Research on frequency regulation strategy of battery energy storage system supporting power system

Due to the large-scale grid connection of new energy, the inertia of the power system has decreased, seriously affecting the frequency stability of the power grid, and there is an urgent need for effective frequency support technology. In response to the above issues, this article proposes a frequency control strategy for battery energy storage systems to support power systems. Firstly, establish a battery equivalent circuit model to simulate the dynamic and static performance as well as external characteristics of the battery; Secondly, two frequency modulation strategies, droop control and accelerated droop control, are used to control the active power emitted by the battery energy storage system. Finally, a four-machine two-zone system with a 25% proportion of wind power was built using MATLAB for simulation analysis in two scenarios: sudden reduction of wind power and rapid fluctuation of wind power. The results showed that the frequency modulation strategy proposed in this paper can effectively improve the lowest and stable point frequencies of the system, and can slow down the rate of frequency decline, thereby improving the frequency stability of the power grid.


Introduction
In response to the dual carbon goal, China's new energy development momentum is rapid [1].The increase in the installed capacity of new energy generation and the gradual withdrawal of thermal power units will lead to a decrease in the inertia of the power system.Low inertia power systems will face challenges such as high-frequency change rates, large frequency deviations, distributed photovoltaic trips, and distributed generator trips.Therefore, the continuous integration of new energy poses a huge challenge to the frequency stability of the power grid, and it is necessary to upgrade and transform frequency active support technology to improve the frequency of the power grid [2].
Energy storage has become the most visionary and important resource for frequency regulation in the power system due to its strong new energy absorption capacity, long life cycle, and wide operating temperature range [3].At the same time, energy storage has the potential to become an excellent frequency assistance service method due to its fast response and continuous active output ability.Among the numerous existing energy storage methods, battery energy storage systems are suitable for power grid frequency regulation due to their technical characteristics such as fast response, stable performance, and flexible control.
At present, scholars both domestically and internationally have conducted extensive research on how to regulate the frequency of the power grid for battery energy storage.The joint participation of traditional thermal power units and energy storage systems in the frequency support process of the power grid is studied and models for analysis are established in [4], but there is a lack of research on frequency regulation strategies.The use of a droop control strategy to improve grid frequency is proposed in [5] but does not consider the State of Charge (SOC) of the battery itself.In [6], frequency modulation strategies are innovated, proposing the use of both droop control and virtual inertial control to further improve power grid stability.
Taking the above research results as a reference, this article researches the participation of batteries in frequency regulation of a high proportion of new energy grids.Firstly, a primary frequency modulation model for battery-based energy storage systems was established.Secondly, two control methods, droop control, and accelerated droop control, were proposed to enable the battery to quickly respond to changes in grid frequency and improve grid frequency stability.Finally, the effectiveness of the proposed method was verified through simulation.

Battery model
The commonly used mathematical models for batteries currently include internal resistance models, equivalent circuit models, genetic algorithm models, neural network models, and electrochemical models.Due to the relatively simple structure of the equivalent circuit model and the fact that the battery can reflect some of the characteristics of resistance and capacitance.Therefore, this article selects an equivalent circuit model to simulate the dynamic and static performance and external characteristics of batteries.The battery electromotive force can be expressed as: where E represents the internal potential of the battery, SOC is the state of charge of the battery; E0 is the initial internal potential, K is the polarization voltage constant; Au is the coefficient of voltage variation; Bc is the coefficient of change in battery capacity; Qn is the rated capacity of the battery; i(t) is the charging and discharging current; Ct is the polarization effect coefficient; Tb is the battery temperature; SOC(0) is the initial state of charge of the battery; Nb-s is the number of batteries in series in the battery pack.
Detecting the current state of charge of the battery can provide real-time feedback on the remaining capacity of the battery.The size of SOC directly reflects the operating status of the battery, whether it is overcharged, discharged, or in normal operation.The current integration method is currently one of the most commonly used SOC estimation methods in the field of battery management systems.By using the current integration method, the SOC at any time can be obtained by integrating the cumulative current over time and the initial value of the battery's state of charge.Therefore, the battery SOC can be expressed as: where SOC (0) is the initial state of charge of the battery; Nb-p is the number of parallel batteries in the battery pack.

