1

1

Advanced Combustion and Emission Control (ACEC)

Activities

February 8, 2005

2

FreedomCAR Advanced Combustion and Emission Control (ACEC) Goals

• Reduce petroleum dependence by removing critical technical barriers to the mass commercialization of high-efficiency, emissions-compliant internal combustion engine (ICE) powertrains

• Primary directions− ICE efficiency improvements for cars and light-duty trucks through low-

temperature combustion and minimization of thermal and parasitic losses − Emission control development integrated with combustion strategies for

emissions compliance and minimization of efficiency penalty− Increased efficiency and power density for hydrogen ICE− Coordination with fuels R&D to enable clean, high-efficiency engines using

both hydrocarbon-based fuel (petroleum and non-petroleum) and H2

• 2010 technical goals– Best engine brake thermal efficiency 45%– Powertrain cost $30/kW – NOx & PM emissions Tier II Bin5

2

3

ACEC Goals and Status

2004 Status Goals by Fiscal Year PFI DI 2007 2010 2013 2015

Goals for HC Fuel

ICE peak brake thermal efficiency % 30 41 43 45 46

ICE Powertrain cost (1) $/kW 20 30 35 30 30

Projected vehicle emissions Tier 2 < Bin 10 Bin 10 Bin 5 Bin 5 Bin 5

Emission control fuel economy penalty (2) % <5 <4 <3

Emissions Durability k miles 120 120 120 120 120 120

Goals for H2 Fuel

H2 ICE peak brake thermal efficiency % 38 45 45

H2 ICE Powertrain cost (1) $/kW 45 30

Projected vehicle emissions Tier 2 < Bin 5 Bin 5 Bin 5

(1) High volume production of 500,000 per year

(2) Fuel economy penalty over combined Federal Test Procedure due to emission control relative to diesel vehicle with 2003 emissions

Goals reflect strategy to move from gasoline PFI to diesel/gasoline/H2 DI

4

US Diesel Emission StandardsFuel Sulfur Regulation: current - 500 ppm, future - 15 ppm starting in 2006

0 2.0 4.00.2

0.01

0.1

NOx g/bhp-hr

PM g

/bhp

-hr

Fed. 1998Fed 2004

NOx+HC

Fed 2007-10

Heavy Duty Diesel Truck Emission Standards (>8500 lbs)

CA 2002

CurrentDiesel

P.M. g/ mi

NOx , g/ mi

Bin9

Bin10

EURO 4 Cars

EURO 3 Cars

Bin8

0.08

0.06

0.04

0.02

0

0.01

0.07 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

Bin5

Diesel Tailpipe Emissions Regulations (< 8500 lb)

CurrentDiesel

• By 2009, same standards for all US cars & light-duty trucks • Fleet average of Bin 5, maximum emissions Bin 8• In states with California emissions, only Bin 5 diesels

• US standards are more stringent than current EU standards• HD diesels have emission challenges similar to LD• Paths to Bin 5:

• With existing engine-out emissions, must achieve 90% PM/NOx reduction• By further reducing engine-out emissions, aftertreatment burden reduced

3

5

Research StrategyPNGV FutureTodayEmission Control

Low TempCombustion

Combustion OptimizationReduce Thermal/Mechanical Losses

H2 ICE

Combustion & Emissions HC Fuels

Improved Efficiency

Enablers

Combustion Understanding

Direct InjectionCombustion Optimization

Modeling, kinetics, fundamentals

2003LTC at Light Load

2006Durable Bin 5 System

2010Bin 5 Emissions45% PeakEfficiency

201045% Peak Efficiency

2007Expanded Load Range

2005Sulfur Trap Feasibility

2007Combustion Control Sensors

2004PM Measurement Systems

2015$30/kW

2010$30/kW

2003Bin 5 Performance for Fresh Catalysts

EmissionsFor LTC

6

Combustion• Focus on new processes with potential for improved efficiency

and reduced emissions– Low Temperature Combustion (LTC) strategies such as Homogeneous

