A comparative study to determine the photovoltaic annual energy generation in the Brazilian Northeast

Abstract

Renewable energy sources have been growing worldwide and solar energy is a significant part of such sources. An essential step in planning studies, including the problem of optimal location and sizing of photovoltaic (PV) generators, is estimating how much energy the panels will generate over time. For that, two aspects must be considered: the stochastic nature of the input variables and the way to calculate the generator's output power. A brief literature review identified 12 approaches to determine the PV output power. Therefore, this article compares such methods through actual meteorological data and generated energy over 1 year by a solar power plant located in a specific site in the Brazilian Northeast. An analysis is carried out on the influence of correlations between meteorological variables in estimating the generated energy. There is no comparative work in the literature on the analytical models for calculating the PV output power in conjunction with the correlation analysis. As expected, models that consider the influence of ambient temperature on output power performed better than those that do not. How the energy generated by the PV panel is estimated can influence the economic viability of a project since oversizing the PV system entails unnecessary additional costs. Monte Carlo simulations generate data samples used by the studied models for power output calculation. According to our findings, considering the correlation between meteorological variables reduces the error in estimating the generated energy by PV panels.