Electric Power Systems Research 79 (2009) 57–64

Contents lists available at ScienceDirect

Electric Power Systems Research

journal homepage: www.elsevier.com/locate/epsr

Procedures for fault location and isolation to solve protection selectivity

problems in MV distribution networks with dispersed generation
Stefania Conti ∗ , Sebastiano Nicotra
Dip. di Ingegneria Elettrica, Elettronica e dei Sistemi (D.I.E.E.S.), Università degli Studi di Catania, Viale Andrea Doria 6, 95125 Catania, Italy

a r t i c l e i n f o a b s t r a c t

Article history: Present planning and operation criteria used for MV passive radial distribution networks are, in general,
Received 23 March 2007 not suitable to cope with the presence of a significant distributed generation (DG) capacity. In fact, many
Received in revised form 6 March 2008 technical problems are still to be solved in order to increase DG penetration and to provide, at the same
Accepted 21 May 2008
time, high service quality levels to customers.
Available online 14 July 2008
In this context, the present paper illustrates the results of the first part of a research project, conducted in
collaboration with ENEL Distribuzione S.p.A. (Italy), which analyses possible solutions for various protection
Keywords:
issues arising in distribution networks in presence of DG. The paper deals, in particular, with the problem
Electrical distribution networks
Dispersed generation
of incorrect operation of traditional protection schemes for multi-phase and phase-to-ground faults in
Distribution protection MV networks, and proposes a protection philosophy based on innovative technical solutions to solve the
problem of lack of protective devices coordination in presence of DG to improve service continuity.
© 2008 Elsevier B.V. All rights reserved.

1. Introduction provide a high quality service to customers, various issues such as

voltage control, power quality, lines thermal condition, and system
The general view is that dispersed generation1 (DG) is expected protection have to be taken into consideration [4].
to play an important role in future electrical energy systems, as rec- Conflicts that may arise between DG and traditional distribution
ognized, for example, by the developers of the European Technology network protections are discussed in a companion paper [5], which
Platform for the Electricity Networks of the Future “Smart Energy is intended to be read before the present one, in order to ease the
Networks” in FP7 [1]. Increased penetration of DG is expected, comprehension of the presented concepts, as it presents a prelimi-
especially in distribution networks, due to a growing interest and nary analysis to the development of a research project, carried out
need to exploit renewable energy sources and CHP generation in in collaboration with ENEL Distribuzione S.p.A. (the major Italian
the framework of environmental issues, emerging power markets distribution system operator), aimed at finding innovative solu-
and government economic incentives for renewables in most coun- tions for protection problems arising in distribution networks in
tries [2,3]. However, besides a number of benefits, there are still presence of DG. Practical cases have been examined with reference
some technical, economical and regulatory issues with DG. From to unforeseen increase in short-circuit currents, lack of protec-
a technical viewpoint, the presence of a significant DG capacity tive devices coordination, ineffectiveness of line reclosing after a
in distribution networks would result in some conflicts with the fault using automatic reclosing devices, undesired islanding and
correct operation of the present networks, mainly because, unlike untimely tripping of DG interface protections. Moreover, assum-
the meshed transmission system, the distribution system is usually ing the possibility to manage portions of distribution networks as
designed as a “passive” radial system, which is conceived with nei- intentional islands, the potential difficulty in setting up an effective
ther generators operating in parallel nor power flow control. As it and selective protection scheme for faults occurring in the island
is known, in order to maintain correct distribution operation and has been considered.
Then, the present paper summarizes the results of the first
part of the aforementioned research project. Based on the analy-
sis presented in [5], a new protection philosophy will be proposed,
∗ Corresponding author. Tel.: +39 0957382604; fax: +39 0957387944. along with its possible implementation, to cope with the prob-
E-mail addresses: [email protected], [email protected] (S. Conti). lem of effective and selective protection against multi-phase and
1
Generally speaking, DG is characterised by relatively small rated power sources
phase-to-ground faults in distribution networks in presence of
(typically, from a few kW to tens of MW), connected to LV or MV distribution net-
works, privately owned, neither centrally planned nor dispatched and unable to DG. In Section 2, the critical analysis of some technical solutions
contribute to frequency or voltage control. presented in literature will be reported. Then, in Section 3, inno-

