0% found this document useful (0 votes)
6 views

Papers

Uploaded by

Shailesh Patil
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
6 views

Papers

Uploaded by

Shailesh Patil
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

1

Circulating currents in closed loop structure, a


new problematic in distribution networks
M. Loos, Student Member, IEEE, S. Werben, Siemens AG, and J.C. Maun, Member, IEEE

protection algorithms could malfunction due to such zero


Abstract-- Loop operation of distribution networks is more and sequence currents. Lot of protection algorithms dealing with
more common. The increasing penetration of distributed the zero sequence measurement supposes that there is no
generation and higher power quality requirement are making the interaction between the load and the zero sequence system. For
loop structure more interesting. We study the impact of the example, some devices measure the active zero sequence
unbalanced phase impedances. Measurements and simulations current level to detect if the feeder is faulty. They consider that
show that it creates a mutual coupling between the positive
there is only capacitive current on the sound feeder in the zero
sequence and the zero sequence system. Balanced load creates a
zero sequence voltage between two feeders. Their connection of sequence system. But in case of circulating current, active
the both ends to form a loop results in circulation of a zero current will be measured on sound feeder too.
sequence current. In case of an earth fault this current will be This paper explains the phenomena of unbalanced serial
added to the fault current and can negatively influence the impedance that could produce such circulating current. Firstly,
protection devices. Healthy lines do not have active current in the we present the theoretical context of the circulating current and
zero sequence system because there is only capacitance connection explain how they are produced in a simple loop and multi loop
with the earth but the circulating current change this hypothesis. systems. Recordings of earth fault in 20kV nominal voltage in
We propose a study of this problematic current for simple and distribution network in Germany are presented. This network
multi loop in distribution networks.
is a compensated network with a Petersen coil; also known as
Index Terms-- Circulating current, earth fault, isolated
resonant network. This network has a loop of two feeders and
network, compensated network, closed loop a loop of four feeders. Then simulations have been made with
the ATP/EMTP software. Based on a distributed line model of
I. INTRODUCTION overhead lines and cables, the coupling between the positive
and zero sequence system has been reproduced. The network
D istribution networks in Europe have been changing since
several years. The increasing demand for distributed
studied is based on the real network parameters. Finally, some
solutions to avoid the malfunctioning are suggested in the
generation systems is changing the philosophy of the topic of transient relays in isolated and compensated networks
distribution network use. The previous passive contribution of [7].
the network is turning into a more active one with a real
contribution in the electricity production [1]. Generally, most II. DISTRIBUTION NETWORKS WITH CLOSED LOOP STRUCTURE
of the distribution networks are operated in open loop due to
security and protection limits. However the increased demand Networks can have different kinds of loop. The more
for a better power quality makes the distribution networks common one is the loop made of two feeders. Generally, there
operate more and more in closed loop mode instead of open are two feeders feeding the same area and connected at their
loop mode [2], [3], [4]. both ends. The others loops are called multi loop and contains
In Europe, the use of three phases transformers and their more than two feeders. These feeders are also connected at
connection in delta creates no or a very small zero sequence their both ends and are built to increase the power capacity of
current during healthy operation. However, the closed loop the line. This paper investigates these two kinds of loop and
structure might produce circulation of zero-sequence current their differences.
because of unbalanced serial impedance in the line [1], [5] or A. Closed loop distribution networks
coupling with parallel lines [6]. State of the art earth fault
In Germany, distribution network voltage is between 6 kV
and 30kV. Usually they are operated in radial mode or with
The authors gratefully acknowledge the support of Siemens AG and loop. In such networks, underground cables are often replacing
Pfalzwerke Netzgesellschaft GmbH for the recordings. the overhead lines. These cables are usually buried in trefoil
M. Loos is PhD student in the Department of Electrical Engineering,
position or in flat position. The impedance of a three phase
BEAMS Energy, Université Libre de Bruxelles, Brussels, Belgium (e-mail:
[email protected]). feeder can be expressed by the equation (1) with the matrix
J.-C. Maun is Professor in the Department of Electrical Engineering, Zabc.
BEAMS Energy, Université Libre de Bruxelles, Brussels, Belgium (e-mail:
[email protected]).
S. Werben is with Siemens AG, Energy Sector, Power Distribution
Division, 59 Humboldstr. 90459 Nurnbeg, Germany. (e-mail:
[email protected])
2

