Thermoelectric Generators

Jeffrey M. Weisse
October 24, 2010

Submitted as coursework for Physics 240, Stanford University, Fall 2010

Fig. 1: Configuration of a Thermoelectric Module.

The increasing price of oil, concerts about climate, and the depletion of our natural resources have drawn much attention to renewable energy technology. In 2007 the world consumed roughly 500 quadrillion BTUs of energy and is expected to increase at 1.4% per year. [1] About 90% of this energy was generated through fossil fuel combustion with a typical efficiency of 30-40%. The remaining 60-70% of the energy was lost to the environment via automotive exhaust, industrial processes, and more. [2] It is highly desired to use the wasted heat to improve the overall efficiency of energy conversion. The energy of the wasted heat can be collected and transformed to electricity through a thermoelectric device.

What is a Thermoelectric?

Thermoelectrics are solid state heat engines with materials properties that enable them to convert waste heat into electricity. Their fundamental principle is based on charge carriers: electrons in N-type materials and holes in P-type materials have the ability to move freely through metals and semiconductors. In the presence of a temperature gradient, charge carriers diffuse from hot to cold until an equilibrium is reached between the diffusion potential and the electrostatic repulsion potential, resulting in a buildup of charge carriers known as the Seebeck effect. [3] Typical thermoelectric devices consist of alternating P and N-type semiconductors connected electrically in series and thermally in parallel (Fig. 1). This allows the holes and electrons to flow in opposite directions forming an electric current for power generation. [4]

The efficiency of a thermoelectric device is closely related to the semiconductor's material properties. Ideally, a good thermoelectric device should behave as a "phonon glass" minimizing the thermal conductivity to maintain a high temperature gradient, but also as an "electric crystal" maximizing the electrical conductivity and Seebeck coefficient [3]. The thermoelectric effectiveness is often described by its figure of merit, ZT, which is a dimensionless unit depending on the Seebeck coefficient (α), absolute temperature (T), electrical conductivity (σ), and thermal conductivity (κ): [3]

Fig. 2: Thermoelectric efficiency as a function of ZT and source temperature compared to competitive heat engines. [7]

In order for a thermoelectric device to be competitive with current power generation methods it must possess at ZT great than 3. However, over past five decades the room temperature ZT of materials with our best available technology has only slightly increased from 0.6 to about 1.0. [5] Materials exist with "phonon glass" or "electric crystal" properties, however obtaining both of these simultaneously is a challenge. The issues arise that the Seebeck coefficient, and thermal and electric conductivity properties are all effected by flow of electrons which conducts both heat and charge. Simply increasing the electrical conductivity simultaneously decreases the Seebeck coefficient and increases the thermal conductivity, limiting the potential improvement in ZT for bulk materials. The best performing materials to optimize ZT tend to be heavily doped semiconductors. [4]

State-of-the-art thermoelectric research is investigating new materials to independently tailor these properties. Theoretical ZT improvements exist in decoupling the Seebeck coefficient from the electrical conductivity using engineered heterostructures and independently reducing the thermal conductivity in high atomic weight and nanostructures materials. [3,5] Reports show there has been as much as a 100-fold decrease in thermal conductivity in silicon nanowires, thereby opening the door to significant improvements in ZT. [2,6] Despite all these efforts, there have been only three reports of producing a ZT greater than 2 including the best ZT roughly 3.5 at 575 °K. Transition from these laboratory results to actually devices does not seem to be likely anytime in the near future. [7]

Thermoelectric Market

Unfortunately thermoelectric efficiency (electrical power produced over waste heat in) is currently only about 1/6 of the maximum Carnot efficiency. As shown in Fig. 2, a comparison of thermoelectric efficiencies as a function of ZT and operating temperature are compared to several common heat engines. For thermoelectrics to contend with large scale power production technology (>1000kW), such as a solar thermal heat engine would require a ZT between 8 and 20 which is not foreseeable to be competitive anytime soon. [7] In addition, besides the low efficiency, the cost per watt of power generated of current thermoelectrics has been relatively too high to even assist in large scale energy production. [4]