Charge Compression Ignition (HCCI).• Barriers

– Poor understanding of LTC fundamentals– LTC operating range limited to low to moderate loads– Control of combustion and heat-release rates for LTC for steady and

transient conditions• Strategy

– LTC to reduce engine-out emissions which reduce requirements for aftertreatment system

– Expand LTC to cover more of the driving schedule – Improve engine efficiency – Study interaction of fuel properties on LTC to determine synergies– Improve understanding of combustion fundamentals and thermodynamic

processes– Synergy with hybrids

4

7

Combustion Research is Focused on Developing the Knowledge-Base Needed to Implement LTC.

Adiabatic flame temperaturein air

CO to CO2conversionstops

1000 1400 1800 2200 2600 30000

1

2

3

4

5

6

Temperature [K]

Equi

vale

nce

ratio

Soot li

mit

Soot

NOX

Diesel combustion– controlled heat release (mixing)– controlled combustion timing– wide load range– high efficiency (relative to Sl)– NOx and PM emissions

Akihama et. al 2001 SAE2001-01-0655

SI combustion (stoichiometric) – controlled heat release

(flame propagation) – controlled combustion timing– wide load range– three-way catalyst– throttling losses– lower efficiency

(relative to diesel)

HCCI

Low-Temperature Combustion– low NOx and particulate emissions– offers diesel-like efficiency – load range?– combustion timing?– heat release rate?– transient control?– robust combustion system?– fuel?

HCCI

8

1000 1400 1800 2200 2600 30000

1

2

3

4

5

6

Temperature [K]

Equi

vale

nce

ratio

Soot li

mit

Soot

NOX

• Importance & Status– New methods for LTC with diesel-like, mixing-controlled heat release were

discovered - one approach uses no EGR.– Since combustion is mixing-controlled, unpredictable heat-release rates in

premixed systems are avoided. • Next steps

– Determine the boundaries & robustness of mixing-controlled LTC – Investigate the feasibility of using in an engine.

Mixing-Controlled LTC Demonstrated in a Combustion Vessel (SNL)

8% 10%

21% O2

15%

- Higher mixing before combustion prevents NOx & soot formation.

- No EGR (21% O2).

- Low-temperature rich combustion prevents soot and NOx formation

- High EGR with less mixing

- Mixing before combustion prevents soot formation.

- EGR used to lower NOx.

Conventional diesel combustion

Axial distance from injector [mm]6040200

180 µm

Image of Flame Location(injector on left of figure)

5

9

• Importance– Late-injection LTC often sacrifices fuel

economy for low emissions.– Previously, kinetics largely thought to

dominate the entire combustion process. • Status

– Variable swirl alters the in-cylinder air motion and can produce counter-rotating vortices

• At optimal conditions, fuel and air meet at interface with high mixing and combustion phasing is improved

– Best fuel economy and lowest soot found between swirl ratios, Rs, of 2.5 and 3.5.

• Next steps– Investigate early injection strategies and

the potential of flow-structure enhanced mixing to extend the speed/load range.

In-Cylinder Flow Structures Promotes Air-Fuel Mixing in Late-Injection Diesel LTC (SNL and UW).

Fuel

Air

Rs = 1.44Non-optimal

Rs = 2.59Optimal

Fuel

Air

In-Cylinder Air Motion During Combustion at Two Swirl Levels

BoreCenter

CylinderWall

10

Achieving and Transitioning to Efficient Low Temperature Combustion (ORNL)

• Accomplishment– Achieved LTC with reduction of

engine-out NOx (90%) and soot (50%) at 20% load.

– Equal engine efficiencies during LTC and normal combustion.

– The transition between modes was several seconds. This was faster than other studies but not sufficient for a production vehicle. The ability to achieve the LTC condition depended on the transition path.

• Importance– The combustion process is not

degraded and HC emissions are not excessive.

• Next Steps– Demonstrate load expansion with

low-pressure EGR and partially premixed charge preparation.