Z Z Z
L1 M 12 M 13
abc
Z Z Z Z 1
M 21 L2 M 23
Z Z Z
M 31 M 32 L3
The matrix Zabc represents the impedance series and mutual
coupling of a three phase conductor. ZM12 represents the
impedance between the phase L1 and the phase L2. ZL1
Fig. 1: a) Single conductors in trefoil
represents the impedance of the phase L1. This impedance can
b) Single conductors in parallel
be described in the symmetrical component with the
transformation of the matrix with the matrix A. If the mutual coupling is not neglected as in equation (6),
1 1 1 then equation (7) proves that a positive sequence current will
A 1 a2 a 2 create a zero sequence voltage. This coupling is represented in
Fig. 2 by M01.
1 a a2
U0 I0
with the symbol a; a phase angle of 120°. 012
U1 Z I1 7
j2
a e 3 3 U2 I2

The impedance matrix of the symmetrical components is Fig. 2 illustrates that even in sound operation a zero
calculated with equation (4): sequence voltage can be measured only because a perfectly
balanced load current is supplied.
012 1 abc
Z A Z A 4

Usually, the mutual coupling between the phase impedance


is considered as the same for every phase. With such
assumptions, the matrix Z012 is diagonal and there is no
coupling between the symmetrical systems.
Z0 0 0
012
Z 0 Z1 0 5
0 0 Z2
where Z0, Z1 and Z2 are respectively the zero-sequence
impedance, the positive sequence impedance and the negative
sequence impedance
The position of the cables influences the serial impedance
of the feeder. The mutual coupling between the phases is the
same if the cables are laid in a trefoil position as shown in Fig.
1 b). This is obviously not the same in case of Fig. 1 a) where
the cables are laid in a parallel position. They have an
unbalanced impedance matrix. The same problem can occur on
overhead lines but usually the phases are transposed [8], [9]. Fig. 2: The positive sequence system and the zero sequence system are
connected by mutual impedance. Zero sequence voltage is measured during
The unbalance in the serial impedance of the feeder creates healthy operation.
a mutual coupling in the matrix impedance. The zero sequence
and the positive sequence systems are connected even during During healthy operation, the zero-sequence voltage and
healthy condition. Generally, the symmetrical component current are very small in European distribution networks
impedance matrix is diagonal as in equation (5) but if it is because of the use of three phase transformers in the MV/LV.
assumed that ZM12, ZM13 and ZM23 are different a coupling A lot of protection devices use the zero-sequence components
between the systems appears. The matrix Z012 is not diagonal to detect an earth fault. Presence of zero sequence voltage
anymore as stated by equation (6). (U0) can be a criterion for the detection of an earth fault in the
Z 0 M 01 M 02 network. The unbalance of the phases is creating a small zero
012 sequence voltage that is not very high. In radial networks, only
Z M 10 Z 1 M 12 6
M 20 M 21 Z 2 a small capacitive current is produced because U0 applied on
the phase to earth capacitance of the feeder.
3

Fig. 3: Connection of two feeders with unbalanced phase impedance causes


circulating current.

In case of loop structure caused by e.g. connecting two


feeders, a zero sequence voltage is applied between the two
feeders and a zero sequence current can circulate in the loop in
addition to the capacitive current. This explanation is
illustrated on Fig. 3 for a closed loop of two feeders in a
compensated network. Only because of the serial impedance of
the lines, such currents may show amplitudes up to several
Amps primary. Such values are enough to negatively influence
fault detection algorithms.
Fig. 2 and Fig. 3 prove that even a perfect positive
sequence system with balanced load can produce zero
sequence current, which is usually considered as an indicator
for earth faults.
The circulating current depends on the positive sequence Fig. 4: In addition to mutual coupling between the positive sequence system
load and the unbalance of the serial phase impedances. The and the zero sequence system, a mutual coupling in the zero sequence system
itself is also to consider in the multi loop system.
unbalance does not change during the life of the network and
the load is generally very stable; it does not change within a
short period of time. The amplitude of this current will be
III. FIELD MEASUREMENTS
relatively constant compared to the evolution of an earth fault.
Measurements in a German compensated network have
B. Multi loop networks been recorded in the purpose to obtain data for single phase
Loop structures are more and more used; some distribution earth fault. There circulating currents have been measured.
system operators now have loops of more than two feeders. Some results are shown in this paper. The aim of the
This is sometimes due to a needed increase of the power recordings was to record fault information but circulating
capacity. E.g. several feeders are contained in the loop and currents are clearly visible before the fault inception.
connected to the same underlying busbar. In case of The recordings have been done by a digital fault recorder
underground cables, the feeder cables can be placed in the SIMEAS-R. The recordings of the zero sequence voltage show
same trench. The cables of different feeders might be placed in no voltage before the fault. The SIMEAS-R has a threshold
parallel for kilometers with a distance of around half a meter to level and if the voltage is below the threshold no recording
each other. The mutual zero sequence coupling between the takes place. According to the simulation, the voltage is small
loops is no more negligible. The current circulating in a ring (around 20V primary) before the fault inception which is
has an impact on the other rings. Higher amplitudes of the below the recording level of the SIMEAS-R. The current is
current can be expected in case of such multi loop systems. circulating only in the serial impedance of the loop which is
The Fig. 4 illustrates this mutual coupling between the small. A very small voltage is then needed to make the current
feeders of the same loop and the coupling with the positive circulate.
sequence system in a compensated network.
A. The network
The network is a compensated distribution network
operated by Pfalzwerke Netzgesellschaft GmbH. The nominal
voltage is 20kV. The network consists of two busbars, each
supplied by a distribution transformer. A scheme of this
network is shown in Fig. 5.
4

