Reactive Power Management and Flexibility Aggregation in Active Distribution Networks

Power systems are experiencing significant changes due to the ongoing energy transition. These changes impact all areas of power systems: generation, transmission, distribution, and consumption. For example, there is a shift from large centralized fossil fuel-based power plants connected to the transmission system to small-scale environmentally friendly distributed energy resources connected to the distribution system. On the consumption side, there are new types of power electronics (PE) loads. Electric vehicles (EVs), heat pumps (HPs) and data centers are examples of such loads. In the power transmission and distribution sector, the use of underground cables is increasing to enhance reliability and minimize their impact on the scenery. All of these changes, along with their benefits, bring new challenges to the operation of power systems. One of such challenges is capacitive reactive power (CRP) flow in distribution networks. This thesis begins by introducing the issue of CRP flows, presenting actual measurement data from the Dutch distribution system operator (DSO) and the transmission system operator (TSO). The data indicates a downward trend in reactive power flowing from the transmission system to the system, along with a rise in reactive power flow in the opposite direction. Chapter 2 provides the causes of CRP flows, the reactive power limits specified by the European Network of Transmission System Operators for electricity (ENTSO-E) and the Dutch electricity regulator Authority for Consumers and Markets (ACM). Furthermore, this chapter examines the impact of CRP flows on network parameters, such as voltages and line losses, as well as on the reactive power limits specified in grid codes. The need for TSO-DSO coordination to address CRP flows and other challenges in power systems is emphasized. Chapter 3 presents the control scheme for the smart photovoltaic (PV) inverters connected in the distribution network to manage CRP flows. PV inverters generate active power during the day and remain idle during nighttime. The remaining available capacity of smart PV inverters apart from the capacity used for active power generation could be considered as an asset to handle CRP flow issues. Smart PV inverters with the proposed control scheme can work as a partial static synchronous compensator (STATCOM) during the day and as a full STATCOM during the nighttime. For the nighttime operation of the proposed control scheme there is a need to have an extra charging resistor and disconnecting switch between DC-link and solar panels to avoid the damage caused by the flow of inrush current. By controlling smart PV inverters in the distribution network in the proposed manner, the reactive power flow can be managed at the network boundaries. Active and reactive power flexibility available from the flexibility providing units (FPUs) in active distribution network (ADN) could be used to provide services. The flexibility provided by an individual FPUs is often insufficient, making aggregation necessary for more effective utilization. Feasible operation region (FOR) is an approach that represents the feasible aggregated flexibility of a network in the form of a P −Q chart. The FOR of an ADN is influenced by factors such as the network topology and the tap position of the transformer which are collectively referred to as network settings in this thesis. In Chapter 4, we analyze how network settings can influence the size and shape of FOR of ADN. Moreover, to aid network operators in decision making network settings can be ranked according to the size of the FORs obtained. This ranking can act as a decision tool for operators to properly use flexibility available from an ADN. The FOR of a network is fundamentally influenced by the nodal voltage limits and the cable loading constraints. Chapter 5 presents the proposed mathematical analysis which can represent the nodal voltage limits and cable loading constraints in the form of interconnection power flows (IPFs) and power injections. This analysis is used to examine the impact of nodal voltage limits and cable loading constraints on FOR of the network. The results of the analysis show that the impact of the nodal voltage limits on the FOR boundary can be represented in the form of a straight line and the impact of cable loading constraints can be represented by sets of circles. Chapter 6 presents an analytical method to find accurate boundaries of the FOR under nodal voltage constraints. The proposed method consists of a bucket filling sampling method, a first order loss approximation of the power flow and an analytical analysis. The proposed method is validated with numerical simulations for the LV and MV radial distribution networks. The results show that the proposed method can accurately describe the nodal voltage boundaries of the FOR of an ADN.FOR of a network may become non-convex when the system is operated under stressed conditions. Chapter 7 introduces a methodology for aggregating FORs from different distribution networks at their vertical interconnection. The proposed approach is capable of handling both convex and non-convex FORs. The methodology is based on a decomposition based Minkowski addition of convex hulls. For non-convex input FORs, the method partitions them into convex subregions, aggregates these subregions, and applies Minkowski summation to their hulls to obtain the aggregated FOR. The methodology was validated on a standard distribution network model incorporating both loads and generators, and the results demonstrate its effectiveness in accurately aggregating flexibility across multiple networks. Chapter 8 presents conclusions and recommendations for future work. Overall, this thesis provides reactive power management and flexibility aggregation methods for distribution networks. The contributions made in the thesis are twofold. Firstly, it introduces the issue of CRP flows, examines their effects on distribution networks, and proposes solutions for reactive power management, including the use of PV STATCOM. Secondly, it explores the FOR as a tool for network aware flexibility aggregation, analyzing how network constraints and network settings impact its size and shape. Furthermore, it also explores how to aggregate FORs coming from different system operators without the need for their detailed network information.