IET Generation, Transmission & Distribution

Research Article

Protection coordination for networked ISSN 1751-8687

Received on 10th April 2019

microgrids using single and dual setting

Revised 27th March 2020
Accepted on 20th April 2020
E-First on 20th May 2020
overcurrent relays doi: 10.1049/iet-gtd.2019.0557
www.ietdl.org

Mahamad Nabab Alam1 , Ramakrishna Gokaraju2, Saikat Chakrabarti1

1Department of Electrical Engineering, Indian Institute of Technology, Kanpur-208016, India
2Department of Electrical and Computing Engineering, University of Saskatchewan, Saskatoon, SK S7N 5A9, Canada
E-mail: [email protected]

Abstract: The operation of multiple microgrids (MGs) with distribution system enhances reliability and resiliency of the power
supply significantly by enabling high penetration of locally available distributed energy resources (DERs). However, there are
several issues in the proper operation of multiple MGs that need to be addressed. Providing proper protection coordination is
one of the most important issues in the coordinated operation of networked MGs (NMGs). In this study, an efficient protection
coordination scheme for NMGs is proposed by utilising the commonly used numerical directional overcurrent relays (DOCRs)
with single and dual settings. Issues related to the operation of NMGs such as the change in network topology, connection
status of the point of common coupling and low voltage fault ride through of DERs have been considered. The protection
coordination problem has been formulated as an optimisation problem and solved using the interior point method. The proposed
protection coordination scheme of NMGs has been tested on the 69-bus radial distribution system with the integration of three
MGs, which is not reported in earlier literature. The suitability and effectiveness of the proposed DOCRs coordination scheme
have been demonstrated.

1 Introduction protection coordination among the protective devices in NMGs is a

challenging task.
Networking of multiple self-governed microgrids (MGs) are Reclosers and fuses are mostly used for protection of radial DSs
emerging as one of the best alternatives to improve resiliency and because they are effective, cheaper and sufficient. However, proper
reliability of power system networks by utilising locally available protection coordination using recloser and fuses is difficult to
cost-effective and environment-friendly electric power generated achieve in radial DSs when multiple distributed generations (DGs)
from distributed energy resources (DERs) [1]. Photovoltaic (PV) are present [14–16]. These studies suggest that the location and
and wind turbines (WTs) are the two major sources of energy in penetration of DGs are responsible for such miscoordinations.
DERs. Advanced control and communication technologies enable Thus, reclosers and fuses may not be able to provide sufficient
the operation of multiple MGs with the distribution system (DS) to protection coordination in NMGs.
supply day-to-day growing electrical energy requirement On primary DSs, overcurrent protection is predominantly used
economically and efficiently. Networking of multiple MGs either [17]. Recently, several studies have been reported for protection of
to the grid or another grid-connected or isolated MG can enable interconnected DSs using directional overcurrent relays (DOCRs)
more reliable and economic power supply to the consumers. in the presence of DGs [18–20]. These studies demonstrate that
During emergencies, multiple MGs may continue supply to the DOCRs can provide proper coordination in the presence of one or
critical loads for longer periods, depending on the capacity of the more DGs in interconnected DSs. Although settings of DOCRs
energy storage system (ESS) of each MGs, and help to restore need to be revised based on DG's location and maximum short-
supply as soon as possible by providing black-start support to the circuit level, proper protection coordination is always maintained.
conventional power stations [2]. Also, it can provide the In [21], a time-current–voltage characteristic based DOCR has
opportunity for competitive ancillary market [3]. However, been proposed to protect an autonomous MG. Additionally, for
operating multiple MGs requires them to maintain the same protection of interconnected DSs with an MG, several studies have
operational reliability as a conventional power system before they been presented [22–24]. In [22], protection of MG using DOCRs
become more common. System of multiple MGs connected to the with agent-based communication has been proposed. Proper
DS is termed as networked MGs (NMGs). coordination is dependent on a reliable communication link in this
Various issues related to the planning and operation of NMGs scheme. Also, intelligent electronic devices, synchrophasor and
have been discussed in the literature. Optimum energy supervisory control and data acquisition system are required, which
management of NMGs has been analysed in [4–6]. Different types may make the overall scheme uneconomical for MG of small size.
of control strategies suitable for the operation of NMGs have been In [23], protection coordination of DOCRs has been proposed for
discussed in [7–9]. Resiliency and reliability improvement of the the protection of DS with the grid-connected and islanded mode of
overall system with the networking of multiple MGs have been operation of MG by utilising fault current limiter at the point of
analysed in [10–12]. However, little attention has been given to the common coupling (PCC). In [24], an improved protection
protection coordination issues of NMGs [13]. Proper protection coordination scheme by utilising dual-setting DOCRs has been
coordination among the protective devices of NMGs is the key to discussed for protection of DS with both grid-connected and
the successful operation of the entire system. Any miscoordination islanded mode of MG operation without fault current limiter. A
during an electrical fault can compromise the security of the entire significant reduction in the overall operating time of primary and
system. Short-circuit level of the distribution feeder increases backup relays has been shown in this study without any
significantly once multiple MGs are integrated to it. Further, the miscoordination. However, in all these studies, only one MG have
unidirectional flow of currents no longer exists in the feeder been considered, whereas NMGs may have several MGs. Also, two
because of the presence of multiple MGs. Thus, providing proper DOCRs have been used for protection of each line, which makes
the protection scheme costly.

IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828 2818
© The Institution of Engineering and Technology 2020
 17518695, 2020, 14, Downloaded from https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-gtd.2019.0557 by Cochrane Portugal, Wiley Online Library on [30/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
single MG in the DS. This scheme with some required
modifications can be extended for the protection of NMGs.
In this study, an efficient protection coordination scheme for
NMGs is proposed by utilising single-setting and dual-setting
DOCRs. The novelty of the proposed methodology compared to
the earlier methods proposed in the literature is that it finds suitable
locations for placing single and dual-setting DOCRs and also uses
a minimum number of the two schemes for protection. All possible
Fig. 1 A typical NMG system topologies of NMGs have been considered in the scheme.
Protection coordination issues of NMGs has been formulated as an
Table 1 Connection status of NMG system optimisation problem and solved using Open Source OPTI Toolbox
Particular Status of switches Topological remarks in MATLAB environment [25]. The proposed scheme has been
SW1 SW2 SW3 tested on the 69-bus radial DS in the presence of three MGs. Also,
the suitability and effectiveness of the proposed protection scheme
case 1 on off on multiple MGs on a local feeder
have been demonstrated.
case 2 off on on multiple MGs on main feeder This article is organised as follows: Section 2 discusses
case 3 on on off MGs on different feeders protection coordination issues of NMGs. Section 3 explains the
case 4 on on on multiple MGs in loop problem formulation of NMGs protection. The details of the test
case 5 off off on islanded but interconnected MGs systems and simulation set-up are given in Section 4. Results and
case 6 on off off one connected, other islanded discussions are given in Section 5, and the conclusions are
case 7 off on off one connected, other islanded presented in Section 6.
case 8 off off off islanded MGs
2 Protection coordination issues with networked
microgrids
Proper protection coordination is one of the key requirements to
run NMGs with high reliability. Change in network topology may
be frequent in these systems, which causes protection coordination
issues. The magnitude and direction of fault currents can be
different in various topologies of the system [26]. Fig. 1 shows a
typical NMG system. This system is having two local feeders along
with the main feeder supplied from the grid through a step-down
transformer, and two MGs. There are three switches to enable
connection of MGs with the DS. The possible connection status of
the MGs in the DS is given in Table 1. The connection diagrams of
the first four topologies mentioned in Table 1 are shown in Fig. 2.
Further, protective devices and three fault locations are shown in
each sub-figure of this figure. It is to be noted that although three
fault location have been indicated in these sub-figure but only a
single fault location to be considered at once for the analysis.
Simultaneous occurrence of multiple faults are rare in the system.
From Fig. 2, it can be observed that during normal and fault
conditions, the current will be contributed from the main grid and
MGs in all cases. Figs. 2a and b show the operation of multiple
MGs through feeder 1 and the main feeder, respectively. Fig. 2c
shows the operation of MG1 connected to feeder 1 and MG2 to the
main feeder, whereas Fig. 2d displays looped interconnection of
the two MGs. It is to be noted that all these configurations are
decided time-to-time by the distribution system operator (DSO),
depending on the optimum energy scheduling of the entire
distribution network. Table 2 shows the description of fault
currents passing through various protective devices placed in the
system under all these four cases. All the faults are considered to
be bolted three-phase fault. For other types of faults (such as
phase-ground, phase-phase, phase-phase-ground), only magnitude
would be different but the fault current pattern will remain the
same as have been mentioned in Table 2.

