ARB’s Study of Emissions from “Late-model” Diesel and CNG Heavy-duty Transit Buses:

Toxic Compounds and PM Emissions

Alberto Ayala, Norman Kado, Robert Okamoto, and Paul Rieger

Technical Collaborators:

Dr. B. Holmen (UCD), Dr. L. Zafonte, Dr. M. Gebel, H. Porter (CAVTC), K. Stiglitz (CAVTC), F. Gonzalez (CAVTC), P. Kuzmicky (UCD), Reiko Kobayashi (UCD), K. Sahay, G. Gatt, N. Verma,1 C. Maddox, Dr. B. Dharmawardhana, Dr. S. Paulson (UCLA)

Briefing # 2 Content

RESULTS: • Speciation of Gas-Phase Hydrocarbons • Carbonyl Compounds • Organic and Elemental Carbon and Elemental

Composition of PM • Phase distribution of PAH’s • Mutagenicity: Ames Bioassay Analysis • Emission Factors Summary Table

2

Project Update

• Majority of results already in …

• Remaining PAH results expected early ‘02 • Remaining bioassay results expected early ’02 • ELPI results expected Dec ‘01

3

Test Fleet “CNG”

“CNG re-test” “Diesel (OEM)”

“CRT”

Model 2000 DDC Series 50G

1998 DDC Series 50

1998 DDC Series 50

Aftertreatment None OEM Catalyzed

Muffler

CRT

Fuel CNG ECD-1 ECD-1 Odometer 19,629 15,169 15,569

Weight 33,150 lbs 30,510 30,510

• Los Angeles County Metropolitan Transit Authority fleet • 8.5 liter, 4-stroke, turbocharged, 4-cylinder, New Flyer Low 40

passenger transit buses

NOTE: CNG retest fuel sample below specification 4

Toxic Gas-Phase HC’s - Sampling Methodology

Target Analytes - 1,3-Butadiene - Benzene -Toluene -Ethylbenzene -m,p-xylene -o-xylene -Styrene

Tedlar Bag Collection On-site GC-FID’s

5

1,3 Butadiene Vehicle Emission (range of values for multiple tests denoted)

0

2

4

6

8

10

12

Vehi

cle

Emis

sion

(mg/

mile

)

CBD cycle (LOD ~ 0.6)

UDDS cycle (LOD ~ 0.4)

NYC cycle (LOD ~ 2.1)

SS test (LOD ~ 0.1)

CNG Diesel Diesel CNG vehicle OEM w/CRT re-test

Benzene Vehicle Emission (range of values for multiple tests denoted) Total BTEX Vehicle Emission

0

1

2

3

4

5

6

7

8

9

10

Vehi

cle

Emis

sion

(mg/

mile

)

CBD cycle (LOD ~ 0.6)

UDDS cycle (LOD ~ 0.4)

NYC cycle (LOD ~ 2.1)

SS test (LOD ~ 0.1)

CNG Diesel Diesel CNG

(range of values for multiple tests denoted)

0

2

4

6

8

10

12

14

16

18

Vehi

cle

Emis

sion

(mg/

mile

) CBD cycle (LOD ~ 0.6)

UDDS cycle (LOD ~ 0.4)

NYC cycle (LOD ~ 2.1)

SS test (LOD ~ 0.1)

CNG Diesel Diesel CNG

vehicle OEM w/CRT re-test vehicle OEM w/CRT re-test

6

Carbonyl Compounds

Sampling Methodology and Analysis

- Collection on DNPH cartridges - High-precision Liquid Chromatography Analysis

Target Analytes - Formaldehyde - Acetaldehyde - Acetone - Acrolein - Propionaldehyde -Crotonaldehyde - Methyl ethyl ketone - Methacrolein - Butyaldehyde - Benzaldehyde - Valeraldehyde - M-tolualdehyde - Hexanal

Aldehydes Bench

7

Carbonyl Emission for CBD Cycle (range of values for multiple tests denoted)

