Thermoelectrics: power from waste heat · Thermoelectric modules can not only convert heat directly...

A compact guide to energy

A service from FIZ Karlsruhe GmbH

Themeninfo I/2016

Thermoelectrics: power from waste heat Thermoelectric generators supply autonomous systems and save energy

http://www.bine.info/en/topics/

2

Zur SacheWeit hinten, hinter den Wortbergen, fern der Länder Vokalien und Konsonantien leben die Blindtexte. Abgeschieden wohnen sie in Buchstabhausen an der Küste des Seman-tik, eines großen Sprachozeans. Ein kleines Bächlein namens Duden fließt durch ihren Ort und versorgt sie mit den nötigen Regelialien. Es ist ein paradiesmatisches Land, in dem einem gebratene Satzteile in den Mund fliegen. Nicht einmal von der allmäch-tigen Interpunktion werden die Blindtexte beherrscht – ein geradezu unorthogra-phisches Leben. Eines Tages aber beschloß eine kleine Zeile Blindtext, ihr Name war Lorem Ipsum, hinaus zu gehen in die weite Grammatik.

Der große Oxmox riet ihr davon ab, da es dort wimmele von bösen Kommata, wilden Fragezeichen und hinterhältigen Semikoli, doch das Blindtextchen ließ sich nicht beirren. Es packte seine sieben Versalien, schob sich sein Initial in den Gürtel und machte sich auf den Weg. Als es die ersten Hügel des Kursivgebirges erklommen hatte, warf es einen letzten Blick zurück auf die Skyline seiner Heimatstadt Buchstabhausen, die Headline von Alphabetdorf und die Subline seiner eigenen Straße, der Zeilengas-se. Wehmütig lief ihm eine rhetorische Frage über die Wange, dann setzte es seinen Weg fort. Unterwegs traf es eine Copy. Die Copy warnte das Blindtextchen, da, wo sie herkäme wäre sie zigmal umgeschrieben worden und alles, was von ihrem Ursprung noch übrig wäre, sei das Wort „und“ und das Blindtextchen solle umkehren und wieder in sein eigenes, sicheres Land zurückkehren.

Doch alles Gutzureden konnte es nicht überzeugen und so dauerte es nicht lange, bis ihm ein paar heimtückische Werbetexter auflauerten, es mit Longe und Parole betrun-ken machten und es dann in ihre Agentur schleppten, wo sie es für ihre Projekte wieder und wieder mißbrauchten. Und wenn es nicht umgeschrieben wurde, dann benutzen Sie es immernoch. Weit hinten, hinter den Wortbergen, fern der Länder Vokalien und Konsonantien leben die Blindtexte. Abgeschieden wohnen sie in Buchstabhausen an der Küste des Semantik, eines großen Sprachozeans. Ein kleines Bächlein namens Duden fließt durch ihren Ort und versorgt sie mit den nötigen Regelialien.

Ihre BINE-Redaktion wünscht Ihnen eine anregende Lektüre

„“

Inhalt 3 Anforderungen von allen Seiten

4 Luftqualität im Schulalltag

7 Aus der Praxis: Motorisch unterstützte Fensterlüftung

8 Systeme und Anlagentechnik

13 Aus der Praxis: Schule Olbersdorf

13 En passant: Schachtlüftung

14 Gute Akustik in Klassenräumen

16 Erfahrungen und Kommunikation

19 Standpunkte: Lüftungsanlagen in Schulen – Luxus oder Notwendigkeit

19 Aus der Praxis: Realschule Lindau

20 Lüftung und Energie

Authors Dr Jan D. König, Fraunhofer-Institut für

Physikalische Messtechnik IPM, Coordination

In cooperation with

Dr Kilian Bartholomé, Fraunhofer IPM

Dr Harald Böttner, Fraunhofer IPM

Daniel Jänsch,

Ingenieurgesellschaft Auto und Verkehr IAV

Mirko Klein Altstedde, DLR

Martin Köhne, Bosch

Dr Joachim Nurnus, Micropelt

Dr Aljoscha Roch, Fraunhofer IWS

Dr Karina Tarantik, Fraunhofer IPM

Editor Gerhard Hirn

CopyrightText and illustrations from this publication can only be used if permission has been granted by the BINE editorial team. We would be delighted to hear from you.

Cover image: NASA / JPL-Caltech /Malin Space Science SystemsLead photos:P. 3: DLR (CC-BY 3.0)P. 6, 8: Fraunhofer IPMP. 10: Bastian Ehl, MPIP. 14: MicropeltP. 18: DLR - FK AutoP. 22: VDEh Betriebsforschungsinstitut BFI

BINE-Themeninfo I/2016

Contents 3 Transforming waste heat directly into electricity

6 How does thermoelectrics work?

8 Thermoelectrics make sensors energy-self sufficient

10 Researching the optimal materials

14 Manufacturing TE materials and TEGs

15 Points of view: How thermoelectrics can emerge from the niche market

17 En passant: With thermoelectrics through the galaxy

18 Thermoelectrics in cars help to save fuel

22 Utilising waste heat saves energy and CO2

In practice: Thin-film TEGs, TEG for vehicles, ThermoHeusler project, waste heat utilisation

Straight to the pointWaste heat is produced in all areas of daily life – in industry, homes and transport. In Germany alone, this adds up to a waste heat potential of 300 TWh per year. This energy volume corresponds to nearly half the total energy consumed by German industry. Thermoelectric generators (TEGs) can tap into this huge energy reservoir and transform the „waste energy“ without moving parts into a higher form of energy. The TEG utilise waste heat and already produce electricity from small temperature differen-ces. This recycling of waste or ambient heat that would otherwise be lost is also known as energy harvesting, which in future will make an important contribution to achieving energy efficiency and reducing CO2 emissions.

Thermoelectric modules can not only convert heat directly into electricity but also, in the form of an electric heat pump, use electricity to provide direct cooling and heating: these days, industry uses Peltier coolers for controlling the temperature of car seats or for stabilising the temperature of electronic components. In the consumer goods sector, these coolers can also be found, for example, in camping cool boxes and silent hotel fridges. In their simplest form, thermoelectric modules consist of a thermocouple that is used in many applications as a temperature sensor. This is constructed from two thermoelectric materials whose electrical junctions are at different temperature levels. The first thermoelectric materials could only be produced at great expense and generated only a few watts. These days systems are now available that generate up to 1000 watts. New materials and processing methods enable greater temperature differences to be utilised, whereby the power output will also continue to rise.

Researchers and manufacturers are working to increase the efficiency of thermoelectric generators and develop appropriate manufacturing processes for mass production. Similar to photovoltaics, the way has now been paved for using a technology originally developed to supply vehicles in outer space, such as the Mars rover, Curiosity, for a wide range of uses in various applications. This Themeninfo brochure provides an overview of how the technology functions and its application areas, as well as the materials that are being developed and optimised for different temperature ranges.

Your BINE editorial team wishes you an enjoyable read

„“

Kaiserstraße 185-197, 53113 Bonn, Germany Phone +49 228 92379-0 Fax +49 228 92379-29 [email protected] www.bine.info

3BINE-Themeninfo I/2016

Thermoelectrics can be used for both

recovering energy from waste heat as well as

for cooling and temperature regulation.

The direct energy conversion has no moving

parts and works in both directions:

thermoelectrics can convert thermal energy into

electrical energy or use electric energy for cooling.

Transforming waste heat into electricity

Thermocouples are mainly used as temperature sensors, as so-called Peltier elements for temperature control, and as thermoelectric generators for recovering waste heat.

With the aim of enabling a wider use of thermoelectrics, researchers are working to improve the system efficiency of the energy conversion from thermal to electrical energy, which currently amounts to between 2 and 7 % depend-ing on the temperature range. This will be achieved on the one hand by improving the quality of the materials com-mercially used but also by minimising the losses in the module and the system.

Many technical processes use only 30 to 40 % of the energy deployed. The rest is usually lost as waste heat. It is therefore economically and environmentally worthwhile using this heat with high waste heat temperatures. Until now, mechanical equipment such as turbines or Stirling engines have mostly been used for this purpose. The dis-advantages of these technologies are the moving compo-nents and the associated limited service life as well as the resulting maintenance costs.

Thermoelectric converters, on the other hand, operate vibration-free and noiselessly with no moving parts. They convert heat directly into electrical energy and can be integrated into existing heat exchangers. Despite the many advantages, energy recovery with thermoelectric generators is still largely unknown. This is because such generators have previously been mainly used in niche ap-plications owing to their relatively low efficiency.

Durable and maintenance-free

Thermoelectric generators are durable and work without requiring maintenance. For more than five decades, space technology has used them to power spacecraft. For exam-ple, the Voyager probes have performed flawlessly since their launch in 1977. Such long periods of uninterrupted and maintenance-free operation are generally extremely rare for technical applications and thus a striking feature of thermoelectrics. Another advantage is that small temper-

ature differences can be used for recycling residual heat that are not suitable for other means of converting heat into usable forms of energy. Because they are very small and compact, thermoelectric modules can be easily inte-grated into existing systems. Until now, however, high pro-duction costs and their low overall efficiency have hindered the more widespread use of the technology.

From a niche to a mass market product

Thermoelectrics is still a niche technology and although estimates and market analyses indicate considerable growth potential, there are considerable disparities in terms of the figures. Whereas a market analysis in 2003 indicates a global market volume of 160 million euros with a strong upward trend, an estimate from 2008 suggests 500 million euros. A more conservative estimate predicts that approximately 100 million US dollars will be reached in 2016 (Fig. 4).

