Primary rated current 75 A

1106 Secondary rated current 1 A

1111 Matching factorIee/Iph forearth current

0.800

1112 CT nom. current IEEp, sec 1 A

1113 Nominal VT voltage, primary 6.00 kV

1114 Nominal VT voltage, secondary 100 V

1115 Reverse power direction off

1116 Threshold c ircuit breaker c losed 0.10 I /In

1118 Nominal current ofmotorin relation to Tr.C.

0.6

1119 Startup current in relationto nominal current

6.5

1120 Maximum startup time 8.0 s

1121 Reset thermal image on startup yes

1134 Minimumtripcommand duration

0.15 s

1135 Maximum closecommand duration

1.00 s

Connection of the 7SJ602 for an isolated orresonant-earth system with earth-fault directiondetection.

2.

A 75 A/1 A current transformer is chosen for thephases and a core-balance current transformer60 A/1 A, as well as 6 kV/100 V voltage trans-formers.

Fig. 1 Connection with 3 current inputs IA(L1)-IB(L2)-I4(EE)for sensitive earth-fault protection anddisplacement voltage input Ve-n

Size Value

Power 460 kW

Rated voltage 6 kV

Rated motor current 48 A

Max. starting current (at 100 %VN) 312 A = 6.5 IN, M

Starting time (at 100 % VN) 2.1 sec

Heating time constant 40 min

Cooling time constant at standstill 240 min

Max.continuousthermalratingcurrent 58 A = 1.2 IN, M

Max. locked rotor time 18 sec at 6.5 IN, M


35/80


Low-PowerMotors

4. Protection functions

4.1 Time-overcurrent protection, phase

Dynamic thresholds are used, i.e. during theincreased current in motor starting, the over-current stagesI>>> (1303) andI>> (1305)are raised appropriately (1304:I>>> dyn,1306:I>> dyn) and are lowered again after start-ing (hold time 1302: 10s). Advantage: a sensitivesetting with a short tripping time is possibleduring normal operation. As an additional reservestage, stageI> is operated with a tripping time(1310) longer than the starting time.All stages are set to 1.5 times the maximum operat-ing current, e.g.I>>, dyn = 1.5 6.5 0.64 = 6.24;

6.5: max. starting current,0.64: motor/transformer rated current conver-sion.I>> = 1.5 1.2 0.64 = 1.15.1.2: max. continuous thermal current,0.64: motor/transformer rated current conver-sion.

4.2 Negative-sequence protection

No data is available from the motor manufactur-er, and so the following settings can be used asrecommended values:

No. Parameter Value

1300 O/C PROTECTION PHASE FAULTS

1301 O/C protection for phase faults on

1302 Duration oftemporary pickup value c/o

10.00s

1303 Pickup value ofthe high-set inst.stage I>>>

1.2I/In

1304 Pickupval.of high-setinst. stage I>>> (dyn)

6.3I/In

1305 Pickupvalueof thehigh-set stage I>>

1.2I/In

1306 Pickupvalueof thehigh-set stage I>> (dyn) 6.3I/In

1307 Trip timedelayof thehigh-set stage I>>

0.05 s

1308 Pickupvalueof theovercurrentstage I>

1.2I/In

1309 Pickupvalueof the O/C stage I>(dyn)

1.2I/In

1310 Trip timedelay oftheovercurrent stage I>

10.00s

1311 Measurement repetition no

1319 Manual close I>>undelayed

The values refer to the rated transformer current.1502 and 1504 are set to approx. 10 % or 40 % ofthe rated motor current. Further information canbe found in the chapter entitled Thermal stressof motors and necessary protection functions.

4.3 Thermal overload protection

Thermal overload protection serves to protect thestator against inadmissible temperature rise.

All values are referred to rated transformer current.2702: The figure can be obtained from the mo-tors thermal withstand curve and is the asymp-

totic value of the characteristic at the bottom.In the example,

kN, M

= 1 2. I

I

Referred to rated transformer current:

k N, M

N, CT

= = =1 2 1 2 0 64 0 77. . . .I

I

2703 and 2704 are taken from the data sheet;important: the thermal time constant duringcontinuous operation and not the copper timeconstant for short-time operation must be usedfor 2703.

2705: The thermal warning stage picks up whenthe calculated temperature reaches 90 % of themax. permissible temperature.

No. Parameter Value

2700 THERMAL OVERLOADPROTECTION

2701 State ofthermaloverload protection

on

2702 K-factorfor thermaloverload protection

0.77

2703 Timeconstant for thermaloverload protection

60.0 min

2704 Multiplierof timeconstant at standstill

4.00

2705 Thermal warning stage 90 %

No. Parameter Value

1500 UNBALANCEDLOAD

PROTECTION

1501 State of the unbalanced-loadprotection

on

1502 Pickupvalue ofneg.seq.I low-set stage I2>

8 %

1503 Tripdelayof neg.seq.I low-set stage TI2>

20.00s

1504 Pickupvalue for highcurrent stage

26 %

1505 Triptime delay for highcurrent stage

3.00 s


36/80


Low-PowerMotors

4.4 Starting time supervision

This serves to protect the rotor during motorstarting. It is advisable to permit a maximum

starting time that lies between the real startingtime (2.1 s) and the maximum locked-rotor time(18 s), on the one hand to avoid responding toosensitively in the event of long starting times, andthus deactivating the motor unnecessarily, on theother hand, though, to also avoid reaching the ro-tor insulations absolute thermal limit. In the caseof the 7SJ602, both values are located in the sys-tem data parameter block, where they are de-scribed.

