1.Introduction

As a broadly applied power source for portable devices,lithiumion battery has advantages of high energy density, no memory effects, long cycle life, being environmental friendly,etc. After small lithium-ion batteries are dominant in the consumer electronies area, large lithium-ion batteries are developed, marching into automobile and power grid applications.

The application of lithium-ion battery brings about fire accidents and explosions, many researchers in the field of battery chemistry have studied and analyzed the reasons of battery failures under various conditions from a chemistry point of view. According to such studics, continuous improvements of battery components were made: various anode and cathode materials were developed to improve the chemical stability; a multilayer separator was designed to limit thermal runaway; proper additives were introduced into electrolyte to block chemical reactions or to discharge the battery itself for mitigation of overcharge risk without affecting normal charging, etc. Manufacturing and assembling techniques were also improved to decrease the defect probability. However, detailed substances of battery components and the quality of battery assembling, which highly affect the safety of the battery, remain unclear to electrical engineers operating battery energy storage system. Therefore, principles of lithium ion battery needs to be presented, then a basic perspective on risks caused by lithiunrion battery and reasons of such risks can be gained. This perspective will provide engineers operating BESS with proper ways for well-regulated safety.

2.Abuse Test of Lithium ion Battery Cell

Practically, the failure of lithiumion cell is acomprehensive process, which can start with any of the above-mentioned exothermic reactions while ended with different dangers such as battery body expansion, electrolyte leakage,gas venting, fire, explosion, ete. To estimate the safety level of commercial lithium ion batteries, abuse test items in mechanical, electrical and thermal aspects are designed according to standards from UL and IEC For lithiumrion BESS used in the grid application, large batteries with pouch type or prismatic type design are preferred.In this paper, lithium-ion battery cells of those two types were tested. All the cells were fully charged before abuse tests according to the standards.

2.1 Thermal Abuse

Objective cells were heated in a temperature chamber. The ambient temperature of the chamber was set to 130 ℃ with an increased rate of 5℃/min. After the ambient temperature in the chamber reached 130 ℃, it was kept for 10 min and then the samples were observed. Under the temperature,potential risks were brought by SEI failure the melt of the separator and an increased gas pressure from the electrolyte. After the test, the leakage, gas venting and the voltage drop of the tested battery cells were not observed.

Therefore, there was no thermal runaway of the tested cells.It is observed from Fig.2 that the cell body expansion of both prismatic type cells and pouch type cells occurs. The body expansion might be caused by vaporizing clectrolyte. The body expansion rate under the test depends on the amount of the low-boiling point solvent in the electrolyte. And the exact amount and the proportion are unclear to users. However, according to the observation, it is concluded that the pouch type sample shown in Fig.2(c) has better performance than that in Fig.2(b), which indicates a higher safety level. The prismatic type sample in Fig.2 (a) shows good performance due to its high thermal resistance caused by the thickness.

2.2 Nail Penetration

A ∅5 mm nail was penetrated into the samples at 20 mm/s. It was then pulled out after 1 min. Under this testcondition, an internal short circuit caused by the direct contact of positive and negative materials may happen. The heat brought by the internal short circuit may lead to decomposition reactions of battery components.

During the test, electrolyte spray and serious gas venting were observed for all the prismatic type cells. The measured voltage and surface temperature of one cell are shown in Fig.3(a). The curves showed an occurrence of the internal short circuit which caused a release of the stored energy and a drop of the cell voltage. Furthermore, the temperature rose up to 130 8 ℃ due to the released energy. The surface temperature then dropped to a relatively safe range, which means exothermic chain reactions did not occur and the thermal runaway was avoided after the test.

For pouch type cells, temperature rise, electrolyte spray or gas venting were not found during the test, except one of five samples. In Fig.3 (b), the measured voltage and the surface temperature of the problematic one are plotted. Part of the stored energy was released through the internal short circuit. And the internal short circuit was terminated bv the gas released from the electrolyte, which expended the cell body and formed an isolation layer between the positive and negative materials and the separator in the penetrated area.

This incomplete internal short circuit only led to a slightly reduced cell voltage and a peak temperature of 90. 5 ℃.As the cell surface temperature dropped, the phenomenon of thermal runaway did not occur after the test. The venting protection of the prismatic type cell was triggered due to the high internal pressure, as shown in Fig.4(a). For the pouch type cell, body expansions could be observed, as shown in Fig.4(b). Generally, pouch type cells show higher safety level compared with prismatic type cells.

2.3 Overcharge

The samples are overcharged with a current of 0.05 C. Once the cell voltage reaches 5 V or the charging time reaches 30 min, the test ends. 1 C is defined as the current rate at which the battery cell is fully discharged in 1 hour, which means 1 C equals 40 A for a battery with the capacity of 40 Ah.

For all the tested samples, there were no clectrolyte leakage, venting gas or other dangers observed. The body cxpansion of the samples can be observed after the test. According to the measured cell voltage, the ambient and cell surface temperatures shown in Fig.5, it is concluded that the phenomenon of thermal runaway did not occur.

2.4 External Short Circuit

A circuit contactor was connected between sample electrodes, and the short circuit resistance was set to 5 mΩ. In the initial trials of the test, fires were found on the cable or the contactor, as shown in Fig.6. Therefore. Cables and contactors with a current of 1 500 A were chosen in the test afterward to avoid the fire in the test circuit.

The body expansion, electrolyte leakage and gas venting were found during the test of all the prismatic type samples as well as some pouch type samples. After the test circuit contactor was closed, the cell surface temperature rose up to around 100°C. With an elevated temperature, cell body expanded (as shown in Fig. 7, area A )with gas released from the electrolyte (as shown in Fig.7, area B)and clectrolyte leaked (as shown in Fig.7, area C). Finally,a fierce gas venting with electrolyte (as shown in Fig.7, area D)occurred. According to the previous introduction of the electrolyte, the vented gas and electrolyte is ignitable. After around 10 min, the cell surface temperature began to drop. There was no fire or explosion during and after the test. Therefore, the phenomenon of thermal runaway did not occur.

Besides,the current collector of the positive electrode was melt down immediately, terminating the external short circuit. The phenomenon was observed in most of the pouch type samples and one prismatic type sample. Fig.8 showed that the melting down of prismatic type samples was fiercer than that of pouch type samples. Metallic sparks in Fig. 8 splashed from the positive current collector. Those metallic sparks can ignite the venting gas or leaking electrolyte and then cause fire.