A multifactor coupled rarly-warning model for pipeline landslide hazards based on the dynamic evolutionary process of disaster

Long oil and gas pipelines inevitably cross mountainous regions with complex geological environment and are threatened by various geological disasters, the most severe of which is landslides. To achieve early-warning of pipeline landslide hazards, multiple monitoring and early-warning indicators are typically established. However, the failure to account for the correlation between various indicators has led to a high frequency of false alarms. Based on the analysis of the dynamic evolutionary process of pipeline landslide hazards, this study proposes four stages of pipeline landslide hazards: (1) rainfall on the landslide leads to surface infiltration; (2) the groundwater level rises increasing the slope self-weight increases and reducing the geotechnical parameter; (3) the slope has undergone local deformation; (4) the landslide slides entirely and threatens pipeline safety. Three key monitoring indicators of landslide hazard deformations, external induced indicators, and pipeline stress-strain were extracted to construct a monitoring index system for pipeline landslide hazards. On this basis, a multifactor coupled early-warning model for pipeline landslide hazards is proposed. This model integrates the relationships between landslide deformation indicators, external triggers, and pipeline mechanical responses, creating a multidimensional early-warning criterion. This early-warning model was applied in a natural gas pipeline landslide in Guizhou Province, and successfully issued blue, yellow, and red early-warning information, providing sufficient time for emergency response and rescue of the landslide, ensuring the safe operation of the pipeline. This model can avoid the problem of high early-warning levels due to excessive rainfall, further improving the accuracy of pipeline landslide hazards early-warning and making the early-warning results more realistic. This study can provide a valuable reference for the monitoring and early-warning work on pipeline landslide hazards.