comparison of ambient measurements to emissions ... · comparison of ambient measurements to...

Comparison of Ambient Measurements to Emissions Representations in Modeling

Presented by:Lyle R. ChinkinStephen B. Reid

Sonoma Technology, Inc.Petaluma, CA

Presented to:The CCOS Technical Committee

Sacramento, CADecember 15, 2006

905044.12-2838

2

To provide corroborative evidence, with

sufficient justification, that can potentially

explain disagreements between emission

inventory and observed pollutant

concentrations.

Project Objective

3

Preview of Findings (1 of 3)

� Overall, the emissions data show better agreement with ambient data than previous emission inventories have. 0

1

2

3

4

5

6

7

8

Summer 1996 Summer 2000

TN

MO

C/N

Ox

rati

o

0

2

4

6

8

10

12

14

16

18


CO

/NO

x ra

tio

Ambient

Emissions

Sacramento Del Paso

0

1

2

3

4

5

6


TN

MO

C/N

Ox

rati

o

0

2

4

6

8

10

12

14

16

18

20


CO

/NO

x ra

tio

Ambient

Emissions

Fresno First Street

0

1

2

3

4

5

6

7

8

Summer 1996 Summer 2000T

NM

OC

/NO

x ra

tio

Ambient

Emissions

Folsom

0

1

2

3

4

5

6

7

8


TN

MO

C/N

Ox

rati

o

Clovis

2.6 - 4.32.93.9--Kern

1.4 - 7.21.92.63.6Fresno

1.4 - 2.41.71.62.2Sacramento

STI 2000*DRI 2000DRI 1995DRI 1990

Ambient ratio/Emissions ratioAir Basin

*This column shows the range of results from all sites evaluated in each air basin—including both urban and rural sites.

4


� At some sites, the emissions data correlate with ambient data as closely as could be expected given analyses limitations*.

0

2

4

6

8

10

12

14

16

18

20

All Days Weekday Weekend

Day of Week

CO

/NO

x R

atio

Ambient - Median

EI - With ElevatedSources

EI - Low Level Only

Del Paso Manor Site

* “ARB staff believes that an assessment such as this should only be expected to produce ambient/emissions ratios that are within approximately +/- 25 to 50% of 1.0.” (ARB, 1997)

5


� EI generally under-predicts pollutant ratios

� Urbanized Sacramento area:

• Good agreement on weekdays

• Poorer agreement on weekends

� Urbanized Fresno area:

• Good agreement on weekdays and weekends

� Urbanized Bakersfield area:

• Poor agreement on weekdays and weekends

� Rural sites don’t fully meet underlying assumptions of analysis techniques

6

Preview of Recommendations

� Improve accuracy of weekend emission estimates.

� Correct spatial distributions of emissions (e.g., from livestock waste and heavy-duty engines).

� Further investigate the poor agreement in Kern County.

� Collect more ambient data at Bay Area sites to enable more robust evaluations.

7

Outline

� Overview of Approach

� Sites Selected

� Results

� Findings and Recommendations

� Questions & Discussion

8

Overview of Approach (1 of 2)

� Comparisons of ambient data to the emission inventory include:

• TNMOC-to-NOx ratios

• CO-to-NOx ratios

• Ratios of individual species

• Chemical composition of hydrocarbons

9

Overview of Approach (2 of 2)

� Spatial and temporal comparisons done by:

• Weekday vs. weekend

• Wind quadrants

Wind Quadrant 1 (1-90°) Wind Quadrant 2 (91-180°) Win d Quadrant 3 (181-270°) Wind Quadrant 4 (271-360°)Wind Quadrant 1 (1-90°) Wind Quadrant 2 (91-180°) Win d Quadrant 3 (181-270°) Wind Quadrant 4 (271-360°)

10

Monitoring Sites Selected

XGranite Bay Stn.Sacto4GNBY

XBakersfield Stn. (California Ave.)SJV4BAC

XTurlock Stn.SJV3TSM

XSan Jose Stn. (4th St.)Bay Area3SJ4

XXXShafter Stn.SJV3SHA

XXXMadera Stn.SJV3M29

XXElk Grove Stn.Sacto3ELK

XXXXArvin Stn.SJV3ARV

XXXParlier Stn.SJV2PLR

XXXFolsom Stn.Sacto2FLN

XSunol Stn.Bay Area1SUN

XXXSacramento Stn. (Del Paso Manor)Sacto1SDP

XXXXSacramento/Natomas Stn.Sacto1NAT

XXFresno Stn. (First St.)SJV1FSF

XXXXClovis Stn.SJV1CLO

XXXXBakersfield Stn. (Golden State)SJV1BGS

VOC fingerprints

Species ratiosCO/NOxTNMOC/NOxSite NameDistrictTierSite

11

Comparison of TNMOC/NOx Ratios

Elk Grove (1.4)

Folsom (1.5)

Sacramento – Natomas (1.2)

Sacramento – Del Paso (1.4 - 2.3)

Clovis (1.3 - 2.1)

Fresno – First St. (1.3)

Madera (6.3) [rural]

