Thermal to Electric Energy Conversion for Cyclic Heat Loads

Benjamin E Whitehead, Purdue University

Abstract

Today, we find cyclic heat loads almost everywhere. When we drive our cars, the engines heat up while we are driving and cool while parked. Processors heat while the computer is in use at the office and cool when idle at night. The sun heats the earth during the day and the earth radiates that heat into space at night. With modern technology, we have access to a number of methods to take that heat and convert it into electricity, but, before selecting one, we need to identify the parameters that inform decision making. The majority of the parameters for most systems include duty cycle, total cost, weight, size, thermal efficiency, and electrical efficiency. However, the importance of each of these will depend on the application. Size and weight take priority in a handheld device, while efficiency dominates in a power plant, and duty cycle is likely to dominate in highly demanding heat pump applications. Over the past decade, developments in semiconductor technology has led to the creation of the thermoelectric generator. With no moving parts and a nearly endlessly scalable nature, these generators present interesting opportunities for taking advantage of any source of waste heat. However, these generators are typically only capable of 5-8% efficiency from conversion of thermal to electric energy. [1] Similarly, advancements in photovoltaic cells has led to the development of thermophotovoltaics. By heating an emitter to a temperature so it radiates light, a thermophotovoltaic cell then converts that light into electricity. By selecting materials that emit light in the optimal ranges of the appropriate photovoltaic cells, thermophotovoltaic systems can potentially exceed the current maximum of 10% efficiency. [2] By pressurizing certain metal powders with hydrogen, hydrogen can be bound to the metal, creating a metal hydride, from which hydrogen can be later re-extracted under the correct pressure and temperature conditions. Since this hydriding reaction is exothermic, and dehydriding is endothermic, we can use the reaction to control temperature and store or release energy as desired. Connecting the liberated hydrogen gas to a hydrogen/air or hydrogen/oxygen fuel cell can then generate useful electrical power. A fuel cell operates by flowing hydrogen and oxygen over a membrane that only allows protons through. This process creates a voltage through the separation of the negatively charged electrons and positively charged water. Typical fuel cells operate at 30-40% efficiency with research aiming to increase that number to 65% with solid oxide fuel cells. [3] In this thesis, I develop several models to size metal hydride systems, identify the critical design parameters of a metal hydride system, and predict hydrogen production for a given heat source. The first model consists of a lumped parameter treatment that analyzes how the effects of varying metal hydrides and heat source values change the dehydriding process. The second model uses COMSOL® Multiphysics to create a higher fidelity simulation of the heat transfer within a metal hydride bed by calculating the spatial heat transfer as well as the porous nature of the system. The Comsol model shows that thermal conductivity is the highest sensitivity parameter of those studied, and therefore should be the primary focus for system design. The model also shows that the efficiency of the system is relatively independent of the duty cycle of the heat source.

Degree

M.S.A.A.

Advisors

Pourpoint, Purdue University.

Subject Area

Aerospace engineering|Energy

COinS