Protective Relaying Coordination in Power Systems Comprising Renewable Sources: Challenges and Future Insights

Abstract

This article provides a comprehensive review of optimal relay coordination (ORC) in distribution networks (DNs) that include distributed generators (DGs). The integration of DGs into DNs has become a real challenge for power system protection, as the power flow changes from unidirectional to bidirectional, which complicates the relay settings. The introduction of DGs in DNs requires changes and modifications in the protective schemes to maintain proper operation, reliability, stability, and security of the system. This paper focuses on the impacts of DGs penetration into DNs, including the effects on protective scheme coordination. Various expressions for characterizing the overcurrent (OC) coordination problem, as well as related solution attempts, are discussed. Several optimization strategies and techniques are suggested by scientists to deal with coordination optimization problems aiming to achieve less computation time and better accuracy. All these efforts ultimately aim to define optimal relay settings to achieve ORC by generating the optimal setting of cascading relative OC relays. This comprehensive review provides a broad overview of the contributions of scholars in recent publications in this field, with more than 210 articles reviewed and analyzed. It is a valuable resource for other researchers in the same field who aim to tackle ORC problems in their future endeavors.

1. Introduction

Global climate change is driving the need for countries to integrate different and reliable generation technologies into their power networks. Solar, wind, fuel cell, and other renewable energy sources (RESs) technologies are various forms of distributed generators (DGs) that can be integrated into low-voltage networks or in the medium-voltage side to meet future power demands. The integration of DGs into DNs can increase efficiency, reduce power outages, serve client demands, and decrease electricity costs, offering economic, ecological, and technical benefits [1,2,3].

However, the high penetration of DGs into DNs can also produce problems and disturbances for the network. Governments around the world are disseminating energy policies to encourage the exploitation and use of RESs, including the adoption of highly developed technologies to authorize the incorporation and accounting of energy production and consumption [4,5,6].

In [7], the basis for increasing the penetration of DGs into DNs is presented. This highlights the need for proper management of the integration of DGs into DNs to ensure optimal and reliable operation of the power system. As the demand for electricity continues to grow, the integration of DGs into DNs will become increasingly important in meeting the needs of society while also addressing global climate change challenges.

The effects of DGs involve the consequence that is produced by DGs in voltage regulation, providing security, and nourishing reliability, as well as impacts on equipment control and disturbances in safety and impacts on the islanding situation clarified in [8,9,10]; the authors of [11] estimate transient and small-signal stability of DGs connected to utility DNs, voltage regulation is illustrated in [12], and power losses and voltage profile issues are discussed in [13]. In [13,14,15,16], distributed power generation is discussed in detail. The potential of DG systems equipped with an appropriate power electronic interface is illustrated [17,18,19]. Other impacts of DGs on DNs are fault current level and its direction and the protection system issues [20,21,22,23], while system stability issues are discussed in [24]. The aforementioned technical concerns can be alleviated by the way of appropriate system protection and control [10,25,26].

The effects of DG on a present DN may be positive or negative relying on the position and the size of the DGs, as presented in [15,16,20,27,28,29]. Imitation of the impacts of DG capacity and location of the DN must be conducted [30]. In [31,32], researchers discuss the influence of DG capacity as well as the location in DNs through stable situations and abnormal situations. Furthermore, there are many different methods including analytical approaches and intelligent algorithms to define the optimum sizing with placement for DGs into the DNs [33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52].

DGs significantly impact fault current level contributions, as discussed in [27]. Since synchronous generators have much higher short-circuit levels than inverter-based DGs, they consequently give rise to a much stronger impact on the protection systems [53]. DGs produce power with capacities from several kilowatts to a few megawatts [54], and researchers in [55] illustrate the most common classification of DG capacities. The more important problem is the miscoordination between overcurrent protection devices during abnormal conditions due to bidirectional power flows within the DS [56,57,58,59].

DOCRs are the economical selection to protect power systems and can be used as primary or backup protection [60,61,62]. Topological analyses using graphical theoretic as well as function techniques and optimization frameworks are applied to solve the coordination matter for DOCRs [63]. Protection coordination schemes are used to cascade the operations of main and backup relay schemes [64]. Coordination problems could be solved by using adaptive techniques [65].

The features of this effort may be stated as follows: (i) this review paper is focused on presenting issues relating to the integration of DGs into DNs, (ii) this paper presents several coordination optimization techniques that are utilized for fixing the miscoordination matter [66,67], and (iii) this paper shows a survey of protection coordination strategies and strategies using the optimization methods.

The other sections of this paper are as follows: Section 2 illustrates the impacts of the penetration of DGs onto DNs. In Section 3, the influences of penetration of DGs into DN on protection systems are presented. Section 4 discusses ORC techniques and optimization frameworks, showing the ORC problem formulation and various optimization techniques for ORC. Section 5 announces a sample of the representative results in regard to IEEE 15-bus electric power systems. Lastly, the conclusion that completes the paper with some guidance for future research is presented in Section 6.

2. Impacts of Penetration of DGs on DNs

DGs are known as energy sources linked to DNs [68]. Nowadays, there is a large trend and a huge increase in the use of DGs in DNs because of their many important benefits, including a reduction in cost of energy generation and high efficiency, improved voltage profile, power loss reduction, increase in network reliability, and contribution in the preservation of the environment [69,70,71,72,73]. Although the integration of DGs has many profitable advantages, it puts DNs in front of many challenges because of the significant problems that require urgent solutions [72,73]. The influence of DGs on DNs is based on placement, size, working principle, and the characteristics of DGs [52]. There are many types of DGs, as shown in [71,74]. In the following subsections, some of these problems are briefly presented.

