Use of Selected Machine Learning in Detecting Leakage in Piping Systems

Leakage in the piping system is a serious problem for the oil and gas sector since it can result in accidents, fatalities, and environmental degradation. The oil and gas industry generates a lot of data on a daily basis and this data being generated can be used to proffer mathematical model solutions with the use of Machine Learning. Engineers working in the petroleum and gas industry are starting to research the ways that techniques like data science, machine learning/ intelligent models and artificial intelligence might be used to address issues that arise in the sector. Using simple parameters like pressure, temperature, and flow rate, this project aims to study the capacity of intelligent models to forecast corrosion rates that will ultimately result in leakages in a piping system. It also compares intelligent models to ascertain which is best and more accurate at identifying leaks in pipes in the oil and gas industry. In this project, a well data stream will be utilized for examining five intelligent models (Extra tree classifier, Light Gradient Boosting, Random Forest, Nu Support vector model and Support Vector Machine models). The result of this project would be useful to the engineers in the oil and gas industry, researchers and academicians who are interested in using Data Science to solve problems around piping system leakage in the oil and gas industry.