Energy Harvest Predictions for a Spectrally Tuned Multiple Quantum Well Device Utilising Measured and Modelled Solar Spectra

We present a comparison between modelled and measured solar spectra in Nicosia, Cyprus. The modelled spectra were generated using the SMARTS model, driven by various sources of measured atmospheric data, updated every 30 minutes or less. A comparison with measured spectra reveals the most reliable source of data for that location and period. The spectral simulations demonstrate that both aerosol optical depth and precipitable water content must be accurately known at the location of interest in order to realistically recreate the shape and power of measured spectra accurately. Energy harvest calculations of four triple junction (3J) solar cell designs were performed using the simulated spectra. The model predicts a 5.7% increase in energy harvest when multiple quantum wells are included in a standard 3J device. Simple modifications made to the quantum well device to ‘tune’ the cell to the incident spectra result in a 6.3% increase in predicted energy production over the standard 3J device.