2.1. Voltage Rises and Reverse Power Flow

The DGs penetration into DNs changes the loading conditions of the network to the point of common connection [74], resulting in significant influences in voltage profiles and the operation of voltage regulator of the feeder [63]. The voltage level of the DN has boundaries defined by global standards such as ANSI C84.1 [75] and IEC 60038 [76], and when these standard voltage boundaries are exceeded, the network stability deteriorates [77].

Increasing penetration of DGs leads to a raised voltage level [78]. For maintaining the DN voltage in permissible boundaries, reactive power is required [69]. Many researchers discuss the size of DG penetration into the DN keeping the voltage level as it is [74,75]. In [74,79,80,81,82,83,84,85,86,87,88,89,90,91,92], researchers present different methods for voltage rise control in a DS with DG penetration to regulate voltage level.

2.2. Power Losses

The poor selection of placement and size of DGs in DN could result in more losses in comparison with the losses occurring in the case of no DGs [93,94]. So, a significant amount of research [16,20,95,96] presents studies to decide the optimum place where DGs should be interconnected to a DN and the amount that can be integrated into the DN to decrease power losses and keep voltage levels inside acceptable boundaries. In [67], the OPF is applied to define the optimum DG capacity.

2.3. Islanding Issues

Islanding of DGs has negative effects on a network and weakens the reliability of the source for DNs [97,98]. The requirements for planning islanding are introduced in IEEE 1547.4 [99]. This approach presents a comprehensive presentation of requirements and different forms for islanding schemes. In [100,101], the advancement of control concepts for DGs during islanding conditions is discussed.

The control and protection scheme should be coordinated to reduce the tripping of DGs and loads while keeping both voltage and frequency within standard levels [102,103].

3. The Impact of Adding DGs into DN on Protection Systems

The introduction of DGs in a DN causes challenges related to ORC in the DN. The augmentation in the short-circuit currents with changes in their direction in the DN [104,105,106,107] affects the ORC of relays installed in the DNs and upsets the selectivity of those protective devices. Furthermore, blinding of protective devices, false tripping, unintended islanding, and asynchronous reclosing are considered as major difficulties in the protection problems resulting from integrating DGs in DNs [108]. After integration of DGs into a DN, the network changes from a radial structure to a loop structure. This results in the miscoordination of the existing network protection devices. Some of protection challenges are discussed as the following.

3.1. Overcurrent Protection and Short-Circuit Levels

DG penetration changes power flow from unidirectional to bidirectional [109] and requires modifications in protective device coordination [110]. This has an influence on the short-circuit current magnitudes, influencing both the ORC and device ratings [12]; meanwhile, the main protective and fault detection element in a distribution network is overcurrent protection, so in every protective device, directional elements must be inserted for the proper operation and to face the bidirectional problem [111,112]. Short circuits lead to very high current flows and are required to be interrupted prior to any damage to network equipment [113]. The alteration of short-circuit current level and the direction of this fault lead to miscoordination as well as incorrect tripping [20,21,22,23]. Moreover, the short-circuit capacity level depends on the DG unit type [27,114], location, capacity, and mode in DG linked or islanded styles. Then, in DG linked style, the fault current become very high because the fault is supplied by both the utility and the DGs, but in islanded mode, the fault current is highly decreased, whereas the utility is the only source feeding the fault [115]. With practical experience, it has been noted that overcurrent protection is very sensitive to bad discrimination in networks with high penetration of DGs [116,117].

Miscoordination of overcurrent protection due to changes in short-circuit levels after integration of DGs into networks can appear in two ways. Firstly, in “false tripping”, a protective device can sense and trip unnecessarily for a fault in an adjacent protection zone which has a faulty feeder, not in its own protection zone which has a healthy feeder. The impact of false tripping is illustrated in Figure 1, where a DG is connected to feeder 2 and a fault occurs at feeder 1. The fault current that flows through feeder 1 is the summation of short-circuit current supplied from the utility and short-circuit current contributed by DG1, and with high penetration of DG, the current flow in feeder 2 would exceed the operating value for relay R2 and result in R2 operating before the operation of the relay R1 which had to work first [116].

Secondly, in “blinding of protection”, a protective device fails to trip even though a fault occurs within its protection zone. Since the fault happens at the lower side of the line, considering that the impedance of the network would directly be proportional to the impedance of DG in increasing or decreasing, the short-circuit current that flows through the feeder relay (R1) from the utility will be below the Ip of this relay, as shown in Figure 2 making the relay fail to detect a fault and system exhibit blinding of protection [116].

