Optimail Strategy for Over Current Relay Coordination Using Genetic Algorithm
Rania A. Swief, Almoataz Y. Abdelaziz" Member IEEE, A. Nagy
Department of Electrical Power & Machines, Faculty of Engineering, Ain Shams University, Cairo, Egypt
Abstract - In conventional methods coordination of Over
Current relays (OCR) is obtained through careful time
grading. Due to the expansion of electric systems, the need
for an efficient coordinated protective system is crucial.
The optimal time for protection given the coordination
problem turned to be little complex to be calculated. To
solve the problem of the coordination of OC relays,
Genetic Algorithm (GA) technique is applied. The purpose
of the OC relay coordination is to find an optimum relay
setting to minimize the time dial setting (TDS) and
calculate the pickup current (Ip). Setting of the relay is the
core of the coordination study to satisfy the primary and
secondary operation. This paper presents an application of
GA technique for optimal coordination ofOC relays to a 6-bus ring system.
Keywords - Over current relay coordination, Genetic algorithms , power system protection
I. INTRODUCTION
Directional over-current relay is an important protective device in power system. It is used to protect electric power equipment in power system when a fault occurs. Power system protection is mainly divided into protective zones. Each zone is responsible for prevention and protections operate in separate zone of responsibility as quickly as possible from the system when fault occur in the system. Over current (OC) relays are commonly used for the protection of interconnected sub transmission systems, and distribution systems [1]. This level of protection is called primary protection system and if primary protection fails or does not operate a secondary protection must be operated, which is responsible for backup the operations of the primary protection. Therefore the relay position has served as primary relay or backup relay in case of a fault occurred. To provide more effective protection, relays must be coordinated in power systems [2].
OC time setting characteristics have three criterions: constant time, instantaneous and inverse time characteristics. OC with inverse time characteristics is having fast fault clearing times, as the magnitude of the current increases. These relays are provided in electrical power system to isolate only the fault lines of the faulted section from the system.
Relay is a logical element and issues a trip signal to circuit breaker if a fault occurs within the relay protective zone and is placed at both ends of each line. Relay co
.ordination proble� is
to determine the sequence of relay operatIOns for each possIble fault location so that faulted section is isolated [3]. Good OC relay coordination is very important for industrial plant, poor
978-1-4799-5807-8/14/$31.00 @2014 IEEE
coordination may spread fault zones wider caused unnecessary power blackout or damage equipment which are avoidable , or even more affect backup utility substations [4].
The setting of OC relays has to satisfy all possible network configurations subject to type and location of all faults. It is not easy to find a proper OC relay setting to meet this requirement by traditional methods [5]. The setting of the OC relay is consisted of defining two terms, the pick-up current which the relay starts to operate and the time delay setting TDS. For the difficulty of defming the protection problems given many constraints at the same time, an essential need arise for using intelligent techniques like GA to select the suitable pick up current (Ip) and operating time (TDS). The structure of the fundamental protective functions is met under the requirements of sensitivity, selectivity, reliability and speed [6].
Many search technique have been implemented to solve the coordination problem starting from linear programming [5] till the artificial intelligent techniques like particle swarm [1, 7], differential evolution [3, 6, 8], genetic algorithm [4] and multiagent system [9]. GA is different from other search techniques in several aspects. First, the algorithm is a multipath that searches many peaks parallel, hence reducing the possibility of local minimum trapping. Second, the GA works with a coding of parameters instead of parameters themselves. The coding of parameter will help the genetic operator to evolve the current state into the next state with minimum computations. Hence GA gives the global optimum solution. GA optimization method has been employed to many power system problems [9-11] and it is applied in this paper to solve the optimum coordination of OC relays.
This paper consists of five sections. Section I presents a review for the coordination problem and the need for artificial intelligence techniques to set the optimization problem. Section II discusses the construction of fitness function and constraints of the coordination problem. Section III describes the GA outlines for OC coordination problem Section IV describes the application of GA to solve the OCR optimal coordination problem on IEEE 6-bus ring system. Section V discusses the results and conclusion.
II. COORDINA nON OF OC RELAYS
A. Objective function
In the coordination problem, the purpose is to minimize the TDS and calculates Ip of each relay, so that the sum of the operating time of the primary relay, for near end fault, is to �e minimized. The objective function can be defined as follows m equation 1:
(1)
Where; n is the number of relays and ti is the operating time of the i relay for near-end fault. The weight wi depends upon the probability of a given fault occurring in protection zone and usually set to one.