0378-7796/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.epsr.2008.05.003
58 S. Conti, S. Nicotra / Electric Power Systems Research 79 (2009) 57–64

vative technical solutions will be described to solve the problem of As reported in [8], adaptive protection can be defined as “an on-
loss of protective devices coordination for multi-phase and phase- line activity that modifies the preferred protective response to a
to-ground fault protection in presence of DG in MV distribution change in system conditions or requirements. It is usually auto-
networks. The proposed method is based on two algorithms that matic, but can include timely human intervention”. An adaptive
define the devices and the sequence of operations required for fault relay can be defined as “a relay that can have its settings, charac-
location and isolation. Implementation issues of the proposed pro- teristics or logic functions changed on-line in a timely manner by
cedures in present Italian distribution networks will be concisely means of externally generated signals or control action”. In other
discussed in Section 4, in order to highlight how the solution could words, it could be possible to use adaptive relays in order to adjust
by easily applied to an actual scenario. The potential economic protection response to changing conditions, including the presence
advantage for the distribution operator is also briefly discussed. of generators in the network. In [7] the authors consider a typi-
cal case in which recloser and fuse are coordinated for a fault on
2. Related work laterals without DG. When DG is connected between the recloser
and the fuse, for a fault on the lateral, the fuse will see more cur-
Not many studies have been published in scientific literature on rent than the recloser. This can result into coordination being lost
incorrect operation of protections in distribution networks due to between these devices. Moreover, the recloser would see fault cur-
the presence of DG. It is worth to note that distribution protection rent for upstream faults too. They conclude that the coordination
is actually a wide field to be investigated and the few solutions in the presence of DG can be achieved with microprocessor-based
proposed so far do not meet all practical cases. As for the prob- reclosers. The expected requirements from the recloser are, basi-
lem of difficult fault location in presence of DG, for example, in [6] cally, to allow different operating curves (included user-defined
a distribution protection scheme based on distance relays is pre- curves) and to give a variety of curve-choices to achieve coordina-
sented specifically for 11 kV circuits typically used in U.K., in which, tion. Further, the recloser is required to be “adaptive” to changing
as stated by the Authors, the connection of DG causes problems system conditions and it has to be made directional toward the
with existing non-directional overcurrent and earth-fault protec- downstream side of the feeder.
tion. The proposed solution helps minimising these problems. Actually, it can be demonstrated that the use of adaptive protec-
More in general, two approaches have been proposed to deal tions does not ensure that correct coordination is always obtained.
with the problem of lack of coordination among protective devices To show this, an example is reported, which is referred to the net-
and fault location, such as adaptive protection, and use of databases work described in [5] and reproposed here for greater convenience
to collect off-line information useful to identify and locate a fault. in Fig. 1. Let us consider, for example, a fault in section e1 of Line
In [7] the problem of fuse-recloser coordination in presence of E (F1). When DG is present in the network, the fault is supplied
DG is discussed and the use of adaptive protections is proposed. The by two sources at the same time (the primary substation, PS, from
concept of adaptive protection has been presented in the eighties one side and the generators on Line D from the other one). Then,
for improving protection performance in transmission networks. to obtain correct protection coordination in presence of DG, the

Fig. 1. Typical scheme for MV distribution network with dispersed generators.

