Fuel Consumption and NOx Trade-offs on a Port-Fuel-Injected SI Gasoline

Engine Equipped with a Lean-NOx Trap

J.A. Lymburner, R.W. McCabe, and J.R. Theis

Ford Motor Company

2009 DEER Conference

Dearborn, MI

August 4, 2009

Contact: rmccabe@ford.com

Focus of Presentation

Three Main Aspects:

1. Effects of Lean-burn, EGR, and deVCT on BSFC

3 combustion features

and BSNOx at 1200 rpm, 4 bar BMEP

2. NOx trap purge fuel requirements at several of the best BSFC/BSNOx points

3. Trade-offs between tailpipe NOx emissions and “effective” BSFC* (and implications for lean-burn engine operation in general)

= Cycle-weighted BSFC for lean operation (NOx storage) *BSFCeffective

and rich operation (NOx trap purge)

Engine: 2005 5.4L 3V V-8* with: • dual equal VCT (deVCT)

• non-production external EGR system (see below)

*Stein, et al., The Combustion System of the Ford 5.4L 3-Valve Engine Proc.

2003 Global Powertrain Congress, Vol. 24, Ann Arbor, MI Sept. 23-25, 2003

What’s new? • lean operation

• EGR system (non-production)

• aftertreatment (TWC+LNT)

But…… • 1200 rpm, 4 bar BMEP only

• 250 combinations of lean AF, cam

retard, and % EGR

250

0

2

4

6

8

10

12

14

16

BS

NO

x (

g/k

W-h

r)

245

255

260

265

270

275

BS

FC

(g

/kW

-hr)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

RM

S C

OV

of

IME

P (

%)

Combustion

becomes

unstable

NOx & BSFC

both too high

NOx

Fuel

Consumption

Combustion

Stability

5.4L PFI V-8

Engine @1200

rpm/4 bar BMEP

lean rich

~8%

Lean PFI has narrow

operating range

compared to stratified

gasoline and diesel

Gasoline lean-burn

can default to stoich

or rich conditions

The Lean

Challenge

14 15 16 17 18 19 20 21 22

AFR

Series1 289.1 270.3 273.4 262.0 262.9 251.4 259.6 253.5

0 0 0 0 0 0 0 0

BSNOx 13.50 7.76 4.90 2.36 13.31 8.14 3.99 4.20

BSFC vs BSNOx comparison for various combustion features (1200 rpm/4 bar)

230

240

250

260

270

280

290

300

BS

FC

(g

/kW

-h)

Base VCT EGR EGR +

VCT Base VCT EGR

EGR +

VCT

Stoichiometric Lean

AF R 14.7 14.7 14.7 14.7 20.1 19.1 20.2 17.1

CAM 0 45 0 40 0 40 0 30

EG R 0 0 15 17 0 0 10 17 COV 0.45 0.73 0.64 1.07 0.83 1.40 1.66 1.61

16

14

12

10

8

6

4

2

0

BS

NO

x (

g/k

W-h

)

Fuel breakdown analysis

St oich

Stoich

w/

deVCT

Stoich

w/ EGR

Stoi ch

w/

EGR

w/

deVCT

Lean

(20:1)

Lea n +

deVCT

17 AFR

30 CAM

16.7EGR

BSNOx (g/kW-h) 19.14 10.34 5 .80 2.93 13.30 11.14 4.28

BSFC (g /kW-h) 287.3 273.0 274.2 2 62.1 261.1 253.2 253.9

% decrease vs. stoich

( measured) 5 .00 4 .58 8.79 9.12 1 1.89 11.64

cal cu lated % decrease in

BSFC due to:

y PMEP 4 .54 1 .89 6.37 1.69 5.88 6.28

y Dilut ion 0 .45 2 .58 2.84 6.90 5.62 5.05

y HC 0 .45 -0 .35 -0.29 -0.66 -0.43 -0.68

y CO -0 .93 0 .35 -0.01 1.34 1.37 1.39

y Comb ustion

Phasing 0 .13 0 .09 0.13 0.08 0.13 0.12

y FMEP 0 .36 0 .02 -0.25 -0.23 -0.67 -0.52

~ 3% BSFC benefit for best lean case vs best stoich case

How lean-burn helps fuel economy

• Decreases pumping losses (i.e. PMEP), but no incremental benefit over VCT + EGR