0

200

400

600

800

1000

1200

Vehi

cle

Emis

sion

(mg/

mile

) Formaldehyde

Acetaldehyde

Total Carbonyls

Carbonyl LOD ~ 0.9

Carbonyl Emission for SS tests CNG Diesel CNG (range of values for multiple tests denoted)

vehicle w/CRT re-test

Veh

icle

Em

issi

on (m

g/m

ile) 100

80

60

40

20

0

Formaldehyde

Acetaldehyde

Total Carbonyls

Carbonyl LOD ~ 0.1

CNG Diesel vehicle w/CRT

e e ein e e e in e de e al n ydn d yd yd d

Hexane

Aceto hy o yy l

Acrol

cro h eh heh htede e de e Kd ald d daal ral l l l

hyl

Crotona a uah

on nz

Valer

Mety

m-TolEt

But eip Blo

thyPr

Me

Additional Carbonyls for CBD Cycle

0

2

4

6

8

10

12

14

16

Vehi

cle

Emis

sion

(mg/

mile

)

CNG vehicle Diesel w/CRT CNG re-test

Carbonyl LOD ~ 0.1

9

EC/OC and Elemental Analysis

EC/OC Procedure • Quartz-Filter Collection of PM

• DRI/IMPROVE Optical/Thermal Analysis

Elemental Analysis • Teflon-Filter Collection of PM

• X-ray Fluorescence

Sampling Probes

Primary Dilution Tunnel

10

11

Average Composition of PM Average Fractions in CNG Exhaust

100%

80%

60%

40%

20%

0% TB IDLE SS CBD NYBC UDDS

EC/TPM OC/TPM Elements/TPM

• OC dominates CNG PM composition across all cycles

• Similar tunnel blank composition

Average Fractions in CNG retest Exhaust

0%

20%

40%

60%

80%

100%

TB SS CBD NYBC UDDS

EC/TPM OC/TPM Elements/TPM

NOTE: TPM=Total PM= EC+OC+Elements

12

Average Composition of PM (cont’d)

• EC/OC fraction in Diesel (OEM) PM shows strong cycle dependence

• OC dominates CRT PM composition across all cycles

Average Fractions in Diesel (OEM) Exhaust

0%

20%

40%

60%

80%

100%

TB IDLE SS NYBC UDDS CBD

EC/TPM OC/TPM Elements/TPM

Average Fractions in CRT Exhaust

0%

20%

40%

60%

80%

100%

TB IDLE SS CBD NYBC UDDS

EC/TPM OC/TPM Elements/TPM

NOTE: TPM=Total PM= EC+OC+Elements

• In general, emissions:

NOTE: Cumulative results per test sequence, not per cycle 13

EC/OC/Elements-Total Emissions

CNG

0

20

40

60

80

100

120

140

TB idle SS TB

CBD

CBD

CBD TB TB

NYBC

NYBC TB

UDDS TB

CBD

g/fil

ter

OC EC TC Elements

CNG retest

0

20

40

60

80

100

120

140

CBD

CBD TB TB

NVBC

NYBC TB TB

UDDS

UDDS

UDDS TB TB SS SS

g/fil

ter

OC EC TC Elements

Diesel (OEM)

0

50

100

150

200

250

300

350

400

TB

IDLE

IDLE SS SS

NYBC TB

UDDS

UDDS TB

CBD

CBD TB

g/fil

ter

OC EC

CRT

g/fil

ter

Diesel (OEM) > CNG > CRT TB ~ Idle < other cycles

TC Elements

0

10 20

30

40 50

60

70

80 90

100

CBD

CBD TB SS SS TB

NYBC

NYBC TB

UDDS IDLE TB

UDDS

UDDS TB

OC EC TC Elements

• EC/OC/Elements fractions show cycle dependence and variability

m

umum se

on n c ne l us n r e u no Tiu k Zifi

anga

ne

hosp

horr ci i ln or om cIc SuSiliNii alm l

ChC hrlu CA M P

Elemental Analysis Results Tunnel Blank

CNG

0

300

600

900

1200

1500

Aluminu

mCalc

ium

Chlorin

e Chro

mium

Coppe

r Cob

alt

Nickel

Manga

nese

Iron

Phos

phoru

s Po

tassiu

mSilic

on

Sulfur

Zinc

ng/fi

lter

TB3(UDDSX2) SS(UDDSX2) CBDX2 CBDX2 CDBX2 CBDX4 TB1(CBDX2) TB2(CBDX2)

CNGS Ca Cu Fe K Al

• Ca, Cl, P, Zn, S are oil components • Fe from engine wear • Si source unknown • Si emissions: Diesel (OEM) >> CNG ~ CRT • In general, TB << SS and CBD

NOTE: Cumulative results per test sequence, not per cycle

14

Diesel (OEM)