 S. Conti, S. Nicotra / Electric Power Systems Research 79 (2009) 57–64 59

general, not only the complete knowledge of the network struc-

ture and characteristics but, also, the consideration of all possible
locations and conditions for each fault type. Voltage and/or current
measures should be gathered each time a fault occurs and they
should be compared to the stored information in order to assess
the fault type and its location. Then, the most appropriate devices
should be operated to isolate the fault.
In [12] a similar approach has been followed to cope with the
negative effect of high DG penetration on coordination of protec-
tive devices (fuses and reclosers). An adaptive protection scheme,
supported by the use of a “relay database”, has been suggested as
a solution to the problem. As stated by the authors “this method
requires a load flow study and a complete short-circuit analysis for
all types of faults involving different phases. In addition to fault
current for different types of faults at each bus, this analysis should
Fig. 2. Correct protections coordination for any fault in the network without DG or also find out fault current contribution from each DG and from the
for fault F2 in presence of DG. main source”. Further, “the load flow and the short-circuit analysis
need to be updated after every significant change in load, DG, or
the system configuration”.
It is to be remarked that the methods that rely on databases have
two major drawbacks. Firstly, they can sometimes fail in identify-
ing the fault, since, in actual networks, different fault events may
occur with similar characteristics. In [12] the authors admit this
possibility, especially if DG penetration is relatively low, so that
the distribution system keeps its radial nature in some parts of it.
Further, it should be noted that it is practically impossible to anal-
yse all possible faults considering all different fault impedances.
Moreover, the above methods would require the collection of an
enormous amount of data due to the hugeness of the set of con-
ditions to be considered. This makes totally impractical to take
into account the possibility of simultaneous faults (even though
they are relatively unlikely to occur). Trying to do this would
mean to increase the number of simulations beyond a realistic
Fig. 3. Correct protections coordination for fault F1 in presence of DG. limit.
In conclusion, though interesting at a conceptual level, these
methods have their weak point in the possibility of practical imple-
required characteristics of protective devices e1, d-e, d2 and d7
mentation. On the contrary, the method proposed in this paper is
should be changed from the ones shown in Fig. 2 (typically used
relatively easy to be implemented and it is suitable for networks
without DG) to those in Fig. 3. In fact, as indicated by the tripping
with low as well as high DG penetration.
sequence of Fig. 3, protection d-e should trip faster than protection
d2 in order to leave the Satellite Centre2 (SC), SC D, the possibility
3. Innovative method for fault location and isolation in
to re-energize Line D through d1 (this is only possible if the breaker
presence of DG
in d2 is closed).
In these conditions, on the other hand, should a fault occur in
In order to solve the problem of difficult fault location due to
feeder section d2 (F2), again we would observe incorrect coordi-
lack of coordination among multi-phase fault protective devices in
nation since the protection in d-e would trip before the one in d2,
presence of DG (as discussed in [5]), and to overcome the limita-
thus failing the objective to isolate the minimum faulted area (in
tions of the approaches mentioned in the previous section, a new
fact, the loads connected to branch d11 would be unnecessarily left
protection scheme has been developed. It is based on software pro-
unsupplied until the fault is located and isolated). In conclusion,
cedures required to establish the control actions performed by an
adaptive protections cannot be considered a general solution for
automatic telecontrol system. Such control actions take place after
lack of coordination in presence of DG.
line protective devices have tripped following a fault. This is eco-
The possibility to use databases to identify and locate a faulty
nomically acceptable in distribution networks, such as the Italian
section of a distribution feeder is referred by a few papers in liter-
ones, which rely essentially on line circuit breakers without using
ature, e.g. [9–11]. The basic idea is to perform an off-line extensive
fuses. In fact, the aim of the procedure is to “correct” the operation
study on a given network, typically by using a simulation software,
of the devices that have lost the capacity to be selective, due to the
in order to collect a big amount of information on the possible val-
presence of DG, and to quickly identify the switching devices able
ues of electrical quantities characterising fault events (short-circuit
to isolate the faulty feeder section.
currents, amplitude and phase of voltage sags, symmetrical com-
Even though, in case of phase-to-ground faults, the presence of
ponents of unbalanced voltages and currents, etc.). This implies, in
earth-isolated generators connected in parallel to the network does
not prevent installed protections from operating correctly, a single
procedure has been developed to be applied in case of both multi-
2
We recall from [5] (in which a compete description of the distribution network phase and phase-to-ground faults. The procedure implements the
structure is given) that the purpose of the SC is to increase network reliability thanks
to installation of line protections at the sending-end of each feeder supplied by the
same algorithms but, obviously, relies on different measures per-
SC itself. Further, it is typically possible to ensure emergency supply to SCs by means formed by the sensing devices installed in the network according
of energy alternative paths. to the fault being a multi-phase or a phase-to-ground one. For this
60 S. Conti, S. Nicotra / Electric Power Systems Research 79 (2009) 57–64

reason, the effectiveness of the proposed method is not dependent PS, since it can be sited in the SCs along the line) and the fault
on the fault typology. sensing devices have sent a signal to the control system. The num-
bering procedure is schematically described by the flow chart in
3.1. Preliminary procedure for fault location Fig. 4.
It is important to assess the “direction” of the fault current that
When a fault occurs on a feeder, firstly, a numbering procedure has been sensed by each protective device in order to establish
is performed involving the fault sensing devices installed on the whether the fault is upstream or downstream from it. Commonly,
feeder itself. The control system, which “knows” the network con- in case of multi-phase fault, the current flowing from the PS to the
figuration, automatically starts assigning an identification code (i.c.) feeder receiving-end is considered “positive”. For the sake of conci-
to the devices immediately after at least one protective device has sion, in the flow chart describing the numbering procedure, we will
tripped (note that this device is not necessarily installed in the use the conventional definitions of “P direction” and “N direction”

Fig. 4. Flow chart describing the numbering procedure.