Fig. 3: Effect of efficiency on power generation size. [7]

Even though it seems unlikely that thermoelectric devices will have a role in large scale energy production, they do have some benefits over current technologies. Their solid state technology offers several large advantages compared to other technologies. They produce electrical energy with no moving parts, which makes them silent and highly reliable while also decreasing operation, maintenance, and potentially capital costs. This allows them to be placed in harsh or remote environments where there reliability justifies their lower efficiency and higher costs [4]. However, the biggest advantage that favors thermoelectric devices is their scalability. Typical coal or other mechanical engines significantly drop in efficiency as they are reduced in size or power level, as schematically demonstrated in Fig. 3. Yet, thermoelectrics maintain their efficiency regardless of power level even on the milliwatt level. This leads to a cross over point where thermoelectric devices are actually more efficient. An increase in ZT will only increase the range of applications where thermoelectrics are more efficient. [7]

Recovering vehicle waste heat appears to have the most potential to implement thermoelectric generators, improving fuel economy and conserving natural resources. [4] The internal combustion engine of typical vehicles is rather inefficient, utilizing about 25% of the energy produced during the fuel combustion process. [8] The rest of the energy is lost as waste heat through friction, coolant, and exhaust gas. About 40% can be harvested as it leaves the exhaust manifold at an average temperate over 600 °K. A thermoelectric device with a ZT of 1.25 will have an efficiency of around 10% and could be used to generate useable power directly towards the vehicle's operations. [4] This could increase fuel efficiency as much as up to 16%.

In 2007, passenger transportation (5.75 x 1019 joules, or 54.5 quadrillion BTUs) accounted for roughly 30% of the world's total liquid fuel consumption (1.92 x 1020 joules, or 182.3 quadrillion BTUs) [1]. It is expected that by 2035 fuel consumed for passenger transportation will further increase to roughly 70.6 quadrillion BTUs [1]. If every vehicle implemented 10% efficient thermoelectric generators, it would reduce passenger transportation fuel consumption as much as 16%. In the future if thermoelectric were able to achieve a ZT greater than 4 they would have an efficiency around 30%, comparable to solar power. If these were installed on every vehicle they would result in as much as a 50% reduction in passenger transport fuel consumption.


Thermoelectric generators are an intriguing way to generate renewable energy directly from waste heat. However, their efficiencies are limited due to their thermal and electrical properties being dependent on each other. Nevertheless, their solid state scalable technology makes them appealing and even more efficient in selective applications. Implementing thermoelectric generators on vehicle exhaust manifolds would help reduce fuel consumption, which in turn would help preserve the world natural resources and reduce carbon emissions.

© Jeffrey M. Weisse. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.


[1] International Energy Outlook 2010, (U.S. Energy Information Administration, 2010).

[2] A. I. Hochbaum, "Enhanced Thermoelectric Performance of Rough Silicon Nanowires," Nature 451, 163 (2007).

[3] G. J. Snyder, "Complex Thermoelectric Materials," Nature Materials 7, 105 (2008).

[4] L. E. Bell, "Cooling, Heating, Generating Power, and Recovering Waste Heat with Thermoelectric Systems," Science 321, 1457 (2008).

[5] A. Majumdar, "Thermoelectricity in semiconductor nanostructures," Science 303, 777 (2004).

[6] A. I. Boukai, "Silicon Nanowires as Efficient Thermoelectric Materials," Nature 451, 168 (2008).

[7] C. B. Vining, "An Inconvenient Truth About Thermoelectrics," Nature Materials 8, 83 (2009).

[8] C. Yu, "Thermoelectric Automotive Waste Heat Energy Recovery Using Maximum Power Point Tracking," Energy Conversion and Management 50, 1506 (2009).