Maximizing exergy and energy efficiency in PV/TEG hybrid systems through beam splitting under best concentration conditions

High-concentration photovoltaic (PV) systems frequently undergo severe thermal losses, which negatively impact cell operation and accelerate material degradation. This work investigates a hybrid PV–thermoelectric generator (TEG) system with spectral beam splitting under concentrated sunlight conditions. A cold mirror beam splitter redirects visible light to the PV cells while allowing infrared (IR) radiation to pass through to the TEG modules. This approach effectively maintains lower temperatures within the PV cells and thus recovers the heat produced by IR radiation. Notably, the Best Concentration Condition (BCC)—defined as the highest concentration ratio at which passive cooling is feasible—was experimentally determined at 16 suns for the hybrid system, compared to 8 suns for the PV-only system. The hybrid system achieved a peak power density of 1,342 W/m² at the 16 suns concentration, representing a 42 % increase compared to 942 W/m² measured by the PV-only system at its BCC. The use of spectral splitting raised the electrical conversion efficiency of the PV component from 14 % (without splitting) to 18 %, while the TEG module achieved an IR-based energy recovery efficiency of 3 %. Consequently, the integrated system delivered a combined electrical efficiency between 11 % and 12 % and an exergy efficiency between 8 % and 10 %, with the PV segment performing at an exergy efficiency of 22 % when using spectral splitting (versus 14 % without splitting). A techno-economic analysis showed a 43 % increase in annual energy production (584 vs. 409 kWh/m²), all while maintaining a comparatively steady levelized cost of electricity (∼0.045 USD/kWh), and a total reduction in CO₂ emissions of 2.2 t/m² over 25 years. The findings illustrate that operation at the BCC, when combined with spectral beam splitting, considerably enhances both energy and exergy performance without the addition of mechanical complexity.