The refining industry is faced with medium and long term planning problems (e.g. crude supply, stocks, capacity decisions) where an important part of the data is uncertain. Deterministic optimization may in this case lead to non optimal decisions. In this paper, we discuss the potential contribution of the stochastic approaches, discuss their inherent complications and means to overcome them, and present development possibilities and computational procedures.
Paru en mars 1991 , 23 pages