evaluation of atmospheric impacts of coatings vocs · pdf filew. p. l. carter 04/11/2005...
TRANSCRIPT
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 1
EVALUATION OF ATMOSPHERICIMPACTS OF COATINGS VOCs
William P. L. CarterCE-CERT, University of California, Riverside
April 13, 2005
Outline• Background and Objectives• Environmental Chamber Experiments – Ozone Impacts• Environmental Chamber Experiments – PM Impacts• Exploratory Glycol Availability Experiments• Estimation of Hydrocarbon Solvent Reactivities• Direct Reactivity Measurement• Summary and Conclusions
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 2
Recent UCR Coatings Reactivity ProjectsEvaluation of Atmospheric Impacts of Selected Coatings VOCs• CARB Contract 00-333
• Objective: Reduce uncertainties in estimations of ozone impactsof coatings VOCs
• Final report at http://www.cert.ucr.edu/~carter/coatings
Environmental Chamber Studies of VOC Species inArchitectural Coatings and Mobile Source Emissions• SCAQMD Contract No. 03468
• Relevant Objectives:• Evaluate O3 impacts of selected water-based coatings VOCs• Determine PM impacts for Coating VOCs studied for CARB• Evaluate use of chamber for availability studies
• Final report in preparation
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 3
Components of Coatings Projects• Environmental chamber studies
• Six complex hydrocarbon solvents and four water-basedcoatings VOC compounds chosen for study
• UCR EPA chamber employed• Chamber results used to evaluate ozone reactivity
predictions of the SAPRC-99 mechanism• PM measurement results used to derive qualitative
estimates of relative PM impacts of solvents studied• Exploratory studies of effects of aerosol and humidity on
glycol availability
• Development and evaluation of general procedures to estimatereactivities of complex hydrocarbon solvents
• Further development and evaluation of direct reactivitymeasurement methods
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 4
Measurement or Calculationof Ozone Reactivities of VOCs
• VOC Reactivities measured in chamber experiments are notexactly the same as VOC reactivities in the atmosphere.• Impractical to duplicate all relevant conditions• Chamber experiments have wall effects, static conditions,
higher levels of test VOCs, etc.
• Atmospheric Ozone impacts of VOCs must be calculated usingcomputer airshed models, given:• Models for airshed conditions• Chemical mechanism for VOC’s Atmospheric Reactions
• BUT mechanisms have uncertainties. reactivity calculations canbe no more reliable than the chemical mechanism used.
• Therefore, the purpose of chamber experiments is to testthe ability of the mechanisms to predict reactivity in models
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 5
Diagram of UCR EPA Chamber
20 ft. 20 ft.
20 ft.
This volume kept clear to maintain light
uniformity
Temperature controlled room flushed with purified air and with reflective
material on all inner surfaces
Dual Teflon Reactors
Two air Handlers are located in the corners on each side of the light
(not shown).
Gas sample lines to laboratory below Access
Door
200 KW Arc Light
2 Banks of Blacklights
SEMS (PM) Instrument
Floor Frame
Movable top frame allows reactors to collapse
under pressure control
Mixing SystemUnder floor of reactors
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 6
Photographs of Chamber and LightsLooking Towards Reactors (from light) Looking Towards Lights and Air Inlet
ArcLight
PartiallyFilledReactors
BlackLights
Reflectivewalls
AirIntake
(Alsoone on
otherside)
Reflectivewalls and
Ceiling
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 7
Incremental Reactivity Experiments
• Objective is to determine effects of VOC’s reactions in chemicalenvironments representing a range of atmospheric conditions.
• Approach is to conduct two simultaneous experiments in thedual chamber• Base Case Experiment: Irradiate surrogate ROG – NOx
mixture simulating an ambient chemical environment• Test Experiment: Same as base case experiment except
that a test compound or solvent added
• Effect of added VOC on O3, radicals, etc, provides a means totest model predictions reactivities of VOCs under similarconditions in the atmosphere
• Base case experiments should reflect range of atmosphericchemical conditions relevant to VOC reactivity.