B. Relay characteristics
In this study all relays are assumed to be identical. The characteristic equation can be defined as follows in equation 2:
_ O.14xTDSi lik -1
- ( )0.02
Ipi
(2)
Where Iik is the short circuit current passing through the relay and Ipi is the pickup current settings of relay Ri
C. Coordination constraints 1. Selectivity constraints for all relay pairs
Selectivity means that the faulted line is the only part to be disconnected which means that the primary time must be greater than the secondary time with certain delay as shown in equation 3.
Tbackup -Tprimary 2: eTI (3)
Where Tbackup is operating time of backup relay.
Tprimary is the operating time of primary relay. cn is
coordination time interval, is equal to 0.3 seconds. This number based on using electromechanical relays and can be reduced using the electronic relays.
2. Bounds on TDS
There setting boundaries must be fulfilled as described in equation 4.
TDS; . ::;; TDS; ::;; TDS; <mln < <max (4)
Where TDSimin is the lower limit and TDSimax is upper
limit ofTDSi. These limits are 0.05 and l.l respectively.
3. Limits on primary operation times
This constraint imposes constraint on each term TDS of the objective function to lie between 0.05 and l.2.
III. OUTLINES OF GENETIC ALGORITHM PROGRAM FOR
OCR COORDINATION
Figure 1 shows a flow chart which describes the outlines of the genetic algorithm program as applied to the OCR coordination problem. The setting parameters for the genetic algorithm are as in Table I.
Table I - GA Parameters Number of generation 100 Number of population 100 Crossover scattered
Selection uniform
Formulate the objective function (operating time) and the constraints
GA initialization and parameters' setting
Output pick up currents and the time setting
chromosome
Figure 1 - Genetic Algorithm Program Flow Chart for Coordination of Over Current Relay
IV. THE IEEE 6-BUS MODEL
The system under study is the IEEE 6-bus ring system which is shown in Figure 2 [3]. The proposed numbering of the over current relays is shown in Figure 2. Values of relay setting parameters depend upon the number of turns in the equipment current transformer (CT). CT is used to reduce the level of the current so that relay can withstand it. The parameters of primary relays, their back up relays and their coordination are presented in Tables II & III with the relay currents and their fault currents.
Two strategies are adopted in this study. The fIrst one is to take into consideration both near and far end times (Part I) or the only the far end time delay setting (Part II).
(2) -.2 1+
7
e 12
4
The value of constants, a and c, are maximum fault currents while band d are load currents. These parameters are of primary relay conditions. The constraints are shown in equations 8, 9, and 10. They are summarized in Table II.
T�ackuP - T�rimary - CT I :2: 0 (8)
(9)
(10)
The ps is the plug setting value and this value is one of the unknown in the objective equation. The values of constants, e and g, are the maximum fault currents while f and h are the load currents. These parameters are of backup relay conditions which are summarized in Table II & Table III.
t TABLE II - SOME PARAMETERS OF PRIMARY RELAY
11
3 5
0 -.
0 14 13
Figure 2 - A typical IEEE 6-bus OC relays coordination problem model
In Figure 2, 14 over-current relays are equipped in a 6-bus system [3].
• Part I
The fIrst strategy is to calculate the sum of the near end and the far end time as follows in equations 5,6, and 7.
OBI = If�l TJr_cun + II!l T�rJar_bus
Where
ri . =
O.14*TDSi pr JUn (�)O.02
ps!*b!
