Mobile emissions processing

1
Workshop for Air Toxics and Health
Effects October 17-18, 2005 University of Houston
Recent Developments in Air Quality Modeling
Techniques for studying Air Toxics in the
Houston-Galveston Area
Prof. Daewon W. Byun Dr. Soontae Kim, Ms.
Violeta Coarfa Dr. Peter Percell University of
Houston Institute for Multidimensional Air
Quality Studies Dr Graciela Lubertino, Houston
Galveston Area Council Jason Ching, U.S. EPA
http//www.imaqs.uh.edu
2
Why modeling air toxics?
Fate-Transport Modeling Based on First Principles

• To better understand their impacts on human
  health and air quality
• To protect public health by limiting their
  emissions from man-made sources
• To help communities prepare to better respond in
  case of chemical spills of such substances
• Air toxics assessment activities include
• - assessment of emissions monitoring and
  modeling of air quality
• - programs for reducing exposure and emissions of
  pollutants
• - development and implementation of control
  strategies of emissions
• - emergency response in case of serious events.

3
Linking Air Quality and Exposure Modeling
Ching, 2005 presentation
4
Air toxics modeling Follows Ozone PM
approach
US Continent
Meteorology
Where we are now.
Regional/State
Air Quality
HG area
What we need to work on
Neighborhood scale
For neighborhood scale modeling Method 1
Combine CMAQ with ASPEN (following EPAs
Philadelphia Study) Method 2 Apply trajectory
adaptive grid (TAG) method
HAPEM
Ching, 2005 presentation
5
Key Issue How to improve Air Toxics Emissions
Inventories

• Air toxics Emissions Processing Method
• Separate processing of toxics species that do
  not get involved in the main chemistry (i.e.
  CMAQ/HAPS)
• Combined processing of toxics with other
  photochemical (O3) and PM model species with full
  interaction (i.e. extended SAPRC) CMAQ/Air-Toxics
• Which Emissions Inventory?
• Which Model Species?
• How to process??

http//www.imaqs.uh.edu
6
Toxic Emissions Inventories

• NEI Criteria VOC emissions
• Needs proper speciation profiles
• NEI HAP emissions
• One-to-one mapping
• Texas Criteria VOC emissions
• Needs proper speciation profiles
• Texas PSDB
• Speciated One-to-one mapping
• MOBILE6.2 Toxic emissions from HGAC

EPA air toxics Inventories available
What we did
What we are working on now
Additional Efforts Required
7
Preliminary CMAQ/AT Results

• Preliminary studies have been done using three
  different inventories
• Emissions Data
• 1. EPAs National Emissions Inventory NEI99
  
• Criteria (CAPs) Hazardous (HAPs)
• Tools SMOKE CMAQ/Air-Toxics PAVE
• Domain 4km_83x65
• Period Aug, 22 Aug, 31, 2000
• Simulation results were compared with the
  observations

8
Emissions for toluene and ethylbenzene
On-road
Point
Area
Non-road
Non-road
Point
Area
On-road
Air Toxics Emissions
NEI99 with SMOKE2.1 Result
August 25, 2000 1800 CDT
All Sources
All Sources
9
CMAQ/AT 4.4 results for aromatics
August 25, 2000 0800 CDT
August 25, 2000 1600 CDT
Air Concentrations
August 25, 2000 _at_ 1 pm GMT and 9 pm GMT
10
Comparison with hourly observations at Clinton
site
11
Comparison with hourly observations at Clinton
site
12
Comparison with hourly observations at Clinton
site
13
Comparison with observations at Clinton site
Some success some failures But modeling
advances are very promising if and only if we can
improve emissions inventories

• 2. TEI00 NEI99
• ? TEIb4a HAPs
• NEI99 TRI00
• ? CAPs ( HAPs TRI00)

