1. Introduction
Due to the environmental and global warming issues, people have been recommending and projecting ideas on the smart grid. Today, the smart grid is a significant concept in the power policy national wide [
1,
2]. After the smart grid is introduced, the next generation of the grid is the microgrid. The microgrid power system is small-scale with various distribution models [
3]. Its main idea is to incorporate a limited number of distributed generators to ideally control them without making the system complex [
4]. The distribution models can be hybrid, radial, or interconnected. One prominent advantage of the microgrid is that it has two modes of operation. They are grid-connected and islanded-mode. This is a key player in the next generation of the power network.
There is increased growth of distributed energy resources (DER) in the power grid [
5]. The reason is that it can reduce electricity costs as well as greenhouse gas emissions [
6]. It can also assist the consumers through reducing the dependence on the main electricity grid. DER has a very low value to generate power [
7]. It is usually connected to the low and medium voltage of the distribution network using an inverter.
The key point of adaptive protection is that when a fault occurs, the relay can adjust to the power network conditions [
8]. It means that the relay can adjust to the current operational condition of the power system. The adaptive protection relay itself is a digital relay. Since this relay has an automatic setting, this technique applies to do so. In the traditional relay, the setting is fixed with one parameter. The relay does not adapt to the surrounding environment, as it does not have the function of real- time. The adaptive protection relay has a microprocessor, so it can use the real-time function. The faults that cannot be protected by the traditional relay, the adaptive protection, can protect the system with a no-fault error.
During the modeling of a modern power system, it recognized two factors that could affect the system. They are the analyzation and isolation of the fault. If a fault happens, it should be isolated fully. To keep the load safely, the affected regions in the main system or the neighboring areas should isolate the fault very fast, no matter how big or small the fault is. The existing protection methods and relay setting can lead to inappropriate tripping, as the evolvement of the grid needs protection such as adaptive protection, which can get the most use of the protection [
6]. As the technologies get advanced day by day, the protection relay also has to be advanced. Therefore, adaptive overcurrent protection is one of the effective solutions, as it can be used for traditional networks as well as future grids.
This paper presents the data on the digital relay and the challenges of the microgrid. This paper contributes to an algorithm that is developed for the adaptive overcurrent relay protection to overcome the challenges of the microgrid with distributed energy resources. The adaptive overcurrent protection will be shown in the microgrid connected to the IEEE-9 bus system simulation at the same time other protection types are showcased as well.
The remaining of the paper is organized in the following manner:
Section 2 contains a discussion on the challenges of the microgrid.
Section 3 will discuss the possible solutions for the challenges.
Section 4 will emphasize the distributed energy resources and impacts.
Section 5 will explain the proposed algorithm protection.
Section 6 presents the validation and verification of microgrid simulation studies.
Section 7 discusses the results. The conclusion is summarized in
Section 8.
3. Possible Solutions for the Challenges
The following are the possible solutions for the above-mentioned challenges.
3.1. Adaptive Protection
Each grid-connected and islanded mode will have its own set of relay settings. When the microgrid switches the operational mode, the relay will choose the best setting for that particular mode. For the islanded mode, the time overcurrent characteristic curve of the relay will be changed to instantaneous and/or definite time overcurrent relay settings to low fault currents. Since the adaptive protection relay is a digital relay, the relay’s setting would adjust the operational mode of the network. One example would be when there is a large voltage depression case, the restraint voltage overcurrent protection would decrease delay tripping, which is the time dial in the relay setting.
3.2. Differential Protection
This protection is used on the coupled differential directional relays. It can accurately identify the location and remove the fault without disrupting the components in the system. This protection is used in traditional protection technologies. However, it can be used for both islanded mode and grid-connected in the microgrid. Differential protection can be either localized or centralized. The definition of centralized means it is coordinated and monitored by the central controls. The definition of localization means it is created from the local communication among the relays. The localized scheme, which is more acceptable for applications of the industry, allows communication directly among the relays where the fault will have the fastest solution for it. In the centralized scheme, the topology of the microgrid and power device’s setting operation is monitored by the central controllers. Then, the tripping of the command is sent to the power device when the fault is sensed. This scheme provides precise results with an unacceptable time delay to do calculation, which is required by the central controllers.
3.3. Real-Time Closed-Loop Test to Adaptive Protection in a Smart-Grid Context
The purpose of this research is to show the viability of real-time and automated adjustment in the setting of commercial relay protection. To illustrate it, a commercial protection relay and Real-Time Digital Simulator (RTDS) with a closed-loop test rig used to set up. The commercial relay protection uses the theory of adaptive protection while the RTDS set is real-time automated with various functions of protection.
