energies

Review

Protection Schemes of Meshed DistributionNetworks for Smart Grids and Electric Vehicles

Stavros Lazarou , Vasiliki Vita and Lambros Ekonomou *

Department of Electrical and Electronic Engineering Educators, School of Pedagogical & TechnologicalEducation (ASPETE), Heraklion Attikis, 141 21 Athens, Greece; slazarou@ta.aspete.gr (S.L.);vasvita@aspete.gr (V.V.)* Correspondence: leekonomou@aspete.gr; Tel.: +30-210-2896-927

Received: 20 October 2018; Accepted: 6 November 2018; Published: 10 November 2018�����������������

Abstract: This paper reviews protection schemes for meshed distribution networks. It gives emphasisto the increasing penetration of electric vehicles, their charging patterns, and to the increasing valueof distributed generators, especially from renewables. It includes a preliminary analysis on systemplanning with electric vehicles that is studied probabilistically and a more detailed analysis of theexpected changes introduced by these new loads. Finally, a real time hardware-in-the-loop reviewanalysis for protection systems and the open source networks available for protection studies fromseveral sources are also provided. This work could be useful as a collective review of the recentbibliography on protection for meshed networks, giving emphasis to electric vehicles and their realtime simulation.

Keywords: protection; distribution networks; smart grids; electric vehicles; meshed configuration

1. Introduction

Electric vehicles, storage devices, and distributed generation are changing the operationalcharacteristics of distribution grids across the world [1]. These new components could createunder specific operational conditions bidirectional energy flows. This affects grid planning, sincethey are optimized for one direction flows and their protection. Based on these circumstances,the established procedures of organizing distribution networks on radial configurations cannot beholistically applied [2], even if they are comparatively simple and well tested. Therefore, it is expectedthat meshed network systems will be increasingly adopted [3], offering the capability of bidirectionalflows, higher penetration of renewables, and lower expansion requirements for accommodating newdemanding loads such as electric vehicles.

Apart from the challenges created from the connection of these new components, theadvancements achieved through the improved technological capabilities of today, are able to furtherprovide protection and control enhancements. These include an extended online communicationsystem that is able to safely support distribution grid expansion and higher computational capabilitiesapplied to all system components. As such, distribution network planning has been naturally evolvingfrom radial to the more complex meshed configuration, taking advantage of these communicationand computational enhancements. The optimization methods for network planning are evolvingaccordingly [4]. The optimization patterns are increasingly calculated, on generalized level, throughwider use of supercomputers. This knowledge is transferred to the system operation level, and new,more complicated patterns are being implemented [5]. Application of the aforementioned patterns isincreasingly leading to a meshed network.

As expected, given that relevant challenges have been overcome, there are operational benefits.Several studies demonstrate the improvements of meshing the network [6,7] in terms of performance

Energies 2018, 11, 3106 ; doi:10.3390/en11113106 www.mdpi.com/journal/energies

Energies 2018, 11, 3106 2 of 17

and efficiency [8]. Distributed generation, which is mostly applied to networks compared to electricvehicles and storage [9], has already received much attention [10]. It is nowadays a main component inthe procedure of network planning [11] and being investigated on a variety of cases [12]. Nowadays,the distribution network is increasingly simulated as a whole system of lines, loads, and production,intermittent or not, that is connected to the transmission network to receive the services is not able todomestically create. It has to be mentioned that several optimization methods still consider networkradiality as a major characteristic [13], however these approaches are decreasing in time.

The main new operational challenges are connected to the electric vehicles’ load and productionfrom renewable sources. These components demonstrate substantial intermittency [14] caused by theinherited performance of primary renewable energy sources such as solar and wind, as well as by thebehavioral patterns of electric vehicles’ owners. Especially for the charging patterns of electric vehicles,it is important to investigate how it will affect the operation of the distribution system. As a matter offact, networks are already under strain due to the existing loads, and electric vehicles will increasinglyrequire more energy. Behavioral technologies that aggregate the consumption, moving it to highrenewable production or low overall consumption times, could delay expansion needs, improvingquality of life using the possible limited resources. The research community is able to capture thisphenomenon using probabilistic simulation methods. The most random probabilistic method is MonteCarlo [15]. For steady state probabilistic load flows it is the method of choice [16], but also for otherpower system probabilistic phenomena during planning [17] and/or expansion [18,19]. Except fromdistribution networks planning, of course, this method is used to a wide variety of applications [20],relevant to power systems.

Continuing this analysis, one of the reasons meshed networks were not widely used in the pastis that they demonstrate technical challenges [21] as far as their protection is concerned. Protectionpatterns need to take into consideration the special characteristics of the meshed network combinedwith its new components as they operate in a given moment. System topology, and hence protectionprocedure, change all the time based on how production is connected, which electric vehicles arecharging, and how the other loads are behaving. Given this, the operational characteristics of electricvehicles, distributed generations, and storage need to be considered. Specifically, microgeneratordynamic performance and topology altering are concerns due to renewables intermittency faultcurrent variations [22]. The above could lead to protection nonselectivity, resynchronization issuesand spurious separations or false trips [23]. The phenomenon of nonselectivity happens when largerportions of the network than necessary are disconnected to clear a given fault. In this case, exceptfrom the inconvenience of the unserved loads, it more difficult to locate the fault and consequentlyrepair it. Resynchronization issues could happen in the case of islanding operation of the system,even in a short time. Distribution generators continue instantaneously to serve the load withoutrespecting synchronization principles of voltage and frequency at the disconnected system node.Spurious separations and false trips could happen under several operational conditions of high loador production or to system branches that under specific conditions are overloaded. Several studieswere conducted to counter these challenges and protection operational limitations for the distributiongrid have been specifically defined and quantified [24]. However, all these need to be tested and safeproved under all operational conditions. This is an exercise that requires substantial computationalresources, even for a comparatively small network.

Due to the above, a concept of adaptive protection in the general framework of increasingintelligence for the distribution grid [25] has emerged. Researchers differentiate protection systemsto active, passive, and hybrid [26], based on their capability to take decisions using measureddata at other nodes than the ones they are connected to or communicated with them. A centraldecision-making center connected to each piece of protection equipment could act as the brain of thesystem. This configuration requires robust decision-making procedures, availability of calculationpower, and the capability to communicate its decisions and receive measurements and commands.Adaptive protection is a prerequisite for the dynamic update of protection settings using a network

Energies 2018, 11, 3106 3 of 17

configuration [27] that takes into consideration problem specificities such as imbalances at thedistribution level and communication constraints [28], however it can possibly lead to suboptimaloperation. An alternative could be to apply, on a distribution level, the traditional protection schemesused at the transmission level. This solution could also demonstrate some challenges [29]: Differentialprotection schemes need redundancy for communication failures. Distance protection relays may haveinaccurate reads for short distances. Voltage protection systems may trip under normal voltage dropsand they strongly depend on grid topology. Standardization of these procedures remains an issue ofparamount importance [30].

Having mentioned the above, this paper reviews recent work on the protection of distributionnetworks, giving emphasis to electromobility. It is organized as follows. Section 2 is dedicated toelectric grid expected changes due to the connection of electric vehicles. Section 3 provides informationon new protection techniques as well as on the research related to relays coordination that takesinto consideration distributed generators, transportation loads, and storage. Section 4 investigateshardware-in-the-loop real time simulation protection research, giving emphasis to the implementationsthat meet the emerging smart grids needs. Finally, Section 5 provides technical literature networks thatcan be used by researchers to attain knowledge in this specific field.

2. Electric Vehicles Connected to Smart Grids

Electric vehicles are expected to substantially affect power system planning and its operation(Figure 1) due to their extended requirements of energy, however they are able to offer enhanceddemand response capabilities in combination to the connected distributed generators [1] as well asstorage and other ancillary services. Ancillary services from electric vehicles could be offered throughscheduled charging and storage. Being parked most of the time, they may charge when electricityprices are lower. This may occur when renewables intermittent production is momentarily higher.Another option is when other loads suddenly decrease and inflexible power stations such as lignite,coal, and nuclear are not able to immediately adjust their production. Storage could also be analternative. Electric vehicles are equipped with large batteries that can retransfer the energy storedto the grid, under the procedure widely known as vehicle to grid (V2G). However, this additionalflow of energy increases electricity losses, there also exists the possibility that charging cannot beaccommodated from the existing electricity network. Henceforth, potential reinforcements have to beplanned wisely. A generic approach is proposed in this review (Figure 2), facilitating the researchers tounderstand the important emerging factors for enhancing electric vehicles’ penetration to the grid.

Energies 2018, 11, x FOR PEER REVIEW 3 of 16

distribution level and communication constraints [28], however it can possibly lead to suboptimal

operation. An alternative could be to apply, on a distribution level, the traditional protection schemes

used at the transmission level. This solution could also demonstrate some challenges [29]: Differential

protection schemes need redundancy for communication failures. Distance protection relays may

have inaccurate reads for short distances. Voltage protection systems may trip under normal voltage

drops and they strongly depend on grid topology. Standardization of these procedures remains an

issue of paramount importance [30].