2.1 Fault current under different network topologies

From Table 2, it is observed that the fault current passing through
protective device A is IS for fault 1, −IN for fault 2 and fault 3 in
case 1. Further, the fault current passing through A is IS + IN for
Fig. 2 Some possible topologies of NMG system and fault locations (single
fault to be considered)
fault 1 and zero for fault 2 and fault 3 in case 2. Also, the fault
(a) Case 1: multiple MGs on feeder 1, (b) Case 2: multiple MGs on the main feeder,
current passing through A is IS + IM2 for fault 1 and −IM1 for fault 2
(c) Case 3: mG1 on feeder 1 and MG2 on the main feeder, (d) Case 4: multiple MGs in
and fault 3 in case 3. This indicates that both the magnitude and
loop with local and the main feeder
direction of fault currents passing through protective device A
varies, as configuration of NMGs changes. Similar observations
From the above discussion, it is clear that protection can be made for protective devices B and D. Also, it can be
coordination issues of NMGs have not been fully addressed yet. observed from Fig. 2 that the flow of current through various
Also, recloser and fuses cannot guarantee proper protection switches (SW1, SW2 and SW3) reverses as the NMG's topology
coordination in NMGs. Dual-setting DOCRs have the ability to changes. On the other hand, only the magnitude of the fault
provide excellent protection coordination in the presence of a

IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828 2819
© The Institution of Engineering and Technology 2020
 17518695, 2020, 14, Downloaded from https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-gtd.2019.0557 by Cochrane Portugal, Wiley Online Library on [30/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Table 2 Fault current description in various configurations 2.2 Cases of miscoordination
of NMGs
The coordination time interval (CTI) of a relay pair is defined as
Configuration Fault Fault current description
the time gap between the operating time of primary relay (top) and
location
its backup relay (tob). Mathematically, it is expressed as
case 1 fault 1 IF = IS + I N , I R = IS
I A = IS, IB = 0, IC = 0, ID = 0 CTI = tob − top (1)
fault 2 I F = I S + I N , I R = I S, I D = 0
I A = − I N , I B = IF , IC = IF The protection coordination status of a primary-backup relay
fault 3 I F = I S + I N , I R = I S, I C = 0 pair is identified as follows:
I A = − I N , I B = IF , I D = IF
• CTI ≥ MCT: coordination holds
case 2 fault 1 I F = I S + I N , I R = I S, I C = 0 • CTI < MCT: coordination lost
I A = IF , I B = − I N , I D = − I N
fault 2 I F = I S + I N , I R = I S, I A = 0 Here, minimum coordination time (MCT) is the minimum time
I B = I S, I C = I F , I D = − I N gap required between operating times of primary and its backup
relay for proper coordination. The value of MCT lies in 0.2–0.3 s
fault 3 I F = I S + I N , I R = I S, I A = 0
[20].
IB = IS, IC = 0, ID = IF The change in magnitude and direction of fault currents under
case 3 fault 1 IF = IS + IM1 + IM2, IR = IS, IC = 0 different NMG topology makes conventional overcurrent
I A = IS + IM2, IB = ID = − IM2 protection schemes extremely difficult to coordinate properly [14–
16]. Protection coordination issues of networking MGs with DS is
fault 2 IF = IS + IM1 + IM2, IR = IS, I A = − IM1
given in Table 3. Here, first three cases of Table 1 (Figs. 2a–c) has
IB = IS + IM1, IC = IF, ID = − IM2 been analysed. Coordination status of primary-backup relay pairs
fault 3 IF = IS + IM1 + IM2, IR = IS, IC = 0 A − R, C − B and D − B has been shown in Figs. 3a–c,
I A = − IM1, IB = IS + IM1, ID = IF respectively. It is assumed that all these relays follow inverse
case 4 fault 1 IF = IS + IM1 + IM2, IR = IS, IC = 0 definite minimum time (IDMT) characteristic.
From Table 3 and Fig. 3a, it is observed that the fault current
I A, IB, and ID will be mixed of IS, IM1 and
passing through backup relay R remains the same (i.e. IS) in
IM2
different networked topologies, whereas the current through the
fault 2 IF = IS + IM1 + IM2, IR = IS, IB = IF primary relay A is higher. This causes the primary relay to operate
I A, IB, and ID will be mixed of IS, IM1 and faster; whereas the backup relay takes the same amount of time as
IM2 it was taking earlier (without MGs). Thus, there is not a single case
fault 3 IF = IS + IM1 + IM2, IR = IS, IC = 0 of miscoordination for relay pair A–R. On the other hand, from
Table 3 and Fig. 3b, it is observed that the fault current passing
I A, IB, and ID will be mixed of IS, IM1 and
through primary relay C is very high (i.e. IS + IN ) and that through
IM2
backup B is also higher. Thus, the operating times of both primary
Positive direction of the flow of current is from substation towards and backup relays are reduced as shown in Fig. 3b, which causes
load. miscoordination in various cases. Similarly, it is observed from
Table 3 and Fig. 3c that the fault current passing through both
primary and backup relays are higher and thus their operating times
Table 3 Protection coordination status of NMG system get reduced. These reductions in the operating times will reduce
Particular Relay pair Fault currents through Coordination CTI of the pair because of inverse time-current characteristics
(primary- Primary Backup Status (TCCs) of relays and thus this pair miscoordinates in various cases.
backup) From the above discussion, it is clear that the existing
fault 1, case 1 A−R IS IS holds overcurrent protection scheme cannot coordinate properly when
multiple MGs are connected to the DS and operated as NMGs.
fault 1, case 2 A−R IS + I N IS holds
Therefore, a new overcurrent protection scheme based on single
fault 1, case 3 A−R IS + I M 2 IS holds and dual-setting DOCR for NMGs is proposed in the next section.
fault 2, case 1 C−B IS + I N IS + I N lost
fault 2, case 2 C−B IS + I N IS holds 3 Formulation of protection coordination problem
fault 2, case 3 C−B IS + I N IS + I M 1 lost for networked microgrids
fault 3, case 1 D−B IS + I N IS + I N lost Commonly, protection coordination problem is to make sure that
fault 3, case 2 D−B IS IS holds there is no misccordination among any primary-backup relay pairs
fault 3, case 3 D−B IS + I M 1 IS + I M 1 lost
for any network scenarios. In addition, the operating times of all
the relays should be as small as possible so that it does not result in
equipment (line/cable) failure and stability issues in the system.
Therefore, the protection coordination problem of NMGs using
currents passing through protective device C is different under DOCRs can be formulated as an optimisation problem, where the
changing configurations of NMGs. objective is to minimise the sum of operating times of all the relays
It is to be noted that reversal of current through protective subject to maintaining proper coordination among them under
devices A, B and D may cause serious protection issues. Therefore, various NMG topologies. Mathematically, this problem can be
these devices must have the ability to discriminate the fault expressed as
occurred on either side. Also, these devices should discriminate the
fault with different pickup current values depending on the fault N m n
side of device. It is because the magnitude of fault current would OF = min ∑ ∑ topR ik + ∑
, tfwR, jk + trvR, jk (2)
be different depending on where the fault is upstream or k=1 i=1 j=1
downstream of the device. Dual-setting DOCRs can be the ideal
choice to provide proper discrimination in such cases [24]. subject to

tobR, l − topR, i ≥ MCT; ∀i, ∀k (3)

