Thermoelectric power generation

Publisher Website
Google Scholar
Abstract
Using existing semiconductor materials, it should be possible to produce thermoelectric generators with efficiencies of about 14%. Despite future improvement in material properties it appears unlikely that an efficiency greater than 40% of that of the Carnot cycle can be obtained. This renders thermoelectric devices unsuitable for the large scale production of electric power, and limits their usefulness to certain specialized applications where a suitable heat source and sink are available. When this is the case a generator which would have a power-to-weight ratio comparable with that of conventional small d.c. power sources seems feasible.

This publication has 3 references indexed in Scilit: