national air toxics trends sites data characterization ... · national air toxics trends sites data...
TRANSCRIPT
National Air Toxics Trends Sites Data Characterization,
2004-2007Regi Oommen
Eastern Research Group, Inc (ERG)
1600 Perimeter Park
Morrisville, NC 27560
November 4, 2009
National Air Monitoring Conference – Nashville, TN
Acknowledgements• U.S. EPA
• Barbara Driscoll, Work Assignment Manager• Mike Jones
• ERG• Jaime Hauser• Julie Swift• Heather Perez• Karla Faught
• RTI –
Jennifer Lloyd• Work funded under U.S. EPA Contract EP-D-08-047
National Air Monitoring Conference – Nashville, TN
• Dave Dayton• Steve Mendenhall• Kerry Fountain• Jody Tisano
Oommen – pg 2
Overview of Presentation
• NATTS Overview
• Data Set
• Results
• Data Quality Challenges
• Next Steps
National Air Monitoring Conference – Nashville, TN Oommen – pg 3
NATTS•
Created to obtain long-term ambient air toxics concentration
data at specific fixed sites across the country
• Purpose: • Validating modeling results and emission inventories (NATA)
• Assessing current regulatory benchmarks
•
Establishing policies for reducing the risk of developing cancerous and noncancerous health effects
•
Evaluate two 3-year period averages for comparison (2004-06 and 2007-09)
• Used recently to help assess School Air Toxics data studyNational Air Monitoring Conference – Nashville, TN Oommen – pg 4
NATTS Sites• This presentation covers data from sites from 2004-2007
National Air Monitoring Conference – Nashville, TN
= Rural Site
2004-06: 23 sites2007: 25 sites2008: 28 sites
Oommen – pg 5
NATTS Sites (all)
National Air Monitoring Conference – Nashville, TN
Site Code AQS Code Location Site Code AQS Code Location
BOMA 25-025-0042 Boston, MA PLORb 41-051-0246 Portland, OR
BTUT 49-011-0004 Bountiful, UT PRRI 44-007-0022 Providence, RI
BXNY 36-005-0110 Bronx, NY PXSS 04-013-9997 Phoenix, AZ
CAMS 35 48-201-1039 Deer Park, TX RIVAb 51-087-0014 Richmond, VA
CAMS 85 48-203-0002 Karnack, TX ROCH 36-055-1007 Rochester, NY
CELAa 06-037-1103 Los Angeles, CA RUCAa 06-065-8001 Rubidoux, CA
CHSC 45-025-0001 Chesterfield, SC S4MO 29-510-0085 St. Louis, MO
DEMI 26-163-0033 Dearborn, MI SDGA 13-089-0002 Atlanta, GA
GLKYb 21-043-0500 Grayson Lake, KY SEWA 53-033-0080 Seattle, WA
GPCO 08-077-0017/18 Grand Junction, CO SJCA 06-085-0005 San Jose, CA
HAKYc 21-193-0003 Hazard, KY SKFL 12-103-0026 Pinellas Park, FL
LAOR 41-061-0119 La Grande, OR SYFL 12-057-3002 Tampa, FL
MVWI 55-027-0007 Mayville, WI UNVT 50-007-0007 Underhill, VT
NBIL 17-031-4201 Northbrook, IL WADC 11-001-0043 Washington, D.C.
a: Added 2007b: Added 2008c: Discontinued 2008
Oommen – pg 6
NATTS Pollutants• Pollutant Groups:
• VOCs
(TO-15)• Carbonyls (TO-11A)• Metals, PM10
(IO-3.5)• Hexavalent
Chromium (EPA-Approved)
• PAHs
(TO-13)
• PAHs
began sampling in 2007 (not included in this presentation)
• Hexavalent
Chromium began sampling in 2005
• Report focuses on HAPs
only
National Air Monitoring Conference – Nashville, TN Oommen – pg 7
NATTS MQO HAPs• Method Quality Objective (MQO) HAPs
• Acetaldehyde• Acrolein• Arsenic• Benzene• Benzo(a)Pyrene• Beryllium• 1,3-Butadiene• Cadmium• Carbon Tetrachloride• Chloroform
• Naphthalene and benzo(a)pyrene
not included in this assessment
National Air Monitoring Conference – Nashville, TN
• Formaldehyde• Hexavalent
Chromium
• Lead• Manganase• Naphthalene• Nickel• Tetrachloroethylene• Trichloethylene• Vinyl Chloride
Oommen – pg 8
Data Retrieval• Phase VI Data Archive:
• Consisted of data from 1973-2007 from AQS, IMPROVE, etc.
