集装箱式锂离子电池储能系统消防安全早期预警方法

doi:10.19799/j.cnki.2095-4239.2023.0950

摘要/Abstract

摘要：

为了降低集装箱式锂离子电池储能系统火灾发生概率，提出消防安全早期预警方法。分析锂离子电池储能系统的生热过程，确定温度、电压等因素与储能系统消防安全的影响关系。采集电压、电流、温度、声信号等电池储能系统工作数据，检测表面健康程序，根据工作数据与表面健康状态的变化规律，综合考虑影响因素，预测储能系统的运行走势，设置电池储能系统消防安全早期预警等级，完成电池储能系统消防安全的早期预警。通过性能测试实验得出结论：与传统预警方法相比，优化设计方法的误警率和漏警率降低了0.90%以上，不同预警等级下，该方法仅出现1个预警错误，而且不同表面破损位置下的预警准确率达到了99.7%，具备有效性和可靠性。

关键词: 集装箱式电池, 锂离子电池, 储能系统, 消防安全, 早期预警

Abstract:

To mitigate the risk of fires in containerized lithium-ion battery energy storage systems, we propose an early warning method for fire safety. This method involves analyzing the heat generation process in such systems and establishing the correlation between temperature, voltage, and the fire safety of the storage system. By collecting operational data, including voltage, current, temperature, and acoustic signals, we assess the system's surface health and predict its operational trends based on changes in the operational data and surface health status. Considering various influencing factors, we establish early warning levels for the system's fire safety and implement a comprehensive early warning system. Performance testing demonstrates that, compared to conventional warning methods, our optimized approach significantly reduces the rates of false positives and negatives by more than 0.90%. Furthermore, the method exhibited only one warning error across different warning levels, and achieved a warning accuracy of 99.7% across various surface damage locations, proving its effectiveness and reliability.

Key words: container type batteries, lithium-ion batteries, energy storage system, fire safety, early warning

中图分类号:

TM 912

李润源, 郭傅傲, 赵钢超. 集装箱式锂离子电池储能系统消防安全早期预警方法[J]. 储能科学与技术, 2024, 13(5): 1595-1602.

Runyuan LI, Fu'ao GUO, Gangchao ZHAO. Early warning method for fire safety of containerized lithium-ion battery energy storage systems[J]. Energy Storage Science and Technology, 2024, 13(5): 1595-1602.

图/表 9

图1

图2

表1

表2

图3

图4

表3

表4

表5

参考文献 12

1	吕力行, 刘骅, 徐雷, 等. 基于数据-模型混合驱动的锂电池储能系统状态估计及预警方法[J]. 热力发电, 2021, 50(8): 64-72.
	LYU L X, LIU H, XU L, et al. State estimation and early warning method for lithium battery energy storage system based on data-model hybrid drive[J]. Thermal Power Generation, 2021, 50(8): 64-72.
2	唐文杰, 姜欣, 刘昊琰, 等. 基于气液逸出物图像识别的锂离子电池火灾早期预警[J]. 高电压技术, 2022, 48(8): 3295-3304.
	TANG W J, JIANG X, LIU H Y, et al. Early warning of lithium-ion battery fire based on image recognition of gas-liquid escape[J]. High Voltage Engineering, 2022, 48(8): 3295-3304.
3	李毓烜, 阚强, 崔海浩. 储能系统用三元锂离子电池热失控火灾特性[J]. 电源技术, 2023, 47(3): 328-331.
	LI Y X, KAN Q, CUI H H. Thermal runaway induced fire characteristics of ternary lithium ion batteries used in energy storage system[J]. Chinese Journal of Power Sources, 2023, 47(3): 328-331.
4	杨梦洁, 杨爱军, 叶奕君, 等. 基于气体分析的锂离子电池热失控早期预警研究进展[J]. 电工技术学报, 2023, 38(17): 4507-4538.
	YANG M J, YANG A J, YE Y J, et al. Research progress on early warning of thermal runaway of Li-ion batteries based on gas analysis[J]. Transactions of China Electrotechnical Society, 2023, 38(17): 4507-4538.
5	林格. 基于AI技术的锂离子电池存储区火灾早期探测技术研究[J]. 消防科学与技术, 2022, 41(5): 686-689, 693.
	LIN G. Research on early fire detection technology of lithium-ion battery storage area based on AI technology[J]. Fire Science and Technology, 2022, 41(5): 686-689, 693.
6	贺佳伟, 邵垒, 杨文举, 等. 基于多指标评估云模型的车载动力锂电池火灾风险研究[J]. 科学技术与工程, 2023, 23(20): 8901-8906.
	HE J W, SHAO L, YANG W J, et al. Fire safety of vehicle power lithium battery based on multi-index assessment cloud model[J]. Science Technology and Engineering, 2023, 23(20): 8901-8906.
7	孟祥飞, 韩季君, 陈杰, 等. 锂离子电池火灾原因解析及消防技术研究[J]. 工业安全与环保, 2021, 47(12): 57-60.
	MENG X F, HAN J J, CHEN J, et al. Study on fire cause analysis and fire fighting technology of lithium ion battery[J]. Industrial Safety and Environmental Protection, 2021, 47(12): 57-60.
8	王博石, 余娟, 杨燕, 等. 基于重构误差计算的数据驱动储能电池热失控预警方法[J]. 中国电机工程学报, 2023, 43(11): 4244-4254.
	WANG B S, YU J, YANG Y, et al. A data-driven thermal runaway early warning method for energy storage battery with reconstruction error calculation[J]. Proceedings of the CSEE, 2023, 43(11): 4244-4254.
9	薛金花, 王德顺, 李硕玮, 等. 基于信号的锂离子电池热失控故障早期检测[J]. 电源技术, 2022, 46(11): 1275-1278.
	XUE J H, WANG D S, LI S W, et al. Early detection of thermal runaway fault in lithium-ion batteries based on signal decomposition[J]. Chinese Journal of Power Sources, 2022, 46(11): 1275-1278.
10	彭晓晗, 马宏忠, 许洪华, 等. 基于振动的储能电池异常工况预警新方法[J]. 电测与仪表, 2023, 60(2): 167-171.
	PENG X H, MA H Z, XU H H, et al. A novel method of early warning for abnormal working conditions of energy storage batteries based on vibration[J]. Electrical Measurement & Instrumentation, 2023, 60(2): 167-171.
11	孙磊, 马勇, 郭东亮, 等. 火灾预警系统监测信号漂移和毛刺抑制研究[J]. 消防科学与技术, 2022, 41(12): 1640-1644.
	SUN L, MA Y, GUO D L, et al. Research on monitoring signal drift and burr suppression of intelligent fire warning system[J]. Fire Science and Technology, 2022, 41(12): 1640-1644.
12	王皆佳, 贾君瑞, 赵彤, 等. 锂离子动力电池热失控特性实验研究[J]. 电子设计工程, 2023, 31(5): 111-115.
	WANG J J, JIA J R, ZHAO T, et al. Experimental research on thermal runaway characteristics of Li-ion power battrey[J]. Electronic Design Engineering, 2023, 31(5): 111-115.

