Risk assessment method combining complex networks with MCDA for multi-facility risk chain and coupling in UUS

With the acceleration of urban development, the urban underground space (UUS) begins to show the characteristics of a complex giant system. In most of the traditional risk assessment frameworks, metrics are developed by treating the object as a whole. There is a lack of applicable objective-based assessment methods when there are correlations and network structures among the internal structures of the object. Here we propose and test a method, a comprehensive risk assessment model E-CN-TOPSIS applicable to UUS, to solve this problem. First, an initial risk assessment framework with 4 primary indicators and 30 sub-indicators was proposed based on MultiCriteria Decision Analysis (MCDA). Second, a risk chain analysis network of UUS containing 201 risk nodes based on real cases was constructed. The principal component analysis method was taken to quantify the chain-coupled risk mechanism of multi-facility within UUS to integrate the external risk assessment framework with the internal risk chain network. Then we analyzed 16 main cities in three major economic circles of China. Our analyses indicate that 'Recovery capability' was the priority in all evaluation indicators. The main factors influencing the UUS risk in three major economic circles are management policies. The UUS risk is highest in Greater Bay Area, followed by Jing-Jin-Ji, and Yangtze River Delta. Finally, some outlooks and suggestions are given. By applying this methodology, useful information can be provided for the fine planning and precise prevention of new complex UUS risks.

Automated summary

A comprehensive risk assessment model was proposed and tested for urban underground spaces, highlighting the importance of management policies as the main factor influencing UUS risks in different economic circles. The study demonstrated variations in UUS risks among different regions in China, with the Greater Bay Area showing the highest risk level.