You are on page 1of 8

Electric Power Systems Research 53 (2000) 23 30

www.elsevier.com/locate/epsr

Wavelet analysis based scheme for fault detection and


classification in underground power cable systems
W. Zhao a, Y.H. Song a,*, Y. Min b
b

a
Department of Electrical Engineering and Electronics, Brunel Uni6ersity, Uxbridge UB8 3PH, UK
Department of Electrical Engineering, Tsinghua Uni6ersity, Bejing 100084 , Peoples Republic of China

Received 6 August 1998; accepted 22 December 1998

Abstract
This paper presents a new method for detecting and classifying fault transients in underground cable systems based on the use
of discrete wavelet transform. A 400 kV underground cable system is simulated by ATP/EMTP (electro-magnetic transients
program) under various system and fault conditions. Daubechies eight wavelet transform is employed to analyze fault transients
for the development of a novel fault detection and classification scheme. Simulation results show that this scheme is effective and
robust. 2000 Elsevier Science S.A. All rights reserved.
Keywords: Wavelet analysis; Fault detection; Underground power cable systems

1. Introduction
For various reasons, there is an increasing tendency
towards the use of underground cables at all voltage
levels. This inevitably presents additional problems to
power transmission protection since the underground
cables have different electrical properties from those of
the overhead lines. No matter how well the cables are
designed and manufactured, underground cable systems
are prone to numerous hazards which may lead to their
failure. In most of the power system relay schemes, the
first step always involves fault detection and classification. The information provided by this step is essential
for the fault location scheme.
It is widely known that faults on underground cable
systems produce a wide range of frequency components. Previous studies on distorted waveform generated by cable faults mainly use Fourier Transform [1].
The basis functions used in Fourier analysis, sine waves
and cosine waves, are precisely located in frequency,
but exist for all time. Although the short-time Fourier
Transform (STFT) which windows the input signal
overcomes the time location problem to a large extent,
* Corresponding author. Fax: +44-1895-258728.
E-mail address: y.h.song@brunel.ac.uk (Y.H. Song)

it does not provide multiple resolution in time and


frequency because the window is fixed. As the frequency is increased, there are more and more cycles
within the windows. Thus, individual frequency components are not treated in the same way. Wavelets, however, can be chosen with very desirable frequency and
time characteristic as compared to Fourier techniques.
In contrast to the STFT, which uses a single analysis
window, the wavelet transform uses short windows at
high frequencies and long windows at low frequencies.
Thus the windowing of wavelet transform is adjusted
automatically for low or high frequencies and each
frequency component gets treated in the same manner
without any reinterpretation of the results.
The wavelet analysis procedure is to adopt a wavelet
prototype function, called a mother wavelet. Temporal
analysis is performed with a contracted, high frequency
version of the prototype wavelet, while frequency analysis is performed with a dilated, low frequency version
of the prototype wavelet. Wavelet has been successfully
applied in several fields, such as image and signal
analysis, radar, acoustics, and mechanical vibrations.
Recently, several papers have been presented proposing
the use of wavelets for power system analysis [28].
The aim of this paper is to investigate the application
of wavelets in cable fault transients analysis and thus to

0378-7796/00/$ - see front matter 2000 Elsevier Science S.A. All rights reserved.
PII: S 0 3 7 8 - 7 7 9 6 ( 9 9 ) 0 0 0 3 3 - 4

W. Zhao et al. / Electric Power Systems Research 53 (2000) 2330

24

provide a new concept on cable faults detection and


classification, based on some initial results reported in
ref. [8]. Section 2 contains a brief introduction to
wavelet and the discrete wavelet transform adopted in
the paper. The results presented in this paper relate
to a typical 400 kV underground cable system, the
faulted responses of which are attained using ATP. In
Section 3, fault transients are analyzed by using
wavelet transform. Then a novel fault detection and
classification scheme based on discrete wavelet transform analysis, together with some test results, is presented in Section 4.

2. Theory of wavelet analysis


Wavelets are functions that satisfy certain requirements. The very name wa6elet comes from the requirement that they should integrate to zero, waving
above and below the x-axis. The diminutive connotation of wa6elet suggests the function has to be well
localized. Other requirements are technical and
needed mostly to insure quick and easy calculation of
the direct and inverse wavelet transform. Compared
with traditional Fourier method, there are some important differences between them. First Fourier basis
functions are localized in frequency but not in time
while wavelets are localized in both frequency (via
dilation) and time (via translation). Moreover,
wavelets can provide multiple resolution in time
and frequency. Second, many classes of functions
can be represented by wavelets in more compact
way. For example, functions with discontinuities
and functions with sharp spikes usually take substantially fewer wavelet basis functions than sine-cosine
basis functions to achieve a comparable approximation.
There are many types of wavelets [9,10], such as
Harr, Daubechies 4, Daubechies 8, Coiflet 3, Symmlet
8 and so on. One can choose between them depending on a particular application. As with the discrete
Fourier transform, the wavelet transform has a digitally implementable counterpart, the discrete wavelet
transform (DWT). If the discrete analysis is pursuing
on the discrete time, the DWT is defined as
C( j, k)= % s(n)gj, k (n)
nZ

