Experimental Studies of Circuit Breaker Drives and

Mechanisms Diagnostics

Zbigniew Pochanke, Waldemar Chmielak, Tadeusz Daszczyński

Warsaw University of Technology
Institute of Electrical Power Engineering
Warsaw, Poland
[email protected], [email protected], [email protected]

Abstract—Diagnostics of a circuit breaker is usually Presently, despite of longtime attempts, diagnostic is far
performed using a digital processing of voltage and current from the above guidelines due to the few reasons [8]:
signals, that gives the assessment of the mechanical work
delivered by the drives. Results of such systematic and long-term • lack of methodologies based on proper measurements
measurements, gathered for large enough population and associated with advanced numerical models of the
analyzed using statistic methods make possible rational mechanism,
management of circuit breaker (CB) assets. This Paper shows • weak interests of users in systematic deeper diagnostic
experimental studies and analysis of symptoms of changes in the tests,
state of mechanism comprised in motion characteristics made on • complete unconcern of manufacturer even to
medium voltage circuit breaker with introduced defects. functionality tests.

Keywords— high voltage circuit breaker, circuit breaker Reference [9] presents one methodology aimed to reveal
diagnostics,circuit breaker drives and mechanism. and assess mechanical resistance of the mechanism and
efficiency of the drive. In this approach the basic kinematic
quantity – apart from the displacement - is velocity, which can
I. INTRODUCTION (HEADING 1) be acquired in different way based on the measurements
The term diagnostics of a circuit breaker drives usually is driving force and of motion quantities. There are three basic
understood as verification of functionality of the circuit breaker methods to acquire velocity signal in tests of a mechanism:
[1]. Diagnosing of drives and mechanisms is important through differentiation, by a direct measurement and through
because, as operational statistics shows, the drives are the integration of the acceleration signal. In this paper the results
most unreliable part of the circuit-breaker [2-4]. In mechanical of experimental studies of the proposed methodology for
aspect in the recording travel of contacts it consists of, signals medium voltage circuit breaker with introduced defects are
of current in tripping coils and moment of contacts close/open also presented.
[5]. In time graph it is possible to read out series of parameters
which allow for assessing [6]:
DIAGNOSTICS OF CIRCUIT BREAKER DRIVES
II.
• ability of the circuit breaker to perform its switching AND MECHANISMS
duties: time of the operation, velocity at the contacts Basic way to identify motion of a CB is use of a
touch or separation, contacts stroke, contacts travel in displacement gauge fixed to specific characteristic element of
the close position, the mechanism (e.g. main shaft, driving link). At the moment,
• some aspects of the mechanism state, e.g. damping or practically the only gauges in use are potentiometers. Modern
presence and parameters of oscillations. designs, devoted to control system, provide smooth signals
However the above presented examination cannot be with high linearity. Velocity may be derived by differentiation
considered as a true diagnostics, because its results do not of the displacement signal this involves high sampling rate and
explain causes of an abnormal state. Diagnostics performed in high accuracy (high resolution – low disturbances); Authors
this situation should [7]: suggest combinations 20 kS/s – 12 bits, 50 kS/s – 10 bits, 100
kS/s – 8 bits as tentative data, which should be verified in
• follow and evaluate in terms of some quantities aging every particular case.
and/or degradation processes -friction, losses in driving
force, deformations of structure and elements. A. Differentiation of displacement signal.
• reveal irregularities in the motion - impacts, plays, Differentiation of the time series of displacement signal
seizure {xi} with si=s(ih) is combined with its smoothing by the local
• work out data for decisions related to the maintenance approximation trinomial over the range i:[-l:l]. This can be
and assets management. performed as finite impulse response (FIR) filtering.
Coefficients of the local trinomial square are: smoothed value
of the displacement, velocity and acceleration.

978-1-5090-5310-0/16/$31.00 ©2016 IEEE

 Effect of such a filtering, at given sampling rate strongly
depends on the length of the interval of approximation 2l+1
samples. Frequency analysis of this filtering seems to be
useless because Fourier transform of displacement signal does
not exists. Selection of the filter length should be better done
practically following the smoothness of the derived velocity.
Two cases presented in Fig.1 shows different roughness.
Despite shorter filter (higher pass frequency), up case has no
oscillations, the significant roughness was the testimony of
some impulse obstacle of the motion. In the graph of velocity
there are oscillation suggesting elastic deformation of some
link, support of the mechanism or vibrations of the gauge
fixing. In general – fast changes of the velocity observed only
in one signal cannot be interpreted unambiguously.

