Negative pressure wave based pipeline Leak Detection: Challenges and algorithms

Abstract

Negative pressure wave is a popular method to detect the occurrence and location of leak incidents in oil/gas pipeline. Three core technical challenges and related algorithm are discussed in this paper. The first is data quality. The balance between noise level and locating precision is discussed in filter design. The second one is dynamic slope in anomaly detection, whence a bi-SPC (Static Process Control) algorithms is proposed to make the threshold be adaptive. The third one is the false alarm caused normal working condition changes. Multiple-sensor paring algorithms is presented. With these algorithms, the robustness and locating precision of LDS (Leak Detection System) can be ensured. © 2012 IEEE.