2820 IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828
© The Institution of Engineering and Technology 2020
 17518695, 2020, 14, Downloaded from https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-gtd.2019.0557 by Cochrane Portugal, Wiley Online Library on [30/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
In (2), topR, ik is the operating time of relay Ri in the kth topology, commonly set to 0.14, 0.02 and 0, respectively, for IDMT relay
tfwR, jk and trvR, jk are the operating times in forward and reverse [27]. The value of TMS for a DOCR is specified by the
direction mode, respectively, of dual-setting relay R j in the kth manufacturers within the range of [0.1, 1.1], whereas the value of
topology, m and n are the number of single and dual-setting PS is determined based on the maximum load current passing
DOCRs, respectively, N is the total number of possible system through the relay. The following equation is used to determine PS
topology in the DS with NMGs, and OF is the objective function to value of single-setting DOCR Ri:
be minimised. In (3), MCT is time required between the operating
times of primary and the corresponding backup relays and tobR, l is PSop, i = OLF × ILmax, op, i (5)
the operating time of backup relay l for a primary relay i. It is to be
noted that tobR ∈ {topR, tfwR, trvR} and index l is a set of backup where ILmax, op, i is the maximum value of load current passing
relays for each primary relay i. In order words, a backup relay for a through Ri and OLF is the overload factor, which is normally
primary relay is either another single-setting relay or dual-setting considered as 1.25.
relay in forward/reverse direction. In these two equations, index i is Similarly, the operating times of dual-setting DOCR in forward
the set of single-setting primary relays, index j is the set of dual- and reverse direction can be expressed as follows:
setting primary relays, and index k is the set of possible system
topologies. α
tfwR, j = TMSfw, j × +γ (6)
The operating time of single-setting DOCR is defined as β
Ifw, j /PSfw, j − 1
follows [27]:
α
α trvR, j = TMSrv, j × +γ (7)
(4)
β
topR, i = TMSop, i × β
+γ Irv, j /PSrv, j − 1
Iop, i /PSop, i − 1
In (6), TMSfw, j, PSfw, j and Ifw, j are the values of TMS PS and
In (4), Iop, i is the maximum fault current passing through relay the maximum fault current passing through relay R j in forward
Ri; TMSop, i and PSop, i are time-multiplier setting (TMS) and plug- direction, whereas in (7), TMSrv, j, PSrv, j and Irv, j are the values of
setting (PS), respectively, of relay Ri; and three constants, α, β and TMS, PS and the maximum fault current passing through relay R j
γ are the characteristic coefficients of a DOCR. The parameters α, in reverse direction. Also, PSfw, j and PSrv, j for dual-setting DOCR
β and γ are fixed for a particular type of relay, which are
R j are determined as:

PSfw, j = OLF × ILmax, fw, j (8)

PSrv, j = OLF × ILmax, rv, j (9)

where ILmax, fw, j and ILmax, rv, j are the maximum value of load
current passing in forward and reverse direction, respectively, of
the dual-setting DOCR R j. In the proposed formulation, a
constraint is imposed on TMS values of each DOCR to keep it
within the range as follows:

TMSmin ≤ TMSop, i ≤ TMSmax (10)

TMSmin ≤ TMSfw, j ≤ TMSmax (11)

TMSmin ≤ TMSrv, j ≤ TMSmax (12)

where TMSmin and TMSmax are the minimum and maximum limits
on TMS values of DOCRs, respectively.

3.1 Protection coordination under various types of faults

Satisfying coordination constraint (3) for the maximum fault
current (i.e. three-phase fault) ensures that the relays coordinate
well for the lower values of fault currents. This is because of the
inverse nature of TCCs of the relays. The other types of faults,
such as double-line-ground, line-line and line-ground faults are not
as severe as the three-phase fault and cause lower fault currents to
flow through the relays. The relays not only discriminate these
faults but also coordinate well for all topologies. Fig. 4 shows the
typical TCC of a primary-backup relay pair. The minimum and
maximum fault values of currents passing through the primary and
backup relays of a pair for all of the different network topologies
for all types of faults have been shown in the figure. PS multiplier
(PSM), i.e. PS times of the fault current passing through the relay
pair has been shown on the horizontal axis. A relay cannot operate
unless the value of PSM becomes greater than unity.
From Fig. 4, it can be observed that the value of CTI (shown
with double-headed arrows) increases as the characteristics move
from right to the left. This indicates that once the CTI requirement
is maintained at the maximum values of fault current, then for the
Fig. 3 Operating times of relays under different topologies for lower values of fault currents, it would definitely be satisfied.
(a) Relay pair A − R, (b) Relay pair C − B, (c) Relay pair D − B Therefore, if coordination constraint of a pair is maintained for the

IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828 2821
© The Institution of Engineering and Technology 2020
 17518695, 2020, 14, Downloaded from https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-gtd.2019.0557 by Cochrane Portugal, Wiley Online Library on [30/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Fig. 4 Typical TCC of a primary-backup pair under different topologies