The first test is about testing smart substation control (SSC) using a real-time closed-loop test. The equipment for the setup is an RTDS and EFACEC TPU S220, which is the commercial relay. Using the network’s actual state, the SSC can verify whether the network is in a sympathetic tripping situation. The SSC collects real-time information from the local sensors to know the network’s actual state and the sympathetic tripping evaluation. Based on the SSC mode of operation, to avoid sympathetic tripping, the protection group values of the feeder relay might alter.
The second test is the adaptive protection test. For this test, the researcher has done two case studies: one where the SSC was halted, and the other one was SSC activated. The preset of the parameters and functions of the protection relay is controlled by the SSC to change automatically when the SSC is activated. Under the SSC, there are two setting groups. They are default settings and sympathetic tripping. The default setting is named group 1, while the setting for sympathetic tripping is group 2. Once the SSC senses the possible occurrence of synthetic tripping, it will change from group 1 to group 2.
3.4. Dual Complex Adaptive Protection Algorithm for Microgrid
The power demand has been growing, and the traditional energy sources are depleting. So, the penetration of DERs has been introduced to the microgrid. Therefore, the microgrid cannot use the conventional power protection scheme [
10]. The power flow of the microgrid is bidirectional. Therefore, it affects the operation of the protection as well as coordination. In the traditional power grid, the power flow is one-directional.
Therefore, one of the researchers has projected the dual simplex algorithm for the novel adaptive overcurrent protection for a microgrid. The projected algorithm should be able to detect the operation time and time multiplier setting (TMS) of the relay for a given microgrid type. This information is saved in the lookup table, which is in the central protection system. The lookup table should find suitable protection coordination when there is a fault in the microgrid.
5. Algorithm for Adaptive Overcurrent Relay Protection
Adaptive overcurrent protection is more efficient than traditional overcurrent protection. The traditional overcurrent protection relay is not flexible in adjusting to the network parameters automatically. It must be done manually. This might lead to the wrong isolation of the fault. However, for adaptive overcurrent protection, it uses real-time values automatically and isolates the fault very fast. Therefore, there is less or no wrong isolation of the fault. The reason why adaptive protection is more effective than traditional protection will be explained under discussion using the four key points.
An algorithm is developed for adaptive overcurrent relay, which is explained as follows.
The flowchart of the algorithm is given in
Figure 1. The new adaptive relay connected to the network will firstly do an auto-calibrate. After that, the parameters of
, Time Dial Setting (TDS), and the fault current will be preset to zero. Following, it will do a Newton–Raphson analysis, which is the load flow and short-circuit analysis. Then, it will calculate the
. The 1.5 in the
represents the safety margin [
13].
Then, the relay will check for any changes in the system. If there is no change, it will calculate for new TDS. However, there is a change to the system when the load value changes or the distributed generator (DG) connection value changes. All the necessary changes for load or DG are communicated through the relay. This results in a new set of parameters for and TDS generated.
The next step is identifying the fault and its location by using the overcurrent equation, which is denoted by “(1)” in the algorithm.
The principle of overcurrent protection is shown in (1).
In the overcurrent principle, there are two conditions. The first condition is “Normal Condition”, where there is no fault. The relay will not operate for circuit breakers (CB) to trip, which means that I’ is smaller than the . T is the representation of the second condition, which is the short-circuit fault condition. When I’ is greater the , the relay determines the mode of operation.
After determining the mode of protection, it issues a command for which CB to trip. Once the fault is cleared, it will restore to the original condition. It continues to do the same process for the next iteration where .
Implementation of the Algorithm in an Existing Distribution System
The developed algorithm can be implemented in the actual distribution system using supervisory control and data acquisition (SCADA) and ETAP Real-Time. The latest ETAP software should be used for the implementation.
The information system portrays a key importance in the application of new ideas in the control and monitoring of the power network [
14]. The new feature in the SCADA system is the state estimation concept. It supports the advancement of the information system and automation schemes. The SCADA system uses real-time in ETAP software which, in turn, allows the execution of the state estimate.
Real-time simulation is one of the needs that Engineers use these days. It uses the execution of new solutions in monitoring, the control system, and automation. This concept is created as the foundation for a combination of two technologies for the simulation of control systems and monitoring.
