Multi-resource dynamic coordinated planning of flexible distribution network

Rui Wang, Haoran Ji, Peng Li, Hao Yu, Jinli Zhao, Liang Zhao, Yue Zhou, Jianzhong Wu, Linquan Bai, Jinyue Yan, Chengshan WangNature Communications 2024 · DOI

TL;DR

This paper proposes a novel multi-resource dynamic coordinated planning method for flexible distribution networks (FDNs) that integrates soft open points (SOPs), electric vehicle charging stations (EVCSs), and photovoltaic (PV) systems. By employing a probabilistic framework to handle source-load uncertainties and a multi-stage planning approach, the method aims to optimize investment and operational costs while ensuring grid security.

What Problem Does This Paper Address?

Modern distribution networks are facing significant challenges due to the increasing integration of distributed energy resources (DERs) like PVs and the rapid growth of new loads such as electric vehicles (EVs). The traditional radial structure of these networks struggles to manage the bidirectional power flow and inherent uncertainties introduced by these elements, leading to security issues like voltage violations and line overloads. Conventional planning approaches, which often involve costly infrastructure upgrades, are becoming increasingly unaffordable and inflexible. This paper addresses the need for a more dynamic, coordinated, and cost-effective planning strategy that can accommodate these changes while maintaining grid reliability.

Methods

The core of the proposed method is a multi-resource dynamic coordinated planning framework for FDNs. Key methodological components include:

  • Flexible Distribution Network (FDN) Modeling: This involves modeling the network's topology evolution over time, incorporating flexible interconnections enabled by Soft Open Points (SOPs).
  • Probabilistic Framework for Uncertainty: To address source-load uncertainties, the paper utilizes a chance-constrained programming (CCP) approach. This allows for a specified probability of constraint violations, enabling a trade-off between investment and operational security.
  • Uncertainty Quantification and Propagation: Gaussian Mixture Models (GMMs) are used to quantify uncertainties in loads and generation, and Nataf transformation is employed for uncertainty propagation.
  • Modified Iterative Algorithm: A novel iterative algorithm is developed to efficiently solve the stochastic optimization problem arising from CCP. This algorithm iteratively updates security constraint margins to achieve convergence.
  • Multi-Stage Planning: The planning horizon is divided into stages, allowing for a phased reinforcement of the distribution network and deferral of investment. SOPs, EVCSs, and PVs are planned and sized dynamically across these stages.

Datasets

The paper evaluates its proposed method using a modified practical distribution network case study. This network comprises 11 feeders of 11.4 kV and is planned over a 20-year horizon divided into four stages. Specific parameters for load growth, EV and PV penetration rates, and equipment costs are provided. No new benchmark datasets are introduced; rather, a realistic network scenario is used for validation.

Key Insights and Ideas

  • FDNs with SOPs offer enhanced controllability: SOPs enable flexible interconnections between feeders, allowing for better power flow regulation and resource sharing, which is crucial for managing DERs and EV loads.
  • Dynamic, multi-stage planning is crucial: Deferring investments and adapting the network topology over time, rather than a single large upfront investment, leads to significant cost benefits and improved resource utilization.
  • Probabilistic planning is essential for security: Directly addressing uncertainties through chance constraints allows for a more realistic assessment of risks and a controlled trade-off between investment costs and operational security.
  • Coordinated planning of resources is key: Simultaneously planning SOPs, EVCSs, and PVs, rather than optimizing them in isolation, leads to better overall system performance and cost-effectiveness.
  • The proposed iterative algorithm is efficient: The modified iterative algorithm significantly reduces computation time compared to traditional methods, making the complex optimization problem tractable.

Key Results

  • The proposed multi-resource dynamic coordinated planning method (Case I) demonstrated significant cost savings compared to traditional planning approaches (Case IV) and other scenarios, reducing total costs by up to 55.02%.
  • The multi-stage planning approach effectively defers investment, leading to improved cash flow and resource utilization.
  • By adjusting the acceptable violation probability ($\gamma$), a trade-off between investment efficiency and operational security can be achieved. Lower $\gamma$ values lead to earlier and more extensive SOP deployment, increasing investment but reducing operational penalties.
  • The proposed method successfully managed voltage violations and line overloads within predefined risk levels, ensuring the secure hosting of high PV and EV penetrations.
  • The modified iterative algorithm achieved convergence in 6 iterations requiring 9.40 hours, a substantial improvement over a general iterative algorithm that took 57.46 hours.

Limitations and Future Work

  • Limited scope of controllable resources: The current study focuses on SOPs, EVCSs, and PVs. Future work could incorporate other controllable resources like energy storage systems (ESSs) and demand response.
  • Game theory and Pareto optimization: While the paper frames the problem as a multi-objective optimization, it simplifies it into a single objective. Future research could explore Pareto frontier analysis or game theory to address multi-stakeholder interests more explicitly.
  • Reliability enhancement: The paper mentions that FDNs with SOPs can improve load recovery during failures. Further investigation into optimizing SOP transfer capacity for enhanced reliability is suggested.
  • Generalization of the FDN model: The paper lays the groundwork for "honeycomb FDNs," suggesting future work on more advanced, segmented grid structures.
  • Real-time operational considerations: While planning is addressed, real-time operational adjustments and dynamic control strategies for FDNs could be a future research direction.

Summary updated

Abstract

The flexible distribution network presents a promising architecture to accommodate highly integrated distributed generators and increasing loads in an efficient and cost-effective way. The distribution network is characterised by flexible interconnections and expansions based on soft open points, which enables it to dispatch power flow over the entire system with enhanced controllability and compatibility. Herein, we propose a multi-resource dynamic coordinated planning method of flexible distribution network that allows allocation strategies to be determined over a long-term planning period. Additionally, we establish a probabilistic framework to address source-load uncertainties, which mitigates the security risks of voltage violations and line overloads. A practical distribution network is adopted for flexible upgrading based on soft open points, and its cost benefits are evaluated and compared with that of traditional planning approaches. By adjusting the acceptable violation probability in chance constraints, a trade-off between investment efficiency and operational security can be realised.
BibTeX
@inproceedings{wang2024multi,
  author = {Rui Wang and Haoran Ji and Peng Li and Hao Yu and Jinli Zhao and Liang Zhao and Yue Zhou and Jianzhong Wu and Linquan Bai and Jinyue Yan and Chengshan Wang},
  title = {{Multi-resource dynamic coordinated planning of flexible distribution network}},
  booktitle = {Nature Communications},
  year = {2024},
  doi = {10.1038/s41467-024-48862-5},
}
Figure