1. Introduction
In recent years, the depletion of fossil fuels and environmental pollution have become important matters of global concern because they cause the depletion of energy resources and global warming. Accordingly, the demand for the use of renewable energy sources, such as wind power, solar energy, and biomass, has rapidly increased [
1,
2,
3]. To meet the demand and reduce the problems associated with the depletion of fossil fuels and environmental pollution, numerous studies have been conducted for green energy and renewable energy development [
4,
5,
6]. The common renewable energy systems are wind power generation (WPG), which uses wind turbines, and solar energy generation (SEG), which uses photovoltaic cells. However, sources such as wind power and solar energy might be unreliable and unpredictable because of changes in environmental conditions. Therefore, energy storage systems (ESSs) are used for conserving energy generated by the renewable energy sources in battery systems.
The grid-connected ESS usually generates and supplies power by connecting to a grid. It is used for conserving the additional energy with a reasonable cost, such as at night. Moreover, it can improve the energy quality and maximize its efficiency by supplying the conserved energy on requirement. The ESS has been made commercially available as typical renewable energy and energy conservation systems [
7,
8,
9,
10,
11].
Power generation systems connected to the grid can have a negative impact on the grid in case the power scale of the systems becomes large. Therefore, several countries suggest grid-code regulations to mitigate the negative impacts on the grid. Grid-code regulations differ from country to country; Germany’s grid-code regulations are more rigorous than the grid-code regulations of other countries. In general, the grid-code regulations define that systems such as WPG and SEG have to remain connected to the grid when the voltage drops for a specified time and support the grid with a reactive current. This requirement, known as a low-voltage ride-through (LVRT), is needed to avoid grid blackouts.
The LVRT requirement is crucial for grid-code regulations.
Figure 1 shows the LVRT requirements in Germany and China. The requirements are different regarding a fault duration and the injection quantity of the reactive power depending on the voltage drop ratio. This requirement ensures that the system connected to the grid operates properly when the grid voltage drops [
12,
13,
14,
15,
16,
17,
18]. The LVRT requirement contributes to the recovery of the grid voltage by supplying the reactive current in the designated voltage range. In order to satisfy the LVRT requirement, a proper control of the reactive current in the grid-connected ESS is necessary. To control the reactive power and the current for the grid-connected inverters in the WPG, SEG, and ESS is crucial. Numerous studies have been conducted for issues related to the reactive power and current compensation capabilities of such systems [
19,
20,
21,
22,
23,
24,
25].
The WPG systems using wind turbines and the SEG systems using photovoltaic cells usually have a large power scale. Therefore, the LVRT requirement needs to be thoroughly implemented in these systems. The LVRT control strategies researched recently comply with the LVRT requirement. In addition, these strategies focus on the additional operations that are effective and economical, which include the continuously stable operations with dynamic breakers in the WPG systems [
26,
27,
28,
29,
30] and the maximum power generation with low voltages in the SEG systems [
31,
32,
33]. These methods for additional operations are presented by considering the characteristics of the source type. The research on the LVRT control strategies for the WPG and SEG systems is actively progressed, however, the research on the LVRT control strategy for the grid-connected ESSs is insufficient.
The WPG and SEG systems are characterized by unidirectional power flow, which are unlike the grid-connected ESSs that are characterized by bidirectional power flow. Therefore, the LVRT control strategy used for the grid-connected ESSs needs to comply with the LVRT requirement, additionally, the characteristics of bidirectional power flow, i.e., the charging conditions of the grid-connected ESSs, need to be considered in the LVRT control strategy. The LVRT control strategy for the ESS is presented in [
34]. In [
34], when the grid voltage drops under charging condition of the grid-connected ESS, the LVRT control strategy is applied after the condition of the grid-connected ESS is changed to the discharging condition. It contributes to increase the voltage at the point of common coupling (PCC). This LVRT control strategy can be applied in applications improving the grid voltage durability such as frequency regulator that improves the grid frequency stability [
35], [
36]. However, in applications with other system connected to the DC-link of the grid-connected ESS, the charging condition of the grid-connected ESS must be maintained although the grid voltage drops. This paper presents the LVRT control strategy for grid-connected ESSs by considering the characteristic of bidirectional power flow. The proposed LVRT control strategy complies with the LVRT requirement and operates for contributing to increase voltage at the PCC. In other words, when the grid voltage drops, the proposed LVRT control strategy determines the injection quantity of the active and reactive currents depending on the voltage drop ratio of the three-phase grid. Additionally, in this paper, we analyzed the variations of the PCC voltage depending on the phase of the reactive current of the grid-connected ESS during the charging and discharging conditions. The validity of the proposed LVRT control strategy is verified and the variations of the PCC voltage of the grid-connected ESS are analyzed by PSIM simulation and experimental results.