预警等级	温度/℃	表面破损面积/mm²	声信号强度/dB
Ⅰ	[80,200]	[50,200]	[60,150]
Ⅱ	[60,80]	[30,50]	[50,60]
Ⅲ	[50,60]	[20,30]	[40,50]
Ⅳ	[40,50]	[10,20]	[30,40]
Ⅴ	[30,40]	[0,10]	[0,30]

电池样品编号	电压/V	电流/A	温度/℃	表面破损面积/mm²	声信号强度/dB	预期预警等级
01	230	2.5	69	31	55	Ⅱ
02	250	3.8	125	121	71	Ⅰ
03	228	2.4	56	23	46	Ⅲ
04	226	1.8	52	21	42	Ⅲ
05	238	2.7	73	45	58	Ⅱ
06	231	2.4	42	13	34	Ⅳ
07	229	2.3	35	7	21	Ⅴ

实验次数	预期预警样品数量/个	基于数据-模型混合驱动的电池储能系统早期预警方法		基于气液逸出物图像识别的电池储能系统早期预警方法		优化设计集装箱式锂离子电池储能系统消防安全早期预警方法
实验次数	预期预警样品数量/个	预警错误样品数量/个	实际预警样品数量/个	预警错误样品数量/个	实际预警样品数量/个	预警错误样品数量/个	实际预警样品数量/个
1	297	3	292	5	292	0	297
2	293	2	290	1	291	0	293
3	299	3	296	7	292	1	298
4	295	3	292	2	291	0	295
5	294	4	289	5	288	0	294

预警等级	预期预警样品数量/个	预警正确数量/个
预警等级	预期预警样品数量/个	基于数据-模型混合驱动的电池储能系统早期预警方法	基于气液逸出物图像识别的电池储能系统早期预警方法	优化设计集装箱式锂离子电池储能系统消防安全早期预警方法
Ⅰ	60	58(2个全部误判为Ⅱ)	59(误判为Ⅱ)	60
Ⅱ	60	58(1个误判为Ⅰ，1个误判为Ⅲ)	58(2个全部误判为Ⅲ)	60
Ⅲ	60	59(误判为Ⅱ)	58(2个全部误判为Ⅳ)	60
Ⅳ	60	60	58(2个全部误判为Ⅴ)	59(误判为Ⅲ)
Ⅴ	60	59(误判为无需预警)	58(2个全部误判为无需预警)	60

样本数量/个	预警准确率/%
样本数量/个	基于数据-模型混合驱动的电池储能系统早期预警方法	基于气液逸出物图像识别的电池储能系统早期预警方法	优化设计集装箱式锂离子电池储能系统消防安全早期预警方法
50	100.0	100.0	100.0
100	99.0	98.0	100.0
150	98.7	98.0	100.0
200	98.5	97.5	100.0
250	98.4	96.8	100.0