Abstract:

Scenarios screening is the key to ensure the effectiveness of layer of protection analysis (LOPA). To screen out the scenarios accurately which need to carry out the LOPA analysis, this paper proposes a scenarios identification and screening method based on the most credible scene on the basis of Bow-tie model. First, use the Fault tree and Event tree to establish a simplified Bow-tie model which centers on loss events to identify the potential hazards of the analysis object, and obtain a list of hazard scenarios. Secondly, improve the scenario screening method based on the most credible scenario identification method. Use the improved screening method to filter the list of hazardous scenarios. Simultaneously determine the scenarios that need carry on LOPA analysis in the scenarios list. Finally, apply this method in the 10⁵m3 crude oil storage tank of a reserve storage and use risk analysis screening tool (RAST) to verify the feasibility and effectiveness of the proposed method. The results show that the proposed method can better consider the actual operation of the equipment and screen scenarios according to the internal risk standards of the enterprise, which makes the actual screening results obtained more instructive for the subsequent LOPA analysis.

Key words: petroleum, safety, model reduction, scenarios screening, Bow-tie model

摘要：

场景筛选是确保保护层分析（LOPA）有效性的关键，为准确筛选出需要进行LOPA分析的场景，本文在Bow-tie模型的基础上，提出了基于最大可信场景的场景辨识及筛选方法。首先，利用事故树与事件树建立以损失事件为中心的简化Bow-tie模型辨识分析对象的潜在危害，得到危害场景列表；其次，以最大可信场景辨识方法为基础对场景筛选方法进行改进，利用改进后的筛选方法对危害场景列表进行筛选，确定场景列表中最终需要进行LOPA分析的场景；最后，将此方法应用于某商储库10⁵m³原油储罐风险分析的场景筛选中，并利用风险分析筛选软件（RAST）来验证所提方法的可行性和有效性。结果表明：所提方法能够较好地考虑设备实际运行情况，且根据企业内部的风险标准进行筛选，得到的实际筛选结果对后续的LOPA分析更具指导意义。

关键词: 石油, 安全, 模型简化, 场景筛选, Bow-tie模型