3. Proposed LVRT Control Strategy for Grid-Connected ESSs and Analysis of PCC Voltage Variation
Figure 4 shows the circuit configurations of the grid-connected ESS using a voltage source inverter (VSI). It comprises a battery, DC-link capacitors (
CDC), a VSI, a filter (
Lf), the resistive impedance (
RG) and inductive impedance (
LG) elements of the grid, and a three-phase grid. The VSI comprises six insulated gate bipolar transistors (IGBTs) with antiparallel diodes. In addition, there is a PCC between the filter and the three-phase grid. The three-phase grid generates balanced three-phase grid voltages (
VR,
VS, and
VT) having constant frequencies.
3.1. Proposed LVRT Control Strategy for a Grid-Connected ESS
The LVRT control strategy needs to be applied in the grid-connected ESS having a large power scale. This strategy generally considers the characteristic of the source type or the application of the system. However, regardless of the characteristics or application, the main purpose of the LVRT control strategy is to comply with the LVRT requirement and inject reactive power through the grid-side inverter on the basis of the gird-code regulations.
The characteristics of the grid-connected ESS are different from those of the WPG and SEG systems; the grid-connected ESS with the battery in the DC-link has bidirectional power flow depending on the charging and discharging conditions, unlike the WPG and SEG systems, which have unidirectional power flow. Therefore, the LVRT control strategy for the grid-connected ESS needs to consider the charging and discharging conditions.
When the grid voltage drops during ESS operation under the discharging and charging conditions, the proposed LVRT control strategy for the grid-connected ESS becomes similar to the strategy for the WPG and SEG systems. The proposed LVRT control strategy determines the injection quantity of the reactive current for injecting reactive power into the three-phase grid depending on the grid-code regulation. In addition, the active current for transferring active power is determined within the range of the rating current of the grid-connected ESS.
The method for determining the injection quantity of the active and reactive currents depends on the voltage drop ratio of the three-phase grid as shown in
Figure 5. The voltage-level (
VLEVEL) of the three-phase grid is calculated by the voltage-level calculation process using the three-phase grid voltages
VR,
VS and
VT.
VLEVEL is classified into three parts depending on the LVRT requirement of the grid-code regulations, and the method for determining the injection quantity of the active and reactive currents is selected by each part. If
VLEVEL is greater than 90% of the three-phase grid voltage under normal conditions, the reactive current injected into the three-phase grid is zero, and the active current becomes the reference current. If
VLEVEL is greater than 50% but less than 90% of the three-phase grid voltage, then reactive current to be injected into the three-phase grid is determined based on the voltage drop ratio of the three-phase grid. In addition, the active current is calculated using the reactive current and the rating current of the grid-connected ESS. If
VLEVEL is less than 50% of the three-phase grid voltage, the reactive current injected into the three-phase grid is the rating current of the grid-connected ESS, and the active current is zero.
Figure 6 shows the voltage-level calculation process using
VR,
VS and
VT as the three-phase grid voltages. The voltage magnitudes (
VR(mag),
VS(mag) and
VT(mag)) of each phase in the three-phase grid is calculated by
VR,
VS and,
VT and the three-phase grid voltages (
VR(shift),
VS(shift) and
VT(shift)), which are transformed by the orthogonal signal generator such as the all-pass filter. In addition,
VLEVEL is determined by
VR(mag),
VS(mag), and
VT(mag) using the maximum value estimation. The
VLEVEL obtained by the voltage-level calculation process that uses the three-phase grid voltage magnitude is important because it is used for detecting the LVRT requirement.
As a result, the proposed LVRT control strategy for a grid-connected ESS determines the injection quantity of the active and reactive currents depending on the voltage drop ratio of the three-phase grid, which is based on the grid-code regulation.
Figure 7 shows the control block diagram of the grid-connected ESS with the proposed LVRT control strategy. The
VR,
VS and
VT are used to detect the phase angle of the three-phase grid using the phase-locked loop (PLL). Additionally, they are used for the voltage-level calculation process, which determines
VLEVEL. Using
VLEVEL and the LVRT control strategy, the active and reactive reference currents are determined.
