Thermoelectric Technology forThermoelectric Technology for Automotive Waste Heat RecoveryAutomotive Waste Heat Recovery

Jihui YangGM Research & Development Center2007 DEER Conference, Detroit, MI

August 15, 2007 Outline � Introduction � Engineering Highlights � Materials Research Highlights � Automotive TE cooling?

- environmental impact of R134a Sponsored by � Summary

US Department of Energy Energy Efficiency Renewable Energy (EERE)

Waste Heat Recovery and Utilization Research and Development for Passenger Vehicle and Light/Heavy Duty Truck Applications

DE-FC26-04NT42278

Thermoelectric Power Generation fromThermoelectric Power Generation from Automotive Waste Heat RecoveryAutomotive Waste Heat Recovery

• Target : 10% fuel economy improvement without increasing emissions

• Partnering: – GM – materials research, subsystem design, integration, modeling, and

validation – GE – TE module, subsystem design and manufacturing – Oak Ridge National Lab – high temperature material property

measurement and validation – RTI – superlattice-based thin film materials and modules – University of Michigan – bulk materials : filled skutterudites, nano-

composites,… – University of South Florida – bulk materials: clathrates, nano-grain

PbTe, … – Michigan State University – bulk PbTe-based materials …

DOE Program High-Level ProcessDOE Program High-Level Process

Science: GM, RTI, UM, MSU & USF

Engineering: GE, GM, & RTI

Target: Cost-effective 10% FE Improvement

Material Data Validation: ORNL

Techno

logy D

evelop

ment

Microstructural

From Science, Engineering, toFrom Science, Engineering, to TechnologyTechnology 0.12 Wout0.0.1212 Wout

A Candidate TEA Candidate TE MaterialMaterial 1818

0.1 234.80.0.11 234.8451.45 61.6

0.08 668.40.0.0808 668.4885.88 25.2

0.06 11020.0.0606 110213191319153615360.040.0.04041752175219691969

0.020.0.020221862186

000350 400 450 500 550 600 65035350 400 450 500 550 600 6500 400 450 500 550 600 650

Exhaust Flow Input Temperature [C]Exhaust Flow Input Temperature [C]Continuum

Exha

ust M

ass

Flow

[kg

/s]

Exha

ustM

ass

Flow

[kg/

s]

Competency

Engineering

Cold-side heat exchanger Exhaust

outlet to muffler

TEm odules

Materials Science Exhaust heat

exchanger Cold-side heatAtomistic exchanger

Hot exhaust inlet

Chemistry Electronic

log(Length scale) Physics

nm μm mm m

Standby/Idle 17.2 (3.6)%

Engine Losses 62.4 (69.2)%

Accessories 2.2 (1.5)%

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%)

Standby/Idle17.2 (3.6)%

Engine Losses62.4 (69.2)%

Engine Losses62.4 (69.2)%

Accessories2.2 (1.5)%

Accessories2.2 (1.5)%

Driveline Losses5.6 (5.4)%

Driveline Losses5.6 (5.4)%

Aero2.6 (10.9)%

Aero2.6 (10.9)%

Rolling4.2 (7.1)%Rolling4.2 (7.1)%

KineticKinetic

Braking5.8 (2.2)%Braking5.8 (2.2)%

EngineEngine100% 18.2%

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

PNGV source

• Today’s ICE-based vehicles: < 20% of fuel energy is used for propulsion

• > 60% of gasoline energy (waste heat) is not utilized

Energy Distribution - Typical Mid-Size Vehicle on the Federal Test Procedure (FTP) Schedule Urban (Highway) % energy use

Energy Distribution - Typical Mid-Size Vehicle on the Federal Test Procedure (FTP) Schedule Urban (Highway) % energy use

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0 500 1000 1500 2000 2500 3000

TIME (SEC)

EXH

AUS

T G

AS F

LOW

RAT

E (k

g/s)

0

200

400

600

800

1000

1200

TEM

PER

ATUR

E (K

)

