International Vehicle Emissions Modeling

Design And Measurements

IVE Modeling Goals

• Define low-cost, easy to use methodologies for developing key motor vehicle related data.

• Provide a sophisticated model that is:– Flexible and easy to use.– Adaptable to multiple international locations.– Useful for analyzing policy decisions and vehicle

growth impacts.– Provides a broad range of criteria, toxic, and global

warming pollutant data.

Results• Have developed data collection methodologies to

supplement local data that can be completed in 2-3 weeks using about 12-15 participants.

• Have completed development of a computer based emissions model that allows consideration of local geographic information, fleet technologies, and driving patterns.

• Criteria pollutants, toxic pollutants, and global warming gases can be estimated.

0

5

10

15

20

25

0.7 0.9 1.1 1.3 1.5

Air/Fuel Ratio (Lambda)

Rel

ativ

e E

mis

sio

ns

CO

HC

NOx

Emissions Compared to Air/Fuel Ratio

Stochiometric

Most Torque

Better Fuel Economy

Source: Bosch Automotive Handbook

0

2

4

6

8

10

12

14

16

0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5

Air/Fuel Ratio (Lambda)

Rela

tive E

mis

sio

ns

20 deg

30 deg

40 deg

50 deg

NOx Emissions and Ignition Timing

Highest Torque

Best Fuel Economy

Source: Bosch Automotive Handbook

Catalyst Efficiency verses Catalyst Temperature and Time

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 100 200 300 400 500 600 700 800

Catalyst Temperature (deg C)

Cat

alys

t E

ffic

ienc

y

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 20 40 60 80 100 120 140 160 180 200

Time (se c)

Cat

aly

st E

ffic

ienc

y

Vehicle Emissions Are Dynamic

• Change second by second.

• Depend on– Vehicle technology.– Engine operating parameters.– Temperature, altitude, and humidity.– Fuel characteristics

Keys To Vehicle Emissions

Engine Technology• Ignition Type and Compression Ratios

– Spark ignition, compression ignition, high or low compression ratios.

• Air Fuel Control– Carburetor, Single-Point or Multi-Point Fuel

Injection.– Oxygen Sensor/Computer

• Ignition Control– Mechanical, Computer

Evaporative Control

• Positive Crankcase Ventilation– Vapors from engine oil system collected and

fed to combustion cylinders

• Fuel Tank Vapor Capture– Carbon Canisters that regenerate when

engine is operated– Sealed fuel tank

• Improved Engine Seals

Tailpipe Control• Air pump

– Adds combustion air to tailpipe.

• 2-Way Catalyst (CO/VOC Control)– Catalyst design and active metal content.

• 3-Way Catalyst (CO/VOC/NOx Control)– Catalyst design and active metal content.

• Location of Catalyst near engine– Heats catalyst faster to provide emissions control

sooner

• Pre-capture pollutants during first few minutes of vehicle operation– Captures pollutants until catalyst is hot

Fuel Quality• Eliminate lead (Pb)

– Reduces public lead exposure, Allows catalyst.

• Reduce Sulfur– Reduces public exposure to sulfur oxides and

resulting particulate matter, Increases catalyst efficiency and life.

• Improve combustion qualities of fuel– Modify mixture of hydrocarbons in the fuel, Use

easier to combust fuel (ethanol, natural gas), Pre-mix fuel for 2-stroke to reduce oil in fuel, Add water to diesel fuel to improve combustion (lowers power).

• Lower oxygen demand of the fuel– Add oxygen containing organics such as ethanol.

Local Geographic Parameters

• Temperature

• Altitude

• Humidity

• Road Grade

Emissions Broken Into Two Categories• Start-up emissions

– Excess emissions beyond normal hot-running emissions that occur while engine is warming up.

– Occur typically in the first 200 seconds of vehicle operation.

• Running emissions– Those emissions that occur during vehicle

operations including idle.– There are both running and start-up emissions

during the first 200 seconds of vehicle operation.

