Name: | Description: | Size: | Format: | |
---|---|---|---|---|
10.34 MB | Adobe PDF |
Authors
Abstract(s)
As baterias de Lítio–Fosfato de Ferro (LiFePO4) têm vindo a assumir-se como uma das mais
promissoras tecnologias de armazenamento de energia, particularmente para aplicações de
pequena escala relacionadas com a mobilidade elétrica e com os sistemas de aproveitamento
de energias renováveis. As características que estas baterias apresentam são favoráveis a tais
aplicações, nomeadamente o facto de esta tecnologia ser segura, durável, apresentar reduzido
impacto ambiental, possuir um rendimento elevado e uma taxa de autodescarga reduzida.
O presente trabalho apresenta uma descrição da tecnologia LiFePO4 e descreve as respetivas
vantagens e inconvenientes face a outras tecnologias de armazenamento. Adicionalmente é
apresentado um conjunto de ensaios efetuados com baterias LiFePO4, os quais permitiram
caracterizar o comportamento de baterias deste tipo, nomeadamente no que se refere à tensão
(em carga e em vazio) e ao rendimento do processo de carga e descarga. Para o efeito foi
desenvolvida uma plataforma de ensaios dotada de um banco de cargas, um carregador e um
sistema de controlo, monitorização e registo de dados.
A referida plataforma e os dados recolhidos nos ensaios serviram de suporte ao
desenvolvimento e teste de um algoritmo capaz de estimar e transmitir, a um sistema de
gestão de energia, o estado de um sistema de armazenamento baseado em baterias LiFePO4,
nomeadamente no que se refere ao estado de carga (SoC) e de saúde (SoH). O algoritmo
desenvolvido baseia-se no método da contagem de Coulomb, tendo sido integrado um
mecanismo de ajuste automático e dinâmico.
O algoritmo foi testado usando a plataforma desenvolvida, a qual foi parametrizada para
simular um consumidor doméstico alimentado por um sistema isolado da rede elétrica
constituído por um gerador fotovoltaico e por um sistema de armazenamento baseado em
baterias LiFePO4. Os testes efetuados mostram que o algoritmo permitiu estimar o SoC e o
SoH de uma bateria com um erro máximo de 1,5 e 3%, respetivamente.
ABSTRACT: Lithium-Iron Phosphate (LiFePO4) batteries have been considered one of the most promising technologies for energy storage, particularly concerning the small-scale applications related to electric mobility and renewable energy systems. The features of these batteries favor such applications, namely the safety and durability of this technology, the high efficiency and low self-discharge rate and the low environmental impact resulting from the use of that technology. This dissertation presents a description of the LiFePO4 technology and describes its advantages and disadvantages in comparison to other storage technologies. Moreover, it presents a set of experiments carried out with LiFePO4 batteries, which allowed to characterize the behavior of the tested batteries, namely regarding the voltage (open-circuit and load voltage) and the efficiency of the charging and discharging processes. For this purpose, a test platform equipped with a load bank, a charger circuit, and a system devoted to control the charging/discharging processes as well as to monitor and store relevant data, was developed. The platform and the data collected in the experiments were then used to support the implementation and test of an algorithm able to estimate, in real time, the state of a storage system based on LiFePO4 batteries, specifically concerning the state of charge (SoC) and the state of Health (SoH). That algorithm is based on the Coulomb counting method. In order to mitigate the difficulties that characterize this method, an automatic and dynamic adjustment mechanism was included. Additionally, the implemented solution has the ability to communicate the SoC and the SoH to a power management system (Power Box). The algorithm was tested using the developed platform, which has been parameterized to mimic a domestic grid-isolated consumer with a photovoltaic (PV) generator as the supply source. In order to satisfy the consumption that occurs at periods without PV generation, an energy storage system composed by LiFePO4 batteries was assumed. The performed tests have shown that the algorithm estimated the SoC and the SoH of the battery bank with a maximum error of 1,5 and 3%, respectively.
ABSTRACT: Lithium-Iron Phosphate (LiFePO4) batteries have been considered one of the most promising technologies for energy storage, particularly concerning the small-scale applications related to electric mobility and renewable energy systems. The features of these batteries favor such applications, namely the safety and durability of this technology, the high efficiency and low self-discharge rate and the low environmental impact resulting from the use of that technology. This dissertation presents a description of the LiFePO4 technology and describes its advantages and disadvantages in comparison to other storage technologies. Moreover, it presents a set of experiments carried out with LiFePO4 batteries, which allowed to characterize the behavior of the tested batteries, namely regarding the voltage (open-circuit and load voltage) and the efficiency of the charging and discharging processes. For this purpose, a test platform equipped with a load bank, a charger circuit, and a system devoted to control the charging/discharging processes as well as to monitor and store relevant data, was developed. The platform and the data collected in the experiments were then used to support the implementation and test of an algorithm able to estimate, in real time, the state of a storage system based on LiFePO4 batteries, specifically concerning the state of charge (SoC) and the state of Health (SoH). That algorithm is based on the Coulomb counting method. In order to mitigate the difficulties that characterize this method, an automatic and dynamic adjustment mechanism was included. Additionally, the implemented solution has the ability to communicate the SoC and the SoH to a power management system (Power Box). The algorithm was tested using the developed platform, which has been parameterized to mimic a domestic grid-isolated consumer with a photovoltaic (PV) generator as the supply source. In order to satisfy the consumption that occurs at periods without PV generation, an energy storage system composed by LiFePO4 batteries was assumed. The performed tests have shown that the algorithm estimated the SoC and the SoH of the battery bank with a maximum error of 1,5 and 3%, respectively.
Description
Dissertação de Mestrado em Eng. Eletrotécnica-Energia e Automação Industrial
Keywords
Sistemas de Armazenamento de Energia (ESS) Baterias LiFePO4 Estado de carga (SoC) Estado de saúde (SoH) Contagem de Coulomb Energia renovável Sistemas autónomos
Citation
Publisher
Instituto Politécnico de Viseu. Escola Superior de Tecnologia e Gestão de Viseu