摘要
The research addresses the prevalence of gassy soil, containing methane (CH4), within the soil particles of southeast coastal areas of China, such as the Quaternary deposit in the Hangzhou Bay area. This soil exhibits spatial variability in the distribution of gas pressure, posing a potential threat of engineering disasters, including fire outbreaks and blasting, during the construction of underground projects. Consequently, it is crucial to assess the risk state of gas pressure, involving accurate identification and reduction of associated uncertainty, through site investigation. This is indispensable prior to the commencement of underground projects. However, during the site investigation stage, the random field parameters that quantify the spatial variability distribution of gas pressure (e.g., mean value, standard deviations, and scale of fluctuation) are unknown, introducing corresponding statistical uncertainty. Therefore, the most significant consideration for planning site investigation from an engineering perspective involves determining the risk state of gas pressure while considering the statistical uncertainty of these random field parameters. This consideration heavily relies on the engineering experience gained from current site investigation practices. To address this challenge, the study introduces a probabilistic site investigation optimization method designed for planning the site investigation scheme for gassy soils, including determining the number and locations of boreholes. The method is based on the expected state-identification probability, representing the probability of identifying the risk state of gas pressure, and takes into account the statistical uncertainty of random field parameters. The proposed method aims to determine an optimal investigation scheme before conducting the site investigation, leveraging prior knowledge. This optimal scheme is identified using Subset Simulation Optimization (SSO) in the space of candidate site investigations, maximizing the value of the expected state-identification probability at the minimal value point. Finally, the paper illustrates the proposed approach through a case study.
The research addresses the prevalence of gassy soil, containing methane (CH4), within the soil particles of southeast coastal areas of China, such as the Quaternary deposit in the Hangzhou Bay area. This soil exhibits spatial variability in the distribution of gas pressure, posing a potential threat of engineering disasters, including fire outbreaks and blasting, during the construction of underground projects. Consequently, it is crucial to assess the risk state of gas pressure, involving accurate identification and reduction of associated uncertainty, through site investigation. This is indispensable prior to the commencement of underground projects. However, during the site investigation stage, the random field parameters that quantify the spatial variability distribution of gas pressure (e.g., mean value, standard deviations, and scale of fluctuation) are unknown, introducing corresponding statistical uncertainty. Therefore, the most significant consideration for planning site investigation from an engineering perspective involves determining the risk state of gas pressure while considering the statistical uncertainty of these random field parameters. This consideration heavily relies on the engineering experience gained from current site investigation practices. To address this challenge, the study introduces a probabilistic site investigation optimization method designed for planning the site investigation scheme for gassy soils, including determining the number and locations of boreholes. The method is based on the expected state-identification probability, representing the probability of identifying the risk state of gas pressure, and takes into account the statistical uncertainty of random field parameters. The proposed method aims to determine an optimal investigation scheme before conducting the site investigation, leveraging prior knowledge. This optimal scheme is identified using Subset Simulation Optimization (SSO) in the space of candidate site investigations, maximizing the value of the expected state-identification probability at the minimal value point. Finally, the paper illustrates the proposed approach through a case study.
作者
Shaolin Ding
Quanhong Li
Shaolin Ding;Quanhong Li(Key Laboratory of Geotechnical Mechanics and Engineering of the Ministry of Water Resources, Yangtze River Scientific Research Institute, Wuhan, China;Mid-Route Source of South-to-North Water Transfer Corp., Ltd., Danjiangkou, China)