0

50000

100000

150000

200000

250000

300000

0 50 100 150 200 250 300 350

Time from Start (seconds)

CO

Em

iss

ion

s (

ug

/se

c)

Example of Start-Up Emissions

Carbon Monoxide

LEV Vehicle

0.00

2000.00

4000.00

6000.00

8000.00

10000.00

12000.00

0 50 100 150 200 250 300

Time from Start (seconds)

NO

(u

g/s

)Example of Start-Up Emissions (cont.)

Nitrogen Oxide

LEV Vehicle

Modeling Start-Up in IVE Model• Established 10 categories of start-up.

– ¼, ½, 1, 2, 3, 4, 6, 8, 12, 18 hours engine off.– Refer to the 10 start-up categories as bins.

• Set start-up emission factor for each vehicle technology.

• Set up a start-up adjustment factor for each vehicle technology for each of the 10 start-up bins.

• Initial start-up factors estimated from Mobile 6 data.

Determination of Vehicle Start-Up Patterns

Measuring Start-Up Patterns

• Survey– Ask drivers to fill out forms about

daily driving habits.– Provide booklet to drivers to fill out

throughout the day.

• Instrumentation

Voltage Based Start-Up Monitor

Designed and built by Global Sustainable Systems Research

Voltage monitored in cigarette lighter.

VOCE Unit records second by second voltage that is used to determine driving times and start-up information.

US/ Chilean Start-Up Patterns

0%

5%

10%

15%

20%

25%

30%

35%

40%

Fra

ctio

n o

f Sta

rt-U

ps

0.06-0.5Hour

0.51-1.0Hour

1.1-2.0 Hour

2.1-3.0Hour

3.1-4.0Hour

4.1-5.0Hour

6.1-8.0Hour

8.1-12.0Hour

12.1-18.0Hour

>18.0Hour

Time Vehicle Shut-Off Before Start

Santiago Overall U.S. Comparison

Chilean Vehicle Starts by HourWeekday

Pre-Start Category Overall

6:00 to 10:00

10:00 to 14:00

14:00 to18:00

18:00 to 22:00

22:00 to 2:00

2:00 to 6:00

U.S. Comparison

0.25 Hour 38% 37% 47% 43% 34% 36% 60% 19%0.5 Hour 10% 5% 18% 13% 9% 5% 0% 17%1 Hour 12% 3% 9% 10% 19% 23% 0% 11%2 Hour 5% 0% 12% 5% 4% 9% 0% 11%3 Hour 7% 2% 2% 15% 4% 14% 0% 6%4 Hour 6% 3% 7% 3% 10% 9% 40% 10%6 Hour 4% 5% 0% 3% 5% 5% 0% 4%8 Hour 11% 29% 2% 8% 11% 0% 0% 4%12 Hour 5% 15% 2% 0% 2% 0% 0% 6%18 Hour 1% 2% 2% 3% 1% 0% 0% 12%Starts 6.3 23.8% 20.6% 14.3% 31.8% 7.9% 1.6% 7.0

WeekendPre-Start Category Overall

6:00 to 10:00

10:00 to 14:00

14:00 to18:00

18:00 to 22:00

22:00 to 2:00

2:00 to 6:00

U.S. Comparison

0.25 Hour 35% 33% 36% 45% 32% 34% 31% 19%0.5 Hour 14% 5% 20% 14% 11% 20% 0% 17%1 Hour 17% 7% 13% 16% 22% 20% 23% 11%2 Hour 4% 0% 3% 4% 5% 6% 8% 11%3 Hour 6% 0% 7% 6% 5% 6% 23% 6%4 Hour 5% 2% 1% 6% 9% 4% 15% 10%6 Hour 3% 7% 1% 0% 4% 4% 0% 4%8 Hour 8% 28% 5% 1% 7% 2% 0% 4%12 Hour 6% 12% 13% 4% 1% 4% 0% 6%18 Hour 3% 7% 1% 3% 3% 0% 0% 12%Starts 6.0 11.7% 25.0% 18.3% 28.4% 13.3% 3.3% 7.0

Distribution of Daily Passenger Vehicle Driving

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

40.0%

Fra

ctio

n o

f D

rivi

ng

06:00-10:00 10:00-14:00 14:00-18:00 18:00-22:00 22:00-02:00 02:00-06:00

Time Range

Weekday Weekend

Sample Running Emissions

0

20

40

60

80

100

120

1000 1050 1100 1150 1200 1250 1300

Time from Start (seconds)

CO

Em

issio

ns (

ug

/s)

0

5

10

15

20

25

30

35

40

45

50

CO Vehicle Speed

Sample Running Emissions (cont.)

