Pipeline leakage Prediction

Supervised/Unsupervised Deep Learning

Pipeline integrity (PI) programs are in place to ensure the integrity of TC Energy’s (TCE) pipeline operations are reliable, efficient and safe with minimal environmental impacts. The PI program consists of threat identification (i.e. inspections, corrosion or leak identification etc..) and maintenance work that focuses on repairs and replacements.

TC Energy’s Canadian Gas Division, has made progress in late 2020 with their piMiner initiative and they were able to develop an ML model around their Keystone Pipeline Asset using a single datasource. With some of the data migration work completed there are new data sources which could complement and help refine the model. The purpose of the engagement is to refine, or rebuild the model to enable Canada Gas to prove or disprove the hypothesis they are working to deliver. The engagement would set the foundation for understanding on key next steps around moving forward on a broader project.“