Workshop on Air Quality Data Analysis and Interpretation

1
Workshop on Air Quality Data Analysis and
Interpretation

• Evaluation of Emission Inventory

2
Emission Inventories

• Emission inventories are routinely used for
  planning purposes and as input to comprehensive
  photochemical air quality models.
• Significant biases in either VOC or NOx emission
  estimates can lead to poor baseline photochemical
  model performance and erroneous estimates of the
  effects of control strategies.
• Essential top-down emission inventory
  evaluation procedure comparison of emission
  estimates with ambient air quality data.
• Caution Ambient/emission inventory comparisons
  are useful for examining the relative composition
  of emission inventories they are not useful for
  verifying absolute amounts unless they are
  combined with bottom-up evaluations.

3
Approach

• Perform the following three tasks 
• Compare early morning (e.g., 0700-0900 LT)
  ambient- and emissions-derived NMOC/NOx and
  CO/NOx ratios.
• Compare early morning ambient- and
  emissions-derived relative compositions of
  individual chemical species and species groups. 
• Compare early morning ambient- and
  emissions-derived relative reactivities of
  individual chemical species and species groups.
• Early morning sampling periods are more
  appropriate to use in these evaluations because
  they have the best potential to minimize the
  effects of upwind transport and photochemistry.
  Emissions are generally high, mixing depths are
  low, winds are usually light, and photochemical
  reactions are minimized.
• Conduct a second evaluation following the
  incorporation of the recommendations made in the
  first evaluation, in order to verify improvement.

4
NMHC/NOx Emissions

• PCD 1997 (Bangkok Inventory)
• Total NMHC (MW14 g/mol) on a per C basis
• (268,882 ton/yr)x(1000 kg/ton)/(0.014 kg/mol)
  19.2 x 109 mol/yr
• Total NOx (MW46 g/mol as NO2)
• (329,161 ton/yr)x(1000 kg/ton)/(0.046 kg/mol)
  7.16 x 109 mol/yr
• NMHC/NOx 2.7 (ppbC/ppb)

5
NMHC/NOx Emissions - Mobile

• PCD 1997 (Bangkok Inventory)
• Mobile NMHC (MW14 g/mol C)
• (232,973 ton/yr)x(1000 kg/ton)/(0.014 kg/mol)
  16.6 x 109 mol/yr
• Mobile NOx (MW46 g/mol)
• (264,648 ton/yr)x(1000 kg/ton)/(0.046 kg/mol)
  5.75 x 109 mol/yr
• NMHC/NOx 2.9 (ppbC/ppb)

6
Bangkok Emission Inventory Comparison

• NOx/CO
• Ambient 30 70 ppb/ppm
• Inventory (Total) 430
• Inventory (Mobile) 460
• NMHC/NOx
• Ambient (slope) 9.3 ppbC/ppb
• Ambient mean, median 22.9, 17.2
• Inventory (Total) 2.7
• Inventory (Mobile) 2.9

7
Lets look at the NMHC/CO ratio in emissions!

• Total Emissions
• NMHC/CO (19.2 x 109 mol/yr)/ (16.5 x 109
  mol/yr) 1.2 ppbC/ppb
• Mobile Emissions
• NMHC/CO (16.6 x 109 mol/yr)/ (12.5 x 109
  mol/yr) 1.3 ppbC/ppb
• Ambient
• NMHC/CO 0.5 (slope of scatter plot)
• NMHC/CO 1.3, 0.9 (Mean, median of ratio at
  National Housing 10T)

8
Bangkok Emissions Inventory Conclusions

• NOx/CO lower for ambient than inventory
• NMHC/NOx higher for ambient than inventory
• NMHC/CO reasonably close in ambient to
  inventory
• These results make one question the NOx portion
  of the inventory specifically. It seems to be
  high in the inventory relative to both CO and
  NMHC.