3.2. Recloser–Fuse Miscoordination

The recloser is one of the most important protective devices in the network as it guarantees achieving high reliability for the DN as most faults on DNs are temporary [117]. Provisional faults represent 70% to 80% of faults taking place in a DS. The fuse used in a lateral feeder isolates a fault by burning out, so this fuse must be replaced; that is the reason for using a fuse with a permanent fault. To achieve a proper process, the fuse should be in coordination with the corresponding recloser, as illustrated in Figure 3 and Figure 4 showing the recloser–fuse arrangement. The coordination between the recloser and fuse is commonly accomplished by considering fuse-saving basics [21]. In recloser–fuse coordination, just the fuse must work for a lasting fault, while the recloser must operate with a fast operation for a provisional fault to give the fault a chance to be erased. If the fuse cannot operate for a permanent fault, the recloser becomes a backup for the fuse through slow mode and finally locking out [15]. Figure 3 shows the range where the recloser and fuse achieve coordination for all fault currents within $I_{fmin}$ and $I_{fmax}$ as discussed in [60].

After integration of a DG, the upper and lower values for fault currents regarding the loading feeder fault current, representing the coordination range boundaries as shown in Figure 3, will change, and the current passing in a fuse during fault may be higher in comparison with current passing in recloser, resulting in miscoordination between the recloser and fuse depending on the distance of the DG from the recloser, as introduced in [116]. The miscoordination between the recloser and fuse increases the probability of a fuse blowing out on a temporary fault before the fast operation of recloser [15].

To tackle this problem, the modification of the coordination curves of the recloser and fuse by moving the fast curve for the recloser to work first is proposed. In [118], researchers proposed an FCL to mitigate the miscoordination between the fuse and recloser, so this problem of coordination could be eliminated by correctly sizing the reactor. In [60,119,120], another solution is presented to avoid recloser–fuse miscoordination by determining the maximum permissible capacity of a DG that guarantees no coordination failure.

3.3. Anti-Islanding Protection

Islanding can occur when the main utility disconnects and continues to be energized by its DGs [121]. Intentional islanding is prohibited by AS4777.3 [122]. So, it is imperative that islanding is detected, and each DG unit is always provided with anti-islanding protection to detect an islanding condition. Bear in mind that DGs must be disconnected before upstream network reconnection [123,124].

Anti-islanding protection can be categorized into three different types as discussed in [125,126,127,128]. In [127], two-level islanding detection procedures are discussed for DGs in DNs.

4. ORC and Optimization Techniques

Due to the aforementioned significant issues and problems resulting from adding DGs in a DN, especially the protection challenges, it was necessary to review and change relaying coordination to maintain the reliable, stable, and proper operation of DNs. In DNs, the widespread elementary protection is overcurrent security [128]. However, with looped networks including various sources where power can pass in two directions, the most appropriate choice capable of dealing with these conditions is DOCRs [129]. The target of the optimal coordination of DOCRs is determining proper relay settings to ensure the elimination of faults in protection as fast as possible through the primary relay which is responsible for protecting this area, but if the running of the primary relay is unsuccessful, the responsible backup relay operates with a delay called the CTI [130]. The optimal coordination must achieve the fundamental protective function represented by sensitivity, selectivity, reliability, and speed. Subsequently, the DOCRs are adapted to form an optimization problem. This problem is highly constrained with various forms of constraints, and it can be formulated in different ways.

4.1. ORC Problem Formulation

The problem of coordination of DOCRs in a meshed network can be formed as an optimization problem. Thus, this optimization problem is solved with various techniques, as will be discussed in the next section. The coordination problem must satisfy some important requirements that guarantee the reliability, stability, and proper operation of DNs as introduced in [131]:

✓

Reduce the overall operation times for primary relays considering achieving coordination between all relays.

✓

The PS and TDS settings must be adapted with various conceivable topological and working strategies.

✓

The optimization technicality must have the capability of reaching near-universal optimum settings.

The relaying coordination problem is represented by defining the concatenation of relay operations to any conceivable fault site for isolating the faulty part, with sufficient coordination margins including the operating time of the corresponding circuit breaker and security margin [132], and the reliability in system protection requires at least two protection levels in order. Based on that, the coordination problem aims to determine optimal TDS as well as PS with a minimal operation time for all relays and under some constraints [133].

The coordination problem could be in two formulations; the first is an LP problem while the second is an NLP problem. In the linear problem, Ip is constant and known, ranging between the upper value for load current and the minimum current during fault, while only the optimization value for TDS is determined. In the nonlinear problem, both TDS and PS values should be optimized with each other. Many authors use nonlinear optimization problems for TDS and PS optimization [133,134,135,136,137,138,139,140]. In [141], NLP is used to solve the problem of coordination to optimize PS as well as TDS by optimizing the coefficients for the inverse-time relay curve, providing more flexibility to adjust the characteristic of the relay while taking fluctuation in topology as well as loading into consideration. The PS is firstly determined with the evolutionary algorithm; then, the PS becomes known and is used in LP to optimize the TDS [142].

The optimization of coordination for protective relays in power systems is a complex problem that involves multiple objectives, constraints, and variables. Researchers have proposed various approaches and techniques to solve this problem, including NLP models, weighted optimization, and LP techniques.

In one approach, researchers formulated an NLP model that includes the fault current limiter (FCL) as a variable to be calculated [143]. In another study, the problem of coordination for DOCRs considers changes in the topology of the network and considers the primacy of constraints through specifying weights [144,145].

To avoid the complexity of mixed integer NLP techniques, some authors and approaches have proposed using PS for DOCRs with TDS parameters, which transforms the problem of coordination into a linear form [146,147,148,149]. In another study, interval LP was used to resolve the problem of coordination for DOCRs, taking into consideration changes in the topological conditions of the network [150].

