IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 53, NO.

6, NOVEMBER/DECEMBER 2017 5217

An Adaptive Overcurrent Coordination Scheme to

Improve Relay Sensitivity and Overcome Drawbacks
due to Distributed Generation in Smart Grids
Meng Yen Shih , Arturo Conde, Member, IEEE, Zbigniew Leonowicz, Senior Member, IEEE,
and Luigi Martirano , Senior Member, IEEE

Abstract—Distributed generation (DG) brought new challenges Despite the numerous advantages of having DGs installed
for protection engineers since standard relay settings of traditional in the network, there are also new challenges [9]–[14] and
system may no longer function properly under increasing presence negative impacts on the protective overcurrent relays (OCRs).
of DG. The extreme case is coordination loss between primary and
backup relays. The directional overcurrent relay (DOCR), which is These are mainly increase of short-circuit current during fault
the most implemented protective device in the electrical network, condition and the bidirectional load flow in radial lines which
also suffers performance degradation in the presence of DG. There- the elements of the network are not designed to operate under
fore, this paper proposes the mitigation of DG impact on DOCR these new conditions. Possible consequences to the protection
coordination employing adaptive protection scheme (APS) using system are false tripping, under/over reach of relays, and co-
differential evolution algorithm while improving overall sensitiv-
ity of relays. The impacts of DG prior and after the application ordination loss between primary and backup relays [9], [10],
of APS are presented based on interconnected 6 bus and IEEE 14 [15]–[22].
bus system. As a consequence, general sensitivity improvement and Several solutions have been proposed to mitigate the impact of
mitigation scheme is proposed. DG penetration on subtransmission and distribution networks,
Index Terms—Adaptive protection scheme (APS), directional such as the following:
overcurrent relay (DOCR) coordination, distributed generation 1) disconnection of DGs immediately after fault detection
(DG), differential evolution (DE) algorithm, smart grid. [23];
I. INTRODUCTION 2) limitation of installed DGs capacity [24]–[26];
3) modification of the protection system by installing more
ISTRIBUTED generation (DG) in the form of renewable
D energy sources has become one of the most discussed top-
ics nowadays. The scope to depart from traditional generation
breakers for sectionalization, reconfiguration of networks
or the use of distance relays, and/or directional OCRs
(DOCRs) [27]–[30];
plants for long-term economic and environmental benefits has 4) installation of fault current limiters (FCLs) to pre-
made a massive increase of interests in DG technologies. More- serve/restore the original relay settings [18], [31]–
over, DGs can contribute to important aspects such as: network [42];
reliability, line congestion relief, overall loss reduction, and gen- 5) fault ride through control strategy of inverter based DGs
eration cost reduction in smart grid [1]–[8]. [43];
6) fault current control by solid-state-switch-based field dis-
Manuscript received January 22, 2017; revised May 28, 2017; accepted June charge circuit for synchronous DGs [44];
14, 2017. Date of publication June 20, 2017; date of current version Novem- 7) adaptive protection schemes (APS) [20], [45]–[54].
ber 20, 2017. Paper 2017-PSPC-0107.R1, presented at the 2016 IEEE 16th
International Conference on Environment and Electrical Engineering, Florence,
Although these methods can adequately mitigate the negative
Italy, Jun. 6–8, and approved for publication in the IEEE TRANSACTIONS ON impacts of DGs penetration on performance of the protective re-
INDUSTRY APPLICATIONS by the Power System Protection Committee of the lays, they suffer several limitations as well. Disconnecting large
IEEE Industry Applications Society. (Corresponding author: Meng Yen Shih.)
M. Y. Shih is with the Division of Science and Engineering, Uni-
DGs immediately after fault detection may lead to severe volt-
versity of Quintana Roo, Quintana Roo 77019, Mexico (e-mail: sono- age sags as the contribution of reactive power from DGs will be
fafriend@gmail.com). cut off. Moreover, most faults are temporary, thus disconnecting
A. Conde is with the Faculty of Mechanical and Electrical Engineering,
Autonomous University of Nuevo Leon, Nuevo Leon 66451, Mexico (e-mail:
the DGs is not economically beneficial since the DGs will need
con_de@yahoo.com). to be reconnected to the network after the clearance of temporal
Z. Leonowicz is with the Faculty of Electrical Engineering, Wroclaw Uni- fault in order to profit from the renewable energy. Also, stability
versity of Science and Technology, Wroclaw 50-370, Poland (e-mail: zbig-
niew.leonowicz@pwr.edu.pl).
problem may occur if there were high penetrations of DGs in
L. Martirano is with the Department of Astronautics, Electrical, and the network.
Energetics, Sapienza University of Rome, Rome 00185, Italy (e-mail: Limiting the DGs capacity is a provisional solution, since
luigi.martirano@uniroma1.it).
Color versions of one or more of the figures in this paper are available online
renewable energy is cheap, it should be fully exploited to gain
at http://ieeexplore.ieee.org. more profit and also to avoid excess CO2 emission mostly gen-
Digital Object Identifier 10.1109/TIA.2017.2717880 erated from conventional power plants.
0093-9994 © 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 5218 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 53, NO. 6, NOVEMBER/DECEMBER 2017

Modifying the protection scheme by installing more breakers

for sectionalization, reconfiguration of networks, or change of
protection principles is costly, and also the use of numerous
protection principles in a certain area of the power system may
lead to more complicated protection coordination scenario and
difficult postevent analysis.
Installing FCLs to preserve/restore the original relay settings
are practical since this device has almost zero impedance in
steady-state operation mode and inserts high impedance in se-
ries with the system when a fault occurs to limit the fault magni-
tude to its previous magnitude before DG installation. But this
implies an advanced study of optimal impedance and location
of the FCLs. Moreover, the major drawback of broad imple-
mentation of FCLs is the additional and elevated cost which is
undesirable for both utility and DG owners.
Both the fault ride-through control strategy of inverter-based
DGs and fault current control by solid-state-switch-based field
discharge circuit for synchronous DGs are low-cost solution
compared to the previous ones. The first consists of a commu-
tation control strategy of the inverter switches in order to limit
the fault current contribution. The second consists of installing a
solid-state-switch-based field discharge circuit for synchronous Fig. 1. Smart grid: general schematic presentation.
DGs in order to drain the excess fault currents. However, both
are only partial solutions to the problem since the first solution
is only applicable to inverter-based DGs and the second only to
synchronous DGs. These shortcomings lead to another option,
the APS.
The APS proposed in this research [54] consists of automatic
online re-adjustment of relay settings so that the relays are best
attuned for different network operating conditions due to dis-
patch or natural condition. Such changes are variations of inputs
and outputs of generators and transmission lines that affect the
load flow and fault current distribution. The APS may require a
central host with powerful computer that is linked by communi-
cation channels to send/receive data to/from relays prior or after Fig. 2. Loss of coordination due to DG penetration.
disturbance. Integration of substation control and data acquisi-
tion (SCADA) and energy management system will be needed sources [1]–[8]. An illustration of smart grid scheme is presented
to effectively implement this scheme. Contemporary DOCRs in Fig. 1.
have memory capacity and can be remotely re-adjusted through A general definition follows the Electric Power Research In-
communication channels. Hence, the DOCRs, SCADA system stitute (EPRI) Smart Grid Resource Center.
along with an appropriate online optimization algorithm for co-
ordination of DOCRs, can potentially improve the degraded per- “A Smart Grid is one that incorporates information and communica-
tions technology into every aspect of electricity generation, delivery
formance of relays caused by DGs. Therefore, this method can and consumption in order to minimize environmental impact, en-
be very beneficial in the long-term view for modern smart grids. hance markets, improve reliability and service, and reduce costs and
Differential evolution algorithm (DE) was selected for the improve efficiency” [58].
coordination study since it has been reported to be very effi-
Therefore, the proposal will rely on these advanced informa-
cient in different areas [55], [56]. DE had outperformed genetic
tion and communication technologies to perform adaptive and
algorithm, particle swarm optimization, harmony search algo-
online coordination of DOCRs.
rithm, and seeker optimization algorithm in the coordination
study [57].
B. Impact of DG on Protective Overcurrent Relay
II. SMART GRID AND DG IMPACT ON PROTECTIVE RELAY 1) Coordination Loss: The loss of coordination is defined as
the violation of coordination time interval (CTI) constraint be-
A. Smart Grid tween the primary and backup relays [18], [20], [57]. Example:
Smart grid targets highly reliable, self-healing, self- for a given fault at point F in Fig. 2, the coordination pairs to
regulating, demand response, efficient, and cutting edge network be analyzed in this scenario are R7–R8 and R7–R9 [primary-
that allows integration of high penetration of renewable energy backup]. Due to the penetration of DG, all these relays sense an

