Monthly schedule of crude refinery: Multi-scale strategy and multi-criteria objective

•Optimization of the monthly schedule of a realistic industrial refinery.•Downstream residue demands to largely determine the optimal solution.•Formulation of an exact penalty model to optimize the sequence of distillation.•Multi-scale solution strategy in both space and time leads to fast solution...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Computers & chemical engineering Jg. 164; S. 107866
Hauptverfasser: Zhang, Chenguang, Ricoux, Philippe
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier Ltd 01.08.2022
Schlagworte:
Crude refinery
Distillation sequence
Multi-scale analysis
Residue tanks
Scheduling optimization
Crude refinery
Multi-scale analysis
Residue tanks
Scheduling optimization
Distillation sequence
ISSN:0098-1354
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:•Optimization of the monthly schedule of a realistic industrial refinery.•Downstream residue demands to largely determine the optimal solution.•Formulation of an exact penalty model to optimize the sequence of distillation.•Multi-scale solution strategy in both space and time leads to fast solution speed. The optimization of crude refinery schedule, a subject of extensive research, frequently results in large-scale mixed-integer nonlinear programming problems. This paper optimizes a realistic industrial refinery's schedule over a one-month horizon. To achieve a rapid solution time, we developed a novel multi-scale strategy that treats the optimization model at two temporal (longer horizon divided into shorter periods) and two spatial scales (the refinery and its coarse-grained version). Problems at the shorter temporal and coarser spatial scales are solved quickly and contribute to the efficient solution of the original problem. Among other novelties, the model includes the frequently overlooked residue tanks downstream of the CDU, which have a significant influence on the solution property. Additionally, an exact mixed-integer linear formulation of the crude swing penalty was proposed, allowing optimization of the distillation sequence as part of a multi-criteria objective. Several examples demonstrate the improvement in computing time and solution quality.
ISSN:0098-1354
DOI:10.1016/j.compchemeng.2022.107866