Risk Assessment of AC/DC Hybrid Distribution Network considering New Energy and Electric Vehicle Access

Jie Dong, Jianjun Zhao, Ziyue Liu, Yang Li, Zhongshu Yu, Jialin Liu, Zhenhao WangFrontiers in Energy Research 2022 · DOI

Abstract

In view of the operational risk issues such as safety and economy caused by the connection of new energy and multiple types of electric vehicles (EVs) to the AC–DC distribution network, an AC–DC distribution network operational risk assessment method that takes into account multiple risk factors is proposed. First, a probability distribution model of scenery output and EV timing is constructed, and the starting charge state of multiple types of EVs is replaced by the number of daily driving miles subjectively set; then, based on the complex network theory, timing safety indicators, such as voltage overrun risk and branch power overload operation risk, are proposed, and the economic risk is established according to the economic operation of the distribution grid; Furthermore, a risk assessment matrix for grid-connected EVs with different capacities is constructed, and the principal component analysis (PCA) method is used to reduce the dimension of the risk assessment matrix and calculate the objective weight coefficient; finally, taking the improved IEEE 33 node AC/DC power distribution system as an example, the comprehensive risk evaluation based on PCA is compared with the traditional one, and the results show that when the safety and economic risk factors are considered at the same time, the operation risk in a certain range has a downward trend when the proposed method is adopted, which has a positive guiding significance for the planning of EV capacity in a certain area.

Introduction

As global energy systems shift toward sustainability, distribution networks are evolving from traditional AC systems into complex AC/DC hybrids. This transition facilitates the integration of Distributed Generation (DG), such as wind and solar power. However, it also introduces significant operational challenges. The intermittent nature of renewable energy, combined with the volatile charging behavior of various electric vehicle (EV) types—including taxis, buses, and private cars—creates unpredictable loads that can lead to voltage overruns and branch power overloads. Traditional deterministic models often fail to capture these probabilistic risks, necessitating a more robust, data-driven assessment framework.

Methodology

To address these challenges, this paper proposes a multi-layered risk assessment framework. The approach begins with probabilistic modeling of renewable outputs using Weibull distributions for wind speed and Beta distributions for solar intensity. For EVs, the researchers model the "Constant Current–Constant Voltage" (CC–CV) charging process to capture realistic load profiles rather than assuming a simple flat load.

Calculation process for the distribution network operation risk assessment.

The framework utilizes complex network theory to determine "node importance," ensuring that faults at critical hubs are weighted more heavily than those at peripheral branches. Risk indicators are categorized into safety (voltage and power overloads) and economy (line loss and operational profit/loss). Finally, Principal Component Analysis (PCA) is applied to reduce the dimensionality of the risk matrix, providing an objective evaluation index that removes subjective bias in how different risks are weighted.

Experimental Setup

The proposed method was validated using an improved IEEE 33-node AC/DC distribution system. This model was modified to include Voltage Source Converters (VSCs), wind turbines, and solar arrays to simulate a modern hybrid grid environment. EV travel patterns and charging times were modeled based on 2017 and 2018 National Family Travel Survey (NHTS) data to realistically reflect daily mileage and initial charging states.

Topology of the improved IEEE 33-node AC/DC distribution system.

Key Insights and Results

The study revealed several critical findings regarding the integration of new energy and EVs:

  • Stochastic vs. Deterministic Modeling: Traditional deterministic models were found to be overly optimistic. In the test case, a deterministic model identified only 10 nodes with voltage issues, while the proposed stochastic model revealed that nearly the entire network (nodes 4–18) faced some level of voltage overrun risk.
  • The 12 MW "Sweet Spot": Interestingly, the comprehensive risk did not increase linearly with EV capacity. At approximately 12 MW of EV integration, the risk actually trended downward, suggesting that balanced EV loads can mitigate some grid stresses when synchronized with renewable output.
  • Temporal Risk Peaks: The highest operational risks occurred between 20:00 and 22:00, coinciding with the overlap of conventional peak loads and EV charging cycles.
  • Information Retention: The PCA method successfully condensed the complex risk data with a 96.28% information retention rate, identifying node voltage violation as the most critical stability factor for hybrid systems.

Conclusion

The proposed PCA-based risk assessment provides a robust tool for grid operators to manage the complexities of modern hybrid networks. By objectively weighing safety and economic factors, the model identifies optimal EV integration levels that minimize overall grid stress. Future work will explore the potential of Vehicle-to-Grid (V2G) technology to further mitigate risks and enhance grid resilience by utilizing EVs as mobile storage units.

Summary updated

Abstract

In view of the operational risk issues such as safety and economy caused by the connection of new energy and multiple types of electric vehicles (EVs) to the AC–DC distribution network, an AC–DC distribution network operational risk assessment method that takes into account multiple risk factors is proposed. First, a probability distribution model of scenery output and EV timing is constructed, and the starting charge state of multiple types of EVs is replaced by the number of daily driving miles subjectively set; then, based on the complex network theory, timing safety indicators, such as voltage overrun risk and branch power overload operation risk, are proposed, and the economic risk is established according to the economic operation of the distribution grid; Furthermore, a risk assessment matrix for grid-connected EVs with different capacities is constructed, and the principal component analysis (PCA) method is used to reduce the dimension of the risk assessment matrix and calculate the objective weight coefficient; finally, taking the improved IEEE 33 node AC/DC power distribution system as an example, the comprehensive risk evaluation based on PCA is compared with the traditional one, and the results show that when the safety and economic risk factors are considered at the same time, the operation risk in a certain range has a downward trend when the proposed method is adopted, which has a positive guiding significance for the planning of EV capacity in a certain area.
BibTeX
@inproceedings{dong2022risk,
  author = {Jie Dong and Jianjun Zhao and Ziyue Liu and Yang Li and Zhongshu Yu and Jialin Liu and Zhenhao Wang},
  title = {{Risk Assessment of AC/DC Hybrid Distribution Network considering New Energy and Electric Vehicle Access}},
  booktitle = {Frontiers in Energy Research},
  year = {2022},
  doi = {10.3389/fenrg.2022.816246},
}
Figure