Edge and fog computing using IoT for direct load optimization and control with flexibility services for citizen energy communities

The grid congestion and imbalance problems are nowadays approached through a new perspective of the energy flexibility services that are usually acquired through a bidding process by aggregators on behalf of grid operators and electricity retailers and traded via Local Flexibility Markets (LFM). They can be also acquired at a fixed price under specific conditions of interruption and operation. The flexibility services are offered by electricity consumers and prosumers that own generation and storage facilities and controllable appliances, adding value to the emerging Information and Communications Technology (ICT) and smart metering technologies. In this paper, we propose an adaptive direct load optimization and control with an Internet of Things (IoT) architecture and compare it with a classic approach of Direct Load Control (DLC). The optimization process consists in a day-ahead scheduling that aims to minimize the electricity expense using programmable appliances (shift) and their operational constraints, whereas the adaptive DLC comes on top of it with additional load control (shed) using flexibility to cope with real-time uncertainties including the temperature comfort of the consumers. As the consumption is recorded using smart meters and appliances that continuously generate large volumes of data at different time resolution and formats requiring fast analytical and decisional processes, the challenge consists in correlating and integrating the optimization requirements with IoT and appliance control within an edge and fog computing architecture to overcome grid congestion, power capacity scarcity and forecast errors. The simulations are performed using real open datasets consisting in 114 single-family houses that form a small community with modern and flexible appliances, providing that the electricity bill reduction is up to 22.62%. Furthermore, to validate, several indicators for consumers and aggregator are proposed: total daily used flexibility decreased on average by 21.05%, number of interruptions also decreased on average by 20.51%, maximum number of interruptions per appliance decreased by 58.33%, while Peak to Average Ratio (PAR) improved by 32% when implementing the proposed DLC architecture. (C) 2021 Elsevier B.V. All rights reserved.