• Large HAP archive of 26 million records
• NATTS sites are required to submit data into AQS
• Over 262,000 data records retrieved
• State-level files are posted at:http://www.epa.gov/ttn/amtic/toxdat.html
National Air Monitoring Conference – Nashville, TN Oommen – pg 9
Question 1 –
What was the data completeness for the network?
• 342 possible data sets for 2004-2007:• VOCs: 23 sites * 4 years = 92 data sets• Carbonyls: 23 sites * 4 years = 92 data sets• Metals: 23 sites * 4 years = 92 data sets• Hex Chrome: 22 sites * 3 years = 66 data sets
• 35 data sets did not meet 85% sampling criteria• NATTS Contract Laboratory: 4 data sets• Independent Laboratories: 31 data sets
• 28 expected data sets were not found in AQS
• 8 data sets “invalidated”
due to suspiciously high values
National Air Monitoring Conference – Nashville, TN Oommen – pg 10
Data Completeness -
Overall
National Air Monitoring Conference – Nashville, TN
VOCs Carbonyls Metals Hex Chrome
Period A B C A B C A B C A B C
2004 1,158 1,273 91 1,026 1,167 88 771 819 94 NA NA NA
2005 1,263 1,354 93 1,113 1,251 89 981 1,022 96 874 915 96
2006 1,262 1,329 95 1,096 1,269 86 1,069 1,106 97 983 1,023 96
2007 1,092 1,172 93 1,168 1,372 85 1,134 1,202 94 1,099 1,155 95
Overall 4,775 5,128 93 4,403 5,059 87 3,955 4,149 95 2,956 3,093 96
A = Number of Valid SamplesB = Total Number of SamplesC = Percentage Completeness
Oommen – pg 11
# Data Sets > 85% completeness
National Air Monitoring Conference – Nashville, TN Oommen – pg 12
Invalidated Data Sets by Site
National Air Monitoring Conference – Nashville, TN
Three of twenty-three sites had a data set that was not used in this analysis; a
total of 8 data sets were not used for this analysis.
Oommen – pg 13
Missing Data Sets by Site
National Air Monitoring Conference – Nashville, TN
Ten of twenty-three sites had an expected data set that was not found in AQS; a total of 28 data sets were not found.
Oommen – pg 14
Question 2 –
What HAPs
were detected most?
National Air Monitoring Conference – Nashville, TN Oommen – pg 15
Detections -
Top 20 HAPs
National Air Monitoring Conference – Nashville, TN Oommen – pg 16
Detections –
Middle 16 HAPs
National Air Monitoring Conference – Nashville, TN Oommen – pg 17
Detections –
Bottom 20 HAPs
National Air Monitoring Conference – Nashville, TN Oommen – pg 18
Question 3 –
As a network, what were the central tendencies and data distributions of the HAP data?
National Air Monitoring Conference – Nashville, TN
HAP Detectsa Averageb
(µg/m3)Minimum(µg/m3)
25th(µg/m3)
50th
(µg/m3)75th
(µg/m3)Maximum
(µg/m3)
Formaldehyde 4,741 3.15 ±
0.10 0.01 1.34 2.33 3.97 91.50
Acetaldehyde 4,733 1.94 ±
0.07 0.01 0.88 1.44 2.36 92.78
Benzene 4,482 1.13 ±
0.03 0.01 0.54 0.86 1.39 10.19
Toluene 4,221 2.71 ±
0.25 0.01 0.83 1.66 3.26 482.53
Xylene, m/p- 4,175 1.29 ±
0.05 0.02 0.39 0.78 1.56 21.41
Carbon Tetrachloride 4,060 0.59 ±
0.01 0.03 0.50 0.57 0.66 1.53
Chloromethane 4,059 1.21 ±
0.01 0.03 1.04 1.18 1.34 19.70
a: minimum 4,000 detectsb: includes confidence interval at alpha = 0.05
Oommen – pg 19
Question 4 –
How do the NATTS measurements compare to short-, intermediate-, and long-term health benchmarks?
