Condition monitoring of slip-ring induction motors

Electric Power Systems Research, 15 (1988) 189 - 195

Condition

Monitoring

189

of Slip-ring Induction

Motors

R. NATARAJAN, J. L. KOHLER and J. SOTTILE

Mine Electrical Laboratory, Department of Mineral Engineering, The Pennsylvania State University, University Park, PA 16802 (U.S.A.) (Received April 15, 1988)

ABSTRACT

A technique for real-time condition monitoring o f slip-ring induction motors is presented. The proposed method utilizes on-line sensing and analysis o f the rotor current to assess the integrity of the motor. The theoretical basis for this approach is presented; the results o f an experimental investigation are given. The paper is concluded with a discussion o f the efficacy of the proposed method in industrial power systems.

INTRODUCTION

The maintenance of electric drives is essential to the efficient and safe operation of any industry. The lack of an effectively designed and implemented maintenance policy will result in increased downtime, increased capital losses from catastrophic failures, and, in some cases, a reduced level of personnel safety. Traditional approaches to maintenance include age-replacement policies and preventive-maintenance testing. The relative success or failure of these approaches is dependent on the specific industrial setting. However, both have inherent limitations. For example, replacement policies tend to increase costs because motors are often taken o u t of service while still in good condition. Off-line testing procedures, such as insulation-resistance measurements, have limited value since they can be difficult to interpret and are applicable to the detection of a limited number of failure modes. On-line and continuous monitoring methods offer significant advantages in this regard, since incipient failures can be detected, and appropriate maintenance actions can be scheduled, thereby avoiding the 0378-7796/88/$3.50

costs and problems associated with unexpected failures. Thus far, research in this area has focused on sensing airgap flux, line currents, vibrations, electromagnetic noise, shaft flux, temperature and speed fluctuations [ 1 - 5]. An alternative hardware and software approach has been under development. In this approach a microcomputer-based monitoring system with rugged and inexpensive voltage and current sensors is used to sense terminal characteristics of power system components. Extensive signal processing techniques are then used to detect incipient failure [6, 7]. The results of a recent investigation suggest the suitability of this approach for condition monitoring of slip-ring induction motors [8]. This paper first presents a theory for the identification of harmonic frequencies present in the rotor flux/current signal. Next, the arrangement for the experimental study is outlined. The results of the machine in the conventional mode and in the simulated deterioration mode are discussed.

THEORY OF ROTOR HARMONICS

The permeance variation of the stator winding distribution due to s t a t o r - r o t o r slotting and saturation is responsible for the production of airgap harmonics in a threephase motor. An analysis of the rotor harmonics is presented here.

Magnetomotive force (MMF) waves For a symmetrical three-phase machine the stator MMF wave can be written as [9] Ms = ~ M n , s cos(np~b + cot)

(1)

© Elsevier Sequoia/Printed in The Netherlands

190

where n = 6q + 1; q = 0, 1, 2, 3 . . . . , The corresponding first- and second-order rotor MMF waves are

_[S (1--s)--(6--5s)]cot}

scot)+Ms. r cos[5p¢

Mr = Ml.r c o s ( p C -

+ B 6 c o s l ( S - - R + 5p)¢

+ (6 -- 5s)cot]

(2)

--IS(1--s)--(6-- 5s)]cotl

The permeance variation due to statorrotor slotting is given by [9] PsR =

PsR,, cos

X (1--s)cot

+ B7 cos t(2S + 5p)¢

[n(S +

-- [2~Sp(1--s) +-(6--5s)lcotI

R)¢ -- n --

1

(3)

+ Bs cos[(2R + 5p)¢ -- (6 -- 5s)cot]

.I

The effects of saturation can be expressed as [ 6 ] ee

P~ = ~ Psa,, c o s [ 2 n ( p ¢ - - co01

(4)

The rotor harmonic flux density waves can be derived by multiplying the MMF by the permeance waves. For a uniform-airgap machine without saturation, the flux density waves can be obtained by multiplying eqns. (2) and (3). In eqn. (3) only the first-order waves and their modulating effects are included. The flux density expression is B = (Ml, r cos(p0 --

scot)

l [

+B~cos

(6)

[1 + m ( 1 - - s ) ] f l

m = 5, 11, 17, ... (7)

(b)

[1--m(1 --s)]fl

m = 1, 7, 13, 19, ...