Frequency response model
To analyze the primary frequency regulation characteristics of the power system, a frequency response model of the energy storage system containing batteries is established.Figure 1 shows a primary frequency regulation model of a system that takes into account battery energy storage and includes a governor, prime mover, and load.

Primary frequency modulation circuit
Figure1 Frequency response model of battery grid connected system.
Among them, M represents the equivalent inertia coefficient of the system, D represents the load damping coefficient; R represents the sag coefficient; Tt and Tg represent the time constants of the governor and prime mover, respectively, Δμ represents the increment of the governor; ΔPM represents the change in output of the prime mover, ΔPL represents the amount of load change, and in the study of large disturbance frequency events, step signals are usually used to represent it; ΔPbat represents the change in battery output.
When the system experiences frequency deviation Δf, the generator's primary frequency regulation circuit responds, and the governor provides an incremental signal to the prime mover Δμ, adjusting the output of the prime mover ΔPM.At the same time, the system load and battery output also vary with frequency deviation ΔPL and ΔPbat, the three calculate the system power variation, and then obtain a new frequency deviation through the inertia module Δf.During the frequency response process, ΔPbat is equivalent to the additional active power injected into the system.Studying frequency modulation strategies for ΔPbat control plays a crucial role in improving system frequency.

Participation of battery energy storage system in power grid frequency regulation strategy
The battery energy storage is connected to the DC bus through a bidirectional half-bridge DC converter, achieving bidirectional energy flow for battery charging and discharging.This article proposes two strategies for battery energy storage to participate in grid frequency regulation: droop control and accelerated droop control.Figure 2  At this point, the Equation for battery output is: where Pbat is the battery output, and P0 is the initial output value before the system disturbance occurs.
From the above equation, it can be seen that the droop control is essentially a proportional coefficient set by the energy storage equipment.When frequency deviation is detected, the active power proportional to the frequency deviation is transmitted to the power grid through the converter in response to the frequency change of the power grid.
Accelerated droop control improves the droop control by not changing its droop coefficient but locking in its maximum frequency deviation.We allow ΔPbat to constantly maintain the maximum absolute value to accelerate the change in battery output and achieve a faster response to frequency changes.At this point, the Equation for battery output is:

Simulation analysis
To verify the effectiveness of the control strategy proposed for the battery energy storage system in this article, a simulation model of the battery energy storage grid connected system was built in MATLAB simulation software.The power grid architecture model constructed in this article is a four-machine twozone system, and a thermal power unit with a capacity of 900 MW is replaced with a wind turbine unit.Among them, the wind turbine model consists of a three-phase voltage source and a converter, which are then connected to the power grid by a grid-connected inverter.The main simulation parameters of thermal power units are as follows:

