Deep Learning Optimization to Improve Refinery Operations

Abstract

Refining companies have invested in control systems, advanced process control and real time optimization applications to optimize their process and operations. To enhance the efficiency already achieved with the current automation platforms, artificial intelligence (AI) offers opportunities to operate at the higher yields and margins by operating closer to product specification and diminishing product giveaway. In addition, AI can enhance the current automation platforms and applications by providing real time predictions and faster identification of operational issues.

This paper describes the implementation of advanced process control and real time optimization through deep learning optimization to achieve the multiple goals of the refining processes while safeguarding the safety and environmental integrity of operations.