Experts believe that the market volume could increase tenfold if the thermoelectric figure of merit for commercial modules can be doubled.

Researchers and developers are working to make thermo-electric materials and generators more efficient and cheaper. In the near future they expect significantly higher efficiencies, whereby considerable increases have already been achieved in the laboratory. Important approaches include improvements in the thermoelectric material properties, the assembly and connection technology, as well as the thermal and electrical integration into the overall system.

Future markets in the thermoelectrics sector will include the decentralised energy supply of sensor systems, decen-tralised energy recovery of waste heat in the automotive sector and industry, energy recovery in energy-intensive technological processes as well as temperature control and air conditioning, for example in the e-mobility sector. Thermoelectricity is now regarded as a potential key tech-nology for recovering energy from waste heat.

4 BINE-Themeninfo I/2016

Fig. 2 Peltier elements can keep drinks fresh in thermo cup holders. Source: Dometic Group

Fig. 3 Thermoelectric sensor. Source: Fraunhofer IPM, Kai-Uwe Wudtke

Fig. 1 Peltier coolers in PCs enable optimal processor performance. Source: pcgameshardware

Precision cooling with Peltier elements

In addition to measuring temperature using thermocouples, thermoelectricity is traditionally mainly used for cooling (approximately 70 – 80 % market share). Peltier elements

work in mobile cooling boxes and silent hotel fridges, they heat seats in cars and cool beverage containers. In electric vehicles, controlling the temperature of the battery can ensure optimal operation and charging.

Peltier coolers enable very effective cooling in the smallest spaces in laser, computer and medical technology. They respond very quickly and achieve a very high power density. This enables accurate thermal management of precision devices such as semiconductor lasers and processes, for example for use in biomedicine: there small cooling sur-faces under 1 mm² are used for extremely rapid tempera-ture cycling of small samples. For instance, temperature ramps with values of 15 K per second can be applied with polymerase chain reaction investigations. This leads to by- product-freer reaction products and thus a more accurate evaluation.

Maintenance-free supply of sensors

Thermoelectrics plays a particularly important role in supplying energy-autonomous sensors and transmission systems with capacities in the milliwatt range. Because their power consumption is continually reducing, thermo-electric generators with their low weight and fast response times are becoming increasingly better at supplying power. This eliminates the need for cables or batteries to supply power to the sensors or send data – the energy required can be obtained directly from the ambient heat. Temperature differences of just a few degrees Celsius are sufficient for this purpose.

In particular, such self-powered wireless sensors enable the reliable monitoring of inaccessible places in vehicles and aircraft as well as in process and plant engineering.

More powerful systems for cars and industry

With vehicles with internal combustion engines, thermo-electrics can generate electricity from waste heat for the on-board network and thus relieve the alternator. This re-

200

180

160

140

120

100

80

60

40

20

Mill

ion

US

dolla

rs

2014 2015 2016 20170

Miscellaneous capital goods industries Other consumersHealth sector Other industriesMilitary, aerospace Wireless sensor networks

Fig. 4 The estimated development of the global market for thermoelectric energy harvesting (in millions of US dollars). Source: IDTechEx


Fig. 5 Researchers are working on a TEG for use in cars. Source: DLR Institute of Vehicle Concepts (FK)

duces the fuel consumption and thus the CO2 emissions: experts believe that TEGs will enable fuel savings of up to 5 %. Automotive manufacturers and suppliers have developed and tested the technology, and are now work-ing on concepts for its cost-effective mass production. For a broader use of thermoelectrics, the automotive sector could therefore provide a door opening function. Possible applications exist in vehicles with combustion engines as well as hybrid and electric drives.

When using industrial waste heat, TEGs have the advantage over other competing technologies that they can be easily integrated into existing systems. In cogeneration (CHP) plants, thermoelectrics can help to increase the electrical power output.

Safety applications with thermoelectrics

Thermoelectric fuses are used in household appliances such as washing machines and gas cookers. A washing machine can only be opened when the water in the wash-ing drum is cooled and pumped out. An electromagnet powered with thermal electricity blocks the opening mechanism, even if all other power supplies have failed. This operational monitoring functions without electrical or mechanical „auxiliary energy“.

With gas cookers, heated thermocouples generate electric-ity and a magnetic field: as long as the flame burns on the gas cooker, a current is generated in the two-wire circuit which keeps the solenoid valve for the gas discharge open. If the flame is extinguished, no more electricity flows in the thermocouple and the gas valve closes. With underground oil and gas pipelines, thermoelectric ele-ments can replace sacrificial anodes as electrochemical corrosion protection.

What is thermoelectrics?

Thermoelectrics refers to the direct conversion of an electrical current flow into a heat flow as well as the heat flow into a current flow – thermoelectrics works both ways. The basic principle was already recognised by Thomas Johann Seebeck in 1821 when he observed that a compass needle near two different, interconnect-ed metal wires is deflected when the temperatures at the junc-tions differ, whereby the degree of deflection is proportional to the temperature difference. This is due to an electrical field that is created by the temperature gradient on the conductors.In 1834, the French scientist Jean Peltier discovered that this effect can be reversed and used as a heat pump: if a current is connected to the interconnected conductors, a temperature gradient is created at the junctions. Thermal energy is transported from one junction to the other. The so-called Peltier effect can be used for heating or cooling. The maximum possible yield when heat is thermoelectrically converted into electrical energy is physically determined by the efficiency of the Carnot cycle process.

Electricalconnection

Referencejunction

Mea

sura

ble

volt

age:

Diff

eren

ce in

the

ther

moe

lect

ric

volt

ages

bet

wee

n m

etal

A a

nd B

Met

al A

and

B a

reel

ectr

ical

ly c

onne

cted

Referencetemperature

Temperaturedifference

Metal A

Metal B

Measurementjunction

Measurementtemperature

V

Fig. 7 How a thermocouple is used for measuring temperature: If p-type material is combined with n-type material, this generates a thermoelectric voltage. This is proportional to the temperature difference between the measurement junction and the reference junction with the known temperature. This voltage is measured and converted into the corresponding temperature. Source: Wikimedia Commons, Arne Hückelheim

Fig. 6 The pioneers of thermoelectrics: Thomas Johann Seebeck (1770 – 1831), left, and Jean Charles Athanase Peltier (1785 – 1845), right. Source: public domain via Wikimedia


The fundamental principle for generating thermoelectric power is the Seebeck effect. A thermoelectric module is constructed from numerous thermocouples electrically connected in series that consist of n- and p-type thermo-electric materials (Fig. 8). Thermally, these thermocouples are connected in parallel so that heat can flow through the thermoelectric materials. Electrical insulation on the hot and cold sides stabilises the thermocouples arranged in this manner. The electrical junctions complete the module.

Seebeck effect: Heat becomes electricity

In an open two-wire circuit consisting of two different con-ductive materials, a temperature gradient between the junctions of the two materials induces an electrical voltage at the free conductor ends.

The thermoelectric voltage depends on the material: for any given temperature difference, two different conductors have different thermoelectric voltages. Here αAB is called the thermopower or the Seebeck coefficient of the con-ductor combination AB. It is usually given in µV/K. The coefficient is positive when the thermal current flows at the warmest of the two transitions flows from material A to material B. According to this convention, n-type conduc-tors (electron conductors) receive a negative Seebeck co-efficient and p-type conductors (hole conductors) receive a positive Seebeck coefficient.

Peltier effect: Direct temperature control with electricity

The Peltier effect enables direct electrical cooling or tem-perature regulation. It works as follows: if an electric current flows through a thermoelectric module, heat is transported from one side to the other. The temperature therefore lowers at one side and increases on the other side. If the direction of the current flow is reversed, the heated and cooled sides of the thermoelectric module are swapped over. This enables the temperature of solids,

liquids and gases to be regulated very efficiently and accurately, whereby the ratio of expended electric power to the achieved cooling power describes the efficiency of the cooler.

Modern thermoelectric converters consist of a large number of thermocouples with positive- and negative-type thermo-electric semiconductor materials that are connected electri-cally in series and thermally in parallel.

Qualitatively, the Peltier effect can be defined by the Peltier coefficient PAB as the ratio between the heat volume Q per unit time generated and discharged at the transition and the electric current I. Its interrelationship with the Seebeck coefficient is known as the Kelvin Relations.

The materials

Thermoelectric materials are made of special alloys or semiconductor compounds. Recently, electrically conductive polymers have also been investigated in terms of their thermoelectrical properties.

En route to the best possible thermoelectric generator, developers need to consider many factors simultaneously. The desired materials require high efficiencies achieved by low thermal conductivity simultaneously combined with high electrical conductivity and a high Seebeck coef-ficient. The materials should be stable in regards to high temperatures and thermomechanical influences, and also meet requirements such as availability and environ-mental compatibility. With rapid temperature changes and severe vibrations, the different application areas make considerable demands on the assembly and connection technology as well as the system integration in terms of achieving long-term stability. At the same time, the researchers are investigating how thermoelectric gene-rators can be largely automated and cheaply manu-factured.

Closely associated with the thermoelectrics

phenomenon are the names of the pioneering scientists

Seebeck and Peltier. The conversion of heat into

electricity is known as the Seebeck effect, whilst the

direct conversion of electricity in temperature gradients

is known as the Peltier effect. How this direct energy

conversion works is explained here.

