Predicting Oil-Water Flow Patterns through Machine Learning
This research was presented at the 7th YSJ conference at Cambridge University on October 2, 2019.
The flow of oil and water in a pipe presents a specific structure also known as the flow regime. The pressure drops of oil-water in pipelines can be dramatically reduced through a water annulus, next to the pipe wall, and an inner oil core. This approach can improve oil transportation efficiency significantly. Current flow regime prediction models have an accuracy of about 70%.