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 SHIH et al.: ADAPTIVE OVERCURRENT COORDINATION SCHEME 5219

Fig. 3. Illustration: Loss of coordination due to DG penetration. Fig. 5. Illustration: Inappropriate relay delay operations due to intentional or
unintentional islanding (DG microgrid).

This can cause further problems such as load or source

tripping. Many industrial motors have under-voltage tripping
protection, so if the fault takes long time to clear then the volt-
age sag duration increases which may lead to a disconnection of
industrial loads. The same applies to some power sources (i.e.:
wind turbine generators), which disconnect from the network
after several seconds for small sag or immediately after big sag.
A graphical illustration of an inappropriate relay delay oper-
ation using inverse time relay characteristic curve for the coor-
dination pair R7–R8 is presented in Fig. 5. It can be clearly seen
that after entering island operation mode, the backup relay R8
Fig. 4. Inappropriate relay delay operations due to intentional or unintentional increases its tripping time due to the decrease of fault current,
islanding (DG microgrid). whereas the primary relay is barely affected because its tripping
time is already located at the horizontal asymptote curve. Hence,
increase of short-circuit current. For R7, this is not critical as it there will be an undesired backup tripping time if a fault occurs
is the primary relay. But for R8 and R9, their CTI with respect during island operation mode.
to R7 may not be fulfilled as when there was absence of DG.
Therefore, there is a loss of coordination between pairs R7–R8 III. PROTECTION COORDINATION SCHEME
and R7–R9.
A graphical illustration of coordination loss using inverse Detailed description of the APS and the formulation of coor-
time relay characteristic curve for the coordination pair R7– dination problem are presented in the following sections.
R8 is presented in Fig. 3. It can be clearly seen that after the
integration of DG, the backup relay R8 accelerates its tripping
time due to the increase of fault current, whereas the primary A. Description of Adaptive Protection Scheme
relay R7 is barely affected because its tripping time is already The APS for coordination of DOCRs including DGs is pre-
located at the horizontal asymptote curve. Hence, there is a sented in Fig. 6.
loss of coordination because CTI is no longer preserved for the The proposed idea [54] for mitigating the impacts of DGs on
coordination pair R7–R8. DOCR coordination is based on a centralized adaptive scheme.
2) Islanding Operation: The islanding operation is defined This protection scheme consists of a centralized processing
as the isolation of a certain part of a network from the main net- server which analyzes and optimizes the data obtained through
work due to dispatch or natural condition [53], [59]–[62]. For a SCADA system of the network that implements overcurrent
given fault at point F in Fig. 4, suppose that relay R10 success- protection principle. The SCADA system monitors the network
fully cleared the permanent fault by tripping the circuit breaker. condition and identifies the operational and topological changes
Then, the remaining circuit from bus 7 to 10 will form an island of the network. As soon as a change in the network is identified,
operation network (microgrid) fed by the DG (assuming that the latest breaker and network configuration and/or the status
the DG has sufficient capacity to maintain stable operation for of DGs are input into the centralized processing server. Based
the islanded network). Under new network operating condition, on the network status data, the server performs load flow, fault,
if a fault occurs at any point along the lines between buses 7 contingency, and sensitivity analysis. Then, it recalculates the
and 10, then both primary and backup relays will suffer signif- pickup current of relays and optimizes the DOCR coordination.
icant time delay in clearing the fault due to the relatively small The new settings are updated to the DOCRs via communication
fault current contribution by the DG. The relays can regain their network so that the DOCRs become best-tuned to the present
operation speed if they were re-adjusted/recoordinated for this network operating condition. A single cycle is then completed.
new network operation and topology For every change of new operating condition, the cycle is exe-
cuted again. The frequency will also be in function of wind and
Isc
DG
< Isc
system
. solar forecast since DGs are intermittent sources.

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 5220 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 53, NO. 6, NOVEMBER/DECEMBER 2017

need to be included in the objective function because it was

observed that the use of only tp a in the objective function for
coordination in larger meshed systems may converge at a result
where there may be higher backup time, higher CTI, and may
have violation of constraints. Therefore, tb b , ECTIL , and NV are
included in the objective function to further improve the results
while maintaining selectivity.
2) Primary and Backup Relay Constraints: To coordinate
the relays, there must be a time difference between the primary
and backup relay. This time difference is called CTI. In this way,
whenever the primary relay fails to extinct a fault, the backup
relay enters and tries to extinct the fault after a prespecified de-
lay. This delay is usually set between 0.2 and 0.5 s, and 0.3 s
is used in this paper. The coordination constraint for every co-
ordination pair is given as

CTI ≤ tb − tp (2)

where CTI is the predefined CTI, tp is the primary operation

time for near-end fault, and tb is the backup operation time for
far-end fault.
There is also a range for each relay setting where feasible
solutions can be found. This is given as

dialmin ≤ dial ≤ dialmax (3)