How does thermoelectrics work?


Thermoelectric efficiency as a measure of effectiveness: ZT

The figure of merit ZT enables the usability of thermoelectric materials to be assessed. The greater the figure of merit and the merit Z of a material, the more effectively it can be used in thermoelectric generators or Peltier coolers. Therefore all approaches for improving thermoelectric materials are aimed at maximising ZT. The materials currently used for cooling and generator purposes achieve average figures of merit between 0.5 and 0.8. The application temperatures and the thermoelectric figure of merit Z (or the dimensionless figure of merit ZT) deter-mine the maximum efficiency of thermoelectric generators ηmax and Peltier coolers φmax .

The thermoelectric figure of merit Z or the dimensionless figure of merit ZT is described physically by

whereby σ denotes the electrical conductivity, λ the thermal conductivity and α the Seebeck coefficient.

0 200 400 600 800 1.0000.0

0.1

0.2

0.3

0.4

0.5

0.6

Tcold= 300 K

Carnot efficiency ZT = 2.5 ZT = 2ZT = 1.5 ZT = 1 ZT = 0.5

Effic

ienc

y η

Δ T [K]

Fig.11 Conversion efficiency of a thermoelectric generator as a function of the temperature difference between the hot and cold sides (cold side temperature fixed at 300 K) and the average thermoelectric figure of merit ZT of the materials. Source: Fraunhofer IPM

n-type TE materialp-type TE material

Heat flowElectrical current flow

Warm

Cold

Thermo-electricalmodule

B

A A

TW TK

UAB

Absorbed heat(cold side)

Waste heat(hot side)

Ceramicsubstrate

Electricaljunctions

p-dopedsemiconductormaterial

n-dopedsemiconductormaterial

Positive (+)Negative (–)

Fig. 8 Design and operation of a thermoelectric generator. The heat flow creates an electric current. Source: Fraunhofer IPM

Fig. 9 Open electric two-wire circuit consisting of the materials A and B with a temperature differenc e TW – TK between the junctions. The thermal voltage UAB is generated as a result of the Seebeck effect. Source: Fraunhofer IPM

Fig. 10 Schematic structure of a Peltier element. If an electric current flows through thermoelectric semiconductor material, one side heats up and the opposite side is cooled. Source: PANCO GmbH

ηmax = φmax =,T1 – T2

T2 – T1 T2 T1

T1

T1 √1 + ZTm – 1

√1 + ZTm + 1√1 + ZTm +

T2 T1

√1 + ZTm –

Z = ZT = Tresp.α2 σ α2 σ

λ λ (T = absolute temperature)


The sensors self-sufficiently supply themselves with power and do not require batteries and maintenance. Specially adapted thermoelectric generators (TEG) generate the re-quired energy from the temperature difference between the environment and the interior space. The difference between body and room temperatures is already sufficient in principle to operate electronic devices. Thermoelectri-cally operated sensors are also conceivable for the wireless monitoring of body functions. This was shown in the late 1990s by thermoelectric watches.

Sensors that self-sufficiently supply themselves with energy, mutually determine the energy consumption, cooperate within one another in local networks and send measurement readings to a central control system are being developed in a joint project entitled “Autonomous flexible monitor-ing units for monitoring technical systems – AMETYST”, which is being funded by the German federal government.

Energy for wireless monitoring of fuselages

The sensors are stuck to the inside of the fuselage of airplanes in the form of an “intelligent patch” that saves on cables and weight. They are energy self-sufficient and can detect wear, fatigue and corrosion during flight. The necessary energy is obtained from the temperature difference between the cold outside air at minus 20 to minus 50 °C and the warm passenger cabin at a tempera-ture of 20 °C.

The thickness of the thermoelectric layers that separate the two temperature levels from one another essentially determines the performance of the device: the aim is to achieve an electrical output of 10 mW with temperature differences of 30 K. To achieve this, the developers are optimising the heat flow. The first prototypes will soon be tested in an aircraft.

Thermoelectrics supply energy-autonomous

sensor systems for monitoring aircraft, high-rise

buildings, power plants, wind turbines and industrial

plants as well as dangerous goods shipments.

These are situated in inaccessible or dangerous

places and monitor components or transmit operational

data to a central unit.

TEGs make sensors energy-self sufficient

!

!

Fig. 12 Energy-autonomous sensors monitor the state of a paper machine. Small, cost-efficient sensor nodes form a wireless network. The sensors can detect and transmit data even in inaccessible places. Source: Fraunhofer IZM


Autonomous sensors for cars, buildings and hearing aids Wireless sensor networks are flexible, secure and inex-pensive to install. That is why they are used, for example, for infrastructure and environmental sensor systems as well as vehicle monitoring.

Athletes could measure their pulses via sensor systems integrated in their T-shirts, and hearing aids could draw their energy from body heat. In the same way as with air-planes, energy-autonomous sensor networks could also help to save weight in cars as heavy cables are eliminated. In order to monitor buildings and bridges, sensors can easily be installed in the masonry where, for example, they can measure the moisture.

The entire system generally consists of a sensor that measures the physical and chemical properties or the material composition, a thermoelectric generator to gen-erate energy, an energy storage system, a charging circuit and a radio module as well as its receiver. This operating principle is schematically depicted in Fig. 13.

Intelligent power management with TEGs

Wireless sensor networks are particularly ideal for appli-cations where batteries are less suitable. Conceivable are production and plant technology, monitoring of structures such as bridges and buildings, means of transport such as ships, airplanes and trains as well as tracking and tracing in logistics. For instance, here they can be used to track goods during their frequently months-long transportation and to monitor their condition.

To provide an application with 10 mW of power for a year with a battery with an energy density of 0.06 Wh/cm3 would require a lithium battery with a volume of 285 cm3, which would have to be replaced each year. A TEG with a total volume of 0.6 cm3 would be sufficient for this, since with a temperature difference of 3 K, TEGs with a volume of 0.01 cm3 deliver a power output of about 160 µW. The use of self-sufficient energy sources is also worthwhile in the low power range when the ambient conditions are right.

Even very small temperature differences of less than 1 K are sufficient for TEGs to supply wireless sensor networks with power. A system developed at Fraunhofer IPM utilis-es the temperature fluctuations associated with the day-night cycle to continuously measure the ambient temper-ature and transmit them to a receiver. The sensor system essentially consists of a heat collector thermally connected to the ambient environment, a thermally decoupled heat storage system and a thermoelectric generator installed between them.

When the heat collector has a different temperature than the heat storage unit, heat flows through the thermoelectric generator and this generates electric power. Via a power management system, it also supplies the network contin-uously with electricity at night.

TEGs make sensors energy-self sufficient

Energy-self sufficient systems and energy harvesting

Energy harvesting refers to the generation of electrical energy from ambient energies such as heat, light, vibrations or wind. In addition to thermoelectric generators, these include piezo-electric and vibration transducers. Piezoelectric transducers use vibrations or pressure fluctuations to produce electricity, while vibration transducers induce a voltage electrodynamically; they use the vibrations and oscillations from manufacturing plants with a system of coils and permanent magnets adapted to the respective resonance frequencies.

Thermoelectric energy harvesting can be used, for example, to supply wireless sensor networks, intelligent building control systems and radiator actuators with maintenance-free power.

Micro-controller

Wirelessdata

transmission

Signalprocessing

Sensor

Solar energy Radio waves

Kinetic energy Thermal energy

Energymanagement

EnergyHarvester

Fig. 13 Schematic representation of various energy harvesting processes for supplying a wireless sensor node. Source: Aalborg University

Fig. 14 Top: The thermal generator of the radiator actuator works from a temperature difference of 4 Kelvin. The warm side is the hot water at the heating valve and the cold side is the radiator actuator housing. Source: Micropelt; Below: An energy-autonomous sensor bracelet measures, for example, the ambient temperature and sends it via Bluetooth to a PC. Source: Fraunhofer IIS / Kurt Fuchs


Which thermoelectric material is best suited depends on technological aspects as well as on the economic, eco-logical and toxicological properties of the material and application system. Researchers need to take into account all these factors for the further development to market maturity.

How TEGs become efficient and cost-effective

In technological terms, a thermoelectric figure of merit ZT that is as high as possible is required. Materials with maximum values of 1.5 are already suitable for use. In addition, the ZT value should only change slightly over a temperature range that is as wide as possible. For mobile applications, for example as thermoelectric generators in

vehicles, the thermoelectric material should have a spe-cific weight that is as low possible. The junction materials for the electrical and thermal connections must be suitable in chemical terms and in relation to the electrical junction resistance. Their physical properties, for example their coefficient of thermal expansion, must also be absolutely compatible with the thermoelectric material.

To ensure that thermoelectric products can be affordable for anyone, the material and production costs must be kept low. In preparation for mass production, it is necessary to estimate the availability and prices of raw materials and technologies for producing materials and systems.

From an environmental point of view, it is important whether and to what extent closed product cycles or at least high material recycling rates can be achieved for everyday thermoelectric products. At each step of the thermoelectric value chain, the toxicological assessment can in itself be sufficient to exclude what would other-wise be perfectly usable materials, technologies and processes.

Material research

To date, thermoelectric research has tended to focus on materials, driven by the desire to achieve ZT values that are as high as possible. In the laboratory, the greatest success has been achieved through the physical-based approach of reducing the thermal conductivity while maintaining application-suitable electrical properties.

