Experimental and theoretical investigations of heat generation rates for a water cooled LiFePO₄ battery

Highlights

• Presents experimental method for heat generation measurement in Li-ion battery.

• Method is based on temperature and heat flux measurement.

• Experimental data agrees with well-established theoretical models.

• Results may contribute towards thermal safety of Li-ion cells.

Abstract

Understanding the rate of heat generation in a lithium-ion cell is critical for its safety and performance behavior. This paper presents in situ measurements of the heat generation rate for a prismatic Lithium-ion battery at 1C, 2C, 3C and 4C discharge rates and 5 °C, 15 °C, 25 °C, and 35 °C boundary conditions (BCs). For this work, an aluminum-laminated battery consisting of LiFePO₄ cathode material with 20 Ah capacity was adopted to investigate the variation of the rate of heat generation as a function of the discharge capacity. Ten thermocouples and three heat flux sensors were applied to the battery surface at distributed locations. The results of this study show that the highest rate of heat generation was found to be 91 W for 4C discharge rate and 5 °C BC while the minimum value was 13 W measured at 1C discharge rate and 35 °C BC. It was also found that the increase in discharge rate and thus the discharge current caused consistent increase in the heat generation rate for equal depth of discharge points. A model is later developed using the neural network approach and validated. The heat generation rate predicted by the model demonstrates an identical behavior with experimental results.

Introduction

Nowadays, the energy crisis is a key issue due to limited fossil fuels sources and concerns over greenhouse emissions [1], [2]. Today, the lithium-ion battery is considered as a suitable energy storage device for alternative energy sources, such as wind and solar, and has many advantages: (i) high specific energy and power densities [3], [4]; (ii) high nominal voltage and low self-discharge rate [5]; and (iii) long cycle-life and no memory effect [6]. That is why; the lithium-ion battery is the most advanced battery technology for electric vehicles (EVs), hybrid electric vehicles (HEVs), and plug-in hybrid electric vehicles (PHEVs). Small consumer products such as laptops, cell phones, toys, radios, laser pointers, slide changers, and many other consumer products also use lithium-ion batteries as the main or secondary power source [7]. However, disadvantageous properties, such as thermal behavior at high discharging rates still remain; therefore, it is crucial to obtain accurate knowledge of battery heat generation and a thermal management system in EV applications [8], [9]. Lithium-ion polymer batteries must be carefully monitored and managed (electrically and thermally) to avoid safety (inflammability) and performance related issues [10], [11], [12].

A lithium-ion battery cells usually has five different layers, namely: the negative current collector, negative electrode (anode), separator, positive electrode (cathode), and positive current collector. The positive electrode materials [13], [14] are typically four types: (i) a metal oxide with a layered structure, such as lithium cobalt oxide (LiCoO₂/LCO) [15]; (ii) a metal with a three dimensional spinal structure, such as lithium manganese oxide (LiMn₂O₄) [16]; (iii) lithium nickel manganese cobalt oxide (LiNiMnCoO₂/NMC); and (iv) a metal with an olivine structure, such as lithium iron phosphate (LiFePO₄/LFP) [17]. The anode is usually made of graphite or a metal oxide. The electrolyte can be liquid, polymer or solid.

At a high temperature environment, lithium ion batteries degrade rapidly, while in a cold temperature environment, the power output and energy are reduced which ultimately results in reduction of performance and driving range [18]. Power fade, capacity fade and self-discharge are well known performance related problems at high temperatures [19]. In addition, safety at high temperatures is a major concern [20]. Therefore, an efficient battery thermal management system (BTMS) is required that uses an optimal thermal design, which relies on a good understanding of the heat generation within the batteries. There are two basic types of cooling systems: (1) air cooling, and (2) water cooling. The advantage of air cooling is electrical safety but it has lower heat transfer coefficient, which makes it more difficult to obtain a uniform temperature distributions. On the other hand, water cooling is more effective and occupies less volume, but has high complexities as well as high cost and weight [21].

Now, to understand the thermal behavior of batteries and its impact on battery performance and life, the first step experimentally is to study the battery temperature distributions and the heat generation profiles at different charge and discharge rates. To make this study relevant to EVs, HEVs, and PHEVs, the charge and discharge rates must be typical of those seen and expected to be seen in vehicles. Fig. 1 shows the surface temperature of a lithium-ion pouch cell at different discharge rates of C/5, C/2, 1C, 2C, 3C and 4C (C-rate is a measure of the charge and discharge current of a battery), on the order of those seen in vehicles. The charge rate between discharges in all cases is 1C. Fig. 1 illustrates the thermal spikes that can accompany discharge. Over a short 20 min time period (short from a vehicle operation viewpoint) for 3C and a 15 min time period for 4C discharge, enough heat is generated to increase the cell temperature to 46 °C (for 3C) and 58 °C (for 4C) from a 22 °C start condition. This value is only for a single pouch cell with a free convection boundary condition, thus an even greater temperature can result when extrapolated to approximately 300 pouch cells in a battery pack of PHEVs, HEVs, and EVs, where there is no free boundary convection, but only conduction between pouch cells. This problem is explained below.

Operating lithium-ion batteries above 50 °C can accelerate the aging process and lead to significant degradation of battery capacity and electric range reduction. Following from Fig. 1, battery cell temperatures above 50 °C are very possible, especially when cells are stacked into modules, and then packs, and if the ambient temperature is closer to 50 °C than the 22 °C used in Fig. 1. The possibility of fire is also a major issue with high operating temperatures where thermal runaway is a possibility. Furthermore, even if thermal runaway does not occur, significant degradation of battery capacity can result by consistently operating at elevated temperature (>50 °C) [22]. Thus, adequate battery cooling and thermal management are an integral part of the vehicle operation during electric mode operation. EVs, HEVs, and PHEVs require a robust battery thermal management system in order to ensure optimal (safe, good performance, and long battery life) vehicle operation.

Various researchers have studied the thermal characteristics by combining the numerical simulation technology and experiment method. There are numerous papers available in the open literature for battery thermal modeling, using different approaches, such as an artificial neural network [23], [24], a finite element model (FEM) [25], [26] or a lumped parameter model (LPM) [27], a linear parameter varying (LPV) model [28], a partial differential equation (PDE) model [29] or a equivalent circuit model (ECM) [30], [31]. Thermal parameters can be determined using analytical relations which need a previous knowledge of the battery [25], [30]. They can also be experimentally determined by adapting a model to experimental data [32]. Here, for the battery modeling, we used a neural network approach. Neural networks are usually organized in layers with nodes or neurons connecting different layers through an activation function. The data or pattern is presented at the input layer which travels to the hidden layers through weighted connections and is finally processed at the output layer which represents the output of the network [23], [24].

In this paper, the research conducted on the variation of the rate of heat generation as a function of the discharge capacity for a particular lithium-ion battery under different discharge rates (1C, 2C, 3C and 4C) and various boundary conditions (BCs) of 5 °C, 15 °C, 25 °C, and 35 °C is presented. First, an experimental study will be presented and the simulated results on the heat generation rate of the lithium-ion battery using a neural network model will then be discussed.