Advanced Analytics for Diesel Flash Point Inferential

Diesel flash point is an extremely important indicator of the quality of diesel products. In order to achieve great quality control, the diesel flash point must be strictly tested and controlled, but this important indicator is difficult to measure due to technical or economic limitations. As an effective solution, data-driven inferential sensors provide stable and reliable online estimation of these variables based on historical measurements of easy-to-measure process variables. In this project, we will utilize advanced data analytics and deep learning technique to build comprehensive inferential sensors to estimate the diesel flash point. Compared with traditional prediction model, this inferential model based on deep learning, with much less design effort and maintenance cost, can fully mine process data information and provide much better estimation. This project will improve the diesel quality, reduce the cost of entire process, and bring huge economic and environmental benefits for petroleum refinery industry.

Faculty Supervisor:

Bhushan Gopaluni

Student:

Liang Cao

Partner:

Parkland Refining (BC) Ltd

Discipline:

Other

Sector:

Manufacturing

University:

University of British Columbia