Having mentioned the above, this paper reviews recent work on the protection of distribution

networks, giving emphasis to electromobility. It is organized as follows. Section 2 is dedicated to

electric grid expected changes due to the connection of electric vehicles. Section 3 provides

information on new protection techniques as well as on the research related to relays coordination

that takes into consideration distributed generators, transportation loads, and storage. Section 4

investigates hardware-in-the-loop real time simulation protection research, giving emphasis to the

implementations that meet the emerging smart grids needs. Finally, Section 5 provides technical

literature networks that can be used by researchers to attain knowledge in this specific field.

2. Electric Vehicles Connected to Smart Grids

Electric vehicles are expected to substantially affect power system planning and its operation

(Figure 1) due to their extended requirements of energy, however they are able to offer enhanced

demand response capabilities in combination to the connected distributed generators [1] as well as

storage and other ancillary services. Ancillary services from electric vehicles could be offered through

scheduled charging and storage. Being parked most of the time, they may charge when electricity

prices are lower. This may occur when renewables intermittent production is momentarily higher.

Another option is when other loads suddenly decrease and inflexible power stations such as lignite,

coal, and nuclear are not able to immediately adjust their production. Storage could also be an

alternative. Electric vehicles are equipped with large batteries that can retransfer the energy stored to

the grid, under the procedure widely known as vehicle to grid (V2G). However, this additional flow

of energy increases electricity losses, there also exists the possibility that charging cannot be

accommodated from the existing electricity network. Henceforth, potential reinforcements have to

be planned wisely. A generic approach is proposed in this review (Figure 2), facilitating the

researchers to understand the important emerging factors for enhancing electric vehicles’ penetration

to the grid.

Figure 1. Power system enhancements due to the connection of electric vehicles. Figure 1. Power system enhancements due to the connection of electric vehicles.

Energies 2018, 11, 3106 4 of 17

Energies 2018, 11, x FOR PEER REVIEW 4 of 16

Figure 2. The expected evolvement towards solving the technical challenges for increasing electric

vehicle penetration to the electricity network.

Several studies, of increasing complexity, strive to meet this challenge. Morais et al. [31]

evaluated the impact of electric vehicles on the smart grid. In this case, three distinctive operational

patterns are considered where electric vehicles charge in an uncoordinated manner, they charge in a

smart manner and/or they charge and discharge in a smart manner. Being able to charge and

discharge, it is possible to store energy and act in an aggregated manner as a provider of ancillary

services, easing bottlenecks in energy flows and improving grid’s performance and its efficiency.

Rahbari et al. [32] proposed a solution for controlled charging of plug-in electric vehicles taking into

consideration the presence of renewables intermittent generation; it is being tested on the IEEE-26

buses case. In this case, the capacity factor for the electricity grid is increasing, providing the

opportunity to provide energy when consumption is lower or energy production from the

renewables is higher. Farid et al. [33] point out the importance of a holistic approach to control all

smart grid elements. Their approach describes the grid with a 3-d structure that includes on the x-

axis production, transmission & distribution, and load; on the y-axis, the fact that system is physical

as well as cyber; and on the z-axis, the dispatchability and the system’s stochastic character. Such an

approach points out the importance of interoperability between stakeholders as well as equipment.

Zakariazadeh et al. [34] proposed a method for electric vehicles to be aggregated to a virtual power

plant. In this way, the offering of ancillary services to the grid becomes better organized to the current

market operation and the capacity factor could be increased. Bhattarai et al. [35] developed an

optimized aggregation method for organizing the distribution network in areas that take into

consideration distribution of loads and costs for optimal participation to the market. Knezović et al.

[36] explained the barriers of electric vehicles’ wider integration to the system and how these can be

lifted. The barriers are not only infrastructure- and technology-related but also regulatory- and

market-oriented. As a matter of fact, as long as technology improves and research advances, technical

challenges can be overcome and only cultural- and market-oriented barriers remain. The authors

propose a roadmap of four stages to overcome these barriers. Abdelsamad et al. [37] proposed an

innovative method for designing the secondary distribution network in a manner to be better able to

accommodate plug-in electric vehicles. They take into consideration system overall costs and

transformers loss of life. The connection of electric vehicles, due to their ability to behave, are storage

devices for energy provide the opportunity to the distribution system operators to delay lines’

expansion costs improving system performance in total. Sadeghi and Kalantar [38] propose a

planning method taking into consideration uncertainties due to distributed generators and economic

conditions. As mentioned before, intermittency is inhered to renewables and electric vehicles, but it

Figure 2. The expected evolvement towards solving the technical challenges for increasing electricvehicle penetration to the electricity network.

Several studies, of increasing complexity, strive to meet this challenge. Morais et al. [31] evaluatedthe impact of electric vehicles on the smart grid. In this case, three distinctive operational patterns areconsidered where electric vehicles charge in an uncoordinated manner, they charge in a smart mannerand/or they charge and discharge in a smart manner. Being able to charge and discharge, it is possibleto store energy and act in an aggregated manner as a provider of ancillary services, easing bottlenecksin energy flows and improving grid’s performance and its efficiency. Rahbari et al. [32] proposed asolution for controlled charging of plug-in electric vehicles taking into consideration the presenceof renewables intermittent generation; it is being tested on the IEEE-26 buses case. In this case, thecapacity factor for the electricity grid is increasing, providing the opportunity to provide energy whenconsumption is lower or energy production from the renewables is higher. Farid et al. [33] point out theimportance of a holistic approach to control all smart grid elements. Their approach describes the gridwith a 3-d structure that includes on the x-axis production, transmission & distribution, and load; onthe y-axis, the fact that system is physical as well as cyber; and on the z-axis, the dispatchability and thesystem’s stochastic character. Such an approach points out the importance of interoperability betweenstakeholders as well as equipment. Zakariazadeh et al. [34] proposed a method for electric vehicles tobe aggregated to a virtual power plant. In this way, the offering of ancillary services to the grid becomesbetter organized to the current market operation and the capacity factor could be increased. Bhattaraiet al. [35] developed an optimized aggregation method for organizing the distribution network in areasthat take into consideration distribution of loads and costs for optimal participation to the market.Knezovic et al. [36] explained the barriers of electric vehicles’ wider integration to the system and howthese can be lifted. The barriers are not only infrastructure- and technology-related but also regulatory-and market-oriented. As a matter of fact, as long as technology improves and research advances,technical challenges can be overcome and only cultural- and market-oriented barriers remain. Theauthors propose a roadmap of four stages to overcome these barriers. Abdelsamad et al. [37] proposedan innovative method for designing the secondary distribution network in a manner to be betterable to accommodate plug-in electric vehicles. They take into consideration system overall costsand transformers loss of life. The connection of electric vehicles, due to their ability to behave, arestorage devices for energy provide the opportunity to the distribution system operators to delay lines’expansion costs improving system performance in total. Sadeghi and Kalantar [38] propose a planningmethod taking into consideration uncertainties due to distributed generators and economic conditions.As mentioned before, intermittency is inhered to renewables and electric vehicles, but it can be alsoobserved in the behavior of other stakeholders such as market participants. These uncertainties are

Energies 2018, 11, 3106 5 of 17

needed also to be studied in order to achieve adequate understanding of the whole energetic systems.Sun et al. [39] characterized plug-in electric vehicles demand based on battery state of charge andtemporal travel purposes. They use a stochastic approach and they are able to demonstrate benefits tonetwork planning. Arias et al. [40] predicted, in a stochastic optimal manner electric vehicle connectedto the distribution network, charging behavior. Wu and Sioshansi [41] provided an optimizationmodel, which is able to schedule electric vehicle charging in a manner that relieves network constraints.Hu et al. [42] proposed an electric vehicle coordinated charging procedure for alleviating smart gridoperational congestions. On a similar pattern, Beaude et al. [43] sought to reduce charging effects of thedistribution network. Xiang et al. [44] develop optimal management procedures for aggregated plug-inelectric vehicles charging to active distribution networks. Sundström and Binding [45] analyzed electricvehicle charging constraints to the distribution network on voltage and power level and how they canbe avoided.

Wang et al. [46] also provided decentralized solutions for electric vehicles and distributedgenerators from renewable sources. Shuai et al. [47] approached electric vehicle charging from theeconomic point of view. Peng et al. [48] reviewed issues of active power management of power gridusing electric vehicles. Pavic et al. [49] increased system flexibility using electric vehicles. This is animportant trait that is transferred to the transmission system. Lazarou et al. [50] studied the connectionof electric vehicles giving priority to the grid. Aggregated cohorts of electric vehicles are able tooffer expensive flexibility services to system that otherwise shall be covered by gas or hydro, whereavailable, power plants. This being said, electric vehicles as such is possible to contribute more activelyin the mitigation of climate change and the decarbonization of power supply in total. Mojdehi andGhosh [51] demonstrate a procedure for compensating electric vehicles as ancillary service providers.Daina et al. [52] explains the potential implications of electric vehicle charging per se. Weiller et al. [53]approached this subject from an industrial point of view. Schmidt et al. [54] researched electricvehicles from the transportation demand angle. Sato et al. [55] analyzed issues of standardization inelectric vehicle charging, because all procedures have to be accepted by the complete group of marketparticipants, which is possible to be achieved efficiently through global standardization procedures.All the above are complex exercises that require substantial computational power.

