Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
net/publication/229018532
CITATIONS READS
14 1,156
3 authors, including:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Sami Repo on 01 June 2014.
Healthy loop 3
Open-point
disconnector
Earth fault protection
Branch
Compensated or The presence of an earth fault can be indicated by the neutral
Feeder A
isolated neutral Open-point
4 disconnector
voltage measurement in the corresponding way as with radial
Faulty loop
Feeder B e.g. high-speed operation. Directional information of the fault can be obtained
switch-disconnector through the same principles as with a radial feeder. Instead of
Fault
U0 Branch the residual current of the individual feeder, the sum of the
residual currents of the looped feeders (Feeders A and B in
Figure 1) is utilized. The sum of residual currents of the
Figure 1. Medium voltage system including looped and radial feeders looped feeders is later called the "total sum current". The total
sum current is referenced to the neutral voltage and the
In this study an economical solution with straightforward directional information can be obtained through the direction
protection and fault indication principles is proposed. With of the total sum current. This method requires horizontal
ring connected feeders one task is how to indicate and locate communication of the feeder terminals of the looped feeder
the faulty ring. One possibility to implement the feeder e.g. via LON bus (Local Operating Network).
protection is that after indication and location of the fault the
loop will be opened by a high-speed switch-disconnector and The loop can be opened either via the remote control system
a conventional feeder protection trips the radial faulty feeder or on account of the neutral voltage measurement present in
(Figure 1). Otherwise, both feeder circuit-breakers of the loop the disconnector substation. Controlling the switch-
must be opened, which broadens the area of the influence of disconnector via the remote control system is normally too
the interruption. slow for protection purposes. If a switch-disconnector must
be controlled independent of other load breaks, the device in
Over-current protection the disconnector substation must be capable of indicating the
faulty loop. The directional function, which indicates the
Over-current protection can function in the same way as with existence of the fault independently and sends an open
a radial feeder. In the event of a fault the breakers of both command to a switch-disconnector, is needed for the
feeders will trip. In this case the disturbance area will be disconnector substation. The same device can close the
extended compared to radial operation. The other philosophy switch-disconnector independently. No communication to the
is that the loop will be opened on account of the fault current disconnector substation is needed. If a 110/20 kV substation
using a high-speed switch-disconnector. After that the feeds only a few looped feeders or if they are used only
conventional over-current protection will trip the faulty radial temporarily during heavy load periods, all the loops can be
feeder. A switch-disconnector receives the open command on opened on account of the neutral voltage measurement.
account of the current measurement present at the
disconnector substation. Thus, the time delay of the data Rapid autoreclosure can be used to extinguish temporary
transmission can be avoided. Autoreclosures operate in the earth faults after opening the switch-disconnector. With
corresponding way as with radial operation. In certain temporary earth faults it is reasonable to use a long enough
conditions, e.g. in summer with thunderstorms when many time delay before tripping in any case within the touch
autoreclosures may occur the large number of breakings of voltage limits [3]. The common tendency is to avoid the
the switch-disconnector must be taken into account. In voltage break of the autoreclosure and use some more
underground cable networks autoreclosures are not in use. developed method (earth fault current compensation with or
without residual current compensation) for extinguishing the
Instantaneous release may cause a problem with short-circuits earth fault arc.
if the dimensioning of the lines does not allow the time
needed for opening the switch-disconnector. If simultaneous The indication of the open conductor faults is easier with
tripping of circuit-breakers of both feeders is used the looped than radial operation because the earth contact is
disturbance area of influence will be extended compared to always on the side of the substation.
radial operation.
IDENTIFICATION OF VERY HIGH-RESISTANCE
Looped operation of feeders has no significant influence on EARTH FAULTS
the dimensioning of the lines. The fault current of the fault
location will increase, however, due to the ring connection
An earth fault in an isolated or compensated system leads to was about 23 A (RF = 0 Ω). The neutral voltage was only 64V
very low fault current with high-resistance earth fault. Thus, in normal operating conditions. The cable current
the detection of such faults is a very challenging task. Early transformers with transformation ratio 50/1 were available at
identification and location of very high-resistance earth faults the substation. The measurements were made using one 8-
are of increasing importance in improving the reliability of channel Yokogawa measuring instrument and two 4-channel
electricity distribution. Thus, accurate supervision of the Tektronix oscilloscopes. The sampling frequency used was 2
distribution systems in order to evaluate the actual state of the kHz, which corresponds to the sampling frequency of modern
network and make the early detection of the developing earth feeder terminals. Staged resistors were used to simulate a
faults possible is becoming increasingly important. New fault resistance. The duration of the artificial fault was
methods for identifying and locating earth faults have been between five and ten seconds. Only fundamental waves of
studied during the past few years in the Institute of Power voltages and currents were used for calculation.
