Optimization under uncertainty of the integrated oil supply chain using stochastic and robust programming

Abstract

This paper proposes the development of a strategic planning model for an integrated oil chain considering three sources of uncertainty: crude oil production, demand for refined products and market prices. To deal with these uncertainties, three formulations are proposed: (1) a two-stage stochastic model with a finite number of realizations, (2) a robust min-max regret model and (3) a max-min model. These models were applied to Brazil's oil chain, comprising 17 refineries and three main petrochemical plants, 16 groups of crude oils, 50 intermediate products, 10 final products, 13 terminals and a logistic network composed of 278 transportation arcs relative to the road, water, rail and pipeline modes. The time horizon analyzed covers 10 years. The results indicate significant financial differences between the three formulations, depending on the agent's risk profile.

Fuzzy sets to model master production effectively in Make to Stock companies with Lack of Homogeneity in the Product

“…Many authors propose handling such LHP uncertainty through stochastic models by considering different scenarios. In the petroleum sector, these are: Carneiro et al [13] consider the composition of intermediate products as uncertain parameters; Luo and Rong [14] define that properties for components are uncertain parameters with continuous probability distribution; Ribas et al [15] deal with density and viscosity as uncertain parameters. In the remanufacturing sector, Aras et al [16], Denizel et al [17] and Zeballos et al [18] consider that returned products are categorized in relation to their quality.…”

“…Constraint (15) calculates the number of orders from each customer class k based on the total amount to be sold and the mean order size of the corresponding customer class. Constraint (16) is analogous to constraint (15), but refers to backorders.…”

Sustainable refined products supply chain: A reliability assessment for demand‐side management in primary distribution processes

“…Researchers have done some relevant studies on the design and planning of the oil supply chain . The mathematical models developed in their studies were used as efficient tools to determine the distribution plans.…”

Finding Common Ground When Experts Disagree: Belief Dominance Over Portfolios of Alternatives

“…This approach differs from the literature on ambiguity aversion and robust optimization in that those methods use some version of a worst-case analysis. Worst-case analyses can range from straightforward Maxmin rules (Gilboa & Schmeidler 1989;Ribas et al 2010), to more sophisticated methods such as Klibanoff's smooth ambiguity (Klibanoff et al 2005) All of these approaches evaluate a set of alternatives against robustness criteria using scenario discovery to identify key uncertainties. They do not generally provide a single best solution, but rather focus on clearly communicating how the implications of different alternatives compare in terms of robustness.…”