Distributed Energy Management in Smart Thermal Grids with Uncertain Demands

Abstract

Smart Thermal Grids (STGs) represent a new concept in the energy sector that involves the use of the smart grid concept in thermal energy networks connecting users, such as households, buildings and greenhouses, to each other via a transport line of thermal energy. In this concept, there exists an energy management system that aims to improve the efficiency, reliability and sustainability of the energy production and the distribution of energy. This highlights the necessity of constructing a high level control unit which sets the operating points of the production units such as boilers, micro Combined Heat Power (CHP) generators, and chillers for every agent. In this thesis, we develop a framework for the energy management which incorporates the model of the smart thermal grid and the energy demand profile of the agents. This framework is based on a Model Predictive Control (MPC) strategy in which the grid is addressed as a large-scale uncertain system. We formulate a mixed-integer chance-constrained optimization problem for the planning of the operation of the production units for all agents in the grid in the presence of uncertain thermal energy demand profile. In order to deal with the chance constraints together with integer variables, the robust randomized method, which was particularly developed for this problem, is employed. This technique allows us to handle mixed-integer problems and stochastic programming in a unified framework and provide a-priori probabilistic guarantees for the obtained solution. Motivated by the need for a more flexible and scalable framework, we then extend this method to distributed computation schemes using the Alternating Direction Method of Multipliers (ADMM). The resulting performance enhancement in terms of the total operational costs observed in the simulation was substantial and comparable with the centralized control approach. Finally, we investigate the opportunity of improving the efficiency of energy usage when seasonal storage systems exist in the thermal grids. We first propose a dynamic model for the seasonal storage systems and then, incorporate it in the energy management problem formulation. Due to the annual cyclic dynamical behavior of the seasonal storage systems, this leads to a multi-rate albeit very complex optimization problem. To this end, we develop a hierarchical MPC to solve such problem together with a discussion on the resulting optimization problems in a receding horizon setting. The technical developments were validated on a realistic benchmark problem (three-agent thermal grid in Utrecht, The Netherlands). The simulation results show that the proposed method was able to provide better usage of the seasonal storage.