Photo-voltaic (PV) power generation is promising from an environmental and economic standpoint. However, increased penetration of PV power into the electric grid can present significant challenges to decision-making in energy markets. To maintain grid robustness, especially with renewable sources like PV power that are highly uncertain and variable, system operators seek the most accurate information available on the generation characteristics. In this regard, a scenario generation and reduction framework that captures the uncertainty and variability of PV power is proposed. This work characterizes the forecast error via a set of uncertainty and variability indices. A large set of scenarios is generated using a pseudo-random number generation process. Next, a scenario reduction framework to improve computational tractability is proposed. Finally, the efficacy of these methodologies is proven by observing their impact on the unit commitment solution via a cost-benefit analysis. The proposed work is tested using measured and forecasted PV power data for a specific geographical location in North Carolina. The entire framework is built using MATLAB/Simulink and GAMS Optimization software. The results indicate that incorporating scenarios in the deterministic unit commitment model improves overall operational costs while capturing the uncertainty and variability of PV generation.