• Improves fuel conversion efficiency via the

dilution (or “gamma”) effect on the burned gas

expansion process

– = 1-(1/r γ-1) [where γ = C /C , and the greater mass ηf c p v

of gas results in lower burned gas T and Cv and thus

greater γ; also less heat transfer and dissociation]

– Partial overlap with EGR and slight overlap with VCT

• Decreases CO/H2 emissions (lost fuel)

Engine-Out Fuel Consumption & NOx Optimization

RMS COV

AFR CAM EGR BSNOx BSFC IMEP

Stoich at Base Cam 14.66 0 0 13.50 289.1 0.45

Stoich at 45 CAM 14.69 45 0 7.76 270.3 0.73

Stoich + VCT + EGR 14.70 40 17 2.36 262.0 1.07

Lean only 20.05 0 0 13.31 0.83

Lean + VCT 19.12 40 0 8.14 1.40

Lean + EGR 20.19 0 10 3.99

262.9

251.4

259.6 1.66

Lean + EGR + VCT 17.05 30 17 4.20 253.5 1.61

• Best BSFC is obtained with Lean & VCT only.

• Adding EGR greatly reduces NOx but at the expense of

higher BSFC than with best Lean+VCT case.

• Best engine operating condition can only be determined based

on tailpipe emissions and fuel consumption.

Part 2: LNT Impact on NOx

Tailpipe Emissions and BSFCeffective

• NOx tailpipe emissions:

– LNT needs low engine-out NOx to operate

with high efficiency

• BSFCeffective

:

– cycle-weighted BSFC for lean storage and

rich purge

• Trade-off between low BSFC and low TP NOx

Lean-rich duty cycle for LNT operation

0 50 100 150 200 250 300

Time (sec.)

0

50

100

150

200

250

300

350

400

BS

FC

(g

/kW

h)

250 g/kWh during

Lean operation (17:1)

330 g/kWh during

Rich operation (12:1)

17 AFR, 16 EGR, 30 CAM

20 lean 10 rich

The Modern IC Engine with Lean-Burn Aftertreatment System –

a rich integration of chemical, mechanical and controls engineering

Gasoline internal combustion engine

Three-way catalyst NOx

storage/reduction

catalyst (LNT)

optional heat exchanger

sensors & strategy for

control and diagnostics

Lean NO x Traps

Flow

Ceramic substrates (cordierite)

400 cells/sq inch

Washcoat – a thin layer

coated onto substrate

Like TWC, washcoat of LNT

contains - PGM (Pt, Rh, optional Pd)

- Al2O3, ceria for OSC, stabilizers

LNT also contains high levels

of NO x storage materials -Alkaline-earth metals (Ba, Mg)

-Alkali metals (K, Cs, Na)

Lean NOx Traps • Convert HC, CO, & NOx at λ=1

• at λ>1 convert HC & CO, store NOx

• Reduce stored NO x to N2 at λ < 1 • NO x storage performance not

as durable as TWC (800oC max)

Automotive Catalytic Converter

(TWC or LNT)

Effect of NO x Concentration & Flow Rate (SV) – Lab Data

LNT Aged 50 hrs 900 C Max

Eval on 60/5 Test at 450 C N

Ox

Bre

akth

rou

gh

(p

pm

)

1200 25k hr

-1 10k hr

-1 25k hr

-1 10k hr

-1

1000 1500 ppm 1500 ppm 500 ppm 500 ppm

Inlet Concentrations

600

800 Larger LNT

Lower FG NOx

500 or 1500 ppm NO,

10% CO2 & H2O,

5% O2 (lean),

5.0% CO+1.7% H2 (rich)

400 Lower FG NOx & larger LNT

200

0

0 500 1000 1500 2000

Time (sec)

Low feedgas NOx and large LNT volumes will be vital for achieving very low NOx emissions

Engine-Out Fuel Consumption & NOx Optimization

RMS COV

AFR CAM EGR BSNOx BSFC IMEP

Stoich at Base Cam 14.66 0 0 13.50 289.1 0.45

Stoich at 45 CAM 14.69 45 0 7.76 270.3 0.73

Stoich + VCT + EGR 14.70 40 17 2.36 262.0 1.07

Lean only 20.05 0 0 13.31

251.4

262.9 0.83

Lean + VCT 19.12 40 0 8.14 1.40

Lean + EGR 20.19 0 10 3.99 259.6 1.66

Lean + EGR + VCT 17.05 30 17 4.20 253.5 1.61

Tailpipe Fuel Consumption & NOx Optimization Lean Purge (Matching Effective BSFC)