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 8
Choice of Base Case forIncremental Reactivity Experiments
• Major atmospheric chemical condition relevant to VOC reactivityis relative availability of NOx (relative ROG/NOx ratio). E.g.,
• Different aspects of the mechanism affect O3 impacts underdifferent NOx conditions• High NOx experiments: test effects of VOCs on radical levels• Low NOx experiments: test effects of VOCs on NOx removal
• Therefore, minimum of two base case experiments is needed
High
Moderate
Low
ROG/NOx
Most sensitive to NOxLowLow NOx
Optimum conditions for O3ModerateMOIR
Most sensitive to VOCsHighMIR
O3 sensitivityRelative NOxCondition
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 9
Base Case Experimentsused in Coatings Reactivity Studies
• NOx levels of 25-30 ppb, based on CARB recommendations ofrange of NOx that represents urban conditions in California
• 8- component ROG surrogate (used previously) employed torepresent major classes of VOCS present in ambient air• Formaldehyde removed in later experiments and other
VOCs increased by 10% to simplify experiments• Calculated to give essentially the same reactivity results as
base ROG mixture used to calculated reactivity scales.
½ MOIR NOx levels. NOx relatively low butnot so low that O3 insensitive to VOCs.
1.025MOIR/2
ROG levels calculated to yield MIRconditions
0.530MIR
DiscussionROG(ppmC)
NOx(ppb)Experiment
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 10
SAPRC-99 Model simulations of representative experiments
0.00
0.04
0.08
0.12
0.16
0 60 120 180 240 300 360
Time (minutes)
Ozo
ne (p
pm)
ModelEPA095BEPA123AEPA124BEPA126A
Experimental and Calculated O3 forRepresentative Base Case Experiments
0.00
0.04
0.08
0.12
0.16
0 60 120 180 240 300 360
Time (minutes)
Ozo
ne (p
pm)
Model
EPA110B
EPA114A
EPA127B
MOIR/2 BASE CASE(NOx=25 ppb, ROG=1 ppmC)
MIR BASE CASE(NOx=30 ppb, ROG=0.5 ppmC)
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 11
Dependence of SAPRC-99 UnderpredictionBias on Relative ROG/NOx Levels
-20%
0%
20%
40%
60%
80%
0.1 1 10
ROG / NOx Ratio Relative to ROG / NOx Giving Maximum O3
([O3 ]-
[NO
]) M
odel
und
erpr
edic
tion
erro
r(E
xper
imen
tal -
Mod
el) /
Exp
erim
enta
l
UCR EPA ChamberCSIRO Outdoor ChamberTVA Chamber
Maximum O3 (MOIR)Approximate MIR
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 12
Aromatic Mechanism Adjustmentsto Improve Base Case Simulations
• Biases in model simulations of base case experiment attributedto aromatics mechanism. So far, no aromatic mechanism foundthat gives satisfactory fits all the available chamber data.
• Base case biases may cause biases in simulations of reactivity.
• To investigate this, mechanisms for aromatics in the base ROGwas adjusted to remove biases in base case simulations.• Yields of aromatic fragmentation products AFG2 and AFG3
for toluene and m-xylene increased by factor of 1.75• Rate of reaction of AFG1 with O3 increased by factor of 10
• Comparing simulations with and without this adjustment showseffects base mechanism biases on reactivity predictions
• Not a “better” aromatics mechanism because this adjustmentmakes fits to single aromatic – NOx experiments worse.