Researchers are investigating, for example, alloy scattering, the incorporation of scattering centres in thermoelectric materials from so-called cage structures and different nanostructure technologies. They are creating the nanos-tructural properties with special processing technologies such as ball milling. With compacting processes such as spark plasma sintering or hot pressing, configurable, semi-finished products are created in the form of cylindrical pellets. Currently it is still unclear as to what extent the improvements achieved in the ZT value will be retained

The challenge for developers

is that optimal thermo electric materials

must meet a partly contradictory property

spectrum.

The need is for high electrical and low thermal

conductivity as well as high thermo power

(Seebeck coefficient).

Researching the optimal materials

Fig. 15 Temperature-dependent course of the figure of merit ZT for different thermoelectric materials (status of 2014). Source: Fraunhofer IPM

2000 400 600 800 1,000

2.0

1.5

1.0

0.5

0

Bi-Te (p)Bi-Te (n)

Half-Heuslers (p)

Si-Ge (p)

Fe-Si (n)Zn-O (n)

Pb-Te (p) Skutterudites (p)Pb-Te (n)

Mg2(Sn,Si) (n)

Skutterudites (n)

Half-Heuslers (n)

Si-Ge (n)

Na-Ca-Co-O (p)

ZT

Temperature [°C]

Originalgröße bitte bessere Vorlage


under conditions of use, particularly with Bi2Te3 and oper-ational temperatures above approximately 300 °C.

The materials investigated are as varied as the factors that influence their properties: besides the classic V2-VI3 ma-terials (from the groups V and VI of the periodic table of elements) with the basic compound Bi2Te3, the IV-VI ma-terials with the basic compound PbTe and the IV/V alloys from the SiGe system, there are also many other com-pounds available. Fig. 15 shows the figure of merit ZT as a function of temperature for n-type and p-type materials.

Thermoelectric materials and material families that are currently being investigated include:

• Chalcogenides with derivatives such as Bi2Te3, PbTe, AgPb18SbTe20, LaTe1,45, Cu12Sb4S13

• Half-Heusler alloys based on TiNiSn • Silicides such as Mg2Si0,8Sn0,2, MnSi1,7, FeSi2, Si0,8Ge0,2 • Clathrates such as Ba8Ga16Ge30

• Skutterudites based on CoSb3

• Zintl phases such as Zn4Sb3, Yb14Mn1Sb11

• Bi-Sb alloys• Oxides with perovskite structures such as NaxCoO2 and

[Ca2CoO3]0.62[CoO2]• Polymers

Many of the n-type and p-type materials achieve or sur-pass the “magic” value ZT=1. Laboratory data show that materials are available for all application-relevant tem-perature ranges. Bi2Te3-based materials have been exclu-sively used for the room temperature range for about 50 years.

Commodity prices as a selection criterion

Since there is no reliable data on the costs for producing thermoelectric compounds and for further processing them into thermoelectric components, the raw material price for each established compound in the specified sto-ichiometry is being drawn upon as a further criterion for optimum materials.

Fig. 17 Electrical output of the thermoelectric generator as a function of the thermal resistance KTEG of the generator. The maximum of the generated output is Kopt . Source: Fraunhofer IPM

Fig. 18 In the „TEG line“, scientists at the Institute for Materials Research at the German Aerospace Centre (Deutsches Zentrum für Luft- und Raumfahrt, DLR) are working on the further development of thermoelectric materials and generators. Source: DLR

Desired properties of thermoelectric materials

Most thermoelectric materials achieve a peak value for the figure of merit ZT within a narrow temperature range: this means that they must be correspondingly selected for the intended tempera-ture range. The materials should have the following properties if possible:

• A high electrical conductivity σ in order to keep the electrical losses (Joulean heat) small

• A low thermal conductivity λ in order to achieve a large temperature gradient dT

• A high Seebeck coefficient α in order to provide as large a voltage at the output of the generator as possible

Thermal and electrical energy flows follow similar rules. In order to generate maximum electric or thermal power, both the electrical and thermal power as well as the resistances from the thermoelectric generator and system must be precisely matched with one another.

Fig. 16 Peltier element (manufacturer: Peltron). Source: Fraunhofer IFAM

KTEG

KZ Kopt

Pow

er


Leading the way are Zn4Sb3, the silicides, skutterudites, half-Heusler compounds and oxides. Trailing behind them are the classics Bi2Te3 and PbTe. Considerably more expensive is the “space material” SiGe.

However, at higher temperatures Zn4Sb3 tends to become chemically instable with Sb losses. Despite significant improvements in quality in recent years, the oxides are still a long way from attaining technically usable ZT values and from providing suitable, long-term-stable junction technologies. Beyond the operating temperature for Bi2Te3, the silicides, skutterudites and half-Heusler materials cover currently what is probably the most important temperature range of up to about 600 °C. For future applications, it would be desirable to research new materials with high figures of merit for the temperature range between 100 and 300 °C in order to close the exist-ing gap.

Even with price-wise highly promising materials, supply security and substantial fluctuations in the commodity prices must be taken into consideration. With the previous-ly best half-Heusler material (Ti0.5(Zr0.5Hf0.5)0.5NiSn0.998Sb0.002), in particular the prices for the main component hafnium have soared in recent years. Most of the elements used for producing thermoelectric materials belong to the list of Critical Raw Materials (CRMs) deemed essential for secur-ing supplies in the European Union. The few exceptions relate primarily to the silicides and the half-Heusler base compound TiNiSn.

Highly promising TE materials

With the aim of making thermoelectrics applicable for the mass market, both basic research and application-orient-ed developments are increasingly focussing on the mate-rial families for silicides, skutterudites and half-Heusler compounds. This is also due to the environmental and toxicological classification of these so-called medi-um-temperature materials.

Bi2Te3 is problematic in that it is classified as toxic, whereby this can hardly refer to the poorly soluble compound Bi2Te3 itself. This also applies to the likewise poorly soluble compound PbTe. Despite diverse improvements in quality and decades of technological experience, it is foreseeable that PbTe-based bulk products will not achieve social ac-ceptance. The EU exemption for PbTe-based thermoelec-tric generators in motor vehicles until 2018 will probably not change this.

In terms of the state of technology and the prospects for the large-scale manufacture of thermoelectric materials, the classics are technologically secured. Half-Heusler materials, skutterudites and PbTe can currently be pro-duced at a pilot-scale (several tens of kilograms), and magnesium and manganese silicides in the single-digit kilogram range. There are reports that the production of skutterudites with “reasonable” ZT values has been achieved at the tonne-scale.

Researchers are currently examining promising synthetic methods and materials that have already achieved fig-ures of merit of ZT = 1.5 to 2.5 in the laboratory.

The latest detailed information on the materials is sum-marised in Fig. 19.

The magic threshold for the figure of merit value ZT ~1 has been reached or safely exceeded for all major material families with the exception of manganese silicides and oxides. The technically important temperature ranges up to 600 °C are covered. There is still need for research and development in regards to the long-term stability as well as the mechanical, thermal and chemical stability. Not all materials have sufficient data on their toxicity. The issues concerning Critical Raw Materials apply to almost all ma-terials. The production on a pilot-plant scale has recently made considerable progress. To date it is still open as to which of the medium-temperature materials will be the optimal material. It will probably be found among the silicides, skutterudites or half-Heusler. materials.

Fig. 19 Aspects for determining optimal thermoelectric materials, as of 2014. Source: Fraunhofer IPM

Materials Bi2Te3 PbTe SiGe MnSi1.73 Mg2-SiSn CoSb3 Oxides Half-Heuslers

Commercial and space modules Research modules Prototypes

Figure of merit (ZT) >1.0 >1.0 <1.0 ≥ 1.0 ≥ 1.0 > 1.0 < 1.0 ≥ 1.0

Operational temperature < 300 °C < 500 °C < 900 °C < 550 °C < 550 °C < 520 °C < 700 °C < 550 °C

Long-term stability ■ ■ ■ ■ ■ ■ ■ ■

Mechanical stability ■ ■ ■ ■ ■ ■ ■ ■

Thermal stability ■ ■ ■ ■ ■ ■ ■Chemical stability ■ ■ ■ ■ ■ ■ ■ ■Toxicity ■ ■ ■ ■ ■ ■ ■ ■Environmental aspects ■ ■ ■ ■ ■ ■ ■ ■Raw material availability ■ ■ ■ ■ ■ ■ ■ ■Large-scale manufacture ■ ■ ■ ■ ■ ■ ■ ■ ■ Positive assessment ■ Negative assessment ■ Contradictory data


The path from material to module

The material dictates the maximum thermoelectric effi-ciency. With the semiconductor compounds mainly used, the desired properties are created by targeted doping. Tailored to room temperatures, the bismuth telluride (Bi2Te3) material system has been successfully used for decades.

To ensure that the TE modules can withstand temperature differences of several hundred Kelvin during operation, the developers need to take into account the different expansion coefficients of the thermoelectric material, the metallic junctions and the ceramic insulation during assembly.

Other important factors that need to be considered when developing efficient modules: the materials need to be processed without damage, electrically and thermally connected, and firmly bonded.

The connected materials are linked to the heat source and heat sink so that in thermal terms they are as resist-ance-free as possible. Here it should be ensured that the coupling is achieved with electrical insulation with a ther-mal conductivity that is as high as possible, for example using ceramics such as aluminium nitride.