0

20

40

60

80

100

120

140

160

180

200

1000 1050 1100 1150 1200 1250 1300

Time from Start (seconds)

NO

Em

iss

ion

s (

ug

/s)

0

5

10

15

20

25

30

35

40

45

50

NO Vehicle Speed

Development of Driving Factors

• Second by second emissions and associated driving data were reviewed to determine best driving factors to estimate emissions.

• Statistical analysis indicates that vehicle power requirement is the best overall predictor of emissions.

Vehicle Power DemandVSP (kW/ton) can be calculated for each second of data using

the following equation [Jimenez-Palacios, 1999]:

VSP = v[1.1a + 9.81 (atan(sin(grade)))+0.132] + 0.000302v3

grade = (ht=0 – ht=-1)/ v (t=-1to0) v = velocity (m/s) a = acceleration (m/s2) h = Altitude (m)

Engine Stress (unitless) = RPMIndex + (0.08 ton/kW)*PreaveragePower

PreaveragePower = Average(VSPt=-5sec to –25 sec) (kW/ton) RPMIndex = Velocityt=0/SpeedDivider (unitless) Minimum RPMIndex = 0.9

Engine stress is related to vehicle power load requirements over the past 20 seconds of operation and engine RPM:

Vehicle Emissions & Power DemandCO2 Emissions Increase Power Curves

0

1

2

3

4

5

6

7

8

0 5 10 15 20

Low Stress

Med Stress

High Stress

CO Emissions Increase Power Curves

0

50

100

150

200

250

0 5 10 15 20

Low Stress

Med Stress

High Stress

VOC Emissions Increase Power Curves

0

5

10

15

20

25

30

35

40

0 5 10 15 20

Low Stress

Med Stress

High Stress

NOx Emissions Increase Power Curves

0

2

4

6

8

10

12

14

16

18

0 5 10 15 20

Low Stress

Med Stress

High Stress

Estimates for the IVE Model• Divided driving patterns into 20 power bins.

– 15 power bins may be sufficient to represent power data.

• Established parameter called vehicle stress.– Weights Implied RPM and Pre-Power and

sums them to produce an index.– Weighting established by statistical analysis

to give the best emission results.– Pre-Power is least important and may

deserve to be dropped.– Three Stress bins were established.

Power Bins for IVE Model

• 60 total bins used for model.

• 3 stress bins and 20 power bins set up to estimate driving impacts on emissions.

• 60 bin process allows alternative binning modifications in future model improvements.

Advantage of Power Binning

• Power statistics easily drawn from measured vehicle speed patterns.

• Road grade and air conditioning loads can easily be included in the binning process.

IVE Model Uses 3 Key Input Files• Location file describes applicable

– Altitude, road grade(optional), temperature, humidity

– Diesel and gasoline fuel quality

– Driving and start patterns and amounts

• Fleet file describes applicable– Distribution of vehicle technologies (2 X 1243)

– Allows Normal Use and Multi-stop vehicles

• Country adjustment file describes applicable– Adjustments to base emission factors

Model Calculation Process

Model Calculation Process Allows

• Single or Multiple times of day

• Single or Multiple locations or links in an urban area

• Use of local emission factors

• Setup of independent adjustment calculations for each technology and pollutant

Location Page• Specifies data pertaining to the

immediate location including– General Information such as average

altitude, type of I/M program if any, fraction of persons using a/c at 80 deg F, fossil fuel characteristics.

