Thermoelectric devices can convert thermal energy directly into electrical energy, or they can work in reverse and use electrical energy to create a temperature gradient for cooling or heating applications. The absence of moving parts, wide range of operating temperatures, scalability, and modular capabilities makes thermoelectrics attractive for energy generation applications. They have been considered for use with vehicle exhausts, co-generation, and other energy recovery from lost heat in thermodynamic cycles. Thermoelectric devices have relatively low efficiencies but there have been recent advances in thermoelectric materials potentially opening the door to more power applications. As material advancements continue and a wider range of power generation applications will be considered, module and system level modeling becomes critical for the design of the next generation of thermoelectric systems. The overall focus of this thesis is to further develop the current thermoelectric power unit models available, validate the developed model, and implement the model into a simulation environment for feasibility and optimization studies. Current models found in the literature are often based on very specific applications or are too general in nature to truly explore the optimization of a wide range of potential thermoelectric applications. The model developed in this work is highly customizable permitting the optimization of a large number of varying systems. Module mismatch and heat spreading in three dimensions are explored, modeled, and considered for incorporation into the new thermoelectric power unit model to allow for more accurate performance prediction. Mismatch models are developed for thermoelectric modules electrically connected in series and in parallel. Predecessors in the research community believe power output to be hampered due to variation of module specific parameters. Developed models for performance prediction display the extent of the mismatch effect. Experimental validation of the model shows the high level of exactness in prediction. Often, a one dimensional heat transfer assumption is made when analyzing system performance. This is known to cause discrepancies in the thermoelectric power unit model predictions. For thermal resistances with regards to the thermoelectric modules internal to the thermoelectric power unit, a heat spreading effect is observed. This effect requires an analytical method for quantifying the extent of the three dimensional resistance. The analytical model is described and applied to the thermoelectric power unit. Numerical comparison and experimentation are performed to verify the accuracy of the model. The model for heat spreading in three dimensions is incorporated into the new thermoelectric power unit model to allow for more accurate performance prediction. The advanced thermoelectric power unit model is coded in the ThermoElectric Power System Simulator (TEPSS) environment for a project sponsored by New York State Energy Research and Development Authority (NYSERDA). The programmed module will serve as the key component of the software package that will predict the performance of the thermoelectric heat recovery unit used in common thermodynamic cycles. Experimental validation of the advanced model is performed. Data is compared to simulation results from the TEPSS tepowerunit component. The data fits are agreeable when the uncertainty of various system parameters is considered.

Library of Congress Subject Headings

Thermoelectric generators--Computer simulation; Heat recovery

Publication Date


Document Type


Department, Program, or Center

Mechanical Engineering (KGCOE)


Stevens, Robert


Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: TK2950 .F74 2011


RIT – Main Campus