This work addresses primarily the scheduling of a refinery importing various types of crude oil. The refinery operation discussed in this paper involves the unloading of crude oil to storage tanks, the transfer and blending from storage tanks to charging tanks and crude oil distillation units, and the arrangement of the temperature cut-points for each distillation unit. The paper describes a simulation-based approach to the refinery operation, which is modelled as a pooling problem. The proposed approach uses a random-search formulation, which allows for controlling search depth, breadth and solution quality, as well as computational effort. An implementation of the proposed approach is presented in a real case scenario.