(8)

(d)

+cos[2S¢ -- 2 S (1-- s)cot] f (1--s)+s

- - (6 - - 5s)cot]

Is

f

n-- ( 1 - - s) -+ [ 1 - - m ( 1 - - s)] fl P m = 1 , 7 , 13, ...

(9)

n = 1, 2, 3, ...

s(1-s)cot]

P

l

+ B10 c o s [ 5 p ¢

(a)

(c)

B=BlCOS ( S + R + p ) ¢ - -

scot)

where the flux density (B) magnitudes are due to the multiplication of the MMF and permeance waves. The frequencies of interest are with low pole-pair numbers, and a summary of the important frequencies is as follows.

+ Ms, r cos[5p~b + (6 -- 5s)cot]}

X Psa, l cos (S + R ) ¢ - -

+ B9 cos(pC --

(5)

tnS(1--s)+[l+m(1--s)]lfl

(10)

m = 5, 11, 17, ... n = 1, 2, 3 . . . .

]f

cot

(S -- R -+ p ) ¢ - - ~ ( 1 - - s ) _ + s

cot

The frequencies of eqns. (7) and (8) are due to the MMF; and those in eqns. (9) and (10) are due to the interaction of the MMF and slotting effects. These frequency expressions are in agreement with the earlier results [10], derived using other approaches.

Effect of saturationon field frequencies + B3 cosl(2S + p ) ¢ - - [ ~

+ B4 cos[(2R

(l--s)+slcot

+P)O+scot]

+ Bs cos t(S + R +- 5p)¢ !

I

Most of the present-day machines are operating at the saturation level, and they produce additional magnetic frequencies. The permeance equation, taking into account the slotting and saturation, is given by P = eqn. (5) + P3 cos[2(p¢ -- cot)]

191 + - -

2

cos

+ B3 cos I (S -- R -- 3p)~b

(8 + R + 2p)¢

--[S(1--s) +7--5s]cot1 +

PIPa cos l (S + R -- 2p)~b 2

+ B4 cos l (S -- R -- p ) ¢

P2P3 l( S - - R + 2 p ) ¢

+ Ba cos I (S -- R -- 3p)¢

P2P3 t(S -- R -- 2p)¢

+ Bs c o s t ( S - - R + 3p)~b

+--cos 2

+ --cos 2

1

--[pS--(1--s)--7 +

The flux density equations can be obtained b y multiplying the MMF equation (2) by the permeance equation (11). In the resulting equation only the c o m p o n e n t s of low pole-pair numbers are demonstrated because they produce frequencies of dominant nature. The flux density wave is B = eqn. (6) + B, c o s [ 3 p ¢ -- (1 + + B 1 cos[p~b -- (1 -- s ) ~ t ]

s)~t]

5s]cotf

+ Bs c o s l ( S - - R -- 7p)¢ (12)

From eqn. (12) the new harmonic frequencies due to saturation in the MMF frequency region and slot harmonic sidebands can be identified. They are:

(a) [1 + k ( 1 - - s ) ] f l

(13)

+ B 2 cosl(8 - - R + 3p)¢

1

due to MMF and saturation, k = 3, 5, 7, 11, ... (b)

Is

np(1--s)--m(1

]

+ s ) fx

(14)

+ B 2 COSI(8--R +p)~b sidebands due to saturation, k = 3, 5, 7, 11, ... (c)

+ B 3 cosl(8 - - R + 7p)~b

+s)]ffl In--(1--s)--[2+m(l s

(15)

P

sidebands due to saturation, m = 5, 7, 11, ... These frequencies will appear on either side of the slot harmonic frequencies and are important in saturated machines.

