Abstract

Chile stands out for its exceptional potential for non-conventional renewable energy production, positioning the country as a leading global hub for green hydrogen generation. However, this opportunity carries the critical responsibility of addressing the socio-environmental conflicts (SEC) that have historically affected energy projects in the country. This study proposes a novel self-organizing map (SOM)-based methodology to identify and classify optimal areas for the installation of photovoltaic plants dedicated to green hydrogen production in the Antofagasta region of northern Chile. This methodology explicitly integrates socio-environmental and technical factors within a single unsupervised learning approach. The SOM clusters potential areas into three conflict-sensitive typologies: (i) ideal zones, areas close to industrial infrastructure but distant from socio-environmental hotspots; (ii) promising zones, areas with moderate proximity to both industrial and conflict zones; and (iii) challenging zones, areas near SEC hotspots that present significant barriers for development. This typology supports early-stage, conflict-aware planning by highlighting areas where photovoltaic (PV)-based hydrogen projects can help decarbonize sacrifice zones while reducing the risk of new conflicts. The methodology delivers a transferable decision-support tool for the strategic placement of solar-powered hydrogen production facilities that aligns with principles of just energy transition and environmental justice, ensuring that socio-environmental considerations are systematically incorporated into green hydrogen planning.