In order to achieve a high efficiency, the thermal manage-ment of the overall system is crucial. In order to optimally integrate thermoelectric modules into a system, both the heat exchanger and the thermal and electric power need to be adapted.

For a good thermoelectric contact the following require-ments apply:

1. The electrical contact resistance should be small compared with the electrical resistance of the thermocouple leg.

2. The junction should be mechanically stable and resistant to temperature changes. A high cycle stability is particularly required for use in vehicles.

3. The diffusion between the semiconductor and metal electrode should be as small as possible to ensure that the properties of the semiconductor are not changed by material diffusing in. The electrode material should not create any doping effect in the semiconductor.

4. The softening point of the junction layer should not be considerably below the melting point of the thermoelectric material to ensure that the operational temperature range is not reduced.

5. The temperature jump at the junction should be small compared with the drop in temperature at the thermocouple leg.

6. The coefficients of thermal expansion should be sufficiently identical in order to ensure the mechanical

(long-term) stability and to avoid stresses and shearing forces due to different thermal expansion.

7. The thermoelectric material should not be damaged during the contacting process.

8. For industrial use, the assembly and connection technologies adapted to the materials should be easy to industrialise.

200 250 300 350 400 450 500

22

20

18

16

14

12

10

8

6

4

2

0

Bi2Te3 PbSe SiGe

Carn

ot v

alue

[%]

T (K)

Fig. 20 Thermoelectric bars Source: Fraunhofer IPM

Fig. 21 Bi2Te3 is ideally suited for applications in the room temperature range (300 K). Source: Micropelt


The diverse variety of thermoelectric materials

is also reflected in the wide range of manufacturing

processes. These range from the foundry-based

industrial production of alloys such as the

half-Heusler compounds to the print-based or

galvanic production of thermoelectric materials

to ceramic processes.

Manufacturing TE materials and TEGs

p-type polymer (PEDOT:Tos) to achieve ZT values between around 0.2 and 0.3.

However, the biggest challenge in the development of conductive polymers is the synthesis of n-type polymers that are highly stable in air. Compared with purely organic materials, the polymer approach using organometallic complexes has the advantage that, depending on the metal component used, both n- and p-type polymers can be developed. The ZT values for organometallic complex-es therefore need to be further improved; they are around 0.01 for p-type materials and 0.2 for n-type materials.

TEG parameters (polymers)

With a stack of 35 thermocouples, TEGs from organome-tallic polymers have achieved surface power densities of approximately 1 µW/cm2 at a temperature difference of 25 K. For a PEDOT:Tos-based TEG, the surface power den-sity is around 0.27 µW/cm2 (ΔT=30 K). One reason for the low power density is believed to be the relatively high junction resistance between organic material and metal, which can amount to several ohms. For example, the junction resistance of PEDOT:PSS/Ag is about 5 Ohm*mm2.

With the discovery of electrically conductive polymers, organic materials were also investigated in terms of their suitability for thermoelectric applications. For research-ers they are of particular interest because they have a low thermal conductivity and are also flexible and non-toxic. Since rare elements are not required for their synthesis, large-scale applications are also conceivable. The poly-mers can be processed by continuous and automated technologies such as printing.

State of technology: Polymer material development

In recent years considerable progress has been made in the development of polymers. There are publications on the synthesis, modification and characterisation of a diverse range of polymers. Both purely organic materials (PANI, PPV, PPy:Tos, PEDOT:PSS or PEDOT:Tos) and orga-nometallic complexes have been investigated. Most of the organic polymers developed are p-type and can be doped by different molecules and methods. For instance, varying the oxidation states enables, for example, the charge carrier concentration in the polymer layers to be adjusted chemically or electrochemically with considera-tion precision. This enables the currently most promising

Fig. 22 (From left to right) 1 Pre-textured substrate material, 2 Printing and filling the porous material with, 3 and 4, p- and n-type polymer-filled material, 5 Printing all contacts, 6 Flexible polymer TEG with a size of 10 x 10 cm². Source: Fraunhofer IWS


Points of view

How thermoelectrics can emerge from the niche market

Despite the diverse global efforts that have resulted from the technologically interesting figures of merit, high-temperature thermo-electric generators capable of producing high power densities in converting otherwise unused waste heat have still not found their way into the mass market. This is mainly due to two challenges that developers are currently facing.

Firstly, a prerequisite for the serial use of TEGs is the high availability of highly efficient function materials with long-term stability for development work and production. Secondly, the contacts in modules manufactured with series production processes must also have long-term stability.

Professor Eckhard Müller German Aerospace Centre (DLR), Institute of Materials Research, Thermoelectric Materials and Systems, Cologne

To enable thermoelectrics to support the CO2 emission target for vehicles from 2020 onwards, the automotive industry needs to develop TEGs now and plan their integration into future vehicles. This will require a sufficient volume of TE modules for the development that have the potential to meet the special requirements of the automotive industry with regard to efficiency, quality, quantity and price by the start of production at the latest. And there must be durable strategies as to how this potential can be tapped.

In addition to achieving high temperature and change resistance as well as durability, a particular challenge is that considerable module efficiency is also required with the low exhaust temperatures in city operation in order to achieve a relevant reduction of CO2 in the test cycle.

Daniel Jänsch Senior Project Management, Program Management, Hybrid and Energy Recovery, MD Mechatronik Diesel IAV GmbH, Berlin Fig. 24 Dispenser print of a Monoleg TEG

Source: Fraunhofer IWS

Research on new thin-film thermocouples

Fig. 23 Microstructure of miniaturised, thin-film thermoelectric generators: more than 100 leg pairs per square millimetre are possible. Source: Micropelt

It is intended that micro cooling elements should achieve even better heat pump outputs. With this aim, researchers working in the joint Nanopelt project have investigated how further enhanced manufacturing technology and new thermoelectric coating materials can be used to produce more efficient thermoelectric thin-film components. The challenge is to fix a series of ultrathin layers extremely precisely, distortion-free and permanently so that they are resistant to loads caused by high temperature fluctuations.

The researchers replaced the silicon normally used as a substrate with highly polished copper wafers with very good thermal conducting properties. The chips manufactured in thin-film processes are separated by a laser process. To ensure that the wafers do not deform, the laser separation needs to be refined before it can go into industrial production. With the improved manufacturing process on copper wafers, the researchers are aiming to achieve temperature differences > 60 K and a heat pump output of 150 W/cm².

In practice


In 2013, Fujifilm for the first time presented a prototype of a polymer-based TEG for generating electrical energy from human body heat. It was sufficiently flexible to enclose curved surfaces. This TEG was exclusively assembled from a p-type polymer material (ZT> 0.27).

Printing thermoelectric materials

In addition to their interesting material properties, poly-mers have the advantage that they can be processed with different printing technologies. Various Fraunhofer insti-tutes are working on constructing thermoelectric genera-tors using printing technologies. At Fraunhofer IWS, re-searchers are synthesising and modifying n- and p-type polymers such as PEDOT:Tos for printed thermoelectrics. Dispenser technology enables them to manufacture gen-erators from only one material type with a mono-leg de-sign. Here lines from the conductive polymer are printed on polyimide strips and then later positioned upright (Fig. 24). A similar structure can also be used for TEGs made from p- and n-type polymer composites.

In addition to the variant with a printed, vertically mounted strip, other designs can also be realised which are adapt-ed from the classic TEG structure. One is based on a porous material (for example a fleece), which is partially filled with a polymer. This enables a robust but also flexible TEG structure. Very simple 3D structures can therefore be gen-erated (Fig. 22). Such a flexible polymer TEG can be used in the room temperature range up to about 200 °C.

The polymer poly-3.4-ethylenedioxythiophene (PEDOT, also known as PEDT) is flexible, and by doping becomes conductive with a negatively charged counter-ion. The polymer that results from combining it with polystyrene sulphonate (PSS) is then PEDOT: PSS.

Polyaniline (PANI) is a conductive polymer that, due to its properties, is also referred to as an organic metal.

Poly(p-phenylene-vinylene) (PPV) and polypyrrole (PPy) are also conductive polymers.

Producing TEGs with thin-film processes

Until now, thermoelectric components have mainly been labour-intensively produced from solid materials. A pro-cess developed at Fraunhofer IPM transfers semiconduc-tor production methods to thermoelectrics. The Micropelt technology is based on a patented “double-wafer concept”. Here n- and p-type Bi2Te3-based semiconductor materials are respectively deposited on the two separate wafers with pre-textured junction elements. The thermoelectric properties are then optimised by means of thermal post-treatment. A solder layer is then deposited on the wafers. The layer stack metallisation solder – the thermo-electric material – is then textured by dry etching.

Depending on the layout of the masks, thermoelectric structures can currently be realised with dimensions be-tween 35 and 600 µm. The component sizes for the final products typically range between 0.5 and 25 mm². Finally, the p- and n-components are then adjusted and soldered by means of a flip-chip bond.

The new thin-film process enables the wafer-based mass production of more efficient thermoelectric coolers and generators (see Fig. 26). These thin-film thermoelectric devices, which are just a few square millimetres thick, achieve very high power densities. They are ten times smaller than conventional elements producing compara-ble outputs. They can therefore be combined to form larg-er networks – enabling new compact designs.

