Puget Sound Area Ozone Modeling NW AIRQUEST December 4, 2006

1
Puget Sound Area Ozone ModelingNW
AIRQUESTDecember 4, 2006

• Washington State University
• Puget Sound Clean Air Agency
• Washington Department of Ecology

2
Overview/Disclaimer

• Presentation will NOT be an in-depth technical
  exploration of the models.
• Presentation WILL show how a local agency tapped
  into local (NW AIRQUEST) modeling resources to
  help answer internal policy/planning questions.

3
Background Ozone Levels in the Puget Sound Area
4
Background Stakeholder charge Future Factors

• Previous WSU modeling, based on a 1996 ozone
  event, contributed to stakeholder process and
  adopted strategies
• Low Reid Vapor Pressure (RVP) gasoline
• Stage II vapor recovery
• Future Factors
• Moderate population and VMT growth
• Dramatic growth projected for local ports

• Although the modeling results demonstrated
  minimal reductions in ozone concentration as a
  result of implementing the strategies the air
  quality benefits produced by the strategies could
  be an important contribution to remaining in
  attainment.
• prudence requires that early identification of
  any potential impediments to remaining in
  attainment be vigorously pursued

5
Objectives and Improvements

• OBJECTIVES
• To determine if the Puget Sound area is likely to
  remain in attainment in the future, using more
  updated inventory and methods.
• To better understand the impact that low RVP
  gasoline has on ozone formation in the Puget
  Sound area.
• To better understand ozone formation through
  sensitivity analyses.

• IMPROVEMENTS/CHANGES
• CMAQ dispersion model
• (CALGRID used in previous)
• MOBILE6 mobile model
• (MOBILE 5B used in previous)
• Land surface model MM5
• (simple soil MM5 in previous)
• 1998 ozone event
• (1998 concentrations higher than 1996)
• 1999 emission estimates (1996 NEI)

6
Modeling Project Process and Design

• Emissions Inventory Technical Group
• PSCAA, Ecology, WSU
• Inventory based largely on EPAs 1999 National
  Emissions Inventory
• Exception is non-road mobile (marine)
• Base Case Year (1998)
• Future Case Year (2015)
• Two scenarios run, with and without low Reid
  Vapor Pressure (RVP) agreements
• Sensitivity Runs
• Runs increasing and decreasing VOC and NOX in
  different combinations

7
Emissions Inventory SummaryVOC, NOX emissions
reduce 20, 25
Inventory for PSCAA Counties - King, Kitsap,
Pierce, Snohomish
8
Results High Ozone Sites

• 2015 ozone levels lower than 1998 at highest
  ozone sites
• High or low RVP gasoline made small difference
  in 2015 levels, less than 0.3 ppb at a monitor
• Poor model performance on second and third day
  of event

1-hour ozone at Enumclaw. Red 1998
observations, blue 1998 model, green 2015
model (low RVP)
9
Results Urban Core Sites

• Beacon Hill, near urban core, showed higher
  concentrations, in 2015 than 1998
• Higher concentrations moving closer to urban
  area
• Ozone/NOX ratios increasing from 1998 to 2015,
  smaller band VOC-limited
• Not important for attainment status, but may
  potentially have public health implication

1-hour ozone at Beacon Hill. Red 1998
observations, blue 1998 model, green 2015
model
10
Results Low RVP Agreement
Difference between 2015 high and low RVP
scenarios less than 1 ppb
At individual monitoring sites, the difference is
less than 0.3 ppb
11
Results Sensitivity Analyses (based on 1998)

• Reducing VOC (with NOX constant or increasing)
  yields most optimal ozone reductions

12
Uncertainties

• Uncertainties in emissions inventory and
  projections
• Uncertainties in meteorological and dispersion
  modeling
• Under-prediction in second, third day of event
• Future uncertainties
• Lower future ozone NAAQS?
• Background concentrations from developing
  countries
• Climate change effect?

13
Results Objectives Met?

• Modeling confirms previous results. The Puget
  Sound area is likely to remain in attainment in
  future years (with the current standard).
• Voluntary low RVP gasoline will make little
  difference as cleaner cars are incorporated into
  the fleet in future years.
• The region is VOC limited. VOC emissions
  reductions contribute to lower ozone
  concentrations. (caveat sensitivity runs based
  on 1998 scenario).

14
Post-Model Discussions/Input

• Results inform PSCAA Programs
• Voluntary Low Reid Vapor Pressure (RVP) gasoline
  in summer months
• Vapor recovery (Stage II)

• Continuous Improvement
• Possible model metrics (in addition to EPAs
  guidance)
• Model improvements

15
Acknowledgements and Questions

• Many thanks to all who contributed
• Washington State Department of Ecology
• Washington State University
• Region 10 Environmental Protection Agency
• Puget Sound Regional Council
• Puget Sound Clean Air Agency

16
Results Pack Forest
1-hour ozone at Pack Forest. Red 1998
observations, blue 1998 model, green 2015
model (low RVP)
17
Results Model Performance

• Overall, CMAQ model performance statistics met
  EPA guidance
• 1-hour mean normalized bias and error of -3
  and 21 (guidance is 15, 35)
• 8-hour normalized bias of 1 (matched cell) and
  9 (5 x 5 cells)
• Very good agreement at Pack Forest
• Under-prediction at Enumclaw sites
• Likely due to meteorology and over-prediction of
  planetary boundary layer
• Negative bias around 15
• When include surrounding cells, negative bias
  around 10
• Scatter diagram, quantile-quantile plot show
  good agreement between observed and predicted
• MM5 modeling generally shows an index of
  agreement 0.8 for wind speed, direction,
  temperature, and humidity.

18
Results Test of Future AttainmentWeyerhauser
Mill, Lake Sammamish, Pack Forest

• Designed to take model uncertainty account, using
  model in a relative sense. Very Conservative as
  it takes 5 years into account.
• Calculate current design values (DVC) based on
  4th high 8-hour concentrations
• Calculate the relative reduction factor from
  modeling
• RRF average 2015 8 hour max
• average 1998 8 hour max
• Calculate future design value DVC RRF
• SITE RRF Future Design Value
• WM 0.90 68 ppb
• LS 0.97 58 ppb
• PF 0.94 68 ppb

19
Results Sensitivity Analyses
8-Hour Ozone Standard Exceeded with Entire Source
Categories Removed
20
Emissions Inventory SummaryTotal VOCs estimated
to decrease 20
Total 950 tons/day
Inventory for PSCAA Counties - King, Kitsap,
Pierce, Snohomish
Total 750 tons/day
21
Emissions Inventory SummaryNOX Emissions
Estimated to Decrease 25
Inventory for PSCAA Counties - King, Kitsap,
Pierce, Snohomish
Total 400 Tons/Day