192

the rotor flux density is very involved, since the rotor is in motion during normal operation. However, the slip-ring currents are proportional to the rotor flux density. Hence, current sensing probes were installed arou~ld the slip-ring o u t p u t terminals. The signals were analyzed on a universal waveform analyzer (Analogic Data 6000), a versatile instrument with data acquisition, storage, and processing capabilities. The experimental set-up and the instrumentation systems are shown in Fig. 1. This approach ensures t h a t the transducer is nonintrusive so that on-line measurements can be taken with a minimum of disturbance to the normal operation of the motor.

Fig. 1. E x p e r i m e n t a l set-up: UM -- universal m a c h i n e ; U A -- universal w a v e f o r m a n a l y z e r ; S R = c u r r e n t sensor.

Time and frequency domain analysis The rotor current was monitored with the aid of a clip-on current transformer and the signal processed in the time and frequency domain. With 100 V, 3.45 A, and 3240 rev min -1 (slip = 0.1), the time domain signal is shown in Fig. 2. Clearly, the fundamental has a period of 0.167 s, which amounts to

EXPERIMENTAL ANALYSIS

The experimental verification of the analysis was carried out on a 2 hp, 220 V, 7 A, 3600 rev min -1 universal machine. The number of stator and rotor slots were 24 and 36 respectively. Measurement of

4D.

20-

o

D-

--P.D -

--~,D -

I D.D

I

I

I D. 2

J

I

Time,

I

I D.,6

s

Fig. 2. Time d o m a i n analysis o f t h e r o t o r c u r r e n t signal a t 1 0 0 V, 3.45 A a n d 3 2 4 0 rev m i n -1.

I

193

~o ,r',h

~7

~ E

~ v ,-4

,-..~ II

II i[1. 4-

I

II E

u~

,

~.q

II

II

,~i~

~

U

II

I oO o"x ~

I:zl

n.

0 >

~"

2-

II

II.

~

~

,-I

,-4

I I

II I

~

II

4-

~

I

I

'~

~

r~

~

I

,~

"-"

,'-4

I

~

~

~

,.~

~ l

~ I

l

I

I

i

Frequency, domain

9 II

II ° 1 2

i

/

~,2'

k 4' 12' 6 '

6 5'

l

I ='DDO

Hz

analysis of the current signal at I 0 0 V, 3.45 A a n d 3 2 4 0 rev m i n -I.

O

'°/

I

I r',Do

Fig. 3. F r e q u e n c y

I

-~-

I~

~

~-"

nD

47-

~

cg •~

Ul

~

on

S

Fig. 4. S i m u l a t e d phase-to-phase fault in the stator w i n d i n g using an external resistance RE.

6 Hz in the frequency domain. The slot harmonic ripples modulate the sine wave with high frequencies. The frequency domain analysis of the above current signal is shown

in Fig. 3. The major frequencies identified are classified as follows. (a) MMF harmonics: 6, 318, 642, 966, 1290, ... Hz according to eqn. (7). (b) MMF harmonics: 330, 654, 978, 1302, ... Hz according to eqn. (8). (c) First-order slot harmonics (1290, 1302), (1614, 978), ... Hz according to eqn. (9). Second- and higher order slot harmonics are not seen. (d) Saturation harmonics: {102, 292), (210, 330), (318, 438), (534, 634), ... Hz. Some of the saturation frequencies are clear. Some other frequencies coincide with MMF! slot harmonics. These are in accordance with eqn. (13). Thus there is good agreement between the calculated values using the proposed theory

194

oof

0.4

0.2 u') I.J 0>

to

}-

0.0

J 0>

O.C

-0.2

-0.4

(a)

--0.5

I 0.3

I

I 0.4

I

I

0.5

I

I

0.6

I

I

I

0.7

(a)