Furthermore, LP-style optimization techniques were applied to obtain the optimal PS and TDS for OCRs in [151]. These methods help to solve the problem of coordination for protective relays in power systems and improve the reliability and safety of the power system operation.

The OCR characteristic is a nonlinear equation with two control variables: PS and TDS. The operating time equation for the relay includes PS, TDS, and the fault current that flows through the relay and is formed as expressed in (1) [148,150,151,152,153,154,155,156].

$$t_{op}=TDS\left(\frac{A}{{\left(\frac{I_{fault}}{I_P}\right)}^B-1}\right)$$

(1)

where $t_{op}$: the time of relay operation, $I_{fault}$: abnormal current passing in the protective relay, $TDS$: time dial setting, and A and B: characteristic constants.

According to the type of OCR used, the constants A and B are determined [157]. Figure 5 indicates the four basic characteristics according to IEC 60255-3. They are normal inverse (NI), very inverse (VI), extremely inverse (EI), and long inverse (LI). However, the effect of varying TDS is depicted in Figure 6.

The objective function (OF) for the problem of coordination represents reducing the overall sum of the times of operation of all primary relays inside the network and keeping all relays coordinated [132,134,135,136], but it can be formulated in various expressions. In [137], the OF is used for decreasing the overall sum of times of operation of all primary and backup relays. The authors of [138] presented the OF as minimizing times of operation among main and backup operating relays (CTI). Some researchers have taken different approaches in the introduction of the OF, including reducing the overall sum of times of operation of all primary relays only to faults of the nearby end [132,134,135,136]. However, the OF has two parts representing the times of operation of all primary relays with faults on the nearby end as well as the far end, as shown in [135,136,138,139]. The common and widely used OF is described in (2).

$$OF=minimize\left(\sum _{i=1}^N{t}_{pi}+\sum _{r=1}^{TP}{t}_{br}\right)$$

(2)

where $t_{pi}$: the operating time of the primary relay, $t_{br}$: the operating time of the backup relay, $N$: number of primary relays in the pairs, and $TP$: number of backup relays in the pairs.

Increasing the degree of nonlinearity in optimal relay coordination (ORC) involves incorporating more constraints into the coordination problem formulation. This can help to improve the accuracy and effectiveness of the coordination solutions by considering the complex interactions between protective relays and other system elements.

One approach to increasing the degree of nonlinearity in ORC is to incorporate additional constraints based on the characteristics of the power system and the protective relays. For example, the coordination problem formulation can include constraints related to the fault current magnitude, fault duration, fault location, load variations, and other system parameters. These constraints can help to ensure that the protective relays operate in a coordinated manner under a wide range of operating conditions.

Another way to increase the degree of nonlinearity in ORC is to use more advanced optimization techniques. For example, evolutionary algorithms, neural networks, and fuzzy logic can be used to solve complex coordination problems that involve multiple objectives, conflicting constraints, and nonlinear interactions between system components. These techniques can help to improve the accuracy and efficiency of the coordination solutions and provide more robust and reliable protection for the power system.

In addition, the use of advanced simulation tools and models can also help to increase the degree of nonlinearity in ORC. For example, the use of real-time simulation tools that incorporate detailed models of the power system and protective relays can help to identify and address complex coordination issues that may not be apparent in traditional analysis methods.

The coordination problem represented in the OF to achieve the optimal solution must be subjected to groups of constraints, as detailed in the following subsections.

4.1.1. Constraint of Relay Operating Time

The relay is required to operate in minimum time and at the same time does not take much time to work to achieve the basics of protection. The limitation mathematically can be expressed as follows:

$$t_{i,min}\le t_{i,k}\le t_{i,max}$$

(3)

where $t_{i,min}$: minimal time of operation of relay on site i to the fault in any site inside its area for working, and $t_{i,max}$: upper time of operation of relay on site i to the fault in any site inside its area for working.

In [135,136], a limitation in primary times of operation has been enforced for every parameter in OF; this limitation ranges from 0.05 s to 1 s.

4.1.2. Constraint of PS of Relays

To ensure the achievement of the sensitivity of the relay to the minimum fault current, the minimal Ip setting is greater than or equal to 1.5 from the highest current of load, while the highest Ip setting is less than or equal to two-thirds from the lowest current of fault [137]. PS limits are bounded by

$${PS}_{i,min}\le {PS}_i\le {PS}_{i,max}$$

(4)

where ${PS}_{i,min}$: minimal value for PS for ith relay, ${PS}_{i,max}$: highest value for PS for ith relay, and ${PS}_{i,min}$ and ${PS}_{i,max}$ values are taken as 1.25 and 2.0 pu, respectively [135,136,137].

4.1.3. Constraint of TDS of Relays

The TDS has importance in coordination for relays and affects the operating time of relays. TDS limits are

$${TDS}_{i,min}\le {TDS}_i\le {TDS}_{i,max}$$

(5)

where ${TDS}_{i,min}$: minimal value for TDS for ith relay, ${TDS}_{i,max}$: highest value for TDS for ith relay; ${TDS}_{i,min}$ and ${TDS}_{i,max}$ values are taken as 0.05 and 1.1, respectively in [135,136], and assumed as 0.025 and 1.2, respectively, in [138,139,140], and in [140], authors use values of 0.01 and 1.0, respectively.