8000

9000

SS(UDDSX1) SS(UDDSX1) TB1(UDDSX1) CBDX2 CBDX2 TB(CBDX2)

Diesel (OEM)

Al Si Zn Fe

CNG retest

0

300

600

900

1200

1500

Aluminu

m Calc

ium

Chromium

Chlo

rine

Cobalt

Iron

Phos

phoru

s

Silicon

Sulfur Zinc

ng/fi

lter

SS1 SS2 CBDX4 CBDX4 TB(CDBX4) TB(CBDX4)

CNG retest

S

CRT

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

Aluminu

m

Chlorin

e

Iron

Silicon

Sulfur Zinc

ng/fi

lter

SS1(UDDSX4)

SS2(UDDSX4)

TB1(UDDSX2)

CBDX8

CBDX8

CRT

Si

7000

6000

5000

4000

3000

2000

1000

0

ng/fi

lter

Polycyclic Aromatic Hydrocarbons

Particle Associated OEHHA Unit risk for PAH's cancer by inhalation

per million (ug/m3)E-1

Benz[a]anthracene Chrysene 11

Benzo[b]fluoranthene 110 Benzo[k]fluoranthene 110

Benzo[a]pyrene 1100 Dibenz[ah]anthracene 1200

Expected PAH phase distribution in ambient and CARB diesel exhaust samples

TARG ET PAHS

Partic le Assciated PAHs Benzo[ghi]perylene D ibenz[ah]anthracene Indeno[1,2,3-cd]pryene Perylene Benzo[a]pyrene Benzo[e]pyrene Benzo[k]fluoranthene Benzo[b]fluoranthene Chrysene Benz[a]Anthracene

Sem i-Volatile PAHs Pyrene Fluoranthene M ethyl Phenanthrene Anthracene Phenanthrene Fluorene

Volatile PAHs Dim ethyl naphthalene Acenaphthene Acenaphthylene D im ethyl naphthalene B iphenyl 1-m ethyl naphthalene 2-m ethyl naphthalene Naphthalene

FILTER

PUF

XA D

15

PAH ANALYSIS

GC/MS PUF

GC/MS FIL

FIL PUF XAD

BIOASSAY SAMPLES RANGE FINDING

GC/MS AN

POOL & CON

GC/MS AN

EXTRACT DCM

GC/MS AN

POOL

EXTRACT TOLUENE ECD-WO

FILTER IS

GC/MS AN

EXTRACT ACETONE

PUF IS

GC/MS AN

EXTRACT DCM

XAD IS

PAH SAMPLES

16

*All results not corrected for tunnel blanks and XAD values corrected for background contamination

Partic le As s o c iated PAHs -CBD and S S Cyc les *

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

B(a)

A

Chry

B(b)

F

B(k)

F

B(e)

P

B(a)

P

Per

In(1

,2,3

-cd

)P

DiB(

a ,h)

A

B(g,

h,i)P

ug/m

ile

FILTER CNG-CBD FIL CNG-S S FILTER D OEM-CBD FIL D OEM-S S FILTER D W/CRT-CBD FIL D W/CRT-S S FILTER CNG RT-CBD

Filter and PUF PAHs -CBD and S S Cyc le s *

0.0

5.0

10.0

15.0

20.0

25.0

Phen

1-M

e-Ph

en

Fuor

an Pyr

ug/m

ileCNG-CBD TOTAL CNG-S S F+P D OEM-CBD F+P D OEM-S S F+P D W/CRT-CBD F+P ECD-WT D W/CRT CNG-RT-CBD F+P

PUF and XAD PAHs -CBD and S S Cyc le s *

0.0

100.0

200.0

300.0

400.0

500.0

600.0

Na p

(T)

2-M

e-N

ap

1-M

e-N

ap

ug/ m

ile

CNG-CBD PUF+XAD CNG-SS PUF+XAD D OEM-CBD PUF+XAD D OEM-SS PUF+XAD

D w /CRT-CBD PUF+XAD D w /CRT-SS PUF+XAD CNG R-CBD PUR+XAD

17

P AH

PAH

PAH

18

•CBD and SS Results PAHs in PM –Diesel (OEM)-Most PAHs Detected –CNG CBD - Most PAHs m.w. 252 Not Detected except for BaP –CNG SS- All PAHs m.w. 252 Not Detected –CRT- CBD and SS Only Benz[a]anthrancene and Chrysene Detected