Ceramic thermoelectric materials

Ceramic materials offer an alternative to the semi-metals and semiconductors previously mainly used in thermoe-

Fig. 25 Left: Coiled mono-leg TEG: The dark stripes are PEDOT:PSS, the grey stripes are the printed, flexible metal contacts. Source: Fraunhofer IWS, Centre: The thermocouple is created: The n- and p-type half-chips are joined together. Source: Micropelt, Right: Successful miniaturisation: The small element is used for stabilising the temperature of semiconductor lasers, the larger one for to supplying energy-autonomous radiator actuators. Source: Micropelt


En passant

Fig. 27 NASA is exploring the surface of Mars with its Mars rover Curiosity. The aim is to find out if organic life has ever existed on the planet. Source: NASA

With thermoelectrics through the galaxy

Space exploration needs electricity. Thermoelectric generators travel with spacecraft on journeys into the darkest regions of the universe remote from suns. They produce electrical energy from the heat generated by radioisotope batteries. This solution is used when a photovoltaic power supply is not possible or appropriate, for example with the Voyager spacecraft and the Mars rover Curiosity. Its energy source is a radioisotope generator of the latest generation, a so-called “Multi-Mission Radioisotope Thermoelectric Generator” (MMRTG). Such thermoelectric generators have long been used as reliable energy sources – TEGs were also used for the moon landings.

Fig. 29 NASA technicians installing the thermoelectric radioisotope generator in the rear of the Mars rover Curiosity. Source: NASA / Cory Huston

Fig. 28 Installing the third RTG in the Saturn probe Cassini. Source: NASA

lectrics such as bismuth telluride, lead telluride or silicon germanium. The ceramic materials are thermally stable and can also be used at high temperatures. For this pur-pose, researchers at Fraunhofer IKTS have developed car-bide, nitride and oxide ceramics as well as mixed ceram-ics. They have produced an optimised S-SiC material with high electrical and low thermal conductivity of less than 15 W/mK, as well as CVD thin-film SiC elements with a high Seebeck coefficient.

Galvanic production of thin-film TEGs

Thin films of bismuth telluride (Bi2Te3) can be produced by electrochemical deposition. The process enables homogeneous layers with thicknesses ranging from just a few nanometres to several millimetres to be realised over large areas. Galvanostatic, potentiostatic, electroless or pulsed methods are also described in literature.

Using pulsed potentiostatic deposition, researchers at Fraunhofer IPM have succeeded in producing thin films made of Bi2Te3 and (Bi,Sb)2Te3 on substrates up to 4 inches in size. The subsequent texturing and heat treatment of the layers has also been successfully achieved.

In the next step, the textured thermoelectric p-type materials (Fig. 26), which are provided with a diffusion barrier and gold contacts, can be soldered together with the counterpart consisting of n-type textured materials according to the Micropelt principle to form a thermoelectric module.

In practice

Fig. 26 Substrate with textured gold contacts, nickel diffusion barriers and textured Bi2Te3 legs. Source: Fraunhofer IPM


Internal combustion engines only utilise about one third of the energy contained in fuel in creating forward mo-mentum – most of the energy is currently lost as waste heat via the cooling and exhaust systems. Especially the energy contained in the hot exhaust gas could be partial-ly converted using TEGs into electrical energy in order to relieve alternators, charge batteries or operate auxiliary units in vehicles. That could reduce the fuel consumption and thus the CO2 emissions. In collaboration with research institutions, automotive manufacturers and suppliers have already demonstrated the use of thermoelectric generators in exhaust systems.

Generators in exhaust systems

The assembly and experimental testing of such TEGs is preceded by elaborate investigations and simulations of the system. TEGs interact with other components and must be adapted precisely to the respective application environment.

Since the engine cooling system is also used as the mod-ule cold side, this provides an additional heat input. On

the one hand this helps the combustion engine reach the operating temperature more quickly from a cold start, but on the other also contributes to increased cooling loads during hill journeys in summer. The combustion process is also directly affected by the TEG. By means of a heat-transfer structure, heat is drawn from the exhaust gas and added to the thermoelectric material. This results in an increased exhaust gas back pressure in front of the TEG, which can affect the charge exchange dynamics of the combustion processes.

Suitable for the respective vehicle, researchers are devel-oping a TEG system using experimental investigations on customised chassis dynamometers and in real driving tests. Based on these data and the occurring interactions, they are deriving designs aimed at improving the efficiency of the power-train.

Development of TEG components

In designing the TEG component, the developers are facing a major challenge: if the combustion engine is operated at high loads (for example, motorway driving), there is considerable potential for utilising the waste heat waste using the TEG. If, on the other hand, the evaluation of the system is in a more low-load driving cycle, the complexity of the sizing increases. Current work is therefore concerned with achieving the highest possible performance even in low-load driving conditions. The following measures are therefore being pursued:

• Consistent function and component integration

• Increased performance through better thermodynamic design of the exhaust and cooling water heat exchanger

• Improving the thermal junction resistance• Reducing parasitic heat flows

The long-term stability as well as the production-oriented assembly of the components are critical variables that need to be considered in the design. Encapsulating the TEM to

Automotive manufacturers need to improve

the energy efficiency of vehicle engines in order to

comply with strict CO2 limits. This requires more

than just internal engine modifications and lightweight

construction. Researchers believe that converting

previously unused waste heat will enable fuel savings

of up to 5 % in future.

Thermoelectrics in cars help to save fuel

Fig. 30 Usage and performance data from the customers‘ and manufacturers‘ viewpoints. Source: IAV

Reference vehicle Customer viewpoint Manufacturer viewpoint (Customer operation) (Approval cycle)

∅ Fuel consumption 10.0 l/100 km 7.5 l/100 km

∅ Fuel saving through TEG 2.50 % 1.75 %

∅ Fuel saving through TEG 0.2500 l/100 km 0.1312 l/100 km

∅ Fuel price 1.50 euros/l 1.50 euros/l

∅ Fuel costs saving 0.3750 euros/100 km 0.1968 euros/100 km

∅ CO2 reduction through TEG 6.3 g/km 3.3 g/km

∅ Cost of reference system (TEG) 250 euros 164 euros

∅ Distance covered until amortisation 67,000 km

∅ Cost per gram CO2 50 euros/g CO2


TEGs with high-temperature heat storage

In real driving conditions, TEGs only work for a fraction of the time at the optimum operating point when the heat transferring structure of the hot gas heat exchanger reaches the maximum hot side temperature of the thermoelectric modules. In load conditions above this, part of the exhaust flow is drawn off unused via a bypass in order to protect the TEG. With load conditions below this, the maximum temperature potential of the exhaust gas cannot be utilised.

To achieve a uniform heat supply, the researchers are integrating high-temperature heat storage material from a special metal alloy into the TEG’s laminar structure. This latent storage material, which is tailored to the high operating temperatures, can absorb and release heat at an almost constant temperature level. The researchers are investigating which materials are suitable, how they can be integrated into the system, and which savings can be achieved.

Exhaust gas mass flow

Coolant flow

Thermoelectric generator

Electricity fed into onboard network

1

1

2

3

4

2

3

4

Fig. 32 Schematic representation of the TEG integration into the exhaust system of a combustion engine-powered vehicle. Source: DLR - FK

Fig. 31 CAD model of a TEG component with a highly integrated design with integrated bypass system. Source: DLR - FK

enable it to be operated in a protective atmosphere pro-vides decisive advantages in this respect. An example of this optimisation method can be seen in Fig. 31.

What should TEGs in vehicles cost?

To achieve the CO2 emission limits for passenger cars in Europe, the automotive industry is constantly improving the energy efficiency of their vehicles. However, the pre-scribed CO2 targets for 2020+ have still yet to be achieved and further efforts are needed. Significant technical ad-vances are still possible, but must be available at a price acceptable to the market. All solutions that are simple and cheap are already being deployed, but their further development is reaching technical and economic limits. The measures still required entail higher costs with less effect. New concepts and technologies are needed, but these cannot be obtained cheaply or without incurring additional costs.

The costs of innovative technologies are initially borne by the vehicle manufacturers, who then pass on the addi-tional costs to the customer. Even if customers do not see the costs of fuel consumption- and emission-reducing measures, because they are incorporated in the overall vehicle costs, each one must be economically feasible in itself.

For customers buying new cars, lower operating costs and low fuel consumption are the most important pur-chasing criteria. More than half of new vehicles are used in the business customer segment. Here as well as with commercial vehicles, economic rationale is almost exclu-sively decisive for the adoption of technical innovations. Technologies only prevail in the market if the higher ac-quisition costs are balanced out by lower operating costs.

The question as to what a TEG may cost needs to be an-swered by customers and manufacturers (Fig. 30). The following explanations relate to a large and heavy, top-of-the-range reference vehicle. Here the pressure to use in-novative efficiency technologies is significantly higher.

Fig. 33 Prototype of a TEG with integrated latent heat storage: The more uniform heat provision is intended to create a greater electricity yield. Source: DLR - FK

In practice


Pyroelectric materials, magneto- and electrocaloric materials – magnetic cooling

How a magnetocaloric heat pump works: Magnetocaloric (MC) materials heat up due to magnetic ordering occurring in the material when it is exposed to a magnetic field (H-field). The heated MC material is connected to a heat sink so that heat can be dissipated. If the magnetic field is removed, the material cools down and is at a lower temperature than at the start of the cycle, enabling it to absorb heat. Magnetocaloric cooling systems can theoretically achieve up to 30 % higher efficiencies relative to compressor-based systems. Compared with thermoelectric cooling, the achievable temperature differentials per material are still small, however, which means that several magnetocaloric materials have to be connected in series.

