Coordination of Directional overcurrent, Distance, and Breaker failure relays using Genetic Algorithm including pilot protection.

Protection relays play an important role in the power systems to maintain stability, reliability, selectivity, and security for the power systems. In this paper, genetic algorithm (GA) optimization technique with pilot protection (PP) and without pilot protection (WPP) has been used to obtain proper coordination and the optimum value of transmission lines protection relays between distance relays (DRs) and directional overcurrent relays (DOCRs), as well as the critical case for fails in high voltage circuit breaker (HVCB) during faults, so using breaker failure relays (BFRs) for IEEE-8 bus system. The main aim of the used PP to reduce the overall time of DOCRs. A comparison is made with previous work in literature to show the efficiency and accuracy of the proposed algorithm with PP and the total operation time of DOCRs in the network is minimized.


INTRODUCTION
High voltage circuit breaker may be fails to clear faults during a fault. Therefore, it requires strong, secure, and correct coordination between protection relays especially at transmission and sub transmission lines due to frequent faults on this part of the power system.
Transmission and sub-transmission lines contain both distance relay (DR) and directional overcurrent relay (DOCR) widely [1]. The DR is used as the main protection relay and the DOCR as a backup protection relay in transmission and sub-transmission systems [2]. Protection relays detect a fault in which, part of a power system during abnormal conditions, and it is operating to isolate this faulty part of a system as fast as possible [3]. Therefore, proper coordination between main and backup protection relays have required, in case fails main protection relay to isolate HVCB during a fault, the Backup relay should isolate HVCB in the faulty part from the other healthy parts [4]. To improve power system reliability is required a functional duplicate of HVCB. This is a breaker failure relay (BFR) protection role, which distinguishes when HVCB fails to interrupt current after receiving a tripping signal from the main protection relay and operates with a suitable coordinate time delay to isolate backup breakers which feed HVCB faulty [5], [6]. Protection relays should have special specifications such as sensitivity, selectivity, speed, and proper setting for each relay to obtain reliability in the power systems [7] - [10]. Conventional methods such as (simplex, dual simplex, etc) have been used previously [11] - [15]. Nowadays, evaluation algorithms such as GA and particle swarm optimization are used as intelligent optimization methods to coordinate of overcurrent relays [16] - [18]. In this paper, GA optimization technique has used to obtain proper coordination between protection relays on transmission lines, also to obtain optimum values of TMS for DOCR, TZ2 for DR with PP and WPP as well as a time of BFR for IEEE-8 bus system.

PROBLEM FORMULATION
Transmission lines have contained on DR as the main protection, DOCR as backup protection, and BFR as local backup protection in the case of HVCB fails to isolate faulty part during faults as illustrated in Figure 1. There are four scenarios that should be studied to achieve proper coordination of protection relays in the transmission lines. These scenarios are coordination between (main and backup DR, main and backup DOCR, local DOCR and TZ2 for backup DR, and critical case coordination between BFR and main DR and local DOCR at case fails HVCB. Figure 2 illustrates a logic diagram of BFR. This BFR should exceed pickup current setting and waiting to receive a trip signal from any protection relays DR or local DOCR during faults, as well as HVCB, it must stay in the closing position. BFR has contained two-timer, the timer 1 operated re trip for the same HVCB but via second tripping coil. timer 2 is used as a backup trip to all HVCB which neighboring local HVCB at the case this local HVCB faulty. 3 that unnecessary faults. While, timer 2 should be set 0.1-0.11 sec to be greater than interrupting time of HVCB, current detector reset, and safety margin [19]. Figure 3 illustrates the coordination of main and backup DOCR at near and far-end faults with the constraints of the following equations: At the near-end faults (F1), the constraint is:    PP communication signal permissive under reach transfer trip (PUTT) has used to coordinate of main DR and backup DOCR in Matlab code simulation. The purpose of PUTT signal to accelerate and minimize tripping time between distance protection relays Used to protect the same transmission line. The calculation of DR divided into three zones, zone 1 is set instantaneous tripping time, while zone 2 is set with an operating time equal to 0.4 sec. Therefore, PUTT used to accelerate the trip and minimize the time from 0.4 sec to 0.02-0.04 sec, and zone 2 in Figure 4 will become illustrates in Figure 6 with the same constraint but with different operating time.