EXHAUST MASS FLOW

EXHAUST GAS TEMPERATURE

COOLANT TEMPERATURE

LD9 ENGINE --EPA FTP CYCLE

Exhaust Flow and Temperatures for a 4 Cylinder Engine Exhaust Flow and Temperatures for a 4 Cylinder Engine

IC Engine

Electrical Storage Transmission

Electric Power

Controls

Powertrain Control

Electronics

Catalytic Converter Heat Collector

Thermoelectric Device

Cooling

Exhaust

Vehicle Electrical System

Alternator

• Engine coolant is also a possible heat source, but smaller ΔT

Thermoelectric Energy Recovery Augmented Electrical System Thermoelectric Energy Recovery Augmented Electrical System

Nano-derived materials give additional controlNano-derived materials give additional control� Enhanced density of states due to quantum confinement effects

⇒Increase S (sharp edge of electron D.O.S. near Fermi surface EF) ⇒without reducing σ ( very high electron D.O.S. near Fermi surface EF)

� Boundary scattering at interfaces reduces κL more than σ � Possibility of materials engineering to further improve ZT

3D E 2D E

EF 1D E 0D E

Bulk semiconductor Quantum well Quantum wire Quantum dot

conductor thermopower temperature electrical conductivityinsulator

2

κ ρ κ κ electronic

L e thermal

TE Figure S T S 2σT of Merit ZT = =

+electronelectron phonon T conductivity

total thermal conductivity electrical resistivity lattice thermal conductivity

D. O

. S.

D. O

. S.

D. O

. S.

D. O

. S.

Thermoelectric Waste Heat RecoveryThermoelectric Waste Heat RecoveryEfficiency:

T − T 1 + ZT − 1ε = H C

TT H 1 + ZT + C

HT

Year 1940 1950 1960 1970 1980 1990 2000 2010

ZT

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

ZnSb PbTe

Bi2Te3 Zn4Sb3

filled skutterudites

Bi2Te3/Sb2Te3 superlattices

PbSeTe/PbTe quantum dots

AgPb18SbTe20

• Many recent material advances are nano-based

$ / W – a Program Metric$ / W – a Program Metric� $/W (not only ZT) is used for balancing various material, module, and subsystem options

� $/W can be readily converted to $/Δmpg, and $/Δmpg < Savings/Δmpg is necessary to provide consumer value

FE (mpg) 0 10 20 30 40 50 60 70

3 Y

ear F

uel S

avin

g pe

r 1m

pg F

E im

prov

emen

t

0

200

400

600

800

1000

1200

1400

1600

1800

2000

$2/gallon $3/gallon $4/gallon

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

� plenty of space for accommodating TE subsystem � a lot of waste heat: exhaust and radiator � current muffler: 610 x 310 x235 (mm) � available envelope: 840 x 360 x 255 (mm)

Vehicle Selection – Full Size TruckVehicle Selection – Full Size Truck

Typical Exhaust Heat -Ci ty Driving Cycle

0

10

20

30

40

50

60

70

80

0 500 1000 1500 2000 2500

Test Time (s)

kW

Fuel Economy Analysis - OFuel Economy Analysis - verdriveOverdriveThrottle PositionThrottle Position Cycle Vehicle SpeedCycle Vehicle Speed 19.6

EngineEngine DrivDri erver19.4

Midsize TruckFuel &Fuel &Fuel & Throttle Transmission BrakingThThrottlerottle Transmission BrakingControlContControlrol BodyBodBodyy 19.2

VehicleVehicle19.0

Intake ShiftIntaIntakeke ShiftShift Wheel SlipWhWheel Slipeel SlipManifold ControlMaManifoldnifold ControlControl & Rolling& R& Rollingolling

FE (m

pg)

18.8

4167 lbs18.6

CyliCylindernder TorquTorquTorqueee GeaGeaGearrr Fore/AFore/AFore/Afffttt 18.4 BlBlockock ConvertConvertConvertererer CaCaCassseee DynaDynaDynammmicsicsics 4417 lbs

18.2 4667 lbs

SpinSpinSpin AeroAeroAero18.0

LossesLossesLosses DraDraDraggg 17.8 0 200 400 600

EL (W)

10% fuel economy improvement for a full size truck would require (for a hybrid full size truck): � 1.65 kW – city � 2.5 kW – Highway

350 W is the minimum requirement � equal to the base electrical load of today's generator on FTP, and would improve its composite

Urban/Highway fuel economy by ~ 3%

Subsystem Model SchematicSubsystem Model Schematic

Subsystem ModelingSubsystem Modeling

Exhaust: TH, TC, flow & fluid properties

Radiator: TH, TC, flow & fluid properties

Elec. Load: Effective R, Vreq,

GM’s Inputs GE’s System Model

Device config. • Structure / design Heat transfer • Hot & cold exchangers • Interfaces Power conditioning • DC/DCc onverter

Requirements: Cost, Power, Life

GM’s MPG modelOutput

cost

Δmass, ΔP, power

% FESavings

GM’s FETo ols

Materials Dev. • α(T),ρ(T),λ(T) •Cost & mass •Geom. req.

Subsystem Modeling – 2 D Thermoelectric Network Model Subsystem Modeling – 2 D Thermoelectric Network Model

n p

metal

metalmetal

metal

conductor

Cont act Resist ances

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.

n p

metal

metalmetal

metal

conductor

ContactResistan

n p

meta

ces

conductor conductor

insulator

insulator

Hot side heat exchanger

Cold side heat exchanger

thermal

thermal

Electrical &thermal

thermal

thermal

volumeresistivity

Electrical &thermal

Thermalshuntpath

Joule & Peltier terms

l

metalmetal

metal

conductor

ContactResistances

conductor conductor

insulator

insulator

Hot side heat exchanger

Cold side heat exchanger

thermal

thermal

Electrical &thermal

thermal

thermal

volumeresistivity

Electrical &thermal

Thermalshuntpath

Joule & Peltier terms

Joule heating from all electrical contacts are accounted for.

2007 Year End Module and Subsystem Goals2007 Year End Module and Subsystem Goals

• Thermoelectric Module Design – TE material manufacturing – Thermomechanically robust materials & module design

– Initial module characterization

• Subsystem Design and Optimization – Heat exchanger optimization – Thermal and mechanical integration strategy – Exhaust flow testing facility complete

TE Module Conceptual DesignTE Module Conceptual Design

Identifying and modeling candidate module designs for cost, thermoelectric performance and thermo-mechanical durability • Minimize thermal stresses

pn

• High operating temperature • Good ohmic contact & high thermal conductivity (diffusion barriers, solders/brazes, electrodes)

pn

• Minimize shunt path • No thermo-chemical degradation

common header

pn

Independent hot-side

egg crate

In-line stack

n pn p

Module Design OptimizationModule Design Optimization

Initial Design Study:

• Identify primary module design variables

• Examine effect on primary output variables 1. Power output 2. Cost 3. Thermo-mechanical

durability

Design Space OptimizationDesign Space OptimizationRun Optimized

Training

Scaled Cost

Scaled Power Output

Scal

ed D

urab

ility

) )

Better

Evaluate sensitivities of designs on Pareto frontier

Overlay of TE System Efficiency andOverlay of TE System Efficiency and Expected Exhaust ConditionsExpected Exhaust Conditions

0.12 WoutA Candidate TE Material 18

0.1 234.8 451.6

0.08 668.4 885.2

0.06 1102

1319

15360.041752

19690.02

2186

0350 400 450 500 550 600 650

Exhaust Flow Input Temperature [C]

Exh

aust

Mas

s F l

ow [

kg/s

]

