Baton Rouge Ozone: Cause

1
Baton Rouge Ozone Cause Effect A look into
the Five Parish Baton Rouge Area Ozone Exceedances

• October 2, 2003

2
Agenda

• Welcome
• Concern over Ozone Exceedances
• Ozone Exceedance Assessment
• Monitored Data
• Actual Emissions
• Recent Actions
• Next Steps

3
BTR Ozone History Progress

• In Nov. 1990, BTR was designated as a serious
  ozone non-attainment area. The attainment date
  was set for Nov. 1999.
• The first Ozone Task Force was formed.
• Rate of Progress commitments were made to reduce
  VOC emissions - 15 by 1994 9 by 1996
• Despite progress in reducing emissions and the
  number of exceedance days, BTR failed to meet the
  Nov. 1999 attainment date.
• The second Ozone Task Force convened.
• Effective June 23, 2003, the BTR area was bumped
  up to a severe ozone non-attainment
  classification.

4
5 Parish Baton Rouge Area Point Source VOC
Emissions
Approx tons 1990 39,000 2002 13,000 67
Reduction
5
5 Parish Baton Rouge Area Point Source NOx
Emissions
Approx tons 1990 66,700 2002 42,300 37
Reduction
6
1-hr Exceedances / Design Values
7
8-hr Exceedances/Design Values
85 ppb standard
8
Consequences of the ExceedancesContinued
Non-Attainment

• No chance for attainment with the 1 hour standard
  for 2003 or 2004
• Attainment date is 2005
• Severe non-attainment designation remains
• RFG
• Offsets increased to 1.3 to 1
• Redefinition of Major Source lowered to 25 tpy
• Penalty Fees?

9
Consequences of Continued Non-Attainment Severe
Status Penalty Fees

• CAA Section 185 requires the assessment of
  penalty fees on facilities in an area designated
  as severe if the area fails to reach attainment
  by 2005.
• Fees are assessed for each ton of NOx and VOC
  emissions in excess of 80 of the sources
  baseline amount.
• Fee is equal to 5000 per ton plus an annual CPI
  adjustment predicted to be 7500 per ton.
• Based on actual 2002 emissions, annual fees for
  some local companies will likely equal (in
  millions)
• Dow - 14.8 Shell Chem. - 1.9 PCS - 2
• Exxon Chem. - 6.3 Exxon Ref. - 7.9 Placid - 1.8

10
Ozone Exceedance Assessment Jim Hazlett
11
1-hour exceedances occurred at the following
monitors in 2002
1-hour Max, ppb
(Standard 125 ppb)
8/8/02 9/11/02
LSU 102 154
Capitol 131 164
Port Allen 128 102
Bayou Plaquemine 66 139
12
1-hour exceedances occurred at the following
monitors in 2003
1-hour Max, ppb
(Standard 125 ppb)
4/27/03 7/18/03 8/18/03 9/19/03
LSU 88 174 108 129
Capitol 78 147 139 105
Port Allen 81 129 144 94
Bayou Plaquemine 78 78 77 126
Carville 129 129 74 97
13
Exceedance Notice as of 9/21/03
14
Common Factors During Baton Rouge Ozone Episodes

• Recent ozone events occurred on days predicted to
  be green or low yellow air quality.
• On 8/18/03 the cloud cover was predicted to be
  about 65.
• Afternoon clouds did not arrive, however, and
  cloud cover was actually closer to 10.
• Ozone events characterized by low wind speed and
  temperatures in the 90s except for Sept 19th (
  84º).
• Most of the Ozone events appear to be very
  localized and not regional in nature.
• While the Baton Rouge Capitol Site is dominated
  by mobile sources 90 of the time, recent ozone
  events were preceded by spikes of VOCs not
  characteristic of mobile sources.

