The inevitable growth in demand for energy resources, imposed by times of accelerated economic and social transformation, and the impositions of the climate agenda, encourage interest in the study of more efficient and less polluting processes. In this context, performing simulations in MATLAB/Simulink, this work carries out the sizing, simulation and optimization of a hybrid power generation system, disconnected from the grid (off-grid), and composed by generator sets and battery banks, aiming commercial applications with different levels of demand, in different seasonal scenarios. It is also object of study to evaluate the gains in efficiency and the emissions of atmospheric pollutants, comparing the simulated scenarios from baseline estimates. The systems are composed by diesel generators (baseline) with output power between 600 and 5200kW, Natural Gas generators with output power between 400 and 5088kW, and battery banks with capacities between 900 and 9000kWh for alternative scenarios. An energy analysis of the systems was performed, considering energy variation, state of charge, C rate, engine load levels, specific fuel consumption and efficiency. In addition, the analysis of pollutant emissions from generators. Finally, with the model s output variables, the results were compared between the different cases and the baseline scenarios, in addition to analyzing the influence of the factors used to create the simulated scenarios (types of generation, load variability and multipliers). In comparison between the baseline and the most promising scenario (Natural Gas + battery banks), it was possible to observe average gains with efficiency increases of up to 19.8 percent for the summer, and 26.7 percent for the winter. In addition to detecting increased emissions of unburned hydrocarbons and carbon monoxide, and reduction in emissions of nitrous oxide, particulate matter and carbon dioxide.