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 13
Effects of Mechanism Adjustments onSimulations of Representative Experiments
∆([O
3]-[N
O])
(ppm
)
Surrogate - NOx Runs M-Xylene - NOx RunsEPA188A: Low ROG/NOx (MIR) CTC036: Low ROG/NOx
EPA124B: High ROG/NOx (MOIR/2) CTC029: High ROG/NOx
Time (minutes)
0.0
0.2
0.4
0.6
0 60 120 180 240 300 360
0.0
0.2
0.4
0.6
0 60 120 180 240 300 360
0.00
0.04
0.08
0.12
0.16
0 60 120 180 240 300 360
ExperimentalStandard ModelAdjusted Model
0.00
0.04
0.08
0.12
0.16
0 60 120 180 240 300 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 14
Effect of Aromatic Mechanism Adjustmenton Base Case Model Underprediction Bias
-30%
-20%
-10%
0%
10%
20%
30%
40%
0 10 20 30 40 50 60ROG/NOx
Mod
el u
nder
pred
ictio
n Bi
as
Standard Model Adjusted Aromatics Model
MIR MOIR/2
∆([O
3]-[N
O])
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 15
VOCs Identified in the 2001 CARB Surveyof Water-Based Architectural Coatings
O
O OH
OO
OH
Mixtures of C8-C12 alkanes andaromatics4%Various Hydrocarbon
Solvents
3%Butyl Carbitol;2-(2-Butoxyethoxy)-Ethanol
(2)
3%Benzyl Alcohol
21%Ethylene Glycol
25%Propylene Glycol
(1)31%
Texanol®;Isobutyrate esters of 2,2,4-Trimethylpentyl-1,3-diol
StructuresMass %Compound
OHOH
OHOH
OH O
O
OH
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 16
Results of a Texanol® Injection Test
0.0
0.2
0.4
0.6
0.8
1.0
0 5 10 15 20 25Hours after injection
ppm
C T
exan
ol o
r TH
C
FID Total Hydrocarbon Analyzer DataCalculated amount injected, corrected for dilution using CO dataGC Analysis, Based on liquid calibration (sum of two GC peaks)
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 17
0.0
0.2
0.4
0.6
0.8
1.0
0.0 0.2 0.4 0.6 0.8Ln ([m-Xylene]0 / [m-Xylene]t)
ln([T
est V
OC
]0 /
[[Tes
t VO
C]t )
Texanol 1
Texanol 2 (offset)
Butyl Carbitol / 2 (offset)
Best fit to data
SAPRC-99 Estimate
OH Radical Rate ConstantsDerived from Chamber Data
4.29 x 10-11Butyl Carbitol
1.62 x 10-11Texanol 2
1.29 x 10-11Texanol 1
kOH (cm3
molec-1 s-1)Compound
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 18
Reactivity Data for Texanol®∆[(O3]-[NO]) (ppm) ∆[(O3]-[NO]) Change (ppm) IntOH (ppt-min)
Time (minutes)
EP
A22
9A (M
IR)
EP
A23
2A (M
OIR
/2)
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360 -0.03
-0.02
-0.01
0.00
0.01
0 60 120 180 240 300 360
0
10
20
30
40
0 60 120 180 240 300 360
Test Experiment Standard Model Standard Base ModelBase Experiment Adjusted Aromatic Model
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360 -0.02
-0.01
0.00
0.01
0 60 120 180 240 300 360
0
20
40
60
0 60 120 180 240 300 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 19
Comparison of Chamber and AmbientReactivity Calculation For Texanol®
"MIR" Incremental Reactivity Chamber experiment
"Averaged Conditions" MIRBox Model Airshed Scenario
Both Cases: Moles Texanol added = 5% of moles Carbon in Base Case ROGsBoth simulations predict measurable effect of Texanol on OH radical levels.