Ipickupmin ≤ Ipickup ≤ Ipickupmax (4)
Fig. 6. Adaptive protection scheme for DOCRs including DGs. where dial is setting within range [0.05–10] for most numerical
relays, and Ipickup is the relay pickup current that consists of
B. Formulation of Coordination Problem a temporal overload and security factor (k) that multiplies the
1) Objective Function: The purpose of formulating the co- maximum load current. The k value is normally set to be between
ordination of DOCRs as an optimization problem is to minimize 140% and 160% [63].
the primary and backup operation time of relays while main- The minimum and maximum values of dial and pickup set-
taining selectivity of relays’ operation. It is of great importance tings are both hardware limitations. The dial parameter is a
to establish appropriate objective function that evaluates the fit- family of curves of the same type, which moves up or down to
ness of the settings because this is the key to ensure optimum enable coordination among relays for certain tripping time [63],
solution using optimization algorithms. The fitness function is [64]. Minimum dial settings are often used to obtain faster relay
given as tripping time. But this must be analyzed as it may compromise
the selectivity or coordination of relays. On the other hand, the
   NCP  
NCP

pickup current settings apart from the hardware limitation of
NV a = 1 tp a b =1 tb b
fitness = + ∗h1 + ∗h2 upper and lower bounds must have a minimum bound limita-
NCP NCP NCP
tion which can tolerate common temporal overloading scenarios
 NCP  [63], [64]. Also, an upper bound is set, since as pickup setting

+ ECTIL ∗ h3 (1) increases, sensitivity decreases. And relays may delay too much
L =1 to trip or never trip if pickup setting is set too high when a
two-phase fault occurs.
where h1 , h2 , and h3 are factors that increase or decrease the
3) Relay Characteristic Curve: The time OCR functions are
influence of each subobjective function, NV is the number of
set according to the relay characteristic curve (inverse time
violation of coordination constraints, NCP is the number of
curve). IEEE standard C37.112-1996 [65] is followed in this
coordination pairs, tp a is the primary operation time of relay a
paper and is given as
for near-end fault, tb b is the backup operation time of relay b
⎡ ⎤
for far-end fault, and ECTIL is the CTI error of Lth coordination
A
pair. t =⎣ I p + B ⎦ ∗ dial (5)
tb b minimizes the backup operation time of relays, ECTIL sc3∅ max
I pickup − 1
minimizes the CTI to as close to 0.3 s as possible, NV mini-
mizes the number of violations to zero (avoid converging to a where t is the relay operation time, Isc3∅ max is the maximum
local minimum), and tp a , tb b , and NV are all scaled and di- three-phase short-circuit current, and A, B, p are constants of
vided by NCP to be able to sum together. These different values the IEEE standard.

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 SHIH et al.: ADAPTIVE OVERCURRENT COORDINATION SCHEME 5221

The standard curves are moderate inverse (MI), very inverse formed
     
(VI), and extremely inverse (EI). The IEEE VI curve is chosen   r  + f  r  + f  r 
p =  f X X X (9)
in this paper, but other curves from the IEC standard can be used 1 2 3

 
as well.   r  /p
p1 = f X 1
(10)
 
  r  /p
C. Sensitivity Analysis p2 = f X 2
(11)
 
The sensitivity analysis is the examination of whether the   r  /p
backup relay is sensitive enough to operate for minimum fault p3 =  f X 3
(12)
current located at the far end of its primary relay protection zone.
where f () is the function to be minimized. The trigonomet-
This is computed for every coordination pair and is given as
ric mutation rate г is found within the interval (0, 1) and the
trigonometric mutation scheme is presented as
Isc2∅ backup
sensitivity = (6)   
k ∗ Iloadmax i,G +1 = Xr 1 + Xr 2 + Xr 3 + (p2 − p1 ) ∗
V r − X
X r
1 2
3
where Isc2∅ backup is the current that the backup relay senses for the + (p3 − p2 ) ∗
minimum fault simulated at the far end of its primary relay pro-
tection zone, k is the temporal overload factor of the backup re- r − X
X  r + (p1 − p3 ) ∗ r − X
X r
2 3 3 1

lay, and Iloadmax is the maximum load current of the backup relay.
The sensitivity analysis is a very important matter in the if rand [0, 1] ≤ Γ (13)
coordination study. For coordination pairs whose backup relays i,G +1 = X
r + F r − X
r
V 1
X 2 3
else (14)
do not fulfill the requirement of sensitivity will lead to very long
operation time. i is the donor vector and F is a scalar number that is
where V
Hence, the sensitivity is to be used as a comparative reference
typically found in the interval [0.4,1]. The parameters г and F
for the sensitivity analysis. The sensitivity constraint is given as
are chosen to be 0.5 and 0.8, respectively, in this paper.
3) Binomial Crossover: The crossover operation is per-
sensitivity ≥ 1.5. (7) formed after creating the donor vector via mutation. This
operation enhances the diversity of the population by ex-
changing the components of donor vector with the tar-
D. Differential Evolution Algorithm get vector X  i,G to generate the trial vector U  i,G =
DE algorithm [55]–[57] is a population-based EA consisting [ u1,i,G , u2,i,G , u3,i,G , . . . , uD ,i,G ].
of natural selection of genes. In this algorithm, probabilistic dis- 4) Binomial crossover scheme: whenever a randomly gen-
tribution is not needed for the generation of offspring. Therefore, erated number between 0 and 1 is less than or equal to the
it needs less mathematical operations and execution time com- crossover rate Cr value for each of the D variables, binomial
pared to other EAs. Detailed formulation of DE can be found in crossover is performed. Under this circumstance, there will be
[55]–[57]. a nearly uniform distribution of number of parameters inher-
1) Initial Population: Initiate all parameter vector genes in ited from the donor vector. The binomial crossover scheme is
their feasible range of corresponding relay settings. The initial presented as
population matrix is presented as 
vj,i,G if (randi,j [0, 1] ≤ Cr or j = jrand )
uj,i,G = (15)
⎡ ⎤ xj,i,G otherwise
dial(1,1) . . . dial(1,NR) k(1,NR+1) . . . k(1,NR*2)
⎢ .. .. .. .. .. .. ⎥ where randi,j [0, 1] is a uniformly distributed random number.
P= ⎣ . . . . . . ⎦.
This random function is executed for each jth component of
dial(NP,1) . . . dial(NP,NR) k(NP,NR+1) . . . k(NP,NR*2) the ith parameter vector. Then, a randomly chosen index jrand ∈
(8)  i,G gets at least one
[1, 2, . . . , D] ensures that the trial vector U
The population size can be defined as (NP, D*NR), where NP 
component from the donor vector Vi,G . The crossover operation
represents number of parameter vectors, D number of control
parameter Cr is selected to be 0.5 in this paper.
variables, and NR number of relays.
5) Selection: The selection operation determines whether
2) Trigonometric Mutation: Three different vector numbers
the trial or the target vector get through to the following genera-
are randomly selected from the DE population for each tar-
tion, for example, at generation G + 1. The selection operation
get vector. Suppose that the selected population members are
 r ,G , X r ,G , X
 r ,G for the ith target vector X  i,G . The indices is presented as
X 1 2 3 ⎧
r1 , r2 , and r3 are generated only once for each mutant vec- ⎨U  i,G if f U  i,G ≤ f X  i,G
tor and are mutually exclusive integers randomly chosen from  i,G +1 =
X (16)
the range [1, NP], which are also different from the index i. ⎩X  i,G if f U  i,G > f X  i,G
According to (10)–(12), three weighting coefficients are

