**4. Simulation Results**

The proposed method based on the compound acceleration factor and BAS optimization algorithm aim to improve the speed and coordination of the microgrid protection. The impacts of negative factors on the conventional ITOCR in terms of a low fault level, microgrid operation mode, DG positions, and DG integration status have been notably reduced. In this section, the speed and coordination of the proposed protection method are evaluated. The possible negative influencing factors on the performance of the proposed method, such as fault types, fault positions, and microgrid operation modes, are evaluated. Besides, the comparison between the proposed method and the conventional method is also included in this section.

In the simulation, the model of a typical microgrid shown in Figure 4 is established in DIgSILENT/Power Factory. The detailed parameters of the microgrid are shown in Table 2. All DGs adopted the PQ control during the grid-connected operation mode of the microgrid. During the islanded operation mode, the microgrid applies the master–slave control mode. DG1, which is assumed as the most stable and reliable resource in the microgrid, is denoted as the master DG and adopts the V/f control. In turn, the other DGs kept using the PQ control. The initial state of the DGs in the microgrid is Scenario 1 in Table 1. The change in DG status will be evaluated in this section. Due to the thermal limit of the power inverter, the fault current limit of the DGs is set as 2 p.u. For conventional ITOCRs, the parameters of A and α are set as 0.14 and 0.02, and the minimum allowed *CTI* is set to 0.3 s [22,32]. F1 to F5 in Figure 4 denotes the fault positions.

**Table 2.** Simulation parameters of the microgrid model.


#### *4.1. Optimized Configuration of the Protection Parameters*

The settings parameters of the microgrid protection of Scenario 1 are optimized based on the methods described in Section 3. Table 3 shows the optimal protection parameters obtained according to the BAS algorithm optimization process shown in Figure 3.


**Table 3.** Optimal configuration of the protection parameters in Scenario 1.

#### *4.2. Evaluation of the Acceleration Capability of the Proposed Method*

In this section, the acceleration capability of the proposed method is evaluated with di fferent microgrid operating modes, fault types, and positions, and the comparison with conventional protection methods is carried out.

#### 4.2.1. Evaluations in the Grid-Connected Mode of the Microgrid

(a) Di fferent fault positions: In this case, the simulations are performed for di fferent fault positions in the grid-connected mode. The three-phase faults are applied in this case, while the evaluations of di fferent fault types will be discussed in the following section.

When fault F3 of Line 3 occurred, the fault current is measured as 5.2553 kA, and the acceleration factor *Mi* is calculated to be 0.0417 according to (10). According to the protection parameters in Table 3, the action delay of R3 is 0.0499 s. Figure 5 shows the voltage change of the bus when occurring at F3 of Line 3. From Figure 5, it is found that the operation time of the conventional primary relay is 0.1096 s, and the improved primary relay is 0.0597 s faster than the traditional relay. Table 4 shows the simulation results with different fault positions. The simulation results prove that the speed of the improved protection method can be guaranteed with different fault positions, and both the primary and backup relay's operation times of the I-ITOCR are faster than those of the ITOCR.

**Figure 5.** Bus voltage during the fault at F3.

**Table 4.** Operation times of the different fault positions in the grid-connected mode.


Note: PR and BR represent primary relay and backup relay, respectively.

(b) Different fault types: This section evaluates the speed of the proposed protection method with different fault types in the grid-connected mode. Table 5 shows the operation times of the I-ITOCRs and ITOCRs. The fault is set at the F2 of Line 2. In the table, A, B, C, and D represent a single-phase-to-ground fault, a two-phase fault, a two-phase-to-ground fault, and three-phase fault, respectively. The operation time for the improved primary relays are all less than 0.15 s, and the conventional primary relays are all greater than 0.35 s, which proves that the I-ITOCR can adopt all the fault types and is faster than the ITOCRs under different fault types.

**Table 5.** Operation times of the different fault types in the grid-connected mode.


#### 4.2.2. Evaluations in the Islanded Mode of the Microgrid

(a) Different fault positions: This case tests the speed of the proposed protection method for different fault position in the islanded mode. Please note that the three-phase fault is included in this case; other fault types will be discussed in the following section.

When fault F3 of Line 3 occurred, the fault current is 3.1970 kA, and the acceleration factor *Mi* is 0.0412 according to (10). According to the protection parameters in Table 3, the action delay of R3 is 0.0568 s. Figure 6 shows the voltage of the I-ITOCR and ITOCR of BUS2 when a fault occurs at F3 of Line 3. As can be seen from Figure 6, the operation time of the conventional primary relay is 0.1650 s, and the action delay of the improved primary relay is much shorter than that of the conventional protection method.