0.00
0.05
0.10
0.15
0 2 4 6Irradiation time (hour)
Ozo
ne (p
pm)
Base Case
Added Texanol
0.00
0.05
0.10
0.15
0.20
0.25
8 10 12 14 16 18Simulated Local Time (hour)
Ozo
ne (p
pm)
Base Case
Added Texanol
Two curvesalmost on topof each other
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 20
Reactivity Data for Butyl Carbitol∆[(O3]-[NO]) (ppm) ∆[(O3]-[NO]) Change (ppm) IntOH (ppt-min)
Time (minutes)
EP
A35
2B (M
IR)
EP
A35
3B (M
OIR
/2)
Test Experiment Standard Model Standard Base ModelBase Experiment Adjusted Aromatic Model
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360 -0.02
-0.01
0.00
0.01
0.02
0 60 120 180 240 300 360
0
20
40
60
0 60 120 180 240 300 360
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360 -0.02
-0.01
0.00
0.01
0 60 120 180 240 300 360
0
10
20
30
40
0 60 120 180 240 300 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 21
Reactivity Data for Propylene Glycol∆[(O3]-[NO]) (ppm) ∆[(O3]-[NO]) Change (ppm) IntOH (ppt-min)
Time (minutes)
EP
A27
7A (M
IR)
EP
A25
2A (M
OIR
/2)
Test Experiment Standard Model Standard Base ModelBase Experiment Adjusted Aromatics Model
0.00
0.05
0.10
0.15
0.20
0.25
0 60 120 180 240 300 3600.00
0.02
0.04
0.06
0 60 120 180 240 300 3600
20
40
60
80
0 60 120 180 240 300 360
0.00
0.05
0.10
0.15
0.20
0.25
0 60 120 180 240 300 3600.00
0.02
0.04
0.06
0 60 120 180 240 300 3600
10
20
30
40
0 60 120 180 240 300 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 22
Reactivity Data for Ethylene Glycol∆[(O3]-[NO]) (ppm) ∆[(O3]-[NO]) Change (ppm) IntOH (ppt-min)
Time (minutes)
EP
A27
8B (M
IR)
EP
A25
0B (M
OIR
/2)
Test Experiment Standard Model Standard Base ModelBase Experiment Adjusted Aromatics Model
0.0
0.1
0.2
0.3
0 60 120 180 240 300 3600.00
0.02
0.04
0.06
0.08
0 60 120 180 240 300 3600
20
40
60
0 60 120 180 240 300 360
0.0
0.1
0.1
0.2
0.2
0.3
0 60 120 180 240 300 3600.00
0.01
0.02
0.03
0.04
0.05
0 60 120 180 240 300 3600
10
20
30
40
0 60 120 180 240 300 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 23
Glycol Decay Rates in Reactivity Runs:Comparison with Literature k(OH) Values
Propylene Glycol vs m-Xylene Ethylene Glycol vs. m-Xylene
ln([m-Xylene]0/[m-Xylene]t)
ln([G
lyco
l] 0/[G
lyco
l] t)
-0.1
0.1
0.3
0.5
0.7
0.9
1.1
1.3
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Single Run Single Run Single RunSingle Run Single RunNot used for kOH Fit Best kOH fit to Data Literature kOH Line
0.0
0.2
0.4
0.6
0.8
0.0 0.3 0.6 0.9 1.2
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 24
Glycol Reactivity Data with aNon-Aromatic Surrogate
∆[(O3]-[NO]) (ppm) ∆[(O3]-[NO]) Change (ppm) IntOH (ppt-min)
Time (minutes)
Pro
pyle
ne G
lyco
lE
thyl
ene
Gly
col
Test Experiment Standard ModelBase Experiment Standard Base Model
0.00
0.04
0.08
0.12
0.16
0 60 120 180 240 300 3600.00
0.02
0.04
0.06
0.08
0 60 120 180 240 300 3600
5
10
15
20
25
0 60 120 180 240 300 360
0.00
0.04
0.08
0.12
0.16
0 60 120 180 240 300 3600.00
0.02
0.04
0.06
0 60 120 180 240 300 3600
10
20
30
0 60 120 180 240 300 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 25
Development of aBenzyl Alcohol Mechanism
• Reaction with OH radicals assumed to dominate.
• Single measurement of k(OH) given by Atkinson (1989) used
• Reaction at -CH2OH, forming Benzaldehyde + HO2 assumed tooccur 30% of the time, to fit benzaldehyde data in experiments
• Mechanism of OH addition to ring based on that used fortoluene
• Overall nitrate yield adjusted to be 5% to give best fits to data
• Benzaldehyde – NOx experiments also carried out to provideadditional basis for developing adjusted mechanism.
