Integrated design and operation optimization of multi-energy systems and energy networks with focus on the electric grid model

In the context of the green energy transition, Multi-Energy Systems (MESs) are one of the best routes to reduce carbon emissions and increase the conversion efficiency at competitive costs. However, the promising energy, environmental and economic advantages of these systems depend on their design and operational strategy, which must be optimized. Mixed Integer Linear Programming (MILP) optimization is an effective way of tackling this issue. To formulate the optimization problem, it is necessary to properly represent each element of the MES, namely the energy conversion and storage units, the end users and the energy networks, such as district heating and electrical grids. While the design and operation optimization of the energy conversion and storage units has found many accurate applications in the context of MESs, the optimization of the energy networks has been rarely applied and, when applied, many simplifications have been made to ensure the computational feasibility of the problem. This thesis focuses on modelling and optimizing the electrical network within a MES. The goal consists in identifying a model that can be integrated in the general design and operation optimization problem of a MES while ensuring a good tradeoff between results accuracy and computational simplicity. In the literature, several techniques are available for modelling the electrical network. However, many of them cannot be implemented in the MES optimization problem. For instance, techniques based on iterative approaches, such as the Newton-Raphson method, are inconsistent with numerical optimization, while Alternating Current Optimal Power Flow (ACOPF) frameworks involve nonlinear equations that, on the one hand, increase exponentially the computational complexity and, on the other hand, cannot be integrated in a MILP problem. To overcome these issues, a Direct Current Optimal Power Flow (DCOPF) model of the electrical network is proposed. Since the DCOPF model simplifies the relationship between voltage angles and active power flows, it can be integrated in a MILP framework. Moreover, in order to improve the accuracy of the results, the proposed modelling approach is supplemented with a formulation of the network losses, which are instead neglected in the classical DCOPF formulation. The proposed DCOPF model is initially validated with a reference network layout available in the literature. Subsequently, to prove its effectiveness, it is implemented in the optimization of a district MES and is compared with a baseline model of the electrical network considering only capacity constraints and neglecting the role of voltage angles . The objective function of the problem aims at minimizing the life cycle cost of the MES that is composed of an investment term, related to the installation of the energy conversion and storage units and energy networks, and an operational contribution, associated with the consumption of external energy carriers . Optimization results show how the model of the electrical network influences the MES design. Although the optimal value of the objective function does not vary sensibly, 0.125%, replacing the capacity-based model of the electrical network with the proposed DCOPF one affects both the installed capacity of the network lines and the capacity and location of the installed energy conversion and storage plants, especially photovoltaic panels, and lithium batteries. This is due to the consideration of physical constraints in the network model, such as those associated with voltage angles. Finally, results are strengthened by means of a sensitivity analysis on physical parameters, investment costs and electricity prices .