4.1.4. Constraint of Minimum Operating Time

The minimum operating time of numerical relays should be considered to imitate the real installations as illustrated in (6).

$$t_{op,iP}\ge T_{min}$$

(6)

where $t_{op,iP}$: operating time of ith relay (primary relay for specific relay pair along the network), and $T_{min}$: minimum operating time of digital relay from a practical point of view, i.e., 10–50 mS.

4.1.5. Constraints of Coordination

The secondary relay should be operated as a backup relay after a certain time, which is called CTI. This period includes the operation time for the circuit breaker corresponding to the primary relay in addition to overshooting duration. So, to achieve proper operation for relays, the CTI can be expressed as follows:

$$t_{i,k}-t_{j,m}\ge CTI$$

(7)

where $t_{j,m}$: the operation time for primary jth relay during fault at m, and $t_{i,k}$: the operation time for backup ith relay during the fault itself at m.

4.2. Optimization Techniques for Relaying Coordination

The purpose of relay coordination is to properly protect the power system from abnormal conditions by disconnecting faulty sections immediately to guarantee continued service in the healthy section of the power system [62,158,159,160,161,162,163,164,165,166,167,168,169,170,171].

To maintain the coordination among relays, [135] proposed inserting an FCL to shackle the current from a DG only during the fault; the authors used MATLAB for determining the lowest value of impedance for the FCL while defining the appropriate kind of FCL for different sequences of operation for DN. The FCLs are used to keep the directional OCRs (DOCRs) in coordination status without any modification in OCR settings regardless of DG status [67]. Meanwhile, in [63], researchers presented a reduction in impedance for the FCL to be beneath the critical rating. The applied scheme in [63] has the feature of the capability to be utilized with any network regardless of sizing or number of DGs.

Adaptive relaying is used for optimum coordination of DOCRs in meshed networks; the impacts of DGs are discussed in [170], and ABC was used for resolving the coordination issue. The proposed scheme in [154] acclimates to changes in the network representing generation capacity changes that require new relay settings and network topology changes. In [154], the requirement for adaptive protection is listed and illustrated. In [155], a protection coordination style that depends on the technology of the agent has been proposed. Another form of adaptive protection system for DNs with DGs is applied in [156]; the proposed system depends on a network zoning approach, in which each zone acts as a separate section with the capability of island operation if necessary. In [157], an adaptive protection system for fuse saving in DNs that have DGs is shown. In [158], an expert system is used for protection coordination for DNs that have DGs.

Approaches for relay coordination can be broadly classified into two main categories: topological analysis and optimization techniques [63]. Prior to these classifications, some authors used trial-and-error methods for relay coordination, as discussed in [159]. However, this method can require many iterations, as noted in [160]. In [161], topological analysis was used to determine breakpoints and reduce the number of iterations required for the coordination process, mitigating the disadvantages of the trial-and-error method. The topological analysis approach was introduced by [162]. However, the solution obtained using this method may not be optimal.

Optimization techniques for relay coordination can be complex and time-consuming, especially when using NLP methods. Some researchers have used NLP methods, such as sequential quadratic programming, to optimize the protective relay settings [133]. In other studies, the problem of relay coordination was expressed as mixed integer NLP and solved using software such as the General Algebraic Modeling System [163].

To reduce the complexity of the coordination problem, some researchers have used LP models, such as those discussed in [164,165], to determine the optimal PS and TDS for OCRs. However, the use of LP models can be limited by the need for an initial guess, which can result in the solution being trapped in a local minimum rather than achieving the global optimal solution [166].

To address these challenges, researchers have developed intelligence-based optimizers, which are summarized in subsequent sections. These optimizers can help to solve the problem of relay coordination regardless of the complexity of the optimization problem and improve the reliability and safety of power system operation.

4.2.1. Genetic Algorithm (GA)

GA is a searching method that depends on the basics of genetics and naturalistic choice [167]. GA has the flexibility to make it an attractive method for many optimization problems. It is a heuristic search approach that is applied to various forms of optimization problems. GA is used in [46,168,169,170,171,172,173,174,175,176] to achieve optimal coordination for OCRs to solve coordination problems while minimizing operating times for the relays. The authors of [141] demonstrate solving an NLP type that depends upon GA for a coordination problem with a multiplied source in looped DNs in case of penetration of DGs, and they present a protection system dependent upon a combination of dual setting relays with traditional DOCRs. A GA has a role regarding more than one level of short-circuit current and non-standardized inverse time curves for DOCR coordination [171]. In [172], authors introduce non-dominated sorting GA-II to coordinate OCRs. To overcome the drawback of GA of occasional convergence to values that might not be optimal in ORC, [134] proposed a hybrid GA for determining the optimal PS and TDS for OCRs. In [139], a hybrid GA has been applied for determining an optimal solution for the problem of coordination taking the impacts of various network topologies into consideration.

4.2.2. Particle Swarm Optimization (PSO)

PSO is originally attributed to Kennedy and Eberhart [177]. In [178,179], the convergence of the sequence of solutions has been scrutinized for PSO. PSO is one of the optimization techniques most often used for ORC. In [180], ORC of OC relays has been achieved by utilizing the PSO algorithm. Meanwhile, the authors of [180] discussed how to keep the DOCRs coordinated in case of penetration of DGs into a network.