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 5222 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 53, NO. 6, NOVEMBER/DECEMBER 2017

where f (X) is the fitness of the target vector and f (U  ) is the

fitness of the trail vector. If a lower or equal value of fitness is
obtained from the new trial vector, then the target vector will be
replaced in the next generation; otherwise, the target vector is
kept in the population. By doing so, the population will never
deteriorate since it either gets better or remains the same in
fitness quality.
Finally, a pseudo-code of DE algorithm can be formulated as
follows:

BEGIN (Initial Population)

Generate random initial population of feasible Fig. 7. 6 bus interconnected system.
solutions.
Evaluate fitness of initial population per vector. analysis), the DG10 case (10 MW DG inserted on bus 6 with
WHILE termination criterion not satisfied Xd’ 0.5 used for comparison purpose only since no coordination
DO was carried out), and DG20 case (20 MW DG inserted on bus
FOR 1:NP (Mutation Operation) 6 with Xd’ 0.3 used for comparison purpose only, no coordina-
Randomly select 3 mutually exclusive target tion was carried out). For the DG10 and DG20 cases, we run
vectors. power flow and fault analysis including the penetration of DG
Generate donor vectors by mutation scheme. on bus 6 without contingency analysis and without performing
END FOR coordination. Then, the relay settings of base case are used to
FOR 1:NP (Crossover Operation) determine the new operation time of the relays (influenced by
Execute crossover scheme between each target DG penetration) in order to evaluate the performance of the
and donor vector to form trial vector. relays.
END FOR The same three cases are presented after employing APS. All
FOR 1:NP (Selection Operation) three cases are then coordinated including contingency analysis.
Compare fitness between each target and trial The 6 bus system is presented in Fig. 7.
vector.
If trial vector has better fitness than target vector B. Description of the IEEE 14 Bus System
Target vector is replaced by trial vector.
END FOR A large interconnected IEEE 14 bus system has been chosen
END WHILE when termination criterion is satisfied to study the overall impacts of DGs on relay coordination.
The system consists of 30 active phase relays and 45 coor-
END dination pairs. A DG farm of 30 MW is connected to every
bus.
First, DOCRs of the base case are coordinated with no DGs
IV. TEST SYSTEM connected. Then, the relay settings of base case are used to
determine the new operation time of the relays (influenced by
The fault currents have been calculated with remote bus
DG, connected to each bus) in order to evaluate the overall
breaker opened. The DE algorithm has been simulated with 200
degradation and sensitivity improvements of DOCRs.
individuals and its sensitive parameters of DE Cr, Γ, and F
Finally, sensitivity improvement on a 24-hour basis is pre-
have been selected to be 0.5, 0.5, and 0.8, respectively, in this
sented for the IEEE 14 bus system. The IEEE 14 bus system is
paper.
presented in Fig. 8.
A. Description of a 6 Bus System
V. RESULTS AND DISCUSSIONS
A small interconnected 6 bus system was chosen to study the
A. Impacts of DGs on Directional Overcurrent Relay
impacts of DGs on relay coordination. The impacts of DGs on
the relays are the same in radial and interconnected systems [9]– Coordination for a 6 Bus System
[12]. But as the system under study becomes more complex, the The CTI results and short-circuit currents for the cases before
impacts of DGs may not be clearly seen since the fault current and after the insertion of DGs are presented in Figs. 9 and 10.
contribution of the system may be several times greater than the The threshold line is a visual representation of the pre-
contributions of DGs. established CTI of 0.3 s. It is observed from Figs. 9(a) and
The 6 bus system consists of 10 active phase relays and 16 10(a) that the CTI of all 16 coordination pairs of the base case
coordination pairs. A DG farm of 10 and 20 MW is connected satisfy the constraint presented in (2). The coordination of dif-
on bus 6. ferent pairs of relays for the base case fulfills the expectation
The three cases before employing APS are presented: the base of good selectivity since the majority of CTI values are found
case (DOCRs are coordinated in this case including contingency between 0.3 and 0.5 s, but this is not true for the cases DG10

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 SHIH et al.: ADAPTIVE OVERCURRENT COORDINATION SCHEME 5223

Fig. 10. Tendencies of CTI and backup short-circuit currents of the three
Fig. 8. IEEE 14 bus interconnected system. cases for all 16 coordination pairs. (a) Tendencies of CTI Coordination pairs.
(b) Tendencies of Backup Short Circuit Currents Coordination pairs.

DG is located in-between the two relays. Here, the primary relay

will sense higher current magnitude than the base case while the
backup relay will sense no change of the current magnitude. The
latter effect can be observed from Fig. 10(b) for the pairs: 7, 8,
12, 14, and 15, where the short-circuit current did not increase.
The infeed effect is not a critical issue since the selectivity
is still maintained. The worst scenario is the coordination loss
which can be seen from Fig. 10(a) for the pairs that have CTI
values below the threshold. For the cases DG10 and DG20 the
pairs: 1, 2, 3, 4, 6, 10, 13, and 16 suffered coordination loss. The
cause of coordination loss effect due to the over reach of backup
relay is that the DG is located behind both relays. Hence, both
relays sense an increase of short-circuit current, but since the
primary relay is situated near the horizontal asymptotic region
of the operation characteristic curve, the operation time of the
primary relay is barely affected. On the contrary, the backup
relay is situated farther from the horizontal asymptotic region;
Fig. 9. Tendencies of CTI and primary short-circuit currents of the three
cases for all 16 coordination pairs. a) Tendencies of CTI Coordination pairs. b)
so the operation time of the backup relay is significantly affected.
Tendencies of Primary Short Circuit Currents Coordination pairs. For the pairs: 1, 2, 3, 4, 6, 10, 13, and 16 it can be seen from
Figs. 10(b) and 9(b) that there is a significant increase of both
the backup and primary short-circuit currents; hence, the (CTI)
and DG20. From Figs. 9(a) and 10(a), it is observed that sev- coordination is lost for those pairs.
eral coordination pairs for the cases DG10 and DG20 are found The number of violations and percentage of violations for the
below the threshold value which means that there are violations three cases are presented in Fig. 11. From Fig. 11, it is clearly
of constraints when DGs are inserted to the system. seen that as the capacity of the DG penetration increases, the
Figs. 9(b) and 10(b) are scaled on the vertical axis to have a percentage of number of violations increases as well.
clearer view of the changes of primary and backup short-circuit
currents, respectively. Also, they are plotted with the CTI results
B. Mitigating the Impacts of DGs on Directional Overcurrent
of the same scale on the horizontal axis to observe the infeed
and coordination loss effects. Relay Coordination Using Adaptive Protection Scheme for a 6
From Fig. 9(a), the infeed effect of DG penetration can be Bus System
observed. Whenever there is a significant increase of primary The averaged sensitivity and sensitivity improvement per-
short-circuit current, the CTI increases. This is due to the loca- centage for the three cases are presented in Fig. 12. From
tion of DG in the system. The coordination pairs that suffered Fig. 12, it is clearly seen that as the capacity of DG penetra-
infeed effect are pairs: 7, 8, 12, 14, and 15 as it can be observed tion increases, the average sensitivity of the coordination pairs
from Fig. 9(a). These pairs suffered infeed effect because the increases as well.

