1
A Strategy for Protection Coordination in Radial
Distribution Networks with Distributed Generators
Amin Zamani, Student Member, IEEE, Tarlochan Sidhu, Fellow, IEEE, and Amir Yazdani, Senior Member, IEEE
Abstract—In this paper, a simple and effective strategy is
proposed for the coordination of protective devices in typical
radial distribution networks embedding distributed generation
units. The proposed strategy benefits from microprocessor-based
reclosers and directional elements for feeder relays, and addresses
the issues of blinding of protection, false tripping, recloser-
fuse and fuse-fuse miscoordination, and failed auto-reclosing.
Another salient feature of the proposed strategy is that it does
not require adaptive protective devices. The effectiveness of the
proposed algorithm is demonstrated by simulation case studies
conducted on a model of an example distribution network,
in the PSCAD/EMTDC software environment. The paper also
presents the steps taken to characterize the impact of DGs on
the protection coordination, for host and neighboring feeders.
Index Terms—Distributed generation, distribution system,
fault analysis, protection coordination, recloser, relay.
I. I NTRODUCTION
RADITIONALLY, distribution systems have been de-
T signed to operate radially, that is, the power flows from
the upper voltage levels down to customers connected to radial
feeders. This simple configuration has enabled straightfor-
ward protection strategies. Thus, the conventional distribution
networks are protected by such simple protective devices as
overcurrent relays, reclosers, and fuses [1], [2]. With the intro-
duction of distributed generation, however, the radial structure
of the networks is compromised, and protection coordination
may be affected, or entirely lost in some cases. The issues
include blinding of protection, false tripping, recloser-fuse and
fuse-fuse miscoordination, and failed auto-reclosing [3]-[5].
The aforementioned issues have prompted research activities
to find effective solutions.
The impact of distributed generators (DG)s on the sys-
tem short-circuit current has been investigated in [4]. The
authors recommend that the margin required for preserving
the protection coordination be checked whenever a new DG
is added to the network. The possibility of maintaining the
coordination in radial distribution systems with DGs, if there
is enough margin between relay characteristic curves, has been
demonstrated in [5]; however, the authors have not formulated
the problem. According to [5], the relay parameters have to
be reset to revised relay curves in case the coordination is
lost although no method has been presented for the revision.
Reference [6] proposes the use of adaptive microprocessor-
based reclosers to solve the fuse-recloser coordination problem
in a typical one-feeder test system embedding a DG; however,
The authors are with the University of Western Ontario, Lon-
don, ON, Canada (e-mails: mzamani2@uwo.ca, tsidhu@eng.uwo.ca, ayaz-
dan2@uwo.ca).
978-1-4244-6551-4/10/$26.00 ©2010 IEEE
the other aforementioned issues have not been considered,
and also recloser settings must be changed after the first
reclosure. In reference [7], the idea is to employ fault current
limiters (FCL)s for limiting fault currents to such low levels
that the coordination issue does not manifest itself. Although
consistent with the general sense that devices with limited
fault currents, e.g. electronically-interfaced DGs, do not com-
promise the protection coordination, at least as drastically as
their rotating machine-based counterparts, the study has not
formulated or characterized the problem, nor has it considered
the effectiveness of the scheme for transient regimes.
In this paper, a simple and effective strategy is proposed
for the coordination of protective devices in radial distribution
networks embedding distributed generation units. The pro-
posed strategy benefits from microprocessor-based reclosers
and directional elements for feeder relays. The algorithm
also addresses the issues mentioned above associated with
the introduction of DGs to distribution networks. Another
salient feature of the proposed strategy is that it does not
require adaptive protective devices. The effectiveness of the
proposed algorithm is demonstrated by simulation case studies,
conducted on a model of an example distribution network in
the PSCAD/EMTDC software environment. The paper also
presents the steps taken to characterize the impact of DGs on
the protection coordination, for host and neighboring feeders.
The rest of this paper has been organized in five sections.
Section II briefly reviews the traditional method of protection
coordination, for a typical distribution network. In Section
III, the impacts of addition of distributed generators on the
traditional coordination are investigated, and a methodology is
proposed for coordination of protection that aims at resolving
the issues listed above. In Section IV, the application of
the proposed coordination methodology is illustrated for an
example system. Sections V and VI provide the simulation
results and conclusions, respectively.
II. T RADITIONAL P ROTECTION C OORDINATION
In a feeder, fuses must be coordinated with the recloser
installed at the beginning or middle of the feeder. The coor-
dination means that a fuse must operate only if a permanent
fault impacts the feeder (fuse saving scheme). For a temporary
fault, however, the recloser must rapidly open to isolate the
feeder and to give the fault a chance to self-clear. If the fuse
fails to operate for a permanent fault, the recloser will back it
up by operating in its slow mode. The feeder relay will operate
lastly only if both the recloser and the fuse fail [8], [9].
Fig. 1 illustrates the basics of the conventional coordina-
tion practice for the relay, recloser, and fuses in a typical