The probabilistic approaches for solving electric vehicles and distributed generation challengesare computationally intensive. These is due to the complexity of the problem that needs to involve thebehavioral characteristics of vehicle users, the constraints of the operation, and capacity of the grid, aswell as, the intermittency of the connected production from renewable sources. Adding to the above,the increasing use of storage needs to be considered. High performance computer use has proliferatedto the degree that they are able to be widely used, even for studies to smaller networks. Green et al. [56]approached the problem of smart grids with electric vehicles’ computational requirements usingindustrial solutions such as high-performance computers (HPC). They take into consideration thecomplexities imposed by the combination of several layers of probabilistic phenomena happeningat the same time to a multilayered smart grid. Procopiou et al. [57] used HPC for simulating electricvehicles charging behavior at the secondary distribution network level. Their findings, based on areal system, demonstrate the future requirements of computational resources at smart grids, evenon secondary level. This is an observation of paramount importance, since millions of kilometers ofsecondary network are currently operating, and the extensive computational resources required tooptimize it could potential become an important factor of its operational cost. A generic description isprovided to Table 1.

Energies 2018, 11, 3106 6 of 17

Table 1. Distribution networks protection studies.

Reference Description Research Method Contribution

[1,14,31,36]

Explanation of thebarriers and the benefitsfor increasing thepenetration of electricvehicles, evaluation ofthe impact

Analysis, scenarios’development

Qualifying andquantifying theimportance of electricvehicle barriers for theirconnection to the grid

[15–20,33,37–50]Network planning forincreasing penetration ofelectric vehicles

Power flow analysis,probabilistic andoptimization methods(Monte Carlo), novelapproaches for depictingthe electricity network

Proposals for advancednetwork planning toaccommodate increasingconnection of electricvehicles

[32,34,35,51–55]Electric vehicles chargingthrough aggregators andstandardization

Analysis, marketprinciples, optimizationmethods, Virtual PowerPlants, scenarios’analysis

Principles fordevelopment ofaggregators for EVscharging,standardization

[56,57]

High-performancecomputing fordistribution grid andelectric vehiclessimulation

Probabilistic methods Simulation of extendedoperational scenarios

3. Protection of Smart Grids

The protection schemes currently used at the distribution level are distinguished in active, passive,and hybrid systems [23]. Several approaches have been developed with the most widely usedto be current limiters, centralized protection, distance protection, protection based on variables,differential protection, multi-agent protection schemes, and others, all based on relays (Figure 3).Current limiters are usually semiconducting devices that decrease the maximum current during faults.This is paramount importance, since the equipment, especially at distribution level, is scaled to sustainmaximum currents. Circuit breakers are also not able to interrupt too large current fault. On theother hand, distributed generators, especially the rotating ones (with synchronous or asynchronousgenerators connected), as well as rotating machines could increase fault currents during to theirinherited inertia. Distant protection is based on the fact that electrical characteristics change if afault occurs on a line. Its resistance, impedance, and capacitance change, and, instantaneously, thedistance protection measures the difference and activates the circuit breakers required to clear thefault. Differential protection can be applied to twin lines with the same characteristic and loadingor/and at the start and the end of a given line. When a fault occurs to one of the twin lines andmeasurements change, it is clear that protection shall be activated. Similarly, when a measurementat the beginning is different to the end, after considering line’s electrical characteristics, then theprotection system shall be activated. Other types of protection are more sophisticated and increasinglyused to distribution networks.

Energies 2018, 11, 3106 7 of 17

Energies 2018, 11, x FOR PEER REVIEW 6 of 16

Table 1. Distribution networks protection studies.

Reference Description Research Method Contribution

[1,14,31,36]

Explanation of the barriers and

the benefits for increasing the

penetration of electric vehicles,

evaluation of the impact

Analysis, scenarios’

development

Qualifying and quantifying

the importance of electric

vehicle barriers for their

connection to the grid

[15–20,33,37–50]

Network planning for

increasing penetration of

electric vehicles

Power flow analysis,

probabilistic and optimization

methods (Monte Carlo), novel

approaches for depicting the

electricity network

Proposals for advanced

network planning to

accommodate increasing

connection of electric

vehicles

[32,34,35,51–55]

Electric vehicles charging

through aggregators and

standardization

Analysis, market principles,

optimization methods, Virtual

Power Plants, scenarios’ analysis

Principles for development

of aggregators for EVs

charging, standardization

[56,57]

High-performance computing

for distribution grid and electric

vehicles simulation

Probabilistic methods Simulation of extended

operational scenarios

3. Protection of Smart Grids

The protection schemes currently used at the distribution level are distinguished in active,

passive, and hybrid systems [23]. Several approaches have been developed with the most widely

used to be current limiters, centralized protection, distance protection, protection based on variables,

differential protection, multi-agent protection schemes, and others, all based on relays (Figure 3).

Current limiters are usually semiconducting devices that decrease the maximum current during

faults. This is paramount importance, since the equipment, especially at distribution level, is scaled

to sustain maximum currents. Circuit breakers are also not able to interrupt too large current fault.

On the other hand, distributed generators, especially the rotating ones (with synchronous or

asynchronous generators connected), as well as rotating machines could increase fault currents

during to their inherited inertia. Distant protection is based on the fact that electrical characteristics

change if a fault occurs on a line. Its resistance, impedance, and capacitance change, and,

instantaneously, the distance protection measures the difference and activates the circuit breakers

required to clear the fault. Differential protection can be applied to twin lines with the same

characteristic and loading or/and at the start and the end of a given line. When a fault occurs to one

of the twin lines and measurements change, it is clear that protection shall be activated. Similarly,

when a measurement at the beginning is different to the end, after considering line’s electrical

characteristics, then the protection system shall be activated. Other types of protection are more

sophisticated and increasingly used to distribution networks.

Figure 3. Protection schemes (rewritten from [23]).

At this point it is of paramount importance to coordinate relays’ operation. To achieve this goal

except from the location of the fault, the flow of power has to be known. As an example, at Figure 4

Figure 3. Protection schemes (rewritten from [23]).

At this point it is of paramount importance to coordinate relays’ operation. To achieve thisgoal except from the location of the fault, the flow of power has to be known. As an example, atFigure 4 the same relay RBC is able to interrupt fault F1 in the first case and F2 in the second. Thevice versa faults are to be interrupted only by different relays not depicted at this figure. For mesheddistribution networks this procedure is becoming more complex, however researchers were able todevelop methods with adequate results [58]. Singh et al. [59] propose an adaptive method for relaycoordination based on fuzzy logic. Advanced control techniques could potentially widely used forprotection purposes. This is due to their capability to provide results without the complete descriptionof the system under investigation but only based to previous measurements. However, this approachbears the risks of spurious operation or protection malfunctions.

Ates et al. [60] modify relays’ parametrization to achieve selectivity. Protection selectivity couldbe achieved through adjusting the parametrization of the existing equipment. This has the benefitof lower initial resources deployment, but its application may be limited. Similarly, Chen et al. [61]coordinate relays’ operation to a meshed network with distributed generation. Chabanloo et al. [62]use fault current limiters to achieve protection coverage, having also connected distributed generatorsat an urban distribution grid. Jalilian et al. [63] propose a method based on post-fault current.Costa et al. [64] use a metaheuristic method for achieving coordination. Saleh et al. [65] and Hussainet al. [66] provide the relevant optimization procedures. Corrêa et al. [67] propose an online methodfor coordinating the protection relays. Ojaghi and Ghahremani [68] developed new characteristics forovercurrent relays to cover selectivity issues.

Energies 2018, 11, x FOR PEER REVIEW 7 of 16

the same relay RBC is able to interrupt fault F1 in the first case and F2 in the second. The vice versa

faults are to be interrupted only by different relays not depicted at this figure. For meshed

distribution networks this procedure is becoming more complex, however researchers were able to

develop methods with adequate results [58]. Singh et al. [59] propose an adaptive method for relay

coordination based on fuzzy logic. Advanced control techniques could potentially widely used for

protection purposes. This is due to their capability to provide results without the complete

description of the system under investigation but only based to previous measurements. However,

this approach bears the risks of spurious operation or protection malfunctions.

Ates et al. [60] modify relays’ parametrization to achieve selectivity. Protection selectivity could

be achieved through adjusting the parametrization of the existing equipment. This has the benefit of

lower initial resources deployment, but its application may be limited. Similarly, Chen et al. [61]

coordinate relays’ operation to a meshed network with distributed generation. Chabanloo et al. [62]

use fault current limiters to achieve protection coverage, having also connected distributed

generators at an urban distribution grid. Jalilian et al. [63] propose a method based on post-fault

current. Costa et al. [64] use a metaheuristic method for achieving coordination. Saleh et al. [65] and

Hussain et al. [66] provide the relevant optimization procedures. Corrêa et al. [67] propose an online

method for coordinating the protection relays. Ojaghi and Ghahremani [68] developed new

characteristics for overcurrent relays to cover selectivity issues.

Figure 4. Relay direction vs. power flow direction [63].

Scientists conduct research on several additional issues related to protection of smart grids.