– Hourly (or daily) data including driving patterns, average speeds, travel distance or time, temperature, humidity, start patterns, number of starts

Fleet Technologies• The IVE model allows selection of up to 1243

technologies categorized by vehicle type, size, fuel type, age and emissions control technology

• Several Default files are created for the IVE models from the MOBILE6 data

• Data has been collected for Santiago, Chile and Nairobi, Kenya

• Plans are underway to collect data in Kazakhstan, India, and Guatemala

• Data needs to be collected in China, Thailand, Philippines, Indonesia, and Malaysia

Vehicle Technology Classifications

Carburetor NonePre-Chamber

Inject.None

Carburetor / Mixer

None Carburetor NonePre-Chamber

Inject.None Carburetor None 2-Cycle, FI None

Carburetor 2-WayPre-Chamber

Inject.Improved

Carburetor / Mixer

2-Way Carburetor 2-Way Direct Injection Improved Carburetor2-Way /

EGR4-Cycle, Carb None

Carburetor2-Way /

EGRDirect Injection EGR+

Carburetor / Mixer

2-Way / EGR

Carburetor2-Way /

EGRDirect Injection EGR+ Carburetor

3-Way / EGR

4-Cycle, Carb Catalyst

Carburetor 3-Way FI PM Carburetor /

Mixer3-Way Carburetor 3-Way FI PM FI

3-Way / EGR

4-Cycle, FI None

Carburetor3-Way /

EGRFI PM/NOx

Carburetor / Mixer

3-Way / EGR

Carburetor3-Way /

EGRFI PM/NOx 4-Cycle, FI Catalyst

Single-Pt FI none FI EuroI Single-Pt FI 2-Way FI none FI EuroI

Single-Pt FInone / EGR

FI EuroII Single-Pt FI2-Way /

EGRFI 2-Way FI EuroII

Single-Pt FI 2-Way FI EuroIII Single-Pt FI 3-Way FI2-Way /

EGRFI EuroIII

Single-Pt FI2-Way /

EGRFI EuroIV Single-Pt FI

3-Way / EGR

FI 3-Way FI EuroIV

Single-Pt FI 3-Way FI Hybrid Multi-Pt FI 3-Way FI3-Way /

EGRFI EuroV

Single-Pt FI3-Way /

EGRMulti-Pt FI

3-Way / EGR

FI EuroI FI Hybrid

Multi-Pt FI none Multi-Pt FI3-Way /

EGRFI EuroII

Multi-Pt FInone / EGR

ZEV FI EuroIII

Multi-Pt FI 3-Way FI EuroIV

Multi-Pt FI3-Way /

EGRFI EuroV

Multi-Pt FI3-Way /

EGRMulti-Pt FI LEV

Multi-Pt FI ULEV

Multi-Pt FI SULEV

Multi-Pt FI EuroI

Multi-Pt FI EuroII

Multi-Pt FI EuroIII

Multi-Pt FI EuroIV

Multi-Pt FI Hybrid

Gasoline and Ethanol Motorcycles

Heavy Duty Gasoline Vehicles

Light Duty Gasoline Vehicles

Heavy Duty Diesel VehiclesHeavy Duty Vehicles

(Ethanol, Natural Gas, Propane, etc)

Light Duty Diesel VehiclesLight Duty Vehicles

(Ethanol, Natural Gas, Propane, retrofits, etc)

Each Technology Classification

• Has three size groups associated with it.• Has three use groups associated with it.• Thus, there are 9 sub-groups for each

technology classification.• There are also 45 user defined technologies.• Two technology groups are used. One is for

normal vehicles and the second is for multi-stop vehicles.

Load Effects (Road Grade and A/C)

•Road Grade can either be modeled Directly through the use of the driving patterns or a constant road grade for the entire link may be applied in the Location File. Valid grade inputs range from –14 to +14%•The change in Vehicle Specific Power associated with Road Grade and A/C use is applied to the driving corrections. •The fraction of travel in each VSP bin is prorated for each bin.Example: user inputs road grade of 2% and average velocity of 15m/s => +2.9kW/ton VSP increase. Since each VSP bin range is 4.1kW/tons, the

model would move 72% of the fleet up a bin.

Adjustments to Emission Factors

• Local emission measurements can be used to improve calculations.