BSFC TP Nox

20 sec lean (effective) (ppm)

Lean + VCT 19.12 40 0 8.69 262.00 523.49

Lean + EGR + VCT 17.05 30 17 3.66 262.44 46.47

30 sec lean

Lean + VCT 19.12 40 0 9.49 257.28 709.52

Lean + EGR + VCT 17.05 30 17 3.99 257.08 68.32

4s R

3s R

3s R 2s R

The Effect of Engine Conditions on NOx Conversion Efficiency

Tailpipe Comparison: Best NOx & Best BSFC Cases

0

500

1000

1500

2000

2500

3000

TP

NO

x (

pp

m)

— 20 lean 4 rich

73.9% Conversion Eff.

BSFC = 262 g/kW-h

19:1 AFR, 40 CAM, 0% EGR

— 20 lean 3 rich

94.9% Conversion Eff.

BSFC = 262.4 g/kW-h

17:1 AFR, 30 CAM, 16% EGR Equivalent BSFC

Lean FG NOx = 2200 ppm Lean FG NOx = 970 ppm

Comparison carried out at equivalent effective BSFC

0 100 200

Time (sec.)

Effect of NO x Concentration & Flow Rate (SV) – Lab Data

LNT Aged 50 hrs 900 C Max

Eval on 60/5 Test at 450 C N

Ox

Bre

akth

rou

gh

(p

pm

)

1200 25k hr

-1 10k hr

-1 25k hr

-1 10k hr

-1

1000 1500 ppm 1500 ppm 500 ppm 500 ppm

Inlet Concentrations

600

800 Larger LNT

Lower FG NOx

500 or 1500 ppm NO,

10% CO2 & H2O,

5% O2 (lean),

5.0% CO+1.7% H2 (rich)

400 Lower FG NOx & larger LNT

200

0

0 500 1000 1500 2000

Time (sec)

Low feedgas NOx and large LNT volumes will be vital for achieving very low NOx emissions

Engine-Out Fuel Consumption & NOx Optimization

RMS COV

AFR CAM EGR BSNOx BSFC IMEP

Stoich at Base Cam 14.66 0 0 13.50 289.1 0.45

Stoich at 45 CAM 14.69 45 0 7.76 270.3 0.73

Stoich + VCT + EGR 14.70 40 17 2.36 262.0 1.07

Lean only 20.05 0 0 13.31

251.4

262.9 0.83

Lean + VCT 19.12 40 0 8.14 1.40

Lean + EGR 20.19 0 10 3.99 259.6 1.66

Lean + EGR + VCT 17.05 30 17 4.20 253.5 1.61

Tailpipe Fuel Consumption & NOx Optimization Lean Purge (Matching Effective BSFC)

BSFC TP Nox

20 sec lean (effective) (ppm)

Lean + VCT 19.12 40 0 8.69 262.00 523.49

Lean + EGR + VCT 17.05 30 17 3.66 262.44 46.47

30 sec lean

Lean + VCT 19.12 40 0 9.49 257.28 709.52

Lean + EGR + VCT 17.05 30 17 3.99 257.08 68.32

4s R

3s R

3s R 2s R

Purge fuel breakdown analysis

• Purge not efficient

• Not optimized (L→R step change)

• Richer purge λ best at higher T (Theis, et

al. SAE 2003-01­1159)

Summary

• Lean-burn improves PFI fuel economy by ~3% relative to best stoichiometric VCT/EGR conditions, when used in combination with VCT&EGR.

• The benefit of lean-burn is largely due to improved fuel combustion efficiency owing to dilution.

• Both VCT and (especially) EGR reduce BSNOX, but the extra fuel required to purge a NOx trap gives back virtually all of the BSFC benefit of lean-burn.

• Successful implementation of lean-burn at low emission standards may require: – Engine-out NOx reduction with VCT and EGR

– Alternative to LNT for NOx control (urea-SCR; LNT+in-situSCR)

– New combustion modes (DI stratified or homogeneous lean, HCCI, PCCI) for ultra-low BSNOx

– Engine operation in a fixed or narrow speed-torque range

Acknowledgments

• Robert Wiley and Shawn Bogedain (cell technicians)

• Tom Leone, Bob Stein and Frank Wong (general advice and help with fuel breakdown analysis)

rmccabe@ford.com