Integration of cost-based active management in distribution networks planning and operation

The pursuit of carbon neutrality is significantly advanced by the integration of distributed energy resources (DERs) and electrification. However, if not properly executed, this integration can introduce congestion and present technical challenges to the power system. Conversely, when integrated with careful planning and control, DERs can bring considerable advantages, including enhanced system flexibility. This thesis investigates the integration of cost-based active management within distribution network operations, examining the operational strategies of distribution system operators (DSOs) with a high penetration of DER.
This thesis begins by examining low-voltage distribution networks. In standard practice, DSOs typically have access to only a limited set of measurements. Compounding this challenge, delays and inaccuracies in meter readings often render them less informative for network operations. To pinpoint network congestion effectively, DSOs need to estimate network states even when observability is limited. Utilizing the matrix completion (MC) formulation, this work incorporates current measurements as additional constraints, enhancing estimation accuracy without enlarging the problem’s dimensions. As these current measurements offer a richer perspective on network states, they help diminish estimation uncertainty. Concurrently, the strategic placement of sensors can also influence this accuracy. Given budgetary constraints, allocating every minor branch with a sensor is impractical. Consequently, As a result, this thesis explores the consequences of sensor placements, setting aside the optimal sensor placement for future research.
Following the determination of network states, this thesis pivots to the modeling of aggregated flexibility derived from dispersed DERs within distribution networks. Considering the power capacity of individual flexible DER, the prevalent approach to represent aggregated flexibility is constructing feasible regions on the P-Q chart. This delineated region is characterized by network security boundaries, ensuring that all operational points contained within it comply with network constraints.
This work introduces an innovative methodology to discern such a region. It adeptly sidesteps the overestimation in existing optimization methods without the approximation of nonlinear power flow equations. Encompassing both theoretical development and algorithm design, the thesis lays out multiple sufficient conditions to examine the existence of this feasible space. Moreover, the proposed algorithm anchors itself on well-founded theorems for tangible application. Leveraging exclusively linear operations, the algorithm bypasses the prevalent non-convexity obstacles in optimization. This methodology equips DSOs with the tools to delineate network flexibility and the operating scope of each flexible resource in alignment with their specific network configurations.
Once the DSO determines both the network states and the feasible operational region, also known as aggregated flexibility, it is better equipped to manage the network and address potential congestion. Flexibility service emerges as a pivotal tool, assisting the DSO in navigating these challenges. However, given the multifaceted nature of service parameters and various network configurations, it’s imperative for the DSO to assess the impacts of flexibility services on the network state. Such assessment ensures that the DSO solicits the most efficacious and economically efficient service in the market.
Before studying this, the DSO must model the DER aggregator’s reaction to different services. Yet, bearing in mind the privacy concerns of DER customers, accurately modeling or anticipating consumption behaviors becomes challenging. This thesis introduces a top-down, optimization-centric framework that incorporates service constraints. Consequently, the resultant aggregator response profiles serve as valuable metrics for assessing the influence of flexibility service. Crucially, the data required for this model is gleaned from measurements available to the DSO.
Building on this foundation, the DSO is then able to evaluate the impacts of flexibility service. The thesis investigates the impacts of capacity limitation services, spotlighting pivotal parameters such as capacity limit values, service regions, and service durations. This evaluation adopts a probabilistic approach and is applied to a low-voltage distribution network marked by a high DER penetration. The analysis focuses on the impacts on network bus voltage, transformer loads, and potential rebound effects. For streamlined decision-making, a practical procedure is crafted for the DSO’s benefit.
In conclusion, this thesis provides the DSO with a set of tools from state estimation to service assessment, each dependent on distinct objectives and their combined insights. The DSO’s challenge lies in finding the ideal balance between enhancing network states and the associated costs of acquiring suitable services.