Implementation of a hierarchical cluster model to analyze wind and solar availability in the department of Antioquia, Colombia

Abstract

This research presents a systematic spatiotemporal assessment of solar radiation and wind speed patterns throughout the Department of Antioquia, Colombia, employing advanced machine learning methodologies to evaluate renewable resource potential. The investigation aligns with Colombia's strategic energy transition framework, which targets a 51 % reduction in greenhouse gas emissions by 2030 relative to baseline projections. Through hierarchical clustering analysis, five distinct groups of municipalities were identified, revealing significant solar potential with radiation averages between 4.6 and 5.0 kWh/m2, reaching maximum values of 5.4 kWh/m2 in the Magdalena Medio subregion. This solar availability exceeds 60 % of Colombia's maximum reported values, indicating substantial potential for photovoltaic (PV) and solar thermal projects. In contrast, the wind resource assessment shows limited viability for utility-scale wind energy development, with speeds rarely exceeding 3 m/s even at a height of 100 m, resulting in capacity factors between 4.1 % and 11.8 % in coastal onshore and offshore areas, respectively. Temporal analysis across solstices and equinoxes confirms year-round consistency in solar radiation, attributed to the region's equatorial location, while maintaining stable patterns across the identified clusters. The demonstrated synergy between reliable solar resources and Antioquia's abundant hydroelectric infrastructure supports the development of hybrid systems, particularly solar-hydroelectric integration, with promising applications in both electricity generation and green hydrogen production. Furthermore, the Magdalena Medio and northern Urabá subregions emerge as prime candidates for solar energy projects, offering strategic opportunities for industrial-scale renewable deployment. This research provides crucial insights for Colombia's decarbonization roadmap, emphasizing Antioquia's significant solar energy potential as a cornerstone for transitioning towards a sustainable and diversified energy matrix.

Keywords

• Hierarchical cluster
• Machine learning
• Solar energy
• Wind energy
• Antioquia
• Colombia