Fig. 2. Example of approximation of contact travel by means polynomial:

s(t)=a1t+a2t2+a3t3+a4t4, data 1 – original record, data 2 – approximation, data
3 error of approximation.

B. Direct measurement of the velocity

It requires 5 – 10 kS/s and 10 bits. Basic disadvantage of
this method is the absence of commercial velocity sensors
fitted to mechanism working in impulse regime. Principle of
such a sensor is simple – coil moving in a radial magnetic field
–is easy to built in laboratory conditions. From the point of
view of a range of information, direct measurement of velocity
is the best method. It is because of at mediocre recording
requirement, it offers exact data for displacement and control
over the frequency content of the acceleration signal (Fig. 3).
Unfortunately relatively high weight (100-200 G) makes
difficult its fixing or may cause vibrations of the gauge itself.

Fig. 1. Two cases of velocity traces (data 2) derived from the displacement
signal (data 1); sampling rate 100 kS/s; filters: up – 100 samples, down – 300
samples.

Radical smoothing of a displacement signal is in its Fig. 3. Examples of direct velocity signal (data 2) on the background of
approximation by a polynomial. Through many attempts velocity acquired for the signal of velocity (data 1).
Authors show, that almost in the whole contact travel – C. Integration of the acceleration signal.
excluding final period of mechanism stopping relatively low
order – high accuracy may be achieved by means 4-th order This method requires high dynamics of recording at least
polynomials - cf. Fig. 2. It is worthy to note, that coefficients 12 bits and sampling rate at least 10 kS/s. The velocity signal
of such approximation may be taken as characteristics of given acquired in this way is relatively precise, so exact enough for
state of mechanism. the calculation of the kinetic energy. Examples presented in
Fig. 4 and 5 show that this method leads to doubtful
displacement data.
 offers - here not observed – possibility to detect local, sharp
motion obstacle.

Fig. 4. Signal of acceleration and its integral as velocity in case of vibration

provoked by elastic deformation of the CB mechanism frame.
Fig.6. Motion characteristics (of contact displacement) in close operation;
symbols in annotation refer to the examined states of the mechanism: A – normal
state, B – screws fixing left pole untightened, C – middle pole detached.

Fig. 5. Signal of acceleration and its integral as velocity in case of semi

continuous (friction) disturbances of motion.
Fig.7. . Comparison of motion characteristics s(t) in state A: sA(tS) , and state
C: sC(tC) through function tC(tA).
III. EXPERIMENTAL STUDIES ON MEDIUM VOLTAGE CIRCUIT
BREAKER WITH INTRODUCED DEFECTS.
Tests were carried out for four state of mechanism of
minimum oil circuit breaker:
A: normal state,
B: screws fixing left pole untightened.
C: middle pole detached,
D: middle pole detached and screws fixing left pole
untightened
At each state there were made three tests with direct
measurement of contact velocity.
Analysis of symptoms of changes in the state of mechanism
comprised in motion characteristics was carried out. On the base
of direct form of motion characteristics (Fig. 6) only self-evident
differences - mainly concerning total, effective motion resistances
– can be detected. Quantitative analysis of symptoms of
mechanism changes comprised in motion characteristics can be
made either through analysis of time courses of velocity (Fig.8)
or by comparison of time necessary to reach subsequent positions Fig.8. Sets of three time courses (in second) of contact velocity (in
(Fig.7). The second method meter/second) in four states of mechanism.
 Fig.8. and Fig.9. suggest that velocity characteristics are
repeatable to significant extend, although sets with tightened
pole seem to give more stable results. To verify possibility do
detect a disintegration of mechanism structure (here modeled
by untightened pole) from the velocity courses there were
extracted fast components (irregular and random):
• irregular component is evaluated on the base of slow
velocity component obtained through the
approximation vapp(t) of recorded velocity vrec(t) -
fig.10 and virr(t) = vrec(t) - vapp(t) - cf. fig.11.
• random component is evaluated on the base of average
velocity vave(t) (of three time courses): v rnd(t) = vrec(t) -
vave(t) - cf. fig.12.

Fig.11. Derivation of irregular component of velocity - virr(t) as difference

between velocities recorded - vrec(t) and approximated - vapp(t).

Fig.9. Exemplary displacement-velocity characteristics in three states

of mechanism

Fig. 12. Derivation of random component of velocity - vrnd(t) as difference

between velocities recorded - vrec(t) and averaged - v ave(t); each plot
comprises three tests.

