Petroleum reservoirs are complex process systems defined by intrinsically uncertain data and a distinct pressure gradient. The upstream sector’s assets are with huge uncertainties and high risks. Thus, the investments in these complex geologic process systems majorly suffer severe dynamic risks due to the process’ complex dynamics, process data’s temporal and spatial variabilities, and data/model’s uncertainties. Over time, the complex dynamic risks of the reservoir production system have resulted in unforeseen severe production fluctuations, total process system failures, and/or abrupt well shut-in due to uncontrollable circumstances. Hence, the need to introduce a multipurpose dynamic risk-based smart production prognostic approach to address the outlined inherent petroleum production challenges. This thesis presents dynamic risks assessment models for dynamic risks-based analysis of petroleum reservoir production systems. Different possible production scenarios are captured with the developed adaptive hybrid model with the following highlighted novel scientific contributions. Firstly, a dynamic risk-based predictive model is introduced to forecast production and capture the parameters variabilities, data and model’s uncertainties, and dynamic risks of primary recovery processes. This is followed with an introduced novel model for dynamic risks monitoring and production forecast of secondary recovery processes. A novel model is also presented to incorporate dual reservoir energy support mechanisms in production predictions and associated dynamic risks forecast under lift mechanisms. In addition, a dynamic economic risks analysis model is proposed to consider economic risk assessment of the reservoir production systems. Lastly, a dynamic risks-based smart model is proposed to capture sand face pressure enhancement influence on the reservoir production system with production pump schemes.