You are on page 1of 5

Optimal Combined Overcurrent and Distance Relays

Coordination Using a New Genetic Algorithm
Method
Somayeh.Sadat.Hashemi.Kamangar* Hossein.Askarian.Abyaneh* Farzad.Razavi** Reza.Mohammadi.Chabanloo*
MSc. Student Professor Assistant Professor PhD. Student
soshashemi@gmail.com askarian@aut.ac.ir farzad.razavi@gmail.com reza_rmch@yahoo.com

Abstract—Several optimal coordination of overcurrent relays miscoordination and the other is not having the solution for
have been done in the past by using linear programming such as relays with both discrete and continuous time setting
simplex, two-phase simplex and dual simplex and also intelligent multipliers (TSMs). In [7] the mentioned problems have been
optimization techniques such as genetic algorithm (GA)
techniques. In this paper, a powerful optimal coordination solved.
method based on GA is introduced. The objective function (OF) In the cases that the distance relay is considered to be main
is developed in a such way that in addition to the coordination of and the overcurrent one is the backup relay, it is necessary to
overcurrent relays, the coordination of overcurrent and distance find the critical fault locations. The critical fault locations are
relays is also achieved. In other words, the novelty of the paper the locations at which, the discrimination time (∆t) between
is the modification of the existing objective function of GA, by the backup and main relays is at its minimum. The
adding a new term to OF to fulfill the coordination of both
coordination is made based on the constraints derived from
overcurrent and distance relays. The method is applied to two
power network systems and from the obtained results it is the values of ∆t for critical fault locations.
revealed that the new method is efficient and accurate. In [1] & [10] coordination of overcurrent, distance and
circuit breaker failure (CBF) relays has been done using
Index Terms—Overcurrent Relay, Distance Relay, Optimal
linear programming techniques.
Coordination, Genetic Algorithm.
In this paper a new optimal coordination method based on
GA is introduced. The objective function (OF) is developed
by adding a new term that is the constraint related to the
I. INTRODUCTION coordination of the distance and overcurrent relays when a
Overcurrent (OC) and distance relays are commonly used fault occurs at the critical locations.
in transmission and subtransmission protection systems [1].
To consider comprehensive coordination, all the distance and II. REVIEW OF RECENTLY DEVELOPED GA FOR OPTIMAL OC
overcurrent relays when one of them is considered to be the RELAY COORDINATION
main relay and the other is backup, must be coordinated.
When both main and backup (M/B) relays are distance GA like all other optimization methods needs initial values
relays, the impedances of the three zones of relays are which are chosen randomly. TSMs of relays are the unknown
calculated considering all conditions of the interconnected quantities in the optimization problem. Therefore, the TSMs
network such as connection and disconnection of generators with respect to the number of relays are considered as the
and lines [2]. genomes of the chromosomes in GA. In other words, some
For overcurrent relays the optimal coordination has been TSMs’ sets, i.e. (TSM1, TSM2, TSM3, …, TSMn), (TSM’1,
performed using linear programming techniques, including TSM’2, TSM’3, …, TSM’n), … belonging to relay set (R1, R2,
simplex [3], two-phase simplex [4] and dual simplex [5] R3, …, Rn) are initially randomly selected. The structure of
methods. In reference [6] also, optimal solution is made by the chromosome is shown in Fig. 1. The number of TSMs’
constraints only. The disadvantage of the above optimization sets is referred as the population size. Then, after each
techniques is that they are based on an initial guess and may iteration, the new TSMs’ sets belong to relays R1 to Rn are
be trapped in the local minimum values [7]. Intelligent given to the algorithm. The process is terminated when the
optimization techniques such as genetic algorithm (GA) can number of iterations becomes equal to the generation size [7].
adjust the setting of the relays without the mentioned To evaluate the goodness of each chromosome, it is
difficulties. essential to define an OF. The purpose of optimization is to
In these methods the constraints are included in objective minimize the OF. The chromosomes are evaluated regarding
function [7]. The optimal coordination in [8] has been done the OF and the chromosomes which have more effectiveness
* R1 R2 R3 … Rn
Department of Electrical Engineering, Amirkabir University of
Technology, Tehran, Iran.
** TSM1 TSM2 TSM3 … TSMn
Department of Electrical Engineering, Tafresh University, Tafresh, Iran.
by a method based on GA and in [9] by an evolutionary
algorithm. These methods have two problems. One of them is Fig. 1. Structure of Chromosome