Simulation analysis of the sudden reduction of wind turbines
In the system, the output of thermal power units is 0.778 p.u., 0.787 p.u., and 0.787 p.u. respectively, and the load size is shown in Table 2.The battery output is 300 MW, and the wind turbine output suddenly decreases by 300 MW at 5 s, jumping from 719 MW to 419 MW.The system frequency (detecting the frequency of thermal power unit 3) and battery output are shown in Figure 3: In Figure 3, when the system operates stably, the system frequency is stable at 60 Hz, and when the fan output suddenly decreases by 300 MW, the frequency will suddenly drop.When the frequency modulation strategy is not applied to the battery, the system frequency undergoes a secondary drop due to transient oscillations.After applying the frequency modulation strategy to the battery, the system will not experience a secondary frequency drop phenomenon.This is due to the additional output of the battery, which reduces the power imbalance in the system and weakens system oscillations.
In the three scenarios, the lowest point of the system frequency is 59.835 Hz, 59.850 Hz, and 59.850 Hz.Finally, the system transitions to another stable state, with system frequencies of 59.857 Hz, 59.904 Hz, and 59.930 Hz, respectively.Droop control can effectively improve the lowest point of frequency, increasing it by 0.015 Hz.After the system stabilizes, the droop control stabilizes the frequency at 59.904 Hz, which is 0.047 Hz higher than when there is no frequency modulation strategy.The acceleration droop control stabilizes the frequency at 59.930 Hz, which is 0.073 Hz higher than when there is no frequency modulation strategy.In summary, under this operating condition, both frequency modulation strategies proposed in this article can effectively improve frequency quality, and accelerate droop control, which reflects better frequency modulation characteristics compared to droop control.
As shown in the active power output image of the battery in Figure 3, before the wind turbine power suddenly decreases, the battery stably emits 300 MW of active power.After the sudden decrease in fan power, the droop control and acceleration droop control begin to change the battery output, and after the first peak, there is a control difference between the two.The acceleration droop control enables the battery to continuously emit more active power, which is also the reason why the performance of the acceleration droop control is better than that of the droop control in the scenario of sudden power reduction of the wind turbine.

Simulation analysis of continuous fluctuation of fans
Compared to the simulation analysis of the sudden reduction of the fan, this operating condition only changes the output status of the fan.The fluctuation of fan power, system frequency, and battery output are shown in Figures 4 and 5  In Figures 4 and 5, when the power grid operates stably, the frequency stabilizes at 60 Hz.When the power of the wind turbine is affected by uncontrollable factors and its output continues to slowly change, the frequency of the power grid also slowly changes.
When the grid frequency is below 60 Hz, the acceleration droop control performance is better, and the frequency is closer to 60 Hz.Once the frequency rises above 60 Hz, acceleration droops due to its characteristics.The battery output reference value is always calculated based on the maximum absolute frequency deviation, resulting in no decrease in battery output and a deviation in frequency.The following characteristics of droop control will not cause this problem.Then, when the frequency is stable, the battery adopts droop control and accelerated droop control, and the frequency stabilizes at 59.895 Hz, which is 0.05 Hz higher than the non-frequency modulation strategy.Both droop control and accelerated droop control can slow down the rate of frequency decrease.This indicates that the two frequency modulation strategies proposed in this article still have an improved effect on frequency, but in scenarios where the wind turbine continues to fluctuate, droop control is better than accelerated droop control.

Conclusion
This article designs two frequency control strategies for the battery, droop control and accelerated droop control, for the rapid adjustment of system frequency.A simulation model of a battery grid-connected system with a 25% wind power ratio was built in MATLAB simulation software.And simulation analysis was conducted on two scenarios: sudden reduction of wind turbine power and continuous fluctuation of wind turbine power, and the conclusions are as follows: 1) In both scenarios, this strategy enables the battery to change its active power output in a short period, not only increasing the frequency at the lowest point and achieving a new stable state but also slowing down the rate of frequency decline.In scenarios where wind power output suddenly decreases, the acceleration droop control frequency modulation effect is better, and in scenarios where wind power fluctuates continuously, the droop control frequency modulation effect is better.
2) In the scenario of a sudden reduction in wind power output, the control strategy proposed in this article can quickly reduce the imbalance in the power grid, avoid significant fluctuations in the power grid, and avoid secondary frequency fluctuations in the power grid.

Figure 2
Figure 2 Schematic diagram of battery frequency regulation strategy.

Figure3
Figure3 Simulation results of a sudden reduction of 300 MW in wind turbine power. .

Figure 4 Figure 5
Figure 4 Simulation diagram of grid connection of battery energy storage system.

Table 1
Parameter table of thermal power units.

Table 2
Load parameter table.