3
10
Fuse TCT
Ifmin
2 IR IF
10
Fuse MMT
1
10
time(s)
0
10
Reclose
Slow
1
10
Reclose
Fast
2
10
margin
3
10
2 3
10 10
current(A)
Fig. 1. Typical characteristic curves, their coordination, and the mar-
gin remaining after the addition of a DG
distribution network. The figure illustrates that the devices are
coordinated in such a way that, for all fault currents between
If min (minimum fault current of the feeder) and I f max (max-
imum fault current of the feeder), the fast characteristic curve
of the recloser lies below the fuse Minimum Melting Time
(MMT) curve, whereas the slow characteristic curve of the
recloser lies above the fuse Total Clearing Time (TCT) curve
(descriptions of these curves are provided in the Appendix).
Hence, the recloser opens before the fuse starts to melt and
gives temporary faults a chance to self-clear. However, for
a permanent fault the fuse will operate before the recloser
opens in the slow mode. Fig. 1 also indicates that the relay
provides an overall back-up protection since its characteristic
curve lies above all the other curves. Hence, to maintain the
coordination, one has to ensure that the fault current passing
through the devices remains between I f min and If max .
III. I MPACT OF DG ON T RADITIONAL C OORDINATION
AND THE P ROPOSED S OLUTION
Addition of a DG to a feeder, anywhere downstream of
the relay and the recloser, changes I f max and If min and also
increases the fuse current as compared to the recloser [5].
The immediate consequence is that the coordination will be
lost if the fault current exceeds I f max , due to the DG. The
other consequence is that, since as shown in Fig. 1 the recloser
current IR becomes less than the fuse current I F , the fuse may
melt before the recloser opens in its fast mode of operation.
To accommodate the impact of distributed generation in the
protection coordination, and to overcome the issues mentioned
in Section I, the following algorithm is proposed in this paper.
1) The impact of the DG to be added on the recloser-fuse
coordination of the host feeder or a neighboring feeder
is investigated. This is possible based on the approach
introduced in Section IV-D. If the coordination remains
intact in spite of the addition of a DG, the algorithm is
to be followed from Step 6. Otherwise, the algorithm is
exercised from Step 2.
2) The fault current extremes I f min and If max are calcu-
lated for the feeder under study; the former extreme is
2
calculated with no DGs considered, whereas the latter
one is calculated for the configuration where the DGs
Ifmax are in place. Such a choice of I f min and If max ensures
preservation of the coordination, even, if the DGs are
disconnected from the network, for example due to an
anti-islanding exercise.
Relay
3) The characteristic curves of the devices are coordinated
based on the conventional method explained in Section
II, which requires the values I f min and If max .
4) Since in the presence of DGs the recloser fault current
IR is less than the fuse fault current I F , the recloser fast
characteristic curve, obtained in Step 3, is revised. This
4
requires that the lowest value of I R /IF (less than unity)
10
be calculated, e.g. using an appropriate phase-sequence
equivalent circuit, for the worst-case scenario.
5) The recloser fast characteristic curve is shifted down
through its multiplication by the minimum value of
IR /IF . The revised fast characteristic curve is pro-
grammed in the recloser. Thus, the proposed methodol-
ogy assumes that the recloser is of the microprocessor-
based type and therefore programmable.
6) The impact of the DG on the fuse-fuse coordination
of the neighboring feeders is studied to confirm that
the coordination is preserved. The study methodology,
and a simple solution for the case where coordination
is lost will be introduced in Section IV-D. It should
be emphasized that the fuse-fuse coordination loss is
unlikely to happen in the host feeder. This is due to
our assumption that the DGs get disconnected after the
first operation of the recloser and, thus, the coordination
fulfilled as explained above retains its validity.
The algorithm proposed above assumes that a directional
relay is employed at the beginning of the feeders that embed
DGs, in order to guarantee that false tripping is avoided.
Moreover, it is assumed that the DGs will be dropped out once
the feeder recloser opens the circuit [4]-[6]; this is ensured by
a transfer-trip mechanism and/or the anti-islanding schemes
incorporated in the individual DGs, as also identified by the
IEEE Std. 1547 [10]. Fig. 2 provides a graphical illustration
of the proposed algorithm.
IV. D ESIGN E XAMPLE
To illustrate its effectiveness, the proposed methodology
is applied to, and discussed in the context of, an example
distribution network which, hereinafter, is referred to as the
“test system”.
A. Distribution Network Structure
Fig. 3 illustrates a single-line schematic diagram of the test
system, which consists of a 13.8-kV two-feeder distribution
subsystem that supplies the loads L1 through L4 via two
feeders, i.e. Feeder 1 and Feeder 2. The subsystem is radially
connected to the main grid through a transformer and a 69-
kV line. The grid is represented by a 69-kV bus of 1000-MVA
short-circuit capacity. The loads consist of linear RL branches
of different power factors. The basic system configuration and