Engineering of Tampere University of Technology. The novel
method developed for the identification and location of very Table 1 presents the fault resistances calculated for the faulty
high-resistance faults utilizes the measuring information of looped feeder and healthy radial feeder using MATLAB-
modern numerical feeder terminals and operates software. The loop was arranged before a fault occurrence.
independently as a relay function [4]. Thus, the fault The notation ϕ U refers to phase angle of the phase-to-earth
νF
resistance of all feeders can be monitored separately. The voltage of the faulty phase referenced to phasor ∆U0. The
indication of the earth fault also includes information on the notation ϕ∆U refers to phase angle of phasor ∆U0.
faulty feeder and phase. 0
When developing the appropriate methods and applications TABLE 1 - Calculated fault resistances with looped feeder (faulty) and
one radial feeder (healthy). The total line length of the looped feeder was
the needs of looped operation of MV feeders are also taken 27km.
into consideration. Very high-resistance earth faults can be
indicated with looped feeder applying the same methods as Staged Looped Radial Phase Neutral Phase
those developed for radial systems [4]. Instead of the residual resistance feeder feeder angle voltage angle
current of the single radial feeder the total sum current of (faulty) (healthy)
feeders connected into a closed loop is used. This calls for RF RF RF ϕU νF ∆U0 ϕ ∆U 0
horizontal communication between the protection relays, [kΩ] [kΩ] [kΩ] [deg] [V] [deg]
which is possible nowadays with modern feeder terminals. 10.2 11.2 1343 90.2 557.3 -62.9
According to the notations of Figure 1 the change of sum 21.1 22.1 1882 90.4 289.5 -61.7
current ∆I0i and its phase angle are defined with looped 56.6 55.5 8890 91.8 120.8 -62.2
operation in the following way: 97.0 90.4 4300 91.3 73.0 -61.6
118.0 124.8 10670 90.7 54.0 -60.8
∆ I 0i = ( I 0 A2 + I 0 B 2 ) − ( I 0 A1 + I 0 B1 ) 186.0 196.9 17383 92.0 34.8 -62.1
(1)
ϕ I 0i = arg( ∆ I 0i ) − arg( ∆U 0 )
In all cases an explicit and reliable indication of the fault,
where ∆I0i is the change of the sum current of looped feeder i faulty phase and faulty looped feeder could be achieved up to
I0A1 and I0B1 are the sum currents of Feeders A and B fault resistance 186 kΩ. The calculated shunt resistances of
before a change of U0 the healthy feeder were of a totally different order of
I0A2 and I0B2 are the sum currents of Feeders A and B magnitude (over 1MΩ) and no inaccurate indication occurred.
after a change of U0 The accuracy of the calculated fault resistances was good and
ϕ I 0i is the phase angle of phasor ∆I0i referenced to completely adequate for the practical application.
The only difference between the fault indication on a looped and Short-circuits
radial feeder is that in the looped case the phasor ∆I0i defined by
Formula 1 is used instead of ∆I0i of a radial feeder. According to The calculation of the distance to a short-circuit fault with
field experiments earth faults with fault resistance of a few looped operation may be based on the fault current as with
hundred kilo-ohms can be reliably indicated [4]. radial operation. The main difference is that with looped
operation the distribution ratio of the fault current between
Field Experiments looped feeders A and B (Figure 1) can also be utilized. With
this distribution ratio the fault can be located within the main
The fault indication methods were tested using the measured line of the ring. After that Thevenin's impedance of the circuit
data collected from the field experiments. The field can be calculated using the total fault current. Using the
experiments with staged fault resistances were carried out in distribution ratio of the fault current and Thevenin's
normal network conditions at one 110/20 kV substation of impedance, the estimate for the fault distance can be
Nurmijärvi Electricity, where the neutral of the 20 kV calculated. The number of possible fault locations is reduced
network is isolated from the earth. The network is mainly of compared to fault distance calculation with radial operation.