•CBD and SS Semi-volatile PAHs –Diesel (OEM) Generally the Highest Levels –CNG Similar Levels to Diesel OEM –CRT Lowest Levels

•CBD and SS Volatile PAHs –At Similar Levels

•Fluoranthene and Pyrene Phase Distribution –CBD Diesel(OEM)-Primarily in Filter –SS Distributed more evenly between the Filter and PUF –CRT and CNG-Primarily in PUF

P ha s e Dis tributio n o f F lo ura nthe ne a nd P yrene in S S

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

Phen

1-M

e-Ph

en

P AH CNG FIL CNG P UF D OEM FIL D OEM P UF D w/CRT FIL D w/CRT P UF

Phas e Dis tribution o f Fluo ranthe ne and Pyre ne in CBD

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

Fuo

r an

Py

r

P AH

CN G FILTER CNG PUF D OEM FILTER D OEMPUF D w/ CRT FILTER D w/ CRT PUF CNG RT FILTER CNG RT PUF

Bioassay Analysis

Procedure • Collection of PM on Filter

• Collection of vapor-phase on PUF

• Solvent Extraction

•Salmonella/Microsuspension procedure

• TA98 and TA 100 Tester Strains with and w/o +S9 Metabolic Enzymes

192-stage Sampler

CVS Tunnel

ECD CNG.2 ECD

ECD CNG.2 ECD

Mutagenicity Results

CBD Emissions

0 2 4 6 8

10 12 14

CNG.1 CRT

Vehicle Type

Mut

agen

Em

issi

ons

(TA

98 R

ev/m

ile) x

105

(+S9)

(-S9)

Steady State Emissions

0 1 2 3 4 5 6 7

CNG CRT

Vehicle Type

Mut

agen

Em

issi

ons

(TA

98 R

ev/m

ile) x

105

(+S9)

(-S9)

CNG retest

Diesel (OEM)

Diesel (OEM)

CBD Specific Acivity (-S9)

0

20

40

60

80

100

CNG.1 CRT

Vehicle Type

Spec

ific

Mut

agen

icA

ctiv

ity (R

ev/u

g)

Sample

TBlank

Steady State Specific Activity (-S9)

0

20

40

60

80

100

CNG CRT

Vehicle Type

Spec

ific

Mut

agen

icA

ctiv

ity (

Rev/

ug)

Sample

TBlank

Diesel (OEM)

Diesel (OEM)

CNG retest

20

Emission Factor Summary CBD

CNG emission

rate mg/mi

CRT emission

rate mg/mi

Diesel(OEM) emission

rate mg/mi

OEHHA Unit risk for cancer by inhalation

per million (ug/m3)E-1

Total PM Butadiene Benzene

Formaldehyde Acetaldehyde

40.0 3.5 3.2

780.0 90.0

14.0 0.0 0.6 0.0 0.0

119.0 0.0 1.6 ** **

300* 170 29 6

2.7

110

Total PAH's heavy

semi-vol volatile

0.58 0.0004

0.04 0.54

0.78 0.0001

0.02 0.77

0.80 0.0038

0.05 0.75

B[a]a 9.0E-05 4.0E-05 2.7E-04 Chr 2.3E-04 5.0E-05 3.7E-04 11 B[b]f 0 0 4.1E-04 110 B[k]f 0 0 2.0E-04 110 B[a]p *** *** 4.5E-04 1100

Dib[ah]a 0 0 3.5E-04 1200 Spec.Mut. Activity****

((TA98 Rev/mi)X10E5) 9.6/6 1/2.5 2/1.2

* For diesel total PM only from ARB TAC document ** Data not available *** Emission factor under development ****CBD/Steady State

21

Challenges

• Test Low-emission Technologies: CNG and CRT – Current methodologies needed improvement or required modification

• Emissions Testing – Large Test Matrix – Extremely low levels of PM requiring large number of samples to be

collected – Interferences from the Tunnel Background

• Analytical – Low levels of PM and gaseous emissions require lower detection limits

• Logistics – Short Turn-Around Time – Limited Resources

22

Meeting the Challenge

• Developed ARB in-house Expertise • Modified existing sample collection and analysis

methods: – For PAH’s, 1) concentrated samples, 2) required

different solvent to extract samples, and 3) developed a faster method to extract XAD

• Required Pooling of Samples to Obtain sufficient Sample for analysis

23

What Was Learned From This Study?

