Abstract:
This paper proposes a comprehensive control method for energy storage batteries (ESBs) to enhance their participation in primary frequency regulation of thermal power generators. The method optimizes the power output of ESBs based on primary frequency regulation characteristics and their state of charge (SOC). By analyzing grid frequency modulation characteristics, an improved virtual negative inertia control is derived. In the initial frequency regulation phase, ESBs primarily contribute through droop control, supplemented by inertia control. During the frequency recovery phase, inertia control dominates, with droop control providing support. To optimize ESB power output, an adaptive output coefficient adjustment strategy based on the logistic function is implemented to maintain a healthy SOC. Simulation results using a typical single-area grid model in MATLAB/Simulink demonstrate that the proposed control strategy rapidly and effectively reduces frequency deviations, enabling the power system to swiftly return to the primary frequency regulation steady-state value.
1. Introduction
The extensive use of fossil fuels significantly boosts societal productivity but also exacerbates environmental pollution. China’s “14th Five-Year Plan” emphasizes increasing the utilization of clean energy and achieving carbon peaking by 2030 . As the primary contributor to China’s carbon emissions, the thermal power industry necessitates a transition towards renewable energy sources for sustainable development. Although wind and solar power have immense potential for large-scale grid integration, they introduce new challenges to power system stability due to their lack of rotational inertia and intermittent nature .
Energy storage batteries (ESBs) have emerged as a viable solution to these challenges, exhibiting rapid response times and high regulation accuracy, thereby fulfilling grid frequency regulation requirements . ESBs collaborating with thermal power generators in frequency regulation significantly reduces generator output frequency, alleviates frequency regulation stress, enhances grid resilience against renewable energy intermittency, and minimizes wind/solar curtailment . Furthermore, this approach enhances generator parameters, lowers safety risks and economic losses, thereby safeguarding stable power system operation .
To improve ESBs’ control effectiveness in primary frequency regulation of thermal power generators, researchers have explored various aspects, including control models, joint output strategies, capacity allocation, and economic-security assessments . However, most existing studies focus on preliminary combinations of virtual inertia and droop controls, neglecting diverse control ratios’ impacts and utilizing fixed or linearly segmented SOC adjustment functions, thereby limiting ESB flexibility.
This paper introduces an adaptive control strategy for ESBs participating in primary frequency regulation of thermal power generators. Building upon the conventional virtual inertia control, a virtual negative inertia control is derived, dynamically allocating control proportion coefficients. The strategy seamlessly switches between droop, inertia, and negative inertia controls based on system frequency and deviation rate changes, while maintaining a healthy SOC to prevent overcharging/discharging. Simulation results validate the strategy’s effectiveness under step and continuous load disturbances.
2. Control Strategy for Energy Storage Batteries in Primary Frequency Regulation
2.1 Primary Frequency Regulation Model with Energy Storage Batteries
ESBs, renowned for their rapid response and precise regulation, enhance grid frequency quality when coupled with thermal power generators. This section integrates virtual negative inertia control into the conventional virtual inertia and droop controls to coordinate ESB output. the combined primary frequency regulation model.
The thermal power generator’s primary frequency regulation model comprises a governor and turbine model, as defined in equations (1) and (2), respectively. The ESB model, represented by equation (3), adopts a first-order inertia element for simulation simplicity.
2.2 Classical Control Strategies for Energy Storage Batteries
ESB frequency regulation strategies adjust output based on system frequency deviations and control coefficients returned by the grid [18]. The primary strategies are virtual droop control (equation 4) and virtual inertia control (equation 5).
- Virtual Droop Control: Mimics synchronous generators’ droop response to steady-state frequency deviations, effectively reducing frequency deviations.
- Virtual Inertia Control: Simulates synchronous generators’ inertia response, notably suppressing transient frequency changes.
2.3 Virtual Negative Inertia Control
Although virtual inertia control mitigates frequency deterioration during initial disturbances, it can hinder frequency stabilization during recovery, causing secondary frequency drops. To address this, virtual negative inertia control (equation 6) promotes frequency recovery during this phase, complementing droop control.
3. Energy Storage Battery Output Constraints
3.1 Deadband Settings
To minimize ESB wear, a separate deadband for ESBs (narrower than the generator’s) is set. This approach optimizes frequency regulation while prolonging generator life and reducing operation costs.
3.2 SOC-Based Output Constraints
To prevent ESB overcharging/discharging, output coefficients are dynamically adjusted using the logistic function (equations 7 and 8), considering SOC zones.
4. Comprehensive Control Strategy and Evaluation Metrics
4.1 Adaptive Control Strategy
The proposed adaptive control strategy dynamically selects and proportions control methods based on frequency deviation and rate of change. the control flow.
Control strategy equations (11) and (12) define power increments during different phases.
4.2 Evaluation Metrics
Evaluation metrics assess control performance under step and continuous load disturbances:
- Step Disturbance: Maximum and steady-state frequency deviations (Δf_m, Δf_s), time to maximum deviation (Δt_m), and time to steady state (Δt_s).
- Continuous Disturbance: Root mean square (RMS) of frequency deviation (R_f) and SOC (R_SOC).
5. Simulation Analysis
5.1 Simulation Model
A single-area grid model in MATLAB/Simulink is used for validation, with a 660 MW thermal generator and a 6 MW/6 MWh ESB. Simulations are conducted under step and continuous load disturbances.
5.2 Step Disturbance Analysis
Under a 0.01 p.u. step load disturbance, Compares frequency deviations and SOC changes for no ESB, droop control with fixed K, and the proposed adaptive strategy.
The adaptive strategy exhibits the best performance, with the smallest frequency deviations and fastest recovery, while maintaining a healthy SOC (Table 1).
Table 1: Evaluation Metrics for 0.01 p.u. Step Disturbance
| Method | Δf_m (p.u.) | Δf_s (p.u.) | Δt_m (s) | Δt_s (s) |
|---|---|---|---|---|
| No ESB | -1.62×10^-3 | -1.02×10^-3 | 8.5 | 18.3 |
| Droop | -0.81×10^-3 | -0.76×10^-3 | 7.6 | 17.7 |
| Adaptive | -0.74×10^-3 | -0.68×10^-3 | 7.9 | 15.6 |
5.3 Continuous Disturbance Analysis
Under a 200 s continuous 0.01 p.u. load disturbance, Table 2 present frequency deviation and SOC results.
Table 2: Evaluation Metrics for 0.01 p.u. Continuous Disturbance
| Method | R_f (p.u.) | R_SOC (p.u.) |
|---|---|---|
| No ESB | 4.7096×10^-4 | – |
| Droop | 3.3741×10^-4 | 3.7602×10^-5 |
| Adaptive | 1.6444×10^-4 | 4.2533×10^-5 |
The adaptive strategy outperforms the others, effectively maintaining SOC while optimizing frequency regulation.
6. Conclusion
This paper proposes an adaptive control strategy for ESBs participating in primary frequency regulation of thermal power generators. By integrating virtual droop, inertia, and negative inertia controls, the strategy rapidly reduces frequency deviations while maintaining a healthy SOC. Simulation results demonstrate the strategy’s superiority under step and continuous load disturbances, highlighting its practical applications in modern power systems.