15
Localized Episode
April 27th Exceedance at Carville
1 hr stnd - 125 ppb
16
Ozone Forecast for 7/18/03
1-hr (ppb) 8-hr (ppb)
AM Forecast 91 70
PM Forecast 91 70
Maximum Measured Value 174 119

• Measured values are much higher than
  predicted values for both 1-hr and
  8-hr concentrations.

17
Ozone Forecast for 8/18/03
1-hr (ppb) 8-hr (ppb)
AM Forecast 88 75
PM Forecast 72 60
Maximum Measured Value 144 112

• Measured values are much higher than
  predicted values for both 1-hr and
  8-hr concentrations.

18
Improving the Understanding of Ozone Episodes

• Rate of ozone formation characterization
• Identify VOC species which are the most
  significant contributors to ozone formation
• Improve analysis of monitored VOC variability
  during preceding episodes
• Back trajectory analysis to locate possible
  sources
• Acquire better understanding of plant-to-plant
  emissions variability during ozone episodes

19
How is ozone formed?
O O2 O3
NO2 hv NO O
NO O3 NO2 O2
Reactive VOCs interact with hydroxyl radicals to
convert NO to NO2
O3 hv O2 O
O H2O 2 OH
OH O2 RH RO2 H2O
RO2 NO RO NO2
RO O2 HRHO HO2
HO2 NO NO2 OH
The outcome of these reactions is the additional
NO2 formed photolizes producing more ozone
less NO is available to scavenge the ozone
20
Rate of Ozone FormationIn the hour preceding
exceedances
10-15 ppb/hr typical gt 40 ppb/hr rapid ozone
formation
21
The July 18th event showed very rapid buildup of
ozone at Capitol, Port Allen LSU.
1 hr stnd - 125 ppb
22
Capitol Site 7/18/2003
1 hr stnd - 125 ppb
1200-100 60 ppb ozone increase
23
Port Allen Site 7/18/2003
1 hr stnd - 125 ppb
1100-1200 50 ppb ozone increase
900-1000 50 ppb ozone increase
24
Most Abundant VOC Species Identified at BR
Capitol Site(Top 10 Represent 48 of Total VOC)
propane n-butane ethane isopentane
(2-methylbutane ) ethylene isobutane
(2-methylpropane) propylene n-pentane toluene n-he
xane
Ranked by abundance in ppbc
25
The incremental reactivity characterizes the
ozone formation potential of a VOC specie and all
of its reaction products.
VOC Specie Maximum Incremental Reactivity (MIR)
methane 0.005 mol Ozone/mol C
propylene 2.750 mol Ozone/mol C
butane 0.310 mol Ozone/mol C
benzene 0.110 mol Ozone/mol C
ethane 0.080 mol Ozone/mol C
propane 0.150 mol Ozone/mol C
isoprene 2.580 mol Ozone/mol C
ethylene 2.160 mol Ozone/mol C
toluene 0.740 mol Ozone/mol C
Isopentane 0.410 mol Ozone/mol C
26
VOC Species at BR Capitol SiteRanked by Ozone
Forming Potential(Top 10 Represent 95 of Total
Reactivity)
ethylene propylene isoprene m/p-xylene 1-butene n-
butane isopentane (2-methylbutane
) toluene o-xylene isobutane (2-methylpropane)
Ranked by abundance in ppbc MIR
Red indicates HRVOC species with recently lowered
RQ.
27
VOC Diurnal Profiles
Typically VOC concentrations peak during 6-9am
period and reach lowest levels during late
afternoon hours.
28
Analysis of VOC Profiles

• The Hydrocarbon Profile (fingerprint) of Capitol
  Site is usually very consistent and dominated by
  mobile source emissions, particularly during the
  morning rush hour (6-9am).
• An analysis of peak concentrations observed
  before during ozone events has indicated
  sources of VOCs other than mobile.
• The Capitol site 6-9am sample on 8/17/03 showed
  impact of source very similar to evaporative
  gasoline.
• A sample collected just 3 hours later shows
  entirely different source domination in the VOC
  profile. Profile dominated by very light VOCs
  including the highly reactive species, ethylene
  propylene.