Khorramdel et al. [69] and Khakimzyanov et al. [70] contribute in locating the fault. Locating the exact

location of the fault could the basis for the design of new protection systems, or simply activating the

nearest circuit breakers to clear it. Also, being aware of the exact location of the fault facilitates

potential system repairs if needed. Esmaeili et al. [71] provide a procedure for limiting fault current.

Jamali and Borhani-Bahabadi [72] propose a new protection method for radial and meshed networks,

which have distributed generators connecting, without using communication capabilities. Bukhari et

al. [73] use fuzzy logic for protection purposes. He et al. [74] propose a new method for smart grid

protection purposes. Rahman et al. [75] use multi-agent systems for fault diagnosis at the network

distribution feeder and its reconfiguration. Pathirana et al. [76] developed a new methodology for

active protection systems. Todorović et al. [77] propose a method for converter operation that respects

short circuit levels. Zhang et al. [78] propose a method for protection of closed loop, meshed

networks. Finally, Li et al. [79] and Fazio et al. [80] investigate anti-islanding techniques and

procedures.

Anti-islanding requires that the distribution grid is always connected to the main network. This

increases operational safety, since all distributed generators require to disconnect if main grid is not

available due to any reasoning. On the other, this request decreases the potential for continuously

serving the load, even if the main grid is not available but the connected distributed generators and

storage are able to cover it. Future efforts of this nature will keep production online as much as

possible and additionally take care of resynchronization and addressing issues of quality. A generic

description is provided to Table 2.

Figure 4. Relay direction vs. power flow direction [63].

Scientists conduct research on several additional issues related to protection of smart grids.Khorramdel et al. [69] and Khakimzyanov et al. [70] contribute in locating the fault. Locating the exactlocation of the fault could the basis for the design of new protection systems, or simply activatingthe nearest circuit breakers to clear it. Also, being aware of the exact location of the fault facilitates

Energies 2018, 11, 3106 8 of 17

potential system repairs if needed. Esmaeili et al. [71] provide a procedure for limiting fault current.Jamali and Borhani-Bahabadi [72] propose a new protection method for radial and meshed networks,which have distributed generators connecting, without using communication capabilities. Bukhari etal. [73] use fuzzy logic for protection purposes. He et al. [74] propose a new method for smart gridprotection purposes. Rahman et al. [75] use multi-agent systems for fault diagnosis at the networkdistribution feeder and its reconfiguration. Pathirana et al. [76] developed a new methodology foractive protection systems. Todorovic et al. [77] propose a method for converter operation that respectsshort circuit levels. Zhang et al. [78] propose a method for protection of closed loop, meshed networks.Finally, Li et al. [79] and Fazio et al. [80] investigate anti-islanding techniques and procedures.

Anti-islanding requires that the distribution grid is always connected to the main network. Thisincreases operational safety, since all distributed generators require to disconnect if main grid is notavailable due to any reasoning. On the other, this request decreases the potential for continuouslyserving the load, even if the main grid is not available but the connected distributed generators andstorage are able to cover it. Future efforts of this nature will keep production online as much as possibleand additionally take care of resynchronization and addressing issues of quality. A generic descriptionis provided to Table 2.

Table 2. Distribution networks protection studies.

Reference Description Research Method Contribution

[2–13] Meshed networks in thebibliography

Analysis of thestate-of-the-art situation

Review of thestate-of-the-art on thesubject

[21–30] Protection types andtheir characteristics

Analysis of thestate-of-the-art situation

Description of active,passive and hybridsystems, differential,distance, overcurrentand voltage protection

[58–61,73–75]

Protection methods forcovering meshednetworks, intermittentrenewables and electricvehicles

Fuzzy logic, multi-agent,traditional protectionprocedures

Applying alternativemethods for protection

[62,63,71,77]

Protection procedures forcovering meshednetworks, intermittentrenewables and electricvehicles

Fault current limiters,converter operation

Applying alternativemethods for protection

[64–68,76,78,79]

Protection procedures forcovering meshednetworks, intermittentrenewables and electricvehicles

Metaheuristic methodsand relevantoptimization,overcurrent relays

Applying alternativemethods for protection

[69,70]

Protection procedures forcovering meshednetworks, intermittentrenewables and electricvehicles

Fault locationEnhancing theunderstanding of theoperation of the system

4. Real-Time Simulation Capabilities for Smart Grids Protection

The validation approaches for understanding the exact behavior of smart grids systems, includingwhen it comes to protection operations, are software simulation, laboratory testing, hardware inthe loop, and measurements in the field [81]. Testing the exact behavior of the protection system,taking into consideration the specificities of protection equipment operation, could be a demanding

Energies 2018, 11, 3106 9 of 17

procedure. Traditional digital simulation software could offer good insight into the real operation ofthe system; however, it does not have the capability to calculate hardware limitations that need tobe considered for the safe operation of the system. On the other hand, connecting the real protectionequipment to test it directly to the real system could be very demanding in terms of resources andit could be impossible under certain circumstances. Real time, hardware on the loop simulators areable to combine the benefits of software simulation and hardware testing. The simulator performsthe necessary calculations and then feeds them to its analog inputs and outputs. When the protectionrelays are connected hardware can be tested without the need of the real electricity network.

From the above it appears that hardware in the loop is the prominent and most effective solutionfor developing modern procedures. Real time simulation combines effectively the capabilities ofsimulation software and is able to approach the complexity and specificity of new equipment havinghardware interfaces at signal and network level. Hardware in the loop finds applications on differentaspects, except from protection of power systems but for the specific needs of smart grids it caneffectively perform electromagnetic transients and phasor and voltage control simulations [82]. Apartfrom the above, real time simulators are especially useful for simulating microgrids, they are composedof a computing part which has been uploaded with the simulation model of the equipment notavailable in the laboratory and the proper interfaces to the physical system [83], as depicted to Figure 5.The physical problem is simulated at the processing unit in real time. As digital technology necessitates,the system splits the time in short intervals and calculates problem’s solutions for the next time stepusing the information imported from the previous. Based on this, processing power should be enoughto be able to perform all necessary calculations in time less than it needs the simulated real system.After this, the values are converted to analog signals and if necessary, they are amplified. The output isconnected directly to the equipment to be tested. The behavior of the equipment is measured, and dataare reconverted to digital signals and fed to the computational module for the next simulation step. Inthis manner, the equipment can be tested without having it connected to the real system, facilitatingthe testing procedure in a safe manner. There are several types of real time simulators able to cover allthe needs of smart grids simulations. However, they have their own limitations due to the models andtheir capabilities [84]. On the smart grid level, the work of several researchers has shown very goodperformance of the system [85].

Energies 2018, 11, x FOR PEER REVIEW 9 of 16

have their own limitations due to the models and their capabilities [84]. On the smart grid level, the

work of several researchers has shown very good performance of the system [85].

As far as hardware is concerned, target computers are categorized in two major types [83]. Type

A [86] real time simulators have custom built processors, whereas type B [87] work with general

purpose CPUs. As far as software is concerned, it is developed by the vendor, or it is based on

proprietary software such as MATLAB/Simulink [88]. The target computers’ operating system could

be Windows or Linux or an alternative operating system for special purposes. The solution

methodology is based to the principle that the simulation is done into steps, of which the results are

provided to the Input/Output Interface in such a way that represents the behavior of a real system.

There are limitations are based on the complexity of the model, the processing power, and the time

step. A time step combined with inadequate computational power and a complex model could lead

to simulation instability and inaccurate results, which is the main challenge of the system as a whole.

Having mentioned the above the main vendors are RTDS [86], OPAL-rt [87], MATLAB xPC target

[89], dspace [90], Applied Dynamics International [91], and Typhoon RTDS [92].

Real time simulators of all types able to perform hardware-in-the-loop have similar general

setting. The simulation model is developed on a host personal computer, which serves as the interface

for extracting the simulation results. The compiled model is uploaded on the real-time simulator

target computer. The results are channeled to analog and digital inputs and outputs, being there to

serve as the interface of the simulator to the equipment under study. Complex simulation systems

could also include fast communication networks to operate in parallel with more than one target

computers being remotely located.

The main open source solution to real time simulation is the virtual test bed [93,94]. However,

real time simulation may cover only a part of the requirements for a cyber-physical energy system,

which in fact shall include additional intelligence.

Figure 5. Basic Hardware in the loop simulation concept [83].

As previously mentioned above, real time, hardware-in-loop capable simulators have been a

useful tool used for protection studies. Using these systems, researchers and product development

engineers are able to predict the operational behavior of relays and protection systems in general.

Several laboratories have developed advanced systems and they open them to researchers across the

world for academic purposes [95]. Smaller portable devices are available for restricted simulations

[96]. All these hardware-in-the-loop simulators are optimized for protection purposes [97], for phasor

measurement system [98] testing, to replicate microgrids’ dynamic performance [99] and for other

purposes. Specific examples for protection systems configuration are provided in past works

[100,101].

Another important issue is the islanding behavior of distributed generators and storage devices

at the distribution level [102–104]. To avoid incorrect operation of the grid and maintain safety

precautions, it is not allowed under current standardization for distribution generations to operate if

Figure 5. Basic Hardware in the loop simulation concept [83].

