Risk Assessment of Pipeline Engineering Geological Disaster Based on GIS and WOE-GA-BP Models

Abstract

Oil and gas pipelines are part of long-distance transportation projects which pass through areas with complex geological conditions and which are prone to geological disasters. Geological disasters significantly affect the safety of pipeline operations. Therefore, it is essential to conduct geological disaster risk assessments in areas along pipelines to ensure efficient pipeline operation, and to provide theoretical support for early warning and forecasting of geological disasters. In this study, the pipeline routes of the Sichuan-Chongqing and Western Hubei management offices of the Sichuan-East Gas Transmission Project were studied. Seven topographic factors—surface elevation, topographic slope, topographic aspect, plane curvature, stratum lithology, rainfall, and vegetation coverage index—were superimposed using the laying method with a total of eight evaluation indicators. The quantitative relationships between the factors and geological disasters were obtained using the geographic information system (GIS) and weight of evidence (WOE). The backpropagation neural network (BP) was optimised using a genetic algorithm (GA) to obtain the weight of each evaluation index. The quantified index was then utilized to identify the geological hazard risk zone along the pipeline. The results showed that the laying method, stratum lithology, and normalised difference vegetation index were the factors influencing hazards.