Parlier (1.3 - 2.5) [rural]

Arvin (1.5) [rural]

Bakersfield – Golden St. (3.3)

Shafter (2.6) [rural]

*Numbers represent the ratio of the derived median ambient ratio to the emission inventory ratio

12

Comparison of CO/NOx Ratios

Bakersfield – Golden St. (2.3)

Bakersfield – California Ave (3.9)

Clovis (2.2)

Fresno – First St. (1.7)

Sacramento – Natomas (1.6)

Sacramento – Del Paso (1.1)

Turlock (2.4)

San Jose (1.6)

*Numbers represent the ratio of the derived median ambient ratio to the emission inventory ratio

13

0

2

4

6

8

10

12

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

0

2

4

6

8

10

12


Day of Week

TNM

OC

/NO

x R

atio Ambient


EI - Low Levels Only

Elk Grove

0

1

2

3

4

5

6

7

8

9

10


Day of Week

TNM

OC

/NO

x R

atio

0

1

2

3

4

5

6

7

8

9

10

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

Folsom

Sacramento Area (1 of 3)

0

1

2

3

4

5

6

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

0

1

2

3

4

5

6


Day of Week

TNM

OC

/NO

x R

atio

Sacramento - Natomas

0

2

4

6

8

10

12

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

0

2

4

6

8

10

12


Day of Week

TNM

OC

/NO

x R

atio

Sacramento – Del Paso Manor

� Better agreement on weekdays

� Note poor agreement in wind quadrant 3 at Del Paso Manor

TNMOC/NOx

Error bars represent 25th and 75th percentiles

14

Sacramento (2 of 3)

� Poorest agreement in wind quadrant 3

� Shopping center 1km southwest of site

� Possible issue capturing hot soak emissions in inventory

0

2

4

6

8

10

12

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atioAmbient

EI - W ith ElevatedSources


Del Paso Manor

15

Sacramento Area (3 of 3)

0

5

10

15

20

25

1 2 3 4

Wind Quadrant

CO

/NO

x R

atio

0

5

10

15

20

25


Day of Week

CO

/NO

x R

atio Am bient - Median


EI - Low Level Only

Natomas Site

0

2

4

6

8

10

12

14

16

18

20

1 2 3 4

Wind Quadrant

CO

/NO

x R

atio

0

2

4

6

8

10

12

14

16

18

20


Day of Week

CO

/NO

x R

atio

Ambient - Median


EI - Low Level Only

Del Paso Manor Site

�Weekdays and weekends are similar

�Agricultural or heavy duty sources appear to be misrepresented (too high) in Natomas quadrant 3

�Del Paso ratios agree well

CO/NOx

16

0

1

2

3

4

5

6

7

8

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

0

1

2

3

4

5

6

7

8


Day of Week

TNM

OC

/NO

x R

atio

Fresno First Street

Fresno Area (1 of 3)

0

1

2

3

4

5

6

7

8

9

10

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

0

1

2

3

4

5

6

7

8

9

10


Day of Week

TNM

OC

/NO

x R

atio Ambient



Clovis0

2

4

6

8

10

12

14

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

0

2

4

6

8

10

12

14


Day of Week

TNM

OC

/NO

x R

atio

Madera

0

1

2

3

4

5

6

7

8

9

10

1 2 3 4

Wind Quadrant

TN

MO

C/N

Ox

Ra

tio

0

1

2

3

4

5

6

7

8

9

10


Day of Week

TNM

OC

/NO

x R

atio

Parlier

�Weekdays and weekends are similar

� Changes in real world growth between Clovis and Fresno compared to surrogates (CLO-quadrant 3)

� Rural Madera site with low emission density (doesn’t fully meet assumptions)

TNMOC/NOx

17


0

1

2

3

4

5

6

7

8

9

10

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

Ambient



�Agrees well except in quadrant 3

�Winery in quadrant 3 is not identified as a point source in the inventory

Parlier TNMOC/NOx

18


0

5

10

15

20

25

30

1 2 3 4

Wind Quadrant

CO

/NO

x R

atio

0

5

10

15

20

25

30


Day of Week

CO

/NO

x R

atio

Ambient - Median


EI - Low Level Only

Clovis

0

5

10

15

20

25

1 2 3 4

Wind Quadrant

CO

/NO

x R

atio

0

5

10

15

20

25


Day of Week

CO

/NO

x R

atio

Ambient - Median


EI - Low Level Only

Fresno First Street

� Similar agreement on weekdays and weekend days

�Agreements not as good as in the Sacramento area

�May suggest overestimated heavy-duty NOxemissions in the area

CO/NOx

19

Kern County (1 of 2)

0

1

2

3

4

5

6

7

8

9

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

0

1

2

3

4

5

6

7

8

9


Day of Week

TNM

OC

/NO

x R

atio

Ambient



0

5

10

15

20

25

30

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

0

5

10

15

20

25

30

35


Day of Week

TNM

OC

/NO

x R

atio

Ambient



0

1

2

3

4

5

6

7

8

9

1 2 3 4

Wind Quadrant

TNM

OC

/NO

x R

atio

0

1

2

3

4

5

6

7

8

9


Day of Week

TNM

OC

/NO

x R

atio Ambient



Bakersfield – Golden State

Arvin

Shafter

�Ambient ratios are 2 to 4 times higher than emissions ratio (Much was worse than Sacramento/Fresno)

