Abstract—

Paying increased attention to the safety of nuclear power facilities is a key element for their further development. An approach to analyzing their fire safety with due regard to the uncertainty factors associated with the parameters of applied models and with the fire breakout and development conditions is presented. The two-zone fire modeling code that takes into account the specific features of transportable (floating) nuclear power plants is used, and its basic characteristics are given. An original approach to a dynamic probabilistic safety assessment is applied, which is based on a global optimum search algorithm. This algorithm is used in this study for taking into account the uncertainties associated with combustion models, possible human actions, and change in the equipment’s operation modes. The proposed method incorporates a genetic algorithm for searching the global optimum, the main purpose of which is to find the most hazardous fire development scenarios in accordance with adaptive variation of parameters within their uncertainty boundaries. Application of the method is demonstrated for two versions of fire breakout in the process rooms of a transportable power unit equipped with a nuclear power installation. The basic parameters of the two-zone fire model and dynamic probabilistic safety assessment are given. The results of computations for the dynamic probabilistic safety assessment demonstrate the complex nature of interaction among different processes. The data from analyzing the influence of the main parameters are given, a comparison with the standard calculation version and with the use of uncertainty analysis is carried out, and the conditional probability of the worst scenario is estimated. Owing to the consideration of uncertainties and the search for the most dangerous scenarios, it became possible to establish the fire development scenarios that can lead to essentially more severe consequences than those determined in using the standard parameters of two-zone fire models.

中文翻译：

---

核动力设施火灾危险分析不确定性的考虑

摘要-

更加重视核动力设施的安全是其进一步发展的关键因素。提出了一种在适当考虑与应用模型参数相关的不确定性因素以及火灾突围和发展条件的情况下对其消防安全性进行分析的方法。使用了考虑了可移动（浮动）核电站的特定特征的两区火模型代码，并给出了其基本特征。应用了基于全局最优搜索算法的动态概率安全性评估的原始方法。在本研究中使用该算法来考虑与燃烧模型相关的不确定性，可能的人为行为以及设备运行模式的变化。所提出的方法结合了一种遗传算法来搜索全局最优值，其主要目的是根据不确定性范围内参数的自适应变化来找到最危险的火灾发展场景。演示了该方法在配备核动力装置的可移动动力装置的处理室中用于两种类型的火灾突围的应用。给出了两区火灾模型的基本参数和动态概率安全评估。动态概率安全性评估的计算结果证明了不同过程之间相互作用的复杂性。给出了分析主要参数影响的数据，进行了与标准计算版本的比较以及使用不确定性分析的比较，并估计了最坏情况的条件概率。由于考虑了不确定性和寻找最危险的情况，因此有可能建立起火灾发展的情况，与使用两区火灾模型的标准参数确定的结果相比，这些火灾可能导致更为严重的后果。