– Demonstrate effects of transition path for expanding LTC operation range.

0

10

20

30

40

Torq

ue (f

t.lb)

0

20

40

60

80

EGR

(%)

0

50

100

150

200

NOx

(ppm

)

0

0.2

0.4

0.6

0.8

0 50 100 150 200 250 300

Time (s)

LII P

M (V

)

Base Efficient LTC BaseBase Efficient LTC BaseBSFC 212 g/hp•hr

Soot

(LII)

EGR

(%)

NO

x (p

pm)

Torq

ue (f

t-lb)

BSFC 212 g/hp•hrBSFC 212 g/hp•hr

1500 rpm, 2.6 bar bmep

6

11

University Consortia Developing LTC Optimization and Control Methods (UW, PSU, UM, UCB, MIT, Stanford)• Importance

– Engine control strategies a key to utilizing LTC– Computer simulation used for engine design and

evaluation of proposed control strategies

• Status– Gasoline HCCI transient behavior and mode changing

demonstrated in engine experiments.– Ignition, combustion and heat transfer submodels

developed and incorporated into GT-power.– Diesel LTC under load transients controlled with

injection and intake valve timing control.– Calculated fuel-spray patterns with variable-spray

angle produce minimal diesel fuel on wall and potential for increased combustion efficiency

• Next steps– Demonstrate strategies on laboratory engines– Extend simulation to full transient vehicle environment

Optimized spray angle in each injection pulse gives improved fuel-air mixing, reduced wall fuel

Star

t of C

ombu

stio

n

Intake Valve Closure (IVC) Timing

Controlled SOC, 3% efficiency change

High load -VVT & inj controlHigh loadLow load

Diesel Spray Pattern

Calculated Load Change for Diesel LTC

( 0 = TDC Compression )

12

Combustion Modeling and Chemical Kinetics are Keys to Understanding the Detailed Combustion

ProcessesThe computational fluid dynamics codes and kinetic data developed by National Labs are relevant to and used by the industry as a whole

1200 rpm, 100kPa, T(BDC): Experimental=409K, Model=407- 408K

356.4

356.6

356.8

357.0

357.2

357.4

357.6

357.8

358.0

0.04 0.08 0.12 0.16 0.2 0.24 0.28 0.32Equivalence Ratio

Cra

nk A

ngle

, 10%

Cum

. HR

[°]

Dec and SjobergModel: Mixture-1Model: Mixture-2Model: Mixture-3

Ignition Performance Predicted with Multi-Component Kinetic for Gasoline (LLNL)

Temperature Distribution in a Multi-Component Evaporating Fuel Spray Calculated by KIVA-4 (LANL)

7

13

0 µs

15 µs

29 µs

44 µs

59 µs

74 µs

88 µs

103 µs

118 µs

133 µs

Spray evolution at 500 bar of injection pressure under 5 bar of N2Colors show mass per unit area

X-Ray Measurements Reveal Structure of Diesel Spray Core (ANL)

Importance• X-ray absorption used for the first quantitative

measurements of the fuel-mass near the spray nozzle as functions of time and position

• Mass and momentum data improves the accuracy of spray models, enhancing their ability to reproduce and ultimately predict the structure of the spray under different combustion regimes

Next Steps• Extend scope to measurements under actual

engine conditions• Collaborate with spray modelers to improve

the accuracy and predictive power of spray simulations using unique near-nozzle data

• Use improved modeling to guide fuel-system development for lower emissions and higher efficiency

Value• These measurements were made with a

unique facility and personnel at ANL

14

Co-Development of Fuels and Advanced Combustion Engines is Essential to Ensure Availability of Fuels Optimized for New Combustion Regimes

Work Currently Underway• Characterize sensitivity of

advanced combustion regimes to variations in fuel properties

HCCI Combustion Phasing Affected by Fuel Properties

356357358359360361362363364365

0.04 0.08 0.12 0.16 0.2 0.24 0.28 0.32Equivalence Ratio [φ]