International Journal of Innovations in Energy Systems and Power (Vol. 5 no. 1, April 2010) 17 of 44

The overcurrent relay is the backup of distance relay.3(sec) and 0. number of main and backup overcurrent relays that changes 3) Coordination of distance relays with overcurrent ones. one is minimum. NEW METHOD trapped in local minima. β 3 are the weighting factors. 5 no. will increase the OF = α 1 × ∑ (t i ) 2 run time. To describe the role of this tb (F ) − t Z 2 > CTI (1) new term. The required number of generation varies from i =1 system to system depending on the system complexity and ( ) P1 2 the size of population [7]. k 1 is the 2) Coordination of distance relays with distance ones. TSMs of all Δt mbDISOC k = tbOC k − t mDIS k − CTI 2 2 2 (3) overcurrent relays and the operating time of the second zone of all distance relays must be determined for critical Where tbOC k is the operating time of backup overcurrent conditions.will be used for producing new generation of chromosomes. + β 2 × ∑ Δt mb k1 − Δt mb k1 k1 =1 ( ) P2 2 III. PROBLEM STATEMENT + β 3 × ∑ Δt mbDISOC k2 − Δt mbDISOC k2 k 2 =1 As mentioned in section Ι. The critical condition is defined and shown in Fig. based on OF in GA. i is the number ones. the optimal coordination is done in such a way that all above three cases are satisfied. k 2 is the number of mai distance and backup As mentioned in section II. Then. 1. 2 2. Therefore. relay coordination must be done for the cases below: (2) 1) Coordination of overcurrent relays with overcurrent Where α 1 . F. an overcurrent relay can be the backup of some other distance relays. third term is added to OF to fulfill the requirement of overcurrent and distance relays. In this research. Therefore. coordination between overcurrent and to 2β 3 × Δ t mbDISOC k2 and obviously for positive values of distance relays must be made by adding a new constraint to β 3 the new term will have large values. An overcurrent relay is located at B and a distance one at relay for the fault at the end of the first zone of main distance M. After a fixed number of generations. the OF is formulated as: terminated. second and the third zones of distance relays are calculated. After that. t Z 2 is the operating time of the second zone of the distance OF also has a small value. Mutation in each iteration will cause the algorithm not to be IV. the concept of the evaluation and selection. the discrimination time between the the second zone of main distance relay and CTI is the operating time of overcurrent relay and that of the distance coordination time interval that is equal to 0. after entering the network data. β 2 . The appointed at the critical fault location. 2. 3. are granted more opportunities to be selected for the next iteration. At first. the process will be In the new method. from 1 to P1 . Critical fault location in coordination between overcurrent and distance relays International Journal of Innovations in Energy Systems and Power (Vol. This assumption main distance and backup overcurrent relays which is is a routine protection scheme in power networks. the short circuit currents for the faults exactly close to the circuit breaker (CB) of the main overcurrent relays and for the faults at the end of the first zone of main distance relays( critical fault locations) are calculated. The details of this method are described in section IV. In obtained from the equation below: interconnected networks. the first one has been made in overcurrent relays changing from 1 to P2 .3(sec). assume that Δ t mbDISOC k 2 is positive (fully coordinated). For discrimination time between the main and backup overcurrent the third case it is assumed that the distance relay is the main relays. of overcurrent relays that changes from 1 to N. negative (miscoordination) the mentioned term will be equal As mentioned above. The flowchart of new method is shown in Fig.6(sec). April 2010) 18 of 44 . However if Δt mbDISOC k2 is relay and CTI is the coordination time interval. relay(critical fault locations). the impedances of the first. t mDIS k is the operating time of 2 When a fault occurs at F. GA Fig. the expression below must be Two first terms of (2) are the same as the OF in [7]. Then. Increasing the number of generations will lead to N the better solutions and on the other hand. then the relative term in (2) becomes zero and Where t b (F ) is the operating time of overcurrent relay at F. those values that have more optimal OF values (less value) in the chromosomes. Δt mbDISOC k is the discrimination time between the 2 relay and the overcurrent relay is the backup. Δt mb k1 is the the frame of GA. The operating time of the second and third zones of distance relays are selected respectively 0.