As far as hardware is concerned, target computers are categorized in two major types [83]. TypeA [86] real time simulators have custom built processors, whereas type B [87] work with generalpurpose CPUs. As far as software is concerned, it is developed by the vendor, or it is based onproprietary software such as MATLAB/Simulink [88]. The target computers’ operating system could beWindows or Linux or an alternative operating system for special purposes. The solution methodologyis based to the principle that the simulation is done into steps, of which the results are providedto the Input/Output Interface in such a way that represents the behavior of a real system. There

Energies 2018, 11, 3106 10 of 17

are limitations are based on the complexity of the model, the processing power, and the time step.A time step combined with inadequate computational power and a complex model could lead tosimulation instability and inaccurate results, which is the main challenge of the system as a whole.Having mentioned the above the main vendors are RTDS [86], OPAL-rt [87], MATLAB xPC target [89],dspace [90], Applied Dynamics International [91], and Typhoon RTDS [92].

Real time simulators of all types able to perform hardware-in-the-loop have similar general setting.The simulation model is developed on a host personal computer, which serves as the interface forextracting the simulation results. The compiled model is uploaded on the real-time simulator targetcomputer. The results are channeled to analog and digital inputs and outputs, being there to serveas the interface of the simulator to the equipment under study. Complex simulation systems couldalso include fast communication networks to operate in parallel with more than one target computersbeing remotely located.

The main open source solution to real time simulation is the virtual test bed [93,94]. However,real time simulation may cover only a part of the requirements for a cyber-physical energy system,which in fact shall include additional intelligence.

As previously mentioned above, real time, hardware-in-loop capable simulators have been auseful tool used for protection studies. Using these systems, researchers and product developmentengineers are able to predict the operational behavior of relays and protection systems in general.Several laboratories have developed advanced systems and they open them to researchers across theworld for academic purposes [95]. Smaller portable devices are available for restricted simulations [96].All these hardware-in-the-loop simulators are optimized for protection purposes [97], for phasormeasurement system [98] testing, to replicate microgrids’ dynamic performance [99] and for otherpurposes. Specific examples for protection systems configuration are provided in past works [100,101].

Another important issue is the islanding behavior of distributed generators and storage devicesat the distribution level [102–104]. To avoid incorrect operation of the grid and maintain safetyprecautions, it is not allowed under current standardization for distribution generations to operate ifcentral connection is not available. According to current procedures, this part of the system requires tobe de-energized for safety purposes. The devices and procedures for anti-islanding protection are alsotested in real time simulators [105,106] and in the future could be tested on microgrids, it is expectedthat the islanding operation of the distribution network could be a part of de facto reality. Of course, inislanding operation the distributed generators and the available storage devices have to able to coverthe load and operate safely. A generic description is provided to Table 3.

Table 3. Distribution networks protection studies.

Reference Description Research Method Contribution

[81–85]Real-time simulatorstypes, applications andevolvement

Bibliographical review Technology and itsadvancements review

[79,80,95–101]Real-time simulatorapplications for networkprotection

Hardware-in-the-loopsimulations

Equipment testing anddevelopment fornetworks’ protection

[102–104]Real-time simulatorapplications forislanding protection

Hardware-in-the-loopsimulations

Equipment testing anddevelopment forislanding protection

[86–94]

Real Time DigitalSimulatorsmanufacturers and otherdevelopers

N/A

Availability ofspecialized solutions foreach application,including smart grids’protection

Energies 2018, 11, 3106 11 of 17

5. Open Access Distribution Networks Available for Protection Studies

Open access distribution networks are of paramount importance as case studies for applyingelectricity system research. They are used to test case new procedures and methodologies. For anewcomer on the field, without access to new networks, bibliography offers openly solutions on avariety of voltage levels, complexities, and for several applications. Distribution networks, which arethe main focus of this review, are mostly provided by the major engineering such as IEEE [107,108] orother organizations [109]. Renowned institutions in Europe such as the Joint Research Centre [110]and in the United States the Pacific Northwest National Laboratory [111] have also published theirown models for research purposes. Individual laboratories also propose their own configurations. Forexample, a distribution line is provided at [112] and a model for real time is available at [113].

Studies for smart grids protection systems are usually at the distribution voltage level, which is atthe level of kV up to 35 kV. However, the bibliography has smart grids examples on higher voltagelevel [114]. In this case, the authors use a selection of IEEE high voltage test cases, for example IEEE14 Bus [115] and IEEE 30 Bus [114] that are part of the American Electric Power System, IEEE 17Generator dynamic test case [116], and IEEE 30 Bus dynamic test case [117], which is believed thatis part of the New England power system. Except from the above, individual researchers use linesof their adjacent network to conduct their research. An example of protection studies distributionnetwork in this category is available at [118].

There are journals able to publish only datasets such as Data in Brief from Elsevier [119] andData from MDPI [120], where the researchers are able to find new datasets on every subject includingsimulation networks. Using the same simulation networks in different cases by different researchersenhances the capability of the research community to compare new methodologies. On the otherhand, applications of some methodologies to specific local networks are important to enhance theapplicability of the research and its engineering importance. A generic description is provided toTable 4.

Table 4. A generic description of distribution networks availability for protection studies.

Reference Description Contribution

[113–116] IEEE networks Distribution grids for researchpurposes

[108–111] Laboratories’ networks Distribution grids for researchpurposes

[118,119] Journals for publishing datasets Availability of datasets, includingnetworks for research purposes

6. Conclusions and Future Developments

This paper reviewed protection systems as they evolve due to the expected connection of electricvehicles. The first part of this paper includes a description of network planning in the presence ofelectric vehicles and distributed generators, as well as it incorporates recent relevant research. Thesecond part of this review analyses the availability of real time hardware in the loop capabilities totest protection systems and the open access distribution networks available for studies in the technicalliterature. With the knowledge collected in this manuscript, a researcher is able to understand therecent efforts of the connection of electric vehicles, state-of-art tools, and networks to be used forprotection studies.

Based on the above, the future electricity grid is expected to become the main carrier fortransferring energy to the final users. It will cover, except from the traditional loads, new requirementsof transportation in a wider scale with higher penetration of electric vehicles. However, this emergingsituation will create technical challenges to the existing infrastructure that needs to effectivelyaccommodate these new loads. According to the authors’ perspective, except from reinforcing the

Energies 2018, 11, 3106 12 of 17

existing grid, it is of paramount importance to enhance grid’s performance, applying in a wider scalethe meshed distribution network. Protection issues will eventually arise that could be solved usingreal time simulators. Research community could contribute in this direction using the available opensource real networks.

Author Contributions: L.E. provided the supervision and reviewed the paper. V.V. and S.L. conducted theresearch and wrote the paper.

Funding: The authors acknowledge financial support for this work from the Special Account for Research ofASPETE, under the project “DECA”, through the funding program “Strengthening research of ASPETE facultymembers”.

Acknowledgments: The authors appreciate the productive comments they received from the reviewers of theEnergies journal and those who read this manuscript as a preprint on researchgate.com and provided theirfeedback. Without the support from Evangelos Kotsakis and Heinz Wilkening, this work would have beenimpossible to be produced.

Conflicts of Interest: The authors declare no conflicts of interest.

References

1. Fang, X.; Misra, S.; Xue, G.; Yang, D. Smart Grid—The New and Improved Power Grid: A Survey.IEEE Commun. Surv. Tutor. 2012, 14, 944–980. [CrossRef]

2. Zubo, R.; Mokryani, G.; Rajamani, H.-S.; Aghaei, J.; Niknam, T.; Pillai, P. Operation and planning ofdistribution networks with integration of renewable distributed generators considering uncertainties: Areview. Renew. Sustain. Energy Rev. 2017, 72, 1177–1198. [CrossRef]

3. Dumbrava, V.; Miclescu, T.; Lazaroiu, G.C. Power Distribution Networks Planning Optimization in SmartCities. In City Networks; Springer Optimization and Its Applications; Springer: Cham, Switzerland, 2017;Volume 128, pp. 213–226.

4. Georgilakis, P.S.; Hatziargyriou, N.D. A review of power distribution planning in the modern power systemsera: Models, methods and future research. Electr. Power Syst. Res. 2015, 121, 89–100. [CrossRef]

5. Sedghi, M.; Ahmadian, A.; Aliakbar-Golkar, M. Assessment of optimization algorithms capability indistribution network planning: Review, comparison and modification techniques. Renew. Sustain. EnergyRev. 2016, 66, 415–434. [CrossRef]

6. Alvarez-Herault, M.-C.; N’Doye, N.; Gandioli, C.; Hadjsaid, N.; Tixador, P. Meshed distribution network vs.reinforcement to increase the distributed generation connection. Sustain. Energy Grids Netw. 2015, 1, 20–27.[CrossRef]

7. Novoselnik, B.; Bolfek, M.; Boškovic, M.; Baotic, M. Electrical Power Distribution System Reconfiguration:Case Study of a Real-life Grid in Croatia. IFAC-PapersOnLine 2017, 50, 61–66. [CrossRef]

8. Ehsan, A.; Yang, Q. Optimal integration and planning of renewable distributed generation in the powerdistribution networks: A review of analytical techniques. Appl. Energy 2018, 210, 44–59. [CrossRef]

9. Vita, V. Development of a decision-making algorithm for the optimum size and placement of distributedgeneration units in distribution networks. Energies 2017, 10, 1433. [CrossRef]

10. Singh, B.; Mishra, D.K. A survey on enhancement of power system performances by optimally placed DG indistribution networks. Energy Rep. 2018, 4, 129–158. [CrossRef]

11. Singh, B.; Sharma, J. A review on distributed generation planning. Renew. Sustain. Energy Rev. 2017, 76,529–544. [CrossRef]

12. Hadjiionas, S.; Oikonomou, D.; Fotis, G.; Vita, V.; Ekonomou, L.; Pavlatos, C. Green field planning ofdistribution systems. In Proceedings of the 11th WSEAS International Conference on Automatic Control,Modelling & Simulation (ACMOS’09), Istanbul, Turkey, 30 May 30–1 June 2009.