 S. Conti, S. Nicotra / Electric Power Systems Research 79 (2009) 57–64 61

in order to indicate the characteristic of the fault current sensed by Petersen coils. Then, we will say that the sensed current has P
protections in both multi-phase and phase-to-ground faults. direction if:
Consider that, today, in Italian MV distribution networks, iso-
lated and resonant grounding schemes are both present, even • there is a multi-phase fault and the current has positive
though the situation is evolving towards the exclusive use of direction;

Fig. 5. Flow chart describing the sequence of actions required by fault location.
62 S. Conti, S. Nicotra / Electric Power Systems Research 79 (2009) 57–64

• there is a phase-to-ground fault in an earth-isolated system and

the zero-sequence current lags the zero-sequence voltage;
• there is a phase-to-ground fault in a resonant earthing network
and the zero-sequence current has a resistive component.

Otherwise, we will say that the sensed current has N direction

if:

• there is a multi-phase fault and the current has negative direc-

tion;
• there is a phase-to-ground fault in an earth-isolated system and
the zero-sequence current leads the zero-sequence voltage;
• there is a phase-to-ground fault in a resonant earthing network
and the zero-sequence current has not a resistive component.

In practice, when a protective device trips, two cases can be

considered:

• if the fault current direction through the tripped protection is P, it

will be necessary to apply the numbering procedure only to the
devices installed downstream from the protection on that feeder,
since it is deduced that the fault is downstream from the tripped
breaker (for this reason the protection will be assigned the i.c. 1);
• if the fault current direction through the tripped protection is N,
firstly, it will be necessary to identify the PS circuit breaker in
which the sensed fault current has direction P; then, this circuit
breaker will be opened (if it is still closed) and it will be assigned
the i.c. 1. Consequently, the numbering procedure will proceed
with increasing order towards the devices furthest from the PS,
installed on the concerned feeder.

3.2. Fault location

As described in the previous section, the method proposed for

fault location is based on appropriate numbering of the devices
installed in the faulted feeder and on the acquirement of fault cur-
rent direction. Then, the sequence of operations that are required
for fault location is reported in the flow chart of Fig. 5, which will
be commented with reference to some practical examples.
It is to be noted that the algorithm allows to obtain the best
result in terms of isolating the minimum part of the faulted feeder Fig. 6. Example of faulty line with numbered devices.
when all the Secondary Substations (SSs) have fault sensing devices
and telecontrolled switching breakers. “positive”. The i.c. 1 is assigned to the protective device and the
The proposed algorithm considers the possibility that some numbering procedure goes on along Line B;
devices might not send the required information to the control sys- • “Are there any other devices downstream from device 1?”. Being
tem after a fault event. However, based on the available information, “Yes” the answer, the i.c. 2 is assigned to the subsequent device.
the fault location procedure is able to find the minimum part of the Since 2 is not the last device, it is necessary to consider the sub-
network that is possible to isolate. sequent one, to which the i.c. 3 is assigned, and so on until i.c. 4
and 5 are assigned as well. All devices from 1 to 5 has sensed a
3.2.1. Case 1 “positive” fault current.
Let us consider a multi-phase fault detected on Line B, repre-
sented in Fig. 1, and assume that it has occurred in section b7. As in the case depicted in Fig. 6, it is possible that switching
Following the fault event, the PS circuit breaker of Line B will trip substations3 (or “nodes”) are present along a feeder. In particular,
and the sensing devices affected by the fault current (i.e. installed Node B1 supplies two lines (for the sake of simplicity let us call them
upstream from the fault) will send a signal to the control centre, “left line” and “right line”). Then, after numbering device 5, it has
provided that there is no malfunctioning in the communication to be decided to which line the procedure has to be applied further
system. on. In the considered case (fault in b7) only the devices installed in
In the considered case, the numbering procedure described in the left line are of interest for the numbering procedure, since the
Section 3.1 will result in the automatic attribution of the identifi- first device has sensed a fault current with positive direction, while
cation codes as reported in Fig. 6, as the control system operates in the right line (where no generators are installed) the protection
according to the following logic:

• “Has a PS line circuit breaker tripped?”. The answer is “Yes” and 3

Substations without voltage transformation, where some lines may be protected
the system checks the direction of the fault current, which is and some others may not.
 S. Conti, S. Nicotra / Electric Power Systems Research 79 (2009) 57–64 63

devices has not sensed the fault and sent any signal. However, if negative direction for an upstream fault). Supposing, for argument’s
generators were installed there, the sensing devices would see a sake, that all generators did not provide any contribution to the fault
fault current with N direction and this would prevent the number- current, the fault location procedure would be effective anyway,
ing procedure from being applied to the right line, as this condition simply as if the network was passive (no generators installed).
would be, again, indicative of a fault on the other line. Then, the
numbering procedure identifies devices from 6 to 10 (which is the
4. Implementation issues of the proposed method
last one) on the left line. Note that devices 8, 9 and 10 have not
sensed any fault current and, consequently, have not sent to the
The implementation of the proposed procedures is based on
control system a signal with information on fault current direction.
the availability of an efficient monitoring system and a bidirec-
This implies that the faulty feeder section is b7.
tional communication between control centres and fault sensing
However, the control logic considers the possibility that, for
devices. The procedures can be easily implemented thanks to the
example, device 8 has not sent any signal due to a communication
technology nowadays available in distribution systems, such as, for
system failure rather than to the absence of a fault current affecting
example, in Italy.
it. In order to verify this eventuality, since device 8 is not the last
As discussed in [5], during the last years, Italian distribution
one, the control system will check whether the subsequent devices
system has undergone a wide restructuring process that has been
(9 and 10) have sent a fault signal or not. The first device that has not
essentially characterised by the introduction of telecontrol and
sent any signal (in this case device 8) will be questioned about its
automation technologies, along with new management policies,
status and when its answer confirms that no fault has been sensed,
to cope with the need for higher service continuity levels. Con-
it will be concluded that the fault has occurred in section b7.
sequently, in order to realize the proposed protection scheme, it
would only be necessary to update the software running on the MV
3.2.2. Case 2
network Telecontrol System (MV-TS) and the software of the field
Under the same fault conditions of the previous case, let us
devices installed in the peripheral units (PUs). These devices (called
assume that a generator (not represented in Fig. 6), is connected to
RGDAT, which is the Italian acronym for directional fault and lack of
the feeder at node Cb6. Due to the presence of this generator, a neg-
voltage detectors) are currently used by ENEL for phase-to-ground
ative fault current is sensed by device 8 before protection tripping.
fault detection in both earth-isolated and resonant networks. How-
In this case, after receiving information from device 8 and verifying
ever, they are also suitable to provide information on the direction
that device 7 has sensed a positive fault current, the control system
of multi-phase fault currents. So, according to what was previously
will identify the faulty section, i.e. b7.
described, the MV-TS would perform the fault location procedure
Now, let us assume that device 7 is unable to send the required
on the ground of the information provided by the field sensing
signals to the control system. As said before, since it is not the last
devices, notwithstanding the type of fault.
device in the feeder, the procedure considers information provided
It would be also necessary to adapt the connection plant of the
by device 8. Being the fault current sensed by this devices negative,
independent producers by adding telecontrolled switching break-
it can be concluded that the fault is located upstream from 8, but
ers on the utility side of the producer’s substation.