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 5224 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 53, NO. 6, NOVEMBER/DECEMBER 2017

Fig. 13. Tendencies of CTI of the three cases for all 16 coordination pairs
employing adaptive protection scheme.

Fig. 11. Number of violations before APS: Comparison among base case,
DG10 case, and DG20 case. (a) Total number of violations before APS.
(b) Total percentage number of violations before APS.

Fig. 14. Tendencies of two-phase short-circuit currents, pickup currents, and

sensitivities of the three cases for all 16 coordination pairs employing adaptive
protection scheme. (a) Tendencies of Isc-2-ph Coordination pairs. (b) Ten-
dencies of Pickup Current Coordination pairs. (c) Tendencies of Sensitivity
Fig. 12. Averaged sensitivity after APS: Comparison among base case, DG10 Coordination pairs.
case, and DG20 case. (a) Averaged sensitivity after APS. (b) Total percentage
of averaged sensitivity improvement after APS.
improvement of overall sensitivity as presented in Fig. 14(c)
which coincides with (6) and (7).
The CTI results for the three cases after the insertion of DGs The reason why cases DG10 and DG20 have greater CTI
using APS are presented in Fig. 13. The results of two-phase values compared to base case, as observed in Fig. 13, is because
short-circuit currents, pickup currents, and sensitivity are pre- of the combined effect of increased short-circuit current and
sented Fig. 14. decreased pickup current caused by DG penetration, as shown
It is observed from Fig. 13 that by employing the APS, mitiga- in Fig. 3 and (5). Both effects lead to shorter operation time;
tion of coordination loss due to penetration of DG is successfully hence, APS recoordinated the system again with increased dial
achieved since there is no violation of coordination constraints parameters to maintain coordination.
for both DG10 and DG20 cases.
When APS is employed for the mitigation of DG penetration
impacts, additional benefit can be obtained other than maintain- C. DG Impact and Mitigation on Directional Overcurrent
ing selectivity for all coordination pairs, namely the increase of Relay Coordination Using Adaptive Protection Scheme for the
IEEE 14 Bus System
sensitivity. From Fig. 14(a), it can be observed that the two-phase
short-circuit current increases as the capacity of DG increases. In this section, an evaluation of DOCRs on the IEEE 14 bus
Also from Fig. 14(b), it can be observed that the pickup current system including DG of 30 MW (on each bus) is presented. The
tends to decrease as the capacity of DG increases. The resulting causes and effects of DGs are explained in previous sections;
effect of the observations drawn from Fig. 14(a) and (b) is the therefore, the essence of this section is to show the overall view

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 SHIH et al.: ADAPTIVE OVERCURRENT COORDINATION SCHEME 5225

Fig. 17. 24-hour profile of the IEEE 14 bus system: (a) load profile; and
Fig. 15. Number of violations before APS: Comparison among base case and (b) sensitivity profile for comparison between fixed/conventional relay sensitiv-
DG30 cases on every bus. (a) Total number of violations before APS Buses. ity and APS sensitivity. (a) 24 Hour Load Profile Hours. (b) 24 Hour Sensitivity
(b) Total percentage number of violations before APS Buses. Profile Hours.

as DG moves away from Gen 1 and 2. The small improvement

should not be under-estimated since this effect will be magnified
as installed DG capacity will grow.

D. Evaluation of Directional Overcurrent Relay Sensitivity

Improvement With the Presence of DG Using Adaptive
Protection Scheme on a 24-Hour Basis
In this section, an evaluation of the APS on the IEEE 14 bus
system including DG on bus 13 is presented. The intention is
not to show advantage of implementation of APS in a system
with DG for mitigation of certain effects (this has been shown
clearly in previous sections), but to show the overall possible
improvement of sensitivity during the 24-hour period.
Fig. 16. Averaged sensitivity after APS: Comparison among base case and The 24-hour load profile is a real load profile from the Mex-
DG30 cases on every bus. (a) Averaged sensitivity after APS Buses. (b) Total ican National Interconnected System (SIN) demand CENACE
percentage of averaged sensitivity improvement after APS Buses. on the 3rd of April 2016 [66]. This 24-hour profile is applied to
the IEEE 14 bus system to approximate the real operation of the
of DG impacts before and after employing APS on the whole system. The 24-hour load profile of the IEEE 14 bus system is
system instead of one fixed location. presented in Fig. 17(a).
From Fig. 15, it can be seen that as DG is located farther from From Fig. 17(b), it can be observed that the fixed sensitivity
Gen 1 and 2, more violations appear. This is because the fault (conventional coordination approach) has a constant sensitivity
contribution of DG is far less than Gen 1 and 2; hence, there was throughout 24 h. On the other hand, the sensitivity of relays using
no coordination loss when DG is located on buses 1–9 that are APS increases as the load profile decreases, which yields much
relatively close to Gen 1 and 2. On the other hand, since contri- better relay sensitivity than using the conventional approach.
bution of fault currents from Gen 1 and 2 for far buses decreases Since the conventional coordination approach uses maximum
due to electrical distance (buses 10–14), the integration of DG load profile to coordinate the relays, the coordination will be
on those buses will degrade protective relay performance and maintained for the different load variations as long as the actual
lead to coordination loss. A clear comparison of DG impacts on load flow does not exceed the maximum load profile. But as
large or small system can be drawn by observing Figs. 15(b) and it can be observed from Fig. 17(b), the peak of load profile is
11(b). In Fig. 15(b), a smaller percentage of violations occurred rather short; hence, substantial overall sensitivity enhancement
compared to Fig. 11(b) since larger system is generally more may be achieved if APS is implemented. The use of maximum
robust. load profile as reference for the coordination of protective relays
The system has zero violation for all cases when DG is con- is a prudent approach since advanced communication and con-
nected on each bus when employing APS. Also, sensitivity im- trol schemes were not available decades ago. However, modern
provement can be observed. This is presented in Fig. 16(a) and technology permits the implementation of proposed APS which
(b) where one can see a slight increase of sensitivity percentage can potentially improve different aspects of relay performance.