�Agreement does not vary between weekdays and weekends

� TNMOC emissions in quadrant 2 of Arvin are dominated by biogenics

TNMOC/NOx

20

Kern County (2 of 2)

0

2

4

6

8

10

12

14

16

1 2 3 4

Wind Quadrant

CO

/NO

x R

atio

0

2

4

6

8

10

12

14

16


Day of Week

CO

/NO

x R

atio

Ambient - Median


EI - Low Level Only

0

5

10

15

20

25

30

1 2 3 4

Wind Quadrant

CO

/NO

x R

atio

0

5

10

15

20

25

30


Day of Week

CO

/NO

x R

atio

Ambient - Median


EI - Low Level Only

Bakersfield – California Ave.

Bakersfield – Golden State

�At Calif. Ave., best agreement in wind quadrants 3 and 4

�Poor overall agreement at the Golden State site

CO/NOx

21

Individual Species Ratios

� Compared amounts of individual hydrocarbons in the ambient data and emission inventory

• Acetylene/benzene

• Acetylene/propylene

• Benzene/m- and p-xylene

• Benzene/o-xylene

• Benzene/toluene

• Toluene/m- and p-xlyene

• Toluene/o-xylene

� Emission ratios generally show good agreement with ambient ratios

22

Fingerprint Comparisons (1 of 5)

� Comparisons were performed for 10 sites

� Analyses showed:

• Speciation of emission inventory is generally representative of the TNMOC composition detected by ambient monitoring sites

• Ethane is consistently higher in the emission inventory

• Propane is consistently lower in the emission inventory

• Isoprene is consistently higher in the emission inventory

23

Fingerprint C

ompariso

ns (2

of 5

)

0% 5%10%15%20%25%30%35%40%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

Am

bient

EI - W

ithE

levated Sources

EI - Low

LevelO

nly

M29

0% 5%

10%

15%

20%

25%

30%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

PL

R

0% 5%10%15%20%

25%30%35%40%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

Am

bient

EI - W

ithE

levated Sources

EI - Low

LevelO

nly

0% 5%

10%

15%

20%

25%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

Clovis

Madera

Parlie

rBakersfie

ld –Golden State

Ethane: liv

esto

ck emissio

ns sp

atially allocated usin

g populationdensity

.

24

Fingerprint C

ompariso

ns (3

of 5

)

0% 5%

10%

15%

20%

25%

30%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

Am

bient

EI - W

ithE

levated Sources

EI - Low

LevelO

nly

M29

0% 5%

10%

15%

20%

25%

30%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

PL

R

0% 5%10%15%20%

25%30%35%40%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

Am

bient

EI - W

ithE

levated Sources

EI - Low

LevelO

nly

0% 5%

10%

15%

20%

25%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

Elk Grove

Madera

Shafte

rParlie

r

Generally Propane is a

ssocia

ted with oil a

nd gas e

xtractio

n, except fo

r major

distrib

ution ce

nter lo

cated at E

lk Grove.

25

Fingerprint C

ompariso

ns (4

of 5

)

FL

N

0% 5%

10%

15%

20%

25%

30%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

Am

bient

EI - W

ithE

levated Sources

EI - Low

LevelO

nly

0% 5%

10%

15%

20%

25%

30%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

SD

P

0% 5%

10%

15%

20%

25%

30%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

Am

bient

EI - W

ithE

levated Sources

EI - Low

LevelO

nly

0%

10%

20%

30%

40%

50%

60%

70%

EthyleneBenzene

C6+C4+C8+

StyrenesEthane

PropaneButanes

PentanesPropylene

TolueneAcetylene

IsopreneC6-C11

TNMOC Weight%

Arvin

Sacramento -Natomas

Sacramento –Del Paso Manor

Folsom

Isoprene: re

activ

ity issu

es; p

ossib

le overestim

ated biogenics

at A

rvin.

26

Fingerprint Comparisons (5 of 5)

27

Overall Findings

� Overall, the emissions data show better agreement with ambient data than previous emission inventories have.

� At some sites, the emissions data correlate with ambient data as closely as could be expected given analyses limitations.

� EI generally under-predicts pollutant ratios

28

Overall Findings

� Urbanized Sacramento area:• Good agreement on weekdays

• Poorer agreement on weekends

� Urbanized Fresno area:• Good agreement on weekdays and weekends

� Urbanized Bakersfield:• Poor agreement on weekdays and weekends

29

Recommendations

� Improve accuracy of weekend emission estimates.

� Correct spatial distribution of emissions (e.g., from livestock waste, and heavy duty engines).

� Further investigate the poor agreement in Kern County.

� Collect more ambient data at Bay Area sites to enable more robust evaluations.

30

Questions

and

Discussion

comparison of ambient measurements to emissions ... · comparison of ambient measurements to...

Documents