Quantitative comparison of effects of changes in

mechanism state on the signals defined above may be obtained
as sum of their squares SQ:
Fig.10. Approximation of time-displacement s(t) and its effect on the
time-velocity characteristics
TABLE I. SQV OF IRREGULAR COMPONENT

Left pole tightened Left pole untightened

Three poles A: 438.2 B: 431.0
Two poles C: 476.0 D: 467.4

Fact, that in the normal state A index SQV is not the

smallest one makes it useless as a symptom of failure.
 TABLE II. SQR OF RANDOM COMPONENT models of drives and mechanisms. Within this idea this paper
Left pole tightened Left pole untightened proposes analysis of the work-energy balance. Development of
Three poles A: 31.1 B: 43.3 this approach requires:
• Commercial systems of diagnostics should be equipped
Two poles C: 47.8 D: 42.2
with procedures allowing to identify numerically
motion characteristics and its derivatives; at the present
Supposing that disintegration of mechanism causes state of measurement and recording art, software shows
its random functioning so it is expected that values in the left lack of methodologies based on the proper
column (states A and C) will be less than in the right (states B measurements associated with advanced numerical
and D). This is through for typical structure of the CB. It raise models of the mechanism,
hope of possible feasibility of this approach.
• It is advisable to support those systems with hardware
Results of the pole top acceleration were analyzed to allowing faster and more precise A/D converter,
supplement presented research. In this case none of earlier • manufacturers share the interests with the users in
data processing was needed to reveal parasitic vibration of the efforts to improve assets management and open their
pole, which may comprise symptoms of the state of data and expertise on diagnostics
mechanism – Fig. 13. • Measurement of displacement characteristics s(t)
Quantitative comparison of effects of changes in carried out with high accuracy (time and magnitude
mechanism state on the signals of pole top acceleration is resolution) and properly processed offers possibility to
given as a sum of their squares SQA: diagnose state of CB mechanisms in aspects of general
resistances of motions.
TABLE III. SQA OF POLE TOP ACCELERATION • Proper signal processing must provide approximation
Left pole tightened Left pole untightened of displacement and velocity as its derivative.
Three poles A: 141 B:154
• Inferences on the processes of disintegration of the
mechanism requires dynamic model of the mechanism,
Two poles C: 248 D: 157 which allow for distinguishing in velocity signal
between vibration associated with elastic, casual
Results are similar to the case of SQR. deformations of the structure and components caused
by random displacement.
REFERENCES
[1] W. Chmielak, Z. Pochanke, “Diagnostic of vacuum on the basis of the
dielectric strength” Electrical Review 12b/2012, pp. 311-314.
[2] Working Group A3.06 CIGRÉ “Final Report of the 2004 – 2007
International Enquiry on Reliability of High Voltage Equipment. Part 1 -
Summary and General Matters.” No. 509, 2012.
[3] Working Group A3.06 CIGRÉ (2012) “Final Report of the 2004 – 2007
International Enquiry on Reliability of High Voltage Equipment. Part 2 -
Reliability of High Voltage SF6 Circuit Breakers.” No. 510, 2012.
[4] S.M. Strachan, S.D.J. McArthur, B Stephen, J.R. McDonald, A Cambell,
“Providing Decision Support for the Condition-Based Maintenance of
Circuit Breakers Throught Data Mining of Trip Coil Current Signatures.”
IEEE Trans. Of Power Delivery, vol. 22, No. 1, 2007.

[5] Working Group 13.09 CIGRÉ “User guide for the application of
monitoring and diagnostic techniques for switching equipment for rated
voltages of 72,5 kV and above.” No. 167, 2010.
[6] J. Maksymiuk,Z. Pochanke, “Calculations and diagnostic testing of
electrical apparatus” WNT Warszawa, 2001 (in Polish).
[7] Z. Pochanke “Reliability diagnostic of high voltage circuit breakers.”
Fig. 13. Sets of three time courses (in second) of acceleration of the pole (in Electrical Review Z.2, 1993, pp. 29-32 (in Polish).
arbitrary units) in four states of mechanism. [8] Z. Pochanke “Drive systems diagnosis of medium voltage circuit
breakers.” IV Conf. Diagnostics in industrial power networks, 2003, pp.
IV. CONCLUSIONS. 21-25 (in Polish).
[9] W. Chmielak, Z. Pochanke “Diagnostics of Circuit Breaker Drives and
It is possible and reasonable to introduce the usage of CB Mechanisms” International Conference on Condition Monitoring,
diagnostic methodologies associating with measurements and Diagnosis and Maintenance - CMDM 2013, Romania, Sept. 2013 pp.
101-108.