• Not Routine…Research Effort • Tunnel Background Interferences

– To evaluate accurately low-emission technologies will require new sampling methodologies including different types of dilution tunnels to address tunnel background interferences

• Changes to Sampling Methodology – Need samplers that can collect greater PM mass – Apply what was learned from this study to improve on future

studies

• Analytical Techniques – Methodology for particle number/size characterization is needed – For PAH’s, may need to develop additional clean-up procedures to

eliminate interferences – For PAH’s, automate sample preparation – Develop techniques that lower the Method Detection Limit

24

Final Remarks

Toxic Hydrocarbons and Carbonyl Compounds

• Butadiene was only detected in CNG vehicle exhaust (with 1 exception: Diesel without trap idle test).

• Generally, BTEX concentrations in CVS exhaust samples were close to ambient levels

• Generally, BTEX emission follows the order: CNG > Diesel (OEM) > CRT

• Carbonyl emissions from CNG vehicle were much higher than from CRT-equipped vehicle

• Total carbonyl emissions (by mass) from CNG vehicles are two orders of magnitude higher than BTEX and 1,3 Butadiene emissions

• CNG vehicle carbonyl emissions are dominated (>80%) by formaldehyde

25

Final Remarks (cont’d) Composition of PM • OC dominates CNG PM composition across all cycles • Similar tunnel blank composition • EC/OC fraction in Diesel (EOM) PM shows strong cycle dependence • OC dominates CRT PM composition across all cycles • EC/OC/Elements fractions show cycle dependence and variability • Emission of EC/OC/Elements: Diesel (OEM) > CNG > CRT • Ca, Cl, P, Zn, S are oil components • Fe from engine wear • Si source unknown • Si emissions: Diesel (OEM) >> CNG ~ CRT

26

Final Remarks (cont’d)

PAH’s and Bioassay • Emission rates (ug/mi) for most PAH’s were higher in the CBD than SS • Emission rates for CNG retest were generally higher than CNG • Differences were observed in the properties of PM from CNG, Diesel

(OEM), and CRT • New substrate is needed to replace XAD because of backgrounds of some

PAH’s approach the low PAH levels in tunnel blank and samples • CRT PAH levels are similar levels to TB’s • Generally, CNG and Diesel (OEM) are higher than TB’s • Emissions of mutagenic compounds showed cycle dependence • For CBD, bioassay follows: CNG > Diesel (OEM) > CRT • For SS, bioassay follows: CNG > CRT > Diesel (OEM)

27

“Bottom Line” and Next Step • Thus far, the cumulative results in this study suggest that the “adverse” impact from emissions follows:

Total PM Mass: Diesel(OEM)>CNG>CRT

Secondary PM: CRT>Diesel (OEM)> CNG

Particle Size/Counts: CNG>Diesel(OEM)>CRT ?Toxicity/Mutagenicity: CNG>CRT~Diesel(OEM)

• Results show cycle dependence

• PM is not PM and is not PM !

What is next?

• SCAQMD Briefing

• Finish data analysis and subject it to peer review

• Report back once publications accepted 28

IDEAS FOR FOLLOW-UP STUDY: Investigation of Toxics from Interim and “2007” Heavy-Duty

Emission Controls

Scope Cuts Across Applications:

• Baseline

• CRT and SCRT on bus or truck (in-house)

• DPX and DPX/EGR on bus or truck (in-house)

• SCR/DPF engine tests (piggy back on current work: DOE or EPA)

• NOx Adsorber/DPF engine tests (piggy back on current work: DOE or EPA)

• CNG/Catalyst, “Advanced” CNG, CNG/H2, or CNG/Trap?? bus (in-house)

• Other? (i.e., fuel-borne cat, active PM trap)

Approach:

• Duplicate samples, one fuel, one oil, one or two duty cycles 29

FOLLOW-UP STUDY (cont’d)

Emissions of Concern:

• DPF Ash

• Regulated Emissions

• Elements

• Toxic VOC’s and Carbonyls

• PAH’s and Nitro-PAH’s

• Bioassay Analysis

• Characterization of Particle Number and Size

• Dioxins

• Other? (HONO, N2O, etc.)

Participants:

• Lead Team: RD, SSD, MSCD

• In-house Testing: RD, MSCD

• Analysis: UCD, MLD, NREL, EPA, Private Lab

• Data Reduction: RD, SSD, MSCD

30