β 2 = 200. is applied to two That means. All Δt mb k1 and V. the output results for TSMs are obtained. k2 between 0.0222 150 the relay characteristic is used [11]: TABLE II ⎛ K ⎞ (4) t = TSM ⎜ + L⎟ GENERATORS’ INFORMATION OF SAMPLE NETWORK ⎝ M a −1 ⎠ X (pu) V (kV) Generator 0. These TSMs are given in Table V. 1 circuit currents.0018 0. International Journal of Innovations in Energy Systems and Power (Vol. t is the relay operating time and K . ti . Fig.3(sec) and 0. Δt mb k and Δ t mbDISOC are calculated short values of TSMs are small and they are in valid range. April 2010) 19 of 44 .e.02 150 All distance relays have moho characteristic. Flowchart of new method From the second column of Table V. Sample network It is also assumed that TSMs of the relays are continuous and TSMs vary from 0 to 1. the values of α 1 .0018 0. 5 no.0222 150 3 0. it can be seen that the will start. II and III. β 3 = 200 (5) By applying the GA with selected values. 0.6(sec). 4.0018 0. discrimination time between the main distance and backup overcurrent relays are given in Table VI. α. 5 0. The new method proposed in this paper. TEST CASES Δt mbDISOC k 2 values are positive and most of them are small. β 3 are mentioned as below: α 1 = 1. 4.1 10 Where M is the ratio of short circuit current ( I sc ) to the pickup current ( I b ) of relay ( M = I sc / I b ). It is assumed that all the 2 0. i. 4 0. The information data of Line R (pu) X (pu) V (kV) the network is given in Tables I.02 150 lines are protected by both overcurrent and distance relays.02 150 operating times.0018 0.0022 0. The control parameters of GA are listed in Table IV.0222 150 are based on 100 MVA and 150 kV. Fig. 7 lines. α and L the normal inverse characteristic will be obtained for overcurrent relays that are used in this paper[11]. a more common formula for approximating 7 0. 3.14.02 150 It should be noted that for finding the overcurrent relays 6 0. fit and have not any miscoordination. To compose OF.0022 0. power networks: 1) Case 1 TABLE I LINES’ INFORMATION OF SAMPLE NETWORK This network consists of 6 buses. The other steps of GA are described in Discrimination time between M/B overcurrent relays and section II and in [7]. The operating time of the second and third zones of distance relays are selected respectively 0. 2 transformers and 2 generators and is shown in Fig. the relay settings are accurate.05 and 1. By the values of 0. β 2 .02 and 0 respectively for K . R (pu) and X (pu) 1 0.0022 0. 1. L are the scalar quantities.

1689 0.1691 14 0.255 1 6 0. Only a few Size of population 150 numbers of the Δtmb k1 and Δt mbDISOC k 2 have negative Initial Population Random Mutation 1 values that are very small. The detailed relative information of all relays is not given because of space limitation. 11 12 0. β 2 and β 3 given 12 13 0.2201 given in [12].0000 12 0. The generator.0110 0.05 2 7 0.05 successful and accurate. transmission lines and transformer information are 12 14 0.0026 0.0035 0.122 been tested on two power system networks. International Journal of Innovations in Energy Systems and Power (Vol.0000 lines.1443 0.0037 0.0000 This network is IEEE 30 buses system which can be 10 11 0. SAMPLE 1 VI. CONCLUSION Relay TSM A new computer program for distance and overcurrent Numbers relay coordination based on GA has been developed. In the 1 0.126 5 4 0. CASE 1 6 0. TABLE V TSMS OF OVERCURRENT RELAYS.0023 0. has been applied to the network of Fig. and Transformer 0. 5 0. That means.0015 11 0. 5. shows the system. From the obtained results.0000 Again GA with selected values of α 1 .2613 4 3 0.0000 10 0.2109 13 0.1362 0. The values of the second and the third TABLE IV columns are respectively the numbers of M/B relays and the CONTROL PARAMETERS OF GA average of the related discrimination times.110 3 2 0.0000 0. The computer program has 3 0.0000 in (5) and GA parameters given in Table IV. 4 transformers and 86 overcurrent relays 6 14 0. 2 1 0. April 2010) 20 of 44 .0000 and 86 distance relays. 6 Generators.2781 0.114(sec) that is small and within the range.05 TABLE VI DISCRIMINATION TIMES OF M/B RELAYS.183 overcurrent relay coordination. It can be seen that GA Parameters Value most of the Δtmb k1 and Δt mbDISOC k 2 have positive values Number of Generations 3000 and the related M/B relays are fully coordinated.083 Main Backup Relay Relay Δt mb k1 (sec) ΔtmbDISOC k (sec) 2 8 0.1067 0.02666 Δt mb k1 and Δt mbDISOC k 2 smaller than zero (miscoordination). The average value of TSMs is 0.0523 2) Case 2 9 10 0.0021 0. TABLE III TRANSFORMERS’ INFORMATION OF SAMPLE NETWORK The first column of Table VII consist of Δt mb k1 and X (pu) Δt mbDISOC k 2 greater than zero (fully coordination). Fig.05 14 1 0. the OF has been modified by adding a new term which presents the constraint for distance and 2 0.2736 0.1000 0. it has been shown that the new method is 4 0.194 7 0.147 9 0. 5 no. 5.091 proposed method.0024 13 8 0. 1.0030 0.101 8 7 0.2024 The network consists of 30 buses (132 and 33 kV buses). we almost have no miscordination.0171 considered as a meshed subtransmission /distribution system. The summary of the discrimination times are given in Table VII. 37 7 13 0.6351 0.