Fig. 2. Flowchart of the proposed method
parameters are extracted from the benchmark system of the
IEEE Std. 399 [11].
The test system described above also includes two DGs,
DG1 and DG2, of the corresponding capacities of 5.4 MVA
and 3.8 MVA, interfaced with Feeder 1 and Feeder 2, respec-
tively. The DGs are considered to be based on the synchronous
machine, since electronically-interfaced DGs are limited in
terms of fault current and thus may not reveal the protection
coordination issues.
B. Coordination of Protection
Let us first assume that DG2 is not present and, as such,
the protective devices are coordinated for the configuration
in which only DG1 exists on Feeder 1. We also assume that
DG1 can disrupt the initial coordination of Feeder1, achieved
through the traditional practice (Section II). Thus, the worst-
case scenario that can affect the protection coordination is
considered, and then the coordination strategy of Section
III is employed to reclaim the coordination for Feeder 1.
3
Fig. 3. Single-line schematic diagram of the studied test system
To consider the worst condition, we assume the capacity of
DG1 to be 200% of the maximum feeder load; any larger
capacity is considered unrealistic [6]. We further assume that
the impedance of DG1 is 0.1 pu, as the worst condition [1].
Moreover, the leakage reactance of the corresponding interface
transformer is assumed to be conservatively low, i.e. 5% [6], to
further contribute to the worst-case condition. The foregoing
scenario corresponds to the maximum contribution of DG1
to the fault current (for our test system) and results in the
largest disparity between the fault currents passing through
the fuses and the recloser. If the coordination can be fulfill for
this scenario, it can be achieved for any other condition.
The first step of the coordination process is a fault analysis
based on Step 2 of the proposed algorithm (see Section III),
in order to determine I f min and If max . For the test system
of Fig. 3, it can be shown that the maximum fault current
corresponds to a fault taking place at the fuse end of Feeder
1, i.e. location Flt1, when DG1 is in place, whereas the fault
current is minimum when the fault takes place at the end of
the load feeder, location Flt2, in the absence of DG1; such
a calculation is made separately for phase and earth faults,
since different settings are usually employed for phase and
earth relays. Table I reports the fault analysis results for single-
and three-phase faults impacting Feeder 1, obtained through
simulation of the test system. The table also includes the value
of IR /IF , for each fault scenario. The fault analysis can also
 4

3
10
Ifmin
100E Fuse 3052A
TCT Ifmax
2 4661A
10
Relay
100E Fuse (Ext.Inverse,
MMT TD3)
1
10

time(s)
Recloser Slow
0 (Curve8*,TD16)
10

Fig. 4. Equivalent circuit for (IR /IF ) calculation 10

1

Recloser Fast
Original
TABLE I 10
2
(N104,TD2,Min. 0.018)

FAULT CALCULATION RESULTS FOR THE TEST SYSTEM Recloser Fast

Revised
(N104,TD1,Min. 0.012)
IR /IF
3
Fault Fault No DG with DG1 10
2 3 4
10 10 10
Location Type IF (A) IR (A) IF (A) IR (A) with DG1
current(A)
Flt1 ABC 3219 3219 4661 3011 0.646
A-G 1814 1814 2196 1548 0.705 Fig. 5. Coordination of protective devices of Feeder 1 for phase unit
Flt2 ABC 3052 3052 4231 2733 0.646
A-G 1737 1737 1901 1334.5 0.702
3
10
Ifmin
1737A Ifmax
100E Fuse
TCT 2196A
2
10
be conducted based on appropriate phase-sequence equivalent 100E Fuse
MMT
circuits, such as the one shown in Fig. 4 which corresponds 1
10
to the three-phase fault worst case; a similar equivalent circuit Relay
(Ext.Inverse,
time(s)
can also be constructed for a single-phase fault. It can be 0 Recloser Slow
TD1)
10
observed that the value of I R /IF based on the equivalent (Curve8*,TD10)

