Hi Ta lei: A good question. Heat flow = heat current is usually measured over a macroscopic distance > 100 micron m, requiring accuracy in the sample 3D dimensional measurements of this level or higher. HY 

I think, you must wait a few years.

Hi Chou,

Three thermoelectric (TE) properties characterize  the "TE conversion efficiency" of TE materials,  Seebeck coefficient, \alpha,  electrical resistivity, \rho, and thermal conductivity, \kappa,  by the figure of merit Z. "Bulky modules" are designed with different configurations and TE materials for increasing the thermal efficiency. Experimental determination of the heat flow  in TE modules requires different conditions in according  with required accuracy.

Any TE module has an approximately square shape and therefore 'edges'. The efficiency at the edges, being next to the cooling system is much better than in the centre of the module. The efficiency is based on Delta T which varies across the TE module depending on where the cooling is more efficient. Unless the TE is treated as a whole, no single point will behave in a 'characteristic' manner. Only the TEG viewed as a system has 'average' performance.

Each individual element making up the TEG contributes to the output, but they are wired in sets and the sets are connected in parallel. Only the whole set of sets has a rated performance based of average Delta T. Hence, it is treated as a single object. 

To represent 'real performance' you would scale the analysis to the size of the individual elements (so you would have to know what they are).