Cra

nk A

ngle

, 50%

Cum

. HR

Fire 19/1, PRF80Fire 19/1, iso-OctaneFire 19/1, Gasoline

Fuel-chemistry effects isolated

Future Plans• Through parametric analysis and modeling, identify relevance of fuel properties at a

molecular level to advanced combustion regimes

• Identify bulk fuel properties (e.g., an analog to cetane) that characterize fuels optimal for advanced combustion regime operation

8

15

H2 ICE Research• Objective

– Develop high-efficiency, low-emission H2 SI engines which bridge toward the future H2 economy

• Barriers− Low power density for ultra-low emission operation− Unknown combustion stability, pre-ignition and knock

characteristics − Lack of understanding of direct-injection mixing processes− NOx emissions at higher power density

• Strategy− Achieve power density comparable to gasoline PFI with direct

injection, pressure boosting, intercooling− Achieve bin 5 or lower emissions with a combination of lean and

stoichiometric combustion− Develop the fundamental knowledge-base and simulation tools − Other technical areas develop infrastructure and on-board storage

16

Emission Control Enables High-Efficiency, Lean-Burn Engines

• NOx aftertreatment approaches−LNT: lean NOx trap, also called NOx adsorber−Urea SCR: selective catalytic reduction with urea (ammonia)−HC SCR: selective catalytic reduction with hydrocarbons−Plasma: electrical energy creates ions to react with exhaust gases

• Barriers−LNT: fuel economy penalty, degradation by sulfur and high temp−Urea-SCR: infrastructure, enforcement −PM filter: fuel economy penalty from back pressure and effective

regeneration strategies• Strategy

−LNT: improve conversion at lower temperatures, determine degradation mechanisms due to poisons and ways to reverse effect

−Urea: system control to prevent breakthrough & unregulated emissions−PM filter: new regeneration methods−Modeling: improved fundamental understanding and simulations

9

17

Typical Emission Control System

Urea added

SCR/LNTcatalyst

HC, CO control NOx control

OxCat

PMfilter

PM control

NOx sensor NOx sensor

feedbackfeedforward deltaP

Diesel Engine

O2 sensorPM sensor

Controller

HC added

•Separate devices shown for HC, NOx and PM control•Components depend on strategy:

With SCR, urea is addedWith LNT, HC is added in exhaust or by engine

•HC or external source necessary to heat PM filter for regeneration

18

0

10

20

30

40

50

60

70

80

90

100

0 100 200 300 400 500 600 700

Catalyst Temperature (°C)

NOx

Conv

ersi

on (%

)

HC SCR

LNT

Urea SCR

Plasma

US06FTP

Typical Performance Curves of Slightly Aged Catalysts (Ford)• Widest temperature range for Urea SCR• Narrower temperature range for LNT which does not cover both FTP & US06• Plasma and HC SCR have lower NOx conversion• HC SCR conversion too low, but might improve with new materials

10

19

0102030405060708090

100

0.0% 2.0% 4.0% 6.0% 8.0% 10.0%FTP Fuel Economy Penalty

FTP

NO

x C

onve

rsio

n (%

)

LNT + filter

HC SCR + filter

Urea SCR + filter

Plasma + filter

Simulated FTP NOx Conversion vs Fuel-Economy Penalty Assuming 2.5L Engine (Ford)

• Lowest FE penalty for urea SCR, higher FE penalties for other strategies• Energy requirement for urea manufacture approximately 0.2%• FE penalty for LNT at 80% conversion exceeds technical target (<5%)

FE Penalty fromPM filter due toengine backpressure& filter regeneration

20

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.1

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8NOx Emission (g/mi)

PM E

mis

sion

(g/m

i)

Truck Passive NOxCar Passive NOxCar Active NOxTruck Active NOx

bin10

bin5

bin8

bin9

Bin 5-8 emission achieved for cars & trucks with active NOx systems, but durability is not proven

