Electrical Power and Energy Systems 63 (2014) 140144

Contents lists available at ScienceDirect

Electrical Power and Energy Systems

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

Adaptive overcurrent relay coordination for off-peak loading in

interconnected power system
Chao-Rong Chen, Cheng-Hung Lee
Department of Electrical Engineering, National Taipei University of Technology, Taiwan, ROC

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

Article history: Due to the substation automation, the supervisory control and data acquisition (SCADA) system is
Received 29 June 2013 spreading that the protective relay settings can be modied through the network automatically. This
Received in revised form 4 May 2014 work presents an overcurrent relay coordination decision for switching the relay settings to adaptive
Accepted 27 May 2014
the off-peak loading in interconnected power system. Overcurrent relays (OCR) including directional
and non-directional protective relays are widely applied in power system as main protective devices
as they are reliable, cost effective and selective. Usually, the OCR protective coordination is coordinated
Keywords:
at the time of maximum load and normal power supply, which may not be adaptive, as generating and
Off-peak loading
SCADA
load power are decreased signicantly during the off-peak period of time compare to peak load that
Overcurrent relay insufcient currents caused cascade lagging to activate OCRs. An example of single-ring interconnected
Relay coordination power system was tested to demonstrate the coordination and conrm that the proposed approach
Adaptive protection method is practical during the off-peak time.
2014 Elsevier Ltd. All rights reserved.

Introduction
The inverse denite minimum time (IDMT) OCR is activated by
over currents through taps setting to get required characteristic
curves. Normally OCR coordination for protecting power systems
is to set the relay curves from the top backup relay to the next
main relay, assuming the maximum short circuit capacity of the
power system while the curve type, current tap and time dials
are chosen to provide adaptive protection. Consequently, approach
can be operated during maximum or normal load on the system in
case of fault. In other words, the OCR will be activated on the spe-
cic fault current is reached and the time elapsed.
In interconnected power systems, due to power consumption
change from peak to off-peak signicantly such as noon and
mid-night, hot summer and cold winter, the load and generating
power are changed, as well as the system capacity and short circuit
current are very different, specically as in Taiwan at work days,
summer peak load on year 2012 was 32,932 MW and winter off-
peak load was 16,173 MW in the case of system generation
decreased, the short circuit current will decrease that system fault
current too small to activate OCR for protection promptly, the
solution may either be uses differential relay or distance relay for
Corresponding author. Address: Department of Electrical Engineering, National
Taipei University of Technology, No. 1, Sec. 3, Jhong-Siao E. Rd., Taipei City 106,
Taiwan, ROC. Tel.: +886 2 2655 2666x1258; fax: +886 2 2655 2556.
E-mail address: t5319002@ntut.edu.tw (C.-H. Lee).
instead which are much more cost than OCR, not only relay itself
but also extra wiring and consideration, otherwise the OCR
protection may need to be adaptive and adjustable for the settings
on off-peak time. The conventional method is to get a compromise
of relay coordination setting plan to deal with peak throughout
off-peak loading and short circuit capacity to meet requirement
of grading margin from 0.2 s through 0.4 s in all loading conditions
without changing the settings which may not be adaptive during
off-peak period of time.
Recently, multifunction numerical communicable relays are
developed products whose design is based on microprocessor tech-
nology using Digital signal processor (DSP) as intelligent electronic
device (IED), SCADA system implements adaptive protection being
applicable by data link for telecontrol application. The SCADA
protocol between master and remote terminal unit (RTU) forms a
viable model for RTU to IED communications. There are several
protocols in use, the most popular are International Electrotechni-
cal Commission (IEC) 61850-6 and 60870-5 series, specically IEC
60870-5-101 and Distributed Network Protocol version 3 (DNP3).
IEDs advanced inter-station communication facilities can also be
applied for protection of ring-type and meshed distribution
networks [15]. A typical large SCADA system network is shown
on Fig. 1 [3]. The protocols as can also be applied to switch the
relay settings of IED. The proposed method is to optimize each
grading margin to as closer to 0.3 s as possible by performing the
adaptive protection in IED and SCADA system during peak and
off-peak conditions to get optimal protection.

