Modeling and optimization of thermoelectric generator for waste heat recovery

• Main Author: Ji, Dongxu
• Other Authors: Josep Pou
• Format: Theses and Dissertations
• Language: English
• Published: 2018
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Electric power
• Online Access: http://hdl.handle.net/10356/75881
• Institution: Nanyang Technological University

Tremendous energy is dissipated as waste heat during industrial processes. Many waste heat recovery technologies have been proposed to convert some waste heat to useful energy. Among those waste heat recovery technologies, thermoelectric generator (TEG) has its distinct advantage of directly converting heat into electricity, reliability, long lifetime, no moving parts and no gas emissions. Thermoelectric materials have been investigated extensively in recent decades and the figure of merit ZT has been continuously enhanced. However, the thermoelectric module (TEM) and system development is rather stagnant, and the overall efficiency is quite low for practical applications. This has raised the urgent need for developing simulation and design tools for TEGs. Design and optimization relies on effective simulation tool development. However, the simulation of TEG, consisting of TEM and heat exchanger (HEX), is a complex problem as it involves thermo-electric coupling effect, solid state heat transfer and convective heat transfer as HEXs are usually adopted. To address this issue, the first part of this thesis proposes a numerical model to simulate the TEM and its heat transfer system. Both 3-D computational fluid dynamics (CFD) model and 1-D numerical model are developed and validated against experiments and the models have proven to be accurate enough. After the numerical model is developed and validated by experiments, optimization and design work are conducted on both TEM and TEG. Many previous optimization works assume fixed temperature boundary condition. However, this assumption is only applicable to limited practical circumstances. The effect of different boundary condition is not adequetely investigated. This thesis studies the effect of different types of boundary conditions on the optimization of TEM and TEG, and it is found that the optimized geometry parameter, thermoelectric elements height, and cross-section area vary significantly under different boundary condition assumption. With this finding, the TEM and HEX are optimized simultaneously by Taguchi method taking into consideration the interactive effect. Also, the contributions of each factor to the output power variance are quantified. This thesis provides a set of numerical models for TEG simulation, and an optimization technique based on Taguchi method for preliminary design of TEG for waste heat recovery from exhaust gas. As low computation cost is required and the ability of considering different boundary conditions, the proposed numerical model and optimization method can be generalized to a broad range of waste heat recovery applications.