Inverse optimization investigation for thermoelectric material from device level

Abstract

Thermoelectric generators (TEGs) with improved conversion efficiency are in great need for low-grade heat recovery. Existing studies primarily optimize the dimensionless figure of merit (ZT) of thermoelectric (TE) materials to improve TEG efficiency. However, TE material with a high ZT cannot guarantee that associated TEG has an optimal performance in practical conditions. To solve this problem, an experiment-verified model is proposed considering the effective temperature difference, optimal matching resistance and contact effect. Sensitivity analysis has been conducted to inversely identify the optimization direction of TE materials at the device level. The impacts of the key physical properties on TEG performance are systematically studied under typical operating conditions. The study results show that reducing lattice thermal conductivity is of more priority than increasing the power factor for improving TEG performance. For power factor improvements of TE materials, Seebeck coefficient optimization is found to be more important than electrical conductivity increase. Besides, the impacts of operating conditions on optimizing TE materials are also investigated. Finally, an optimization process to improve the generation performance of TE materials is proposed, which opens up a new way to lead the development of TE materials from the device and application level. The study results are helpful to effectively improve the power generation performance of TEG through the proposed optimized method of TE materials.