Phase I SummaryPhase I Summary

� quantified electric power requirements for 10% FE improvement

� established $/W as a program metric – automotive and consumer focus; and determined cost target for both exhaust and radiator TE waste heat recovery (materials, modules, subsystems)

� completed the initial exhaust and radiator subsystems design, performance analysis, and cost modeling

� assessed manufacturability of bulk and thin-film modules and subsystems

� validated many known materials’ performance, and enhanced the design of thin-film modules

� explored routes toward higher ZT and low material cost

� exhaust recovery can meet the 350 W requirement with existing materials with high starting purities but at high cost

� radiator recovery alone will not meet the 350 W requirement and is cost prohibitive

Dual-frequency Resonant Phonon Scattering in Skutterudites Dual-frequency Resonant Phonon Scattering in Skutterudites

δ

Filling atom cage

δ

Filling atomcage

Small displacement δ of the filler from it equilibrium x will lead to an increase of the total energy of the system.

Smal

2 31 1( ) ( ) ( ) ( )2 6

E x E x E x E xδ δ δ+ = + + +�� ��� "2

harmonic termharm anharmonic term

In a harmonic approximation, is the spring constant.( )E x��

l displacement δ of the filler from itequilibrium x will lead to an increase of the total energy of the system.

31 1( ) ( ) ( ) ( )2 6

E x E x E x E xδ δ δ+ = + + +�� ��� "

onic term anharmonic term

In a harmonic approximation, is the spring constant.( )E x��

0 ( )E x

m ω =

��

Phonon DOS

ω

P

ωDω

Quasi-localized modes due to void fillingQuasi-localized mhonon DOS

ωD

odes due to void filling

Thermoelectric Enhancement in NanocompositesThermoelectric Enhancement in Nanocomposites

PbTe (room temperature)

J. Martin, G.S. Nolas, W. Zang & L. Chen, APL 90, 222112 (2007)

• Non-conglomerated nanocrystals within the bulk matrix • Power Factor enhancement of 30% at room temperature • Three weeks at 600 K with preservation of nanostructure

Filled Skutterudite NanocompositesFilled Skutterudite Nanocomposites

T (K) 200 300 400 500 600 700 800 900

Pow

er fa

ctor

(μW

/cm

-K2 )

10

15

20

25

30

35

40

45

50

x=0 x=0.05 x=0.1 x=0.2 x=0.3 x=0.4

Yb0.2BaxCo4Sb12

T (K) 200 300 400 500 600 700 800 900

ZT

0.0

.2

.4

.6

.8

1.0

1.2

1.4

1.6 x=0 x=0.05 x=0.1 x=0.2 x=0.3 x=0.4

ZT = 1.36 for x=0.05

HTML at Oak Ridge Natioanl LabHTML at Oak Ridge Natioanl LabORNL efforts will focus on the following areas:

– High temperature transport properties testing – Utilizing ORNL capabilities to continue the search for

high ZT materials – Start mechanical properties testing on selected

materials

– Continue to develop and expand measurement capabilities for TE materials

Environmental Impact of R134a Containing Automobile AC: Global Warming Environmental Impact of R134a Containing Automobile AC: Global Warming� Direct and Indirect Global Warming Impact of R-134a

Direct leakage R-134a: 345 gm/yr * 1,400 = 483 kg/yr

Indirect increase in emissions due to AC R-134a: 23.5 gallons/yr * 8.9 kg CO2/gal = 209 kg/yr

� Total Global Warming Impact of R-134a, and Entire Gas-Powered Vehicle

R-134a: 483 + 209 = 692 kg/yr

Entire Vehicle: avg. 696 gal/yr * 8.9 kgCO2/gal = 6194 kg/yr

* M. S. Bhatti, “A critical look at R-744 and R-134a mobile air-conditioning systems”, SAE SAE-970527

SummarySummary

� Automotive TE waste heat recovery is promising for FE improvement

� $/W or $/Δmpg needs to used to assess its commercial feasibility

� Major challenges in materials and engineering

� Automotive TE cooling needs to be evaluated