Defesa de Exame de Qualificação – Leonardo Salsano de Assis – 19/05/2017

19/05/2017 16:53
Defesa de Exame de Qualificação
Aluno Leonardo Salsano de Assis
Orientado Prof. Eduardo Camponogara, Dr. – DAS/UFSC
Data

Local

19/05/2017  14h30   (sexta-feira)

Sala PPGEAS I (piso superior)

  Prof. Jomi Fred Hübner, Dr. – DAS/UFSC (presidente)

Prof. Guilherme Ernani Vieira,  Dr. – EPS/UFSC

Prof. Roger Rocha, Dr. – Petrobras

Prof. Werner Kraus Junior, Dr. – DAS/UFSC (suplente)

Título

 

Operational Management of Crude Oil Supply: Models and Algoritms
Abstract: An important problem faced by oil companies is the supply of crude oil from platforms to refineries, namely the crude oil supply planning problem. This problem is usually found in vertically integrated oil companies, which control production, transportation, storage and refining.

In offshore platforms, crude oil is produced and then transferred to onshore crude oil terminals by sub-sea pipelines or shuttle tankers. After being delivered at the terminals, the crudes are sent to refineries and will be processed in crude distillation units (CDUs) to produce oil products, which will be delivered to chemical, pharmaceutical and energy industries, and end consumers. Since oil pipelines are not available in deep-water offshore oilfields, the oil company relies on platforms, to produce and store crude oil. For the large number of platforms, a fleet of shuttle tankers is needed due to high volume of oil that must be transferred from the platforms to crude oil terminals. After arriving to the terminals, shuttle tankers upload crude oil through a pipeline into storage tanks. Then, crude oil can be pumped to the refinery. At the refinery, the crude oil arriving from the pipeline is stored in charging tanks and subsequently sent to the CDUs. The mixture or blend of different types of crudes is tracked by bilinear terms in blending equations, which are non-convex non-linear functions.

Among the decisions in each decision level of the crude oil supply planning problem, the strategic level is responsible for defining the demands (total volume and type of crude oil) of the refinery for the long-term, as well as crude oil import/export decisions. The tactical level includes more detailed constraints and is concerned with the medium-term resource allocation. This level decides which platforms will feed each crude oil terminal; which terminal will supply each refinery; and the vessel fleet composition. Also, decisions involving material flow such as the shipments of crude oil between platforms and the crude oil terminal, as well as between the crude oil terminal and the refinery are carried out for each period of the planning horizon. The operational level is fed with upper level decisions, which are concerned with routing and scheduling of operations.

This work intends to provide a systematic scheme composed of models and algorithms for the operational management of crude oil supply. This problem can be classified as a combined short-sea maritime inventory routing and crude oil scheduling problem. It integrates entities from the upstream (i.e., production platforms, oil vessels and the crude oil terminal) and midstream (i.e., refinery) segments, as well as tactical (i.e., inventory management and inventory allocation) and operational (i.e., scheduling) decisions, resulting into a non-convex mixed-integer non-linear programming problem (non-convex MINLP).