circuit of Fig. 4 closely agrees with the minimum value of 1

10
IR /IF in Table I. Based on Fig. 4, the value of I R /IF is Recloser Fast
Original
(N104,TD3)
IR ZDG1 2
10
= = 0.65,
Recloser Fast
(1) Revised
IF Zth + ZDG1 (N104,TD2,Min. 0.06)
3
10
where ZDG1 is the aggregate of the impedance of DG1 and 10
2
10
3
10
4

the impedance of the transformer that connects DG1 to the current(A)

network, and Z th is the network Thevenin impedance viewed
from the DG1 point of interconnection.
After selection of appropriate characteristic curves based on
Step 3 of the proposed methodology, Section III, the original
recloser fast characteristic curve is multiplied by the minimum
value of IR /IF , and the new curve is stored in the program
of the recloser. The revised characteristic curve is effective
even after the first opening of the recloser when DG1 gets
disconnected. The rationale is that any fault current that might
develop after the first action of the recloser will certainly be
between If min and If max , and, therefore, the coordination
made in Steps 2 through 5, Section III, is still applicable;
it is remembered that If max has been calculated with the
contribution of DG1 considered, whereas I f min is obtained
in the absence of DG1.
Figs. 5 and 6 illustrate the original and revised characteristic
curves employed for the phase and earth units. It should be
noted that the recloser must be coordinated with the fastest
fuse of the feeder (100E in this example). In this study, the
characteristics curves of the fuse and recloser are constructed
based on the analytical equations given in the Appendix.
C. Design of Directional Element
As pointed out in Section III, the proposed algorithm
requires a directional element on each feeder that embeds DGs,
to preclude false tripping when a fault impacts a neighboring
Fig. 6. Coordination of protective devices of Feeder 1 for earth unit
feeder. In such a scenario, the directional element of the feeder
under study rapidly block its own circuit breaker, for a pre-
specified period of time, to allow the protective devices of the
faulty feeder to isolate the fault on their own. The blockage,
however, does not continue any longer than the duration of
the relay reverse definite time; afterwards, the circuit breaker
is commanded to open, as a back-up protection. This concept
is illustrated in Fig. 7 where t d and td + tgm denote the
relay forward and reverse definite times, respectively, with D
denoting the output of the directional element. It is noted that
the reverse definite time is larger than its forward counterpart
by a suitable grading margin, t gm .
One potential issue that may arise due to the inclusion
of directional elements is that of protection loss in case
a fault directly impacts the substation. To circumvent this
issue, all circuit breakers are commanded to open once their
directional elements simultaneously detect a reverse fault. For
the test system of Fig. 3, this function is realized by the logic
circuit illustrated in Fig. 8, where D 1 and D2 signify the
outputs of the directional elements for Feeder 1 and Feeder 2,
respectively. The directional elements and their parameters are
introduced in more detail in the Appendix. Fig. 9 illustrates
the characteristic of the directional elements used in this paper,
for the test system.