TIME, S

0.006

01 J 0>

0.3

I

I

I

0.4

I

0.5 TIME, S

I

I

I

0.6

I

0.7

0.006

0.004

0.004 ,_1 o>

0.002

0.002

oooo

I 0

I

.... I 400

, . I I I I SO0 1200 F R E Q U E N C Y , Hz

hl I,

I

I

I 1600

I

0.000 . . . . I I 0

~=

l I

,

1 I 400 800 FREQUENCY,

, I 1200 Hz

I

I 1600

(b) Fig. 5. Rotor current of the test motor under normal operating conditions at 220 V, 8.5 A and 3450 rev rain-1 (full load): (a) time domain signal; (b) amplitude spectrum of the signal.

(b) Fig. 6. Rotor current of the test motor during simulated phase deterioration at 220 V, 8.5 A and 3450 rev min -1.

TABLE 1 Effect of motor unbalance on the double-frequency component of the rotor current (winding illustrated in Fig. 4, no load)

Study of simulated deterioration

If (A)

Ia2 (A)

Rotor 119 Hz component (mA)

0 0.33 0.80 1.25 1.38 2.73 4.07

0.016 0.126 0.210 0.337 0.384 0.745 1.143

90 128 212 342 348 703 997

and t h e m e a s u r e m e n t s in a n o r m a l l y o p e r a t i n g machine. Also, it is perceived t h a t m o n i t o r i n g t h e line c u r r e n t is a d e q u a t e t o i d e n t i f y the integral behavior o f a m o t o r . In o r d e r t o gain a b e t t e r u n d e r s t a n d i n g o f these frequencies, t h e m o t o r was l o a d e d t o d i f f e r e n t levels and t h e c u r r e n t signal analyzed. T h o u g h the magnitudes vary, t h e general relations satisfy t h e t h e o r e t i c a l predictions.

A n y m a j o r winding o r insulation failure in a three-phase m o t o r is usually initiated b y a t u r n - t o - t u r n f a u l t in the same phase, b e t w e e n the phases, or t o t h e ground. F u r t h e r m o r e , in the incipient stage such failures are o f high resistance in nature. T h e r e f o r e , it is v e r y difficult t o i d e n t i f y these failures simply b y , m e a s u r e m e n t o f the terminal quantities. Instead, a detailed t i m e and f r e q u e n c y analysis o f t h e signal m a y provide the necessary i n f o r m a t i o n . F o r example, Fig. 4 simulates an incipient failure in t h e s t a t o r o f the universal machine. T h e resistance R E (30 ~ ) is m u c h higher t h a n t h e phase resistance o f the m o t o r (1.24 ~2) and R E represents a high resistance fault, such as a t u r n - t o - t u r n fault b e t w e e n phases. By the t h e o r y o f s y m m e t r i c a l c o m p o n e n t s , the negative-frequency c u r r e n t s in t h e stator i n d u c e c u r r e n t s of d o u b l e f r e q u e n c y in t h e r o t o r circuit [5]. F o r i n d u c t i o n m o t o r s , this d o u b l e f r e q u e n c y is 120 -- /2, w h e r e f2 is the f r e q u e n c y o f the r o t o r c u r r e n t . T h e labor a t o r y tests have indicated t h a t this double-

195 f r e q u e n c y c o m p o n e n t is, indeed, v e r y sensitive t o m o t o r u n b a l a n c e . Figures 5 a n d 6 illustrate this p o i n t v e r y clearly. Figure 5 s h o w s t h e r o t o r c u r r e n t w a v e f o r m a n d its associated spectral content for a normal m o t o r at r a t e d load; t h e f r e q u e n c i e s n o t e d in this s p e c t r u m are t h e s a m e as t h o s e rec o r d e d in Fig. 3. Figure 6 illustrates t h e same information for a motor undergoing s i m u l a t e d d e t e r i o r a t i o n . T h e e f f e c t o f this i m b a l a n c e is clearly seen o n t h e d o u b l e frequency component of the rotor current. Also, T a b l e 1 s h o w s t h e r o t o r d o u b l e frequency component and negative-sequence c u r r e n t f o r several levels o f s i m u l a t e d d e t e r i o r a t i o n . S o m e o f t h e o t h e r i n c i p i e n t failures t h a t c a n be a n a l y z e d using this a p p r o a c h are (a) u n b a l a n c e in t h e s t a t o r v o l t a g e / r e s i s t a n c e , (b) p h a s e - t o - p h a s e faults, a n d (c) p h a s e - t o g r o u n d faults. M o r e r e s e a r c h is u n d e r w a y a l o n g t h e s e lines t o e x p l o i t this p r i n c i p l e as an on-line m o n i t o r f o r slip-ring m a c h i n e s .