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 5226 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 53, NO. 6, NOVEMBER/DECEMBER 2017

Hence, the APS is proposed and viewed as an important im- ACKNOWLEDGMENT

provement for the future smart grid protective schemes.
Dr. M. Y. Shih was a Visiting Scholar with the Wro-
claw University of Science and Technology and would like to
VI. CONCLUSION thank Consejo Nacional de Ciencia y Tecnologı́a for the Ph.D.
Integration of DG in the network surely added numerous oper- scholarship.
ational benefits but at the same time degraded the existing relay
performance. The degradation varies depending on the size and
REFERENCES
location of DG in the network. Likewise, the sensitivity im-
provement also depends on size and location of DG. Therefore, [1] M. E. El-Hawary, “The smart grid—State-of-the-art and future trends,”
Electr. Power Compon. Syst., vol. 42, nos. 3–4, pp. 239–250, 2014.
exploiting the advanced features of relays and with the help of [2] Y. Zhang, N. Rahbari-Asr, J. Duan, and M. Y. Chow, “Day-ahead smart
SCADA system, the APS using DE algorithm is proposed for grid cooperative distributed energy scheduling with renewable and storage
mitigation of DG impacts on electrical networks. Besides, the integration,” IEEE Trans. Sustain. Energy, vol. 7, no. 4, pp. 1739–1748,
Oct. 2016.
proposed APS adequately manages both impacts of steady-state [3] A. Gholian, H. Mohsenian-Rad, and Y. Hua, “Optimal industrial
load current and fault-state short-circuit current contributed by load control in smart grid,” IEEE Trans. Smart Grid, vol. 7, no. 5,
DGs and improves the overall sensitivity of relays. pp. 2305–2316, Sep. 2016.
[4] X. Su, M. A. S. Masoum, and P. J. Wolfs, “Multi-objective hierarchical
The proposed scheme offers multiple advantages such as: au- control of unbalanced distribution networks to accommodate more renew-
tomatic online coordination, fulfillment of selectivity require- able connections in the smart grid era,” IEEE Trans. Power Syst., vol. 31,
ment, and overall sensitivity improvement, which at the same no. 5, pp. 3924–3936, Sep. 2016.
[5] Y. Gong, Y. Cai, Y. Guo, and Y. Fang, “A privacy-preserving scheme for
time increase the probability of high impedance fault detection. incentive-based demand response in the smart grid,” IEEE Trans. Smart
The scheme is robust for future system operational and topolog- Grid, vol. 7, no. 3, pp. 1304–1313, May 2016.
ical changes. [6] F. Kamyab, M. Amini, S. Sheykhha, M. Hasanpour, and M. M. Jalali,
“Demand response program in smart grid using supply function bidding
mechanism,” IEEE Trans. Smart Grid, vol. 7, no. 3, pp. 1277–1284, May
APPENDIX 2016.
[7] N. Rahbari-Asr, Y. Zhang, and M. Y. Chow, “Consensus-based distributed
Numerical comparison tables between “conventional protec- scheduling for cooperative operation of distributed energy resources and
tion scheme (CPS)” and “APS” are presented in this section to storage devices in smart grids,” IET Gener., Transm. Distrib., vol. 10,
no. 5, pp. 1268–1277, 2016.
complement Section V. [8] N. Rahbari-Asr, Y. Zhang, and M. Y. Chow, “Reactive power optimisation
considering wind farms based on an optimal scenario method,” IET Gener.,
A. Numerical Comparison Between Conventional Protection Transm. Distrib., vol. 10, no. 15, pp. 3736–3744, 2016.
[9] P. T. Manditereza and R. Bansal, “Renewable distributed generation: The
Scheme and Adaptive Protection Scheme for the 6 Bus System hidden challenges—A review from the protection perspective,” Renew.
Sustain. Energy Rev., vol. 58, pp. 1457–1465, May 2016.
[10] M. M. Eissa, “Protection techniques with renewable resources and
6 Bus System (DG on Bus 6) smart grids—A survey,” Renew. Sustain. Energy Rev., vol. 52,
pp. 1645–1667, Dec. 2015.
Base DG10 DG20 [11] I. Colak, S. Sagiroglu, G. Fulli, M. Yesibudak, and C. F. Covrig, “A survey
on the critical issues in smart grid technologies,” Renew. Sustain. Energy
NV 0 7 8 CPS Rev., vol. 54, pp. 396–405, Feb. 2016.
%NV – 44% 50% [12] X. Liang, “Emerging power quality challenges due to integration of renew-
Sen 2.87 2.87 2.87 able energy sources,” IEEE Trans. Ind. Appl., vol. 53, no. 2, pp. 855–866,
%Sen – – – Mar./Apr. 2017.
NV 0 0 0 APS [13] S. Montoya-Bueno, J. I. Munoz-Hernandez, and J. Contreras, “Uncer-
%NV – 0% 0% tainty management of renewable distributed generation,” J. Clean. Prod.,
Sen 2.87 3.35 3.69 vol. 138, pp. 103–118, Dec. 2016.
%Sen – 17% 29% [14] S. S. Kaddah, M. M. El-Saadawi, and D. M. El-Hassanin, “Influence
of distributed generation on distribution networks during faults,” Elect.
Power Compon. Syst., vol. 43, no. 16, pp. 1781–1792, 2015.
[15] M. A. Haj-ahmed and M. S. Illindala, “The influence of inverter-based DGs
B. Numerical Comparison Between Conventional Protection and their controllers on distribution network protection,” IEEE Trans. Ind.
Appl., vol. 50, no. 4, pp. 2928–2937, Jul./Aug. 2014.
Scheme and Adaptive Protection Scheme for the IEEE 14 Bus [16] A. Kamel, M. A. Alaam, A. M. Azmy, and A. Y. Abdelaziz, “Protec-
System tion coordination for distribution systems in presence of distributed gen-
erators,” Elect. Power Compon. Syst., vol. 41, no. 15, pp. 1555–1566,
2013.
IEEE 14 Bus System (DG on Bus 12) [17] H. Zayandehroodi, A. Mohamed, H. Shareef, and M. Mohammadjafari,
“Impact of distributed generations on power system protection perfor-
Base DG30 mance,” Int. J. Phys. Sci., vol. 6, no. 16, pp. 3873–3881, Aug. 2011.
[18] W. El-Khattam and T. S. Sidhu, “Restoration of directional overcurrent
NV 0 4 CPS relay coordination in distributed generation systems utilizing fault current
%NV – 9% limiter,” IEEE Trans. Power Del., vol. 23, no. 2, pp. 576–585, Apr. 2008.
Sen 3.4 3.4 [19] J. C. Gómez and M. M. Morcos, “Coordination of voltage sag and over-
%Sen – – current protection in DG system,” IEEE Trans. Power Del., vol. 20, no. 1,
NV 0 0 APS pp. 214–218, Jan. 2005.
%NV – 0% [20] W. El-Khattam and T. S. Sidhu, “Resolving the impact of distributed
Sen 3.4 3.7 renewable generation on directional overcurrent relay coordination: A
%Sen – 10% case study,” IET Renew. Power Gener., vol. 3, no. 4, pp. 415–425,
2009.