Fig. 7. Diagrams showing a) forward and reverse definite times for
relays, and b)logics of implementation
Fig. 8. Symbolic logic circuit for substation protection
D. Impact of a DG on the Protection Coordination
Thus far, we have ignored the impact of DG2 on the
protection coordination. It is thus imperative to revisit the
design, in order to ensure that the coordination is preserved if
DG2 also comes on line. We start this by investigating the
impact of DG2 on its own feeder protection, and then its
impact on the coordination of Feeder 1 will be studied.
1) Impact on the recloser-fuse coordination of the host
feeder (Feeder 2): To investigate the impact of DG2 on
Feeder 2, the worst condition in which a three-phase fault
takes place downstream of DG2 is considered. Considering
DG1, a Thevenin equivalent circuit is derived for the network
Fig. 9. Characteristic of the directional element (phase comparison)
5
0.6
Zth=0.15pu
Z =0.2pu
0.4 th
Z =0.25pu
th
Zth=0.3pu
0.2
tR2 − M M TF 2 (s)
0
0.2
0.4
0.6 Decreasing Network Z =3.71pu
Thevenin Impedance DG2
0.8
0 0.1 0.2 0.3 0.4 0.5
(ZDG2 )−1 (pu)
Fig. 10. Curves characterizing the recloser-fuse grading time as a
function of (the inverse of) ZDG2
seen from the point of DG2 connection. The result will look
something similar to the equivalent circuit of Fig. 4, in which
subscript “1” is replaced by “2”. The equivalent circuit is used
for calculation of I F and IR which are substituted in (2) and
(3), Appendix, to determine the minimum melting time of
fuse F 2, i.e. M M TF 2 , and the operating time of recloser R2,
i.e. tR2 . The coordination is preserved if the recloser operates
faster than the fuse, that is, t R2 − M M TF 2 ≤ 0.
Fig. 10 illustrates the grading time t R2 − M M TF 2 as a
function of (the inverse of) Z DG2 [Fig. 3], for different network
Thevenin impedances. The figure shows that, for a given
network Thevenin impedance, the recloser-fuse coordination
is preserved if ZDG2 is larger than a minimum value. In
our example, this value is 3.71 pu for the network Thevenin
impedance of 0.2 pu. Also, for a given Z DG2 , the coordination
is preserved if the network Thevenin impedance is larger than
a threshold.
As will be discussed in Section IV-D3, the fuse-fuse coor-
dination of a feeder cannot be compromised by DGs on that
feeder, and therefore is not dealt with in this paper.
2) Impact on the recloser-fuse coordination of a neighbor-
ing feeder (Feeder 1): An investigation similar to the one
introduced in Section IV-D1 can be conducted to identify the
impact of DG2 on the coordination between recloser R1 and
fuse F 1, of Feeder 1. The results indicate that the value of
ZDG2 that causes miscoordination between R1 and F 1 is
considerably smaller than that needed for miscoordination in
Feeder 2. In our test system, this value is 0.395 pu for the
network Thevenin impedance of 0.2 pu. Thus, it is expected
in general that any impedance of a DG (Z DG ) that does
not disrupt the recloser-fuse coordination of its host feeder,
will not compromise that coordination of the neighboring
feeders embedding DGs if the proposed algorithm is used
for the protection coordination of the feeders. However, the
impedance of a DG may be so low that the recloser-fuse
coordination of the host feeder is lost. This may also result in
disruption of the recloser-fuse coordination in a neighboring
feeder. If this is the case, the coordination algorithm of Section
III must be executed for the neighboring feeder, as well as the
 6

0.06 3
Zth=0.1pu 10 Ifmax
2279A Ifmax
Z =0.15pu 80E Fuse 3570A
0.04 th TCT
2
Zth=0.2pu 10 Relay
80E Fuse (Ext.Inverse,
0.02 Zth=0.25pu TD2)
MMT
T CTF 12 − M M TF 1 (s)

1
10
0

time(s)
0 Recloser Slow
0.02 10 (Curve8*,TD12)

1
0.04 10
Recloser Fast
Decreasing Network Original
Thevenin Impedance (Curve8*,TD2,Min. 0.02)
0.06 2
10
Recloser Fast
Revised
0.08 (Curve8+,TD1,Min. 0.013)
Z =0.088pu 3
10
DG2
2 3 4
10 10 10
0.1
0 5 10 15 20 current(A)
(ZDG2 )−1 (pu)
Fig. 12. Coordination of protective devices of Feeder 2 for phase unit
Fig. 11. Curves characterizing the fuse-fuse grading time as a function
of (the inverse of) ZDG2
3
10 Ifmin
80E Fuse
TCT 1461A Ifmax
2004A
host feeder. 2
10
3) Impact on the fuse-fuse coordination of a neighboring 80E Fuse
MMT
feeder (Feeder 1): As pointed out in Section I, DGs may 1
10
Relay

time(s)
compromise fuse-fuse coordinations. Based on the design (Ext. Inverse,
TD1)
0 Recloser Slow
10
procedure of Section III, DGs cannot disrupt the coordinations (Curve8*,TD8)

between the fuses of their host feeder, since for any fault 1 Recloser Fast
10
Original
impacting the host feeder, the recloser is supposed to operate (B117,TD2)

and the DGs get dropped out. However, DGs may affect the 2
10
Recloser Fast
Revise
fuse-fuse coordination of neighboring feeders. For example, if (Curve8*,TD1.5,Min. 0.04)