Fig. 5 Fault ride-through capability of DERs

(a) LVFRT time–voltage profile, (b) Corresponding reactive current requirements

maximum fault current (bolted three-phase faults), the coordination (PF) of 0.9 and 0.85, respectively, whereas the total load at MG3 is
would automatically be maintained for the other type of faults. It is 1.2 MVA at PF 0.8. A standby diesel engine generator (DEG) is
because the value of current passing through the relays will be available with each MG to fully supply local loads during an
lesser in these cases as compared to the three-phase fault. The same emergency (when PV, WT and ESS are unable to supply the loads
is true for all possible network topologies. properly). The substation has a short-circuit capacity of 50 MVA,
whereas MG1, MG2 and MG3 have a short-circuit capacity of 6, 6
3.2 Low voltage fault ride-through capability of DERs and 10 MVA, respectively. The maximum short-circuit level of
each MG is about five times their generation, which is justifiable as
The DERs (PV and WT) have low voltage fault ride through the power is getting supplied by PV systems, WTs and standby
(LVFRT) capability [17]. Hence, the DERs will remain connected diesel engine based synchronous generators (SGs) [21, 34, 35]. The
during any transient fault in distribution network [28, 29]. In the SGs are the major contributor to the fault current in MGs as
study, the German grid codes are taken into consideration for the compared to PV panels and WTs. The DERs (PV and WT) have
DERs during LVFRT, which is shown in Fig. 5 [30, 31]. The dynamic voltage ride-through capability as mentioned in the
connection status of a DER during transient has been considered as previous section. The ESS in each MG is considered to supply their
shown in the figure. Additionally, for a fault in the distribution own critical loads alone for 4-hours and with DEG for 24-hours.
network, the MGs get islanded without disconnecting any of the For performing short-circuit analysis, all the inverter-based devices
DERs within the MG. However, for internal faults within the MG, need to modelled as discussed in [36].
DERs will be disconnected (as discussed in 5), and islanding will It is to be noted that protection of NMGs using single and dual-
also take place. settings DOCRs is the main objective of the presented work.
Fig. 5a indicates that following a grid fault, DERs should Stability, control and energy management of the system are not the
remain connected if their terminal voltage is above the time- focus of this study. Therefore, necessary assumptions have been
voltage profile, else they get disconnected. It can be seen from Fig. made for them as mentioned above.
5a that the DERs get disconnected at 0.22 s which is as per the To provide complete protection of this system, a total of 11
IEEE 1547-2018 standard [17]. The necessary reactive current single-setting DOCRs and 9 dual-setting DOCRs are required
required to maintain the voltage profile shown in Fig. 5a is given in which have been shown in Fig. 6. Selection of single-setting and
Fig. 5b. The above LVFRT codes have been considered to calculate dual-setting DOCRs have been done so that the entire system
various fault currents required for protection coordination. remains protected with proper coordination using the least number
of DOCR. The DOCRs through which the flow of current may be
4 Test system and simulation setup bi-directional have been considered as dual-setting DOCR. These
DOCRs can operate for the fault on any side (downstream or
The proposed protection coordination scheme for NMGs is tested upstream) of the relay with different settings [24]. The maximum
on the modified 69-bus radial DS with three MGs [32, 33]. Fig. 6 load current passing through all the relays have been calculated
shows the modified 69-bus radial DS with tie-switches and MGs. using backward-forward sweep method [37] in all radial topologies
There are five PCCs in the three MGs, which form a system of and Netwon–Raphson load flow method in all meshed topologies
networked MGs. Negative sequence and zero sequence values of [38]. The maximum fault current passing through all the relays
various parameters of this three-phase system are assumed to be have been calculated using bus impedance matrix (Zbus) method
equal and 2.5 times the corresponding values of positive sequence
[38]. The other types of faults such as double-line-to-ground,
parameters, respectively. The parameters of connecting lines of
single-line-to-ground and line-to-line faults have also been
MGs is the same as the parameter values of tie-lines of the 69-bus
calculated using the method discussed in [38]. These methods for
system. The line resistance and reactance of each line in all MGs
calculating maximum load and fault currents have been applied
are set to 0.0044 and 0.0108 Ω, respectively. The DERs of each
repeatedly in each possible NMG configuration. Software
MG have solar PV panels, doubly fed induction generator (DFIG)-
programs have been written in MATLAB environment for
based WTs, ESS with enough capacity to supply its own critical
calculating various load and fault currents.
loads for 4-hours and also has the potential to contribute some
power to the external network. Total power generation through the
DERs at MG1 and MG2 is 1.2 MW each, and that at MG3 is 2
MW. The total load at MG1 and MG2 is 0.8 MVA at power factor

2822 IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828
© The Institution of Engineering and Technology 2020
 17518695, 2020, 14, Downloaded from https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-gtd.2019.0557 by Cochrane Portugal, Wiley Online Library on [30/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Fig. 6 Modified 69-bus radial distribution system with multiple MGs

Table 4 Some important NMG topologies in the distribution on the economic dispatch of the entire system. The flow of currents
network created with opening and closing of PCCs during faults will be very different in both the magnitude and
Sl. No. Status of PCCs direction, which requires special attention while obtaining
PCC1 PCC2 PCC3 PCC4 PCC5 optimum setting of the relays protecting the entire system.
1 closed open open open open
2 open closed open open open
4.2 Primary-backup relay pairs in the system
3 open open closed open open In the looped configuration of the network (Sl. No. 12 of Table 4),
4 closed closed open open open there are 70 primary-backup relay pairs caused by all near-end and
5 closed open closed open open far-end faults in the system. Once coordination is maintained for
6 open closed closed open open
these pairs, the entire system will remain protected in this
configuration. To maintain protection coordination for all 32
7 closed closed open closed open
topologies, the constraint related to primary-backup relay pairs in
8 open closed closed closed open these cases must be considered in the optimisation problem. The
9 closed closed closed open open numbers of primary-backup relay pairs in the other topologies are
10 closed closed closed open closed less because of disconnection of MGs and opening of lines. The
11 closed closed closed closed open total number of effective primary-backup relay pairs is 1623.
12 closed closed closed closed closed
4.3 Necessary considerations and optimisation method
Dual-setting relays are indicated as R7, R9, R12 and R15–R20 in
4.1 Possible NMG topologies Fig. 6. The operation in reverse direction of these dual-setting
In the 69-bus radial network with three MGs and five PCCs, a total DOCRs has been represented as R7*, R9*, R12* and R15*–R20*.
of 32 NMG topologies are possible. The total number of possible Thus, R7 and R7* indicate forward and reverse direction operation
NMG topologies can be calculated using the following equation of relay R7, respectively. The CTI for all primary-backup relay
[13]. pairs is considered as 0.2 s. To solve the proposed protection
coordination problem, open source IPOPT solver available in OPTI
(13) Toolbox [25] has been used in the MATLAB environment. IPOPT
PT = 2NPCC
solver gives the global optimum solution efficiently and is
independent of the initial solution for this LP problem.
where NPCC is the total number of PCCs and PT is the total number
of possible topologies. Some of the most important topologies of
this system are given in Table 4. In the first three cases, only one 5 Results and discussion
MG gets integrated to the DS, whereas for cases of Sl. No. 4–6, In this section, the optimum settings considering all the topologies
two MGs, and in the rest of the cases, all the three MGs are added mentioned in the previous section are presented. Also, protection
to the DS. For cases Sl. No. 7-8 and 11-12, all the three MGs are coordination assessment is performed for different types of fault
integrated with DS and operated as NMG. The closing and opening under all possible topologies of NMGs.
status of each PCC is decided by the DSO from time-to-time based

IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828 2823
© The Institution of Engineering and Technology 2020
 17518695, 2020, 14, Downloaded from https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-gtd.2019.0557 by Cochrane Portugal, Wiley Online Library on [30/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Table 5 Optimum setting of DOCRs operation in the last case (case#12) of Table 4 (i.e. looped
Relays Forward direction Relays Reverse direction configuration) is given in Table 7. For the other cases, the same has
PS, A TMS, s PS, A TMS, s not been given for the sake of brevity. The primary-backup relay
R1 280 0.4855 — — —
pairs in all the configurations coordinate well for all possible fault
locations in the system.
R2 260 0.4311 — — —
From Table 6, it is observed that the sum of operating times of
R3 200 0.3991 — — — primary and backup relays are different in each case. It is because
R4 189 0.3371 — — — the number of primary-backup relay pairs and the magnitude of
R5 80 0.3639 — — — fault currents are different in each case. Further, it can be observed
R6 60 0.2913 — — — that the sum of the operating times of primary and backup relays
R7 50 0.2011 R7* 49 0.2348 for near-end faults are less as compared to those for far-end faults
in each case for the same number of pairs. As DOCRs follow
R8 80 0.1 — — —
inverse TCC and for far-end faults, the magnitude of currents are
R9 95 0.1781 R9* 29 0.6382 lower as compared to the corresponding near-end faults. The
R10 80 0.1 — — — operating times of all the relays (primary and backup) for near-end
R11 80 0.1 — — — faults are relatively faster as compared to those for far-end faults.
R12 131 0.1195 R12* 15 0.5132 The detailed operating times of primary and backup relays for near-
R13 80 0.1 — — — end and far-end faults are given in Table 7 for case 12 of Table 6.
R14 60 0.1 — — — From Table 7, it can be observed that the CTI requirement of 0.2 s
is maintained between the operating times of primary and backup
R15 56 0.159 R15* 62 0.1
relays in each pair for both near-end and far-end faults.
R16 82 0.1 R16* 80 0.1
R17 69 0.1 R17* 30 0.1 5.3 Protection coordination assessment
R18 45 0.2949 R18* 62 0.1
R19 83 0.1573 R19* 50 0.1 5.3.1 Fault in distribution feeder: Fig. 7 shows the TCCs of
optimum settings of relays R6, R7, R15 and R19. The operating
R20 63 0.2478 R20* 40 0.1
times of these relays have been displayed for a bolted three-phase
—, No dual-setting; *Operation of the relay in reverse direction. fault occurred at bus 20 in the main feeder of the DS.
For a fault at bus 20 (Fig. 6), R7, R15 and R19 operate as
primary relay; whereas R6 operates as the backup relay for R7.
Table 6 Sum of operating times of primary-backup relay From Fig. 7a, it is observed that CTI between operating times of
pairs R7 and R6 is 0.48 s and thus minimum CTI requirement is
Cases Near-end fault Far-end fault achieved. The fault current passing through the relay pair R7–R6
No. of ∑ top, ∑ tob, No. of ∑ top, ∑ tob, may increase or decrease because of change in the networked
relay pairs s s relay pairs s s topology. This will shift operating points of the relays on TCC
plots either left or right depending on the magnitude of current and
1 25 13.5794 24.3655 21 13.8682 27.6997
the CTI will vary. However, the minimum CTI requirement of 0.2 s
2 17 13.5219 16.5999 13 13.0823 16.8474 will always be maintained because the maximum possible value of
3 25 13.3250 24.7029 21 13.4072 27.1923 fault passing through the relay pair is 1600 A and the
4 26 13.7256 22.7328 22 13.9993 24.8254 corresponding CTI is more than the required value. On the other
5 26 13.4199 23.6095 24 14.1819 26.6857 hand, R15 operates as the main relay and stops supplying fault in
6 25 13.3250 24.7029 21 13.4072 27.1923 0.375 s, if PCC1 fails to island MG1. Similarly, R19 operates as the
main relay, if PCC5 fails to island MG2 after 0.395 s. There is no
7 28 13.3640 26.3686 25 13.9062 29.3869
coordination issue with R15 and R19 because these operate as the
8 27 13.4143 37.5442 24 14.2485 65.6958 main relay only and their concerned PCCs work independently for
9 27 13.6020 23.4948 25 14.3196 26.4609 any fault in the distribution feeder.
10 33 13.4165 29.8579 32 14.7180 34.0168 Table 8 gives the operating times of the other types of fault such
11 29 13.0578 27.0639 28 14.6158 30.9816 as, double-line-to-ground, single-line-to-ground and double-line
12 35 13.0575 36.9251 35 15.9698 47.5840 faults. From this table it can be observed that the operating times of
relay pair R7–R6 is much higher as compared to those mentioned
in Fig. 7. The time gap between the operating times of relay pair
5.1 Optimum settings R7–R6 for the other types of faults is more than that of three-phase
fault. Thus, protection coordination is easily maintained for the
The optimum coordination settings of single and dual-setting of other types of faults. It can be observed that the current passing
DOCRs for the system is given in Table 5. These settings are able through relay pair R7–R6 are less for the other types of faults (as
to provide proper protection coordination among the relays for all compared to the three-phase fault) and consequently the operating
types of near-end and far-end faults. times of the relays shift towards the left at their TCC plots (Fig. 7)
From Table 5, it is observed that the TMS value of single- and thus the CTI requirement is always maintained.
setting DOCR, which works only as primary relay (such as R8,
R10, R11, R13 and R14), is 0.1. Also, the TMS value of dual- 5.3.2 Fault in microgrid: Fig. 8 shows the TCCs of optimum
setting DOCR, which works only as primary relay (such as R15*– settings of relays R7, R9, R15, R16* and R19*. The operating
R20*), is 0.1. The lowest value of TMS for these relays are times of these relays have been displayed for a bolted three-phase
desirable because they have to operate as quickly as possible to fault in MG 2 at bus 80. Fig. 8a shows TCCs for relay pairs R19*–
clear the fault in their protection zones. The other DOCRs work as R7 and R19*–R15, while Fig. 8b shows TCCs for relay pair R16*–
primary and backup relay depending on fault location may not R9.
have the lowest value of TMS because they have to maintain From Fig. 8a it is observed that the minimum CTI requirement
coordination among their primary-backup relay pairs. for relay pairs R19*–R7 and R19*–R15 are maintained. Similarly,
from Fig. 8b it is observed that the minimum CTI requirement for
5.2 Operating times of relays in different topologies relay pair R16*–R9 is maintained for bolted three-phase fault in
MG 2 at bus 80. With the change in network topologies, operating
With these optimum settings, the sum of operating times of the
points of the relays may shift towards the right, particularly for the
primary-backup relay pairs in the 12 cases, discussed in Table 4,
three-phase faults in MG2, however, coordination will be
are given in Table 6 for both near-end and far-end faults. The
operating times of all the relays in the primary and backup mode of
2824 IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828
© The Institution of Engineering and Technology 2020
 17518695, 2020, 14, Downloaded from https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-gtd.2019.0557 by Cochrane Portugal, Wiley Online Library on [30/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Table 7 Operating times of primary and backup relays in looped configuration (case 12)
Sl. No. Primary Backup Operating times of relay pairs, s
relay relay For near-end fault For far-end fault
top, s tob, s top, s tob, s
1 R2 R1 1.2572 1.5871 1.2529 1.5881
2 R2 R9* 1.2572 1.5670 1.2529 1.5695
3 R3 R2 1.0446 1.2529 1.1941 1.4599
4 R4 R3 0.9857 1.1941 1.0012 1.2040
5 R5 R4 0.7240 1.0012 0.8047 1.1458
6 R5 R12* 0.7240 1.0036 0.8047 1.1506
7 R6 R5 0.5836 0.8047 0.6294 0.8657
8 R7 R6 0.4061 0.6294 1.6091 2.7581
9 R8 R1 0.1840 1.5871 0.2029 1.8511
10 R8 R7* 0.1840 1.8988 0.2029 3.3020
11 R8 R9* 0.1840 1.5670 0.2029 1.8369
12 R8 R12* 0.1840 1.1663 0.2029 1.3735
13 R9 R1 0.3614 1.5871 0.4735 1.8498
14 R9 R7* 0.3614 1.8988 0.4735 0.7461
15 R9 R12* 0.3614 1.1663 0.4735 1.7434
16 R10 R2 0.1840 1.2529 0.2073 1.5030
17 R10 R7* 0.1840 1.8797 0.2073 3.7992
18 R10 R12* 0.1840 1.1650 0.2073 1.3974
19 R11 R3 0.1902 1.1941 0.1967 1.2476
20 R11 R7* 0.1902 0.8012 0.1967 0.8473
21 R11 R12* 0.1902 1.0100 0.1967 1.0480
22 R12 R4 0.2837 1.0012 0.6597 2.1864
23 R12 R7* 0.2837 0.7872 0.6597 3.0725
24 R13 R5 0.1980 0.8047 0.2020 0.8174
25 R13 R7* 0.1980 0.6088 0.2020 0.6205
26 R14 R6 0.1852 0.6294 0.1959 0.6667
27 R14 R7* 0.1852 0.5541 0.1959 0.5911
28 R15* R7 0.2453 1.4472 0.2540 1.4589
29 R15* R19 0.2453 0.4594 0.2540 0.4597
30 R16* R9 0.2502 0.4735 0.2529 0.4741
31 R17* R12 0.2122 0.4939 0.2207 0.4947
32 R18* R20 0.2071 0.7234 0.2417 0.7243
33 R19* R7 0.2210 0.6429 0.2218 0.6444
34 R19* R15 0.2210 0.4489 0.2218 0.4495
35 R20* R18 0.1964 0.6362 0.1946 0.6368