TEIb4aNEI99 gives a better agreement between the
simulation results and Clinton observational data
for benzene
14
MOBILE 6.2 emissions
What we are working on now ? Link-based
Example species BENZ, MTBE, BUTA, FORM, ACETA,
ACROL, NAPTHALENE, ETHYLBENZE, N-HEXENE, STYRENE,
TOLUENE, XYLENE, LEAD
MSATS mobile source air toxics species
15
How mobile emissions are estimated?
16
MOBILE6 Input/Ouput
INPUT
OUTPUT

• Registration distribution- TTI
• Gasoline content TCEQ
• Control programs TCEQ
• Trip data H-GAC
• Temperature, humidity SIP
• Diesel sales fractions TTI
• Calendar year

• Emission factors (g/mile) for different air
  chemical species

17
Mobile Source Air Toxics (MSAT) Compounds

• By default Benzene, 1,3-Butadiene, Formaldehyde,
  Acetaldehyde, Acrolein, MTBE
• Extended
• Arsenic compounds, Chromium compounds,
• Dioxin/Furans, Ethylbenzene, n-Hexane, Lead
  compounds,
• Manganese compounds, Mercury compounds,
  Naphthalene,
• Nickel compounds, Polycyclic Organic Matter,
  Styrene,
• Toluene, Xylene, Diesel, Particulate Matter

18
Additional Parameters needed for Air Toxic
Calculations

• GAS AROMATIC aromatic content of gasoline on a
  percentage of total volume basis
• GAS OLEFIN olefin content of gasoline on a
  percentage of total volume basis
• GAS BENZENE benzene content of gasoline on a
  percentage of total volume basis
• E200 Percentage of vapor a given gasoline
  produces at 200F
• E300 percentage of vapor a given gasoline
  produces at 300F
• OXYGENATE oxygenate type and content of gasoline
  on a percentage of total volume basis. There are
  four oxygenate types in the model MTBE, ETBE,
  ETOH, TAME

19
Houston Road Network
What are required to estimate mobile source air
toxics emissions?
20
Road Network, Nodes Links
21
Processing is not always easy the link data
can be messed up
22
Link Nodes inside HGB 8 Counties
23
Link Nodes inside Harris County
24
Example of MOBILE6 outputs
25
Example Output File

• Anode Bnode Roadtype Pollutant EmissionType
  grams emissions by vehicle type
• 1653 10893 8 BENZ COMPOSITE
  1.1234693 1.3399E-1
• 1653 10893 8 BENZ EXH_RUNNING
  2.8073E-1 3.7492E-2
• 1653 10893 8 BENZ START
  7.1150E-1 8.3485E-2
• 1653 10893 8 BENZ HOT_SOAK
  3.3991E-2 3.0365E-3
• 1653 10893 8 BENZ REST_LOSS
  8.5768E-2 8.8688E-3
• 1653 10893 8 BENZ RUN_LOSS
  1.1484E-2 1.1099E-3
• 1653 10893 8 MTBE COMPOSITE
  3.4736070 3.6643E-1
• 1653 10893 8 MTBE EXH_RUNNING
  1.3982E-1 2.0251E-2
• 1653 10893 8 MTBE START
  4.3882E-1 5.9574E-2
• 1653 10893 8 MTBE HOT_SOAK
  8.3029E-1 7.4172E-2
• 1653 10893 8 MTBE REST_LOSS
  1.9068174 1.9717E-1
• 1653 10893 8 MTBE RUN_LOSS
  1.5786E-1 1.5257E-2
• 1653 10893 8 BUTA COMPOSITE
  1.5084E-1 1.7860E-2
• 1653 10893 8 BUTA EXH_RUNNING
  4.1085E-2 5.1978E-3
• 1653 10893 8 BUTA START
  1.0976E-1 1.2663E-2
• 1653 10893 8 FORM COMPOSITE
  4.3185E-1 6.3634E-2