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 26
Reactivity Data for Benzyl Alcohol∆[(O3]-[NO]) (ppm) ∆[(O3]-[NO]) Change (ppm) IntOH (ppt-min)
Time (minutes)
EP
A32
3B (M
IR)
EP
A30
2 (M
OIR
/2)
Test Experiment Standard Model Standard Base ModelBase Experiment Adjusted Aromatic Model
0.00
0.05
0.10
0.15
0 60 120 180 240 300 360 -0.02
0.00
0.02
0.04
0.06
0 60 120 180 240 300 360 0
20
40
60
0 60 120 180 240 300 360
0.00
0.04
0.08
0.12
0.16
0 60 120 180 -0.005
0.000
0.005
0.010
0.015
0.020
0.025
0 60 120 1800
10
20
30
0 60 120 180
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 27
Simulations of RepresentativeBenzaldehyde – NOx Experiments
∆([O3]-[NO]) (ppm) Benzyl Alcohol (ppm) Benzaldehyde (ppm)EPA322A (0.41 ppm Benzyl alcohol, 26 ppb NOx)
EPA325A (0.27 ppm Benzyl alcohol, 55 ppb NOx)
Irradiation time (minutes)
0.000
0.005
0.010
0.015
0.020
0.025
0 120 240 3600.0
0.1
0.2
0.3
0 120 240 360
ExperimentalModel
0.0
0.1
0.2
0.3
0.4
0 120 240 360
0.00
0.05
0.10
0.15
0.20
0.25
0 120 240 360
0.0
0.1
0.2
0.3
0 120 240 3600.00
0.01
0.02
0.03
0 120 240 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 28
Summary of Mechanism Evaluation Resultsfor Water-Based Coatings VOCs
2.9Previous mechanism simulated datasatisfactorily. Not changed.
Butyl Carbitol
4.9No previous mechanism. Parameter-ized mechanism adjusted to fit data.
Benzyl Alcohol
≥ 3.4Mechanism may underpredict ozoneimpact, but uncertain whether changeis appropriate. Not changed.
Ethylene Glycol
≥ 2.7Mechanism may underpredict ozoneimpact, but uncertain whether changeis appropriate. Not changed.
Propylene Glycol
0.88Previous mechanisms simulated datasatisfactorily. Not changed.
Texanol® isomers
MIR *(mass basis)
Mechanism Performance andModificationsCompound
* MIR of base ROG (Ambient Mixture) = 3.7 gm O3 / gm VOC
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 29
Representative Hydrocarbon MixturesChosen For Reactivity Experiments
11-9-12Dearomatized MixedAlkanes
ASTM-1C
60.2%8-9VMP NaphthaVMP-NAPH
AROM-100
ASTM-1A
ASTM-1B
ASTM-3C1
Designation
12-Mostly 11Synthetic isoparrafinicalkane mixture
22
15
14
CARBBin No.
AromaticContent
Carbon No.Range
Description
100%Mostly 9Aromatic 100
19%9-12Regular Mineral Spirits
6%9-12Reduced AromaticsMineral Spirits
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 30
Chemical Type Distributionsfor Hydrocarbon Mixtures Studied
n-Alkane Br-Alkane Cyc-Alkane Aromatic
Aromatic 100ASTM-1AASTM-1B
ASTM-3C1ASTM-1CVMP Naphtha
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 31
Reactivity Data for Dearomatized MixedAlkanes (ASTM-1C)
∆[(O3]-[NO]) (ppm) ∆[(O3]-[NO]) Change (ppm) IntOH (ppt-min)
Time (minutes)
EP
A16
8A (M
IR)
EP
A15
2B (M
OIR
/2)
Test Experiment Standard Model Standard Base ModelBase Experiment Adjusted Aromatic Model
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360 -0.06
-0.04
-0.02
0.000 60 120 180 240 300 360
0
20
40
60
0 60 120 180 240 300 360
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360 -0.05
-0.04
-0.03
-0.02
-0.01
0.000 60 120 180 240 300 360
0
10
20
30
40
0 60 120 180 240 300 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 32
Comparison of Chamber and AmbientReactivity Calculation for ASTM-1C
"MIR" Incremental Reactivity Chamber experiment
"Averaged Conditions" MIRBox Model Airshed Scenario
Moles Carbon Mixture added = 25% of Moles Carbon in Base Case ROGsBoth simulations predict measurable effect of Mixture on OH radical levels.
