Using Artificial Intelligence Algorithms to Identify Factors of Methane Leaks from Gas Transmission Assets

EasyChair Preprint 8244, version 3

Abstract

GRTgaz has launched a major program for improving the assessment and the reduction of methane emissions due to leaks from its gas transmission network and facilities. The industrial assets of GRTgaz notably include thousands of gas delivery units, each containing pipes, gas pressure regulator(s), filter(s), shutoff valve(s), safety relief valve(s)… The leak detection campaign on the gas delivery units then requires significant resources and a lot of time. Targeting the assets that are the most likely to leak is therefore an important challenge for improving the campaign efficiency, moving forward the implementation of corrective measures, and reducing the methane emissions.

This work aims to explore XGBoost (Extreme Gradient Boosting) Cox survival regression model, associated with SHAP (SHapley Additive exPlanations) method, and Bayesian Networks using BayesiaLab software, to identify and explain the effect of different features on reliability of natural gas transmission assets, based on field feedback data.

Keyphrases: Artificial Intelligence, Predictive Maintenance, gas transmission, machine learning, neural networks

@booklet{EasyChair:8244,
  author    = {Amel Belounnas and Florent Brissaud and Elodie Rousset},
  title     = {Using Artificial Intelligence Algorithms to Identify Factors of Methane Leaks from Gas Transmission Assets},
  howpublished = {EasyChair Preprint 8244},
  year      = {EasyChair, 2022}}