(j  N, k  Z)

(1)

where, s(n) is the signal to be analyzed and gj,k (n) is


discrete wavelet function, which is defined by
j/2
j
gj, k (n)= a
g(a
0
0 n kb0)

(2)

Select a0 and b0 carefully, the family of scaled and


shifted mother wavelets constitute an orthonormal basis of l 2(Z) (set of signals of finite energy). When

simply choose a0 = 2 and b0 = 1, a dyadic-orthonormal wavelet transform is obtained. With this choice,
there exists an elegant algorithm, the multiresolution
signal decomposition (MSD) technique [11], which
can decompose a signal into levels with different time
and frequency resolution. At each level j, approximation and detail signals Aj, Dj can be built. The words
approximation and detail are justified by the fact
that Aj is an approximation of Aj 1 taking into account the low frequency of Aj 1, whereas the detail
Dj corresponds to the high frequency correction. The
original signal can be considered as the approximation at level 0.
The coefficients C ( j,k) generated by the DWT are
something like the resemblance indexes between the
signal and the wavelet. If the index is large, the resemblance is strong, otherwise it is slight. The signal
then can be represented by its DWT coefficients as
s(n)= %

% C( j, k)gj, k (n)

(3)

jN kZ

When fix j and sum on k, a detail Dj is defined as


Dj (n)= % C( j, k)gj,k (n)

(4)

kZ

Then sum on j,the signal is the sum of all the details


s(n)= % Dj (n)

(5)

jN

Take a reference level called J, there are two sorts


of details. Those associated with indices j5 J correspond to the scales 2 j 5 2J, which are the fine details.
The others, which correspond to j \J, are the coarser
details. If these latter details are grouped into
Aj = % Dj

(6)

j\J

which defines an approximation of the signals. Connect the details and an approximation, the equality
s= AJ + % Dj

(7)

j5J

which signifies that s is the sum of its approximation


AJ and of its fine details. The coefficients produced
by DWT, therefore, can be divided into two categories: one is detail coefficient, the other is approximation coefficient. To obtain them, MSD provides an
efficient algorithm known as a two channel sub-band
coder using quadrature mirror filters [12]. Then the
detail part is still represented by wavelets, which can
be regarded as series of band-pass filters, whereas the
approximation is represented by the dilation and
translation of a scaling function, which can be regarded as a low-pass filter.

W. Zhao et al. / Electric Power Systems Research 53 (2000) 2330

25

Fig. 3. Section connection of a crossbonded cable.


Fig. 1. Cross-section.

3. Wavelet analysis of fault transients

3.1. Cable system simulation


Single-conductor coaxial cable (Fig. 1) is generally
used for EHV applications because they are the way of
carrying the necessary load current and maintaining the
insulation requirements. The underground cables modeled in this paper are based on 400 kV oil-filled cables.
Fig. 2 illustrates a model system. A 400 kV source
with the frequency of 50 Hz is applied to the system.
Three phase balanced loads are presented in the system.
The cables are buried in a horizontal (flat) configuration 1 m below the surface. The 4.8 km long cable line
is split into four of 1.2 km major sections, each of
which consists of three minor sections. At the end of
each minor section, the core of phase a is connected to
the core of phase b, the core of phase b is connected
to the core of phase c, and core of phase c is
connected to the core of phase a(Fig. 3). The cables
are modeled using the cable constants routine of the
ATP and the parameter is simulated using J. Marti,
frequency dependent traveling wave model. Based on
this cable system, transients of four types of fault
(single-phase to ground, two- phase fault, two-phase to
ground and three-phase fault) under different fault
conditions are simulated by ATP.