overhead construction. The earth fault current of the network
The fault distance can also be calculated with radial operation determined without tripping if the fulfilling of the safety
if the loop is opened by a high-speed switch-disconnector. regulations can be guaranteed. The fault resistance and fault
current can be determined by the methods used for fault
Earth faults indication. The earthing voltage can be calculated when the earth
resistances of the protective earthings are known. When a fault is
Estimation of the distance to an earth fault is problematic in caused by a tree the hazard voltage emerging cannot be
isolated and compensated networks. The longitudinal impedance determined. If the safety regulations do not allow fault distance
of the line between the fault location and the substation has estimation during a fault, the ring is connected only for the short
practically no influence on the magnitude of the fundamental period needed for collecting measurement data for the
frequency fault current. Thus, the earth fault distance cannot be calculation. The minimum duration of the ring connection is the
calculated using the same methods as with short-circuits. In rural time needed for the measurements for the earth fault distance
MV networks the length of the feeders may be several tens of calculation. After the measurements the ring is opened and the
kilometres, sometimes over one hundred kilometres. In urban faulty feeder is allowed to be tripped. This practice also reduces
underground cable networks distances are small, but faults are the time needed for the determination of the fault distance. The
almost always permanent and irreversible. Thus they must be hazard voltage regulations may require a compensation of the
located using some measurement technique. Computational earth fault current. In many cases, the compensation makes it
methods for distance estimation of earth faults reduce the time possible to locate the fault without outages to the customers.
needed for the fault clearing and the effect of the disturbance
because the trial switchings can be avoided in most cases. The The planned loops can be connected either by a remote or
goal is to locate the faulty line section in order to reduce the time manually controlled switching device. In many cases, even the
required to remove the earth fault. use of a manually controlled disconnector considerably reduces
the time needed for the determination of the fault location
Especially in isolated systems permissible touch voltages need to compared to pure trial switchings.
be considered [3]. If the ring operation mode is the normal
switching state then the measuring information is obtained VOLTAGE DROP AND POWER LOSSES
immediately after the fault occurrence. If radial operation is the
normal operation mode then the ring can be connected for the Radial MV feeders are of different length and consist of
earth fault distance estimation only. Then the information on the conductors with different values of resistance and reactance.
earth fault distance up to high fault resistances can be obtained Total loading of separate feeders is different and loads are
by rearranging the faulty feeder into a closed ring over some unevenly distributed. Thus, the universal solution or formula
healthy feeder fed from the same bus bar. when is it worthwhile to use feeders looped is difficult to
determine. The study of the voltage drop and power losses
Earth fault distance can be determined by measuring the sum requires more detailed network calculations. On average the
currents of the looped feeders and neutral voltage during an maximum voltage drop and power losses are lower in looped
earth fault. The sum current fed by the background network than in radial operation. This is mainly a consequence of
and possible neutral-to-earth impedance distributes between smoother distribution of the load current. The size of the
looped feeders in inverse proportion to the longitudinal benefit depends mainly on distribution of the load between
impedances between the fault location and the primary looped feeders, the length of the looped feeders and conductor
substation. If Feeders A and B (Figure 1) are looped when the types.
fault occurs, the influence of the capacitive unbalance can be
eliminated by determining the earth fault current using the Reducing the voltage drop incurs costs (conductor
change of the sum currents and neutral voltage. The distances reinforcements, construction of primary substations,
to a fault along both feeders can be determined by utilizing the compensation of reactive power, step-up transformers). Thus,
distribution ratio of the earth fault current between looped the utilizing of the ring connection may in some cases be a
feeders. good method to postpone investments. Voltage drops vary
locally and temporarily according to load variations. Voltage
The other possibility is to deviate the zero sequence impedance drops are normally highest in winter during heavy loading.
