Optimal Retrofitting of Conventional Oil Refinery into Sustainable Bio-refinery under Uncertainty

This paper focuses on a novel multistage retrofitting problem from conventional fossil-based refinery into biomass-based refinery with the aim of sustainable development of fuels. Given a typical refinery and potential biomass-based technologies, the problem has an objective to integrate the latter into the former by making use of existing units in the new process(es). A mixed-integer linear programming model that considers a ten-year long retrofit planning is formulated. Furthermore, the deterministic problem is extended to a multi-stage stochastic programming problem under both endogenous yield uncertainty and exogenous demand uncertainty and solved via a Lagrangean decomposition algorithm. The results provide flexible design alternatives by determining the units that should be added or modified for the selected biomass-based technologies. Regarding uncertainty, different schemes with strategic and operations decisions are determined. The results show advantages in considering retrofitting problem with detailed operation constraints for each year.