二次應用之電池健康管理技術研究

Abstract

基於節能減碳的綠色產業蓬勃發展，車輛載具使用鋰電池模組為主要動力來源已成為重要趨勢。由於鋰電池的製造成本與回收成本很高，許多的研究機構試圖找出如何充分發揮鋰電池模組的經濟效益，延長電池在役時間的技術或是經濟模式。鋰電池二次應用產業模式的研究便是其中一項引起綠色能源主要推動國家或地區的熱門主題。 鋰電池二次應用產業的基本概念是將在電動車應用中淘汰的老化鋰電池轉移到電池特性需求較低的應用領域以延後電池的除役時間來提高電池的經濟效益。其中最受歡迎的二次應用領域即是與再生能源結合的電池儲能系統。此儲能系統可利用二次電池儲存風力發電或是太陽能發電的電能，之後再將這些電能配合整體電網來滿足尖峰用電以降低電力成本。在發展鋰電池二次應用產業的過程中，電池壽命管理為一項關鍵技術用以估測電動車淘汰電池芯的剩餘價值與延長電池在二次產業的除役時間。本論文即針對此兩項議題分別提出解決辦法。 首先，為了公正與客觀地評估淘汰電池芯的剩餘價值以提供電動車製造商與二次電池買家做為交易評估，本論文於第一部份(第三章與第四章)提出了一項適用於高功率高容量鋰電池(例如磷酸鐵鋰電池芯)的離線式健康狀態檢測技術。此離線式檢測技術根據電荷傳遞電阻資訊來估測已汰役電池芯當下的壽命。根據第三章的實驗結果，電荷傳遞電阻因受電池老化影響甚劇，但卻較不受電池電量狀態與外部電阻變異所影響，因此在單一時間常數電池等校電路模型的阻抗參數中是最適合作為健康狀態評估依據。另外於第四章中，本論文更進一步提出一項快速且有效率的三點式阻抗量測法從電池暫態電壓波形中準確地擷取出此電荷傳遞電阻值。根據電池長時間循環測試的驗證結果，由此阻抗量測法所擷取得到的電荷傳遞電阻確實能夠準確的估測電池的健康狀態，其所產生的誤差僅6.1%。在單一時間常數等校電路模型的現有阻抗估測結果當中，此誤差為最低者。 為有效延長淘汰電池於二次產業的除役時間，本論文於第二部分(第五章)便針對此問題提出一個具有電池放電深度管理功能的雙二次電池儲能系統。此系統總共由一電能管理系統、一個雙電池儲能模組、一個燃料電池堆與兩個直流對直流的單向功率轉換器所組成。不同於傳統的電池管理系統，雙電池儲能模組在所提出的系統中為主要電力來源，而燃料電池堆反而為次要電源。電能管理系統藉由狀態機的控制機制實現了雙電池交換策略以有效限制電池的放電深度，因此電池過度充電與過度放電現象均可有效避免，並且延長儲能系統中的電池壽命。這樣的管理演算法能有效率地分配雙電池與燃料電池堆之間的電力輸出，並且能穩定且連續的輸出電能至負載。為了使系統能夠依照電池需求來調整燃料電池的輸出電能，此章亦提出一個具有電流控制功能的升壓式直流對直流功率轉換器以控制燃料電池的電流來達到輸出功率調整的目的。最後，所實現的雙二次電池儲能系統是由20瓦的質子交換膜燃料電池與12安培小時的磷酸鐵鋰電池所組成。不同的放電深度測試結果證實，如將放電深度設定於60%，則於所提出的雙二次電池儲能系統中，電池的壽命將有66.93%的改善率。

Battery second-use (B2U) is a new business model in the future electric vehicle (EV) market to offset the high EV cost to customers by repurposing used EV batteries to a second-life application, such as a battery energy storage system (BESS) with a renewable energy source (RES), for maximizing its economic benefits. For B2U, battery state-of-health (SOH) managements are essential technology for evaluating the remaining value of an aging EV battery and maximizing the battery second lifetime. In this dissertation, an off-line SOH estimation based on the charge transfer resistance is proposed for high power and high capacity lithium-ion battery cells, such as lithium iron phosphate or LFP battery cells, to estimate an unbiased and reliable SOH, so it can be a useful information for EV manufacturers and second-hand battery buyers in trading evaluation. As shown in the experimental results, the charge transfer resistance has a great aging change with battery degradation and good abilities against the state-of-charge (SOC) effect and external resistance variation in impedance parameters of a single time-constant equivalent circuit model (ECM) including ohmic resistance, charge transfer resistance, double-layer capacitance, and time constant, for SOH estimation. A fast and efficient three-point impedance extraction (TPIE) method is further proposed in this dissertation for accurately extracting the charge transfer resistance in off-line SOH estimation. The results of long-term cycling test demonstrate that the TPIE method can successfully indicate the SOH of LFP battery cells with low estimation error of 6.1%. This dissertation also presents a dual second-life battery energy storage system (DSLBESS) with the battery depth-of-discharge (DOD) management for extending battery cycle life and system service time. The proposed system consists of a power management system, a dual-battery energy storage unit (ESU), a fuel cell (FC) stack, and two DC-DC unidirectional power converters. Unlike conventional BESSs, the dual-battery ESU in the DSLBESS serves as the primary power source, while the FC stack serves as the secondary power source. The proposed power management is based on a state-machine-based mechanism, using a battery interchanging strategy to limit batteries operating within a given DOD range. The state-machine-based mechanism is employed for reducing the effects of battery overcharging or overdischarging; thus, it can significantly improve the battery life time. In addition, this algorithm has the ability to distribute the power flow efficiently between the dual batteries and the FC stack, and also to keep providing a stable and continuous energy output for the load. A current-mode control DC-DC boost converter is designed to regulate the FC current and try to maximize the FC power for battery charging time reduction. The proposed system is designed and implemented with a 20W proton exchange membrane fuel cell (PEMFC) stack and two 12Ah LiFePO4 batteries. The experimental results of capacity fades on various DOD tests verify the usefulness of the proposed DSLBESS. The comparison of life improvement shows that the cycle life has a 66.93% improvement when the normal operating region is limited within 60% DOD.