• Adjustment page provided in model to correct base emission factors.

• 45 undefined technologies provided in model.

Determining Vehicle Fleet Technology Distributions

Registration Data

• California: vehicles registered each year.– California registration database divided by

ZIP code location for fleet by location.

• Problems with approach– ZIP code of registration address may not

represent actual location of vehicle operation.

– No time of day or day of week operations available.

– Often does not indicate vehicle use by age.

On-Road Evaluations

• Observers used to manually record road traffic– Vehicles missed with large number of vehicles.

– Only simple technology evaluations possible.

• Video tape option– Allows freeze-frame to make more accurate count.

– Second zoomed in camera allows improved technology identification.

• Applied to Los Angeles, Santiago, Nairobi

Video Taping of Traffic

Parking Lot Surveys

• Allow improved collection of vehicle parameters.

• Los Angeles studies show good agreement with on-road data.

Parking Lot Survey in Nairobi

Los Angeles Registration, Videotape, Parking Lot

Los Angeles County

0

20

40

60

80

100

Model Year

Perc

ent

DMV

SELEV

Unreg

Model Year Distributions

0.00%

2.00%

4.00%

6.00%

8.00%

10.00%

12.00%

14.00%

Fra

ctio

n/Y

ear

Pre 1980 1980-1989 1990-1992 1993-1995 1996-1998 1999-2001

Model Year Group

Nairobi Santiago US

Santiago Average = 6 yrs

US Average = 10 yrs

Nairobi Average = 12 yrs

General On-Road TechnologiesPassenger Veh. Trucks Buses Motorcycles

Buruburu (Kenya) 86.85% 9.77% 1.83% 1.55%Maipu (Chile) 90.71% 5.40% 3.30% 0.59%

Puente Hills (U.S.) 92.95% 6.02% 0.90% 0.12%

Central Nairobi (Kenya) 90.74% 3.06% 4.46% 1.75%Santiago (Chile) 83.36% 6.76% 8.14% 1.73%Riverside (U.S.) 96.41% 3.00% 0.45% 0.14%

Mathaiga (Kenya) 90.53% 5.28% 3.10% 1.09%Vitacura (Chile) 91.57% 1.91% 4.79% 1.74%

Yorba Linda (U.S.) 96.76% 2.68% 0.40% 0.16%

Overall Los Angeles 95.37% 3.90% 0.59% 0.14%

Overall Nairobi 89.37% 6.03% 3.13% 1.46%Overall Santiago 88.55% 4.69% 5.41% 1.35%

Passenger Vehicle Size Distribution

Location Small Medium LargeBuruburu 57.0% 39.0% 4.0%

Maipu 45.2% 50.3% 4.5%Central Nairobi 68.4% 27.3% 4.3%

Central Santiago 33.8% 59.8% 6.4%Mathaiga 40.4% 40.0% 19.6%Vitacura 23.7% 62.5% 13.8%Overall 44.8% 46.5% 8.8%

Vehicle Use in Santiago

y = -297.95x2 + 16734x

R2 = 0.9746

0

50000

100000

150000

200000

250000

0 5 10 15 20 25 30

Age of Vehicle (years)

Ac

cu

mu

late

d U

se

(K

m)

Vehicle Use in Nairobiy = -531.49x2 + 20494x

R2 = 0.8294

0

50,000

100,000

150,000

200,000

250,000

300,000

0 2 4 6 8 10 12 14 16 18 20

Age of Vehicles (years)

Od

om

eter

(K

ms)

Comparison of Vehicle Use

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

400,000

0 5 10 15 20 25 30

Santiago Nairobi LA

Acc

um

ula

ted

Dri

vin

g (k

m)

US=24,000 km/yr

Santiago=16,000 km/yr

Nairobi=17,000 km/yr

Determining Vehicle Driving Patterns

Objectives

• Develop a methodology for determining power demand on vehicle.– Must know vehicle speed and

acceleration second by second.

– Need to know road grade.

– Need to know air conditioning use.

Global Positioning Satellite Units• Easy to Use• Provide second by second speed, location, and

altitude.• Units combined with microprocessor and flash

memory to store up to one-week of driving data at a time.

