| 초록 |
Most of refinery operation focuses on the fuel-side products, and crude inventory is generally controlled by designating the tanks into high/low sulfur content in crude oil. However, from the viewpoint of lube base oil production, chemical composition of the crude oil is much more important than bulk properties such as sulfur content. Therefore, the new operational strategy is highly necessary for the lube refinery. This work addresses procurement planning and crude inventory control of the lube refinery. The problem is formulated as a mixed integer non-linear programming (MINLP), which is to minimize the deviance between the real supply and the target demand of crude oils into the refinery. To reduce the computational burden of the large-size MINLP problem, two-stage decomposition method was used, in which the original MINLP problem is first relaxed into MILP. The crude purchasing decision which is obtained from solving the MILP is then fixed, and NLP problem is sequentially solved. As a case study, optimization problems for sulfur-based heuristic operation and the operation without restraints are solved, and their results are analyzed. |
