The thermoelectric part

Here you can find an explanation on why commercial thermoelectric generators (TEGs) are not suitable for hybrid thermolectric - photovoltaic systems.

The light that the Sun irradiates in the space is not monochromatic but is spectrum with wavelengths ranging between 200 and 4000 nm. Above the Earth’s atmosphere the total incident power density (the integral over wavelength of the spectral power density) is around 1300 W/m². This value reduces to ∼ 1000 W/m² at the sea level, because of the various absorptions coming from the different components of the atmosphere (in this case the spectrum is called AM 1.5, namely Air Mass 1.5).

Ignoring heat losses, if we take the incoming power density and the typical thermal resistance of commercial thermoelectric generators (TEG), we can determine the difference of temperature across the thermoelectric generator using Fourier’s law

ΔT = Φ Rteg

where ΔT is the difference of temperature across the TEG sides, Φ is the incoming power density, and Rteg the TEG thermal resistance. Inputting 1000 W/m² as the incident power density and 0.001–0.005 m² K/W as the typical TEG thermal resistance we obtain a ΔT ranging between 1 and 5 K. With such a small ΔT, the resulting thermoelectric efficiency is only 0.05–0.3%. It is clear that these efficiency values are too small to arouse any interest for practical applications of STEGs. That's the main reason why commercial TEG devices ar enot suitable for PV hybridization.

In order to increase the ΔT and thus the STEG efficiency, one can act on the system configuration. In particular, from the above equation it is clear that the two possible solutions are 1) increase the value of the incoming power density Φ, and 2) increase the TEG thermal resistance Rteg. In the first case it is possible to implement optical concentration, by means of lens or mirrors. In the second case it is possible to use thermal concentration increasing the ratio between the absorbent area, and the active thermoelectric area, given by the sum of all the thermoelectric leg areas, as shown in the figure.

In both cases a concentration ratio of ten, means a tenfold increase of the ΔT, and a consequent increase of the STEG efficiency. It is also possible to use both strategies together. In that case a concentration ratio of ten for both concentration options means a 100fold increase of the ΔT. Taking the above example, this would mean a ΔT of 100–500K, and a STEG efficiency ranging between 5 and 14%. In this perspectives STEGs are much more appealing.

Of course this is valid only because in the balance of the above equation we are not considering any source of heat losses. Thus the only parameter setting up ΔT is Rteg. In real cases it is not so easy to increase the difference of temperature across the TEG sides since different source of heat losses (such as convective and radiative heat losses) take place. That’s the reason why normally solar thermoelectric system are designed to work with a proper encapsulation which prevent heat losses.