29
The ozone event on 7/18/2003 was preceded by
above normal spikes of VOCs.
Capitol Site 300 pm exceedance 147ppb
30

Note the peaks in VOC concentration throughout
the ozone exceedance.
31
1 hr stnd - 125 ppb
32
The ozone event on 8/18/2003 was also preceded by
above normal spikes of VOCs.
Capitol Site 200 pm exceedance 139 ppb
33
Normal mobile-dominated source profile versus
profile of sample containing above normal amounts
of HRVOCs
Isobutane
Ethylene
Propane
Propylene
34
The ozone event on 9/19/2003 was preceded by a
large above normal spike of VOCs.
Bayou Plaq. 1200 pm exceedance 126 ppb
35
The ozone event on 9/19/2003 was preceded by a
long duration above normal spike of VOCs.
Exceedances 1200 Bayou Plaq. 126ppb 100 LSU
129ppb
36
(No Transcript)
37
Normal mobile-dominated source profile versus
profile of sample containing above normal amounts
of HRVOCs
1,3-butadiene
Ethylene
Propylene
Isobutane
38
Hour preceding exceedance dominated by above
normal amounts of HRVOCs
Ethylene
Ethane
Isobutane
39
Southern University Site detects frequent hits of
industrial emissions.

• Continuous VOC monitor triggers canister
  collection during periods of above normal VOC
  concentrations.
• Sampler has triggered 43 times so far during
  2003.
• Multiple VOC profiles have been observed.
• Total VOC concentrations up to nearly 7 ppm have
  been observed with hits of MTBE hits up to 400
  ppbv.
• Concentrations of highly reactive VOCs, ethylene
  propylene, commonly observed in the 30 60
  ppbc range.

40
The night before the July 18th ozone event, the
Southern monitor was triggered by a source
contributing ethylene isobutane.
Isobutane
Ethylene
41
The capitol site sample preceding 8/18/03 ozone
event Southern spike sample preceding 7/18/03
ozone event appear to have been impacted by the
same source emitting primarily ethylene
isobutane.
Isobutane
Ethylene
42
Isobutane
Propane
Propylene
Ethylene
43
Wind Direction Analysis
An analysis of the wind direction when
ethylene/propylene spikes occur indicates two
local sources one source to the north-northeast
of the site and a second source to the northwest
of the site.
44
Wind Direction Analysis
The south Scotlandville event monitor is
triggered less often but indicates a source to
the south with another possible source north to
northeast of the site.
45
Iberville
Ascension
West BTR
East BTR
46
(No Transcript)
47
(No Transcript)
48
Recent Actions Next Steps Hall Bohlinger
49
Recent Actions

• Emergency rule to lower HRVOC RQs
• Effective 8/12/03
• 18 of 20 notices received as a result of
  emergency rule
• Administrative Order 9/3/03
• Maintain and submit HRVOC data by 10/3/03
• Additional monitoring
• Speciated VOC event monitors added at Capitol,
  Port Allen, LSU and Carville
• Discussions with EPA on collaborative effort to
  study BTRs ozone exceedances

50
Next Steps

• Permanent rule to lower HRVOC RQs
• Required speciation of HRVOCs in permits
• Ethylene, propylene butenes
• Form HRVOC Technical Committee - Convene by
  October
• Evaluate additional data collected (Monitors,
  RQs AO)
• Design studies and determine needed resources by
  12/03
• Oversee contracted research projects to begin by
  5/1/04 and be complete by 11/04
• Recommend control strategies, if needed

51
For more information

• James Hazlett
• 225-219-3554
• Manop Vanichchagorn
• 225-219-3489
• Jennifer Walton
• 225-219-3486
• LDEQ Ozone Website
• http//www.deq.state.la.us/evaluation/ozone/index.
  htm