Develop Thermoelectric TechnologyDevelop Thermoelectric Technology for Automotive Waste Heat Recoveryfor Automotive Waste Heat Recovery

Sponsored by

US Department of Energy

Energy Efficiency Renewable Energy (EERE) DE-FC26-04NT42278

Outline � Background

− Objective, partnering, … − Motivation: fuel economy

� Technology Development Jihui Yang − Subsystem modeling

GM Research & Development Center − Cost − Cost-effective materials

� Summary

1

DOE Program ObjectiveDOE Program Objective

Demonstrate a 10% fuel economy improvement using thermoelectric waste heat recovery technology, without increased emissions and at a cost-effective way.

2

PartneringPartnering• GM

– Materials Research – Subsystem design, integration, modeling, and validation

• GE – TE module design and construction – Subsystem design and construction

• Oak Ridge National Lab – High temperature materials properties measurement

• RTI – Superlattice-base materials and modules

• University of South Florida – bulk materials development: clathrates, nano-grain PbTe, …

• University of Michigan – Bulk materials development, skutterudites, nano-composites,…

3

Accessories2.2 (1.5)%

Accessories2.2 (1.5)%

Energy Distribution –Energy Distribution –Typical Mid-Size Vehicle on Federal Test Procedure (FTypical Mid-Size Vehicle on Federal Test Procedure (F-- Urban (Highway) % energy usUrban (Highway) % energy use

4

Standby/Idle 17.2 (3.6)%

Engine Losses 62.4 (69.2)%

Driveline Losses 5.6 (5.4)%

Aero 2.6 (10.9)%

Rolling 4.2 (7.1)%

Kinetic

Braking 5.8 (2.2)%

Engine 100% 18.2%

(25.6%) D/L 12.6% (20.2%)

PNGV source

Standby/Idle 17.2 (3.6)%

Engine Losses62.4 (69.2)%

Driveline Losses5.6 (5.4)%

Aero2.6 (10.9)%

Rolling4.2 (7.1)%

Kinetic

Braking5.8 (2.2)%

Engine100% 18.2%

(25.6%) D/L 12.6%(20.2%)

PNGV source

Example of an Energy RecoveryExample of an Energy RecoverySystemSystem

TransmissionEngine

Catalyst ERS

Radiator

Alternate: make the radiator into an energy recovery device.

(smaller ΔT)5

Exhaust as Potential TE Heat SourceExhaust as Potential TE Heat Source

kW

Exhaust Heat

80

70

60

50

40

30

20

10

0

0 500 1000 1500 2000 2500

Test Time (s)

6

Thermoelectric Energy ConversionThermoelectric Energy Conversion

4.0 room temperature

3.5 PbSexTe1-x/PbTe quantum dots

3.0 Bi2Te3/Sb2Te3 superlattices

2.5

2.0

CeyCoxFe4-xSb121.5 Yb0.19Co4Sb12 Ba0.3Co3.95Ni0.05Sb12Bi2Te3 La CoFe3Sb121.0

0.9 SiGe PbTe

Efficiency: ε = THT

− TC 1 + ZT − T 1 C

0.5

H 1 + ZT + 0.0 TH

0 200 400 600 800 1000 1200 1400 T (K)

ZT

7

8

IC Engine

Electrical Storage Transmission

Electric Power

Controls

Powertrain Control

Electronics

Catalytic Converter Heat Collector

Thermoelectric Device

Cooling

Exhaust

Vehicle Electrical System

Alternator

Generic Concept for a Thermoelectric Energy Recovery Augmented Electrical System

Generic Concept for a Thermoelectric Energy Recovery Augmented Electrical System

9

Subsystem ModelingSubsystem Modeling

10

TE Module ModelingTE Module Modeling

n p

metal

metalmetal

metal

conductor

Contact Resistances

conductor conductor

insulator

insulator

Hot side heat exchanger

Cold side heat exchanger

thermal

thermal

Electrical & thermal

thermal

thermal

volume resistivity

Electrical & thermal

Thermal shunt path

Joule & Peltier terms

Joule heating from all electrical contacts are accounted for.

11

0

1

2

3 4n4p

5p 5n

6n6p

7n7p 8

9

10

11

Preliminary Temperature ProfilePreliminary Temperature Profile

0 100 200 300 400 500 600 700 800 900

0 1 2 3 4 5 6 7 8 9 10 11

Tem

pera

ture

(K)

20% 15% 5%

ΔTE ΔHX

Position in TE device (node)

60 kW exhaust flow

Validation of GE’s subsystem ModelValidation of GE’s subsystem Model with GEN I TE Generatorwith GEN I TE Generator

Experiment Model Output Power 109W 121W Vload 16.5V 17.4V Current, I 6.6A 7.0A Hot side temperature 183C 189C Cold side temperature 35C 35C

12

On the Cost of Fuel EconomyOn the Cost of Fuel Economy

2000

1800 $2/gallon$3/gallon1600 � Cost of fuel economy - $/Δmpg$4/gallon

1400 •This kind of calculation can be used to balance technology options1200

1000

� $/Δmpg ≤ Savings/Δmpg is necessary800

to provide consumer value600

400

200

00 10 20 30 40 50 60 70

FE (mpg)

