Abstract:
In this prospective paper, we sought to theoretically assess the magnitude order of the benefits related to an in-depth analysis of a high-profile decision-making problem of the oil industry. A simplified representation of the crude oil allocation planning problem in the State of São Paulo is focused on. The scope of the problem comprises 4 oil refineries, which are responsible for approximately one-half of the current Brazilian refining capacity, more than 260 crude oil types and 5 key properties of interest for primary processing in crude distillation units (CDUs). The optimization criterion was economic and considered the crude oil pricing, pumping costs through long crude oil pipelines, and the incomes from intermediary cuts in CDUs. Real- -world data, whenever available, were used to run the computation experiments. The major finding of this study shows that, depending on the problem instance considered, the planning model based on monthly time discretization can overestimate the profitability of the entire system by less than 1%, on average, in relation to its weekly discretized representation, the so-called “plannuling” model.
Reference:
JOLY, Marcel; MORIMOTO, Aline Heymi; ONODA, Kevin Akira; MIYAKE, Mario Yoshikazu; Le ROUX, Galo Antonio Carrillo. Refinery “plannuling”. Revista IPT Tecnologia e Inovação, v.5, n.17, p.25-44, ago., 2021.
Access to the article on the Journal website:
http://revista.ipt.br/index.php/revistaIPT/article/view/136/157
In this prospective paper, we sought to theoretically assess the magnitude order of the benefits related to an in-depth analysis of a high-profile decision-making problem of the oil industry. A simplified representation of the crude oil allocation planning problem in the State of São Paulo is focused on. The scope of the problem comprises 4 oil refineries, which are responsible for approximately one-half of the current Brazilian refining capacity, more than 260 crude oil types and 5 key properties of interest for primary processing in crude distillation units (CDUs). The optimization criterion was economic and considered the crude oil pricing, pumping costs through long crude oil pipelines, and the incomes from intermediary cuts in CDUs. Real- -world data, whenever available, were used to run the computation experiments. The major finding of this study shows that, depending on the problem instance considered, the planning model based on monthly time discretization can overestimate the profitability of the entire system by less than 1%, on average, in relation to its weekly discretized representation, the so-called “plannuling” model.
Reference:
JOLY, Marcel; MORIMOTO, Aline Heymi; ONODA, Kevin Akira; MIYAKE, Mario Yoshikazu; Le ROUX, Galo Antonio Carrillo. Refinery “plannuling”. Revista IPT Tecnologia e Inovação, v.5, n.17, p.25-44, ago., 2021.
Access to the article on the Journal website:
http://revista.ipt.br/index.php/revistaIPT/article/view/136/157
