Desirable SPR Policies in Coping with Various Market Uncertainties: A Markov Decision Process Analysis

AbstractThis paper examines the impacts of market uncertainties on Strategic Petroleum Reserves (SPR) policies by suing a Markov decision process model. Four market uncertain factors have been considered including oil price, disruption probability, and disruption magnitude and disruption duration. A representative numerical case is employed to validate the practicality of our model. The proposed method and case study contribute to support China SPR decision making in three aspects: (1) with given information of. oil price and disruption probability, the model finds optimal SPR size for China; (2) in normal state, the model helps to decide oil acquisition amount based on current SPR size and market condition; (3) in disruption state, the model can help to find optimal SPR drawdown and refilling size by putting in disruption magnitude and duration