In the context of the global energy transition toward low-carbon systems, the large-scale integration of renewable energy sources such as photovoltaic and wind power has introduced significant challenges to power system stability. The inherent low-inertia characteristic of inverter-based resources makes the grid more vulnerable to transient voltage disturbances. To address this issue, our research focuses on a hybrid energy storage system combining flywheel energy storage and battery energy storage systems, integrated with virtual synchronous generator (VSG) control. We aim to enhance the transient voltage support capability of high-penetration renewable energy grids. In this paper, we present our modeling, control strategy, and simulation results to demonstrate the effectiveness of the proposed approach.
The motivation for this work stems from recent large-scale blackout events, such as the Iberian Peninsula blackout in April 2025, which highlighted the risks associated with high shares of variable renewable generation. Traditional synchronous generators provide inertia and damping, but photovoltaic and wind plants lack such capabilities. Therefore, we need innovative solutions to mimic the behavior of synchronous machines. Virtual synchronous generator control is a promising grid-forming technique that emulates the swing equation and excitation system of a synchronous generator, thereby providing virtual inertia and reactive power support. By coupling VSG with hybrid energy storage, we can achieve fast and sustained voltage regulation during grid faults.
Virtual Synchronous Generator Control Principle
VSG control belongs to the grid-forming category, where the power electronic converter behaves as a controlled voltage source with inherent inertia and damping. The core of the VSG is the rotor swing equation of a synchronous generator:
$$
J\frac{d\omega}{dt} = \frac{P_m}{\omega} – \frac{P_e}{\omega} – D(\omega – \omega_n) \tag{1}
$$
where \(J\) is the virtual rotational inertia, \(\omega\) and \(\omega_n\) are the actual and rated angular frequencies, \(P_m\) is the virtual mechanical power, \(P_e\) is the output active power, and \(D\) is the damping coefficient. The primary frequency regulation is expressed as:
$$
P_m = P_{ref} + K_\omega (\omega – \omega_n) \tag{2}
$$
Combining (1) and (2) yields the active power-frequency control module of the VSG. For voltage regulation, the reactive power-voltage control is described by:
$$
E_n = (U_{ref} – U_{in})K_{ut} + (Q_{ref} – Q_{in})\frac{K_q}{s} + E_{empty} \tag{3}
$$
where \(E_n\) is the excitation electromotive force, \(U_{ref}\) and \(U_{in}\) are the reference and measured voltages, \(Q_{ref}\) and \(Q_{in}\) are the reference and measured reactive powers, \(K_{ut}\) and \(K_q\) are the voltage and reactive power droop gains, and \(E_{empty}\) is the no-load voltage.
Under the assumption of a purely inductive grid impedance \(X_f\), the active and reactive power outputs of the VSG can be approximated by:
$$
P_e \approx \frac{E_0 U_1}{X_f} \delta, \quad Q_e \approx \frac{E_0 U_1 \cos\delta – U_1^2}{X_f} \tag{4}
$$
where \(E_0\angle\delta\) is the converter output voltage and \(U_1\angle0^\circ\) is the grid voltage. The small-signal model of the VSG is used to select appropriate values for virtual inertia \(J\) and virtual impedance \(D\).
Photovoltaic Generation System Modeling
We employed the built-in photovoltaic module in MATLAB/Simulink. Since solar irradiance significantly affects the conversion efficiency, we implemented the perturb-and-observe (P&O) maximum power point tracking (MPPT) algorithm. The P&O method perturbs the output voltage or duty cycle and compares the power change before and after the perturbation to determine the direction toward the maximum power point (MPP). The control block diagram for MPPT is shown in Figure 1.
Figure 1 illustrates the MPPT control module used in our simulation. The algorithm provides a reference voltage that is tracked by the DC-DC converter to extract maximum power under varying irradiance.