3.2. Transients analysis


By using wavelets analysis, sub-band information can
be extracted from the simulated transients, which contain useful fault features. By analyzing these features of
the detail signals, different types of fault can be detected and classified. As mentioned earlier, the choice of
analyzing wavelets plays a significant role in fault detec-

Fig. 2. Configuration of a 400 kV cable system.

tion and identification. Since Daubechies 8 is localized,


i.e. compactly supported, in time, it is good for short
and fast transients analysis. After examinations of several types of wavelet, Daubechies 8 is chosen in this
scheme.
Fig. 4 illustrates the detailed wavelet analysis results
of faulty phase-a current for a phase-a to ground fault,
which happened at the left 200 m of section 2 (D=
0.125 pu) with the fault resistance (R) of 5 V. The three
phase current waveforms are original signals of Bus1
simulated by ATP. Each original signal has 2048 sample points, which means the total simulation time duration is 51.2 ms with a sampling time of 0.025 ms. Since
2048= 211, when they are decomposed into Daubechies
8 wavelet components, there are 11 levels. From the
results of an extensive series of studies using wavelet
analysis, the time and frequency localization characteristics of wavelet decomposition of fault transients associated with different fault conditions have been
obtained. It is well known [13] that when a fault occurs
on an underground cable, different frequency components are produced. The wavelet levels to be selected
must best reflect the fault characteristics under various
system and fault conditions. In this respect, according
to the analyses of different wavelet levels of current
waveform as shown in Fig. 4, the level 4 (D4) and level
9 (D9) details, are utilized to extract some useful features. This is because the level 4 details generally reflect
the dominant non-frequency transient generated by
faults. Since level 9 details contain most of the fundamental harmonic, which is of 50 Hz in this system, the
sum of three phase of them (D9a, D9b, and D9c) have
similar characteristics of zero component which can be
used to differ phase-to-ground fault and phase-phase
fault, two-phase to ground fault and three-phase fault.
These two level details of three phase currents for an
a-phase-to-ground fault are presented in Fig. 5. Also
shown is the sum of three phase level 9 details, D9.
Figs. 68 present wavelet analysis results of other types
of fault, which are two-phase fault, two phase to
ground fault and three phase fault, respectively. It can
be seen that the level 4 detail signals in Figs. 58 are
very small during the normal operation time. The
spikes which are seen at the initial and final edges are

26

W. Zhao et al. / Electric Power Systems Research 53 (2000) 2330

caused by discontinuities or edge effects inherent in the


analysis process that can be easily eliminated by using
proper window scheme. When a fault happened, several
spikes can be seen in the level 4 signals, which is due to

Fig. 5. Wavelet analysis of three phase currents for an A G fault:


D =0.125 pu, R =5 V.

the high harmonics generated in the fault conditions.


By detecting the fault generated spikes, fault can be
detected. Figs. 58 show some differences among these
four types of fault as well. It is apparent that these
faults have their own characteristics in the level 4
details of three phase (D4a, D4b, D4c) and the sum of
three phase level 9 (D9) details, which can be successfully used to classify different types of fault, as will be
seen in next section.

4. Fault detection and classification scheme


Fig. 4. Wavelet analysis of phase-a current for an AG fault:
D =0.125 pu, R= 5 V.

After analyzing numerous fault transients simulated


by ATP using wavelet transform, several fault related

W. Zhao et al. / Electric Power Systems Research 53 (2000) 2330

27

features have been identified. Based on these information, a novel fault detection and classification scheme as
shown in Fig. 9 is presented, which has been summarized as the following 20 rules.
1. Variable initialization.
2. Calculate three phase level 4 details:
D4p,

p(a, b, c)

3. For any one of the three phases, if there are three


consecutive absolute values (excluding the spikes in
the initial and final edges) of level-4 details are
greater than t1, claim a fault. Otherwise move
analysis window and go to step (1).
4. Calculate the maximum absolute value (excluding
the spikes in the initial and final edges) of level 4
details in each phase:
xp = max( D4p(i) ),

p= {a, b, c}

Fig. 7. Wavelet analysis of three phase currents for an A B G fault:


D =0.125 pu, R =5 V.

Then, obtain the maximum one of the three: x=


max (xa, xb, xc).
5. Check which phase x belongs to. That is, if x=xp,
let k= ip, ip is the index of xp in vector D4p, p
stands for phase a, b or c.
6. Let
Ep(n)= D4p(k+n 1)
where p = {a, b, c}, n= 1, 2, 3. Calculate the sum
of Ep (1), Ep (2), and Ep (3):
5

Fig. 6. Wavelet analysis of three phase currents for an AB fault:


D =0.125 pu, R= 5 V.

ep = % Ep(n)
n=1

W. Zhao et al. / Electric Power Systems Research 53 (2000) 2330

28

p (a, b, c)
7. Let e=max (ea, eb, ec). Calculate the ratio of each
phase: rp = ep/e, where p={a, b, c}.
8. Calculate the sum of three phase level 9 details:
D9 = D9a + D9b + D9c, let
md9= max( D9(i ) ).
9. If
ra rb rb rc \t2,

ra rc rb rc \ t2

and md9\t3, its an A G fault.