National Air Monitoring Conference – Nashville, TN
• EPA’s Risk Screening Approach
• ATSDR’s
Minimal Risk Levels (MRLs)
• Cancer/Noncancer
Risk Approximations
Oommen – pg 20
Risk Evaluation –
Risk Screening
National Air Monitoring Conference – Nashville, TN
• Risk Screening Approach •
EPA has published a guidance document outlining a risk
screening approach that utilizes a risk-based methodology to perform an initial screen of ambient air toxics monitoring data sets.
•
This screening process provides a risk-based methodology for analysts and interested parties to identify which pollutants may pose a health risk in their area.
• Not all NATTS pollutants have screening values.
• m/p-Xylene
and o-Xylene
were summed together
Oommen – pg 21
Risk Evaluation –
Risk Screening
National Air Monitoring Conference – Nashville, TN
Pollutant# of Failed
Screens
# ofMeasuredDetections
% ofFailed Screens
% of Total
Failures
Acetaldehyde 4,461 4,733 94.25% 11.09%
Benzene 4,421 4,482 98.64% 10.99%
Carbon Tetrachloride 4,056 4,060 99.90% 10.08%
Formaldehyde 3,977 4,741 83.89% 9.88%
Chromium Compounds 3,423 3,493 98.00% 8.51%
Arsenic Compounds 3,133 3,365 93.11% 7.79%
Butadiene, 1,3- 2,852 3,282 86.90% 7.09%
Tetrachloroethylene 1,894 3,401 55.69% 4.71%
Manganese Compounds 1,769 3,684 48.02% 4.40%
Dichlorobenzene, p- 1,678 2,436 68.88% 4.17%
Acrolein 1,470 1,471 99.93% 3.65%
Nickel Compounds 1,374 3,397 40.45% 3.41%
41 Other HAPs 5,729 71,019 8.07% 14.24%
TOTAL 40,237 113,564 35.43% NA
Oommen – pg 22
Risk Evaluation –
Risk Screening
National Air Monitoring Conference – Nashville, TN
• Summary • Twelve HAPs
contributed to the Top 85% of all failed screens
• Thirteen HAPs
did not fail any screens
• Ten HAPs
failed >90% of their screens:• Acetaldehyde• Benzene• Carbon Tetrachloride• Chromium compounds• Arsenic compounds
• Acrolein• 1,1,2,2-Tetrachloroethane• Ethylene Dibromide• Acrylonitrile• α-Chlorotoluene
Oommen – pg 23
Risk Evaluation –
MRLs
National Air Monitoring Conference – Nashville, TN
• Minimal Risk Levels (MRLs)•
An MRL is a concentration of a hazardous substance that
is “without appreciable risk of adverse noncancer health effects over a specified duration of exposure”
• MRLs
are intended to be used as screening tools
• ATSDR defines MRLs
for three durations of exposure:• acute (1 to 14 days)• intermediate (15 to 364 days)• chronic exposure (1 year or greater).