rj = O.14*TDSj prJar_bus (�)O.02
psJ *dJ
(5)
(6)
(7)
T�TJUn TjT-!aT_bus
TDS a' b' TDS c
' d
TDSI 2.5311 0.2585 TDS2 5.9495 0.2585
TDS2 2.7376 0.2585 TDSI 5.3752 0.2585
TDS3 2.9723 0.4863 TDS4 6.6641 0.4863
TDS4 4.1477 0.4863 TDS3 4.5897 0.4863
TDS5 1.9545 0.7138 TDS6 6.2345 0.7138
TDS6 2.7678 0.7138 TDS5 4.2573 0.7138
TDS7 3.8423 1.746 TDS8 6.3694 1.746
TDS8 5.618 l.746 TDS7 4.1783 l.746
TDS9 4.6538 1.0424 TDSI0 3.87 1.0424
TDSIO 3.5261 1.0424 TDS9 5.2696 1.0424
TDSII 2.584 0.7729 TDS12 6.1144 0.7729
TDS12 3.8006 0.7729 TDSII 3.9005 0.7729
TDS13 2.4143 0.5879 TDS14 2.9011 0.5879
TDS14 5.3541 0.5879 TDS13 4.335 0.5879
TABLE III - SOME PARAMETERS OF BACKUP RELAY
T�aCkUP T�TimaTY
P e' f Q g' hi
8 4.0909 l.746 1 5.3752 0.2585
11 1.2886 0.7729 1 5.3752 0.2585
8 2.9323 l.746 1 2.5311 0.2585
3 1.6658 0.4863 2 2.7376 0.2585
3 1.6658 0.4863 2 5.9495 0.2585
lO 0.0923 1.0424 3 4.5897 0.4863
10 2.561 1.0424 3 2.9723 0.4863
13 1.4995 0.5879 3 4.5897 0.4863
1 0.8869 0.2585 4 4.l477 0.4863
1 1.5243 0.2585 4 6.6641 0.4863
12 2.5444 0.7729 5 4.2573 0.7138
12 1.4549 0.7729 5 1.9545 0.7138
14 1.7142 0.5879 5 4.2573 0.7138
3 1.4658 0.4863 6 6.2345 0.7138
3 1.1231 0.2585 6 6.2345 0.7138
11 2.l436 0.7729 7 4.1783 1.746
2 2.0355 0.2585 7 4.1783 1.746
11 1.9712 0.7729 7 3.8423 1.746
2 1.8718 0.2585 7 3.8423 1.746
13 1.8321 0.5879 9 5.2696 1.0424
4 3.4386 0,4863 9 5.2696 1.0424
13 1.618 0.5879 9 4.6538 1.0424
4 3.0368 0.4863 9 4.6538 1.0424
14 2.0871 0.5879 11 3.9005 0.7729
6 1.8138 0.7138 11 3.9005 0.7729
14 1.4744 0.5879 11 2.584 0.7729
6 1.1099 0.7138 11 2.584 0.7729
8 3.3286 1.746 12 3.8006 0.7729
2 0.4734 0.2585 12 3.8006 0.7729
8 4.5736 1.746 12 6.1144 0.7729
2 1.5432 0.2585 12 6.1144 0.7729
12 2.7269 0.7729 13 4.335 0.5879
6 1.6085 0.7138 13 4.335 0.5879
12 1.836 0.7729 13 2.4143 0.5879
lO 2.026 1.0424 14 2.9011 0.5879
4 0.8757 0.4863 14 2.9011 0.5879
lO 2.7784 1.0424 14 5.3541 0.5879
4 2.5823 0.4863 14 5.3541 0.5879
Tables IV shows the results of time delay settings and pick up currents relative to the fIrst proposed criterion (part 1) using the genetic algorithm and satisfying each and every constraint.
TABLE IV - THE OPTIMUM VALUES OF (TDS) AND (lp) OBTAINED USING GA TECHNIQUE - FIRST
CRITERION
Relay Number TDS Ip
Rl 0.12064 1.26713
R2 0.20392 1.30078
R3 0.l0198 l.25512
R4 0.11339 1.25
R5 0.05 1.25452
R6 0.05 1.26615
R7 0.05 1.25002
R8 0.05123 1.2539
R9 0.05 1.25055
RIO 0.06547 l.25182
R11 0.08488 1.25
R12 0.06172 1.29608
R13 0.05624 l.3133
R14 0.09684 1.25926
The optimum total time setting as it is described in details in Table IV is equal to:
FITNESS FUNCTION VALUE: lO.734481982442581
• Part II
Part II is dealing with the optimum function based on the far end values as revealed on equations 11 and 12.
OBi = Lf!l T�rJar_bus Where
(11)
(12)
Tables IV shows the results of time delay settings and pick up currents relative to the second proposed criterion (Part II) using the genetic algorithm and satisfying each and every constraint.