10. If
rb rc ra rc \t2,

rb ra ra rc \t2

Fig. 9. Procedure of fault detection and classification scheme.

and md9\ t3, its a BG fault.


11. If
rc ra rb ra \ t2,

Fig. 8. Wavelet analysis of three phase currents for a three phase


fault: D= 0.125 pu, R= 5 V.

12.
13.
14.
15.
16.
17.
18.

rc rb rb ra \ t2

and md9\ t3, it is a CG fault.


If ra \ t4, rb \ t4 and md9B t3, it is an AB fault.
If rc \ t4, rd \ t4 and md9Bt3, it is a CB fault.
If ra \ t4, rc \ t4 and md9B t3, it is an AC fault.
If md9B t3, it is a three-phase fault.
Let r= min (ra, rb, rc).
If r=ra, it is a BCG fault.
If r= rb, it is an ACG fault.

a=90

Fault conditions

a =45

a =0

Type

D (pu)

R (V)

ra

rb

Rc

md9 (A)

ra

Rb

rc

md9 (A)

ra

Rb

rc

md9 (A)

AG

0.125
0.125
0.875
0.875

5
500
5
500

1.00
1.00
1.00
1.00

0.000
0.001
0.000
0.001

0.000
0.002
0.000
0.001

3.5104
5.2102
3.4104
5.1102

1.00
1.00
1.00
1.00

0.007
0.003
0.002
0.000

0.007
0.001
0.002
0.001

3.8104
5.4102
3.6104
5.4102

1.00
1.00
1.00
1.00

0.002
0.013
0.000
0.012

0.002
0.007
0.001
0.007

3.8104
5.2102
3.7104
5.2102

AB

0.125
0.125
0.875
0.875

5
500
5
500

.998
1.00
1.00
1.00

1.00
1.00
0.999
0.999

0.002
0.001
0.000
0.000

2.235
1.186
2.232
1.173

.998
.999
1.00
1.00

1.00
1.00
0.997
0.999

0.005
0.001
0.001
0.000

2.211
1.189
2.333
1.172

0.997
1.00
0.998
1.00

1.00
0.997
1.00
0.998

0.001
0.001
0.001
0.001

2.202
1.184
2.397
1.172

ABG

0.125
0.125
0.875
0.875

5
500
5
500

1.00
1.00
1.00
1.00

0.602
0.983
0.600
0.982

0.001
0.002
0.000
0.000

2.7104
3.7102
2.6104
3.7102

.677
.973
.687
.980

1.00
1.00
1.00
1.00

0.005
0.006
0.001
0.001

2.7104
3.7102
2.6104
2.6104

0.238
1.00
0.202
1.00

1.00
0.949
1.00
0.941

0.012
0.002
0.000
0.001

2.8104
3.8102
2.7104
3.8102

Three-phase

0.125
0.125
0.875
0.875

5
500
5
500

1.00
1.00
1.00
1.00

0.528
0.528
0.530
0.530

0.470
0.471
0.469
0.469

20.65
2.641
16.27
2.640

.771
.768
.768
.766

1.00
1.00
1.00
1.00

0.236
0.234
0.237
0.235

21.11
2.647
16.16
1.705

0.156
0.156
0.150
0.151

0.848
0.848
0.850
0.851

1.00
1.00
1.00
1.00

20.99
2.592
15.48
2.737

W. Zhao et al. / Electric Power Systems Research 53 (2000) 2330

Table 1
Ratios in different types of fault

29

30

W. Zhao et al. / Electric Power Systems Research 53 (2000) 2330

19. If r = rc, it is an AB G fault.


20. End.
t1, t2 t3 and t4 are thresholds which are selected according to the detail values in normal and fault operations
in different cable systems. In this paper, t1 = 1, t2
= 0.9, t3 =100 and t4 =0.99.
When the proposed technique is used as part of a
protective relay, an appropriate data window length
must be adopted. Two important aspects must be considered: the window contains enough information to
obtain the required wavelet levels and it is short enough
to produce the expected speed. Initial studies suggest
that the proposed technique, based on the two wavelet
levels, should be able to have acceptable performance in
terms of speed and accuracy. However, detailed studies
are required to fully assess the speed and security
requirements of a protective relay. This is out of the
scope of the paper.
Since the fault resistance, fault location and fault
inception angle have important effects on analysis results, they have been considered in this scheme. Some
ratios associated with different fault conditions are
presented in Table 1 which clearly show that this
scheme can be used to identify different types of fault
effectively in underground cable systems. The parameters D and R in the table stand for fault location and
resistance, respectively, ra, rb, rc and md9 have already
been defined above in the rules.
Suppose an AB fault happened, which is in left 200
m of section 11 (D =0.875pu) of the cable line and the
fault resistance of it is 5 V, the parameters are obtained
from Table 1 as follows: ra =1.000, rb =0.999, rc =
0.000 and md9= 2.232. They are obviously not satisfactory to the rule (9), (10) and (11) which are used to
identify single-phase fault in which two of the three
ratios (ra, rb, rc) need to be very small and be almost
