Oil production forecasts and their uncertainty analyses

Abstract

Abstract The Hubbert, multi-cyclic Gauss models can be used to forecast oil production of Ordos Basin in central China and Bohai Bay Basin in eastern China. Oil accumulation trap types (structure vs. stratigraphic traps) and exploitation technique variations contribute to different cycles of oil production for Ordos Basin. Oil production from different depressions with various time lags give rise to multiple cycles of total oil production for Bohai Bay Basin. The results show that production predictions are sensitive to the choice of model and that different models give different forecasts. The forecasts of the Hubbert and Gauss models predict that the next peak annual oil production of 35 million tons occurs around 2026 for Ordos Basin, and the next peak of 75 million tons occurs around 2020 for Bohai Bay Basin. Many economic, political, geological and technological variables may affect future oil production. Those include unstable markets, changeable rate of demand, discontinuous exploration efforts, unconventional oil estimates, and new exploration plays. If any of these variables differs significantly from the assumption on which the predictions are made, the forecasts will be different from the predictions in this paper. Production forecasts are undoubtedly important to petroleum exploration and exploitation planning by oil companies and a government’s economic forecasting. However, the end users of such predictions should know the limitations of the methods and the uncertainties in the resulting predictions.