Leveraging Consumers’ Flexibilityfor the Provision of Ancillary Services

Abstract

This thesis deals with the development and application of models and algorithms to leverage consumers’ flexibility for the provision of ancillary services (AS). The output of most renewable sources is intermittent and can only be predicted with a limited accuracy. Therefore, the increasing penetration of variable renewable sources leads to an unprecedented level of stochasticity and non-linearity in power system dynamics. Such complexities cause various operational challenges for power systems operators by requiring more AS resources. Larger integration of cost-effective renewable sources marginalises the operation of thermal power plants, due to the lower energy prices. As thermal power plants consist of the main source of AS, their retirement will intensify the lack of AS resources. Moreover, as renewable sources become widespread at the distribution level (e.g., with the installation of rooftop photovoltaic panels), AS requirements will extend to the distribution grids, which is unprecedented in existing power systems. For these reasons, it is necessary to look for alternative operational flexibility to serve as new AS for the sake of continuity and security of electricity delivery. In this regard, demand response is a valid solution that leverages demand flexibility to provide services to the grid. To optimally exploit consumers’ flexibility, it is important to account for consumers’ different preferences and constraints. Specifically, studies must approach the heterogeneity of loads and understand what influences consumers’ behaviour. Unfortunately, no study in the technical literature has discussed the aggregate potential of consumers’ flexibility, and estimation studies have been carried out only for specific types of loads. The first part of this thesis intends to fill this gap by proposing methodologies to estimate the potential of consumers’ This thesis deals with the development and application of models and algorithms to leverage consumers’ flexibility for the provision of ancillary services (AS). The output of most renewable sources is intermittent and can only be predicted with a limited accuracy. Therefore, the increasing penetration of variable renewable sources leads to an unprecedented level of stochasticity and non-linearity in power system dynamics. Such complexities cause various operational challenges for power systems operators by requiring more AS resources. Larger integration of cost-effective renewable sources marginalises the operation of thermal power plants, due to the lower energy prices. As thermal power plants consist of the main source of AS, their retirement will intensify the lack of AS resources. Moreover, as renewable sources become widespread at the distribution level (e.g., with the installation of rooftop photovoltaic panels), AS requirements will extend to the distribution grids, which is unprecedented in existing power systems. For these reasons, it is necessary to look for alternative operational flexibility to serve as new AS for the sake of continuity and security of electricity delivery. In this regard, demand response is a valid solution that leverages demand flexibility to provide services to the grid. To optimally exploit consumers’ flexibility, it is important to account for consumers’ different preferences and constraints. Specifically, studies must approach the heterogeneity of loads and understand what influences consumers’ behaviour. Unfortunately, no study in the technical literature has discussed the aggregate potential of consumers’ flexibility, and estimation studies have been carried out only for specific types of loads. The first part of this thesis intends to fill this gap by proposing methodologies to estimate the potential of consumers’ flexibility. To do so, we assume that consumers receive dynamic electricity prices and can autonomously schedule their consumption to minimise the electricity cost. Such studies consider several factors that influence consumers’ price responsiveness (i.e., loads’ rebound effect, outdoor temperature and electricity price) and accounts for a heterogeneous pool of consumers. Finally, these flexibility estimation models account for consumers’ stochastic behaviour toward prices. Consumers are effectively able to provide reliable services only if a proper framework is developed. Such a framework must satisfy different power systems’ requirements, e.g., the provision of services to different levels of the grid, as well as account for consumers’ preferences. The second part of this thesis discusses several alternatives to provide AS in smart grids. From the analysis, none of the existing solutions is capable to optimally leverage consumers’ flexibility. Therefore, we proceed in this research by proposing an innovative framework that can exploit consumers’ flexibility at different grid levels. This solution is based on a one-way communication structure and relies on dynamic electricity prices which are broadcast to consumers. In this thesis, the simulations of this proposed method are carried out to evaluate its potential in supporting frequency and voltage management.