The real-time not only has characteristics for dynamics but also has physical counterparts for a correct simulation time of response. The time required for simulation to produce improvements or results is roughly about the time needed for the actual system. The latest technology has allowed the exchange of data across the telecommunication protocols to be improved. In the electrical system, it is taken as part of the control and monitoring of the system.
The SCADA plays a key role in the execution of an autonomous system. It makes sure the power system has high efficiency and security in the system, the quality, and permanence of the power supply. The current applications of a SCADA system are increasing. It can analyze and obtain the data that is from an online connection.
The state estimator concept is used by the ETAP real-time application. The application needs the power network’s model for load flow calculation as well as measurement integration. It is for collecting online data by using communication protocols.
Based on the above concepts, it should be able to implement the algorithm in the actual power distribution system.
6. Validation and Verification of Microgrid Simulation Studies
Figure 2 represents the IEEE-9 bus system connected to a microgrid system [
15]. In this system, there are three wind turbine generators (WTG) with different parameters being used. Adaptive overcurrent protection and differential transformer protection are being used in this system.
The tripping of the protection relay happens when a fault occurs based on the following equation.
The
is the tripping time of the relay, while
,
, and
are the time multiplier setting (TMS), fault current, and pickup current, respectively. The fault current always depends on the short circuit fault and the system model. The values of
and
are relay settings. The value of
should be greater than the maximum load current. This is measured by the current transformer and leaving some safety factor (SF). Therefore, the pickup current is calculated in the following manner.
As mentioned before, the value of SF is normally set to 1.5. The following is an example of adaptive protection using the algorithm.
Figure 3 shows the steps on how the algorithm works in ETAP software using Line2 having a fault.
In simple terms, this action summarizes that there is a three-phase fault at Line3. Then, CB6 and CB8 will start to operate by opening them. Once the isolation is done, both CB6 and CB8 will close. The system is back to normal. The following results are the during and after the fault.
During the fault, CB6 will trip first, followed by CB8 this is shown in
Figure 4.
The below result is when the fault is isolated. After isolating the fault, the system is back to normal.
The ETAP version 16 does not support real-time values; therefore, some parts are done manually. However, the outcome results are still the same and correct.
When a fault occurs between Bus 4 and Bus 6 as shown in
Figure 5, CB6 will open first as the primary protection followed by CB8 and CB10 as backup protection. The following is an example of when the power output from the wind turbine value changes. The value will change from 1.2 kW to 3 kW. The fault current changes when there is a change in the value, which can be seen below. It can be concluded that when the wind turbine power increases, the fault current also increases. The following results shown in
Figure 5 and
Figure 6 are the comparison between the overcurrent protection and developed algorithm of the adaptive overcurrent protection.
Figure 7 shows the results when WTG1 = 1.2 kW and WTG1 = 3 kW. Parts (a) and (b) in
Figure 7 do not match. The overcurrent relay is fixed at the original parameters, which is at WTG1 = 3 kW. Therefore, the relay needs to recalculate the parameters when WTG1 = 3 kW. The following results are when the developed algorithm of adaptive overcurrent protection is used.
However, in
Figure 8, part (a) and (b) both get the same tripping sequence. The reason is that adaptive overcurrent relay uses the real-time value. Therefore, there is no need for human intervention. The following figure represents a comparison between the two relays when the load value changes. The load value changes from 1 MVA to 3 MVA.
Based on the
Figure 9, it can be concluded that using overcurrent protection is not effective if the load value changes. As the overcurrent protection relay does not have the function of capturing the real values automatically, it needs to recalculate manually. However, for adaptive overcurrent protection, it adjusts automatically to the changes. Therefore, adaptive overcurrent protection is more effective than overcurrent protection.
The following
Table 1 is the comparison between the adaptive overcurrent protection and overcurrent protection. Based on the comparison, it can be said that adaptive overcurrent protection relay is a better option than overcurrent protection. The reason for that is if a fault occurs, adaptive overcurrent relay will isolate the fault accurately and quickly for all the scenarios mentioned. However, for overcurrent protection, when there is no change in the network, the isolation will be correct on its first try. For the rest of the scenarios, the correct isolation will happen after many attempts.
The following
Table 2 is the representation of the tripping sequence between two buses.
The transformers will be protected by differential transformer protection, which is denoted by relay 87. It only protects the internal fault of the transformer and does not protect any external fault. By adding an overcurrent relay, it will protect the external faults of the transformer.
Figure 10 is an example of the external and internal fault of a transformer. These results are based on when WTG1 = 3 kW.
