Contrasting effects of different anthropogenic

source types on ozone formation in Houston

T.B. Ryerson, M. Trainer, D.K. Nicks Jr., A.J. Neuman, R.W. Dissly, J.S.Holloway, R.O. Jakoubek, D.D. Parrish, G.J. Frost, C. Wiedinmyer, D.T.Sueper, W.C. Kuster, P.D. Goldan, G. Huebler, W.M. Angevine, and F.C.Fehsenfeld

Aeronomy Laboratory, NOAA, Boulder, CO

S.G. Donnelly, S. Schauffler, E. Atlas, A.J. Weinheimer, F. Flocke, B.P.Wert, W.T. Potter, and A. Fried

Atmospheric Chemistry Division, NCAR, Boulder, CO

C.J. Senff, L.S. Darby, R.M. Banta, R.J. Alvarez

Environmental Technology Laboratory, NOAA, Boulder, CO

Additional thanks for:

Surface ozone and wind observations (TNRCC)Twin Otter ozone data (Baylor University)

30.5

30.0

29.5

29.0

Lat

itu

de

-96.0 -95.5 -95.0 -94.5Longitude

30.5

30.0

29.5

29.0

Lat

itu

de

-96.0 -95.5 -95.0 -94.5Longitude

30.5

30.0

29.5

29.0

Lat

itu

de

-96.0 -95.5 -95.0 -94.5Longitude

Sunday, August 27(87 ppbv at Conroe)

Monday, August 28(112 ppbv at Conroe)

TexAQS 2000: Electra flight data used to:

• differentiate between source typespetrochemical, urban, and power plant

• assess observed differences inozone production rateozone production yield

• contrast non-exceedance with exceedance days

Friday, August 25(194 ppbv at Crawford)

30.5

30.0

29.5

29.0

Lat

itu

de

-96.0 -95.5 -95.0 -94.5

Longitude

Sweeny Freeport BFreeport A

Choc. Bayou

TX CityParish

29.6

29.4

29.2

29.0

Lat

itu

de

-96.0 -95.8 -95.6 -95.4 -95.2

Longitude

20 ppbv

Sweeny

Freeport B

Freeport A

Choc. Bayou

Parish

Track (3*NOy) Ozone

40

30

20

10

0

EN

Ox , kto

ns/year

543210ENMOC, ktons/year

Use Aug. 27 & 28 to study: • isolated petrochemical plumes S. of Houston • coalesced Ship Channel and TX City plume • contrast source types:

power planturbanpetrochemical

Prompt ozone formation observed downwind ofisolated complexes

• use (speciated VOC)/NOx:- compare to inventories- contrast effects on HOx

• estimate plume ozone

formation rates & yields

12

10

8

6

4

2

0

NO

x, p

rop

ene,

eth

ene

, pp

bv

6:50 PM8/27/00

6:54 PM 6:58 PM 7:02 PM 7:06 PM

70

60

50

40

30

20

10

Ozo

ne, p

pb

v

Sweeny Freeport ChocolateBayou

Ethene

Propene

NOx

Evaluation of alkene emissions inventories

• use measured (ethene/NOx) and (propene/NOx)

• TNRCC PSDB predicts much, much lower(alkene/NOx) ratios from these sources:

(Total NMOC)/NOx, mol/mol:Choc. Bayou: 0 .3Freeport: 0 .1Sweeny: 0 .2

using average NMOC mass of 58 g/mol (C4H10)

Order of magnitude estimates:

measured (propene/NOx) ~ 1measured (ethene/NOx) ~ 1

25

20

15

10

5

0

NO

, NO

2 , and

NO

y , pp

bv

6:50 PM8/27/00

6:54 PM 6:58 PM 7:02 PM 7:06 PM

Time, GMT

70

60

50

40

30

20

10

Ozo

ne, p

pb

v

0.1

2

4

6

81

2

4

6

810

2

k OH *

[X

]

Sweeny Freeport ChocolateBayou

Σ(VOC)

NO2

NO2

NO2

Σ(VOC)Σ(VOC)

Σ(VOC)

25

20

15

10

5

0

NO

y , pp

bv

6:50 PM8/27/00

6:54 PM 6:58 PM 7:02 PM 7:06 PM

Time, GMT

10

8

6

4

2

0

HC

HO

, pp

bv

0.1

2

4

6

81

2

4

6

810

2

k OH *

[X

]

all (kOH*X) data are plotted

isoprene

ethene

propene

HCHO

etheneisoprene

HCHO

ethene

isoprene

propene

etheneisobutane

HCHO

hexane1,3-butadiene

Relative OH reactivities:isolated petrochemical plumes

August 27, 2000

• alkenes dominate plume VOC reactivity • all other compounds (alkanes, aromatics)

do not contribute substantially toobserved rapid ozone formation

25

20

15

10

5

0

NO

, NO

2 , and

NO

y , pp

bv

6:10 PM8/28/00

6:14 PM 6:18 PM 6:22 PM 6:26 PM

Time, GMT

70

60

50

40

30

20

10

Ozo

ne, p

pb

v

0.1

2

4

6

81

2

4

6

810

2

k OH *

[X

]

Sweeny Freeport ChocolateBayou

Σ(VOC)

NO2

Σ(VOC)