Fig. 5: Network builds with a loop of two and four feeders where the field
measurements have been recorded.

The two transformers have their neutral point connected to a


Petersen coil which compensates the capacitive current in case
of an earth fault. This method is greatly reducing the fault
current.
Feeder J03 and J07 form a closed loop and they are a mix
of underground cables and overhead lines. Feeders J14, J15, Fig. 6: Field measurement shows circulating current before a single phase
earth fault on feeder J03 and J07.
J16 and J17 are completely underground cables and are
connected to form one loop. According to the distribution
Due to this circulating current, the earth fault signal is
system operator (DSO), the cables are placed in parallel for 2
corrupted by a current that does not have specific behavior -
kilometers with to distance of approximately 40 cm to each
i.e. capacitive or inductive. In compensated or isolated
other.
networks, several protection devices are using the capacitive
B. The measurements behavior of the sound feeder as reference for sound feeder.
A trigger on the zero sequence voltage started the recorder. Anything that has a non capacitive behavior will be considered
This trigger is reached only during a single phase earth fault as a faulty feeder. For example, an algorithm evaluating the
because this phenomenon produces high voltage. Therefore the active current as a fault indicator will measure an active
circulating currents are only observable alone during the pre- current on every healthy loop. This can lead to wrong fault
fault records. detection.
Circulating currents are measured in both loops. The proof The recording presented next on Fig. 7 is even worse
that they are circulating is that the sum of the four currents is regarding the need of secure fault detection. On feeders 14, 15,
approximately zero. The remaining current is mainly the 16 and 17, the magnitude of the current during the fault is not
capacitive current created by the small zero sequence voltage. higher than the pre fault value. The detection of the faulty
In the loop of two feeders, the circulating current is around 1 feeder is much more difficult in this case because the
A primary. In the loop of four feeders, circulating currents are information from the circulating current is more dominant than
between 2 A and 3 A primary. The voltage shows the presence the zero sequence current due to the fault.
of a fault in the network because the voltage produced by the As stated before, under healthy operation no zero sequence
unbalanced impedance seems to be too small. For a fault, the current or only a very small one can be observed. Thus the
amplitude is varying at each period and the signal is not as detection of high impedance faults is not the biggest challenge
stable as the circulating current. in distribution networks in Europe. However, the detection is
Around 20 earth faults have been recorded from this more difficult in case of closed loop structure because the
network, most of them were temporary. The circulating current circulating current amplitude is the same order than the current
has always been measured on the loop of four feeders with caused by the fault.
different but very stable amplitude. These measurements seem Field measurements show that the zero sequence current
to consolidate the idea of a coupling between the load or can be even decreased during a single phase earth fault
positive system and the zero sequence system. On the loop of because of such a circulating current. In case of high
two feeders, the circulating current was not always measured impedance faults, the current of the faulty feeder can have
and the amplitude was always smaller than on the multi loop. smaller amplitude than the circulating current. If the phase
angle of the faulty current is 180° relative to the circulating
current, the sum of the two will result in a small zero-sequence
current. If a protection device tests the sound behavior of the
feeder, the circulating current will lead to a completely wrong
result.
5

trefoil position has been implemented without symmetrical


position to obtain a mutual coupling between the systems.
There are two different busbars connected to their own
transformer but these transformers have the same neutral and
the Petersen coil is compensating both networks with an
overcompensation of 10%. Each transformer has a nominal
power of 10 MVA.
The four feeders loop consists only of cables and the two
feeders loop consists of a mix of overhead lines and cables.
B. Influence of the load power on the amplitude of the
circulating current
On Fig. 8, the position of the single core cables in ATP are
illustrated. The characteristic of the single core are encoded
and the software is computing the symmetrical component
parameters based on the Bergeron method.