a three-phase fault impacts branch L2, recloser R1 operates 3

first, and DG1 will be disconnected. Therefore, DG1 cannot
affect the coordination between F 1 and F 12; however, DG2
is still on line and can disrupt that coordination. This is
better shown by Fig. 11 where the coordination grading time
between F 1 and F 12, i.e. T CT F 12 − M M TF 1 , is plotted as
a function of (the inverse of) Z DG2 , for a three-phase fault
(worst condition) that has taken place at branch L2. Thus, the
fuse-fuse coordination is maintained if the minimum melting
time of F 1 is less than the total clearing time of F 12, that is,
T CTF 12 − M M TF 1 ≤ 0. As Fig. 11 indicates, this happens
if and only if Z DG2 is larger than 0.088 pu. It is seen again if
ZDG2 can keep coordination between recloser and the fuse, it
will certainly ensure the fuse-fuse coordination for the studied
system.
E. Retrieval of protection Coordination after System Alter-
ations
Let us consider the worst-case scenario in which the ca-
pacity of DG2 is two times the host feeder load, and its
impedance is 0.1 pu. For our test system, it turns out that
this condition results in disruption of the recloser-fuse coor-
dination of both feeders (see Sections IV-D1 and IV-D2), but
it does not affect the fuse-fuse coordination of Feeder 1 (see
Section IV-D3). Consequently, we need to repeat the algorithm
proposed in Section III for both feeders, with both DG1 and
DG2 considered, to retrieve the coordination. Had the fuse-
fuse coordination of Feeder 1 been also lost, we would have
needed to increase Z DG2 as the last step of the coordination
practice.
10
2
10
3
10
4
10
current(A)
Fig. 13. Coordination of protective devices of Feeder 2 for earth unit
Figs. 12 and 13 illustrate the original and the revised char-
acteristic curves, respectively, for the phase and earth relays.
The curves are for the devices of Feeder 2; the characteristic
curves for the devices of Feeder 1 are produced in a similar
fashion, but are not shown here due to space limitations.
V. S IMULATION R ESULTS
To verify the effectiveness of the proposed coordination
algorithm, the test system of Fig. 3 has been modeled and
simulated by PSCAD/EMTDC software. To that end, the same
characteristic curves used for the analytical studies of the
proceeding sections have also been utilized for simulation of
the relays, the reclosers, and the fuses. This section reports the
simulation corresponding to a three-phase fault. The results for
other types of fault are similar and thus not included.
Let us first assume that DG2 is not connected, the devices of
Feeder 2 are coordinated according to the traditional method,
and a permanent three-phase fault impacts the location X1 on
Feeder 2, at t = 0.5 s. Fig. 14 illustrates the current waveforms
of the recloser R1 and the fuse F 2. It can be observed that the
recloser operates twice in the fast mode of operation and then
the fuse interrupts the fault current. It should be noted that
the typical operating sequence of a recloser is fast-fast-slow-
slow, with a time delay between each two consecutive actions
[2], [5], [6]. Fig. 14 shows that the fault causes the current
 7

Current of Recloser R2 Current of Recloser R2

4 4

IR2 (kA)

IR2 (kA)
2 2
0 0
2 2
4 4
0.4 0.6 0.8 1 1.2 1.4 0.4 0.6 0.8 1 1.2 1.4
Current of Fuse F2 Current of Fuse F2
4 5

IF 2 (kA)
IF 2 (kA)

2
0 0
2
4 5
0.4 0.6 0.8 1 1.2 1.4 0.4 0.6 0.8 1 1.2 1.4
Current of Recloser R1 Current of Recloser R1
0.5 0.5

IR1 (kA)
IR1 (kA)

0 0

0.5 0.5
0.4 0.6 0.8 1 1.2 1.4 0.4 0.6 0.8 1 1.2 1.4
time(s) time(s)

Fig. 14. Coordination of devices when DG2 is not in place Fig. 16. Retrieval of coordination after the addition of DG2

Current of Recloser R2 Current of Recloser R2

4 4

IR2 (kA)
IR2 (kA)

2 2
0 0
2 2
4 4
0.4 0.6 0.8 1 1.2 1.4 0.4 0.6 0.8 1 1.2 1.4 1.6
Current of Fuse F2 Current of Fuse F2
5 5

IF 2 (kA)
IF 2 (kA)

0 0

5 5
0.4 0.6 0.8 1 1.2 1.4 0.4 0.6 0.8 1 1.2 1.4 1.6
Current of Recloser R1 Current of Recloser R1
0.5 0.5
IR1 (kA)

IR1 (kA)

0 0

0.5 0.5
0.4 0.6 0.8 1 1.2 1.4 0.4 0.6 0.8 1 1.2 1.4 1.6
time(s) time(s)