NO2

25

20

15

10

5

0

NO

y , pp

bv

6:10 PM8/28/00

6:14 PM 6:18 PM 6:22 PM 6:26 PM

Time, GMT

0.1

2

4

6

81

2

4

6

810

2

k OH *

[X

]

10

8

6

4

2

0

HC

HO

, pp

bv

HCHOHCHOisoprene

ethene

propeneethene

isoprene

Similar picture on the following dayAugust 28, 2000

• No reported emissions upsets on either day • Emissions representative of business-as-usual

Isolated petrochemical plumesAugust 27 and 28, 2000

Rapid photooxidation observed;ozone formed in very high yield

1.0

0.8

0.6

0.4

0.2

0.0

HN

O3/N

Oy

543210

Plume age, hours

1.0

0.8

0.6

0.4

0.2

0.0

NO

x/N

Oy

20

15

10

5

0

O3/(

NO

y-N

Ox)

NOx decayτ =1.5 ± 0.5 hours

HNO3 formation

Ozone yield

0 8 / 2 7 0 8 / 2 8Choc. BayouFreeportSweeney

Source comparisonData from 1995, 1999, and 2000

20

15

10

5

0

Ozo

ne

yiel

d

1086420

Urban plumes(Nashville, TN)

20

15

10

5

0

Ozo

ne

yiel

d

1086420

Power plants

20

15

10

5

0

Ozo

ne

yiel

d

1086420

Plume age, hours

Petrochemicalfacilities

Sweeny

Choc. Bayou

Freeport

8

6

4

2

0

NO

y, p

pb

v

-96.0 -95.8 -95.6 -95.4

Longitude (deg)

NOy

4

3

2

1

0

HN

O3,

pp

bv

-96.0 -95.8 -95.6 -95.4

Longitude (deg)

HNO3

70

60

50

40

30

Ozo

ne,

pp

bv

-96.0 -95.8 -95.6 -95.4

Longitude

Ozone

Lagrangian plume model results(M. Trainer, NOAA)

• Model dispersion and emissions ratios

constrained by Aug. 28 Electra data

SweenyEethene = 5 x ENOx

Epropene = 2 x ENOx

(no alkenes)

FreeportEethene = ENOx

Epropene = ENOx

(no alkenes)

• OH chemistry only;model simulates theobserved HNO3, PANs,HCHO, and ozone

• Rapid ozone formationis consistent withobserved ratios of(alkene/NOx) inthese plumes

30.5

30.0

29.5

29.0

Lat

itu

de

-96.0 -95.5 -95.0 -94.5Longitude

1.0

0.8

0.6

0.4

0.2

0.0

alkene fractio

n

20151050sum(kOHxVOC), s

-1

Coalesced Ship Channel and TX City plumes:relative OH loss rates

WAS and in-situ GC samples below 1500mentire TexAQS mission, August-September

• Reactivity over petrochemical source regionsis extremely high

• Dominated by propene, ethene, 1,3-butadiene • Enhancements consistent with LaPorte data

throughout the TexAQS study period

30.5

30.0

29.5

29.0

Lat

itu

de

-96.0 -95.5 -95.0 -94.5Longitude

Aug. 27

14121086420HCHO, ppbv

30.5

30.0

29.5

29.0

Lat

itu

de

-96.0 -95.5 -95.0 -94.5Longitude

Aug. 28

Formaldehyde data

August 27 and 28, 2000

• HCHO not emitted insubstantial amounts;

• Produced as a result of

(alkene+OH) reactions • Distributions of alkenes,

aldehydes, and ozoneconsistent with differentformation & decay rates

120

110

100

90

80

70

60

50

Ozo

ne,

pp

bv

1086420(NOy - NOx), ppbv

Parishyield ~ 2

Ship Channel/TX Cityyield ~ 12

Houstonyield ~ 5

Ozone yield comparisonElectra data, August 28, 2000

• plumes characterized by (NOx/NOy) = 0.20

• Coalesced Ship Channel/TX City ozone yieldsimilar to those derived in isolated petrochemical plumes on the same day

• No reported upsets at these times;interpreted as "business as usual"for both days

• Large co-located emissions of reactive alkenes

and NOx consistently result in rapid and efficient ozone formation downwind

200

150

100

50

0

Ozo

ne,

pp

bv

12:00 AM8/27/00

6:00 AM 12:00 PM 6:00 PM 12:00 AM8/28/00

Time of day, CST

August 27, 2000TNRCC surface ozone network

120 ppbv

Houston Regional OfficeCrawford

August 27 and 28, 2000:

• ozone concentrations nothing to write home about

• What's so different about an exceedance day?

Preliminary conclusions

• Measurements strongly suggest alkene emissionsare not accurately included in inventories

- models will not accurately simulate observationsuntil the alkene inventories are realistic

• Measurements and models further suggest that, ingeneral, routine emission of alkenes, appropriatemeteorology, and OH chemistry alone aresufficient to explain the observed ozone

- upsets may play a part, but business-as-usualemissions appear to be sufficient

- exceedances are possible on a daily basis; actualoccurrence is dependent upon meteorology

• Anthropogenic emission of large amounts of veryreactive alkenes (ethene, propene, 1,3-butadiene)co-located with NOx emission is the primarycause of ozone exceedances above 200 ppbv inthe Houston area