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 SHIH et al.: ADAPTIVE OVERCURRENT COORDINATION SCHEME 5227

[21] C. J. Mozina, “A tutorial on the impact of distributed generation (DG) on [41] H. Yamaguchi and T. Kataoka, “Current limiting characteristics of trans-
distribution systems,” in Proc. Protective Relay Eng. Annu. Conf., 2008, former type superconducting fault current limiter with shunt impedance
pp. 591–609. and inductive load,” IEEE Trans. Power Del., vol. 23, no. 4, pp. 2545–2554,
[22] R. A. Walling, R. Saint, R. C. Dugan, J. Burke, and L. A. Kojovic, “Sum- Oct. 2008.
mary of distributed resources impact on power delivery system,” IEEE [42] Y. Zhang and R. A. Dougal, “Novel dual-FCL connection for adding
Trans. Power Del., vol. 23, no. 3, pp. 1636–1644, Jul. 2008. distributed generation to a power distribution utility,” IEEE Trans. Appl.
[23] S. Conti, “Analysis of distribution network protection issues in presence Supercond., vol. 21, no. 3, pp. 2179–2183, Jun. 2011.
of dispersed generation,” Elect. Power Syst. Res., vol. 79, pp. 49–56, [43] E. Ebrahimi, M. J. Sanjari, and G. B. Gharehpetian, “Control of three-
2009. phase inverter-based DG system during fault condition without chang-
[24] S. Chaitusaney and A. Yokoyama, “Prevention of reliability degradation ing protection coordination,” Int. J. Elect. Power Energy Syst., vol. 63,
from recloser—Fuse miscoordination due to distributed generation,” IEEE pp. 814–823, 2014.
Trans. Power Del., vol. 23, no. 4, pp. 2545–2554, Oct. 2008. [44] H. Yazdanpanahi, W. Xu, and Y. Wei Li, “A novel fault current con-
[25] J. Chen, R. Fan, X. Duan, and J. Cao, “Penetration level optimization for trol scheme to reduce synchronous DG’s impact on protection co-
DG considering reliable action of relay protection device constrains,” in ordination,” IEEE Trans. Power Del., vol. 29, no. 2, pp. 542–551,
Proc. Int. Conf. Sustain. Power Gen. Supply, 2009, pp. 1–5. Apr. 2014.
[26] T. K. Abdel-Galil et al., “Protection coordination planning with distributed [45] R. Sitharthan, M. Geethanjali, and T. K. S. Pandy, “Adaptive protection
generation,” Canmet Energy Technol. Centre, Varennes, QC, Canada, scheme for smart microgrid with electronically coupled distributed gen-
CETC-Varennes 2007-149 (TR), 2007. erations,” Alexandria Eng. J., vol. 55, no. 3, pp. 2539–2550, Sep. 2016.
[27] H. B. Funmilayo and K. L. Buyler-Purry, “An approach to mitigate the [46] A. Rahmati, M. A. Dimassi, R. Adhami, and D. Bumblauskas, “An over-
impact of distributed generation on the overcurrent protection scheme current protection relay based on local measurements,” IEEE Trans. Ind.
for radial feeders,” in Proc. IEEE Power Syst. Conf. Expo., 2009, Appl., vol. 51, no. 3, pp. 2081–2085, May/Jun. 2015.
pp. 1–11. [47] S. C. Hsieh, C. S. Chen, C. T. Tsai, C. T. Hsu, and C. H. Lin, “Adaptive
[28] F. A. Viawan, D. Karlsson, A. Sannino, and J. Daalde, “Protection scheme relay setting for distribution systems considering operation scenarios of
for meshed distribution systems with high penetration of distributed gen- wind generators,” IEEE Trans. Ind. Appl., vol. 50, no. 2, pp. 1356–1363,
eration,” in Proc. Power Syst. Conf. Adv. Metering, Protection Control, Mar./Apr. 2014.
Commun., Distrib. Resour., 2006, pp. 99–104. [48] S. M. Brahma and A. A. Girgis, “Development of adaptive protec-
[29] I. M. Chilvers, N. Jenkins, and P. A. Crossley, “The use of 11 kV dis- tion scheme for distribution systems with high penetration of dis-
tance protection to increase generation connected to the distribution net- tributed generation,” IEEE Trans. Power Del., vol. 19, no. 1, pp. 56–63,
work,” in Proc. Int. Conf. Develop. Power Syst. Protection, vol. 2, 2004, Jan. 2004.
pp. 551–554. [49] H. Wan, K. K. Li, and K. P. Wong, “An adaptive multiagent approach
[30] D. Uthitsunthorn and T. Kulworawanichpong, “Distance protection of a to protection relay coordination with distributed generators in indus-
renewable energy plant in electric power distribution systems,” in Proc. trial power distribution system,” IEEE Trans. Ind. Appl., vol. 46, no. 5,
Int. Conf. Power Syst. Technol., 2010, pp. 1–4. pp. 2118–2124, Sep./Oct. 2010.
[31] S. B. Naderi, M. Negnevitsky, A. Jalilian, M. Hagh, and K. Muttaqi, “Op- [50] J. Ma, X. Wang, Y. Zhang, Q. Yang, and A. G. Phadke, “A novel adaptive
timum resistive type fault current limiter: An efficient solution to achieve current protection scheme for distribution systems with distributed gen-
maximum fault ride-through capability of fixed speed wind turbines dur- eration,” Int. J. Elect. Power Energy Syst., vol. 43, no. 1, pp. 1460–1466,
ing symmetrical and asymmetrical grid faults,” IEEE Trans. Ind. Appl., 2012.
vol. 53, no. 1, pp. 538–548, Jan./Feb. 2017. [51] J. Ma, J. Li, and Z. Wang, “An adaptive distance protection scheme for
[32] A. Esmaeili, S. Esmaeili, and H. Hojabri, “Short-circuit level control distribution system with distributed generation,” in Proc. 5th Int. Conf.
through a multi-objective feeder reconfiguration using fault current lim- Crit. Infrastruct., 2010, pp. 1–4.
iters in the presence of distributed generations,” IET Gener., Transm. [52] M. Baren and I. El-Markabi, “Adaptive overcurrent protection for distri-
Distrib., vol. 10, no. 14, pp. 3458–3469, 2016. bution feeders with distributed generators,” in Proc. Power Syst. Conf.
[33] T. Ghanbari, E. Farjah, and N. Tashakor, “Thyristor based bridge- Expo., vol. 2, 2004, pp. 715–719.
type fault current limiter for fault current limiting capability enhance- [53] Y. Ates et al., “Adaptive protection scheme for a distribution sys-
ment,” IET Gener., Transm. Distrib., vol. 10, no. 9, pp. 2202–2215, tem considering grid-connected and islanded modes of operation,” En-
2016. ergies, Spec. Issue: Distrib. Renew. Gener., vol. 9, no. 5, pp. 1–18,
[34] W. Guo, L. Xiao, and S. Dai, “Fault current limiter-battery energy storage May 2016.
system for the doubly-fed induction generator: analysis and experimental [54] M. Y. Shih, A. Conde, Z. M. Leonowicz, and L. Martirano, “Mitigating
verification,” IET Gener., Transm. Distrib., vol. 10, no. 3, pp. 653–660, the impact of distributed generation on directional overcurrent relay co-
2016. ordination by adaptive protection scheme,” in Proc. 16th IEEE Int. Conf.
[35] Y. Kim, H. C. Jo, and S. K. Joo, “Analysis of Impacts of superconducting Environ. Elect. Eng., Jun. 2016, pp. 1–6.
fault current limiter (SFCL) placement on distributed generation (DG) [55] S. Das and P. N. Suganthan, “Differential evolution: A survey of the
expansion,” IEEE Trans. Appl. Supercond., vol. 26, no. 4, Jun. 2016, state of the art,” IEEE Trans. Evol. Comput., vol. 15, no. 1, pp. 4–31,
Art. no. 5602305. Feb. 2011.
[36] H. He et al., “Application of a SFCL for fault ride-through capabil- [56] S. Das, S. S. Mullick, and P. N. Suganthan, “Recent advances in differential
ity enhancement of DG in a microgrid system and relay protection co- evolution—An updated survey,” Swarm Evol. Comput., vol. 27, pp. 1–30,
ordination,” IEEE Trans. Appl. Supercond., vol. 26, no. 7, Oct. 2016, Apr. 2016.
Art. no. 5603608. [57] M. N. Alam, B. Das, and V. Pant, “A comparative study of metaheuristic
[37] A. Elmitwally, E. Gouda, and S. Eladawy, “Optimal allocation of fault optimization approaches for directional overcurrent relays coordination,”
current limiters for sustaining overcurrent relays coordination in a power Elect. Power Syst. Res., vol. 128, pp. 39–52, Nov. 2015.
system with distributed generation,” Alexandria Eng. J., vol. 54, no. 4, [58] Electric Power Research Institute (EPRI) Smart Grid Resource Center.
pp. 1077–1089, Dec. 2015. [Online]. Available: http://smartgrid.epri.com/
[38] S. M. Hamid Javadi, A. Mousavi, and M. Khederzadeh, “A novel approach [59] S. A. Saleh, E. Ozkop, and A. S. Aljankawey, “The development of a
to increase FCL application in preservation of over-current relays coor- coordinated anti-islanding protection for collector systems with multi-
dination in presence of asynchronous DGs,” Int. J. Elect. Power Energy ple distributed generation units,” IEEE Trans. Ind. Appl., vol. 52, no. 6,
Syst., vol. 44, pp. 810–815, 2013. pp. 4656–4667, Nov./Dec. 2016.
[39] S.-H. Lim, J.-S. Kim, M.-H. Kim, and J.-C. Kim, “Improvement of pro- [60] A. Tavakoli, M. Negnevitsky, and K. Muttaqi, “A decentralized model
tection coordination of protective devices through application of a SFCL predictive control for operation of multiple distributed generators in an
in a power distribution system with a dispersed generation,” IEEE Trans. islanded mode,” IEEE Trans. Ind. Appl., vol. 53, no. 2, pp. 1466–1475,
Appl. Supercond., vol. 22, no. 3, Jun. 2012, Art. no. 5601004. Mar./Apr. 2017.
[40] S.-H. Lim and J.-C. Kim, “Analysis on protection coordination of protec- [61] G. Marchesan, M. R. Muraro, G. Cardoso Jr., L. Mariotto, and C. D. L. de
tive devices with a SFCL due to the application location of a dispersed Silva, “Method for distributed generation anti-islanding protection based
generation in a power distribution system,” IEEE Trans. Appl. Supercond., on singular value decomposition and linear discrimination analysis,” Elect.
vol. 22, no. 3, Jun. 2012, Art. no. 5601104. Power Syst. Res., vol. 130, pp. 124–131, Jan. 2016.

