Ocean surface albedo (OSA) is the fraction of incoming solar radiation reflected back by the ocean surface, driven mainly by wave geometry, whitecaps, and subsurface components like chlorophyll and mineral particles. The accurate modelling of this parameter is from importance for offshore floating photovoltaics (OFPV), where the bifacial panels can be used to boost the energy yield, impacting the levelized cost of electricity (LCOE) and energy transition strategies.
Current albedo datasets lack accuracy when considering OSA as a constant value, which is an oversimplification that ignores the sea dynamics. Moreover, the spatial resolution of these sources is commonly 1/24° (about 4km), which is a coarse resolution considering that the characteristic length scale of an OFPV plant is in the order of 1km. Empirical models focused specifically on OSA modeling exhibit an improvement compared to the existing datasets; however, they fall short by not including all the parameters that effectively influence OSA, and their empirical nature, which makes their accuracy location-specific. This thesis addresses these gaps, developing and validating a model based on GenPro4 to estimate OSA in realistic ocean scenarios.
The proposed model was tested in two main waterbodies classes: clear and turbid waters, to analyze and determine its accuracy under these two opposite sea states. Root mean square errors (RMSE) values of 0.72% for clear and 3.8% for turbid waters confirmed its robustness and reliability. Furthermore, a sensitivity analysis was executed to identify the key parameters that affect the bifacial gain when considering the dynamic nature of OSA, finding that sea roughness (waves geometry) defined by the significant height wave (Hs) and its peak period (Tp), had a significant effect, producing bifacial gain changes close to 3.5% between calm cases and rough wave scenarios.
Finally, regional simulations on four locations across the globe revealed that despite seasonal variations of bifacial gain, an annual average value ranging from 5 to 8% is obtained regardless of the specific location of the OFPV plant. These findings confirm that bifacial gain varies with different sea states but remains within a predictable and practical range for OFPV design and performance assessment.
Economically, this 5–8% yield uplift offsets the additional CAPEX from bifacial panels, potentially achieving breakeven or lower LCOE values compared to monofacial cases. Dynamic OSA thus drives bifacial viability, confirming competitive OFPV performance across oceans.