equal while another ratio is always equal to one. The
md9 of single-phase fault is always very large compared
with two-phase fault and three-phase fault. The
threshold t3(100) is set to define the large and the
small of md9. Since two ratios associated with two
faulty phases in two-phase fault are almost equal to
1.000, seen from the example presented above (ra =
1.000, rb =0.999), threshold t4(0.99) is set to distinguish
this type of fault from others to some degree. When
two of the three ratios are both greater than t4 and the
value of md9 (2.232 in the example above) is less than
t3, it is a two-phase fault. The ratios of two-phase to
ground fault and three-phase fault change greatly under
different fault conditions which bring difficulties to
identify these two types of fault. Fortunately, they have
different characteristics in the sum of level 9 details,
md9 can be used to distinguish them from each other. If
three ratios of a fault are not satisfactory to rules
(9) (14), that is to say, it is not a single-phase fault nor
a two-phase fault, then check the value of md9. When

md9 is less than t3, it is a three-phase fault, otherwise it


is a two-phase to ground fault. The healthy phase in
two-phase to ground fault can be found easily because
the ratio of it is the smallest one among the three.

5. Conclusions
This paper proposes a novel fault detection and
classification scheme based on discrete wavelet analysis
for power cables. The property of multiresolution in
time and frequency provided by wavelets allows an
accurate time location of fault transients while simultaneously retaining information about the fundamental
frequency and its high-order harmonics, which is powerful to extract characteristics of different types of fault
in underground cable systems. Numerical test results of
the scheme are very encouraging. It must be pointed
out that it is necessary to fully evaluate the proposed
technique as part of protection relays or fault locators
under a wide range of system and fault conditions.
These are currently under investigation.

References
[1] Z.Q. Bo, Q.Y. Xuan, R.K. Aggarwal, A.T. Johns, H.Y. Li, P.J.
Moore, A neural network based protection and fault location
scheme for power distribution cables, APSOM95, Hong Kong,
(1995).
[2] W.A. Wilkinson, M.D. Cox, Discrete wavelet analysis of power
system transients, IEEE Trans. Power Syst. 11 (4) (1996) 2038
2044.
[3] P. Pillay, A. Bhattacharjee, Application of wavelets to model
short-time power system disturbances, IEEE Trans. Power Syst.
11 (4) (1996) 2031 2037.
[4] D.C. Robertson, O.I. Camps, Wavelets and electromagnetic power
system transients, IEEE Trans. Power Deliv. 11 (2) (1996) 1050
1056.
[5] J. Liang, S. Elangovan, J.B. Devotta, A wavelet multiresolution
analysis approach to fault detection and classification in transmission lines, Elec. Power Energy Syst. 20 (5) (1998) 327 332.
[6] S. Santoso, E.J. Power, W.M. Grady, P. Hofmann, Power quality
assessment via wavelet transform analysis, IEEE Trans. Power
Deliv. 11 (5) (1996) 924 930.
[7] I.K. Yu, Y.H. Song, Development of novel adaptive single-pole
autoreclosure schemes for extra high voltage transmission system
using wavelet transform analysis, Elec. Power Syst. Res. 47 (1)
(1998) 11 20.
[8] W. Zhao, Y.H. Song, I.K. Yu, Wavelet analysis of fault generated
transients in power cables, Proceedings of UPEC, UK, 1998.
[9] A. Graps, An introduction to wavelets, IEEE Comp. Sci. Eng.,
2 (2) (1995) 50 61.
[10] A.W. Galli, G.T. Heydt, P.F. Ribeiro, Exploring the power of
wavelet analysis, IEEE Comp. Appl. Power (1996) October 3741.
[11] S.G. Mallat, A theory for multiresolution signal decomposition:
the wavelet representation, IEEE Trans. Pattern Anal. Machine
Intelligence 11 (7) (1989) 674 693.
[12] M. Vetterli, C. Herley, Wavelets and filter banks: theory and
design, IEEE Trans. Signal Proc. 40 (9) (1992) 2207 2232.
[13] G.W. Swift, The spectra of fault-induced transients, IEEE Trans
PAS 98 (3) (1979) 940 947.

You might also like