Mobile emissions processing

1
2006 National Air Quality Conferences February
5-8, 2006, San Antonio, TX
A Case Study of High Ozone Day Emissions for Air
Quality Forecasting System in Houston-Galveston
Area
Soontae Kim1, M. Jang1, C. Song1, F. Cheng1, B.
Czader1, P. Percell1, F. Ngan1, H. Kim1, H. In1,
B. Lambeth2, Daewon Byun1 1 Institute for
Multidimensional Air Quality Studies, Univ. of
Houston 2 Texas Commission on Environmental
Quality
2
Outline

• Overview of Houston-Galveston area
• Emissions inventories used National vs. Texas
• Emissions processing steps for air quality
  forecasting
• Overall Evaluation of AQF results
• A case study for emissions uncertainties
• Conclusions and future works

3
Air Quality Forecasting Setup
ETA data downloading
MM5
Meteorology
MCIP
Emissions
SMOKE
Air Quality
CMAQ
4
Houston-Galveston Area

• Ozone non-attainment
• Houston ? Mobile Area sources
• The Ship-Channel ? Point sources
• Biogenic emissions

Stetson (2003)
5
Air quality in Houston-Galveston-Brazoria (HGB)
and Other Cities in USA
Percent of Federal Standard (120 ppb 1-hr ozone)
Source 2003 South Coast Air Basin Report

6
Emissions from HGA 8 Counties
NOx
VOC
Based on TCEQs Base4a pt_n2o2 emissions on
Aug. 30, 2000
7
Point Sources
-


35.7 of the U.S. capacity for benzene (700
million
The Houston Gulf Coast has 50 of the US
petrochemical production capacity
pounds per year),

56.1 for butadiene (2.4 billion pounds per
year),

52.6 for ethylene (27.6 billion pounds per
year),

62.8 fo
r polymer grade propylene (10.9 billion pounds
per year),

25.3 for toluene (400 million pounds per year),

37.6 for xylenes (4.9 billion pounds per year).

32.6 for polyethylene, and

63.0 of its capacity for polypropylene.


SOURCES Business Houston
2001, Houston Facts 2000,
Chemical Marketing Reporter 1996
-
1997 County Business
Patterns, 1998, U.S. Census Bureau

8
Point Source Emissions
VOC
GWEI
TCEQ
Downtown
The Ship Channel
9
Houston Road Network
Real roads
Links for MOBILE6
Based on link data from TCEQ, HGAC, and TTI
10
Mobile Source Emissions
NOx
(g/hr)
(g/hr)
11
The Houston Ship Channel Emissions
NOx
VOC
NEI
TEI
12
National vs. Texas EI (CB4)
TEI - NEI
Aug. 25, 4pm
13
Surface monitoring sites used to compare NEI and
TEI
C604 HRM-4
C604 HRM-7
C403 Clinton
From a TCEQs web site
14
Simulations with NEI TEI for TexAQS2000
TEI
NEI
15
Ozone production efficiency for Aug. 25, 2000
TEI
NEI
16
Anthropogenic Emissions

• Texas EI from TCEQ
• NOx, CO, and VOC (after imputation)
• TexAQS 2000 (No projection for Year 2005)
• The day of Week
• Aug. 25th ? Friday, Aug. 26th ? Saturday, Aug.
  27th ? Sunday, Aug. 30th ? Monday Thursday
• CB4
• National EI for 1999
• NH3, SO2, and PM
• Outside Texas

17
Simulated O3 Concentrations Regular vs. Imputed
Regular
Imputed
Improvements on high ozone days
La Porte site for August 22 August 31, 2000
18
Simulated ETH Concentrations Regular vs.
Imputed
Regular
Some missing peaks with base emissions
Improvement here
Imputed
Sometimes over-predicted
La Porte site for August 22 August 31, 2000
19
Biogenic Emissions

• LULC data
• TCEQs LULC for biogenic emissions
• Processors
• GloBEIS version 3.1 BEIS version 3.12
  (modified)
• Processing steps
• Step 1 Using GloBEIS 3.1
• Normalized emissions (30 oC, 1000 µmol m-2 s-1)
• Split factors for a chemical mechanism selected
• Step 2 Using BEIS 3.12
• Temperature PASR adjustment

20
TEI processing with SMOKE
SMOKE processing
EI preparation

• Format conversion
• AMS/AFS ? IDA
• Internal database
• Surrogates
• Split factors
• Temporal profiles