Convergence for a local optimum is discussed for PSO in [181], while [182] confirms that PSO requires amendments to ensure the finding of a local optimum. In [183,184], authors provided modifications of the original PSO representing the introduction of a parameter inertia weight into the original particle swarm optimizer. The approach presented in [62] proposed the Nelder–Mead (NM) simple investigation procedure as well as PSO, which is called the NM-PSO method. References [185,186] presented a PSO algorithm with modifications for solving the problem of relaying coordination. A hybrid PSO to achieve optimal coordination for OCRs is proposed in [187]; it has the benefits of solving the discrete value optimization problem as well as smoothness. The coordination for DOCRs is handled by [16] through PSO. The PSO in [161] succeeds in converging to the same optimal setting determined using the simplex method. In [166], modified adaptive PSO (MAPSO) is used for optimum coordination of OCRs as well as distance relay in networks with a series compensation scheme. An efficient variant of the PSO algorithm called time-varying acceleration coefficient (PSO-TVAC) and a hybrid PSO (HPSO) are proposed in [188,189] for solving the ORC of DOCRs.

4.2.3. Differential Evolution Algorithm (DEA)

DEA is an optimization technique that repeatedly modulates a population of candidate solutions for converging to an optimal solution. In [190], the DEA is used to solve the problem of relaying coordination that is expressed in a mixed integer nonlinear problem to optimum coordination for OCRs in DNs that have DGs. In [191], the authors introduce a modification of the DEA depending on local neighborhood search to determine the optimum settings for DOCRs in networks. In [192], the enhancement and development of the DEA are discussed to achieve coordination for DOCRs.

4.2.4. Other Recent Optimization Techniques

Many researchers have made great efforts in the field of ORC through various optimization algorithms, including the honeybee algorithm proposed in [165] which results in less TSM. In [193], ABC has superior speed compared to quasi-Newton and PSO algorithms. The authors of [194] introduced the FA and compared it with GA and LP. Meanwhile, the authors of [195] used FA to solve a constrained mono-objective optimization problem to prevent the inductive FCL from causing miscoordination of the relays. A chaotic FA is used in [196]; the chaotic FA produces better results than the conventional FA in the solution quality, convergence speed, and number of iterations taken to obtain the best solution. Modified swarm FA is employed in [197]. GA and simulated annealing are employed in [198]; the seeker algorithm is introduced in [132], while teaching–learning-based optimization is introduced in [138]. The artificial neural network approach is presented in [199] to improve the optimized DOCR time of response to a short-circuit fault within a looped DN for tackling the problem of miscoordination. The authors of [200] proposed the BSA for solving the DOCR coordination problem that was formulated as a NLP problem as well as a mixed integer NLP issue; the results proved the superiority of BSA over some optimization techniques such as DE, PSO, black hole, electromagnetism-like mechanism, biogeography-based optimizers, and harmony search. Meanwhile, enhanced BSA is implemented in [201], the Box-Muller harmony search algorithm is implemented in [202], the ABC algorithm is introduced in [153], biogeography-based optimization is presented in [203], non-dominated sorting GA-II is presented in [176], an algorithm of improving group search optimization is proposed in [204], and ant colony optimization algorithm is used in [205]. In addition, the stochastic fractal search algorithm [206] is applied to optimize relay settings. Three settings should be defined for OC relays: pickup current, time dial, and type of inverse curve. In addition, the lightning flash algorithm [207] was employed successfully to tackle such mixed-variable optimal relay optimization problems subject to a set of predefined operating and practical constraints, and many more are offered as depicted in [208,209,210,211].

5. Results and Various Scenarios

The performance of the optimizers used [211], namely gorilla troop optimizer (GTO) and WCA [206], is evaluated on two study cases for the IEEE 15-bus system which is deemed as a pure transmission network tackled many times before in the literature. The network is examined using two test cases: test case 1 considers a fixed NI curve as scenario 1; test case 2 uses many curves among the four standard IEC TCCs as depicted in Figure 4, representing scenario 2. The obtained results are compared to the latest effective optimizers and manifest the dominance of the GTO over WCA for solving this complicated optimization problem.

Figure 7 shows the single-line diagram of the IEEE 15-bus network and its system data obtained from [211].

To validate the performance of the GTO and WCA procedures, the following are considered for the study of these two test cases: (i) overloading factor ranges from 1 to 2 pu; (ii) $T_{Di,min}=0.05 \mathrm{s}$ for scenario 1 and 0.01 s for scenario 2 while $T_{Di,max}=1 \mathrm{s}$ for the two test cases; (iii) ${CTM}_{min,j}=0.2 \mathrm{s}$ and ${CTM}_{max,j}=0.4 \mathrm{s}$; and (iv) ${TCC}_{i,min}=1$ and ${TCC}_{i,max}=4$, representing the type of curves (i.e., NI, VI, EI and LI); kindly refer to (1).

Table 1. Cropped settings of the IEEE 15-bus using the GTO and WCA.