For other types of faults such as double-line-to-ground, single-

line-to-ground and double-line faults, operating points of the relays
will shift towards the left on TCCs and coordination among the
pairs will always be maintained. Table 9 gives the operating times
of relay pairs R19*–R7, R19*–R15 and R16*–R9. The higher
values of operating times of all the relays in these pairs indicate
that the operating points have shifted towards the left on their
respective TCC curves of Fig. 8. These will ultimately increase the
time gap between the two TCC curves and results in higher CTI
values. Thus, protection coordination is easily maintained for
double-line-to-ground, single-line-to-ground and double-line faults.

5.3.3 High impedance faults: The single line-to-ground fault

with an impedance of 20 Ω has been considered to check the
performance of optimal settings for high impedance faults [39].
Two fault locations one on the distribution feeder and the other one
inside an MG have been considered.
Fig. 9 shows the TCC curves of all the concerned relays to clear
the fault at bus 20. Here, R7, R15 and R19 are responsible to clear
Fig. 7 Characteristics of relays and the coordination status of relay pair the fault. Relay R7 is backed up by relay R6. Similarly, Fig. 10
for bolted three-phase fault at bus 20 shows the TCC curves of all the concerned relays to clear the fault
(a) Characteristic plots of relay pairR7–R6, (b) Characteristic plot of R15, (c) at bus 80 (inside MG 2). Here, relay pairs R19*–R7, R19*–R15
Characteristic plot of R19 and R16*–R9 are responsible to clear the fault.
From Fig. 9 it can be observed that the operating points of
maintained until the maximum possible fault current, which is relays (R6, R7, R15 and R19) have higher operating times as
1600 A. compared to those on Fig. 7. These clearly indicate that the lower

IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828 2825
© The Institution of Engineering and Technology 2020
 17518695, 2020, 14, Downloaded from https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-gtd.2019.0557 by Cochrane Portugal, Wiley Online Library on [30/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Table 8 Operating times of relays with the obtained settings
for different types faults on feeder at Bus 20 in looped
configuration (case 12)
Fault Fault currents, A Operating times, s
type
Pr Br Pr Br
TLG R7: 290 R6: 290 R7: 0.787 R6: 1.274
R15: 885 — R15: 0.392 —
R19: 1158 — R19: 0.407 —
DLG R7: 260 R6: 260 R7: 0.840 R6: 1.370
R15: 746 — R15: 0.419 —
R19: 1051 — R19: 0.423 —
LL R7: 150 R6: 150 R7:1.267 R6: 2.205
R15: 420 — R15: 0.541 —
R19: 609 — R19: 0.512 —
LG R7: 210 R6: 210 R7: 0.967 R6: 1.607
R15: 487 — R15: 0.504 —
R19: 912 — R19: 0.449 —
Pr, primary relay; Br: backup relay; TLG: three-line-to-ground; LG, Fig. 9 Characteristics of relays and the coordination status of relay pair
line-to-ground; DLG: double-line-to-ground; LL, line-to-line for line-ground fault with impedance 20 Ω on feeder at bus 20
(a) Characteristic plots of relay pair R7–R6, (b) Characteristic plot of R15, (c)
Characteristic plot of R19

values of fault currents with high impedance single-line-to-ground

fault basically shift the operating points of various relays towards
left on TCC curves of Fig. 7 as compared to bolted three-phase
fault. The higher values of operating times of relays results in
higher time gap between the operating times of relays pair R7–R6,
which is 1.62 s (against 0.48 s for bolted three-phase fault). Similar
observations can be made from Fig. 10. The CTI between relay
pairs R19*–R7, R19*–R15 and R16*–R9 in case of Fig. 10 are
more as compared to those on 8. In other words, operating points
of these relays shifted towards left on their respective TCC curves.
Therefore, protection coordination is always maintained with the
obtained settings even for higher impedance faults.
With an increase in fault impedance, the operating times of
relays increases exponentially as long as the fault current is > 1.25
times the maximum load current. It is to be noted that the presented
method will work well as long as the fault current is > 1.25 times
maximum load current that can pass through a DOCR. Even for
Fig. 8 Characteristics of relays and the coordination status of relay pair high impedance faults, the presented scheme will work because the
for bolted three-phase fault in MG 2 at bus 80 fault currents for such a case are expected to be > 1.25 times the
(a) Characteristic plots of relay pairs R19*–R7 and R19*–R15, (b) Characteristic plots maximum load current. If for a case fault current is < 1.25 times,
of relay pair R16*–R9 the maximum load current of the relay, the overcurrent protection
would not operate. For such a scenario, adaptively changing the
Table 9 Operating times of relays with the obtained settings settings of the relays is a possibility but would require very reliable
for different types faults in MG 2 at Bus 80 in looped communication link and several other hardwares associated with
configuration (case 12) the scheme [40, 41].
Fault Fault currents, A Operating times, s
type 5.3.4 Faults with less number of DERs running: The
Pr Br1 Br2 Pr Br1 Br2 performance of obtained optimum settings with the change in
TLG R19*: R7: 424 R15: R19*: R7: 0.644 R15: power generation from DERs has been analysed for looped
1067 630 0.223 0.449 configuration (case 12 of Table 4). For this study, the system is
R16*: R9: 1233 — R16: R9: 0.474 — considered to be operating under lightly loaded condition (with
1183 0.253 40% loading) during a night hour and only WTs inside MGs are
DLG R19*: 896 R7: 374 R15: R19*: R7: 0.687 R15: actively supplying power. In another case, it is considered that
523 0.237 0.487 during a day hour, only solar PV panels are contributing power to
the MGs when 60% loading is there in the system. In these two
R16*: R9: 1121 — R16: R9: 0.493 —
cases, different types of faults are assumed to occur at buses 20 and
1071 0.263
80. The values of fault currents (without impedance) passing
LL R19*: 503 R7: 215 R15: R19*: R7: 0.951 R15: through various relays and the corresponding operating times of
293 0.296 0.662 relays are given in Tables 10 and 11.
R16*: 620 R9: 250 — R16: R9: 1.276 — From Tables 10 and 11, it is observed that the operating times of
0.335 relays is relatively higher when the fault occur with less number of
LG R19*: 571 R7: 296 R15: R19*: R7: 0.778 R15: DER connected to the MGs as compared to those with all DER
317 0.280 0.631 connected as reported in Tables 8 and 9, respectively. This
R16*: 925 R9: 978 — R16: R9: 0.522 — indicates that there is an impact on the operating times of the relays
0.279 when less number of DERs are running. This impact increases the
operating times of all the relays because of reduction in the fault
current with less number of DERs. The protection coordination is
still maintained in such cases with the optimum setting obtained.

