Leveraging the Power of Probabilistic Modelling to Forecast Petroleum Production Using Well Performance

Abstract Accurate forecasting of oil production is critical for optimizing offshore energy projects, where high operational costs and environmental risks demand reliable predictions. Traditional forecasting models, such as Hubbert's Curve and Decline Curve Analysis (DCA), fail to address the dynamic and daily fluctuations in production, limiting their applicability in real-time decision-making. This paper introduces a novel hybrid forecasting framework that integrates Shapley Additive exPlanations (SHAP) and Monte Carlo simulations to enhance predictive accuracy and uncertainty quantification. The framework captures regime-specific production patterns and offers probabilistic forecasts, enabling more informed, risk-aware decision-making. By capturing daily production fluctuations, the methodology improves operational efficiency, resource management, and financial planning. Empirical results demonstrate that the proposed framework outperforms traditional models in both accuracy and uncertainty quantification, making it a transformative tool for the oil and gas industry.