The achievement of targets imposed by Italian Authorities to reduce CO2 emissions and to increase the production of electricity from renewable sources, contained within the “Integrated National Energy and Climate Plan”, has encouraged relevant modifications in the energy sector during the last years. This thesis presents a possible solution to increase the share of renewable sources finding the optimal layout of Multi-Energy Systems (MES) where multiple demands must be satisfied. The mathematical formulation of the Multi-Energy System problem has been provided as an optimization problem of the Total Annual Cost (TAC), which has to be minimized. The mathematical formulation has led to the definition of a Mixed Integer Nonlinear Programming (MINLP) problem due to nonlinearities in the size effects over machine costs and performances, together with nonlinearities in machine performances in off-design. Afterwards, it has been provided the linearization methodologies for the MINLP problem; in particular, performance maps for machines in part load conditions, size effects of costs and performances, and air temperature effects have been linearized. Moreover, to reduce the computational complexity of the problem, it has been implemented an innovative clustering algorithm, called k-MILP, capable of selecting typical days and extreme days, that represent the most challenging conditions for the MES. A rigorous description of the Multi-Energy System has been provided using a Mixed Integer Linear Programming (MILP) formulation. The MILP model implemented is able to consider some fundamental aspects during the optimization process, such as: typical and extreme operating conditions, on-grid and off-grid conditions, and to identify (n-1)-reliable designs that ensure energy supply even during the failure of one component. The multi-energy system, within this formulation, has been defined starting from the declaration of: i) a set of parameters and input profiles; ii) a linear objective function and a set of variables iii) a set of linear constraint equations. The MILP algorithm for multi-energy systems has been applied to the University Campus of Politecnico di Milano, located in Leonardo square (Milano). The results have shown how this approach is capable of identifying reliable designs with a modest increase of the Total Annual Cost of electricity, if compared to unreliable cases.
Il raggiungimento degli obiettivi imposti dalle Autorità italiane per ridurre le emissioni di CO2 e aumentare la produzione di energia elettrica da fonti rinnovabili, contenuta nel "Piano Nazionale Integrato per l’Energia e il Clima", ha favorito negli ultimi anni importanti cambiamenti nel settore energetico. Questa tesi presenta una possibile soluzione per aumentare la quota di fonti rinnovabili trovando il layout ottimale dei Multi-Energy Systems (MES) in cui devono essere soddisfatte più richieste di energia. La formulazione matematica del problema MES è stata fornita come un problema di ottimizzazione del costo annuo totale (TAC), che deve essere minimizzato. La formulazione matematica ha portato alla definizione di un problema MINLP (Mixed Integer Nonlinear Programming) a causa delle non-linearità negli effetti di taglia sui costi e sulle prestazioni delle macchine, e nelle prestazioni della macchina in off-design. Successivamente, sono state fornite le metodologie di linearizzazione per il problema MINLP; in particolare, sono state linearizzate le curve di performance per le macchine in condizioni di carico parziale, gli effetti di taglia sui costi e sulle prestazioni, e gli effetti dovuti alle variazioni della temperatura dell’aria. Inoltre, per ridurre la complessità computazionale del problema, è stato implementato un innovativo algoritmo di Clustering, chiamato k-MILP, in grado di selezionare giorni tipici e giorni estremi, che rappresentano le condizioni più impegnative per il modello MES. È stata fornita una descrizione rigorosa del modello MES utilizzando una formulazione MILP (Mixed Integer Linear Programming). Il modello MILP implementato è in grado di considerare alcuni aspetti fondamentali durante il processo di ottimizzazione, quali: condizioni operative tipiche ed estreme, condizioni on-grid e off-grid, e di identificare configurazioni con affidabilità n-1, in grado di garantire l'approvvigionamento energetico anche durante il guasto di un componente. Il modello è stato definito attraverso la dichiarazione di: i) un insieme di parametri e profili di input; ii) una funzione obiettivo lineare e un insieme di variabili; iii) un insieme di equazioni di vincolo lineari. L'algoritmo MILP per sistemi energetici aggregati è stato applicato al Campus Universitario del Politecnico, situato nella città di Milano, in Piazza Leonardo. I risultati hanno dimostrato come questo approccio sia in grado di identificare configurazioni affidabili con un modesto aumento del costo totale annuo, se confrontato con casi non affidabili.
Optimal design of multi-energy systems with n-1 reliability
Miconi, Marco
2019/2020
Abstract
The achievement of targets imposed by Italian Authorities to reduce CO2 emissions and to increase the production of electricity from renewable sources, contained within the “Integrated National Energy and Climate Plan”, has encouraged relevant modifications in the energy sector during the last years. This thesis presents a possible solution to increase the share of renewable sources finding the optimal layout of Multi-Energy Systems (MES) where multiple demands must be satisfied. The mathematical formulation of the Multi-Energy System problem has been provided as an optimization problem of the Total Annual Cost (TAC), which has to be minimized. The mathematical formulation has led to the definition of a Mixed Integer Nonlinear Programming (MINLP) problem due to nonlinearities in the size effects over machine costs and performances, together with nonlinearities in machine performances in off-design. Afterwards, it has been provided the linearization methodologies for the MINLP problem; in particular, performance maps for machines in part load conditions, size effects of costs and performances, and air temperature effects have been linearized. Moreover, to reduce the computational complexity of the problem, it has been implemented an innovative clustering algorithm, called k-MILP, capable of selecting typical days and extreme days, that represent the most challenging conditions for the MES. A rigorous description of the Multi-Energy System has been provided using a Mixed Integer Linear Programming (MILP) formulation. The MILP model implemented is able to consider some fundamental aspects during the optimization process, such as: typical and extreme operating conditions, on-grid and off-grid conditions, and to identify (n-1)-reliable designs that ensure energy supply even during the failure of one component. The multi-energy system, within this formulation, has been defined starting from the declaration of: i) a set of parameters and input profiles; ii) a linear objective function and a set of variables iii) a set of linear constraint equations. The MILP algorithm for multi-energy systems has been applied to the University Campus of Politecnico di Milano, located in Leonardo square (Milano). The results have shown how this approach is capable of identifying reliable designs with a modest increase of the Total Annual Cost of electricity, if compared to unreliable cases.File | Dimensione | Formato | |
---|---|---|---|
2021_06_Miconi.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
3.03 MB
Formato
Adobe PDF
|
3.03 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/176149