Hybrid Energy Storage System Modeling
Flywheel Energy Storage Model
Flywheel energy storage systems store kinetic energy in a rotating mass. The key components include bearings, flywheel, motor/generator, vacuum enclosure, and power converter. Flywheels offer high efficiency, long cycle life, and fast response. The stored energy is expressed as:
$$
E = \frac{1}{2} J \omega^2 \tag{5}
$$
where \(J\) is the rotational inertia and \(\omega\) is the angular velocity. The usable energy range is determined by the difference between maximum and minimum speeds:
$$
\Delta E = \frac{1}{2} J (\omega_{max}^2 – \omega_{min}^2) \tag{6}
$$
The instantaneous power output is:
$$
P = \frac{d\Delta E}{dt} = J \omega \frac{d\omega}{dt} = T_e \omega \tag{7}
$$
We used a permanent magnet synchronous motor (PMSM) as the flywheel drive due to its high efficiency and wide speed range. The control module for the flywheel system is implemented in Simulink.
Lithium-Ion Battery Model
Battery energy storage systems, specifically lithium-ion batteries, are widely adopted for their high energy density, long cycle life, and low self-discharge. In our hybrid system, the lithium-ion battery complements the flywheel by providing sustained energy capacity, while the flywheel handles rapid power fluctuations. The battery is connected to the DC link via a bidirectional buck-boost converter. The control module for the battery is shown in Figure 2 (not illustrated in this text).
The combination of flywheel and battery energy storage systems creates a hybrid storage that leverages the fast response of the flywheel and the high energy capacity of the battery, effectively smoothing out power imbalances and supporting voltage regulation.
Simulation Results and Analysis
We conducted simulation experiments in MATLAB/Simulink with a total simulation time of 4 seconds. The irradiance profile was set as follows: 1000 W/m² from 0 to 1 s, rising to 1100 W/m² from 1 s to 2.5 s, and then dropping to 800 W/m² from 2.5 s to 4 s. The temperature was held constant at 25°C. The VSG was configured with an active power reference of 150 kW and a reactive power reference of 0 Var. Figure 3 shows the VSG output power waveforms, which closely follow the reference values.
To evaluate transient voltage support, we introduced a 5% grid voltage sag from 2 s to 3 s. During this period, the VSG increased its reactive power output to counteract the voltage drop, as shown in Figure 4. Tables 1 and 2 summarize the machine terminal voltage response with and without VSG control, respectively.
| Scenario | Time Interval (s) | Grid Voltage Sag Amplitude (V) | Terminal Phase Voltage Peak (V) |
|---|---|---|---|
| Normal operation | 0–2 | 0 | ±310.269 |
| Voltage sag | 2–3 | 15.514 | ±294.755 |
| Recovery | 3–4 | 0 | ±310.269 |
| Scenario | Time Interval (s) | Grid Voltage Sag Amplitude (V) | Terminal Phase Voltage Peak (V) |
|---|---|---|---|
| Normal operation | 0–2 | 0 | ±310.566 |
| Voltage sag | 2–3 | 0.797 | ±309.769 |
| Recovery | 3–4 | 0 | ±310.566 |
The results clearly indicate that without VSG control, the voltage sag amplitude reached 15.514 V, causing a significant drop in the terminal voltage. When VSG control was activated, the voltage sag was reduced to only 0.797 V, an improvement of approximately 95%. This demonstrates the effectiveness of our hybrid energy storage system combined with VSG control in providing robust transient voltage support.
Furthermore, the hybrid configuration of flywheel and battery energy storage systems ensures that the system can respond rapidly to disturbances (flywheel) and sustain support over longer durations (battery). This synergy is critical for maintaining grid stability under high-penetration renewable scenarios.
Conclusion
In this work, we have presented a comprehensive study on the use of hybrid energy storage systems—integrating flywheel and battery energy storage systems—with virtual synchronous generator control to enhance transient voltage stability in power systems with high renewable penetration. Our simulation results, under a 5% grid voltage sag condition, show that the proposed scheme reduces the voltage drop from 15.514 V to 0.797 V, a 95% improvement. The combination of fast-acting flywheel storage and high-capacity battery energy storage systems, coupled with the inertia and voltage support capabilities of VSG, provides an effective solution to the low-inertia challenges posed by large-scale renewable integration. Our findings offer a valuable technical reference for building resilient and stable future power systems.