13. Jordehi, A.R. Optimisation of electric distribution systems: A review. Renew. Sustain. Energy Rev. 2015, 51,1088–1100. [CrossRef]

14. Banerjee, B.; Jayaweera, D.; Islam, S. Modelling and Simulation of Power Systems. In Smart Power Systemsand Renewable Energy System Integration; Springer Nature: Basingstoke, UK, 2016; pp. 15–28.

15. Kalos, M.H.; Whitlock, P.A. Monte Carlo Methods; Wiley-VCH: Weinheim, Germany, 2008.

Energies 2018, 11, 3106 13 of 17

16. Prusty, B.R.; Jena, D. A critical review on probabilistic load flow studies in uncertainty constrained powersystems with photovoltaic generation and a new approach. Renew. Sustain. Energy Rev. 2017, 69, 1286–1302.[CrossRef]

17. Mokryani, G. Active distribution networks planning with integration of demand response. Sol. Energy 2015,122, 1362–1370. [CrossRef]

18. Hemmati, R.; Hooshmand, R.-A.; Taheri, N. Distribution network expansion planning and DG placement inthe presence of uncertainties. Electr. Power Energy Syst. 2015, 73, 665–673. [CrossRef]

19. Piao, M.; Li, Y.; Huang, G. Development of a stochastic simulation–optimization model for planning electricpower systems—A case study of Shanghai, China. Energy Convers. Manag. 2014, 86, 111–124. [CrossRef]

20. Marmidis, G.; Lazarou, S.; Pyrgioti, E. Optimal placement of wind turbines in a wind park using MonteCarlo simulation. Renew. Energy 2008, 33, 1455–1460. [CrossRef]

21. Muruganantham, B.; Gnanadass, R.; Padhy, N. Challenges with renewable energy sources and storage inpractical distribution systems. Renew. Sustain. Energy Rev. 2017, 73, 125–134. [CrossRef]

22. Brearley, B.J.; Prabu, R.R. A review on issues and approaches for micro grid protection. Renew. Sustain.Energy Rev. 2017, 67, 988–997. [CrossRef]

23. Memon, A.A.; Kauhaniemi, K. A critical review of AC Microgrid protection issues and available solutions.Electr. Power Syst. Res. 2015, 129, 23–31. [CrossRef]

24. Abdel-Ghany, H.A.; Azmy, A.M. Defining the practical constraints of inserting DG units in distributionsystems regarding protection schemes. Int. Trans. Electr. Energy Syst. 2015, 25, 3618–3629. [CrossRef]

25. Strasser, T.; Andrén, F.; Kathan, J.; Cecati, C.; Buccella, C.; Siano, P.; Leitão, P.; Zhabelova, G.; Vyatkin, V.;Vrba, P.; et al. A Review of Architectures and Concepts for Intelligence in Future Electric Energy Systems.IEEE Trans. Ind. Electron. 2015, 62, 2424–2438. [CrossRef]

26. Palizban, O.; Kauhaniemi, K.; Guerrero, J.M. Microgrids in active network management—Part II: Systemoperation, power quality and protection. Renew. Sustain. Energy Rev. 2014, 36, 440–451. [CrossRef]

27. Kazmi, S.A.A.; Shahzad, M.K.; Khan, A.Z.; Shin, D.R. Smart Distribution Networks: A Review of ModernDistribution Concepts from a Planning Perspective. Energies 2017, 10, 501. [CrossRef]

28. Mirsaeidi, S.; Said, D.M.; Mustafa, M.W.; Habibuddin, M.H.; Ghaffari, K. An analytical literature reviewof the available techniques for the protection of micro-grids. Electr. Power Energy Syst. 2014, 58, 300–306.[CrossRef]

29. Mirsaeidi, S.; Said, D.M.; Mustafa, M.W.; Habibuddin, M.; Ghaffari, K. Progress and problems in micro-gridprotection schemes. Renew. Sustain. Energy Rev. 2014, 37, 834–839. [CrossRef]

30. Norshahrani, M.; Mokhlis, H.; Bakar, H.A.; Jamian, J.J.; Sukumar, S. Progress on Protection Strategies toMitigate the Impact of Renewable Distributed Generation on Distribution Systems. Energies 2017, 10, 1864.[CrossRef]

31. Morais, H.; Sousa, T.; Vale, Z.; Faria, P. Evaluation of the electric vehicle impact in the power demand curvein a smart grid environment. Energy Convers. Manag. 2014, 82, 268–282. [CrossRef]

32. Rahbari, O.; Vafaeipour, M.; Omar, N.; Rosen, M.A.; Hegazy, O.; Timmermans, J.-M.; Heibati, S.;Bossche, P.V.D. An optimal versatile control approach for plug-in electric vehicles to integrate renewableenergy sources and smart grids. Energy 2017, 134, 1053–1067. [CrossRef]

33. Farid, A.; Jiang, B.; Muzhikyan, A.; Youcef-Toumi, K. The need for holistic enterprise control assessmentmethods for the future electricity grid. Renew. Sustain. Energy Rev. 2016, 56, 669–685. [CrossRef]

34. Zakariazadeh, A.; Jadid, S.; Siano, P. Integrated operation of electric vehicles and renewable generation in asmart distribution system. Energy Convers. Manag. 2015, 89, 99–110. [CrossRef]

35. Bhattarai, B.P.; Myers, K.S.; Bak-Jensen, B.; de Cerio Mendaza, I.D.; Turk, R.J.; Gentle, J.P. Optimumaggregation of geographically distributed flexible resources in strategic smart-grid/microgrid locations.Electr. Power Energy Syst. 2017, 92, 193–201. [CrossRef]

36. Knezovic, K.; Marinelli, M.; Zecchino, A.; Andersen, P.B.; Traeholt, C. Supporting involvement of electricvehicles in distribution grids: Lowering the barriers for a proactive integration. Energy 2017, 134, 458–468.[CrossRef]

37. Abdelsamad, S.; Morsi, W.; Sidhu, T. Optimal secondary distribution system design consideringplug-inelectric vehicles. Electr. Power Syst. Res. 2016, 130, 266–276. [CrossRef]

Energies 2018, 11, 3106 14 of 17

38. Sadeghi, M.; Kalantar, M. Multi types DG expansion dynamic planning in distribution system understochastic conditions using Covariance Matrix Adaptation Evolutionary Strategy and Monte-Carlosimulation. Energy Convers. Manag. 2014, 87, 455–471. [CrossRef]

39. Sun, S.; Yang, Q.; Yan, W. A novel Markov-based temporal-SoC analysis for characterizing PEV chargingdemand. IEEE Trans. Ind. Inform. 2018, 14, 156–166. [CrossRef]

40. Arias, A.; Granada, M.; Castro, C.A. Optimal probabilistic charging of electric vehicles in distributionsystems. IET Electr. Syst. Transp. 2017, 7, 246–251. [CrossRef]

41. Wu, F.; Sioshansi, R. A two-stage stochastic optimization model for scheduling electric vehicle chargingloads to relieve distribution-system constraints. Transp. Res. Part B 2017, 102, 55–82. [CrossRef]

42. Hu, J.; You, S.; Lind, M.; Østergaard, J. Coordinated Charging of Electric Vehicles for Congestion Preventionin the Distribution Grid. IEEE Trans. Smart Grid 2014, 5, 703–711. [CrossRef]

43. Beaude, O.; Lasaulce, S.; Hennebel, M.; Mohand-Kaci, I. Reducing the Impact of EV Charging Operations onthe Distribution Network. IEEE Trans. Smart Grid 2016, 7, 2666–2679. [CrossRef]

44. Xiang, Y.; Liua, J.; Liu, Y. Optimal active distribution system management considering aggregated plug-inelectric vehicles. Electr. Power Syst. Res. 2016, 131, 105–115. [CrossRef]

45. Sundström, O.; Binding, C. Flexible Charging Optimization for Electric Vehicles Considering DistributionGrid Constraints. IEEE Trans. Smart Grid 2012, 3, 26–37. [CrossRef]

46. Wang, L.; Sharkh, S.; Chipperfield, A. Optimal decentralized coordination of electric vehicles and renewablegenerators in a distribution network using A∗ search. Electr. Power Energy Syst. 2018, 98, 474–487. [CrossRef]