Emissions reduced with combustion improvements

System Testing with Fresh Catalysts Shows Progress to Bin 5Passive systems, which represent current production, do not achieve Bin 5Active systems studied in this program (LNT, urea SCR) are required for Bin 5Results are for fresh catalysts whose performance deteriorates with use

11

21

NOx Adsorber Aged for 1000 Hours (DOE APBF-DEC)

• Initial conversion met NOx target, but became inconsistent during aging• Some desulfations restore NOx conversion (even @ 1000hr), but others do not

• Remaining issues:• NOx conversion during aging frequently less than target• Process to desulfate, mitigation of fuel penalty, other regulated emissions (NMHC)

• Data guide other studies on mechanisms of catalyst deactivation (sulfur poisoning, thermal deactivation) and desulfation

• LNT aging studies are value added for OEM evaluation of this technology

• NOx adsorber for 1.9L engine aged using D15 fuel with 15-ppm sulfur

• Nine desulfations at 150hr, 300hr, then every 100hrs

• Test planned for an additional 1200hrs (120k mi equivalent)

• Test sequence designed for catalyst aging and not emission certification

• Conversion target equivalent to 0.07 g/mi NOx

Average NOx Conversion on FTP Schedule Simulated with Transient Engine DynamometerGaps in the line show desulfation events

40

60

80

100

0

200

400

600

800

1000

NO

x C

onve

rsio

n (%

)

TARGET

Time (hr)

22

Fast-Throughput Methods Used to Discover New, NOx Reducing Catalytic Materials (GM)

• Importance: Selective Catalytic Reduction (SCR) of NOx using HC has the potential to:- Simplify the emissions control strategy - Minimize fuel economy penalty - Avoid Infrastructure & dispensing issues

• Discovery of this new material is key to developing a HC SCR strategy for NOx reduction

• Status: Bench Reactor tests show that high temperature activity could be sufficient, but low temperature activity needs to be improved.

• Example of work with high risk/high potential reward

0102030405060708090

100

100 200 300 400 500 600Temperature(C)

NO

x Co

nver

sion

(%)

HC SCRNew Material

HC SCRExisting Material

NH3 SCR

0102030405060708090

100

100 200 300 400 500 600Temperature(C)

NO

x Co

nver

sion

(%)

HC SCRNew Material

HC SCRExisting Material

NH3 SCR

• Next Steps: Use the high throughput program to search the very large 3-5 component materials space to discover additional new materials that have good low temperature activity and higher high temperature activity. Move from Bench Reactor to engine exhaust testing.

12

23

Soot Regeneration using Microwave Spot Heating- Soot burning is initiated by localized microwave heating of less

than 5% of DPF volume.- Regeneration is less dependent on engine operation mode and is

controlled by the lower substrate and exhaust gas temperature.- Initiate soot regeneration at lower exhaust gas and DPF substrate

temperatures (more typical of diesel 200-250 o C)- Use un-catalyzed, Cordierite substrate - Minimizes fuel efficiency loss

Diesel Particulate Trap Regenerated by Microwave Spot Heating (GM)

Time, min

DPF Temp

o C

Potential of Uncontrolled Regeneration

1000

750

500

250

0

Current ExhaustHeating

Microwave Spot Heating

Time, min

DPF Temp

o C

Potential of Uncontrolled Regeneration

1000

750

500

250

0

Current ExhaustHeating

Microwave Spot Heating

Current Approach: Regeneration by Exhaust Heating- DPF peak temperature and soot burn rate is controlled by

limiting exhaust O2- Requires excess fuel in combustion chamber or in exhaust- Regeneration cycle is dependent on engine operating mode- Potential for uncontrolled temperature increase

Next Steps- Continue laboratory studies using

vehicle- loaded DPF substrates.- Transition to vehicle testing and

system integration.

Predicted MW Heating

Color shows MW field strength

45

7

Front

1 in.