Electrocaloric and pyroelectric materialsThe electrocaloric effect works in a very similar manner to the magnetocaloric effect except that the magnetic field is replaced by an electric field, such as a plate capacitor. The most common electrocaloric materials are ferroelectric. These exhibit internal electrical polarisation and are also always pyroelectric, i.e. a change in temperature causes a change in the polarisation (the inverse of the electrocaloric effect). If pyroelectric materials are heated, they electrically charge the

opposing surfaces and waste heat can be converted into electricity in a similar manner to thermo electric materials. However, considerable development work is still needed to make this technology more efficient.

Fig. 35 How magnetocaloric cooling works. Source: Fraunhofer IPM

Demagnetisation

Magnetisation

H-field

H-field

Heat absorption

T0

T0

T0 – ΔT

T0 – ΔT

Waste heat

Making range extender vehicles more efficient

The DLR - FK and Fraunhofer IPM are jointly researching the potential use of TEGs in electric vehicles that also use internal combustion engines as so-called range extenders. The TEG converts the waste heat from the engine into electrical energy. This recovered energy increases the limited range while reducing the fuel consumption and CO2 emissions. Researchers from the material development, module design, system develop-ment, simulation, prototype development and vehicle integration areas are working closely together to achieve this goal.

Fig. 34 Range extender test vehicle with TEG at the German Aerospace Centre‘s Institute of Vehicle Concepts. Source: DLR - FK

In practice

From the customer’s viewpoint, the TEG (reference sys-tem) would have paid back the additional 250 euros “at the petrol pump” after almost 67,000 km (Fig. 36). In terms of sensible payback periods, this represents a challenge for customers who travel comparatively short distances. The solid line shows the distance required un-til the return on investment (ROI) is achieved with TEG costs between 125 and 375 euros. The dashed curve shows the effect of varying the fuel savings, consumption and costs.

The manufacturer’s viewpoint: As the fuel consumption under the statutory test conditions is considerably below that achieved in real world traffic conditions, the refer-ence vehicle only consumes 7.5 l/100 km in the approval cycle. Due to the lower load levels here, the average fuel saving is also set lower with 1.75 %, so that on average 0.13 l/100 km of fuel is saved, which corresponds to about 3.3 g CO2/km. From various publications it is known that manufacturers have previously had to invest up to 50 euros for every vehicle and every gram of CO2 saved. This means that a TEG should not cost more than €164 (Fig. 33) if it is to be considered in the current pack-age of measures. However, it is also certain that the costs per gram of CO2 and the efficiency of TEGs will increase further by 2020. It is often commented that the €95/g excess emissions premium would actually then be


applicable. However, that would only be true when the automotive industry has exhausted all measures “cheaper” than €95/g and the fleet objectives are still not reached.

– 50 % – 30 % – 10 % 10 % 30 % 50 %

375 325 275 225 175 125

– 40 % – 20 % 20 % 40 %

115

125

105

95

85

75

65

55

45

35

25

ROI i

n th

ousa

nds

of k

ilom

etre

s

TEG costs (customer) [euros]

Reference system

0

TEG costs

– 50 % – 30 % – 10 % 10 % 30 % 50 %

248 215 182 149 116198231 165 132 99 83

– 40 % – 20 % 20 % 40 %

90

100

80

70

60

50

40

30

20

10

0

Cost

s pe

r red

uced

gra

m o

f CO

2 [eur

os/k

m]

Reference system

TEG costs (manufacturer) [euros]

TEG costs

0

Fig.38 A TEG consists of several of these „all-in-one“ modules. The individual hot side heat exchangers can be seen here that are integrally connected to the TE materials concealed underneath; water connections for cooling are on both sides; the circuit board with the electronics and voltage transducer is situated on the lower side of the module (Fig. 39). Source: IAV

Modular TEGs for vehicles

Working in conjunction with partners in the „TEG 2020“ project funded by the German federal government, the Ingenieurge-sellschaft Auto und Verkehr IAV company has developed concepts and systems that enable waste heat from various indus-trial and drive systems to be used efficiently and economically with thermoelectric generators (TEGs). The development has focussed on a TEG modular concept and innovative „all-in-one“ modules (Fig. 38) where the thermoelectric material is integrally connected to the heat exchanger. This enables TEGs to be scaled for different applications, targeting systems and power classes. Several TEG systems were assembled for the project and investi-gated on test rigs and in the exhaust system of a test vehicle. The sustainability of the developed concepts and the operational ability of the systems and components have been successfully verified.

In practice

Fig. 39 The modular structure of the TEGs enables the system to be sized according to the needs and output of the application. Source: IAV

Fig. 36 Amortisation of TEG costs for customers (variations in the fuel savings or average consumption or fuel costs). Source: IAV

Fig. 37 Manufacturer‘s costs for each reduced gram of CO2 (vari-ations in the fuel savings or average consumption). Source: IAV


Using the example of a Combined Heat and Power plant (CHP) and a single-family home, Fraunhofer IPM is demonstrating the efficiency increase that could be achieved by using TEGs:

At the expense of the amount of heat released, the elec-trical outputs of CHP plants can be increased by up to 5 % by using thermoelectric modules on the exhaust heat ex-changer. For a typical CHP plant with a 145-kW gas input and a total efficiency of about 90 %, this would generate an additional electrical output of about 1.5 kW. With an assumed annual operating time of 90 %, this would be equivalent to approximately 7,700 kWh/p.a. or an annual revenue of around 1,500 euros (with 20 cents/kWh).

The researchers assume that, with low-volume production, the required TE modules would cost about 15,000 euros (15 euros/watt). The investment would pay for itself with-in 10 years. With large quantities, the costs would reduce to about one euro per watt and the investment would pay for itself in less than a year.

Savings potential in the home

The central heating system for a single-family home converts about 20,000 kWh into heat per year.

Thermoelectric modules on the heating boiler would enable 5 % of the heat generated to be converted into electricity, which would generate an additional annual electric output of 1,000 kWh. This could help to relieve the power grid in the cold season when it is most heavily loaded. Based on an electricity price of 20 cents/kWh and minus 6 cents/kWh for the primary energy used, a household would save 140 euros per year. With large quantities, payback would be achieved within five years.

Thermoelectric generators produce electricity without any moving parts. They therefore require very little mainte-nance and remain stable in the long term. They are also easy to install and are easily scalable. This therefore makes them particularly suitable for applications where

heat volumes less than 150 kW are available. Especially in this segment they will be more economical than other energy technologies such as Organic Rankine Cycle (ORC) or Stirling engine systems, and where the installation and maintenance costs are comparatively high as a result of moving parts, their complex structure as well as due to their small operation window.

Very low maintenance

Even with waste heat volumes considerably above 150 kW, TEGs particularly impress in regards to power plants or increasing the efficiency of industrial processes through their low installation and maintenance costs.

However in stationary use, for example with CHP plants, very efficient cycle processes such as the ORC system achieve at a steam temperature of 95 – 200 °C signifi-cantly higher system efficiencies than TEGs that amount to around 10 – 20 %. Therefore thermoelectrics can only be successful in applications where either ORC systems are not permitted for safety reasons, such as with aluminium electrolysis or steel production, or where the easy scala-bility of thermoelectrics enables small systems to be pro-duced at very low cost, for example for grid-independent gas boilers. One reason for this is the short amortisation periods with which industry rates investments.

Good technology needs to be even cheaper

Despite these advantages over other technologies, thermo-electrics is not yet being used on a wide scale for utilising waste heat.

Depending on the underlying conditions, waste heat can, however, already be economically utilised with lower material efficiencies, at a figure of merit ZT at or below 1. In addition to the minimum material efficiency required for the application, it is also important when choosing a thermoelectric material to select cost-effective raw mate-rials, as these constitute a major cost item.

Despite their still low efficiency, the use of

thermoelectrics is worthwhile – and not just for

supplying sensors and special applications:

thermoelectrics can significantly improve the

efficiency of existing processes,

and is also suitable for application fields

where other methods cannot be used.

Utilising waste heat saves energy and CO2


Fig. XX Abbildungsunterschrift Quelle: Name der Quelle

Generating electricity from the waste heat produced by continuous rolling mill trains

High-temperature industrial processes generate large amounts of waste heat. Waste heat flows with temperatures partly exceeding 500 °C are available. Using steel processing as an example, a research project is investigating how high-temperature TEGs can utilise the waste heat from industrial processes.

To this end, the researchers are developing a compact and robust, large-scale TE module. This is being tested and optimised on a laboratory scale at the VDEh-Betriebsforschungsinstitut (BFI research institute) as well as in the industrial environment provided by the Salzgitter Flachstahl steel company. The experts expect that the process will enable 4 – 6 % of the waste heat energy to be converted into electricity.

It is intended that the TE module will be permanently deployed on the roller table belonging to a continuous rolling mill train. The aim is to achieve optimal heat transfer and efficiency under practical conditions. In long-term use, the researchers are studying the influences of the installation location, cooling, and environmental influences such as dusty atmospheres. Currently they are testing the TEG modules on a laboratory test rig in order to optimise the heat transfer and thermal shock resistance of the modules.

The research project is aiming to demonstrate the use of high-temperature TEGs for generating power from waste heat radiation and to derive an optimised application for industry.

Thermoelectric materials on route to achieving industrial maturity – The ThermoHeusler project

Special metal alloys, the so-called half-Heusler compounds, are ideally suited for producing thermoelectric modules. The Vacuumschmelze and Isabellenhütte special alloy manufacturers can now produce them in kilogram quantities and much more efficiently and cost effectively than was previously possible.