Method	GTO [211]					WCA [206]
Relay ID.	Scenario 1		Scenario 2			Scenario 1		Scenario 2
	${\mathit{I}}_{\mathit{p}}$ (A)	${\mathit{T}}_{\mathit{D}}$ (s)	${\mathit{I}}_{\mathit{p}}$ (A)	${\mathit{T}}_{\mathit{D}}$ (s)	Curve	${\mathit{I}}_{\mathit{p}}$ (A)	${\mathit{T}}_{\mathit{D}}$ (s)	${\mathit{I}}_{\mathit{p}}$ (A)	${\mathit{T}}_{\mathit{D}}$ (s)	Curve
R1	274.3358	0.0675	293.9991	0.0215	EI	196.73	0.09	196.00	0.01	LI
R2	313.9608	0.0681	307.6176	0.0244	EI	334.67	0.07	230.00	0.01	LI
R3	478.9256	0.0654	477.1756	0.0221	EI	487.50	0.07	421.85	0.05	VI
R4	440.9214	0.0663	442.4978	0.0294	EI	398.95	0.08	384.54	0.06	VI
R5	489.9721	0.0662	431.7093	0.0297	EI	350.06	0.10	345.21	0.07	VI
R6	373.4484	0.0661	374.9995	0.0264	EI	251.16	0.10	263.69	0.07	EI
R7	345.0252	0.0685	367.4998	0.0270	EI	305.22	0.08	367.50	0.04	VI
R8	556.5137	0.0668	559.5983	0.0205	EI	375.00	0.10	375.14	0.06	VI
R9	432.7932	0.0658	433.3580	0.0184	EI	382.03	0.07	419.21	0.03	VI
R10	403.1725	0.0634	403.8370	0.0199	EI	287.16	0.09	270.01	0.01	LI
R11	449.1812	0.0677	449.9982	0.0269	EI	312.94	0.10	300.00	0.07	VI
R12	483.8421	0.0666	502.4489	0.0210	EI	370.35	0.09	480.90	0.04	VI
R13	509.7451	0.0500	516.4943	0.0131	EI	345.12	0.10	346.25	0.04	VI
R14	394.2057	0.0649	323.7574	0.0338	EI	265.73	0.10	345.23	0.05	VI
R15	344.3698	0.0663	367.4998	0.0251	EI	257.47	0.09	245.01	0.07	VI
R16	253.7067	0.0639	245.1271	0.0232	EI	216.49	0.08	254.99	0.04	VI
R17	179.5477	0.0667	187.4989	0.0351	EI	125.00	0.09	125.00	0.10	EI
R18	457.3094	0.0678	487.9445	0.0204	EI	330.00	0.10	335.22	0.09	VI
R19	425.2535	0.0672	425.6398	0.0251	EI	423.69	0.07	285.00	0.07	VI
R20	620.0474	0.0641	622.3204	0.0213	EI	596.17	0.08	415.06	0.07	VI
R21	539.9664	0.0612	539.9992	0.0166	EI	538.23	0.07	371.42	0.08	VI
R22	202.2495	0.0639	201.8658	0.0244	EI	173.14	0.08	197.70	0.04	VI
R23	395.0120	0.0644	393.4877	0.0243	EI	293.56	0.09	272.73	0.07	VI
R24	251.1717	0.0653	224.4215	0.0314	EI	177.14	0.10	217.11	0.05	NI
R25	348.8178	0.0629	286.8513	0.0287	EI	289.12	0.08	235.00	0.06	EI
R26	304.9423	0.0685	307.2427	0.0383	EI	267.02	0.08	306.65	0.03	EI
R27	352.4827	0.0653	320.5883	0.0275	EI	310.89	0.08	352.50	0.03	EI
R28	447.8116	0.0667	389.5427	0.0293	EI	356.60	0.09	300.00	0.07	VI
R29	671.0048	0.0673	674.2527	0.0263	EI	631.60	0.08	615.42	0.05	VI
R30	201.8216	0.0658	192.8832	0.0310	EI	162.57	0.09	169.11	0.06	VI
R31	300.5795	0.0671	263.0405	0.0283	EI	267.59	0.09	220.01	0.07	VI
R32	278.9333	0.0661	284.9837	0.0161	EI	190.00	0.10	190.00	0.01	LI
R33	351.3740	0.0745	405.0106	0.0227	EI	368.44	0.08	295.00	0.07	VI
R34	298.8261	0.0685	280.2785	0.0308	EI	213.76	0.10	200.00	0.1	EI
R35	364.0380	0.0603	296.5395	0.0290	EI	245.00	0.10	253.12	0.06	VI
R36	433.7195	0.0676	382.2169	0.0261	EI	290.00	0.09	290.06	0.06	VI
R37	553.7038	0.0666	599.9994	0.0276	EI	400.00	0.10	400.21	0.07	VI
R38	299.4094	0.0672	248.6752	0.0331	EI	213.85	0.10	200.00	0.07	VI
R39	285.5946	0.0667	294.4203	0.0225	EI	285.79	0.08	202.22	0.07	VI
R40	560.6371	0.0670	490.9930	0.0293	EI	486.37	0.08	514.52	0.03	EI
R41	297.1019	0.0635	299.9973	0.0268	EI	201.02	0.10	200.00	0.01	LI
R42	337.1191	0.0674	374.1360	0.0164	EI	250.00	0.10	330.10	0.04	VI
TOT (s)	9.0775		1.3962			10.8904		3.4560

presents the cropped best settings of OCRs in the IEEE 15-bus electric system using the GTO for both test cases, including the total operating time (TOT) in seconds (The bold font indicates the best cropped final results). The convergence of the OF trend using the GTO for this extremely imbued DG electric system is depicted in Figure 8. On the other hand, the convergence trend of the WCA is illustrated in Figure 9.