of the system. E.g. a change in the tuning of the compensation The use of the ring connection in wintertime can save
coil during a fault produces a change in the phase-to-earth investment costs if network reinforcement or substation
admittances along both feeders. Alternatively some other investments can be postponed. In summer and autumn when
artificial deviation of neutral voltage can be used. The temporary arc-type faults appear most in overhead line
distribution of the change of the phase-to-earth admittances networks radial operation may be preferred. In addition the
along both feeders can be determined. The distribution of the realization of overvoltage protection affects the number of
total change of the phase-to-earth admittance is proportional to temporary faults.
the distance to a fault via the corresponding feeder. In practice
the estimate for the fault distance can be calculated using the The most significant advantages of looped operation can be
change of the sum currents when the neutral voltage changes. achieved when there is a considerable imbalance in loading,
length or conductor cross section between looped feeders. A
The fault location takes some time and the safety regulations for significant advantage can also be achieved when the load of
the hazard voltages must be fulfilled. The fault distance can be Feeder A is concentrated at the end of the feeder and load of
Feeder B at the beginning of Feeder B. Temporary variation Directional earth fault protection can be arranged with
of loads must also be taken into account (e.g. industry modern relay technology without major modifications. A
feeder/domestic feeder). modern high-speed switch-disconnector equipped with
tripping magnet can be opened in less than 100 milliseconds.
INCREASING THE INTEGRATION CAPACITY OF In the case of an earth fault, looped operation does not
DISTRIBUTED GENERATION normally increase the total tripping delay if a high-speed
switch disconnector is used. The earth fault current can also
The fast propagation of DG at a distribution network will be limited by various methods e.g. by Petersen coil.
have a major influence on distribution networks. The Indication methods for very high-resistance earth faults
structure and operating principle of the distribution network function almost equally with ring and radial operation mode.
will move towards a distribution system (active distribution
network) where there are several active components along the Ring operation mode of MV feeders may offer the following
network. These issues at least should be taken into account in benefits:
the most potential areas of DG in long-term planning studies • Optimal power flow between looped feeders
[5]. DG brings out many new factors which are not usually • Reduced maximum voltage drop
considered and experience of distribution network companies • Reduced power losses
of these topics is so far quite rare. This will necessitate the • Calculation of the distance to low- and high-resistance
development of the existing networks and operation methods earth faults
in order to consider DG in the distribution networks. • Option for dynamic use of ring connection according to
seasonal load variations
The voltage drop of a medium voltage network has been one • Option to increase the integration capacity of DG at
of the main limiting factors in the planning of the rural existing network.
distribution networks in Finland. There are networks where
the length of 20 kV feeders is as much as 100 kilometers and Major drawbacks include:
the demand in sparsely populated areas is very low. The • More complicated feeder protection
impedance of these feeders is very high. If a large DG unit is
− such as autoreclosures of overhead line feeders
installed far from a primary substation, voltage rise problems
• Opening of the ring with a fault case lengthens a tripping
may occur during low demand periods [5]. The voltage rise
delay especially in the case of instantaneous release
problem is acute in coastal and skerries areas due to wind
power installations. The integration capacity of DG units on a • Voltage sag frequency and the severity of sags are
weak distribution network is limited by the voltage rise increased compared to radial operation mode.
problem.
In an underground cable systems failure frequency is very
Ring operation mode of certain feeders offers one option to low. Thus voltage sags have less significance. In systems
increase the integration capacity of DG and network transfer including mainly overhead lines the situation is the opposite.
capability at an existing, often weak distribution network. The
looped operation of distribution feeders will divide power Technically the looped operation proposed in this study can
flows to the feeders more evenly than with radial operation, be realized if the benefits evaluated are adequate compared to
losses will be reduced and voltage rise due to DG could be the drawbacks and the costs incurred through the investments
limited. and the realization of the protection.
CONCLUSIONS
[2] P. Mähönen, M. Moisio, T. Hakola, H. Kuisti, 1996, The
Modern protection, telecommunication and breaker Rogowski coil and the voltage divider in power system
technology forms a good basis for the realization of the protection and monitoring. International Conference on
protection with ring operation mode of feeders. Load flow Large High Voltage Electric Systems, CIGRE, Paris,
and fault current calculation algorithms exist for looped France, Report no. 34-103, 7 p.
distribution networks.