Novel modeling approaches for analyzing the design of distributed multi-energy systems

Questa tesi descrive una serie di miglioramenti metodologici nell'analisi dei sistemi energetici distribuiti e nella loro progettazione, con particolare attenzione ai sistemi che coinvolgono più di un vettore energetico (sistemi multi-energia). La questione della progettazione di tali sistemi sta diventando estremamente importante e necessita di nuovi approcci a causa di uno scenario di fruizione dell'energia che andrà a mutare notevolmente a causa di diversi cambiamenti riconducibili principalmente alla necessità di rendere la fruizione dell'energia più sostenibile e a minor impatto ambientale.This dissertation describes a set of methodological improvements in the analysis of distributed energy systems and their design, with a particular focus on systems involving more than one energy vector (multi-energy systems). The question of the planning of such systems is becoming extremely important and needing of novel approaches due to an energy fruition scenario which is going to greatly mutate due to several changes which can be mostly reconducted to the need of making the fruition of energy more sustainable and with a lesser environmental impact. Such paradigm changes are of diverse nature: coming from both technological trends, social changing paradigms and an always more impellent policy commitment towards reducing the carbon footprint of our society. As an example they can be attributed to the continuously increasing penetration of renewable non-controllable energy conversion sources, the delocalization of energy systems into a distributed paradigm, and the drastic changes of the modality of the fruition of some energy related commodities with the increasing presence of air conditioning and electric vehicles for example. Given the complex nature of such problems, and therefore the computationally demanding nature of the models needed to gain insights about their workings, a renovated focus is put on obtaining models which allow for the analyses to be undertaken in reasonable amounts of time in this novel energy systems context without compromising on the necessary level of detail into the modeling. In this thesis, we define two potential approaches in doing so, tackling two different challenges that emerged from the available literature. We define a novel approach for the consideration of the temporal dimension in the planning of distributed energy systems. Specifically, the approach aims at properly considering the multi-decade timespan over which a decision concerning the potential adoption of energy systems has to be made. In this setting where some relevant parameters such as the capital costs for the investments in the technologies might change even significantly over such long periods. The approach is implemented within a framework having at its core an optimization problem, solved through mixed-integer programming and heuristic techniques. The validity is then tested on a realistic test case modeled by referring to an open dataset of energy related consumptions for a set of households located in the United States, where the goal of the methodology is finding the optimal year of adoption of an electricity iii storage system (if any) considering its dropping costs over the multi-year planning horizon. We define a methodology to consider various sources of uncertainty within simulations performed by means of an already well established model: being the EnergyPLAN model for smart multi-energy systems. This is achieved by defining a framework that models a set of uncertain inputs such as the availability of solar radiation, the uncertainty in the user demands for space heating and sanitary hot water, and finally the uncertain nature of the demands of an electric vehicle fleet. The test case over which the methodology is tested is the city of Osimo, situated in Italy, which can be assumed to be a small scale multi-energy system, and about which much data needed for the modeling is available through its municipal energy company. The goal of the analyses is to understand how two different flexibility assets: namely a large heat pump coupled with a thermal storage system, and a fleet of electric vehicles equipped with smart charging, can aid in welcoming high shares of non controllable renewable energy sources (photovoltaic panels). This to avoid the feeding of a large share of electricity generation surplus to the national distribution system (thus increasing auto consumption) and to understand the impact on carbon emissions to gain a potential policy related insight