• Short-term MRL compared to 24-hour measurement
Oommen – pg 24
Risk Evaluation –
MRLs
National Air Monitoring Conference – Nashville, TN
• Intermediate MRL comparison: comparison with the Seasonal Average
• Seasonal Average: “Subseason
Zone”
Approach• e.g. Spring (March 21-June 20)
• Subseason
Zone 1: March 21-March 31• Subseason
Zone 2: April 1-April 12
• Subseason
Zone 3: April 13-April 24• Subseason
Zone 4: April 25-May 6
• Subseason
Zone 5: May 7-May 18• Subseason
Zone 6: May 19-May 30
• Subseason
Zone 7: May 31-June 11• Subseason
Zone 8: June 12-June 20
•
Must have at least 1 detect within six of eight zones to be valid; used NATTS background concentrations or ½
MDL for missing zones
Oommen – pg 25
Risk Evaluation –
non-Chronic
National Air Monitoring Conference – Nashville, TN
Pollutant
Acute Risk Intermediate Risk
ATSDR MRL
(µg/m3)# of
Exceedances
ATSDR MRL
(µg/m3)
# of Winter
Exceedances
# of Spring
Exceedances
# of Summer
Exceedances
# of Autumn
Exceedances
Formaldehyde 50 3/4,741 40 0/80 0/84 0/85 0/84
Acrolein 7 1/1,471 0.09 23/24 25/26 27/27 30/31
Exceedances
of the MRL
Oommen – pg 26
Risk Evaluation –
Chronic MRLs
National Air Monitoring Conference – Nashville, TN
• Chronic MRL comparison: comparison with the Annual Average
• Annual Average: average of at least 3 Seasonal Averages to be valid.
• No
annual averages were higher than its chronic MRL.
•
Sites that did not have enough acrolein
detects to calculate an Annual Average include: BXNY, CAMS 35, CHSC, HAKY, LAOR, MVWI, ROCH, SDGA, UNVT, and WADC.
Oommen – pg 27
Risk Evaluation –
Cancer Risk Approximations
National Air Monitoring Conference – Nashville, TN
•
Cancer risk is defined as the likelihood of developing cancer as a result of exposure over a 70-year period.
•
Presented as the number of people at risk for cancer per million people
•
Cancer risk approximations for 28 HAPs
were calculated in the following manner:
Cancer Risk Approximation = Annual Average * Cancer URE * 10E-6
• Nearly 45% of the Cancer Risk Approximations were ≤
1 in-a-million
Oommen – pg 28
Acetaldehyde Cancer Risk Approximations
National Air Monitoring Conference – Nashville, TN Oommen – pg 29
Benzene Cancer Risk Approximations
National Air Monitoring Conference – Nashville, TN Oommen – pg 30
Carbon Tetrachloride Cancer Risk Approximations
National Air Monitoring Conference – Nashville, TN Oommen – pg 31
Ethylene Dibromide
Cancer Risk Approximations
National Air Monitoring Conference – Nashville, TN Oommen – pg 32
Nickel Cancer Risk Approximations
National Air Monitoring Conference – Nashville, TN Oommen – pg 33
Cumulative Cancer Risk Approximations
National Air Monitoring Conference – Nashville, TN Oommen – pg 34
Risk Evaluation –
Noncancer
Risk Approximations
National Air Monitoring Conference – Nashville, TN
• Noncancer
risk is presented as the Noncancer
Hazard Quotient (HQ).
•
Noncancer
health effects target certain areas, such as respiratory, neurological, kidneys, and liver systems.
•
If the HQ is calculated to be less than 1.0, then no adverse health effects are expected as a result of exposure, if the HQ is greater than 1.0 the adverse health effects are possible.
•
Noncancer
risk approximations for 10 target areas were calculated in the following manner:
Noncancer
Risk HQ Approximation = Annual Average ÷
Noncancer
RfC
÷
1000
Oommen – pg 35
Neurological Noncancer
Risk Approximations
National Air Monitoring Conference – Nashville, TN Oommen – pg 36
Reproductive Noncancer
Risk Approximations
National Air Monitoring Conference – Nashville, TN Oommen – pg 37
Respiratory Noncancer
Risk Approximations
National Air Monitoring Conference – Nashville, TN Oommen – pg 38
Question 5 –
What were the 3-year period averages?
National Air Monitoring Conference – Nashville, TN
• 3-Yr Period Average:• Average of annual averages• Need to have 2 annual averages to be valid• Could only be conducted for 2004-2006 timeframe• Conducted for only the MQO HAPs• Compared to the 2007 average for preliminary assessment
•
Results: Over 75% of the comparisons realized a decrease in period average concentrations (not statistically significant)
•
Problem: Potential of nearly 400 period averages…only 200 were calculated.