Fig. 7: On some recordings, circulating currents are higher than the fault
current for the four feeders loop. Fault occurs on J03 and J07.

Some protection devices are using only the transient part of


the signal and they filter the 50 Hz component. These devices
should not be disturbed by the circulating current because it
affects mainly the fundamental component. However, these
devices have not a good sensitivity regarding the fault
impedance.
Fig. 8: Position of the single core cables in the soil. Simulation of the four
IV. ANALYSIS BY SIMULATION feeders loop in ATP/EMTP

Influenced of the unbalanced in the matrix impedances have


The simulations have been performed, simulated for
been studied. We have also analyzed the coupling between
different value of the perfectly balanced load from 0 to 50% of
several feeders in case of multi loop system and the impact on
the nominal power of the transformer.
the circulating current amplitude. This section describes the
The zero sequence voltage remains very small and does not
simulated network and the results of the simulations.
disturb the protection devices. Its maximum amplitude has
Simulations have been made with the ATP/EMTP software.
been measured to 0.5% of the nominal voltage. This voltage is
The models used are distributed overhead lines and
measured on the both busbar and the radial feeder will produce
underground cables. Cables are based on a Bergeron model
zero sequence current which is not circulating. However, this
and overhead lines are simulated with a JMarti model. In the
zero sequence current measured on the radial feeder is caused
simulated network, we have tried to obtain the same zero
by the voltage on the zero sequence capacitances of the line.
sequence capacitance than the real network.
This current is a sound one for the protection devices that
A. The network studied protect the network.
A network with a loop of two feeders and a loop of four The circulating current amplitude is proportional to the load
feeders has been simulated trying to reproduce the real because the positive sequence current is directly influencing
network from the field measurement with the data available. the zero sequence system. The zero sequence current can
Underground cables have been studied; the feeders are in represent 6% of the nominal current of the network.
parallel for 2 kilometers with to distance approximately 40 cm
to each other. The length of the lines has been adapted to
match the zero sequence capacitance of each feeder of the real
network and its serial impedances. The position of the cables
has been set in the way to obtain a difference of impedance in
the phase impedance and consider the mutual coupling
between the loops. However, the flat position has not been
implemented because the distribution system operator had
indicated that the cables were laid in a trefoil position. A
6