47. Shuai, W.; Maillé, P.; Pelov, A. Charging Electric Vehicles in the Smart City: A Survey of Economy-DrivenApproaches. IEEE Trans. Intell. Transp. Syst. 2016, 17, 2089–2106. [CrossRef]

48. Peng, C.; Zou, J.; Lian, L. Dispatching strategies of electric vehicles participating in frequency regulation onpower grid: A review. Renew. Sustain. Energy Rev. 2017, 68, 147–152. [CrossRef]

49. Pavic, I.; Capuder, T.; Kuzle, I. A Comprehensive Approach for Maximizing Flexibility Benefits of ElectricVehicles. IEEE Syst. J. 2018, 12, 2882–2893. [CrossRef]

50. Lazarou, S.; Vita, V.; Christodoulou, C.; Ekonomou, L. Calculating operational patterns for electric vehiclecharging on a real distribution network based on renewables’ production. Energies 2018, 11, 2400. [CrossRef]

51. Mojdehi, M.N.; Ghosh, P. An On-Demand Compensation Function for an EV as a Reactive Power ServiceProvider. IEEE Trans. Veh. Technol. 2016, 65, 4572–4583. [CrossRef]

52. Daina, N.; Sivakumar, A.; Polak, J.W. Electric vehicle charging choices: Modelling and implications for smartcharging services. Transp. Res. Part C 2017, 81, 36–56. [CrossRef]

53. Weiller, C.; Neely, A. Using electric vehicles for energy services: Industry perspectives. Energy 2014, 77,194–200. [CrossRef]

54. Schmidt, J.; Lauven, L.-P.; Ihle, N.; Kolbe, L.M. Demand side integration for electric transport vehicles. Int. J.Energy Sect. Manag. 2015, 9, 471–495. [CrossRef]

55. Sato, T.; Kammen, D.M.; Duan, B.; Macuha, M.; Zhou, Z.; Wu, J.; Tariq, M.; Asfaw, S.A. Smart Grid StandardsSpecifications, Requirements and Technologie; John Wiley & Sons Singapore Pte. Ltd.: Singapore, 2015.

56. Green, R.C.; Wang, L.; Alam, M. Applications and Trends of High Performance Computing for ElectricPower Systems: Focusing on Smart Grid. IEEE Trans. Smart Grid 2013, 4, 422–431. [CrossRef]

57. Procopiou, A.T.; Quirós-Tortós, J.; Ochoa, L.F. HPC-Based Probabilistic Analysis of LV Networks with EVs:Impacts and Control. IEEE Trans. Smart Grid 2017, 8, 1479–1487. [CrossRef]

58. Yazdaninejadi, A.; Nazarpour, D.; Golshannavaz, S. Dual-setting directional over-current relays: An optimalcoordination in multiple source meshed distribution networks. Electr. Power Energy Syst. 2017, 86, 163–176.[CrossRef]

59. Singh, M.; Vishnuvardhan, T.; Srivani, S. Adaptive protection coordination scheme for power networksunder penetration of distributed energy resources. IET Gener. Transm. Distrib. 2016, 10, 3919–3929. [CrossRef]

60. Ates, Y.; Uzunoglu, M.; Karakas, A.; Boynuegri, A.R.; Nadar, A.; Dag, B. Implementation of adaptive relaycoordination in distribution systems including distributed generation. J. Clean. Prod. 2016, 112, 2697–2705.[CrossRef]

61. Chen, C.-R.; Lee, C.-H.; Chang, C.-J. Optimal overcurrent relay coordination in power distribution systemusing a new approach. Electr. Power Energy Syst. 2013, 45, 217–222. [CrossRef]

Energies 2018, 11, 3106 15 of 17

62. Chabanloo, R.M.; Maleki, M.G.; Agah, S.M.M.; Habashi, E.M. Comprehensive coordination of radialdistribution network protection in the presence of synchronous distributed generation using fault currentlimiter. Electr. Power Energy Syst. 2018, 99, 214–224. [CrossRef]

63. Jalilian, A.; Hagh, M.T.; Hashemi, S.M. An Innovative Directional Relaying Scheme Based on PostfaultCurrent. IEEE Trans. Power Deliv. 2014, 29, 2640–2646. [CrossRef]

64. Costa, M.H.; Saldanha, R.R.; Ravetti, M.G.; Carrano, E.G. Robust coordination of directional overcurrentrelays using a matheuristic algorithm. IET Gener. Transm. Distrib. 2017, 11, 464–474. [CrossRef]

65. Saleh, K.A.; Zeineldin, H.H.; Al-Hinai, A.; El-Saadany, E.F. Optimal Coordination of Directional OvercurrentRelays Using a New Time–Current–Voltage Characteristic. IEEE Trans. Power Deliv. 2015, 30, 537–544.[CrossRef]

66. Hussain, M.H.; Rahim, S.R.A.; Musirin, I. Optimal Overcurrent Relay Coordination: A Review. Procedia Eng.2013, 53, 332–336. [CrossRef]

67. Corrêa, R.; Cardoso, G., Jr.; de Araújo, O.C.; Mariotto, L. Online coordination of directional overcurrentrelays using binaryinteger programming. Electr. Power Syst. Res. 2015, 127, 118–125. [CrossRef]

68. Ojaghi, M.; Ghahremani, R. Piece–wise Linear Characteristic for Coordinating Numerical Overcurrent Relays.IEEE Trans. Power Deliv. 2017, 32, 145–151. [CrossRef]

69. Khorramdel, B.; Marzooghi, H.; Samet, H.; Pourahmadi-Nakhli, M.; Raoofat, M. Fault locating in largedistribution systems by empirical mode decomposition and core vector regression. Electr. Power Energy Syst.2014, 58, 215–225. [CrossRef]

70. Khakimzyanov, E.; Mustafin, R.; Platonov, P. Method of fault location for double line-to-earth faults indistribution networks. Eng. Sci. Technol. Int. J. 2016, 19, 1668–1671. [CrossRef]

71. Esmaeili, A.; Esmaeili, S.; Hojabri, H. Short-circuit level control through a multiobjective feederreconfiguration using fault current limiters in the presence of distributed generations. IET Gener. Transm.Distrib. 2016, 10, 3458–3469. [CrossRef]

72. Jamali, S.; Borhani-Bahabadi, H. Non-communication protection method for meshed and radial distributionnetworks with synchronous-based DG. Electr. Power Energy Syst. 2017, 93, 468–478. [CrossRef]

73. Bukhari, S.B.A.; Haider, R.; Zaman, M.S.U.; Oh, Y.-S.; Cho, G.-J.; Kim, C.-H. An interval type-2 fuzzy logicbased strategy for microgrid protection. Electr. Power Energy Syst. 2018, 98, 209–218. [CrossRef]

74. He, J.; Liu, L.; Xu, Y.; Ding, F.; Zhang, D. A two-step protection algorithm for smart distribution systemswith DGs. Int. Trans. Electr. Energy Syst. 2018, 28, 1–15. [CrossRef]

75. Rahman, M.; Isherwood, N.; Oo, A. Multi-agent based coordinated protection systems for distribution feederfault diagnosis and reconfiguration. Electr. Power Energy Syst. 2018, 97, 106–119. [CrossRef]

76. Pathirana, A.; Rajapakse, A.; Perera, N. Development of a hybrid protection scheme for activedistributionsystems using polarities of current transients. Electr. Power Syst. Res. 2017, 152, 377–389.[CrossRef]

77. Todorovic, I.; Grabic, S.; Ivanovic, Z. Grid-connected converter active and reactive power productionmaximization with respect to current limitations during grid faults. Electr. Power Energy Syst. 2018, 101,311–322. [CrossRef]

78. Zhang, Z.-H.; Xu, B.-Y.; Crossley, P.; Li, L. Positive-sequence-fault-component-based blocking pilot protectionfor closed-loop distribution network with underground cable. Electr. Power Energy Syst. 2018, 94, 57–66.[CrossRef]

79. Li, S.; Rodolakis, A.J.; El-Arroudi, K.; Joós, G. Islanding protection of multiple distributed resources underadverse islanding conditions. IET Gener. Transm. Distrib. 2016, 10, 1901–1912. [CrossRef]

80. Fazio, A.R.D.; Russo, M.; Valeri, S. A New Protection System for Islanding Detection in LV DistributionSystems. Energies 2015, 8, 3775–3793. [CrossRef]

81. Strasser, T.; Andrén, F.P.; Lauss, G.; Bründlinger, R.; Brunner, H.; Moyo, C.; Seitl, C.; Rohjans, S.; Lehnhoff, S.;Palensky, P.; et al. Towards holistic power distribution system validation and testing—An overview anddiscussion of different possibilities. Elektrotech. Informationstechnik 2017, 134, 71–77. [CrossRef]

82. Faruque, M.D.O.; Strasser, T.; Lauss, G.; Jalili-Marandi, V.; Forsyth, P.; Dufour, C.; Dinavahi, V.; Monti, A.;Kotsampopoulos, P.; Martinez, J.A.; et al. Applications of Real-Time Simulation Technologies in Power andEnergy Systems. IEEE Power Energy Technol. Syst. J. 2015, 2, 103–115.