9
Differences between Emission Inventories and
Ambient are Common
10
Problems with Vehicle Emissions
11
Uncertainties in Evaluation of Emission
Inventories

• EMISSION INVENTORY UNCERTAINTY ISSUES
• Spatial and temporal allocation of activities
• Adjustment of emission rates for temperature and
  day-specific activities
• Assignment of accurate and representative source
  speciation profiles
• AMBIENT MEASUREMENTS UNCERTAINTY ISSUES
• The representativeness of the monitoring sites
• The influence of lower quantifiable limits and
  precision
• The identification, misidentification, or lack of
  identification of all important species
• Potential sampling or handling losses of total
  mass or individual species
• COMPARISONS-RELATED UNCERTAINTY ISSUES
• The matching of emissions and ambient NMOC
  species
• The temporal matching of the emissions and
  ambient data
• The spatial matching of the emissions and ambient
  data
• Meteorological factors such as wind speed and
  direction and mixing height
• The level of ambient background NMOC and NOx
  concentrations
• Chemical reactions

12
VOCs as tracers

•   
•    
•  
•    
•   
•  
•    
•  
•  

13
VOCs as tracers (continued)
14
SPECIATE 3.2

• http//www.epa.gov/ttn/chief/software/speciate/ind
  ex.html
• This is a very useful tool to provide estimates
  of the composition of emissions from a variety of
  sources.
• Speciates the TOC emissions from a few hundred
  different sources into individual organic
  compounds.
• Also, speciates the PM emissions from a few
  hundred different sources into individual
  elemental contributions.
• Source profiles can be exported to the Chemical
  Mass Balance (CMB) model.

15
Source Contributions

• Species contributions to sources are generally
  based on emission source measurements or standard
  source-contributions like SPECIATE.
• Source characterization can be quite expensive
  and representative of operations during test
  conditions.
• We will briefly discuss an option based on
  ambient measurements.

16
Comparison of Source Contributions
17
GRACE/SAFER

• Graphical Ratio Analysis for Composition
  Estimates (GRACE)
• Correlations between acetylene (assumed to be
  emitted solely from vehicle exhaust) and other
  VOC are used to establish the minimum and maximum
  exhaust-related ratios of acetylene to other
  species. GRACE plots of each roadway-corrected
  species versus all others are also examined.
• Source Apportionment by Factors with Explicit
  Restrictions (SAFER)
• SAFER is a multivariate receptor model that
  predicts the number of sources and their
  composition from the ambient data. SAFER requires
  that these predictions be consistent with
  observed intercorrelations of the concentrations
  and with physical constraints and explicit
  constraints derived from GRACE.
• SAFER requires large data sets, thus, the PAMS
  auto-GC data are well suited for this analysis.
• Environ. Sci. Technol., 28, 823-832, 1994.

18
Plots of VOCs vs Acetylene
19
Edge Relationship
Environ. Sci. Technol., 28, 823-832 (1994).
20
Ratios to Acetylene
21
Ambient Data for Emissions Profiles

• GRACE/SAFER RESULTS 1990 ATLANTA OZONE STUDY
• Using ambient data, obtained three source
  profiles roadway emissions (acetylene), whole
  gasoline (roadway-corrected 2,3-dimethylpentane),
  gasoline headspace vapor (n-butane).
• GRACE/SAFER-derived profiles compared well to
  source measurements.
• Source profiles used in subsequent CMB modeling.
• PAMS data well suited for these analyses.

22
VOC Source Contributions
Roadway
Whole Gasoline
Gasoline headspace
White model derived Black source derived
23
Chemical Mass Balance Approach

• The CMB model can be quite useful in identifying
  various source contributions to ambient air
  quality measurements.
• CMB has been used extensively to understand
  source contributions to particulate measurement,
  based on the elemental composition of samples.
• The same approach is quite useful for
  understanding various source contributions to
  ambient VOC measurements, based on speciated VOC
  composition of the samples.