PetroML: Machine learning for accelerating the optimal solution of oil blending problem



Oil blending problem is a “billion dollar” business arising from the petrochemical industry. It involves blending several resources or materials to create one or more products, such as gas, and jet fuels, corresponding to demands. Optimizing the ways the resources are blended can lead to millions of dollars in daily savings. In this project, students will collaborate with ExxonMobil Corporation on developing machine learning algorithms for solving the optimal blending problem. Students will learn using the Python programming language and machine learning algorithms for solving real-world energy systems problems.


Relevant Technologies:

Machine learning, mathematical programming, petroleum engineering


Prerequisite Knowledge/Skills:

  • Fluent in one of the programming languages: Python/C/C++/Julia
  • Knowledge of machine learning will be a plus
  • Instructor permission required


Meeting times:

  • Spring 2023 - TBD