The goal of this MSc. dissertation is to compare the performance of three classical models of numerical simulation (Duffie and Beckman, King et al. and Masters) to generate electricity from the photovoltaic conversion obtained in panels, taking the model of King et al. as a baseline since it was validated vis-àvis experimental data within 2.4 percent to 5.4 percent. The simulations make use of official data (solar hourly and monthly data of solar incident radiation, temperature, wind speed) provided by the National Institute of Meteorology of Brazil and of performance data provided by the manufacturer of the photovoltaic panel. The motivation for development of the work resulted from the perception of the importance of photovoltaic energy as a strategy for diversification of the Brazilian energy matrix. The proposed methodology (i) compares the performance of the three models studied and (ii) applies it for assessing the performance of the photovoltaic conversion in eleven Brazilian cities that exhibit favourable conditions for photovoltaic conversion, thus creating a scale that can be used to establish a national benchmark. The results revealed that the Duffie and Beckman model reproduces the results of the photovoltaic conversion model obtained by King et al. degree within 2.4 percent, while the agreement with the model of Masters over range accuracy (7.4 percent). In conclusion, the use of monthly averaged meteorological data proved to be effective for estimating the energy that results from the process of photovoltaic conversion when hourly data is not available.