Oil refinery procurement strategies: A Bayesian approach

Zhixue Xie,Li Zheng
2012-01-01
Abstract:Current oil refinery procurement strategies do not deal well with purchase price uncertainties on the international crude oil spot market with fluctuating prices. A dynamic programming model was developed to minimize the expected present value of the total cost of oil purchases. Stochastic calculus is used to find the static optimal solution, with a Bayesian decision framework then introduced to find the adaptive strategy. Real data for 26 years of historical prices of one marker on the crude spot market was used to test the effectiveness of the Bayesian-based procurement approach. The results show the potential of this strategy for reducing costs compared with current refinery practices.
What problem does this paper attempt to address?