Fig. 15. Coordination loss when DG2 is added to the test system Fig. 17. Operation of the recloser in the slow mode when the fuse
fails to operate

of Feeder 1 to increase, but does not result in the operation

of Recloser 1; this is due to the blockage exercised by the
directional element of the relay of Feeder 1.
Fig. 15 shows the simulation results for the case where
DG2 is also present, while the devices are coordinated in the
same way as that for the previous case. It is observed that,
in this case the same three-phase fault results in the loss of
coordination, that is, F2 operates before Recloser 2.
Fig. 16 demonstrates the protection response for the case
where DG2 is in place and the coordination has been revised
as explained in Section IV-E. Thus, the proper sequence of
operation has been restored for the protective devices, and re-
closer R2 operates before F2 is damaged. In this scenario, once
the recloser opens for the first time, DG2 will be disconnected
subject to a brief delay and, consequently, the fault current
stops flowing in both the recloser and the fuse. Therefore, the
recloser and the fuse will be subjected to the same current
as in the scenario without DG2, once the recloser makes the
circuit. Nonetheless, the coordination remains effective. If the
fuse fails to operate, the recloser takes action in the slow mode
and break the fault current, as shown in Fig. 17.
VI. C ONCLUSION
This paper investigated the impact of distributed generation
on the protection coordination of radial distribution networks.
First, a simple and effective protection strategy was proposed
that provides coordination between protective devices of a
radial distribution system embedding distributed generators.
The proposed strategy benefits from microprocessor-based
reclosers and directional elements for feeder relays, and ad-
dresses the issues of blinding of protection, false tripping,
fuse-recloser and fuse-fuse miscoordination, and failed auto-
reclosing. Moreover, the proposed strategy does not require
adaptive protective devices. Then, the impact of future addi-
tions of distributed generators on the coordination was investi-
gated and the requirements for preservation of the coordination
were identified. It was further discussed that the proposed
algorithm could be applied for the worst-case scenario, to
reclaim the coordination in case the requirements are not to
be met. The simulation results for a test system indicated the
effectiveness of the method.
A PPENDIX
A. Analytical Equations of Protective Devices
1) Fuse: A Fuse is typically characterized by two character-
istic curves: Minimum Melting Time (MMT) curve and Total
Clearing Time (TCT) curve. The MMT curve indicates the
time period between an overcurrent inception and the instant
when the fuse starts to melt, whereas the TCT curve specifies
 8