0.00
0.05
0.10
0.15
0 2 4 6Irradiation time (hour)
Ozo
ne (p
pm)
Base Case
Added ASTM1C
0.00
0.05
0.10
0.15
0.20
8 10 12 14 16 18Simulated Local Time (hour)
Ozo
ne (p
pm)
Base Case
Added ASTM1C
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 33
Reactivity Data for Aromatic 100∆[(O3]-[NO]) (ppm) ∆[(O3]-[NO]) Change (ppm) IntOH (ppt-min)
Time (minutes)
EP
A24
4B (M
IR)
EP
A23
9A (M
OIR
/2)
Test Experiment Standard Model Standard Base ModelBase Experiment Adjusted Aromatic Model
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 3600.00
0.02
0.04
0.06
0 60 120 180 240 300 3600
20
40
60
80
0 60 120 180 240 300 360
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360 -0.04
-0.02
0.00
0.02
0.04
0 60 120 180 240 300 360
0
10
20
30
40
0 60 120 180 240 300 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 34
Results for Other Petroleum DistillatesChange in ∆([O3]-[NO]) (ppm)
VMP Naphtha ASTM-1B ASTM-1A
Time (minutes)
MIR
MO
IR/2
Test Experiment Standard Model Standard Base ModelBase Experiment Adjusted Aromatic Model
EPA167A
-0.02
-0.01
0.00
0.01
0.02
0 60 120 180 240 300 360
EPA153A
-0.03
-0.02
-0.01
0.000 60 120 180 240 300 360
EPA151A
-0.04
-0.03
-0.02
-0.01
0.000 60 120 180 240 300 360
EPA242B
-0.03
-0.02
-0.01
0.000 60 120 180 240 300 360
EPA238A
-0.05
-0.04
-0.03
-0.02
-0.01
0.000 60 120 180 240 300 360
EPA243B
-0.03
-0.02
-0.01
0.000 60 120 180 240 300 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 35
Reactivity data for SyntheticHydrocarbon Mixture (ASTM-3C1)
∆[(O3]-[NO]) (ppm) ∆[(O3]-[NO]) Change (ppm) IntOH (ppt-min)
Time (minutes)
EP
A16
3A (M
IR)
EP
A23
7B (M
OIR
/2)
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360 -0.05
-0.04
-0.03
-0.02
-0.01
0.000 60 120 180 240 300 360
0
10
20
30
40
0 60 120 180 240 300 360
Test Experiment Standard Model Standard Base ModelBase Experiment Adjusted Aromatic Model
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360 -0.05
-0.04
-0.03
-0.02
-0.01
0.000 60 120 180 240 300 360
0
20
40
60
0 60 120 180 240 300 360
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 36
Assessment of Model Performance andMIRs for the Hydrocarbon Solvents Studied
7.51
1.82
1.26
0.81
0.91
1.41
BinMIR
Consistent with data forMIR. Underpredicts O3inhibition at low NOx
Reasonably consistentwith data.