Objective function for main DRs and backup DOCRs
The problem of DRs and DOCRs coordination in the interrelated power systems can be defined as an optimization problem the main purpose of it to minimize total operating time and the objective function formula as Where OF is Objective function. Ti is Operating time for i th DOCR for near-end fault. N is the total number of DOCRs. j TZ 2 is operating time for the second zone j th DR. M is the total number of DRs.

The setting of DOCRs in optimization problems.
The time multiplier setting (TMS) bounded between two values lower and upper bound to each DOCR, as well as pickup current setting (Ips) to each one depends on lower minimum fault current and max load current.
TMSjMax is maximum bound of TMS for j th DOCR.
Where load IpsjMax− is pickup current setting for max load. According to the bounded value for TMS in equation (8) will obtain the operating time in equations (1) and (2).in this study according to IEC60255 standard, the normal inverse characteristic curve(IDMT) has used with the following equation Where T is operating time for each DOCR. Isc is the value for short circuit current passing during the relay coil, Ips is the pickup current setting for each DOCR.

Flow chart of the tripping sequence of protection relays.
The sequence of tripping to clear fault during faults according to the priority of main and backup protection relays as well as a critical case, in case of CB fails to clear a fault. All these sequences illustrated in Figure 8.

GENETIC ALGORITHM OPTIMZATION TECHINQUE
Flow chart for GA used for coordination between main and backup DOCR and the main local DOCR and DR is illustrated in Figure (9). While Table 1 shows the parameters setting of GA used in this work.  Figure 9. Flow chart for GA to solve the coordination problems between protection relays.

RESULTS AND DISSCUSION
In this paper, the IEEE-8 bus system has been used to coordinate protection relays for transmission lines and BFR for the critical case at case fails HVCB to clear faults. This system consists of two step-up transformers, two generators, an extension network at bus 4 with 400 MVA short circuits, and seven transmission lines as illustrated in Figure (10). The near-end, far-end three-phase short circuit fault current, pickup current for DOCR, and current transformer ratio have been taken from [25]. The system has forty-two protection relays, fourteen DRs, fourteen DOCRs, and fourteen for BFRs according to the number of transmission lines and HVCB in the network. The range of CTI is (0.2-0.5) second [22]- [28]. The CTI1 and CTI2 in equations 1 to 4 have taken 0.2. According to equation 8, TMS lower and upper bound limited values chosen 0.1 to 1.1 to each DOCR. While zones setting time to each DR is chosen, zone 1 = 0 sec. zone 2= 0.4 sec and zone 3= 0.8 sec [29]. The system has tested with GA optimization in an environment Matlab 2017b and compared with the results in [25]. GA has tested with PP and WPP. There are (sixty-eight) linear inequality constraints and (twenty-eight) variable fourteen for DOCRs and fourteen for DRs, all these constraints in MATLAB simulation have achieved. Table 2 and Table 3 illustrate the optimum TMS values for DOCRs from relay 1 (R1) to relay 14 (R14) and TZ2 for DRs from relay 15 (R15) to relay 28 (R28) with proposed GA with PP and WPP respectively.  Table 3 shows The TZ2 results and GA optimization technique is better than NLM optimization in obtained optimal values and performance. The overall time is reduced by about 0.0037 sec.
The time of T BFR is 0.1 in equations (5) and (6) and the BFR from relay 29 (R29) to relay 42 (R42). Therefore, the operating time for each DOCR, TZ2 for DR, the timing of third zone (TZ3) will be (TZ2+ 0.4), the operating time for each BFR, and constraints CTI1, CTI2 are illustrated in Table 4 and represent as a bar chart in Figure 11. According to results if the main zone 1of DR and main DOCR fails to clear faults, TZ2 for backup DR and time of backup DOCR will clear a fault at the same time almost. This study gives more reliability, stability, and sensitivity to the tested system.   Figure 11. The operating time of DOCRs, DRs and BFRs.
The operating time of protection relays in Figure 11 above are matching according to the flow chart in Figure 8 which explains the sequence of operating time protection relays during a fault.

CONCLUSION
In this paper, GA optimization technique to coordinate between protection relays at transmission lines has used. The results obtained with this algorithm have compared with NLM algorithm, and it was better and more accurate for reducing the overall times for all protection relays in the network.
TZ2 of DR has set as autonomous value to each relay and the main aim of that is to obtain proper coordination and give priority for the main local DOCR to clear the fault before TZ2 of backup DR and backup DOCR. It is so important to study critical case for HVCB to fail during faults, so BFR has used to coordinate with DR and DOCR.