For this simulation, Relay 21 operates as the created algorithm adaptive protection. However, the internal fault that is operated by Relay 22 will only trip CB1 and CB2. The results for the rest of the transformers are shown in
Table 3.
There is no backup protection for any transformer internal fault. However, in order to have backup protection, an overcurrent relay connects to it. For this simulation, an adaptive overcurrent protection algorithm is being used.
7. Discussion
To establish a good performance of a relay in the distribution network, four key characteristics are used when choosing a relay. They are selectivity, reliability, sensitivity, and speed [
16].
Selectivity: When operating a fault, only a small number of fault interruption devices should be used.
Reliability: The precision of the functioning relay. The precision is determined by the following equation.
Under reliability, there are two key essentials. They are dependability and security. Unwanted actions are removed under security. Dependability is how precise the tripping of the circuit breaker is when a fault occurs. Security can be improved by the selectivity of the system relay, while dependability can be improved by having a higher sensitivity level in the system relay.
Sensitivity: The protection relay in the system should sense the smallest fault current value. More sensitivity means when the protection relay can sense the smallest value of the fault current. Improving the efficiency of the sensitivity is by looking for that specific thing that makes the fault unique. The faults that happened in the system may not be the same every time.
Speed: How quickly the relay can isolate the fault when a fault occurs, without interrupting the unaffected zones of the network.
Based on the above characteristics, the comparison between two protective relays (overcurrent relay and adaptive overcurrent relay) is shown in
Table 4 below. This comparison is done based on the results of
Figure 7 and
Figure 8.
Based on the above table, it shows that adaptive overcurrent protection is a better option compared to overcurrent protection. The main reason is the adaptive overcurrent protection quickly isolates the fault without disturbing the unaffected area, even when DGs are connected. That is because this relay can achieve real-time values. It is not fixed at one value. However, overcurrent protection is not effective when DGs are connected. The current value changes and is fixed to operate at one value. This can be seen in
Figure 6 and
Figure 7. In addition, the algorithm of adaptive overcurrent relay protection can be used for any type of grid and is not fixed to one. However, it is more effective to use in any smart-grid system.
Before the digital relays, conventional relays were famous. The idea of the digital relay was spread around in 1985 [
17]. One example of the digital relay is the adaptive overcurrent relay. By comparing to the conventional relay, the digital relay has more functions on the monitoring side. In terms of the usage, the functions are flexible [
18]. The digital relay can create complicated functions. The efficiency is high, and it can interconnect with other digital equipment. The digital relay has more advantages over conventional relay. The following are the most noticeable advantages. It is easy to control and set up. The action of the operation is fast. There is a communication line between the various relays and the protection coordination. This helps isolate the fault fully.
7.1. Disturbances in the Electrical Grid
Due to the power quality disturbance in the industrial electrical grid, the algorithm might not work accurately as compared to the ETAP software. The main reason for this is power quality disturbance. The ETAP software assumes all the power quality disturbances are taken care of by the software itself. However, in the industrial electrical grid, the power quality disturbances have to be taken into consideration.
Supraharmonic emission is something new and can affect the infrastructure of any kind of smart grid. PLC communication is one example of it. Due to today’s new electronics, this emission has increased. Energy efficiency devices for domestic uses and power electronics are included as a key source of supraharmonic emission in the smart grid. The following are some of the sources with the emission frequency [
19].
PV inverters: 4–20 kHz
EV chargers: 15–100 kHz
Domestic devices: 2–150 kHz
Streetlamps: Up to 20 kHz
Converter (Industrial-size): 9–150 kHz
It is harder to identify supraharmonic emission when comparing to harmonic emission for a traditional grid. There is a standard harmonic emission set for PV installation and equipment with low voltage. Therefore, they are not categorized as supraharmonic emissions.
7.2. Solutions for DERs Challengers
There are challenges for having many DERs in the network, which will be mentioned in the following [
20]. The first one is that during island mode, the current control strategy is not able to work. The reason for that is no main energy source. The second one is that the microgrid needs fast regulation in the islanded mode because there are many capacities in multiple strategies. The last one is that when the DERs values are high, there is a problem in the distribution network. The problems are the voltage and network frequency being unstable and voltage level rise. The microgrid can solve such problems.
The following are some of the solutions to the challenges. Using the DER’s options, system reliability and flexibility can be improved. Then, the generation efficiency can be improved by the DER’s waste heat. Another one is without disrupting the public grid operation, the integration of DER can be done by permitting the power network to inspect and control the fault more effectively. At the same time, the damage caused by the disruption of the DER should be reduced by constantly feeding critical loads.