A Data-Driven Methodology for Strategic Siting of Offshore Wind Farms in South Africa

South Africa is undergoing a shift toward large-scale renewable energy deployment to meet energy demands and support national decarbonization efforts. Offshore wind holds significant potential to further support the transition toward a more sustainable energy system. However, the inherent intermittency of offshore wind, driven by variable weather conditions, presents challenges for reliable grid integration. This study proposes a methodology for the geospatial allocation of offshore wind energy capacity, using spatially and temporally resolved wind resource and electricity demand data. The approach leverages machine learning techniques, including wind resource clustering and optimization, to align wind energy supply patterns with time-dependent national demand and identify high-value offshore sites. The methodology is applied for a representative offshore region surrounding South Africa, which accommodates floating and fixed-bottom wind technologies. The performance of the methodology is evaluated under three system integration objectives, including minimizing total residual system demand, reducing residual load variability, and maximizing supply during peak demand periods. The results offer insights to guide strategic offshore wind deployment and underscore the value of capacity allocation methodologies in supporting long-term energy system planning. The proposed methodology is adaptable to other regions or integration objectives through context-specific inputs.