A distribution network serves as a critical infrastructure that delivers electricity directly to customers in a power system. Owing to the low-carbon transformation in energy field, the distribution network is developing into a public platform that fulfils diversified user demand and enables clean energy generation. Distribution network planning aims to satisfy the development of sources and loads while ensuring system security over a period by allocating various resources effectively and economically.

Background

In recent years, conventional distribution networks have been inundated with significant challenges. For example, renewable energy sources such as distributed photovoltaics (PVs) are widely integrated into distribution networks, and electric vehicles (EVs) are developing rapidly as an emerging load demand. Considering China’s statistics as an example, the cumulative installed capacity of distributed PVs increased by 46.61% year-on-year to 157.62 million kW in 2022, and the number of new energy vehicles reached 13.1 million, with a year-on-year increase of 67.13%. The radial structure of a conventional distribution network renders it difficult to manage the bidirectional power flow caused by large-scale distributed generators (DGs) and increased loads.  The conventional distribution network is threatened by security risks, such as voltage violations and line overloads. Therefore, a novel architecture for distribution networks and the corresponding planning methodology are required.

As a typical flexible distribution device, the soft open point (SOP) offers multiple advantages, such as spatial power flow regulation and real-time responses to variations. Based on SOPs, a flexible distribution network (FDN) has been established, which presents a promising architecture to accommodate diverse elements and address the source-load uncertainties in a more efficient and cost-effective manner. In FDNs, feeders do not operate in isolation; however, feeders that have complementary resources or suffer from violations can be interconnected flexibly. With the powerful regulation of SOPs, an FDN can operate in closed loop and exchange energy across regions. In addition, the common DC bus of SOP can serve as an interface for future expansion, which enables the topological evolution of FDN to satisfy the growth of sources and loads. Therefore, in contrast to conventional distribution networks, the FDN exhibits a meshed architecture characterised by flexible interconnections and expansions based on SOPs, thereby providing enhanced controllability and compatibility for emerging demands.

Motivation

The motivation behind this work is to explore and design an architecture of distribution networks based on SOPs with the integration of high penetration of DGs and flexible loads. The paper highlights the flexible regulation and interconnection capabilities of SOPs in spatial dimension, which enables an interconnected and extensible architecture for distribution networks. The flexible upgrading of the distribution network can enhance its energy management and DG hosting capability, making it a more cost-effective alternative to constructing new substations or feeders. Therefore, the specific questions that we aim to address are, how to develop a successive FDN planning strategy over a long duration, and how to determine the siting and sizing of SOPs, EV charging stations (EVCSs) and PVs simultaneously, while considering source-load uncertainties.

Contributions

In this paper, we propose a multi-resource dynamic planning method of FDNs, in which the configuration of SOPs, PVs and EVCSs is coordinated over a long-term planning period. The flexible reinforcement of the FDN can be implemented in multiple stages, and favourable cost benefits can be achieved compared with the traditional planning approach. In addition, a probabilistic framework is established to address the strong source-load uncertainties. The security risks are formulated by chance constraints, and the stochastic nonlinear optimisation model is effectively solved based on the modified iterative algorithm. By adjusting the acceptable violation probability in chance constraints, a trade-off between investment efficiency and operational security can be obtained.

Results

The flexibly interconnected and extensible architecture enables FDN to dispatch power flow over the entire system in closed-loop operation. This architecture is based on SOPs that provide strong controllability for wide-area active power transfer and local reactive power compensation. Consequently, the FDN offers a promising way to realize capacity expansion and low-carbon transformation in power systems with highly integrated PVs and EVs. The topology of distribution network is progressively updated and enhanced by segmenting the long planning period into several stages. The establishment of FDN will take years or even decades to fulfil the developing needs of users. A four-terminal SOP may evolve into a six- or eight-terminal structure that encompasses more power supply areas.