*3.3. STATCOM*

The Jeju Island power system has two STATCOM, each with a capacity of 50 MVar. They have an important role with respect to grid voltage stability in its power system because the CSC-HVDCs consume a lot of reactive power. The simulation models of STATCOM consist of a three-level VSC and its controller, as shown in Figure 11.

**Figure 10.** Simulation model of a thermal power plant.

**Figure 11.** Simulation model of STATCOM.

#### *3.4. Existing Wind Farms*

The simulation models of existing wind farms with approximately 250 MW are used as controlled current source equivalent model, which can adjust active and reactive powers easily in a simulation program. Thus, the measured data in the Jeju Island power system will be input data of this model, as seen in Figure 12.

**Figure 12.** Simulation model of an existing wind farm.

#### *3.5. Whole Simulation Model Including the New OWF*

Figure 13 shows the whole simulation model of the Jeju Island power system. The red line represents the connecting point of the new OWF by an electric network interface (ENI), as located in the top left of Figure 13, which supports the parallel computing part in PSCAD/EMTDC. The time scale of the whole simulation is assumed to be 0.14 milli-times by adjusting the time constant of every component. The actual parameters and configurations of the transmission line are applied to this system.

**Figure 13.** Simulation model of overall Jeju Island power system.

#### **4. Simulation Results**

Because there are many components of the Jeju Island power system, the line colors should be noted as the following colors in Table 3. This line color will be applied to every simulation result of active and reactive power.


**Table 3.** Expressions of simulation results.

#### *4.1. Case 1: Normal Operation of the Jeju Power System without 100 MW OWF*

To confirm the accuracy of the base simulation model, this scenario was conducted. Figure 14, representing active and reactive power at the top and bottom, respectively, shows measured data and simulation results. The top and bottom of Figure 15 represent the grid frequency and voltage. Hence, the errors between the simulation results and the measured data were less than 1% by grid frequency and voltage.

**Figure 14.** The simulation results of case 1: (**a**) Measured data in the Jeju Island power system (top: Active power, bottom: Reactive power); (**b**) The simulation model (top: Active power, bottom: Reactive power).

**Figure 15.** The simulation results of case 1: (**a**) Measured data in the Jeju Island power system (top: Grid frequency, bottom: Grid voltage measured at the biggest power load); (**b**) The simulation model (top: Grid frequency, bottom: Grid voltage measured at the biggest power load).

#### *4.2. Case 2: Normal Operation of the Jeju Power System with a New 100 MW OWF Based on DR-HVDC*

In the second scenario, the new 100 MW OWF was connected to the Jeju Island power system newly. In contrast to the first scenario, the CSC-HVDC #2 was operated at a limited minimum power, and the thermal power plants also reduced the output power, because the new OWF had generated additional active power. The CSC-HVDC #1 adjusted output power following demand power load and output from the wind farm to stabilize the grid frequency, as seen in Figure 16a. Due to the operation of CSC-HVDC #1, the grid frequency was in the grid code of South Korea from 59.8 Hz to 60.2 Hz. The maximum variance of frequency was slightly increased to 60.02 Hz, as compared to the first case. The variance of voltage was also higher in case 2 than in case 1 without the operation of the new OWF by approximately 2 kV, as shown in Figure 16b.

Figure 17a shows simulation results focused on DR-HVDC. Its reactive power was maintained to the unity power factor. The DC link voltage of DR-HVDC was in a constant value of 200 kV regardless of the active and reactive power, as seen in Figure 17b.

**Figure 16.** The simulation results of case 2: (**a**) top: Active power, bottom: Reactive power; (**b**) top: Grid frequency, bottom: Grid voltage measured at the biggest power load.

**Figure 17.** The simulation results of case 2: (**a**) top: Active power of DR HVDC, bottom: Reactive power of DR HVDC; (**b**) top: AC voltage measured at DR HVDC connection point, bottom: DC link voltage of DR HVDC.

#### *4.3. Case 3: Disconnection Fault Occurred at the DC Transmission Line*

In the third scenario, the DC submarine cable was disconnected abruptly. Thus, the output power of DR-HVDC was zero suddenly, and then the HVDC #1 compensated, as shown in Figure 18a. From this fault condition, the frequency and voltage were dropped to 59.97 Hz and 158 kV, as seen in Figure 18b, respectively. The DC link voltage of DR-HVDC also had a 5% variance. The AC voltage was recorded at the OWF grid by converter operation of wind turbines, as shown in Figure 19a. Figure 19b presents the onshore output of the DR-HVDC, whose active power was reduced to zero because of disconnection fault.

**Figure 18.** The simulation results of case 4: (**a**) top: Active power, bottom: Reactive power; (**b**) top: Grid frequency, bottom: Grid voltage measured at the biggest power load.

**Figure 19.** The simulation results of case 4: (**a**) top: Instantaneous voltage measure at the OWF grid, middle: Instantaneous current measure at the OWF grid, bottom: DC link voltage of DR-HVDC; (**b**) top: AC voltage measured at DR HVDC connection point, middle: Active power of DC-HVDC, bottom: Reactive power of DR-HVDC.

#### *4.4. Case 4: Single Line Ground Fault Occurred at AC Line in the New 100 MW OWF*

From the simulation results of case 4, the DR-HVDC was able to protect the Jeju Island power system from an OWF side fault. Although the voltage and current of the OWF side power grid were oscillated, as shown in Figure 21a, the voltage and frequency of the Jeju Island power system was stable, as seen in Figure 20b. The active power dropped at 35 MW, as illustrated in Figure 21b, then the CSC-HVDC #1 compensated that immediately, as seen in Figure 20a.

**Figure 20.** The simulation results of case 4: (**a**) top: Active power, bottom: Reactive power; (**b**) top: Grid frequency, bottom: Grid voltage measured at the biggest power load.

**Figure 21.** The simulation results of case 4: (**a**) top: Instantaneous voltage measure at OWF grid, middle: Instantaneous current measure at OWF grid, bottom: DC link voltage of DR-HVDC; (**b**) top: AC voltage measured at DR HVDC connection point, middle: Active power of DC-HVDC, bottom: Reactive power of DR-HVDC.