Spatial allocation 36km, 12km 4km Temporal
allocation hourly emissions Chemical
speciation CB4, SAPRC99 RADM2 Plume rise
Normalized emissions
Met. adjustment
GloBEIS3
BEIS3
21
Imputed VOC Emissions
22
SMOKE processing for AQF
Chemical speciation
Spatial allocation
Temporal allocation
EI reading
Area/ Mobile
SMKINVEN
GRDMAT
TMPRL
MERGE
TMPRL
MERGE
TMPRL
MERGE
SPCMAT
SPCMAT
AGTS
AGTS/MGTS
Point
MERGE
MERGE
MERGE
SMKINVEN
GRDMAT
TMPRL
LAYPNT
TMPRL
LAYPNT
TMPRL
LAYPNT
SPCMAT
SPCMAT
PGTS3D
Time
-
independent components
Biogenic
-
NORMBEIS
MERGE
NORMBEIS
MERGE
BEIS3
BEIS3
Time/Met-dependent components
-
-
BGTS
Merge
AGTS BGTS PGTS3D
EGTS3D
AGTS BGTS PGTS3D
EGTS3D
CMAQ
-
ready
To minimize processing time, time and meteorology
dependent components are processed for each day.
23
Need to check AQF SMOKE processing?
CMAQ w/ EPS2-processed emissions
CMAQ w/ AQF SMOKE emissions
Simulations with TAMU MM5 compared to 19 CAMS
sties for August 23rd August 31st, 2000
24
Daily Evaluation of Hourly O3
June December, 2005
Hurricane Rita
25
Daily Maximum O3 concentrations
50 of simulation days shows less than 10 ppbv
differences in daily maximum O3.
26
Simulated vs. Observed High O3 Conc
27
Model peak of 104 ppb is in area where measured
ozone is near 140 ppb
Ozone is up to 107 ppb at Conroe where model
shows about 80 ppb
Highest measured ozone of 138 ppb is in area of
industrial plumes
Ozone is above 100 ppb in main urban plume area
Source Bryan Lambeth, TCEQ
Large area of net ozone production
Large area where ozone is over 100 ppb and
under-predicted by 30 to 40 ppb
June 2nd, 2005 at 12 CST
28
Observation vs. Forecasting
Simulated ozone concentrations are 30 ppbv lower
than those observed at noon.
Location and concentrations of high ozone match
well in mid-afternoon.
June 2nd, 2005 at 12 CST
June 2nd, 2005 at 15 CST
Observed ozone concentrations at downtown sites
are lower than those predicted.
29
Emissions Sensitivity Tests
A case study on June 2nd, 2005

• 30 NOx reduction in downtown Houston
• Mobile sources
• 30 NOx increase in downtown Houston
• Mobile sources
• 30 HRVOC increase in the Ship Channel
• ETH and OLE from point sources
• 30 ISOP increase in biogenic emissions
• Forests in surrounding area

30
NOx emissions
Emissions on June 2nd, 2005 8 Counties only
Morning rush hour
31
Houston Downtown Targeted emissions NOx
-30 NOx emissions
30 NOx emissions
Base NOx emissions
The difference
Base emissions
The difference
O3 concentrations (Base)
?O3 (NOx decreased - Base)
?O3 (NOx increased - Base)
32
ETH OLE emissions
Emissions on June 2nd, 2005 8 Counties only
Upset emissions
ETH
OLE
33
The Ship Channel Area Targeted emissions HRVOC
Base VOC emissions
30 ETH OLE emissions
?O3 (Ship Channel HRVOC increased - Base)
34
ISOP emissions
Emissions on June 2nd, 2005 04-km domain
35
Biogenic Emissions Targeted Species ISOP
-30 NOx emissions
30 ISOP emissions
Base ISOP emissions
?O3 (ISOP increased - Base)
36
Summary

• Air quality forecast using the MM5-SMOKE-CMAQ
  system has been conducted for Houston-Galveston
  area.
• After evaluating model performance with two
  emissions inventories, Texas EI from TCEQ was
  used to prepare AQM-ready emissions. In order to
  have aerosol species and its precursor emissions,
  NEI99 was used in part.
• Comparisons of predicted hourly O3 concentrations
  to surface measurements show r 0.6 and IOA
  0.7 for the simulation period of June December,
  2005.
• For the period, 50 of simulation days showed
  less than 10 ppbv differences in daily maximum
  O3 concentrations when compared to those
  observed.
• To estimate emissions uncertainties on high ozone
  days, several sensitivities runs were made. The
  results shows less NOx and more VOC emissions are
  favorable to come up with high ozone
  concentrations predicted.