Citation

Xuecai X, Gui F, Shifei S, Xueming S, Jing L, Lida H, Na W. Process. Saf. Environ. Prot. 2024; 182: 1208-1226.

Copyright

(Copyright © 2024, Institution of Chemical Engineers and European Federation of Chemical Engineering, Publisher Hemisphere Publishing)

DOI

10.1016/j.psep.2023.12.042

PMID

unavailable

Abstract

The content and methods of traditional safety training can no longer meet the needs of the industry or accident-targeted safety training. To solve this problem, this study proposes an accidental case data accident causing model-driven safety training method (ACDACM method) based on an analysis of the content and methods of China's safety training. This method is theoretically driven by the accident causation model and data-driven by the accident case information. Through data analysis of historical accident cases, basic information and comprehensive reasons for accidents can be obtained. Using data mining algorithms, the causes and combinations of high-frequency and high-risk accidents can be mined, supported by accident data, and a path map of the accident causes can then be obtained. The developed safety training project is more targeted by comparing accident causes with industry standards and regulations. This study considers coal and gas outburst accidents as an example and uses 24Model to analyze 84 coal and gas outburst accidents in China. The targeted safety training plan had eight parts of training content, and supporting PPTs, handbooks, and videos were produced. The application shows that this method can provide targeted safety training designs for industries or accidents and has industry universality. Finally, the advantages and development of the ACDACM method are elaborated and the future of safety training is discussed. This study provides theoretical and methodological support for targeted safety-training programs.

Language: en

Keywords

24Model; Accurate safety management; Historical accident case data; Safety training; Targeted safety training; Versatility safety training program