• Problems– Loose satellites around tall buildings

– First 3 seconds of acceleration underestimated but then corrects.

Driving Data Collection

Driving Patterns in LA, Nairobi, and Santiago and

Associated Emissions

Driving Data Collection

• In three parts of urban area– Lower income, higher income, commercial

• On three road types– Major highway (freeway), major connecting

roads (arterial), residential streets.

• Each hour from 6am to 9pm

• Using three vehicles and drivers

Results: Velocity Trace for Santiago, Chile

0

20

40

60

80

100

120

0 100 200 300 400 500 600

Time (seconds)

Spe

ed (k

ph)

Freeway

Arterial

Residential

Results: Velocity Trace for Los Angeles, California

0

20

40

60

80

100

120

0 100 200 300 400 500 600

Time (seconds)

Spe

ed (k

ph)

Freeway

Arterial

Residential

Typical Bus Data (Santiago)

0

10

20

30

40

50

60

70

80

90

0 100 200 300 400 500 600

Time (sec)

Sp

eed

(k

ph

)

Comparison of Freeway Driving

0%

10%

20%

30%

40%

50%

60%

70%

Idle Fast Accel Mod Accel Slow Accel Fast Deccel Mod Deccel SlowDeccel

High Cruise Mod Cruise Slow Cruise

Trav

el F

requ

ency

(% T

ime)

Los Angeles, CA

Santiago, Chile

Nairobi, Kenya

Comparison of Arterial Driving

0%

10%

20%

30%

40%

50%

60%

Idle Fast Accel Mod Accel Slow Accel Fast Deccel Mod Deccel Slow Deccel High Cruise Mod Cruise Slow Cruise

Tra

vel F

req

ue

ncy

(%

Tim

e)

Los Angeles, CA

Santiago, Chile

Nairobi, Kenya

Results: Effect of Congestion on FreewaysS

anti

ago

Nai

robi

Los

Ang

eles

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

Power Bin (kW/ton)

Fra

ctio

n o

f D

riv

ing

Freeflow

Moderate Congestion

Heavy Congestion

Low StressModerate Stress

High Stress

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

Power Bin (kW/ton)

Fra

ctio

n o

f D

riv

ing

Freeflow

Moderate Congestion

Heavy Congestion

Low Stress Moderate Stress High Stress

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

Power Bin (kW/ton)

Fra

ctio

n o

f D

riv

ing Freeflow

Moderate Congestion

Heavy Congestion

Low Stress Moderate Stress High Stress

CO Emissions Estimate for LEV

0

1

2

3

4

5

6

7

Freeflow Freeflow Moderate Heavy Freeflow Moderate Heavy

Residential Arterial Freeway

No

rma

lize

d E

mis

sio

ns

Nairobi

Santiago

Los Angeles

Emissions from LA4

cycle

HC Emissions Estimate for LEV

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Freeflow Freeflow Moderate Heavy Freeflow Moderate Heavy

Residential Arterial Freeway

No

rma

lize

d E

mis

sio

ns

Nairobi

Santiago

Los Angeles

Emissions from LA4

cycle

NOx Emissions Estimate for LEV

0.0

0.5

1.0

1.5

2.0

2.5

Freeflow Freeflow Moderate Heavy Freeflow Moderate Heavy

Residential Arterial Freeway

No

rma

lize

d E

mis

sio

ns

Nairobi

Santiago

Los Angeles

Emissions from LA4

cycle

Demonstration of IVE Model

IVE Estimates for SantiagoCarbon Monoxide

0

10

20

30

40

50

60

70

0 2 4 6 8 10 12 14 16 18 20 22

Hour of Day

On

-Ro

ad E

mis

sio

ns

(to

ns/

ho

ur)

Weekday Weekend

Nitrogen Oxides

0

2

4

6

8

10

12

14

0 2 4 6 8 10 12 14 16 18 20 22

Hour of Day

On

-Ro

ad E

mis

sio

ns

(to

ns/

ho

ur)

Weekday Weekend