Energies 2018, 11, 3106 16 of 17

83. Faruque, M.O.; Strasser, T.; Lauss, G.; Jalili-Marandi, V.; Forsyth, P.; Dufour, C.; Dinavahi, V.; Monti, A.;Kotsampopoulos, P.; Martinez, J.A.; et al. Real-Time Simulation Technologies for Power Systems Design,Testing, and Analysis. IEEE Power Energy Technol. Syst. J. 2015, 2, 63–73. [CrossRef]

84. Lauss, G.F.; Faruque, M.O.; Schoder, K.; Dufour, C.; Viehweider, A.; Langston, J. Characteristics and Designof Power Hardware-in-the-Loop Simulations for Electrical Power Systems. IEEE Trans. Ind. Electron. 2016,63, 406–417. [CrossRef]

85. Guo, F.; Herrera, L.; Murawski, R.; Inoa, E.; Wang, C.-L.; Beauchamp, P.; Ekici, E.; Wang, J. ComprehensiveReal-Time Simulation of the Smart Grid. IEEE Trans. Ind. Appl. 2013, 49, 899–908. [CrossRef]

86. RTDS Technologies Inc. Winnipeg, MB, Canada. Available online: https://www.rtds.com (accessed on 5November 2017).

87. OPAL-RT Technologies Inc. Montreal, QC, Canada. Available online: http://www.opal-rt.com (accessed on5 November 2017).

88. MathWorks Inc., Natick, MA, USA. Available online: https://www.mathworks.com (accessed on 5November 2017).

89. xPC Target. MathWorks Inc., Natick, MA, USA. Available online: http://www.mathworks.com (accessed on5 November 2017).

90. dSPACE GmbH. Paderborn, Germany. Available online: http://www.dspace.com (accessed on 5 November2017).

91. ADI RTS. Applied Dynamics International Inc., Arbor, MI, USA. Available online: http://www.adi.com(accessed on 5 November 2017).

92. Tyhoon HIL GmbH. Zürich, Switzerland. Available online: https://www.typhoon-hil.com (accessed on 5November 2017).

93. Lu, B.; Wu, X.; Figueroa, H.; Monti, A. A low-cost real-time hardware in- the-loop testing approach of powerelectronics controls. IEEE Trans. Ind. Electron. 2007, 54, 919–931. [CrossRef]

94. Adler, F.; Benigni, A.; Stagge, H.; Monti, A.; Doncker, R.D. A new versatile hardware platform for digitalreal-time simulation: Verification and evaluation. In Proceedings of the IEEE 13th Workshop ControlModeling Power Electronics (COMPEL), Kyoto, Japan, 10–13 June 2012.

95. Cintuglu, M.H.; Mohammed, O.A.; Akkaya, K.; Uluagac, A.S. A Survey on Smart Grid Cyber-PhysicalSystem Testbeds. IEEE Commun. Surv. Tutor. 2017, 19, 446–464. [CrossRef]

96. Hernandez, M.E.; Ramos, G.A.; Lwin, M.; Siratarnsophon, P.; Santoso, S. Embedded Real-Time SimulationPlatform for Power Distribution Systems. IEEE Access 2017, 6, 6243–6256. [CrossRef]

97. Craciun, O.; Florescu, A.; Munteanu, I.; Bratcu, A.I.; Bacha, S.; Radu, D. Hardware-in-the-loop simulationapplied to protection devices testing. Electr. Power Energy Syst. 2014, 54, 55–64. [CrossRef]

98. Stifter, M.; Cordova, J.; Kazmi, J.; Arghandeh, R. Real-Time Simulation and Hardware-in-the-Loop Testbedfor Distribution Synchrophasor Applications. Energies 2018, 11, 876. [CrossRef]

99. Magro, M.C.; Giannettonib, M.; Pincetic, P.; Vanti, M. Real time simulator for microgrids. Electr. Power Syst.Res. 2018, 160, 381–396. [CrossRef]

100. Papaspiliotopoulos, V.A.; Korres, G.N.; Kleftakis, V.A.; Hatziargyriou, N.D. Hardware-In-the-Loop Designand Optimal Setting of Adaptive Protection Schemes for Distribution Systems with Distributed Generation.IEEE Trans. Power Deliv. 2017, 32, 393–400. [CrossRef]

101. Piesciorovsky, E.C.; Schulz, N.N. Comparison of non-real-time and real-time simulators with relaysin-the-loop for adaptive overcurrent protection. Electr. Power Syst. Res. 2017, 143, 657–668. [CrossRef]

102. Vatu, R.; Ceaki, O.; Mancasi, M.; Porumb, R.; Seritan, G. Power quality issues produced by embedded storagetechnologies in smart grid environment. In Proceedings of the 2015 50th International Universities PowerEngineering Conference (UPEC), Stoke on Trent, UK, 1–4 September 2015; pp. 1–5. [CrossRef]

103. Vatu, R.; Ceaki, O.; Golovanov, N.; Porumb, R.; Serit,an, G. Analysis of storage technologies with smartgrid framework. In Proceedings of the 2014 49th International Universities Power Engineering Conference(UPEC), Cluj-Napoca, Romania, 2–5 September 2014; pp. 1–5, ISBN 978-1-4799-6556-4. [CrossRef]

104. Ekonomou, L.; Fotis, G.; Vita, V.; Mladenov, V. Distributed generation islanding effect on distributionnetworks and end user loads using the master-slave islanding method. J. Power Energy Eng. 2016, 4, 1–24.[CrossRef]

105. Hoke, A.; Nelson, A.; Chakraborty, S.; Bell, F.; McCarty, M. An Islanding Detection Test Platform forMulti-Inverter Islands using Power HIL. IEEE Trans. Ind. Electron. 2018, 65, 7944–7953. [CrossRef]

Energies 2018, 11, 3106 17 of 17

106. Hartmann, N.B.; Santos, R.C.; Grilo, A.P.; de Melo Vieira, J.C. Hardware Implementation and Real-TimeEvaluation of an ANN-Based Algorithm for Anti-Islanding Protection of Distributed Generators. IEEE Trans.Ind. Electron. 2018, 65, 5051–5059. [CrossRef]

107. Schneider, P.; Mather, B.A.; Pal, B.C.; Ten, C.W.; Shirek, G.J.; Zhu, H.; Fuller, J.C.; Pereira, J.L.R.; Ochoa, L.F.;de Araujo, L.R.; et al. Analytic Considerations and Design Basis for the IEEE Distribution Test Feeders. IEEETrans. Power Syst. 2018, 33, 3181–3188. [CrossRef]

108. IEEE PES AMPS DSAS Test Feeder Working Group. Available online: http://sites.ieee.org/pes-testfeeders/resources/ (accessed on 15 May 2018).

109. EPRI’s Green Circuit Project Database. Available online: http://sourceforge.net/p/electricdss/code/HEAD/tree/trunk/Distrib/EPRITestCircuits/ (accessed on 10 September 2018).

110. Mateo, C.; Prettico, G.; Gómez, T.; Cossent, R.; Gangale, F.; Frías, P.; Fulli, G. European representativeelectricity distribution networks. Electr. Power Energy Syst. 2018, 99, 273–280. [CrossRef]

111. Taxonomy of Prototypical Feeders (Pacific Northwest National Laboratory). Available online: http://sourceforge.net/p/gridlab-d/code/HEAD/tree/Taxonomy_Feeders/ (accessed on 5 July 2018).

112. Lazarou, S.; Vita, V.; Ekonomou, L. An open data repository for steady state analysis of a 100-node electricitydistribution network with moderate connection of renewable energy sources. Data Brief 2018, 16, 1095–1101.[CrossRef] [PubMed]

113. Hooshyar, H.; Mahmood, F.; Vanfretti, L.; Baudette, M. Specification, implementation, andhardware-in-the-loop real-time simulation of an active distribution grid. Sustain. Energy Grids Netw. 2015, 3,36–51. [CrossRef]

114. Ntalampiras, S. Fault Diagnosis for Smart Grids in Pragmatic Conditions. IEEE Trans. Smart Grids 2018, 9,1964–1971. [CrossRef]

115. 14 Bus Power Flow Test Case. Available online: https://www2.ee.washington.edu/research/pstca/pf14/pg_tca14bus.htm (accessed on 15 May 2018).

116. 30 Bus Power Flow Test Case. Available online: https://www2.ee.washington.edu/research/pstca/pf30/pg_tca30bus.htm (accessed on 15 May 2018).

117. 17 Generator Dynamic Test Case. Available online: https://www2.ee.washington.edu/research/pstca/dyn17/pg_tcadd17.htm (accessed on 15 May 2018).

118. 30 Bus Dynamic Test Case. Available online: https://www2.ee.washington.edu/research/pstca/dyn30/pg_tcadyn30.htm (accessed on 15 May 2018).

119. Xyngi, I.; Popov, M. An Intelligent Algorithm for the Protection of Smart Power Systems. IEEE Trans. SmartGrid 2013, 4, 1541–1548. [CrossRef]

120. Data in Brief. Available online: https://www.journals.elsevier.com/data-in-brief/ (accessed on15 May 2018).

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open accessarticle distributed under the terms and conditions of the Creative Commons Attribution(CC BY) license (http://creativecommons.org/licenses/by/4.0/).