http://dx.doi.org/10.1016/j.ijepes.2014.05.068
0142-0615/ 2014 Elsevier Ltd. All rights reserved.
 C.-R. Chen, C.-H. Lee / Electrical Power and Energy Systems 63 (2014) 140144
Fig. 1. A typical large SCADA network [3].
IDMT OCR can be adaptive by ne adjusting time and current
settings as polynomial equation [6]. For the currents in solid-state
or microprocessor-based relays are discrete and digital. The
pick-up current setting is discrete; the relay coordination problem
is therefore a mixed integer nonlinear programming problem
(MINLP) [79]. The MINLP is thus easily locally optimized, and
the genetic algorithm (GA) provides global optimal search to locate
optimal solutions [1012].
In the conventional method of relay coordination, the engineer
carefully checks the power system constraints for deciding the
current setting multiplier (CSM) and time setting multiplier
(TSM), those constraints including the maximum fault current,
the load current, the fault clearance time, the current transformer
(CT) ratios and others to yield adaptive curves. Many commercial
computer software packages do not support those checking.
Articial intelligence such as GA, Particle Swarm Optimization
(PSO) have been employed to solve relay coordination problem
in terms of a constrained objective function to optimize coordina-
tion [918], the GA has been applied in OCR coordination in
interconnected or ring-fed power distribution systems, and its
advantages have been demonstrated. However the difference
between peak and off-peak loads for the optimal relay coordina-
tion need to be studied furthermore.
A ring type interconnected power distribution system thus has
been tested for both peak and off-peak loading application of the
study to show the proposed method is applicable in Proposed
method of the article.
Problem formulation
The problem for adaptive relay coordination on peak and off-
peak loadings is to locate the relay CSM and TSM optimal settings
which can be coordinated on either peak or off-peak loads at same
power system with the loadings, that is coordination time interval
(CTI) for activating time of overcurrent relays in series with one
another can be 0.30 s as marginal CTI, when adopting an eight
cycle circuit breaker is used in a static relay, representing a
conservative average of protection, as recommended in IEEE Std
242-2001 [19]. In the radial or ring type systems, the coordination
criteria can be modied from Eq. (2) of [6] as the optimal coordina-
tion criteria:
141
T nk  T ik P CTI 1
where Tnk represents the operating time of the rst backup relay Rn
for fault zone k, and Tik is the operating time of relay Ri, CTI is the
coordination time interval, The CTI of the primary and backup relay
can be easily determined for each pair of relays for coordination.
The OCR curves are simulated in accordance with the equation
that was dened by IEC 60255-3 1989 and similar to that associ-
ated with the IEEE standard C37.112-1996 [20,21] as presented
in Eq. (2)
A  TM
tI 2
IP  1
where t(I) is the relay operating time based on a fault current I, TM
is the dial relay time setting, P is a constant, and I is fault current
during the operation of the relay.
A simple genetic algorithm is adapted to maximum the tness
or objective function of IDMT OCR. The relay equation Eq. (2) is
revised for use with various voltage levels, relay curves and load
types of industrial power systems according to Eq. (3) [22,23]
below:
td  k1
tI a 3
k2  tap  Ictn  1
where t(I) is the same as in Eq. (2), td is the setting of time dial (0.1:
0.05: 1.0), k1 is the relay type constant: for normal inverse (NI) it is
0.14, for very inverse (VI) it is 13.5, and for extreme inverse (EI) it is
80 respectively, tap is current tap of the relay (0.5: 0.1: 1.2), k2 is the
constant of transformer ratio, a is the OCR type constant for NI it is
0.02, for VI it is 1 and for EI it is 2, In is the fault current which
ranges from 2IR to 20IR but should not exceed maximum fault cur-
rent of the system, where IR is the primary rated current of each cur-
rent transformer (CT), and ct is the ratio of CT of each feeder.
The protective relay coordination problem is a linear, a non-
linear, or a mixed integer non-linear problem depending on the
variable of the pick-up setting I, which is a variable that mainly
affects the problem, if I is xed, then the coordination problem
becomes a linear problem. For continuously changing I values,
the problem becomes a nonlinear problem, if I is discrete, the prob-
lem becomes a mixed integer nonlinear Problem (MINLP) [6]. In
the genetic algorithm, the random generation and the initial
142 C.-R. Chen, C.-H. Lee / Electrical Power and Energy Systems 63 (2014) 140144

Abnormal Off-Peak Peak/Normal 6.5

I/O
Loading Loading Loadings
6

5.5

Load Detection and Manual 5

SCADA Verification Confirmation

Fitness Volume.
4.5

Abnormal Peak&Normal 4
Decision
Off-Peak 3.5

Alert- OCR Setting-1 OCR Setting-2 3

IED/OCR Abnormal For Off-Peak For Normal
2.5

Fig. 2. Flow chart of setting decision. 2

1.5
0 20 40 60 80 100 120
population increase the probability of convergence to the global Generation
optimal setting and overcome the limitation of the conventional
Fig. 4. GA simulation curve for OCR coordination.
rule-base methods in optimizing the relay settings.

Proposed method
Optimal relay coordination
The protective relay coordination problem is originally a MINLP,
since CSM is discrete with discrete TDS and some constraints are
nonlinear. Methods for solving MINLP are known potentially to
result in local optimal solution. To solve this problem, GA has been
used to solve the coordination problem. As stated in Introduction,
two constraints should be considered. One is that the CTI should
involve Tnk Tik of equal to or larger than 0.30 s, the other is that
the current setting CSM should have an appropriate weighting fac-
tor in objective function (OF) to yield applicable settings. The OF
presented elsewhere can be redened as Eq. (4) [24]:
" #
1X N X Imax
Objective function OF tn; I w  tap  1:02
N n1 II
min
where the constant N denotes the number of all concerned
RR t(n, I) for averaging the operating time of each relay pair and
total relay operation time; which includes the CB operating time
in the system conguration, w is the coefcient of the contribution
of current taps and should be as close to 1.0 as possible. In this case
a value of four is used, based on trail and error. Additionally, the
total clearance time of the relay operation at maximum fault cur-
rent shall not exceed one second if the momentary breaking capac-
ity of CB is one second rating.
The CSM and TSM were encoded as a chromosome for all OCR
and current transformer (CT) ranges. Fig. 1 of [24] shown the ow
chart of algorithm. Every OCR that is connected to the CT is used to
convert the load current to 5A for relay operating. The load current
in each feeder should be able to carry a full load current of
100120% rated current. The optional coordination is calculated
using Eqs. (1) and (3) from the top Tik to the next Tnk yielding the
required solutions. To evaluate the effectiveness of curve, every

4 point on the curve is dened in Eq. (4).

Fig. 3. Single line diagram ring type power distribution system.

Table 1
7-Bus ring OCR data and settings peak and normal loading.

Main relay Backup relay Primary SC current (A) Backup SC current (A) Grading margin (s) Relay CT ratio Tap setting Time dial (s)
R2 R1 21,672 27,817 0.3153 R1 1200/5 0.5 0.8
R3 R2 15,871 21,672 0.3599 R2 1200/5 0.95 0.8
R4 R3 14,014 15,871 0.3043 R3 1200/5 1.05 0.6
R5 R4 11,562 14,014 0.3981 R4 1200/5 1.0 0.45
R6 R5 9551 11,562 0.3135 R5 1200/5 1.15 0.3
R7 R6 5215 9551 0.3040 R6 1200/5 0.85 0.15
R7 1200/5 1.0 0.05
 C.-R. Chen, C.-H. Lee / Electrical Power and Energy Systems 63 (2014) 140144 143

Table 2
7-Bus system OCR data and settings off-peak loading.

Main relay Backup relay Primary SC current (A) Backup SC current (A) Grading margin (s) Relay CT ratio Tap setting Time dial (s)
R2 R1 13,003 13,909 0.3823 R1 1200/5 0.25 0.8
R3 R2 10,297 13,003 0.3057 R2 1200/5 0.63 0.8
R4 R3 11,211 10,297 0.3165 R3 1200/5 0.36 0.4
R5 R4 10,406 11,211 0.3860 R4 1200/5 0.84 0.45
R6 R5 9506 10,406 0.3167 R5 1200/5 0.72 0.25
R7 R6 5213 9506 0.3026 R6 1200/5 0.9 0.15
R7 1200/5 1.10 0.05

The optimal relay settings for both peak and off-peak loadings of fault during off-peak loading time of the power system, the
to be preset as two settings of CSM and TSM will be selected by grading margins are remain coordinated under those new CMS
SCADA system for adaptive protection. and TMS settings.

