Multi-scale risk assessment method for pipeline slope hazards

Oil and gas pipelines, as linear infrastructure spanning multiple regions, are highly susceptible to geological hazards. Previous research has focused on discrete hazard points, yet failure at a single point can compromise the integrity of the entire system, underscoring a gap in quantitative assessment of systemic risk. This study examines the Guangdong Dapeng liquefied natural gas (LNG) pipeline and proposes a multiscale, coupled risk assessment framework based on a hierarchical "region–section–point" approach. At the regional scale, an information-entropy model maps risk distributions and rapidly flags high-risk sections. At the section scale, UAV-based inspection is combined with in situ monitoring to conduct dynamic slope-stability analysis. At the point scale, monitoring data drive a slope–pipeline coupled finite element model that simulates the pipeline's mechanical response under prospective slope-failure scenarios, enabling early warning of buried-pipeline hazards. Results indicate that the Dapeng pipeline is generally stable. High-risk sections cluster in the Nanshan–Pingshan, Dapeng–Xiasha, and Qingxi–Zhangmutou areas, accounting for 18.9% of the total length. The regional model achieves an area under the ROC curve (AUC) of 0.89, supporting its predictive reliability. UAV inspections detected no visible pipeline damage, and the maximum change in displacement-tangent angle at the Pingshan and Dapeng monitoring sites was −1.83°, consistent with stable slopes. Under current conditions, the Pingshan site remains safe; simulations indicate that damage would initiate near the slope toe when ground displacement reaches about 4 m. Monitoring and simulation exhibit consistent pore-water-pressure trends, with maximum changes of 1.7 Pa and 0.8 Pa during the study period, mutually validating monitoring data quality and simulation credibility. The framework shifts the focus from isolated hazard points to continuous linear segments and provides a methodological basis for systematic, fine-grained geohazard-risk management of linear infrastructure.