the total time elapsing from the beginning of an overcurrent R EFERENCES

to the circuit interruption. The fuse characteristic curves can [1] P. P. Barker, R. W. de Mello, ”Determining the impact of distributed
be approximated by an exponential function whose argument generation on power systems: part I-radial distribution systems”, IEEE
is a polynomial of the natural logarithm of the fuse current Trans. on Power Delivery, vol. 15, pp. 486-493, Apr. 2000.
[2] A. Fazanehrafat, S. A. M. Javadian, S. M. T. Batbaee, M. R. Haghifam
[9], [12]: ”Maintaining the recloser-fuse coordination in distribution systems in
 k  presence of DG by determining DG’s size”, in Proc. IET 9th International
 Conference on Developments in Power System Protection (DPSP), pp.
n
tf = exp an ln (If ) , (2) 132-137, Mar. 2008.
n=0 [3] K. Maki, S. Repo, P. Jarventausta, ”Methods for assessing the protection
impacts of distributed generation in network planning activities”, in Proc.
in which k is the polynomial order, I f is the current mag- IET 9th International Conference on Developments in Power System
nitude, and t f is MMT or TCT; the coefficients a n are Protection (DPSP), pp. 484-489, March 2008.
determined through a curve-fitting exercise. In this study, [4] N. Hadjsaid, J. F. Canard, F. Dumas ”Disperse generation impact on
distribution network”, IEEE Computer Application In Power, Vol. 12,
a fifth-order polynomial function of the natural logarithm Issue 2, pp. 22-28, Apr. 1999.
of the fuse current has been used to approximate the fuse [5] A. Girgis and S. M. Brahma, ”Effect of distributed generation on
characteristic curve, i.e. k = 5, the coefficients of which are protective device coordination in distribution system,” in Proc. Large
Engineering Systems Conference on Power Engineering, pp. 115-119,
given in Table II. Jul. 2001.
2) Recloser and Relay: Similarly, relays and reclosers can [6] S. M. Brahma and A. Girgis, ”Microprocessor-based reclosing to coor-
be characterized by inverse-time characteristic curves. Such dinate fuse and recloser in a system with high penetration of distributed
generation,” in Proc. IEEE Power Eng. Soc. Winter Meeting, vol. 1, pp.
curves are generally represented by the following equation 453-458, 2002.
[13]: [7] G. Tang and M. R. Iravani, ”Application of a fault current limiter to
⎡ ⎤ minimize distributed generation impact on coordinated relay protection,”
A presented at the Int. Conf. Power Systems Transients, Montreal, QC,
top (I) = T D ⎣  p + B⎦ , (3) Canada, Jun. 2005.
I
Ipick−up − 1 [8] ”Power System Protection”, Vol. 1-4, Published by Institution of Electri-
cal Engineers, London, UK, Second Edition, 1995
[9] S. Chaitusaney and A. Yokoyama, ”Impact of Protection Coordination on
where top is the device operating time, I is the device current, Sizes of Several Distributed Generation Sources”, in Proc. 7th Interna-
T D is the time dial setting, Ipick−up is the device current set- tional Power Engineering Conference (IPEC), Vol. 2. pp 669-674, Dec.
point, and A, B and p are constant parameters [13], [14]. 2005
[10] IEEE Standard for Interconnecting Distributed Resources with Electric
TABLE II Power Systems, IEEE Std. 1547-2003
C OEFFICIENTS OF FUSE CHARACTERISTIC CURVES [11] IEEE Recommended Practice for Industrial and Commercial Power
System Analysis, IEEE Std. 399-1997.
Fuse a0 a1 a2 a3 a4 a5 [12] S. Santoso and T. A. Short, ”Identification of fuse and recloser operations
Type in a radial distribution systems”, IEEE Trans. on Power Delivery, Vol.
80E,MMT 2976.4 -5341.2 3836.2 -1376 246.4 -17.6 22, No. 4, pp. 2370-2377, Oct. 2007
80E,TCT -77.68 192.64 -156.9 58.6 -10.46 0.73 [13] IEEE Standard Inverse-Time Characteristic Equations for Over-current
100E,MMT -138.13 354.83 -306.54 122.4 -23.38 1.74 Relays, IEEE Standard C37.112-1996, September 1996.
100E,TCT 568.41 -859.72 528.3 -164 25.58 -1.6 [14] ”ABB PCD Control Protection Curves”, online available at:
125E,MMT -1849.8 3143.8 -2104.6 696.1 -114 7.4 http://library.abb.com/GLOBAL/SCOT/scot235.nsf/VerityDisplay
125E,TCT 498.37 -659.1 348.22 -91.14 11.69 -0.58 [15] T. S. M. Rao, ”Power system protection: static relays with microproces-
150E,MMT -214.59 203.02 -57.515 7.07 -2.69 0.31 sor applications”, 2nd Edition, New Delhi, Tata Mc. Graw-Hill, c1989
150E,TCT 1453.8 -2110.3 1231.1 -359.9 52.61 -3.07
M. Amin Zamani (S’09) received the B.Sc. degree from Iran University
of Science and Technology (IUST), Tehran, Iran, and the M.Sc. degree
form Shahid Chamran University (SCU), Ahvaz, Iran, in 2003 and 2005,
respectively. Presently, he is a Research Assistant with the University of
B. Directional Characteristic Western Ontario, London, Ontario, Canada, working towards the Ph.D. degree.
His research area includes power system protection, control, and monitoring;
The directional characteristic function has been realized in distributed generation, power electronic converters, and microgrids.
this paper through a comparator whose inputs S 1 and S2 are
Tarlochan S. Sidhu (M’90-SM’94-F’04) is a Professor and Chair of the
derived from the feeder voltage and current, as follows [8], Electrical and Computer Engineering Department at the University of Western
[15]: Ontario, London, ON, Canada. He is also the Hydro one Chair in Power
Systems Engineering. From 1990 to 2002, he was with the Department
S1 = |S1 |∠α1 = σV , S2 = |S2 |∠α2 = χI, (4) of Electrical Engineering, University of Saskatchewan, as a Professor and
Graduate Chairman of the department. His research interests include power
where σ is a real number, and χ = |χ|∠θ is a complex number system protection, monitoring, control, and automation.
(of impedance dimension) which, conceptually, emulates the Dr. Sidhu is a Fellow of the IEEE, a Fellow of the Institution of Engineers
(India), and a Fellow of the Institution of Electrical Engineer (U.K.). He is a
feeder impedance and thus converts the feeder current into a Registered Professional Engineer in the Province of Ontario and a Chartered
suitable voltage. The comparator identifies a fault as a forward Engineer in the U.K.
fault, i.e. D = 1, if Amirnaser (Amir) Yazdani (M’05-SM’09) received the Ph.D. degree from
π π the University of Toronto, Toronto, ON, Canada, in 2005. He was with Digital
− ≤ α1 − α2 ≤ or Predictive Systems (DPS) Inc., Mississauga, ON, Canada. Currently, he is an
2 2
π π Assistant Professor with the University of Western Ontario, London, ON,
− + θ ≤ ϕl ≤ + θ, (5) Canada. His research interests include dynamic modeling and control of
2 2 electronic power converters, distributed power generation and energy storage,
where ϕl = ϕv −ϕi is the phase difference between the voltage renewable energy, and microgrids. Dr. Yazdani is a Professional Engineer in
the Province of Ontario.
and the current.