2803: If the measured phase current rises abovethis value, motor starting is detected and thefunction is activated. A reasonable value forstarting detection is 2.5 times the rated motorcurrentIA= 2.5 0.64 = 1.6. 2804 offers the possi-bility of blocking the overcurrent stages to avoidovercurrent tripping during motor starting.Our example, however, uses dynamic thresholdboosting during starting and a reserve stageI> with a tripping time longer than motorstarting, and so this must not be blocked.

4.5 Restart inhibit

Thisfunction preventsexcessively frequentstartingof themotor in succession. As startingtime super-vision does nothave a thermal memory, it wouldpermit several starts in direct succession providedthemaximum start-up time were notexceeded.Generally, though, motor manufacturers permit

only threestarts from thecold operating state(4503+ 4504) and twofrom thewarm state (4503).

Calculation is based on a thermal homogeneous-

body model similar to the standard overloadprotection that operates for the stator winding.The time constant during operation is equal tothe value set in 2703, and an additional value 4505can be set for prolongation during motor stand-still. 4502 is an empirically based value and cangenerally be left set as it is. If the manufacturerspecifies a minimum inhibit time between twostarts, this can be set in 4507. It acts in additionto the blockage derived dynamically from thethermal image. Here, the k factor, which mustbe worked out in the case of overload protection,is determined automatically on the basis of the

number of starts for cold and warm and is nor-mally considerably lower for the rotor than forthe stator.

4.6 Undercurrent monitoring

Undercurrent monitoring serves to protect thedriven load and detects a load loss that might

cause damage, e.g. pumps running dry.The settings depend on the type and size of thedriven load.

4.7 Breaker-failure protection

This serves to repeat (on another command relay)

a TRIP command that does not lead to openingof the circuit-breaker, thus tripping any secondrelease coil of the circuit-breaker or a higher-levelcircuit breaker.

The delay time must be greater than the circuit-breakers run time and the reset time of the BFPincluding safety reserve.

No. Parameter Value

2800 STARTING-TIME SUPERVISION

2801 Supervision of starting time on

2803 Basevalue Istrtof permissiblestart-up curr.

1.6 I/In

2804 Block oftheI>/Ip stagesduringstart-up

no

No. Parameter Value

4500 MOTOR START PROTECTION

4501 Motor start protection on

4502 Temperature equalization time 1.0 min

4503 Maximum permissible numberof warm-starts

2

4504 Difference between no.of cold andwarm-starts

1

4505 Factor for tau during standstill 5.0

4506 Factor for tau during operation 2.0

4507 Minimum blocking timefor motor protection

6.0min

No. Parameter Value

4600 UNDERCURRENTMONITORING

4601 Limit for undercurrentmonitoring

0.20I/In

4602 Delay timefor

undercurrent monitoring

10.0 s

4603 Undercurrent monitoring on

No. Parameter Value

3600 BREAKER-FAILURE PROTECTION

3601 Circuit breaker failure protection on

3602 Delay time T-B/F 0.3 s

3603 Analysisof auxiliarycontacts forB/F

off


37/80


Low-PowerMotors

4.8 Displacement voltage protection VE>

This function serves to detect an earth fault inisolated and compensated power systems and isused in addition to overcurrent sensitive earth-fault protection. Especially at the infeed of abusbar, single-pole faults cannot be detected due

to the zero sequence current and are indicated bytheVEprotection. The fault location can be deter-mined by switching operations and the fault canbe cleared.

Below, earth-fault protection is described indetail, representative for all kinds and sizes ofmotors.

4.9 Earth-fault protection

Earth-fault protection mainly depends on thepower systems star point earthing. An earth faultoccurs in isolated and compensated power sys-tems (Petersen coil at the systems star point);

i.e. the occurring fault currents are relatively lowand therefore do not need to be deactivated in theshortest of times. Power system operators cancontinue operation of their networks for a limitedtime and eliminate the fault through switchingoperations. By contrast, in impedance or solidlyearthed power systems, the fault current is so highthat is has to be switched off straight away.

a) Solid or low-impedance star point earthing

The fault currents are short-circuit currents andmust be eliminated fast. Any resistance installedat the star point serves to limit the earth-fault cur-rents in the event of earth faults (ph-e, ph-ph-e)

to a certain maximum value.

Selectivity criteriaThe short-circuit currents occur only in thefeeder containing the fault. Short-circuitcurrents can be detected that are lower thanthe motors rated current because this isapproximately symmetrical during normal

operation; i.e. it does not contain a negative-sequence and zero-sequence component. Thesetting sensitivity limit is defined by asymme-tries and measuring errors of the current trans-formers especially in a Holmgreen circuit and asymmetries of operation. It is therefore

recommended not to set more sensitivelythan 0.1 0.2IN, CT.

Fig. 2 shows an overv

Optimum Motor Protection en - LIBRO

Documents

Transcript of Optimum Motor Protection en - LIBRO