26
Example of Link-based Emissions

• VOC emissions from Brazoria County

27
Example of Link-based Emissions
Benzene
Toluene
Xylene
Styrene
28
Example of Link-based Emissions
Benzene
http//www.imaqs.uh.edu
29
Further Processing of air toxics emissions with
SMOKE for CMAQ/AT modeling
Link-to-gridded emissions
Input data
EI Processing for AQMs
Temporal profiles
Allocating each link emissions to the covering
cells
Link-based MOBILE6 output
Gridded MOBILE6 emissions
SMOKE
Hourly emissions
Preparing MOBILE6 emission / vehicle types for
temporal allocation
Diurnal temporal x-ref and profile
http//www.imaqs.uh.edu
30
Static Adaptive Fine-mesh Eulerian (SAFE) Grid
Point Source VOC Emissions in Houston-Galeston
31
Point source emissions inventory differences
32
Differences in Benzene Emissions from Point
Sources
TCEQ PSDB 2000
NEI99 HAP
33
Differences in 1,3-Butadiene emissions from
Point Sources
TCEQ PSDB 2000
NEI99 HAP
34
Processing of EI for speciated air toxics
modeling requires speciation profile

• SAPRC99
• 0005 TOG ALK1 0.00083139342
  1 0.025
• 0005 TOG ALK2 0.00030721966
  1 0.008
• 0005 TOG ARO1 0.00007297401
  1 0.01
• 0005 TOG CH4
  0.05162094763 1 0.828
• 0005 TOG ETHENE 0.00417112299
  1 0.117
• 0005 TOG OLE1 0.00007129278
  1 0.003
• SAPRC Extended
• 0005 TOG CH4 0.90174382925
  17.4718304 0.8280000
• 0005 TOG ALK1 0.01452596486
  17.4718304 0.0250000
• 0005 TOG ALK2 0.00536810188
  17.4718304 0.0080000
• 0005 TOG ETHE 0.07286676019
  17.4718304 0.1170000
• 0005 TOG OLE1 0.00124558562
  17.4718304 0.0030000
• 0005 TOG BENZ 0.00424974738
  17.4718304 0.0190000

CMAQ/AT with Extended Aromatic Chemistry Mechanism
35
Current Developments in Air Toxics Modelingat
IMAQS, University of Houston

• An Extended Chemical Mechanism of the EPAs CMAQ
  for Air Toxics Studies (Poster)
• Problem of popular chemical mechanisms many
  chemical species are lumped cannot simulated the
  behavior of individual compounds important in the
  atmospheric chemical processes and/or with
  serious impact on the human health and
  surroundings
• Solution find methods to implement species of
  interest in the chemical mechanisms employed by
  the photochemical models - SAPRC99/extended
  (CMAQ/AT)
• Suitable for studying acute health effects and
  verifying auto GC and canister measurements

http//www.imaqs.uh.edu
36
Current Developments in Air Toxics Modelingat
IMAQS, University of Houston

• A Transport Model for the Air Toxics Studies
  (Poster)
• Long-term simulations (several months, yearly)
  are preferred to better analyze and understand
  the physical and chemical behavior of toxic
  pollutants
• Health effect studies need long-term simulations
  for a proper correlation between pollutant
  concentration and various health conditions need
  a faster model than CMAQ/Air-Toxics
• IMAQS developed an engineering version of
  CMAQ/Air-Toxics, which can simulates seasonal and
  annual simulations (CMAQ/HAPS)
• ? Suitable for studying chronic health effects of
  air toxics
• New method for air quality modeling -- under
  development
• An Eulerian-Lagrangian Hybrid Modeling Method,
  Trajectory Adaptive Grid (TAG) underdevelopment
  (CMAQ/TAG)
• Can handle multiscale air quality issues at
  reasonable computational cost with high accuracy

http//www.imaqs.uh.edu
37
Lagrangian packets to represent movement of
pollutants, but in Eulerian adative grid
Trajectory Adaptive Grid (TAG) Algorithm(very,
very preliminary results as of today)
Eulerian Grid
Eulerian Grid
2-D
3-D
Lagrangian packets
38
Testing of TAG O3 (UTC 2000 Aug 25)
TAG-Result
Preliminary O3 simulation results
Maximum
Packet Average
Eulerian (CMAQ-PPM)
Closest
Minimum