Abstract
With the high penetrated photovoltaics (PVs) accessing distribution networks (DNs) in the future, the overvoltage of distribution network will be more serious, and the voltage control will be more complex. To solve this problem, a cluster partition-based zonal voltage control method for DN is proposed in this study. First, a cluster partition method using a comprehensive performance index is proposed: in terms of DN structure, a global density quality function index is introduced to measure the coupling degree of nodes in clusters. In terms of cluster function, the inconsistency coefficient index is presented to reflect the PV output characteristics, and a node membership function index is presented to limit the scale of the cluster. Then, the DN cluster partition is carried out under the fast Newman (FN) algorithm. Second, a second-order cone programming (SOCP) model of voltage control is established to maximize the PV consumption in each cluster. To achieve the coordinated optimization of clusters, an iterative optimization strategy among clusters is proposed. Finally, the effectiveness and rationality of the proposed method are verified in a 10 kV actual feeder system in Zhejiang Province, China.
Introduction
As the global energy landscape shifts toward carbon neutrality, the integration of rooftop solar photovoltaics (PVs) into distribution networks (DNs) is increasing rapidly. However, traditional distribution networks were not designed to accommodate significant bidirectional power flows. High PV penetration often leads to overvoltage conditions, where voltage levels at specific nodes exceed safety thresholds, risking equipment damage and grid instability.
Managing these voltage fluctuations centrally is computationally intensive, often resulting in high-dimensional models that are too slow for real-time response. Zonal voltage control—partitioning the grid into smaller, autonomous clusters—offers a scalable alternative. However, existing clustering methods frequently overlook the stochastic nature of PV output or fail to create balanced, functionally effective zones.
Methodology
The proposed approach utilizes a two-stage framework: Smart Partitioning followed by Iterative Optimization.
Cluster Partitioning
The researchers developed a comprehensive partitioning index based on three critical metrics:
- Global Density Quality: This measures the electrical coupling strength between nodes to ensure structural integrity.
- Inconsistency Coefficient: This index groups PV units with diverse output characteristics together. By pairing PVs that fluctuate differently (e.g., due to varying cloud cover), the total power fluctuation within a cluster is smoothed.
- Node Membership Function: This prevents the formation of excessively large or small clusters, maintaining computational efficiency.
The Fast Newman (FN) algorithm, a community discovery technique, is employed to execute the clustering based on these indices.
Voltage Control and Optimization
Once clusters are defined, a Second-Order Cone Programming (SOCP) model is applied within each zone. SOCP transforms complex, non-linear power flow equations into convex problems, allowing for rapid, reliable solutions that maximize PV consumption. To ensure global stability, an Iterative Coordination Strategy allows clusters to communicate, ensuring that local voltage corrections do not negatively impact neighboring zones.
Experimental Validation
The model was validated using an actual 10 kV feeder system located in Zhejiang Province, China. The test system comprises 34 nodes with a total installed PV capacity of 4 MW and 1.2 MW of energy storage (ESS). Simulations were conducted using the OpenDSS platform integrated with MATLAB.
Key Results and Insights
- Optimized Partitioning: The 34-node system was successfully divided into six functional clusters. This method produced more balanced zones with superior regulating capabilities compared to traditional modularity-based approaches.
- Voltage Regulation: In scenarios where peak solar output pushed voltages to 1.054 p.u. (surpassing the 1.05 limit), the zonal control successfully restored all nodes to safe operating levels.
- Maximizing Renewables: The system prioritized reactive power and battery storage to manage voltage, only curtailing PV output as a final resort, thereby maximizing renewable energy utilization.
- Computational Efficiency: The iterative strategy demonstrated fast convergence, proving its suitability for real-time grid management where conditions change rapidly.
Conclusion
This study demonstrates that a "divide and conquer" strategy using smart clustering and SOCP optimization can effectively manage the challenges of high PV penetration. Future research will focus on extending this model to three-phase unbalanced systems, accounting for communication latencies in real-world deployments, and integrating a wider variety of distributed energy resources, such as electric vehicle charging stations and flexible industrial loads.
Summary updated