Proposal for a Computational Intelligence in Modelling Leak Detection for Crude Oil Pipelines

EasyChair Preprint 13467

Abstract

The transportation of hydrocarbon deposit beneath the earth surface are mostly safer and cheaper using pipelines. However, these pipelines are frequently challenged with damages such as corrosion, manufacturing errors and environmental degradation that often lead to leakages and can result into gross economic lost and explosion which could also be a threat to lives. Leak detection is a technique used in the Oil & Gas Industry to monitor variation in essential properties of the hydrocarbon product such as pressure or temperature and compare these values at different intervals on the pipeline between the inlet and the outlet. This technique is highly inefficient when a real-time and accurate point of leakage is desired. In most case, multiple excavations need to be done before an exact leak location can be detected. With abundant records of previous leakages and recorded reading of these leak detection system such SCADA, a machine learning architecture is proposed in this paper and a conceptual framework was developed to achieve the paper’s proposed idea.

Keyphrases: Computational Intelligence, Machine Learning., SCADA, leak detection, oil pipelines

@booklet{EasyChair:13467,
  author    = {Ridwan Salahudeen and Wasilah Sada and Aminat Yusuf},
  title     = {Proposal for a Computational Intelligence in Modelling Leak Detection for Crude Oil Pipelines},
  howpublished = {EasyChair Preprint 13467},
  year      = {EasyChair, 2024}}