With a closer look at Table 1, the reader may note with a change in the curve type, the GTO tends to pick up the EI curves for faster operations of the OCRs along the network under study. This is obviously proven when comparing the TOTs of the defined scenarios (i.e., 9.0775 s for the fixed NI curve and 1.3962 s for varying the relay curve).

In addition, verifying relay coordination involves several steps to ensure that the protective relays are functioning correctly and that they are operating in a coordinated manner. Here are some of the steps involved in testing and verifying relay coordination:

• Review the relay coordination settings: Before testing the relay coordination, review the settings of the protective relays to ensure that they are correctly set and coordinated.
• Create fault scenarios: Create fault scenarios based on the system configuration and the types of faults that could occur in the power system. It is important to create a wide range of fault scenarios to test the protective relays under different conditions.
• Simulate faults: Use simulation software to simulate the fault scenarios and analyze the performance of the protective relays. The simulation software can analyze the coordination of the protective relays and determine if they are operating in a coordinated manner.
• Test the protective relays: Perform primary and secondary injection testing of the protective relays to verify their settings and functionality. This involves injecting a test signal into the protective relay and verifying that it operates within its specified settings.
• Verify coordination: Verify that the protective relays are operating in a coordinated manner by analyzing their operating times and ensuring that they are operating in the correct sequence.
• Document results: Document the results of the testing and verification process and make any necessary adjustments to the protective relay settings to ensure ORC.

Overall, testing and verifying relay coordination is a critical step in ensuring the reliable and safe operation of electrical power systems. It helps to identify and correct any issues with the protective relays and ensure that they are operating in a coordinated manner to provide optimal protection to the power system.

6. Overall Discussion

ORC is essential in electrical power systems to ensure the reliable and safe operation of the system. The main objective of relay coordination is to ensure that the protective relays operate in the correct sequence during a fault to isolate the faulty section of the power system while keeping the rest of the system in operation.

Relay coordination involves the selection and setting of protective relays such that they operate in a coordinated manner to isolate the faulty section of the power system. The coordination must be such that the protective relays closest to the fault operate first, followed by the relays located further away from the fault.

There are various methods used for relay coordination, including time–current coordination, impedance coordination, and directional coordination. Time–current coordination involves the use of time–current curves to set the operating times of the protective relays. Impedance coordination involves the use of impedance relays to ensure that the protective relays operate in a coordinated manner. Directional coordination involves the use of directional relays to ensure that the protective relays operate in the correct sequence. To achieve ORC, it is important to consider various factors such as the system configuration, fault types, and the type of protective relays used. Additionally, the coordination must be periodically reviewed and updated to ensure that it remains optimal as the system evolves.

Relay coordination should be periodically reviewed and updated to ensure that it remains optimal as the power system evolves. The frequency of updates depends on various factors such as the complexity of the power system, the type and number of protective relays installed, and the rate of changes in the power system.

As a rule, relay coordination should be reviewed and updated at least once every five years. However, if there are significant changes in the power system, such as the addition of new equipment or the modification of the system configuration, the relay coordination should be reviewed and updated immediately to ensure that it remains optimal.

In addition to periodic reviews, relay coordination should also be tested and verified during commissioning and after any modifications to the power system. This helps to ensure that the protective relays are functioning correctly and that they are operating in a coordinated manner to provide reliable and safe operation of the power system.

Overall, ORC is essential in ensuring the reliable and safe operation of electrical power systems. It helps to minimize downtime, prevent damage to equipment, and ensure the safety of personnel.

7. Conclusions and Future Plans

The integration of DGs into DNs is becoming increasingly important due to the growing demand for electricity. While the integration of DGs into DNs has many benefits, including increased network efficiency, reduced power loss, decreased electricity costs, and increased reliability, it also presents potential challenges and problems for the system.

One of the main challenges of integrating DGs into DNs is the change in power flow from unidirectional to bidirectional. This can lead to issues such as voltage rise and reverse power flow, power losses, and islanding. However, the focus of this paper is on the impacts of DGs on protective relaying coordination in DNs.

The impacts of DGs on protective relaying coordination in DNs depend on their working principle, number, capacity, and location of interconnection. Miscoordination among protective devices can lead to maloperation or failure of the protective relays, which can result in damage to the equipment or even endanger the safety of personnel.

To mitigate these impacts, coordination problem formulations, coordination optimization techniques, and protection schemes have been developed. These include time–current coordination, impedance coordination, directional coordination, and adaptive protection schemes.

The research and development in ORC for DNs with DGs is still a wide subject to be discussed. Therefore, this paper is a contribution to a comprehensive understanding of relaying coordination and DG impacts on DNs. The paper highlights the need to increase the degree of nonlinearity by including more constraints to improve the coordination of protective relays in DNs with DGs.

It can be concluded that with the vast utilization of new heuristic-based optimization strategies in solving ORC problems in power systems, the results are enhanced and promising. However, still, there is room for improvement in modeling and in problem formulations to include more constraints. Among those constraints are motor starting, transformer inrush, and various equipment damage curves. Additionally, the coordination between zone 2 and zone 3 of line distance relay with overcurrent may be of interest to researchers. These issues represent new challenges for researchers to consider which will complicate the modeling and increase the degree of nonlinearity. In addition, the multi-objective formulation considering the minimization of the incident energy simultaneously with ORC processing seems to be a promising challenge for further investigations.