MMF m n P P q R S $ t

magnetomotive force integer integer p e r m e a n c e , p.u. pole numbers slot/pole/phase r o t o r slot n u m b e r s n u m b e r o f s t a t o r slots slip, p.u. time, s

¢

m e c h a n i c a l angle, d e g r e e stator supply angular velocity, rad/s

REFERENCES 1 P. J. Tavner, B. G. Caydon and D. M. Ward, Monitoring generators and large motors, Proc. Inst. Electr. Eng., Part B, 133 (1986) 169 - 180. 2 J. R. Cameron, W. T. Thomson and A. B. Dow, Vibration and current monitoring for detecting airgap eccentricity in large induction motors, Proc. Inst. Electr. Eng., Part B, 133 (1986) 155 163.

SUMMARY

AND CONCLUSIONS

On-line m o n i t o r i n g o f slip-ring i n d u c t i o n m o t o r s c a n be carried o u t using t h e s p e c t r a l c o n t e n t s o f t h e r o t o r c u r r e n t signal. In o r d e r to predetermine the frequencies involved in t h e r o t o r c u r r e n t , a t h e o r y is p r e s e n t e d based on the permeance method. The e x p e c t e d h a r m o n i c f r e q u e n c i e s are i d e n t i f i e d in t e r m s o f t h e s t a t o r , r o t o r slot n u m b e r s , n u m b e r o f poles, a n d t h e s u p p l y f r e q u e n c y . T h e r e is g o o d a g r e e m e n t b e t w e e n t h e calculated and measured frequencies. The usefulness o f this r e s u l t in t h e on-line m o n i t o r i n g o f a slip-ring m a c h i n e is d e m o n s t r a t e d with the help of simulated deterioration.

NOMENCLATURE B fl

flux density frequency, Hz

3 C. D. Rickson, Protecting motors from overload due to asymmetrical fault conditions, Electr. Rev., (Dec.) (1973) 777 - 780. 4 D. H. Ellison, J. L. T. Exon and D. A. Ward, Protection of slip-ring induction motors, IEE Conf. Publ., 185 (1980) 49 - 53. 5 D. J. T. Siyambalapitiya, P. G. McLaren and P. P. Acarnley, A rotor condition monitor for squirrel cage induction machines, Conf. Record o f IEE IAS Annu. Meeting, 1986, pp. 909 - 915. 6 J. L. Kohler, A decision-theoretic method for the classification of incipient failure patterns which are characteristic of mine-power system components, Ph.D. Dissertation, Department of Mining Engineering, Pennsylvania State Univ.,

1983. 7 J. L. Kohler, F. C. Trutt and L. A. Morley, Decision functions for electric power system signals, Electr. P o w e r S y s t . Res,, 11 (1986) 167 - 169. 8 J. Sottile, An experimental analysis of electrical deterioration in three-phase induetion motors, M.S. Thesis, Department of Mining Engineering, Pennsylvania State Univ., 1986. 9 P. L. Alger, Induction Machines, Gordon and Breach, New York, 2nd edn., 1970. 10 K. J. Binns and E. Schmid, Some concepts involved in the analysis of magnetic field in cage induction machines, Proc. Inst. Electr. Eng., 122 (1975) 169 - 175.