Meeting the 8hr Ozone NAAQS in the Houston

1
Meeting the 8-hr OzoneNAAQS in the Houston

• Prepared for
• ExxonMobil Chemical Company
• Baytown, TX 77522
• Prepared by
• T. W. Tesche
• Dennis McNally
• Gregory Stella
• Alpine Geophysics, LLC
• 20 April 2005

AG-90/TS-209-01
2
Study Objectives

• Assess adequacy of Aug-Sept 00 Episode for 8-hr
  SIP modeling
• Estimate emissions reductions beyond current 1-hr
  SIP revision needed to attain 8-hr ozone NAAQS.
• Explore relative effectiveness of controls on
  different source categories in 2007.

2000 Base Case base5b.psito2n2
2007 SIP Base fy07j.CS03_HarCap
3
Technical Approach

• Set up TCEQ base year (2000) and future year
  (2007) modeling frameworks used in 1-hr HGB SIP
  revision.
• Assess model reliability for 8-hr ozone
  attainment demonstrations.
• Model residual 8-hr non-attainment beyond
  controls in 1-hr SIP revision.
• Examine effectiveness of various emissions
  reduction scenarios in 2007.
• Estimate levels of controls likely needed for
  8-hr attainment.
• Consider feasibility of achieving the modeled
  control requirements.

4
CAMx 8-hr Ozone Evaluation
5
Model Performance for 8-hr Ozone at Bayland Park
6
Extent to Which Aug-Sept 00 Episode Includes
Near-DV Days at 20 of 23 HGB 8-hr Ozone
Exceedance Monitors
7
Will 1-hr SIP Satisfy 8-hr NAAQS?

• CAMx exercised with 2007 SIP revision inventory
  to estimate extent of residual nonattainment
• EPAs 1999 8-hr guidance followed since 2005
  document not available at the time
• Several alternative methods for calculating RRFs
  and estimating future 8-hr Design Values (DVs)
  examined.

8
Estimated 8-hr Attainment Results(all days, all
monitors)
2007 Modeled DVs Based on all days in 25 Aug-6
Sept 00 period using current (2001-2003)
measured 8-hr DVs.
9
8-hr Attainment Results Reduced Data Set
2007 Modeled DVs All days in 25 Aug-6 Sept
00 period using current (2001-2003) measured
8-hr DVs. Reduced Data Set 6 Sept
removed. No days with predicted peaks lt 70ppb. No
days with fractional bias or error gt 20.
10
2007 Baseline Assessment

• 1-hr SIP falls far short of 8-hr Attainment
• For All days, All monitors, year 2007 8-hr DVs
  range from 85-98 ppb
• For Reduced Data Set, year 2007 8-hr DVs range
  from 84-98 ppb
• For Valid Days Data Set, year 2007 8-hr DVs
  range from 85-95 ppb
• Implications
• Substantial additional NOx and/or VOC controls
  may be needed beyond the current 1-hr SIP
  controls to show 8-hr attainment by 2010.
• This attainment picture changes little by
  progressive elimination of individual monitors
  and/or modeling days associated with questionable
  8-hr model performance

11
Estimation of Additional Controls Required for
8-hr Ozone Attainment

• Objectives
• Quantify sensitivity of 8-hr DVs to VOC and/or
  NOx controls at the ground and aloft
• Examine RRF calculation sensitivity to various
  WOE approaches
• Seven (7) Across the Board anthropogenic
  emissions reduction runs
• 50 reduction in elevated NOx emissions
• 50 reduction in elevated VOC emissions
• 50 reduction in elevated VOC NOx emissions
• 50 reduction in low level and elevated (i.e.,
  combined) NOx emissions
• 50 reduction in low level and elevated VOC
  emissions
• 50 reduction in low level and elevated VOC NOx
  emissions
• 100 reduction in elevated VOC NOx emissions
  (i.e. point source zero-out)
• See Table 1 for emissions summary

12
RRF Weight of Evidence Methods

• Objective
• Examine RRF Calculation sensitivity to various
  WOE approaches
• Weight of Evidence Methods for Calculating RRFs
• EPA (No-Woe) RRFs calculated from EPA (1999)
  draft 8-hr guidance
• WOE-1 days/hours eliminated from RRF
  calculations days lt 70 ppb, fractional bias gt
  20
• WOE-2 days/hours eliminated from RRF
  calculations days lt 70 ppb, fractional bias gt
  20, entire day of 5 Sept 00
• WOE-3 days/hours eliminated from RRF
  calculations days lt 70 ppb, fractional bias gt
  20, entire day of 5 Sept 00, days where max. O3
  in base case is outside 10 ppb of the 2001-03
  8-hr DV

13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
31 Aug Daily Maximum 8-hr Ozone Concentrations
50 Low-Level NOx Reduction
2007 1-hr SIP Base Case
18
31 Aug Daily Maximum 8-hr Ozone Concentrations
50 Low-Level NOx VOC Reduction
50 Low-Level VOC Reduction
19
(No Transcript)
20
Federally Mandated Motor Vehicle Controls in Texas
Implication While modeling suggests that 50
MV controls might yield attainment, this level of
reduction not scheduled to occur until 2019 - too
late!
Source VISTAS (2004)
21
Findings

• Model performance for Aug-Sept 00 episode
• Adequate in some respects for examining 8-hr
  attainment
• Insufficient number of near-DV days covering
  all exceedance monitors
• Existing performance problems need to be
  addressed.
• Bayland Park is the monitor driving the ozone
  control requirements for this episode.
• If EPA guidance is to be followed
• Additional episodes are needed, possibly one or
  more summer long periods.
• Episodes should be selected following methodology
  laid out in EPA (2005).
• Significant differences in modeled 8-hr ozone
  attainment may arise depending upon RRF method
  employed.

22
Findings (concluded)

• 50 VOC NOx reductions at the ground and aloft
  beyond current 1-hr SIP may be required for 8-hr
  ozone attainment at all monitors.
• Combined (i.e., ground and aloft) emissions
  reductions are most effective.
• Aloft emissions reductions are less effective
  than ground level controls.
• The effectiveness of aloft VOC NOx emissions
  reductions increases with greater percentage
  reductions in aloft emissions,
• but.
• Elimination of all aloft point source VOC NOx
  emissions, regardless of the WOE method, still
  leaves nonattainment at 2-3 monitors (e.g.,
  Bayland Park, Deer Park, and/or Croquet)

23
Recommendations

• Perform CART analyses to identify 8-hr ozone
  exceedance conditions at all monitors
• Establish those conditions adequately addressed
  by Aug-Sept 2000 Episode
• Identify those conditions that must be addressed
  by additional episode(s)
• Select and model additional episodes consistent
  with EPA (2005) guidance
• Consider modeling full summertime periods (e.g.,
  2002 or 2003) to increase robustness of 8-hr
  attainment demonstration.

24
Recommendations (concluded)

• Employ corroborative models (i.e., CMAQ CAMx)
  to strengthen performance appraisal, uncertainty,
  and reliability estimates
• Conduct focused, thoughtful Weight-of-Evidence
  analyses to strengthen 8-hr SIP technical support
  documentation and
• Support strong external peer-review of modeling
  study (independent corroborative analyses
  including modeling not merely paper reviews)