Oommen – pg 39
Question 6 –
How can the results of this assessment be improved?
National Air Monitoring Conference – Nashville, TN
• Challenges with NATTS Data:• Parameter Occurrence Codes: multiple to choose from
• Missing pollutant group data sets
• Questionable data were invalidated
• Under-reporting of pollutants
• MDL Reporting
• MDL Ranges
Oommen – pg 40
National Air Monitoring Conference – Nashville, TN
2004 HAKY = 1,283.62005 HAKY = 1,542.5
0.0
5.0
10.0
15.0
20.0
25.0
BO
MA
BTU
T
BX
NY
CA
MS
35
CA
MS
85
CH
SC
DEM
I
GPC
O
HA
KY
LAO
R
MV
WI
NB
IL
PRR
I
PXSS
RO
CH
S4M
O
SDG
A
SEW
A
SJC
A
SKFL
SYFL
UN
VT
WA
DC
Can
cer
Ris
k A
ppro
xim
atio
n (in
-a-m
illio
n)
2004200520062007
0.0
1.0
2.0
3.0
4.0
5.0
BO
MA
BTU
T
BX
NY
CA
MS
35
CA
MS
85
CH
SC
DEM
I
GPC
O
HA
KY
LAO
R
MV
WI
NB
IL
PRR
I
PXSS
RO
CH
S4M
O
SDG
A
SEW
A
SJC
A
SKFL
SYFL
UN
VT
WA
DC
Can
cer
Ris
k A
ppro
xim
atio
n (in
-a-m
illio
n)2004200520062007
Questionable Data -
Examples
1,1,2-Trichloroethane cancer risk at LAOR: >4 in-a-million
Ethylene dibromide
cancer risk at HAKY: >1200 in-a-million
Oommen – pg 41
Under-reporting of Specific HAPs
Example
National Air Monitoring Conference – Nashville, TN
Benzene Detects by Site
Acrolein
Detects by Site
Oommen – pg 42
Number of HAPs
by Site
National Air Monitoring Conference – Nashville, TN
Under-reporting of Multiple HAPs
Example
Oommen – pg 43
National Air Monitoring Conference – Nashville, TN
MDL Reporting/Ranges• MDL Reporting:
•
Over 42% of the Independent Laboratories MDL values used in this analysis report were not populated in AQS.
• NATTS Contract Laboratory: 100%
• Selected MDL Ranges
Pollutant Units Minimum MDL
Maximum MDL
Average MDL
Average Federal MDL
Acrolein µg/m3 0.002 5.81 0.212 0.079
Benzene µg/m3 0.011 1.597 0.132 0.087
Methylene
Chloride µg/m3 0.035 17.368 0.637 0.322
Acetaldehyde ng/m3 0.004 1.333 0.220 0.221
Arsenic ng/m3 0.003 22 6.834 7.461
Manganese ng/m3 0.004 1.2 0.358 0.295
Nickel ng/m3 0.009 4 0.95 0.814
Oommen – pg 44
National Air Monitoring Conference – Nashville, TN
Next Steps• September 30, 2009 –
Draft 2004-2007 Assessment sent to EPA
• November 4, 2009 –
Report can be e-mailed to states/EPA-regional offices.
•
November 4, 2009 to December 31, 2009 –
Comments due/Opportunity for new/corrected data to be uploaded into AQS.
• December 2009 –
Quarterly NATTS conference call
• January 2, 2010 –
AQS data pull for NATTS data
• February 28, 2010 –
Finalize 2004-2007 Assessment
• August 2010 –
AQS data pull for 2008 and 2009 NATTS data
•
October 2010 –
Draft 2004-2009 Assessment completed; NATTS Network Assessment
Oommen – pg 45
National Air Monitoring Conference – Nashville, TN
Questions/Comments?Regi
OommenEastern Research Group, Inc.
Morrisville, NC919-468-7829
Barbara DriscollU.S. EPARTP, NC
Mike JonesU.S. EPARTP, NC
Oommen – pg 46