John Christiansen and Dr. Matthew Jones, Mechanical Engineering Department
Introduction
In recent years, the concepts of energy, energy production, effects of energy consumption on the environment, and sustainable development have become increasingly significant as members of society become more aware of their level of energy consumption. However, in energy production for home, commercial, or industrial use, up to 60 percent of the chemical energy potential (in the form of fossil or other fuels) is lost in the form of waste heat. In other words, when fossil or other fuels are burned, only 30 to 40 percent of the potential chemical energy is actually harnessed and converted into useful electricity! This occurs because of limitations on the efficiency of energy conversion systems. This truth in the energy production industry has led to an increasing focus on waste heat recovery technologies, or technologies that harness an additional portion of the heat generated from a fuel and convert it into electricity. One of these technologies is known as thermoelectric generation. A thermoelectric generator (a “TEG”) is a device, typically small (1-2 inches square, and less than ¼-inch thick) that converts heat directly into electricity.
Right now, TEG technology is a fringe technology not used frequently in industrial applications. The purpose of this project was to demonstrate the environmental and economic benefits of a thermoelectric waste heat recovery system, and assess the profitability of this type of system.
Results
We followed an established system of economic analysis used in academia and industry to predict the future profitability of a new technology [1]. The analysis is based on the assumption that, as is true for any emerging technology, the cost of purchasing TEG’s will drop with time with increased demand and declining manufacturing cost.
One model that is commonly used to predict declining cost of technology with time is known as the “Learning Curve Model” [2]. The Learning Curve Model is based on trends of increased efficiency observed in factory workers as they progressed on the ‘learning curve.’ The model states that the unit cost of a new technology, Ct, will decrease with time according to the Equation (1), where C0 is the initial cost, qt is the cumulative output (in this case power output), q0 is the initial output, and b is a ‘learning coefficient’ specific to the industry in question. The application of this Learning Curve Model in industry has been justified by other academic work [3], [4].
Other assumptions, tying progress in TEG technology to progress made in a similar technology (photovoltaic technology [5]), leads to a model predicting the increase in efficiency and power production capacity of a TEG system in future years. It is assumed that the cost of electricity will rise in future years (See Figure 1). Using data gathered from various manufactured fin arrays, correlations between energy costs versus energy produced by the TEG system and results used for future predictions. In order to implement the learning curve model (1), future predictions are made for power out. To develop power out of the TEG system, efficiency is again compared to the trends of PV industry. Nemet (2005) noted from recorded data that the efficiency of PV had increased from 8% to 13.5% from 1980 to 2001. The highest efficiency obtainable for a low delta T, used in implemented TEG systems, is the Carnot efficiency, shown by the equation below where Tc is the cold source temperature and Th is the hot source temperature. The results are plotted above in Fig. 2 where a 10-year pay off is used. It can be seen that the expected year for a poor fin design TEG system to have a 10-year payoff is 2053. The expected year for a good fin design TEG system to have a 10-year payoff is 2029. It is shown that good fin design in the TEG system can improve the payoff period by as much as 24 years, and produce almost 280% more power than that of a poor fin design.
References
- Bejan, A., Tsatsaronis, G., and Moran, M., Thermal Design and Optimization, (Wiley, New Jersey, 1995).
- Wright, T.P., 1936. Factors affecting the costs of airplanes. Journal of the Aeronautical Sciences 3.
- Nemet, G.F., 2005. “Beyond the Learning Curve: Factors Influencing Cost Reductions in Photovoltaics.” Energy Policy 34.
- Yelle, L.E., 1979. “The Learning Curve: Historical Review and Comprehensive Survey.” Decision Science 10.
- Gruber, H., 1996. Trade policy and learning by doing: the case of semiconductors. Research Policy 25.