EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

1955. An Iterative Optimization Framework for Streamlining Crude Oil Scheduling in Refinery Operations

Invited abstract in session TD-37: Applications of Multiobjective Optimization, stream Multiobjective Optimization.

Tuesday, 14:30-16:00
Room: 33 (building: 306)

Authors (first author is the speaker)

1. İrem Marttin
R&D, Turkish Petroleum Refineries Corp. (TUPRAS)
2. Özgür Kabak
Industrial Engineering Dept., Istanbul Technical University
3. Melike Kamuran Onat
R&D, TÜPRAŞ
4. Büşra Aydın
Turkish Petroleum Refineries Corp (TUPRAS)
5. Ahmet Nomer
Turkish Petroleum Refineries Corp. (TUPRAS)
6. Oğuz Şüyün
Turkish Petroleum Refineries Corp. (TUPRAS)

Abstract

Refineries are complex industries that take crude oil then fractionate and crack it into valuable products. Feed continuity and quality are very crucial for the production, so the petroleum scheduling is a key step for a stable operation. This study aims to develop an optimization tool addressing operational limits encountered practical contexts to facilitate adept decision-making in crude oil scheduling. To this end, a scheduling model was devised, covering the unloading of crude oil received via pipeline to multiple storage tanks, intra-tank crude oil mixing, and charging the resultant mixture to the crude distillation unit. An event-based continuous-time Mixed Integer Linear Programming formulation was created to effectively navigate the intricacies of the scheduling problem. Given the nonlinear problem stemming from the homogenous crude oil mixture in the tanks, an iterative solution approach was designed. This approach breaks down the problem into manageable sub-problems based on periods of constant tank compositions. The viability of the proposed model has been evaluated through empirical assessments conducted on real refinery cases. The findings revealed satisfactory results within a brief interval, especially given the industrial scale and complexity of the problem covering 4 tanks with around 10 crude oil types. In this regard, the model aims to notably shorten scheduling time while enhancing efficiency through steady operations facilitated by prompt decision-making.

Keywords

Status: accepted

Back to the list of papers