Decision for switching OCR setting Discussion

The decision for switching the OCR setting of peak and off-peak
scheme is as Fig. 2. The system loading status can be detected by
power ow (MW + j MVar) trending with considering clock time
to check if the loading is going to up the peak, normal or going
down to the off-peak loadings. The SCADA will compare the
loading status with preset load value to verify and make decision
to switch the OCR setting, manual conrmation can be added if
required for supervision.
Case study 1 peak and normal loading power system coordination
A 3ph. 69 kV, 7-bus interconnected power distribution system
and protection direction is shown in Fig. 3 [2427],
There are 7 directional OCRs (DOCRs) for relay coordination. The
name of each relay and circuit breaker along with the clockwise
protection on each bus are R1(BUS A 620), R2(BUS B 620),
R3(BUS C 620), R4(BUS D 610), R5(BUS E 610), R6(BUS F 620)
and R7(BUS G 620) respectively, the relay pairs, maximum three-
phase short-circuit current on each bus and generated relay data
by GA are as in Table 1, all CTs are in 1200/5 ratio, using the
proposed method for testing the result that CSM, TSM and grading
margins between each main and back-up relay pairs are
coordinated and satised to the optimal coordination criteria as
in Eq. (1), The GA parameters are as follows:
Crossover rate 1
Mutation rate 0.03
Generation 100
Size of population 50
The simulation curve as shown in Fig. 4.
Each CTI is closed to 0.30 s as shown in center column of the
Table 1.
OCR coordination is a highly professional and requires an engi-
neer to spend much time. The resulting grading accuracy cannot be
easily veried. The GA parameters for a mutation of 0.05 and a
crossover rate of 1.0 to improve the solution of the simulation in
Case study 1 peak and normal loading power system coordina-
tion and Case study 2 off-peak loading power system coordina-
tion both. Tables 1 and 2 presents the generated TSM and CSM
with considering the above constraint and conditions. The results
indicated that the generated settings are applicable in considering
peak and off-peak loading conditions to get adaptive OCR coordi-
nation. The grading margins of the closest points of the relay pairs
in the same protective zone are close to =0.30 s as in grading mar-
gin column in Tables 1 and 2.
It is applicable to use SCADA protocol for place instructions to
IED (OCR) for switching the CSM and TSM settings to get adaptive
cooperation and improve the off-peak loading protection.
Conclusions
Traditional OCR coordination considered the xed load and
maximum short circuit currents only that once the relay set will
not change again unless the load changed. In interconnected power
system, the peak and off-peak loadings are different signicantly
that one xed OCR coordination project may not be protective
promptly due to load currents are different signicantly. In the
modern power system controlled by SCADA system, it is available
to switch the IED (OCR) settings through interconnecting network
by standardization protocol. This study proposed an optimal
OCR coordination method considering peak and off-peak load
conditions for application to the interconnected power system
for SCADA system decision to switch the OCR settings during
different load conditions to get adaptive protection. The result
shows that it is applicable and satised for the OCR coordination.