**Figure 6.** Bus voltage during the fault at F3.

Table 6 shows the simulation results of the other faults in this case. Note that, since the fault level of the microgrid is reduced, when the fault occurs at the F1 point, the operation time of the conventional primary relay is 1.1519 s, and the operation time for the improved primary relay is only 0.0728 s, which is much smaller than the conventional protection method. In addition, the primary and backup relay's operation time at other points of the I-ITOCR is also notably faster than the ITOCR. Therefore, the improved method has a better speed in the islanded mode of the microgrid.


**Table 6.** Operation times of the different fault positions in the islanded mode.

(b) Different fault types: In this case, the speed of the improved protection method for different types of faults in the islanded mode is tested. Table 7 shows the operation times of the ITOCRs and I-ITOCRs when different types of faults occur at the F3 of Line 3 of the microgrid. As can be seen from Table 7, the operation time for the improved primary relays are all less than 0.15 s, and the conventional primary relays are all greater than 0.16 s, which proves that the I-ITOCR is faster than the ITOCRs under different failure types in the islanded mode of the microgrid.


**Table 7.** Operation times of the di fferent fault types in the islanded mode.

#### *4.3. Evaluations of Protection Coordination of the Proposed Method*

In this section, the coordination of the protection methods is evaluated via the aspects of di fferent operating modes, fault positions and types, and DG integration status. In this simulation, the traditional protection parameters are set according to the islanded mode of the microgrid.

4.3.1. Evaluations in the Grid-Connected Mode of the Microgrid

(a) Di fferent fault positions: The simulations are performed with di fferent fault position in the grid-connected mode. The fault type in this section is set as a three-phase fault; other fault types are discussed in the following section.

Table 8 shows the simulation results of the di fferent fault positions. From the table, it is found that the *CTIs* between the primary and backup relays of the conventional protection are all less than 0.195 s, which causes the backup relay to remain active after the primary relay trips (*CTI* < 0.2). As a result, the selectivity of the protection cannot be guaranteed.


**Table 8.** Operation times of the di fferent fault positions in the grid-connected mode

For the proposed I-ITOCR method, the *CTIs* between the primary and backup relays of the improved protection method is all around 0.35 s, which fully satisfy the coordination requirement (0.2 < *CTI* < 0.5). Therefore, improved protection can ensure the selectivity for di fferent fault positions in the grid-connected mode.

(b) Di fferent fault types: This section evaluates the coordination of the proposed protection method for di fferent types of faults in the grid-connected mode. Table 9 shows the operation times of the ITOCRs and I-ITOCRs with di fferent fault types occurring at the F2 of Line 2. From the results, it is found that the *CTIs* of the conventional ITOCR are all less than 0.14 s, which do not satisfy the coordination requirement (*CTI* < 0.2). In turn, the *CTI* of the improved I-ITOCR is all in the range (0.2 < *CTI* < 0.5) for di fferent fault types, which promises the protection coordination and selectivity.


**Table 9.** Operation times of the different fault types in the grid-connected mode.

#### 4.3.2. Evaluations in the Islanded Mode of the Microgrid

(a) Different fault positions: In this case, the simulation is performed for different fault position in the islanded mode. Table 10 shows the simulation results of the example, and the faults are all three-phase faults; other fault types are discussed in the following section. As can be seen from the simulation results, the *CTI* between the primary and backup relays of the improved protection and conventional protection methods meet the condition (0.2 < *CTI* < 0.5), and the selectivity of these protections is not lost for different fault position in the islanded mode.

**Table 10.** Operation times of the different fault positions in the islanded mode.


(b) Different fault types: This section evaluates the coordination of the proposed protection method for different types of faults in the islanded mode. Table 11 shows the operation time of ITOCRs and I-ITOCRs with different fault types occurring at the F2 of Line 2. From the results, it is found that the CTIs of the I-ITOCR and ITOCR are all in the range (0.2<CTI<0.5) for different fault types, which promises the protection coordination and selectivity.


**Table 11.** Operation times of the different fault types in the grid-connected mode.

#### 4.3.3. Evaluations of the DG Connection States

In this case, the simulations were carried out with different DG states in the microgrid and the protection parameter settings in the different scenarios were consistent with the method of Section A. Table 12 shows the operation time of the I-ITOCRs when the fault occurs at F3 of Line 3 of the microgrid for eight scenarios; the faults are all three-phase faults, and G and I represent the grid-connected mode and islanded mode of the microgrid, respectively. As can be seen from the simulation results,