active current on each feeder and if this current is too high, the
feeder is considered as faulty. Such relays could not work
when circulating currents are measured on the loop.
Because these currents depend on the load power, the
amplitude and phase of this current is supposed to be stable
during the faulty process. Estimating this current during the
pre fault state and filtering it for the faulty state with active or
non linear filter could allow a correct detection of the fault.
Another method could be to consider the loop as a single
feeder by summing the current of each feeder. This sum will
eliminate the circulating current part of the signal and keep
Fig. 9: Influence of the load power on the zero sequence circulating current in only the faulty information.
case of loop of two and four feeders.
VI. ACKNOWLEDGMENT
Simulations have proved the link between the positive The authors gratefully acknowledge the help and support of
sequence system and the zero sequence system due to M. Kereit from Siemens AG during the research project.
unbalanced serial impedance in loop. This current can have a Gratitude is also expressed to Pfalzwerke Netzgesellschaft
big impact on protection because this current does not have the GmbH for their field tests in compensated network.
expected behavior for a proper detection of the faulty feeder.
In case of high impedance fault it can be observed that the zero VII. REFERENCES
sequence current caused by the fault is smaller than the zero [1] H. Y. Li, P. A. Crossley, and N. Jenkins, "Transient Directional
sequence circulating current. Protection for Distribution Feeders with Embedded Generations," in
However, this circulating current is probably stable during 14th PSCC, 2002, pp. 24-28
[2] S. Repo, A. Nikander, H. Laaksonen, and P. Järventausta, "A Method to
the whole fault because the load is not disturbed by a single
Increase the Integration Capacity of Distributed Generation on Weak
phase earth fault in such network. The phase to phase voltage Distribution Networks," in 17th International Conference on Electricity
does not vary in case of single phase to earth fault. This Distribution, 2003, no.60, pp. 12-15
current is then stable if the consideration that the load is [3] B. Li, X. Yu, Z. Bo, and S. Member, “Protection schemes for closed
loop distribution network with distributed generator,” in 1st
constant during the time window for the earth fault detection. International Conference on Sustainable Power Generation and
Estimating the phase angle and the amplitude of this current Supply, 2009, no. 8, pp. 1-6, Apr. 2009.
during the pre fault state could allow filtering this component [4] G. Celli, F. Pilo, G. Pisano, R. Cicoria, and a Iaria, “Meshed vs. radial
MV distribution network in presence of large amount of DG,” in IEEE
without filtering the 50Hz component produced by the fault at PES Power Systems Conference and Exposition, 2004., pp. 1357-1362.
least for the first 100 millisecond after the fault inception. [5] G. Druml, R.-W. Klein, and O.Seifert, "New adaptive algorithm for
Active filter and non linear filter could be solution. detecting low- and high ohmic faults in meshed networks," in 20th
International Conference on Electricity Distribution, 2009, no. 631, pp.
Another solution is to consider the loop as a single feeder. 1-5
The sum of the current of each feeder from one loop makes the [6] L. B. Perera, "Directional Earth Fault Relay operation in Mutually
circulating current equals to zero. The effect of this current is Coupled Multiple Circuit Distribution Lines," in 20th Australasian
not measured anymore and the detection of an earth fault can Universities Power Engineering Conference (AUPEC), Christchurch,
New Zealand, 2010.
be done as usual. This solution could have the disadvantage to [7] M. F. Abdel-fattah and M. Lehtonen, “A New Transient Impedance-
not detecting the faulty part of the loop anymore. As a Based Algorithm for Earth Fault Detection in Medium Voltage
consequence a fault would be cleared unselective. Networks,” in International Conference on Power Systems Transients
(IPST), Kyoto, Japan, 2009.
[8] G. Druml, A. Kugi, and B. Parr, "Control of Petersen Coils", presented
V. CONCLUSIONS at XI International Symposium on Theoretical Electrical Engineering,
Linz, Austria, 2001.
This paper presents a problematic of the more and more [9] G. Druml, A. Kugi, and O. Seifert, “New Method to Control Petersen
common closed loop structure in distribution network. The Coils by Injection of Two Frequencies,” in 18th International
position of the cables and overhead lines can create unbalance Conference on Electricity Distribution, 2005.
in the matrix impedance. This unbalanced system has a mutual
coupling between the positive sequence and the zero sequence VIII. BIOGRAPHIES
system which could be neglected in radial structure. A zero
sequence voltage is induced on both feeder and if they are Matthieu Loos was born in Arlon in Belgium, on
July 15, 1987. He studied Electrical Engineering at
connected at both ends, a circulating current will flow in the
the Free University of Brussels (ULB), Belgium and
loop. Multi loops structures are also studied and show that the graduated with his Diploma in 2010.
circulating current can be even bigger in this case due to He is currently working as a PhD student in the
mutual coupling between the zero sequence systems. Electrical Engineering Department (BEAMS) since
2010. His main topic is earth fault detection and
This circulating current can negatively influence state of the location in compensated and isolated distribution
art protection devices because it creates randomly active or network.
reactive current measurement. Some devices are measuring the His working experience is as an intern in the
7

Belgian TSO Elia Engineering in 2010 at Diegem, Belgium and as an intern


in Siemens AG in 2011, Berlin, Germany.

Dipl.-Ing. Stefan Werben was born in Nijmegen,


Netherlands on June 1, 1964. He graduated from
high school Goethe Gymnasium, Einbeck, Germany
in 1983. He studied Electrical Engineering at the
Technical University Braunschweig, Germany and
graduated with his Diploma in 1990. Post-Graduate
studies in Business Management at the Southern
Illinois University, Carbondale, USA tool place in
year 1991.
1992 he joined Siemens AG as a software developer
for digital protection devices. After R&D project management he changed
into the Product Management where he worked as technical specialist for
differential protection. Since 2001 his job is the Product Lifecycle
Management for Overcurrent protection devices.

Prof. Jean-Claude Maun Jean-Claude Maun


received the M.Sc degree in mechanical and
electrical engineering in 1976 and the Ph.D. degree
in Applied Sciences in 1981, both from the
Université Libre de Bruxelles (ULB), Brussels,
Belgium. He joined the Electrical Engineering
Department of this university in 1976 and he is now
professor and Dean of the ULB Engineering School.
He has been leading research projects in the field
of the design of digital protections for Siemens as a
consultant. An emeritus member of the Société des Electriciens et
Electroniciens français and a recipient of many academic awards, Prof. Maun
is an recognized expert in electrical networks’ safety and protection systems
as well as, more generally, in power transport and distribution networks and
in decentralized electricity generation.

You might also like