Case study 2 off-peak loading power system coordination Acknowledgment

The system is the same with in Case study 1 peak and normal
loading power system coordination but loads are in off-peak
status, due to off-peak loading, with compare to Table 1 the
short-circuit currents are decreased as shown in Table 2. In this
case, short currents data are different signicantly that the small
three-phase short-circuit currents on each bus are insufcient for
activating the OCR protection promptly, while the settings were
still coordinated. To cope this problem, a simulation is performed
by using same parameter setting values of the method but use
the small short-circuit currents for generating the new CMS and
TMS. As can be seen in Table 2, the tap settings (CMS) of R1
through R6 were decreased to speed up the OCR activating in case
This work was supported in part by the National Science Coun-
cil in Taiwan, under the Project Title: Caltech-Taiwan collaboration
on energy research- uncertainty mitigation for renewable energy
integration, Project No: NSC 101-3113-P-008 -001.
ReferencesReferences
[1] GE Energy Services. Integrated Substation Control Systems and DNP 3.0 over
Ethernet, vol. 4, SSD0009; 2002. p. 12.
[2] GE Energy Services. Integrated substation control systems and IEC 60870-5-
104, vol. 1, SSD0009A; 2002. p. 1060.
[3] Arash Shoarinejad. Communication protocols in substation automation and
SCADA. GE Energy Network Reliability Products and Services; 2004.
144 C.-R. Chen, C.-H. Lee / Electrical Power and Energy Systems 63 (2014) 140144
[4] Communication Technologies Inc. Supervisory Control and Data Acquisition
(SCADA) Systems. NCS Technical Information Bulletin 041, October 2004.
[5] ABB Company. Feeder protection for overhead lines and cables in isolated
neutral or compensated networks. Version 15F-AB2, October 23, 2009.
[6] Enrquez Arturo Conde, Vzquez-Martnez Ernesto, Altuve-Ferrer Hctor J.
Time overcurrent adaptive relay. Electr Power Energy Syst 2003;25:
8417.
[7] Zeineldin H, El-Saadany EF, Salama MA. Optimal coordination of directional
overcurrent relay coordination. Power engineering society general meeting,
vol. 2, 1216 June 2005, IEEE pages: 11016.
[8] Razavi Farzad, Abyaneh Hossein Askarian, Al-Dabbagh Majid, Mohammadi
Reza, Torkaman Hossein. A new comprehensive genetic algorithm method for
optimal overcurrent relays coordination. Electr Power Res 2008;78:71320.
Elsevier.
[9] Enriquez Arturo Conde, Martinez Ernesto Vazquez. Enhanced time overcurrent
coordination. Electr. Power Res. 2006;76:45765. Elsevier.
[10] So CW, Li KK. Overcurrent relay coordination by evolutionary programming.
Electr Power Res 2000;53:8390. Elsevier.
[11] So CW, Lee KK. Application of genetic algorithm for overcurrent relay
coordination. In: IEE 6th International conference on developments in power
system protection, Nottingham, UK, March 1997, p. 669.
[12] So CW, Lee KK. Application of genetic algorithm to overcurrent relay grading
coordination. IEE 4th international conference on advances in power system
control, operation and management, APSCOM-97, Hong Kong, November
1997, p. 2837.
[13] Zeineldin HH, El-Saadany EF, Salama MMA. Optimal coordination of
overcurrent relays using a modied particle swarm optimization. Electr
Power Res 2006;76:98895. Elsevier.
[14] Bedekar Prashant P, Bhide Sudhir R. Optimum coordination of overcurrent
relay timing using continuous genetic algorithm. Expert Syst Appl 2012;38:
1128692.
[15] Urdaneta AJ, Nadira R, Perez LG. Optimal coordination of directional
overcurrent relay in interconnected power Systems. IEEE Trans Power Deliv
1988;3(July):90311.
[16] Zayandehroodi Hadi, Mohamed Azah, Shareef Hussain, Farhoodnea Masoud. A
novel neural network and backtracking based protection coordination scheme
for distribution system with distributed generation. Electr Power Energy Syst
2012;43(1):86879.
[17] Keil T, Jagar J. Advanced coordination method for overcurrent protection relays
using nonstandard tripping characteristics. IEEE Trans Power Deliv 23 52 57;
2008.
[18] Abyaneh Hossein Askarian, Al-Dabbagh Majid, Karegar Hossein Kazemi,
Sadeghi Seyed Hesameddin Hossein, Khan Rana Abul Jabbar. A new optimal
approach for coordination of overcurrent relays in interconnected power
systems. IEEE Trans Power Deliv 2003;18(April):4304.
[19] Vijayakumar D, Nema RK. A novel optimal setting for directional current relay
coordination using particle swarm optimization. Electr Power Energy Syst Eng;
2008 [1;4 Fall].
[20] Hosseinian Hadi, Askarian Abyaneh, Razavi Farzad, Galvani Sajard,
Mohammadi Reza. Coordination of overcurrent relays using intelligent
methods: a comparative study for GA, PSO and IGA. In: The International
Conference on Electrical Engineering; 2008. No. O-126.
[21] IEEE Standards Board. IEEE Standard Protection and Coordination of Industrial
and Commercial Power Systems, IEEE Std. 242 2001, Chapter 15 pp 602.
[22] IEEE Standard Board. IEEE Standard Inverse-Time Characteristic Equations for
Overcurrent Relays, IEEE Std. C37.112-1996.
[23] IEEE Committee Report. Computer representation of overcurrent relay
characteristics. IEEE Trans Power Deliv 1989;4:165967.
[24] Chen Chao-Rong, Lee Cheng Hung, Chang Chi-Juin. Overcurrent relay
coordination in power distribution system using a new approach. Electr
Power Energy Syst 2013;45:21722.
[25] Lee Cheng-Hung, Chen Chao-Rong. Using genetic algorithm for overcurrent
relay coordination in industrial power system. Intelligent Systems
Applications to Power Systems. ISAP 2007, Kaohsiung; 2007. p. 15.
[26] Chen Chao-Rong, Lee Cheng Hung, Chang Chi-Juin. Overcurrent relay
coordination optimization with partial differentiation approach for the
validation of coordination violation. Electr Power Compon Syst 2011;39:
93347.
[27] Lu Ying, Chung Jarm-Long, Kao Wen-Shiow. A novel approach to fast
eliminating the characteristic curves crossing problem of relay coordination
in radial subtransmission networks. Electr Power Compon Syst 2007;35:
27997.