the lack of information about the status of device 7 does not allow
As far as the telecommunication system is concerned, it would
to isolate the least faulted area. Consequently, switches 6 and 8 will
be possible to use the present one, where communications between
be opened.
MV-TS and PS, as well as between MV-TS and SCs take place using
ISDN technology on dedicated lines (that, in future, will be probably
3.2.3. Case 3
replaced by LAN networks). Communications between MV-TS and
Let us consider a fault in feeder section b5 (Fig. 6) and assume a
PUs are currently performed by means of GSM technology. Obvi-
generator is installed at connection node Cb4. Further, let us assume
ously, the exploitation of the GSM network has the advantage of
that, due to a high impedance fault, the only device that has sensed
using an existing and well-established communication infrastruc-
the relatively small fault current is 6. The presence of the generator
ture, which, however, has the typical limitations of a public utility
results in a negative current detected by the sensing device, and in
(possible unavailability, overload, presence of uncovered areas and
absence of any other information about the status of the devices
fault events linked to electrical supply unavailability). At present,
installed upstream from 6, it is necessary to open switches 6 and 1.
ENEL is studying the possibility to use private lines to implement
Before concluding the presentation of the fault location proce-
Power Line Communications.
dure, it is useful to add some remarks to the examples discussed in
It is important to note that the fault isolation procedure can
this section. The effectiveness of the proposed method is not depen-
normally be concluded in less than 3 min.4 In practice, when a fault
dent on the presence of DG in the network or on the typology of
the generation units, i.e. on the fact that generators are “rotating”
(directly connected synchronous or asynchronous generators) or
4
“static” (connected to the grid through static converters). Generally Since 1999, the Italian Regulator for the Electrical Energy and Gas specifies the
speaking, a generator which supplies a multi-phase fault upstream criteria to record and process the data of continuity of supply to MV and LV customers
and the targets of continuity improvements to be pursued by distribution network
from it may give a contribution to short-circuit current which can operators (DNOs), including the related system of premiums and penalties for the
be felt or not by the sensing devices installed between the generator DNOs. The actual and target continuity levels are calculated in terms of “prearranged
and the fault (the value of the current contributed by the genera- interruptions” and “accidental, long interruptions”. According to Norm EN 50160
tor depends on various factors, such as distance between generator (titled: “Voltage characteristics of electricity supplied by public distribution systems”) a
supply interruption can be classified as prearranged, when consumers are informed
and fault, fault impedance, generator interface, etc.). However, if
in advance, to allow the execution of scheduled works on the distribution system,
the sensing devices installed between the generator and the fault or as accidental, caused by permanent or transient faults, mostly related to external
sense a fault current, the current direction is necessarily negative, events, equipment failures or interference. An accidental interruption is classified
notwithstanding the generator interface typology. as a long interruption (longer than 3 min) caused by a permanent fault, or as a short
If, at the limit, the devices do not sense the generator contri- interruption (up to 3 min) caused by a transient fault. Consequently, it is important
for the distribution operator to keep the duration of an accidental interruption below
bution (this is equivalent to the absence of the generator itself) the threshold of 3 min, thus preventing the interruption event from being recorded
information on fault position will be provided by the last device that as a long interruption and, thus, from contributing the calculation of economical
has sensed the fault (with positive direction for a downstream fault, penalties.
64 S. Conti, S. Nicotra / Electric Power Systems Research 79 (2009) 57–64

