The purpose of this thesis is to create a transient model of a PEM fuel cell system, based on Matlab Simulink, as general, flexible and adaptable as possible, in order to be easily set on different type of systems. The object of the study is the development of the simulation tool, and its validation against literature and experimental data. An important aim of the developed dynamic semi-empirical model is to try to adopt a theoretical physics-based approach whenever possible, in order to have an accurate scientific correlation between experimental output and theoretical laws, without neglecting the accuracy that could be provided by empirical equations. The major work is focused on the fuel cell stack modeling and involves also a large review of literature analysis concerning the simulation of PEM-FCs. In order to guarantee the adaptability of the model, taking inspiration from the latest studies in this field, a differential evolution algorithm is developed to realize the fitting process of the modeled polarization curve, by means of the stack voltage model, on experimental data. This algorithm has a strategic importance for the choice and the setting of the stack voltage equations on the real static performance of the PEM fuel cell system analyzed, with a proved error of about 2-3%. The transient behaviors captured in the model includes flow characteristics, inertia dynamics, lumped-volume manifold filling dynamics, time evolving-homogeneous reactant pressure or mole fraction, membrane humidity and thermal response of fuel cell and cooling system. From one side, the validation against literature data of Section 4 is realized after the development of a general dynamic PEM-FC system model described in Section 2 and 3, comprising all the components normally present in these systems. The comprehensive dynamic model proposed, usually not presented in literature, perform very well respect to the experimental data, comprising the thermic data and the hydration of the membrane, the most important operative parameters but also the most complex ones to simulate. On the other side, the HI-SEA Joint Laboratory, between Fincantieri S.p.A. and the University of Genoa, allows to study a PEM fuel cell system of 8 stacks sized 33 kW each for a total maximum power of 260 kW. The adaptation and the simplification of the dynamic model to this plant layout helps to study a bigger and more complex PEM-FC system and to validate the model to the experimental data. The simplification of the dynamic model starts form the necessity to set the equations only by the commercially available data, usually limited to the datasheet information. This limitation makes the HI-SEA model less detailed but, at the same time, simpler and able to provide different important results, as the stack and cooling system thermal balances, starting from few easily obtainable data.

Analisi e modellazione di sistemi a celle a combustibile di tipo PEM

Analysis and modeling of PEM fuel cell systems

DE CAMPO, MARCO
2019

Abstract

The purpose of this thesis is to create a transient model of a PEM fuel cell system, based on Matlab Simulink, as general, flexible and adaptable as possible, in order to be easily set on different type of systems. The object of the study is the development of the simulation tool, and its validation against literature and experimental data. An important aim of the developed dynamic semi-empirical model is to try to adopt a theoretical physics-based approach whenever possible, in order to have an accurate scientific correlation between experimental output and theoretical laws, without neglecting the accuracy that could be provided by empirical equations. The major work is focused on the fuel cell stack modeling and involves also a large review of literature analysis concerning the simulation of PEM-FCs. In order to guarantee the adaptability of the model, taking inspiration from the latest studies in this field, a differential evolution algorithm is developed to realize the fitting process of the modeled polarization curve, by means of the stack voltage model, on experimental data. This algorithm has a strategic importance for the choice and the setting of the stack voltage equations on the real static performance of the PEM fuel cell system analyzed, with a proved error of about 2-3%. The transient behaviors captured in the model includes flow characteristics, inertia dynamics, lumped-volume manifold filling dynamics, time evolving-homogeneous reactant pressure or mole fraction, membrane humidity and thermal response of fuel cell and cooling system. From one side, the validation against literature data of Section 4 is realized after the development of a general dynamic PEM-FC system model described in Section 2 and 3, comprising all the components normally present in these systems. The comprehensive dynamic model proposed, usually not presented in literature, perform very well respect to the experimental data, comprising the thermic data and the hydration of the membrane, the most important operative parameters but also the most complex ones to simulate. On the other side, the HI-SEA Joint Laboratory, between Fincantieri S.p.A. and the University of Genoa, allows to study a PEM fuel cell system of 8 stacks sized 33 kW each for a total maximum power of 260 kW. The adaptation and the simplification of the dynamic model to this plant layout helps to study a bigger and more complex PEM-FC system and to validate the model to the experimental data. The simplification of the dynamic model starts form the necessity to set the equations only by the commercially available data, usually limited to the datasheet information. This limitation makes the HI-SEA model less detailed but, at the same time, simpler and able to provide different important results, as the stack and cooling system thermal balances, starting from few easily obtainable data.
21-mag-2019
Inglese
Analisi e modellazione di sistemi a celle a combustibile di tipo PEM
MAGISTRI, LOREDANA
CIANCI, ROBERTO
Università degli studi di Genova
File in questo prodotto:
File Dimensione Formato  
phdunige_3348392.pdf

accesso aperto

Dimensione 12.29 MB
Formato Adobe PDF
12.29 MB Adobe PDF Visualizza/Apri

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14242/69460
Il codice NBN di questa tesi è URN:NBN:IT:UNIGE-69460