Assessing complementarity and optimising the combination of intermittent renewable energy sources using ground measurements

The use of intermittent renewable energy sources (IRES) in traditional energy systems must be suitably planned, as their variability affects the reliability and costs of electricity generation. Assessing the complementarity among multiple IRES and determining their optimal mix can mitigate the impacts of variability from independent sources. Although complementarity studies have considered a single time series per source to evaluate complementarity, this approach is not suitable for large areas, because the spatial variability may be lost when multiple time series are merged. In contrast, if data are not merged, optimisation might lead to impractical results. We propose the use of factor analysis for grouping similar data and present two optimisation models that are easily reproducible and not restricted by the number of decision variables. To this end, we considered the south-eastern and central-western areas of Brazil, where the share of IRES remains small but is expected to increase, and includes global solar radiation, wind speed, and river inflow IRES. The river inflow data were clustered into 11 groups. As solar radiation is evenly distributed in the region, rendering models with two or more factors unsuitable, all measurement stations were grouped. In contrast, as wind speeds vary significantly, the corresponding data were not merged but treated separately. The complementarity between data groups was assessed using the Pearson correlation coefficient. Overall, the different IRES exhibited low correlation, suggesting their complementarity. Despite their different objectives, both optimisation models indicate that the photovoltaic generation share in the IRES mix should be around 50%. Obtaining such optimal mixes can guide the expansion of IRES adoption for electricity generation in the region.