The alloys consist of widely available raw materials such as nickel, are considerably more environmentally friendly than the materials previously used, have good thermoelectric properties and withstand high temperatures. In the ThermoHeusler project, the researchers have achieved the best value so far for half-Heusler compounds with a ZT of 1.2.

Thermoelectric modules are composed of blocks just a few millimetres thick. Using a contacting process developed as part of the ThermoHeusler project, developers are producing optimum electrical junctions that withstand large temperature differences, have long-term stability and at the same time low electrical resistance.

In practiceIn practice

Fig. 41 In the ThermoHeusler2 project, thermoelectric modules from half-Heusler compounds are being manufactured for integration in vehicles. Source: Fraunhofer IPM, Kai-Uwe Wudtke

Fig. 40 The blocks used for thermoelectric modules, which are just a few millimetres thick, are sawn from special metal alloys such as, for example, half-Heusler compounds. Source: Fraunhofer IPM

Fig. 42 Roller table for the continuous rolling mill train at Salzgitter Flachstahl. Source: Salzgitter Flachstahl

Fig. 43 It is planned to integrate a demonstrator TEG in the hood of the roller table for the continuous rolling mill train. Source: Salzgitter Flachstahl /VDEh-Betriebsforschungsinstitut

ImpressumProjektorganisation Bundesministerium für Wirtschaft und Energie (BMWi)11019 Berlin

Projektträger Jülich Forschungszentrum Jülich GmbH 52425 Jülich

Förderkennzeichen 00327430M0327430H0327387A-D0335007P

ISSN 1610-8302

Herausgeber FIZ Karlsruhe · Leibniz-Institut für Informationsinfrastruktur GmbH Hermann-von-Helmholtz-Platz 1 76344 Eggenstein-Leopoldshafen


Links und Literatur>> www.XXX.de >> www.XXX.de >> www.XXX.de >> www.XXX.de >> www.XXX.de >> Literaturhinweis>> Literaturhinweis >> Literaturhinweis>> Literaturhinweis >> Literaturhinweis>> Literaturhinweis >> Literaturhinweis>> Literaturhinweis

Mehr vom BINE Informationsdienst>> XXX. BINE-Projektinfo XX/20XX >> XXX. BINE-Projektinfo XX/20XX>> Dieses Themeninfo gibt es auch online und in englischer Sprache unter

www.bine.info/Themeninfo_X_20XX

BINE Informationsdienst berichtet aus Projekten der Energieforschung in seinen Broschürenreihen und dem Newsletter. Diese erhalten Sie im kostenlosen Abonnement unter www.bine.info/abo

ÜberschriftWeit hinten, hinter den Wortbergen, fern der Länder Vokalien und Konsonantien leben die Blindtexte. Abgeschieden wohnen sie in Buchstabhausen an der Küste des Seman-tik, eines großen Sprachozeans. Ein kleines Bächlein namens Duden fließt durch ihren Ort und versorgt sie mit den nötigen Regelialien. Es ist ein paradiesmatisches Land, in dem einem gebratene Satzteile in den Mund fliegen. Nicht einmal von der allmächtigen Interpunktion werden die Blindtexte beherrscht – ein geradezu unorthographisches Leben. Eines Tages aber beschloß eine kleine Zeile Blindtext, ihr Name war Lorem Ip-sum, hinaus zu gehen in die weite Grammatik.

Der große Oxmox riet ihr davon ab, da es dort wimmele von bösen Kommata, wilden Fragezeichen und hinterhältigen Semikoli, doch das Blindtextchen ließ sich nicht beir-ren. Es packte seine sieben Versalien, schob sich sein Initial in den Gürtel und machte sich auf den Weg. Als es die ersten Hügel des Kursivgebirges erklommen hatte, warf es einen letzten Blick zurück auf die Skyline seiner Heimatstadt Buchstabhausen, die Headline von Alphabetdorf und die Subline seiner eigenen Straße, der Zeilengasse. We-hmütig lief ihm eine rhetorische Frage über die Wange, dann setzte es seinen Weg fort.

Die Copy warnte das Blindtextchen, da, wo sie herkäme wäre sie zigmal umgeschrieben worden und alles, was von ihrem Ursprung noch übrig wäre, sei das Wort „und“ und das Blindtextchen solle umkehren und wieder in sein eigenes, sicheres Land zurückkehren. Doch alles Gutzureden konnte es nicht überzeugen und so dauerte es nicht lange, bis ihm ein paar heimtückische Werbetexter auflauerten, es mit Longe und Parole betrunk-en machten und es dann in ihre Agentur schleppten, wo sie es für ihre Projekte wieder und wieder mißbrauchten.

Konz

ept:

iser

unds

chm

idt G

mbH

, Bon

n · S

atz

und

Ges

taltu

ng: K

ERST

IN C

ON

RAD

I Med

ieng

esta

ltung

, Ber

lin

Kontakt · InfoFragen zu diesem Themeninfo? Wir helfen Ihnen weiter:

0228 92379-44 [email protected] Informationsdienst Energieforschung für die PraxisEin Service von FIZ Karlsruhe

Kaiserstraße 185-197 53113 Bonn www.bine.info

ImprintProject organisation Federal Ministry for Economic Affairs and Energy (BMWi) 11019 Berlin Germany

Project Management Jülich Forschungszentrum Jülich GmbH 52425 Jülich Germany

Project number 03ET1164A-C0327863A-G 0327876A-D, G 19U15006A-G 03X4506A

ISSN 1610-8302

Publisher FIZ Karlsruhe · Leibniz Institute for Information Infrastructure GmbH Hermann-von-Helmholtz-Platz 1 76344 Eggenstein-Leopoldshafen Germany


Links and literature>> Deutsche Thermoelektrik-Gesellschaft e.V. (DTG) | www.thermoelektrik.info >> Fraunhofer-Netzwerk Thermoelektrik | www.thermoelektrik.fraunhofer.de >> Fraunhofer Institute for Physical Measurement Techniques IPM | www.ipm.fraunhofer.de >> Fraunhofer Institute for Integrated Circuits IIS | www.iis.fraunhofer.de >> Fraunhofer Institute for Ceramic Technologies and Systems IKTS

www.ikts.fraunhofer.de >> Fraunhofer-Institut für Werkstoff- und Strahltechnik IWS | www.iws.fraunhofer.de>> Fraunhofer Institute for Manufacturing Technology and Advanced Materials IFAM

www.ifam.fraunhofer.de>> German Aerospace Center DLR | www.dlr.de>> Ingenieurgesellschaft Auto und Verkehr IAV | www.iav.com >> Micropelt | www.micropelt.de >> European Thermoelectric Society | www.thermoelectricity.eu/>> International Thermoelectric Society | www.its.org/>> VdEH Betriebsforschungsinstitut | www.bfi.de

More from BINE Information Service>> Energy-saving production of monocrystalline semiconductors. BINE-Projektinfo 1/2015>> This Projektinfo brochure is also available as an online document

at www.bine.info/Themeninfo_I_2016

BINE Information Service reports on energy research projects in its brochure series and the newsletter. You can subscribe to these free of charge at www.bine.info/abo

Thermoelectrics conquers new applications To ensure that energy harvesting is able to diversify beyond niche applications such as supplying energy-autonomous sensor networks, thermoelectric modules must become considerably more efficient and affordable if they are to be widely deployed in, for example, cars, energy systems and industrial plants.

Researchers expect that electricity generated from waste heat using thermoelectrics will make an important contribution to the efficient use of energy in the future. New thermoelectric materials, improved production processes and optimised electrical and thermal management will enable a diverse range of applications. If, for example, print-based methods succeed in producing TEGs inexpensively and over large areas, it will be possible to use them on a large scale for recovering waste heat in the industrial and energy sectors.

Research has focussed until now on improving the thermoelectric materials. Today, new materials such as half-Heusler compounds, skutterudites or silicides achieve significantly greater efficiencies than a few years ago. Intensive work is underway to develop suitable materials and production processes for mass production. Scientists at Fraunhofer IPM have succeeded in halving the amount of materials used for TE modules – for the same performance. This will make them cheaper and lighter in future, which is especially advantageous for mobile applications. This saving of TE material is an important step towards industrialising the module production.

The production of permanently stable TEGs places considerable demands on the selection and combination of suitable materials and on the permanent connection of the individual components to the finished module. Across the entire value chain – from the TE material and module development to modular system concepts, system integration and the construction of demonstrators – researchers in the ThermoHeusler2 project are currently developing the small-scale batch production of thermoelectric generators that are designed to utilise waste heat from internal combustion engines.

Conc

ept:

iser

unds

chm

idt G

mbH

, Bon

n –

Ber

lin, G

erm

any

· Des

ign

KERS

TIN

CO

NRA

DI ·

Med

ieng

esta

ltung

, Ber

lin, G

erm

any

Contact · InfoQuestions regarding this Themeninfo brochure? We will be pleased to help you:

+49 228 92379-44 [email protected] Information Service Energy research for application A service from FIZ Karlsruhe

Kaiserstraße 185-197 53113 Bonn, Germany www.bine.info

http://www.bmwi.de/EN/root.html

Thermoelectrics: power from waste heat · Thermoelectric modules can not only convert heat directly...

Documents

Transcript of Thermoelectrics: power from waste heat · Thermoelectric modules can not only convert heat directly...