occurs in the network and a circuit-breaker trips, the field sensing From the economical viewpoint, the benefits deriving from the
devices immediately send to the MV-TS (through the GSM network) implementation of the method proposed would be significant for
the acquired information about the characteristics of the fault cur- the distribution operator. In fact, the procedure allows to drastically
rents that have affected them. The transmission of this information reduce the time required by fault location, even in presence of DG,
can be performed during the execution, if necessary, of an auto- since the isolation of the faulted feeder section would be done by
matic line reclosing cycle (e.g. a short and a long time reclosure) means of a single action instead of using the present mixed pro-
carried out in order to check whether the fault is temporary or cedure (automatic and manually telecontrolled). In most cases it
not. This reclosing cycle takes, typically, about 31–40 s to be con- would be possible to isolate the fault in less than 180 s, which is
cluded. This time interval is enough for the field information to be the limit over which the interruptions are considered “long” and,
acquired by the MV-TS. Then, if the fault is permanent, the control consequently, recorded in order to calculate the penalties that the
system works out the information provided by the filed devices distribution operator must pay when the target continuity levels in
to locate the minimum faulted area that can be isolated. As the terms of maximum allowed interruption durations are not achieved
algorithm that implements the fault location procedure has a low during a year. On the other hand, the greater service continuity level
computational complexity, this phase does not require particular that the distribution operator would achieve also at the LV load
computation capabilities. Though it is not easy to predict exactly points, thanks to the implementation of the proposed procedures
the calculation duration (which depends on the processor used) it at the MV stage, would be awarded by the Italian Authority.
is reasonable to assume that it would take a time relatively small
with respect to the communication time between the MV-TS and Acknowledgements
the field devices.
As described in Section 3, the lack of information from some PUs The authors would like to thank Concetto Sergio Tosto and
is considered as a possible system malfunctioning. Then, the time Sebastiano Giamboi (ENEL Distribuzione S.p.A., Italy) for their valu-
required to check the status of the field devices has to be consid- able collaboration in the development of the research project.
ered (the time needed to send them the queries and receive back
the answers is about 40 s). Finally, a maximum time of about 40 s References
is required by the MV-TS to establish a new communication with
the telecontrolled sectionalizing switches in order to command the [1] European Technology Platform for the Electricity Networks of the Future
“Smart Energy Networks” in FP7, http://ec.europa.eu/research/energy/
opening of the devices concerned with the faulted area isolation nn/nn rt/nn rt dg/article 2262 en.htm.
and to update the devices status. It can be concluded that the sum [2] Directive 2001/77/EC of the European Parliament and of the Council of Septem-
of the time durations for the above cited operations is normally in ber 27, 2001 on the promotion of electricity produced from renewable energy
sources in the internal electricity market.
the range 120–180 s.
[3] Directive 2003/54/EC of the European Parliament and of the Council of June
26, 2003 concerning common rules for the internal market in electricity and
5. Conclusions repealing Directive 96/92/EC.
[4] N. Jenkins, R. Allan, P. Crossley, D. Kirschen, G. Strbac, Embedded Generation,
The Institution of Electrical Engineers, London, 2000.
The paper presented a new protection philosophy to solve [5] S. Conti, Analysis of distribution network protection issues in presence of dis-
the problem of multi-phase and phase-to-ground fault protec- persed generation, Electric Power System Research Journal, in press.
tion coordination in presence of DG in MV distribution networks. [6] I. Chilvers, N. Jenkins, P. Crossley, Distance relaying of 11 kV circuits to increase
the installed capacity of distributed generators, in: IEE Proceedings on Gener-
The proposed method is based on software procedures required ation, Transmission and Distribution, vol. 152, no. 1, January, 2005.
to establish the control actions performed by an automatic tele- [7] S.M. Brahma, A.A. Girgis, Microprocessor-based reclosing to coordinate fuse and
control system. Two algorithms have been developed to guide the recloser in a system with high penetration of distributed generation, in: Pro-
ceedings of the IEEE Power Eng. Soc. Winter Meeting, vol. 1, 2002, pp. 453–458.
sequence of operations required for correct fault location and iso- [8] G.D. Rockefeller, C.L. Wagner, J.R. Linders, K.L. Hicks, D.T. Rizy, Adaptive trans-
lation. mission relaying concepts for improved performance, IEEE Transactions on
The proposed procedure can be applied in MV distribution Power Delivery 3 (October (4)) (1988) 1446–1458.
[9] M. Lehtonen, A. Matsinen, E. Antila, J. Kuru, P. Vuorenpaa, E. Matinlassi, S. Pet-
networks in presence of DG as well as in passive networks. Its effec- tissalo, Automatic fault management in distribution networks, Proceedings of
tiveness is not influenced by the type of generators installed in the the 16th International Conference and Exhibition on Electricity Distribution
distribution system (with rotating or static grid-interface) and it is (CIRED 2001), June 18–21, 2001 (IEE Conf. Publ No. 482, vol. 3).
[10] H. Li, A.S. Mokhar, N. Jenkins, Automatic fault location on distribution network
independent on the type of fault (multi-phase or phase-to-ground using voltage sags measurements, in: Proceedings of the 18th International
fault). Then, he proposed solution allows to remove one of the main Conference on Electricity Distribution (CIRED2005), June 6–9, 2005.
barriers to the diffusion of a significant number of dispersed gen- [11] A. Campoccia, M.L. Di Silvestre, I. Incontrera, E. Riva Sanseverino, Fault identi-
fication, location and characterization in electrical systems using an analytical
erators in distribution systems. Further, the procedure allows to
model-based approach, in: Proceedings of the 18th International Conference
overcome the present limitation in the maximum number of tele- on Electricity Distribution (CIRED2005), June 6–9, 2005.
controlled switch-breakers that can be used in the same feeder [12] A.A. Girgis, S.M. Brahma, Development of adaptive protection scheme for
in order to perform fault location and isolation procedures, thus distribution systems with high penetration of distributed generation, IEEE
Transaction on Power Delivery 19 (January (1)) (2004).
reducing the extension of the isolated faulted area.