3.2. Analysis of PCC Voltage Variation
When the ESS operates under the discharging condition, the active current is injected into the three-phase grid from the ESS. In this case, the PCC voltages are changed by the phase of the reactive current injected into the three-phase grid.
Figure 8 shows the variations of the PCC voltage in the discharging condition of the ESS depending on the phase of the reactive current.
VG is the grid voltage, and
IG is the grid current flowing between the ESS and the three-phase grid. In this paper, when
IG is positive, it flows to the three-phase grid from the ESS. The
d-axis refers to the reactive component, and the
q-axis refers to the active component.
VPCC is the PCC voltage, which is the sum of
VG and the voltage drops (
VRG and
VLG) of the resistor–inductor of the three-phase grid. In
Figure 8a,
IG includes only the active current, which means that the reactive current is zero.
VPCC is occurred by
VG,
VRG, and
VLG, and it has a phase equal to that of
IG. When the inductive or capacitive reactive current is injected into the three-phase grid,
VPCC is changed by the variation of the phases of
VRG and
VLG, as shown in
Figure 8b,c. In other words,
VPCC decreases because of the injection of the inductive reactive current, and it increases because of the injection of the capacitive reactive current. As a result, in the discharging condition of the ESS, the active current with the capacitive reactive current injected into the three-phase grid contributes to an increase of
VPCC and meets the LVRT requirement.
When the ESS operates under the charging condition, contrary to the discharging condition, the active current flows to the ESS from the three-phase grid.
VPCC is also changed by the phase of the reactive current flowing to the ESS. The variations of
VPCC in the charging condition of the ESS depending on the phase of the reactive current is shown in
Figure 9. In
Figure 9a,
IG includes only the active current without the reactive current. It does not contribute to the increase of
VPCC because
IG has a negative direction. In
Figure 9b,c,
VPCC is changed by the injection of the inductive and capacitive reactive currents, respectively.
VPCC decreases when the inductive reactive current is injected, and it increases when the capacitive reactive current is injected. Therefore, in the charging condition of the ESS, the active current with the capacitive reactive current contributes to the LVRT requirement.
When the grid voltage drops in the grid-connected ESS, the proposed LVRT control strategy is required to comply with the LVRT requirement. The proposed LVRT control strategy determines the injection quantity of the active and reactive currents, and the strategy depends on the voltage drop ratio of the three-phase grid. In addition, regardless of the operating condition of the ESS, such as the charging and discharging conditions, the injection of the capacitive reactive current contributes to the LVRT requirement because of the increase of VPCC. The validity of the proposed LVRT control strategy is demonstrated and the variations of the PCC voltage of the grid-connected ESS are analyzed by simulation and experimental results.
3.3. LVRT Control Strategy Depending on State of Charge of Battery
Contrary to the LVRT control strategy for the WPG and the SEG systems, in the grid-connected ESSs with a battery in the DC-source, shown in
Figure 4, a state of charge (SOC) should be considered for the LVRT control strategy. The SOC is an essential indicator used to regulate the operating decisions and to avoid the over-charge or over-discharge. However, it cannot be measured directly by sensors. In general, the SOC is obtained by the battery management system or various algorithms for estimation of the SOC using the battery model [
37,
38,
39].
In case the SOC is higher or lower than the designated value under the discharging or charging conditions of the grid-connected ESS, the proposed LVRT control strategy, which is mentioned above can be applied. However, in case the SOC is lower or higher than the designated value under the discharging or charging conditions of the grid-connected ESS, the SOC should be considered in the grid-connected ESS with the LVRT control strategy. In other words, if the SOC is lower than the designated value under discharging conditions of the grid-connected ESS, the active current cannot be supplied to the three-phase grid. The other way, if the SOC is higher than designated value under charging conditions, the active current is not required. Therefore, regardless of the grid voltage drops, the grid-connected ESS controls the reactive current as rating current.
4. Simulation Results
To verify the performance of the proposed LVRT control strategy and analyze the variations of the PCC voltage, a simulation that uses the grid-connected ESS as shown in
Figure 4 was conducted using the PSIM software. The simulation parameters are listed in
Table 1. DC-link voltage (
VDC) generated by the battery was 600 V, and the three-phase grid line-to-line voltage (
VG) was 60 Hz/380 V
rms. In addition,
Lf was 3 mH and
RG and
LG were 0.05 Ω and 0.04 mH, respectively.