TABLE V - THE OPTIMUM VALUES OF (TDS) AND (Ip ) OBTAINED USING GA TECHNIQUE
Relay Number TDS Ip
Rl 0.11929 1.25
R2 0.18862 1.47913
R3 0.09688 1.26389
R4 0.11446 1.25078
R5 0.05028 1.25091
R6 0.05079 1.25
R7 0.05 1.25
R8 0.05099 1.25
R9 0.05 1.25
RIO 0.06073 1.25
Rll 0.08467 1.25
R12 0.06392 1.25
R13 0.05781 1.27076
R14 0.09673 1.25731
The optimum total time setting as it is described in detail in Table V is equal to:
FITNESS FUNCTION VALUE: 5.9225973157780185
From Tables IV & V, the results almost show the same pick up currents even if the time is calculated from far end terminal or both near and far end terminals.
CONCLUSION
Coordination of relays is a vital issue in power system studies. Good results are obtained in solving the coordination problem with the help of Genetic Algorithms (GA) on the IEEE 6 bus ring system. Two different time settings calculation strategies were applied, one technique is based on the sum of both far and near end time calculation and the other technique is based on the swn of the far end time setting only. For both strategies, the pickup currents were the same, so the only criterion to differentiate between both strategies is the time setting values. The result shows promising results of the second strategy. The results show that for vital loads which currents can never be allowed to be attained for long time, the designing of the time dial setting (TDS) based on far end time calculations provides better performance.
REFERENCES
[I] A. Rathinam, D. Sattianadan, and K. Vijayakumar, "Optimal Coordination of Directional Overcurrent Relays using Particle Swarm Optimization Technique ", International Journal of Computer Applications (0975 - 8887), Volume 10, No. 2, November 2010.
[2] M. H. Hairi, K. Alias, M. S. M. Aras, M. F. Md. Basar, and S. P. Fah, "inverse Definite Minimum Time Overcurrent Relay Coordination Using Computer Aided Protection Engineering", 4th International Power Engineering and Optimization Conference (PEOCO),2010.
[3] Radha Thangaraj, Millie Pant, and Kusum Deep, "Optimal Coordination of Over-current Relays using Modified Differential Evolution Algorithms", Engineering Applications of Artificial Intelligence 23, 2010, pp. 820-829.
[4] Cheng-Hung Lee, and Chao-Rong Chen, "Using Genetic Algorithm for Overcurrent Relay Coordination in industrial Power System", International Conference on Intelligent Systems Applications to Power Systems (ISAP), Toki Messe, Niigata, 2007.
[5] A. Y. Abdelaziz, H. E. A. Talaat, A. 1. Nosseir and Ammar Hajjar, 'An Adaptive Protection Scheme for Optimal Coordination
of Overcurrent Relays', Electric Power Systems Research Journal, Vol. 61, Issue I, February 2002, pp. 1-9.
[6] C. W. So, and K. K. Li, "Overcurrent relay coordination by evolutionary programming", Electric Power Systems Research Vol. 53,2000, pp. 83-90.
[7] H. H. Zeineldin, E. F. EI-Saadany, and M. M. A. Salama, "Optimal coordination of overcurrent relays using a modified particle swarm optimization", Electric Power Systems Research, Vol. 76,2006, pp. 998-995.
[8] S. Rodporn, D. Uthitsunthorn, T. Kulworawanichpong, R. Oonsivilai, and A. Oonsivilai, "Optimal coordination of over
current relays using differential evolution", International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON), 2012, pp. 1 - 4
[9] D. Uthitsunthorn, and T. Kulworawanichpong, "Adaptive OverCurrent Relay Coordination Based on Multi-Agent System : A Case Study on Transmission Line Outage", Asia-Pacific Power and Energy Engineering Conference (APPEEC), 2012, pp. I - 4.
[10] R. A. Swief, and Mahmoud Mohey EI-Din, "Combining both Plug-in Vehicles and Renewable Energy Resources for Unit Commitment studies in Smart Grid", IOSR, Volume 8 - Issue 3, 2013.
[11] A. Y. Abdelaziz, M. A. EI-Sharkawy and M. A. Attia, "Optimal Location of TCSC in Power Systems for increasing Loadability by Genetic Algorithm ", Electric Power Components and Systems, Vol. 39, No. 13, August 2011, pp. 1373-1387.