摘要
量化实验事故的人因安全水平是预防危险化学品实验事故的关键。该文基于功能共振分析法(functional resonance analysis method,FRAM),融合模糊的认知可靠性和失误分析法(cognitive reliability and error analysis method,CREAM)、人因可靠性分析法、贝叶斯网络,以概率的方式揭示危险化学品实验中的人因安全水平。研究首先依托实验任务人因FRAM模型,系统识别关键功能及其耦合关系;其次,运用模糊CREAM方法量化模型中各环节的人因可靠性;最后,通过贝叶斯网络表述FRAM模型并吸收各环节的人因可靠性水平,实现对系统整体人因安全的概率量化。研究表明,该方法可量化分析危险化学品实验中的人为因素,同时赋予FRAM方法以概率量化分析功能。
[Objective]In recent years,the number of accidents in laboratories working with hazardous chemicals has increased,especially during experiments.Such accidents pose a danger to the safety and health of researchers and also have huge negative impacts on the progress of scientific research and the reputation of research institutes.Statistical reports indicate that over 50%of such accidents are closely related to human factors.Therefore,the objective of this study is to quantify and assess human factor safety levels during experiments involving hazardous chemicals in a laboratory setting.This research has both theoretical significance and practical benefits for laboratory safety.[Methods]The majority of current safety analysis methods are based on the Safety-Ⅰtheory,which focuses primarily on accident prevention and error identification.Safety-Ⅰtheory assumes that tasks are executed strictly according to predefined procedures and that operators follow these procedures in a linear manner.However,extensive studies have shown that in practice,operators flexibly adapt their own working techniques based on the actual working conditions rather than rigidly following pre-designed procedures.Considering the inherent flexibility of operators and the exploratory nature of experiments involving hazardous chemicals,Safety-I theory is not an effective approach for addressing safety issues in experiments.Therefore,in this study,Safety-Ⅱtheory is applied,shifting the focus to identifying critical factors for ensuring task execution and the successful execution of tasks by accounting for nonlinear coupling effects among multiple factors.The Functional Resonance Analysis Method(FRAM),based on Safety-Ⅱtheory,offers significant advantages.However,the FRAM method primarily focuses on qualitative analysis.Despite its strong advantage in analyzing model construction,this method lacks quantitative analysis capability and has limited accuracy for assessing human factor safety levels.To address this limitation,a method is proposed for integrating FRAM with the fuzzy Cognitive Reliability and Error Analysis Method(CREAM),aiming to support the quantitative analysis of human safety factors at each critical point in the FRAM model.Thereafter,using a Bayesian Network(BN)as the link,the FRAM model is transformed into a BN topology structure,and the quantification outputs produced by the fuzzy CREAM method are absorbed as the prior data for the parent node in BN,thereby achieving quantification of the human factor safety level of the overall experimental task.[Results]The proposed method is applied to a real experiment involving hazardous chemicals.A case study employing the FRAM model previously published by the authors,fuzzy CREAM,and BN method,indicates that the control modes of personnel during the experiment are 3%strategic and 97%tactical;thus,the HEP of the overall experiment is 4.9×10^(–2).[Conclusions]From the viewpoint of theoretical research,the proposed method based on Safety-II theory successfully enables quantitative safety analysis and provides a quantitative analysis function for the FRAM method.From an application perspective,the proposed method enables the effective identification of critical points related to human factors and the quantitative analysis of the human factor safety level.Finally,through the selected case study,this study confirms the applicability of the FRAM-fuzzy CREAM method for quantitatively assessing human factor safety during experiments involving hazardous chemicals.
作者
张人友
施梦洁
李冉
刘亭
张洁
陈梓恺
ZHANG Renyou;SHI Mengjie;LI Ran;LIU Ting;ZHANG Jie;CHEN Zikai(School of Safety Engineering,Beijing Institute of Petrochemical Technology,Beijing 102617,China;SINOPEC Safety Engineering Institute Co.,Ltd.,Qingdao 266001,China;Department of Industrial Engineering,Tsinghua University,Beijing 100084,China)
出处
《实验技术与管理》
北大核心
2025年第11期250-256,共7页
Experimental Technology and Management
基金
国家自然科学基金重大项目(T2192933)
化学品安全全国重点实验室开放基金(SKLCS-2024014)
中国石化科技部课题(325053)
北京市教委科技计划项目(KM202210017002)
北京市级大学生创新创业训练计划项目(2024J00072)。
