A new approach to understand metro operation safety by exploring metro operation hazard network (MOHN). (March 2017)

• Record Type: Journal Article
• Authors: Li, Qiming; Song, Liangliang; List, George F.; Deng, Yongliang; Zhou, Zhipeng; Liu, Ping
• Is Part Of: Safety science. Volume 93(2017)
• Page Start: 50
• Page End: 61
• Publication Date: 2017-03
• Subjects: Safety analysis -- Accident analysis -- Metro operation hazard network (MOHN) -- Complex network theory -- Structural characteristics; Industrial accidents -- Periodicals; Accident Prevention -- Periodicals; Safety -- Periodicals; Travail -- Accidents -- Périodiques; 363.11
• DOI: 10.1016/j.ssci.2016.10.010
• Languages: English
• ISSNs: 0925-7535

Abstract

Highlights: An analytical framework is developed to analyze the interactions among metro hazards. Metro operation hazard network (MOHN) is built based on accident database. Complex network theory (CNM) is employed to explore the inherent features of MOHN. MOHN captures the scale-free property as well as the small world property. Abstract: Numerous metro accidents expose the vulnerability of metro system. As cities are rapidly building and expanding metro systems, it is essential to thoroughly explore the nature of metro operation safety. Metro accidents can be regarded as an emergent property that arises from the unusual interactions of system components. These interactions could give rise to the phenomenon that several interrelated hazards simultaneously emerge in one single accident. Understanding these interrelations among hazards is indispensable to study metro operation safety. From this standpoint, 28 hazards and 48 interrelations among hazards were identified from 134 accidents and expert interviews, which were the foundation of establishing metro operation hazard network (MOHN). Whereby complex network theory, seven parameters were applied to further reveal the structural properties of MOHN. The results indicate that the MOHN is a scale-free network for the cumulative degree distribution obeys power-law distribution. The scale-free property of MOHN is indicative of its robustness to random attacks and its vulnerability to deliberate attacks. Nichetargeting controlling hazards of high degrees and betweenness centrality can significantly decrease the metro operation risks. Moreover, MOHN also possesses the small-world property for having a relatively high clustering coefficient and small shortest path length. This indicates that risks would be transmitted very quickly in MOHN. Secondary and derivative hazards should receive enough attention for the rapid propagation of MOHN. Revealing the inherent properties of MOHN assist in making beforehand strategies prior to metro accident and contributes to elevate system safety of metro operation.