on Power Delivery. Electr . . on Power Delivery. “An on-line relay Bahru. pp. J. S. on Power Delivery. pp. no. [7] F. “Application of genetic algorithm for overcurrent relay coordination. 165–173. M. IEEE 30 Buses Network [5] H.” in: Proc. Baltazar. “A new optimal approach for coordination of overcurrent relays in interconnected power systems”. 3. Hashemi. May 2007. Fajardo. Delivery. Mohammadi and Δtmb k1 (sec) < 0 2 -0. IEEE Trans. M. Δt mbDISOC k 2 (sec) > 0 225 0. Y. Palomino and M.3. [3] A. 440–445. H. R. Res. 36.0019 REFERENCES [1] L. W. M.washington. and K. R. Urdaneta. G. 5. IEEE coordination of overcurrent and distance relays considering various Trans. Urdaneta. 2008 PECON Conf. 1. October 1999. Razavi. on Power an adaptive protection system” . on Industry Applications. pp. “Optimal co-ordination of TABLE VII overcurrent relays in power system by dual simplex method. 5 no. Khan. Perez.1995. K. IEEE Trans.385-388. pp. September-October 2000. Chabanloo. Askarian. Resterpo. “Optimal coordination of overcurrent new approach for calculating zone-2 setting of distance relays and its use in relays considering definite time backup relays”. S.” IEEE Trans. K. Abyaneh and R. pp. Abyaneh. 5. S. no. 11. Sanchez and J. in: Proc. Razavi “A new genetic algorithm method for optimal directional overcurrent relays timing using linear programming”.edu/resesrch/pstca.16. and T. no. Abyaneh . 1995 DISCRIMINATION TIMES OF M/B RELAYS. 1.ee. on Power Delivery. Urdaneta. H. no.Fig. April 2003. CASE 2 AUPEC Conf. available at techniques.0625 [8] C. Perth. www. S. M. Chattopadhyay.1051 vol. S. Askarian.0023 H. [11] R. So and K. Li. July 2001. January 2004. Askarian. protection by evolutionary algorithm”. Sachdev. Torkaman. 11. vol.Power Syst. H. “Time coordination method for power system second zone timing in a mixed protection scheme with directional relays”. vol. vol. no. . H Sadeghi Δt M/B Relays Value and R. Power Delivery. G. J. no.1276-1284. January 1996. Lai. S. Al-Dabbagh. So. 1235-1240. 3 -0. Abyaneh. 18. Sachdev.122-129. J. Sidhu . Al-Dabbagh. S. Karegar. W. Numbers of Average [6] H. 1997 IEE Conf. 19. 1. IEEE Trans. Δtmb k1 (sec) > 0 226 0. D. “Optimal computation of distance relays [9] C. pp. pp. Johor [4] B. vol. April 2010) 21 of 44 .4. A. Australia. “A [10] L. Perez and A. 2. vol. A. . J. Keyhani. 1. vol. K. “Coordination of Kamangar. Li. H. Δt mbDISOC k 2 (sec) < 0 Developments in Power System Protection. K. 66–69. coordination algorithm for adaptive protection using linear programming [12] Power system test cases.” in: Proc. vol. 1997. Malaysia. no. pp.70-77. characteristics for overcurrent relays”. H. January 1996.14. F. IEEE Trans. K. Sidhu. IEEE Trans. [2] T. Askarian. H. S. “A new comprehensive genetic algorithm method for overcurrent relays coordination”. Mohammadi. T. December 2008. International Journal of Innovations in Energy Systems and Power (Vol. Fung.