ICT2014 Abstracts

Based on our observation, as smaller the amount of available exergy, the amount of available energy becomes exponentially large. Although heat recovery from automotive exhaust is getting popular, the cost doesn’t match the expectation. In further low temperature heat sources, it has been challenging to find a reasonable cost in $/W and it is a barrier for commercialization.

We present a performance and cost analysis with an optimum design for waste heat recovery in source temperature 60 ^oC, e.g. water drain after cooling computers in data centers in scale of 100s MW of heat, and the available ambient temperature 20 ^oC.

Throughout the sensitivity analysis of thermoelectric properties, thermal conductivity appears to be most influential for the cost per performance rather than the power factor as the product of electrical conductivity and square of Seebeck coefficient. This is due to the internal thermal resistance match similar to external electrical load match. As we know, fractional area coverage (fill factor) of thermoelectric element is a key parameter to reduce the cost. We optimize the fill factor considering a minimum treatment of the module while considering parasitic heat losses and energy payback including pump power for the heat transport. A metric of the cost efficiency is then thermal conductivity divided by the product of the material price in [$/kg] and the heat transfer coefficient [W/m²K]. Considering ZT = 0.5 for the material, the reduction of thermal conductivity will results less than 1.0 [$/W] for the thermoelectric module and the cost dominant is on the heat sinks, which may easily over 10 [$/W] for 100s W/m² of available heat flux. Cost effective heat sink becomes the most challenging issue for the waste heat recovery, if the thermoelectric is engineered and designed to optimum. The results suggest a thermoelectric potential in waste heat recovery.