Figure 10 shows the simulation results of the voltage-level calculation process (given in
Figure 6) that uses the three-phase grid voltages
VR,
VS, and
VT. Under normal conditions,
VR,
VS, and
VT are 60 Hz/310 V
peak. However, in
Figure 10a, the grid voltage drops, and the magnitudes of
VR,
VS, and
VT decrease to 80%, 60%, and 40%, respectively, of the magnitudes of the phase voltages under normal conditions for intervals ranging from 0.1 to 0.5 s. After that, the magnitudes increase to 70% and 100% as compared with the magnitudes of the phase voltages under normal conditions. The magnitude (
Vmag) of
VR,
VS, and
VT is precisely calculated using the voltage-level calculation process as shown in
Figure 6. Moreover,
VLEVEL as the voltage-level of the three-phase grid is determined depending on
Vmag (as shown in
Figure 10b), and
VLEVEL is used for the detecting the LVRT requirement.
Similar to the scenario in
Figure 10,
Figure 11 shows the simulation results of the proposed LVRT control strategy depending on the detecting the LVRT requirement based on the voltage-level.
Figure 11a shows
VLEVEL of the three-phase grid when the voltage drops (as in
Figure 10).
Figure 11b,c show the
d-axis and
q-axis currents (
Ide and
Iqe) and the reference currents (
I*de and
I*qe) of the synchronous reference frame.
Ide and
Iqe stand for the reactive and the active currents in the synchronous reference frame, respectively.
I*de and
I*qe are determined by the proposed LVRT control strategy depending on the detecting the LVRT requirement using
VLEVEL. In other words, the injection quantity of the active and reactive currents is determined from the voltage drop ratio of the three-phase grid using the method as shown in
Figure 5.
If
VLEVEL is 80% during the interval from 0.2 to 0.3 s in
Figure 11a,
I*de as the injection quantity of the reactive current is determined to 40% (approximately 4.28 A) of
Irating in the grid-connected ESS as shown in
Figure 11b. In addition, if
VLEVEL is lower than 50% during the interval from 0.4 to 0.5 s in
Figure 11a,
I*de is determined to 100% (approximately 10.7 A) of
Irating in the grid-connected ESS as shown in
Figure 11b. After determining
I*de using the proposed LVRT control strategy,
I*qe as the injection quantity of the active current is calculated using
I*de and
Irating.
As a result, although the grid voltage drops,
Ide and
Iqe are controlled by
I*de and
I*qe, respectively, using the proposed LVRT control strategy. The proposed LVRT control strategy complies with the LVRT requirement, and the three-phase grid currents (
IR,
IS, and
IT) maintain a sinusoidal waveform, as shown in
Figure 11d.
In this paper, an additional simulation was performed to analyze the variations of the PCC voltage depending on the phase of the reactive current injected into the three-phase grid.
Figure 12 and
Figure 13 show the simulation results of the PCC voltage depending on the phase of the reactive current when the ESS operates under the discharging and charging conditions, respectively. In
Figure 12 and
Figure 13, the magnitudes of
VR,
VS, and
VT are decreased to 60% as compared with those of phase voltages under normal conditions. Therefore,
I*de was determined to 80% (approximately 8.56 A) of
Irating in the grid-connected ESS through the proposed LVRT control strategy.
I*qe as the reference active current is calculated according to (1) as approximately 6.42 A. The
Iqe values as shown in
Figure 12c and
Figure 13c are controlled by
I*qe (6.42 A in
Figure 12c and −6.42 A in
Figure 13c), depending on the operating conditions of the ESS, such as the discharging and charging conditions.
As shown in
Figure 12b and
Figure 13b, the inductive reactive current is injected into the three-phase grid during the interval from 0.3 to 0.5 s, and the capacitive reactive current is injected during the interval from 0.5 to 0.7 s. Depending on the phase of the reactive current injected into the three-phase grid, the PCC voltage (
VPCC) is changed as shown in
Figure 12a and
Figure 13a with an extended simulation waveform. Regardless of the operating conditions of the ESS, the
VPCC magnitude is higher when the capacitive reactive current is injected as compared with the
VPCC magnitude when the inductive reactive current is injected. Therefore, when the grid voltage drops, the capacitive reactive current determined by the proposed LVRT control strategy needs to be injected into the three-phase grid. This contributes to the PCC voltage increase. The three-phase grid currents (
IR,
IS, and
IT) are shown in
Figure 12d and
Figure 13d.
Figure 14 and
Figure 15 show the simulation results of the active and reactive currents of the grid-connected ESS depending on the SOC under discharging and charging conditions. In the discharging and charging conditions of the grid-connected ESS, the proposed LVRT control strategy is applied when the grid voltage drops. However, in
Figure 14, the SOC of the battery in the DC-link is changed from 20 to 10 at 0.3 s. It is lower than designated value, which is decided to 15. Additionally, in
Figure 15, the SOC of the battery is changed from 80 to 90 at 0.3 s. It is higher than designated value, which is decided to 85. Fundamentally, the SOC of the battery cannot change as immediately as a step function because it is a dynamic variable. However, it was simulated to show the performance of the proposed algorithm depending on the SOC of the battery. In these cases, regardless of the grid voltage drops, the grid-connected ESS controls the reactive current as rating current.
5. Experiment Results
To verify the validity of the proposed LVRT control strategy and analyze the variations of the PCC voltage, experiments were performed using the experimental setup as shown in
Figure 16. In the experimental setup, the DC power supply (TC.GSS.20.600.4WR.S) that has characteristic of the bidirectional power flow is used for the battery as shown in
Figure 4. The experimental setup comprised a DC-link, control board, relay, switched mode power supply (SMPS), fan, power board, and magnetic contactor. The SMPS supplies the power to operate the control board and power board. The control board comprised a digital signal processor that used TMS320F28335, and the power board consisted of an inverter stage using the IGBTs and gate drivers. The parameters of the experiment were equal to those of the simulation listed in
Table 1.
Figure 17 shows the experimental results of the voltage-level calculation process. Under normal conditions,
VR as the
R-phase grid voltage is 60 Hz/310 V
peak. However, in
Figure 17, the grid voltage drops. The
VR magnitude decreases from 100% to 80% and further to 60% and 40%; subsequently, it increased to 70% and 100% as compared with the phase voltage magnitude under normal conditions.
Vmag as the magnitude of the three-phase grid voltages is precisely calculated. In addition,
VLEVEL is determined, and it was used for detecting the LVRT requirement.
In a scenario similar to that shown in
Figure 17, the experimental results of the proposed LVRT control strategy depending on the detecting the LVRT requirement are shown in
Figure 18. In
Figure 18,
Ide and
Iqe, which are the reactive and active currents, respectively, are changed by the detecting the LVRT requirement using
VLEVEL and the proposed LVRT control strategy. When
VLEVEL is 80%,
Ide is injected into the three-phase grid with 40% of
Irating as the rating current in the grid-connected ESS. If
VLEVEL is lower than 50%,
Ide is determined to 100% of
Irating. In addition,
Iqe is determined by
Ide and
Irating after determining
Ide using the proposed LVRT control strategy. As a result, when the grid voltage drops, the proposed LVRT control strategy complies with the LVRT requirement, and
IR as the
R-phase grid current is maintained in a sinusoidal waveform, as shown in
Figure 18.
Figure 19 and
Figure 20 show the experimental results of the PCC voltage depending on the phase of the reactive current when the ESS operates under the discharging and charging conditions, respectively. In
Figure 19 and
Figure 20, the magnitudes of the three-phase grid voltages are decreased to 60% as compared with those of phase voltages under normal conditions. Therefore,
Ide is determined to 80% (approximately 8.56 A) of
Irating in the grid-connected ESS using the proposed LVRT control strategy. Additionally,
Iqe is approximately 6.42 A according to (1).
Figure 19 shows that
Iqe is maintained at 6.42 A because the ESS operates under the discharging condition.
Ide is changed to 8.56 from −8.56 A, which means that the reactive current injected into the three-phase grid is changed to the capacitive reactive current from the inductive reactive current. In
Figure 20,
Iqe is maintained at −6.42 A because the ESS operates under the charging condition.
Ide is changed similar to that in
Figure 19. Depending on the phase of the reactive current injected into the three-phase grid,
VPCC is changed. Regardless of the operating conditions of the ESS, the
VPCC magnitude is larger when the capacitive reactive current is injected as compared with the
VPCC magnitude when the inductive reactive current is injected. In other words, the capacitive reactive current helps in increasing
VPCC regardless of the operation conditions. Therefore, when the grid voltage drops, the capacitive reactive current should be injected into the three-phase grid using the proposed LVRT control strategy; this contributes to an increase of the PCC voltage.