Consumer Fuel Savings/Δmpg ≈ $300-400/Δmpg (15000 mile/yr., 3yrs., 18-20 mpg)

3 Y

ear F

uel S

avin

g pe

r 1m

pg F

E im

prov

emen

t

13

14

Skutterudites Nano-101Skutterudites Nano-101 � CoAs3 -based minerals found in region

of Skutterud, Norway � Compounds with the same crystal

structure are known as “skutterudites” � Filled skutterudites are electron-crystal-

phonon-glass materials

2 2

( )T L e

S SZ κ ρ κ κ ρ

= = +

Co Sb

�2Co4Sb12

Filler atoms

Filled skutterudites as CandidateFilled skutterudites as Candidate Materials for Exhaust Heat RecoveryMaterials for Exhaust Heat Recovery

ZT

1.5 room temperature CeyCoxFe4-xSb12 (p-type)

� High ZT values near the exhaustYb0.19Co4Sb12(n-type) Ba0.3Co3.95Ni0.05Sb12

(n-type) temperature1.0

La CoFe3Sb12 (p-type)0.9

� Both high performance n- and p-typeBi2Te3 SiGe PbTe exist – suitable for TE module

0.5

0.0 0 200 400 600 800 1000 1200 1400

T (K)

15

MischmetalMischmetal �Mischmetal: from German - “mixed metals”; a

naturally occurring alloy of rare earth elements � Composition (wt%):

Ce : 50-55 La : 30-35 Nd : 5-10 Pr : 5-10 � Cost Comparison*:

Mischmetal (99.0%): $0.19/g Cerium rod (99.9%): $8.37/g Lanthanum rod (99.9%): $6.40/g

* source: Alfa Aesar

16

Mischmetal Starting MaterialMischmetal Starting Material

Element Wt.%

Ce 52.4 La 23.5 Nd 17.1 Pr 5.9 Si 0.6 Fe 0.3 Al 0.1

Sum 99.9

electron probemicroanalysis

wet chemistry Element Ce La Nd Pr Sample

%

53 23 18 5

17

18

Electron Probe Microanalysis of Mischmetal-filled Skutterudites Electron Probe Microanalysis of Mischmetal-filled Skutterudites

9.1433(2)Mm0.88Fe4Sb12.06MmFe4Sb12 *

9.1294(2)Mm0.82Fe3.51Co0.49Sb11.99Mm0.93Fe3.5Co0.5Sb12 *

9.146(1)Mm0.82Fe4Sb11.96MmFe4Sb12

9.126(1)Mm0.72Fe3.43Co0.57Sb11.97Mm0.93Fe3.5Co0.5Sb12

9.117(1)Mm0.65Fe2.92Co1.08Sb11.98Mm0.82Fe3CoSb12

9.109(1)Mm0.55Fe2.44Co1.56Sb11.96Mm0.71Fe2.5Co1.5Sb12

Room temperature lattice parameter (Å)Actual compositionNominal composition

� nominal comp. chosen according to the optimal filled skutterudites CeyFe4-xCoxSb12 ~ PRL 80, 3551 (1998)

� Mm actual concentration < nominal due to high Mm vapor pressure

� all samples are nearly phase pure � typical secondary phase:

MnSb2 or CoSb2 < 1 vol. %

19

Isotropic Atomic Displacement Parameter – Evidence of Rattling Isotropic Atomic Displacement Parameter – Evidence of Rattling

U is

o (Å

2 ) 0.000

0.005

0.010

0.015

0.020

0.025

T (K) 0 50 100 150 200 250 300

U is

o (Å

2 )

0.000

0.005

0.010

0.015

0.020

Mm0.82Fe3.51Co0.49Sb11.99

Mm0.88Fe4Sb12.06

Mm

Fe/Co

Sb

Mm

Fe

Sb

� low T intercepts of Uiso represent a combination of zero-point vibration and static disorder at the corresponding crystallographic sites

� zero-point vibration ∝ 1/M, expected contribution ~ Mn < Sb < Fe Æstatic disorder at the Sb sites, more so at the Mn sites (Mn vs. �)

� θE ~ 71.5 K ( La: 80 K, Ce: 78 K, Eu: 83 K, and Yb: 65 K)

Room Temperature PropertiesRoom Temperature Properties ComparisonComparison

100 1.2

Ce Fe4-xCo Sb12y x

MmyFe4-xCoxSb12

La0.9Fe3CoSb12

10 Ce0.9Fe3CoSb12

1

Latti

ce th

erm

al re

sist

ivity

(m K

/W)

1.0

0.8

0.6

0.4

0.2

S2 /ρ

(μW

/cm

K2 )

0.1 0.01 10 100 1000 10000 0.0 0.2 0.4 0.6 0.8 1.0 p (1018 cm-3) y

20

21

Summary of Challenges for Automotive TE Waste Heat Recovery Summary of Challenges for Automotive TE Waste Heat Recovery

Outline � Background

− Objective, partnering, … − Motivation: fuel economy

� Technology Development − Subsystem modeling − Cost − Cost-effective materials

� Summary

Need better materials, heat exchanger, contact joining, thermal & mechanical stability, …

Consumer focus: consider $/Δmpg (also valuable for balancing tech. options)

Low cost materials is a must

Thank you ! Questions ?