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.
 5228 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 53, NO. 6, NOVEMBER/DECEMBER 2017

[62] J. Mitra, M. R. Vallem, and C. Singh, “Optimal deployment of distributed Zbigniew Leonowicz (M’03–SM’12) received the
generation using a reliability criterion,” IEEE Trans. Ind. Appl., vol. 52, M.Sc., Ph.D., and Dr. Sci. degrees in electrical en-
no. 3, pp. 1989–1997, May/Jun. 2016. gineering from Wroclaw University of Science and
[63] IEEE Guide for Protective Relay Applications to Distribution Lines, IEEE Technology, Wroclaw, Poland, and Habilitate Doctor-
Std. C37.230-2007, 2007. ate degree from Bialystok University of Technology,
[64] J. Lewis Blackburn and T. J. Domin, Protective Relaying, Principles and Bialystok, Poland in 1997, 2001, and 2012, respec-
Applications, 3rd ed. Boca Raton, FL, USA: CRC Press, 2006. tively.
[65] IEEE Standard Inverse-Time Characteristic Equations for Overcurrent Since 1997, he has been with the Department of
Relays, IEEE Std. C37.112-1996, 1996. Electrical Engineering, Wroclaw University of Sci-
[66] CENACE webpage. [Online]. Available: http://www.cenace.gob.mx/ ence and Technology, Wroclaw, Poland, where he
GraficaDemanda.aspx is currently an Associate Professor. His current re-
search interests include power quality, control and protection of power systems,
renewables, industrial ecology, and applications of advanced signal processing
methods in power systems.
Meng Yen Shih received the B.Eng. degree from
the Instituto Tecnológico de Chetumal, Chetumal,
Mexico, in 2010, and the M.Sc. and Ph.D. degrees
from the Universidad Autónoma de Nuevo León, San
Nicolás de los Garza, Mexico, in 2013 and 2016, re-
spectively, all in electrical engineering.
He is currently a Professor with the Universidad de
Quintana Roo, Quintana Roo, Mexico. His research
interests include power system protection, distributed
Luigi Martirano (M’02–SM’11) received the M.Sc.
generation, and evolutionary algorithms.
and Ph.D. degrees in electrical engineering from Uni-
versity of Rome, Italy, in 1998 and 2003, respectively.
Since 2000, he has been with the Department of
Arturo Conde (M’00) received the M.Sc. and Ph.D. Electrical Engineering, Sapienza University of Rome,
degrees in electric engineering from the Universidad Rome, Italy, where he is currently an Associate Pro-
Autónoma de Nuevo León, San Nicolás de los Garza, fessor. His current research interests include power
Mexico, in 1996 and 2002, respectively. systems design, planning, safety, lightings, home and
Currently, he is a Professor with the Gradu- building automation, and energy management.
ate Program of Electrical Engineering, Universidad Dr. Martirano is a Senior Member of the Industry
Autónoma de Nuevo León. His research interests in- Applications Society, a member of the Italian Electri-
clude adaptive protection of power systems, optimal cal Commission Technical Committees CT205 (Home and Building Electronic
energy management, and smart grid systems. Systems) and CT315 (Energy Efficiency), and a member of the European CEN-
Dr. Conde is a member of the National Research ELEC Joint Working Group CEN/CLC JWG9 “Energy measurement plan for
System of Mexico. organizations.”

Authorized licensed use limited to: Indian Institute of Technology (ISM) Dhanbad. Downloaded on February 14,2022 at 09:58:34 UTC from IEEE Xplore. Restrictions apply.