For reliable operation of the electricity grid, the grid operator needs to know detailed information about the system at any given time. This information usually comes from readings of different devices on the grid, or from different estimation algorithms. Having accurate estimations become more challenging due to increasing amount of PV generation, which increases the potentials of reverse power flow in the grid and elevates the risk of imbalance between generation and demand. This project will involve combining a physical PV model, machine learning tools, weather, and electricity load data to provide probabilistic PV estimations. The probabilistic estimation will provide a confidence band around PV generation that can help Australian Energy Market Operator (AEMO) analysts assess the effects of potential PV variability on operational demand forecasting across the National Electricity Market (NEM).