2 in34

56Back

Entire DPF was Regenerated

Sections along length of DPF after heating front 5%

45

7

Front

1 in.

2 in34

56BackBack

Entire DPF was Regenerated

Sections along length of DPF after heating front 5%

24

• Importance– Kinetic maps summarize the response of new LNT materials to a standard test

protocol– The map data is used to calculate LNT performance for exhaust systems– Industry-wide consensus on LNT device maps simplifies supplier testing and

enables more rapid LNT development• Next steps

– Demonstrate and refine measurement protocol and then transfer the protocol to the aftertreatment suppliers

– Apply LNT experience to DPF and SCR* Crosscut Lean Exhaust Emissions Reduction Simulation

– CLEERS is a collaboration between global OEM’s, DOE, EPA, National Laboratories, Universities & TACOM to share non-competitive data and information

Response Maps for LNT Developed by CLEERS* ActivityNOx Concentration after LNTAbsorption 0-15min, Desorption 15-25 min

13

25

Enabling Technologies

Objective:• Develop sensors and measurement methods necessary for

implementation of control strategies

Barriers:• New combustion and emission control systems require new

sensors for system control and on-board diagnostic requirements

Strategy:• Sensors for NOx and particulates

• Particulate measurement during transients to improve engine control

• Sensors to control combustion

26

Particulate Sensor shows Potential for Engine Control (Honeywell, Univ of Minnesota)

• Importance– Goal is to develop a PM sensor that is suitable for closed-loop control of a

diesel engine and/or an on-board diagnostic monitor– The sensor requirements are low cost, robust to harsh environments and

manufacturable in high volumes. • Status

– Prototype sensors were fabricated and tested for more than 100 hours in the exhaust of different engines without degradation

– Sensor output correlates with traditional measurements and showsrequired temporal resolution

• Next steps– Continue testing, develop signal processing algorithms, correlate output to

individual cylinders

0

2

4

6

8

10

0 1 2 3 4 5 6Bosch Smoke NumberSe

nsor

Res

pons

e (v

olts

RM

S)

0

2

4

6

8

10

0 1 2 3 4 5 6Bosch Smoke NumberSe

nsor

Res

pons

e (v

olts

RM

S)

14

27

30 50 70 90 110 130 1500

10

20

30

40

50

Time (seconds)

LII (

ppb)

& R

PM/1

00

0

20

40

60

80

0

Spee

d (m

ph)

Real-Time PM Diagnostic Development (SNL) Real-time measurements of volatile fractionOn-road LII measurements

• Importance- Improved, real-time diagnostics are needed to investigate PM emissions

during transient operation and to meet the upcoming not-to-exceed regulations- Improved PM measurement sensitivity is needed for 2007 and beyond

• Status- Project enabled commercialization of laser-induced incandescence (LII) instrument

by Atrium Technologies, Inc.- Real-time volatile fraction measurement feasibility has been demonstrated

• Next steps- Develop volatile fraction diagnostic to enable possible commercialization- Develop real-time diagnostic for particulate size

Accelerationfrom a stop

340 360 380 400 420 440 460 480 500 520 540 560

0

1

2

3

4

Time (seconds)

Tota

l PM

–So

lid P

M(a

.u.)

0

20

40

60

80

100

120

Solid

Vol

ume

Frac

tion

(%)

EGRLoadRPM

28

CostObjective:

• $30/kW for engine, transmission and emission systems

Barriers: • Diesel engine costs are high relative to gasoline engines due to

added content • Cost optimization occurs during commercialization by an OEM

Strategy:• First, determine technical feasibility of powertrain and

aftertreatment system

• Then, improve the effectiveness of the system to reduce the costusing strategies such as LTC

15

29

Challenges and Needs

• Open Issues– Improvement of engine efficiency – Robust emission system– $30/kW cost goal

• Research Direction– More involvement of industry through vertically integrated

teams (auto, energy, suppliers) and new focused solicitations

– Continue work to improve understanding at national labs and universities

– Focus on new combustion regimes