摘要
以船舶火灾为背景,建立了适用于船舶舱室火灾风险分析的方法。首先确定了舱室起火频率模型以及主要消防措施的可靠性,然后采用贝叶斯网络建立了火灾场景发生概率模型。其次对船舶舱室火灾的不确定性参数进行了分析,并将蒙特卡洛模拟与双区火灾模型结合对场景发生危险的概率进行计算。最后,对实船典型结构舱室进行了算例。结果表明:舱室人员发生火灾危险的概率(2.011×10-3)大于设备发生火灾危险的概率(7.155×10-4);当舱室起火时,喷淋对舱室人员安全几乎没有影响,无论有无喷淋,烟气沉降到危险临界高度的时间都很短,均值约为9 s;喷淋对舱室设备有很好的保护作用,无喷淋的舱室最高温度均值为292℃,有喷淋的舱室最高温度均值为141℃。
Due to the requirements of risk-based design on ship, a method of fire risk analysis of ship compartment is developed. This method is divided into two parts： analysis of fire scenarios with their occurrence probabilities and analysis of fire consequences of the scenarios in potential danger. In the first part, the ignition frequency per unit area of the SOLAS space category is given and the fire ignition model is constructed. Then the reliabilities of the mainly important fire safety measures is studied. Based on the ignition frequency and the reliabilities of fire safety measures, a Bayesian Networks model is constructed to analyze the fire scenarios and calculate their occurrence probabilities. Through the analysis of fire scenarios, the scenarios in potential danger can be also recognized. In the second part, these potentially dangerous scenarios are analyzed in detail. Firstly, a two-zone model, CFAST, is adopted here to predict fire consequences and a heat release rate model is introduced. Secondly, the statistic characteristics of the input parameters are studied. Thirdly, Latin Hypercube Sampling is used to generate random samples of the uncertainty parameters. Then, a Monte Carlo simulation approach coupling with CFAST is implemented for calculating the distribution characteristics of the predicted safety index, such as smoke layer height, compartment temperature and so on. Finally, the probability of the scenario in danger can be concluded by comparing the predicted safety index with its critical threshold. To demonstrate the proposed method in detail, a typical structure compartment of a ship is presented as a case study. The results indicated that the fire risk to people was 2.011×10^-3 and the fire risk to facilities was 7.155×10^-4 . The sprinklers had no resistance to the descent of smoke layer and had little effect on passenger safety protection. The critical time of smoke layer was about 9 s whether the sprinkler was activated or not. But the sprinkler had an obvious effect on the temperature control and could protect facilities in the compartment. The compartment temperature was about 292℃ when the sprinkler was not activated, and the compartment temperature was about 141℃ when the sprinkler was activated.
出处
《安全与环境学报》
CAS
CSCD
北大核心
2014年第3期132-137,共6页
Journal of Safety and Environment
基金
高等学校博士学科点专项科研项目(20123402110048)
关键词
安全工程
火灾风险
不确定性
贝叶斯网络
船舶舱室
safety engineering
fire risk
uncertainty
Bayesian networks
ship compartment