2826 IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828
© The Institution of Engineering and Technology 2020
 17518695, 2020, 14, Downloaded from https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-gtd.2019.0557 by Cochrane Portugal, Wiley Online Library on [30/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Table 11 Case 12: operating time of the relays with the
different DERs connected and faults at bus 80
Fault DER Fault currents, A Operating times, s
type active Pr Br1 Br2 Pr Br1 Br2
TLG PV R19*: R7: 383 R15: R19*: R7: R15:
749 358 0.252 0.677 0.589
R16*: R9: — R16: R9: —
1128 1176 0.258 0.483
WT R19*: R7: 380 R15: R19*: R7: R15:
725 338 0.255 0.680 0.564
R16*: R9: — R16: R9: —
1123 1171 0.258 0.484
DLG PV R19*: R7: 339 R15: R19*: R7: R15:
618 286 0.271 0.722 0.672
R16*: R9: — R16: R9: —
1023 1070 0.268 0.503
Fig. 10 Characteristics of relays and the coordination status of relay pair WT R19*: R7: 336 R15: R19*: R7: R15:
for line-ground fault with impedance 20 Ω in MG 2 at bus 80 597 268 0.275 0.725 0.670
(a) Characteristic plots of relay pairs R19*–R7 and R19*–R15, (b) Characteristic plot
R16*: R9: — R16: R9: —
of relay pair R16*–R9
1019 1066 0.268 0.503
LL PV R19*: R7: 194 R15: R19*: R7: R15:
Table 10 Case 12: operating time of the relays with the
344 157 0.356 1.024 1.069
different DERs connected and faults at bus 20
R16*: R9: 621 — R16: R9: —
Fault DER Fault currents, A Operating times, s
593 0.342 0.652
type active Pr Br Pr Br
WT R19*: R7: 193 R15: R19*: R7: R15:
TLG PV R7: 275 R6: 275 R7: 0.812 R6: 1.319 333 147 0.362 1.028 1.142
R15: 543 — R15: 0.479 — R16*: R9: 619 — R16: R9: —
R19: 985 — R19: 0.434 — 591 0.343 0.653
WT R7: 273 R6: 273 R7: 0.815 R6: 1.326 LG PV R19*: R7: 269 R15: R19*: R7: R15:
R15: 517 — R15: 0.490 — 360 136 0.348 0.823 1.243
R19: 969 — R19: 0.437 — R16*: R9: 939 — R16: R9: —
DLG PV R7: 243 R6: 243 R7: 0.876 R6: 1.438 888 0.284 0.532
R15: 447 — R15: 0.525 — WT R19*: R7: 261 R15: R19*: R7: R15:
R19: 899 — R19: 0.451 — 344 122 0.356 0.826 1.418
WT R7: 242 R6: 242 R7: 0.879 R6: 1.442 R16*: R9: 936 — R16: R9: —
885 0.284 0.533
R15: 425 — R15: 0.538 —
R19: 885 — R19: 0.454 —
LL PV R7: 140 R6: 140 R7: 1.353 R6: 2.386
6 Conclusions
R15: 249 — R15: 0.735 —
R19: 522 — R19: 0.588 — In the context of modern power systems, distribution feeders are
WT R7: 139 R6: 139 R7: 1.363 R6: 2.407
getting interconnected with DERs or multiple MG. This study
focussed on distribution feeder interconnected with MGs. The
R15: 236 — R15: 0.763 —
studies reported in this study showed that the existing overcurrent
R19: 514 — R19: 0.593 — protection approach could still be used on the current distribution
LG PV R7: 197 R6: 197 R7: 0.991 R6: 1.653 feeders without completely revamping the protection scheme
R15: 259 — R15: 0.716 — altogether, which is often difficult to do in a utility setting. It was
R19: 796 — R19: 0.476 — shown in the study that single and proposed dual-setting DOCRs
WT R7: 197 R6: 197 R7: 1.013 R6: 1.695 could be installed at selected locations and their settings optimised
to achieve proper coordination. This kind of optimisation of single
R15: 242 — R15: 0.749 —
and dual-setting DOCRs for NMG structure has not been reported
R19: 786 — R19: 0.479 — in earlier publications. The low voltage ride through fault
behaviour of the DERs was also taken into account while arriving
at the optimum settings.
Similarly, in other cases where a MG is grid-connected through As a future study, qualitative assessment of reliability and
two or more PCCs, the concerned relay can get sufficiently high resiliency improvement of the NMGs with the proposed single and
fault current to pick-up and remove the fault on either grid-side or dual-setting DOCRs is being carried out and would be reported in a
MG side (as mentioned in Tables 10 and 11). On the other hand, in future study.
such cases where a MG is grid-connected through a single PCC
and operating with a smaller number of DERs (only inverted-
connected sources). Now, if a fault occurs on the grid-side, all the 7 Acknowledgments
relays associated for protecting the grid can clear the fault from the This work was partially funded by the Department of Science and
grid-side, whereas, relay associated to the MG may not pick-up Technology (DST), Ministry of Science and Technology,
because of insufficient fault current through the relay. However, the Government of India, under the project no. DST/EE/2018174.
drop in voltage at the PCC because of the fault, the MG controller
will initiate an operation and self-island MG. Therefore, the 8 References
distribution grid and the islanded MG can work independently after
the faulted part gets removed [1] Li, Z., Shahidehpour, M., Aminifar, F., et al.: ‘Networked microgrids for
enhancing the power system resilience’, Proc. IEEE, 2017, 105, (7), pp.
Thus, protection coordination is maintained with the obtained 1289–1310
optimum settings of single and dual-setting DOCRs for all types of
faults and all possible network topologies in MNGs.

IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828 2827
© The Institution of Engineering and Technology 2020
 17518695, 2020, 14, Downloaded from https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-gtd.2019.0557 by Cochrane Portugal, Wiley Online Library on [30/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
[2] Shahidehpour, M., Li, Z., Bahramirad, S., et al.: ‘Networked microgrids: [22] Cintuglu, M.H., Ma, T., Mohammed, O.A.: ‘Protection of autonomous
exploring the possibilities of the IIT-Bronzeville grid’, IEEE Power Energy microgrids using agent-based distributed communication’, IEEE Trans.
Mag., 2017, 15, (4), pp. 63–71 Power Deliv., 2017, 32, (1), pp. 351–360
[3] Esfahani, M.M., Hariri, A., Mohammed, O.A.: ‘A multiagent-based game- [23] Najy, W.K.A., Zeineldin, H.H., Woon, W.L.: ‘Optimal protection coordination
theoretic and optimization approach for market operation of multimicrogrid for microgrids with grid-connected and islanded capability’, IEEE Trans. Ind.
systems’, IEEE Trans. Ind. Inf., 2019, 15, (1), pp. 280–292 Electron., 2013, 60, (4), pp. 1668–1677
[4] Xu, T., Wu, W., Zheng, W., et al.: ‘Fully distributed quasi-Newton multi-area [24] Sharaf, H.M., Zeineldin, H.H., El-Saadany, E.: ‘Protection coordination for
dynamic economic dispatch method for active distribution networks’, IEEE microgrids with grid-connected and islanded capabilities using
Trans. Power Syst., 2018, 33, (4), pp. 4253–4263 communication assisted dual setting directional overcurrent relays’, IEEE
[5] Jadhav, A.M., Patne, N.R.: ‘Priority-based energy scheduling in a smart Trans. Smart Grid, 2018, 9, (1), pp. 143–151
distributed network with multiple microgrids’, IEEE Trans. Ind. Inf., 2017, [25] OPTI Toolbox Version 2.16. June 2018. Available at
13, (6), pp. 3134–3143 www.inverseproblem.co.nz/OPTI/index.php
[6] Sandgani, M.R., Sirouspour, S.: ‘Priority-based microgrid energy [26] Samet, H., Azhdari, E., Ghanbari, T.: ‘Comprehensive study on different
management in a network environment’, IEEE Trans. Sustain. Enery, 2018, 9, possible operations of multiple grid connected microgrids’, IEEE Trans.
(2), pp. 980–990 Smart Grid, 2018, 9, (2), pp. 1434–1441
[7] Wang, Z., Chen, B., Wang, J., et al.: ‘Coordinated energy management of [27] ‘IEC 60255-151 Measuring relays and protection equipement – Part 151:
networked microgrids in distribution systems’, IEEE Trans. Smart Grid, Functional requirements for over/under current protection’, IEC Std 2009–
2015, 6, (1), pp. 45–53 2010
[8] Gao, H., Liu, J., Wang, L., et al.: ‘Decentralized energy management for [28] Brahma, S., Pragallapati, N., Nagpal, M.: ‘Protection of islanded microgrid
networked microgrids in future distribution systems’, IEEE Trans. Power fed by inverters’. IEEE PES General Meeting 2018, Portland, OR, USA,
Syst., 2018, 33, (4), pp. 3599–3610 August 2018, pp. 1–5
[9] Ma, W.J., Wang, J., Gupta, V., et al.: ‘Distributed energy management for [29] Ambati, B.B., Kanjiya, P., Khadkikar, V.: ‘A low component count series
networked microgrids using online ADMM with regret’, IEEE Trans. Smart voltage compensation scheme for DFIG WTs to enhance fault ride-through
Grid, 2018, 9, (2), pp. 847–856 capability’, IEEE Trans. Energy Conv., 2015, 30, (1), pp. 208–217
[10] Yuan, W., Wang, J., Qiu, F., et al.: ‘Robust optimization-based resilient [30] Lammert, G., Heß, T., Schmidt, M., et al.: ‘Dynamic grid support in low
distribution network planning against natural disasters’, IEEE Trans. Smart voltage grids fault ride-through and reactive power/voltage support during
Grid, 2016, 7, (6), pp. 2817–2826 grid disturbances’. Proc. IEEE Power Systems Computation Conf., Wroclaw,
[11] Hussain, A., Bui, V.H., Kim, H.M.: ‘Resilience-oriented optimal operation of Poland, August 2014, pp. 1–7
networked hybrid microgrids’, IEEE Trans. Smart Grid, 2019, 10, (1), pp. [31] Piya, P., Ebrahimi, M., Karimi-Ghartemani, M., et al.: ‘Fault ride-through
204–215 capability of voltage-controlled inverters’, IEEE Trans. Ind. Electron., 2018,
[12] Schneider, K.P., Tuffner, F.K., Elizondo, M.A., et al.: ‘Enabling resiliency 65, (10), pp. 7933–7943
operations across multiple microgrids with grid friendly appliance [32] Savier, J.S., Das, D.: ‘Impact of network reconfiguration on loss allocation of
controllers’, IEEE Trans. Smart Grid, 2018, 9, (5), pp. 4755–4764 radial distribution systems’, IEEE Trans. Power Deliv., 2007, 22, (4), pp.
[13] Alam, M.N., Chakrabarti, S., Ghosh, A.: ‘Networked microgrids: state-of-the- 2473–2480
art and future perspectives’, IEEE Trans. Ind. Inf., 2019, 15, (3), pp. 1238– [33] Baran, M.E., Wu, F.F.: ‘Optimal capacitor placement on radial distribution
1250 systems’, IEEE Trans. Power Deliv., 1989, 4, (1), pp. 725–734
[14] Naiem, A.F., Hegazy, Y., Abdelaziz, A.Y., et al.: ‘A classification technique [34] Janssen, A., Riet, M.V., Bozelie, J., et al.: ‘Fault current contribution from
for recloser-fuse coordination in distribution systems with distributed state of the art DG's and its limitation’. Proc. Int. Conf. Power Systems
generation’, IEEE Trans. Power Deliv., 2012, 27, (1), pp. 176–185 Transients 2011, Delft University of Technology, Delft, The Netherlands, 14–
[15] Hussain, B., Sharkh, S.M., Hussain, S., et al.: ‘An adaptive relaying scheme 17 June 2019, pp. 1–9
for fuse saving in distribution networks with distributed generation’, IEEE [35] Boutsika, T.N., Papathanassiou, S.A.: ‘Short-circuit calculations in networks
Trans. Power Deliv., 2013, 28, (2), pp. 669–677 with distributed generation’, Electr. Power Syst. Res., 2008, 78, (7), pp. 1181–
[16] Shah, P.H., Bhalja, B.R.: ‘New adaptive digital relaying scheme to tackle 1191
recloser-fuse miscoordination during distributed generation interconnections’, [36] Mathur, A., Das, B., Pant, V.: ‘Fault analysis of unbalanced radial and meshed
IET Gener. Transm. Distrib., 2014, 8, (4), pp. 682–688 distribution system with inverter based distributed generation (IBDG)’, Int. J.
[17] ‘IEEE Standard for Interconnection and Interoperability of Distributed Energy Electr. Power Energy Syst., 2017, 85, pp. 164–177
Resources with Associated Electric Power Systems Interfaces’, IEEE Std [37] Cheng, C.S., Shirmohammadi, D.: ‘A three-phase power flow method for
1547-2018 Std.2018 real-time distribution system analysis’, IEEE Trans. Power Syst., 1995, 10,
[18] Dewadasa, M., Ghosh, A., Ledwich, G., et al.: ‘Fault isolation in distributed (2), pp. 671–679
generation connected distribution networks’, IET Gener. Transm. Distrib., [38] Grainger, J., Stevenson, W.D.: ‘Power system analysis’ (McGraw-Hill Inc.,
2011, 5, (10), pp. 1053–1061 New York, NY, USA, 1994)
[19] Zeineldin, H.H., Sharaf, H.M., Ibrahim, D.K., et al.: ‘Optimal protection [39] Cui, Q., Li, S.: ‘A microgrid protection scheme with conventional relay
coordination for meshed distribution systems with DG using dual setting measurements’. IEEE PES General Meeting 2018, Portland, OR, USA,
directional over-current relays’, IEEE Trans. Smart Grid, 2015, 6, (1), pp. August 2018, pp. 1–5
115–123 [40] Alam, M.N.: ‘Adaptive protection coordination scheme using numerical
[20] Shih, M.Y., Conde, A., Leonowicz, Z., et al.: ‘An adaptive overcurrent directional overcurrent relays’, IEEE Trans. Ind. Inf., 2019, 15, (1), pp. 64–73
coordination scheme to improve relay sensitivity and overcome drawbacks [41] Ghalei Monfared Zanjani, M., Mazlumi, K., Kamwa, I.: ‘Application of μ
due to distributed generation in smart grids’, IEEE Trans. Ind. Appl., 2017, pmus for adaptive protection of overcurrent relays in microgrids’, IET Gener.
53, (6), pp. 5217–5228 Transm. Distrib., 2018, 12, (18), pp. 4061–4068
[21] Saleh, K.A., Zeineldin, H.H., Al-Hinai, A., et al.: ‘Optimal coordination of
directional overcurrent relays using a new time–current–voltage
characteristic’, IEEE Trans. Power Deliv., 2015, 30, (2), pp. 537–544

2828 IET Gener. Transm. Distrib., 2020, Vol. 14 Iss. 14, pp. 2818-2828
© The Institution of Engineering and Technology 2020