Reasonably consistentwith data
Model underestimatesO3 impact by 40-80%
Reasonably consistentwith data
Reasonably consistentwith data
O3 Model Performance
7.70
1.97
1.21
Approx.1.1 – 1.5
0.96
1.35
Best Est.MIR
C9-C12 AlkanesASTM-1C
C8-C9 AlkanesVMPNaphtha
ComponentsDesignation
Methyl ethyl andtrimethylBenzenes
Aromatic100
C9-C12 Alkanes,~20% AromaticsASTM-1A
C9-C12 Alkanes,~6% AromaticsASTM-1B
C11 BranchedAlkanesASTM-3C1
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 37
PM Measurements
• Number densities of particles in 71 size ranges (28 - 730 nm)measured using a a Scanning Electrical Mobility Spectrometer
• Data used to compute total particle number and volume(measured as mass assuming density of H2O) per unit volume
• PM alternately sampled from each of the two reactors, switchingevery 10 minutes (15.3 data points/hour/reactor)
• PM measurements made during most incremental reactivityexperiments for the coatings projects
• Background PM measurements made in experiments where PMprecursors not expected
• Seed aerosol not used in most experiments
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 38
PM Data in Base Case Experiment
EPA233: MOIR/2 Surrogate (Side Equivalency Test)PM Number (cm-3) PM Volume (µg/m3)
Irradiation time (hours)
0
2000
4000
6000
8000
10000
12000
14000
0 1 2 3 4 5 60.0
0.5
1.0
1.5
2.0
0 1 2 3 4 5 6
Side A
Side B
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 39
PM Volume in Background Experiments 5
Hou
r PM
Vol
ume
(µg/
m3 )
Side A Side B
EPA Run Number
0.0
0.5
1.0
1.5
2.0
150 250 350
Pure Air Propene - NOx CO - Air CO - NOx
0.0
0.5
1.0
1.5
2.0
150 250 350
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 40
Effects of Texanol® and the Glycolson 5-Hour PM Volume
5 H
our P
M V
olum
e (µ
g/m
3 )
Side A Side B
EPA Run Number
0.0
0.3
0.6
0.9
1.2
150 200 250 300 3500
1
2
3
4
125 175 225 275 325
Base Experiment Base Fit or Avg. Base SDev.
Texanol Propylene Glycol Ethylene Glycol
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 41
Effects of Hydrocarbon Solvents on5 Hour PM Volume
5 H
our P
M V
olum
e (µ
g/m
3 )
Side A Side B
EPA Run Number
0
2
4
6
8
125 175 225 275 325
Base Experiment Base Fit or Avg. Base SDev.VMP Naphtha ASTM-1C ASTM-3C1ASTM-1B ASTM-1A Aromatics 100
0
1
2
3
4
5
125 175 225 275 325
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 42
Effects of Benzyl Alcohol and Butyl Carbitolon 5-Hour PM Volume
5 H
our P
M V
olum
e (µ
g/m
3 )
Side A Side B
EPA Run Number
0
10
20
30
40
150 200 250 300 350
Base Experiments Base Fit or AverageBenzyl Alcohol Butyl Carbitol
4-hour
0
5
10
15
20
25
30
150 250 350
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 43
Summary of PM Volume Reactivity Results
0 2 4 6 8 10 12 14 16 18 20
Base Case
Propylene Glycol
Ethylene Glycol
Texanol®
ASTM-3C1
ASTM-1C
VMP Naphtha
ASTM-1B
ASTM-1A
Aromatic 100
Butyl Carbitol *
Benzyl Alcohol /2
Average 5 Hour PM Volume µg/m3
Side B Side A
Butyl Carbitol Side A is for 4 hours.
Error Bars are 1-σ standard deviations.No error bar means only one experiment.
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 44
Summary of PM Measurement Results
• Background PM formation in chamber is up to ~1 µg/m3,depending on reactor employed• Probably due to contaminant reacting with OH, forming SOA• Reason for higher background in “A” reactor unknown
• Secondary PM formation from ethylene and propylene glycol,Texanol®, and the ASTM-3C1 synthetic mixture negligible.
• Small but measurable PM from petroleum distillate solvents. Notsimply related to aromatic content.
• Highest secondary PM from butyl carbitol and (especially)benzyl alcohol
• Chamber effects PM model needed before PM data can beused for quantitative mechanism evaluation
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 45
Glycol Availability Screening ExperimentsObjective
• Determine if added aerosols affect gas-phase consumptionrates and reactivities of glycols
Experiments Carried Out
• Dark decay experiments with ethylene and propylene glycol with10 µg/m3 (NH4)2SO4 seed aerosol at 35% RH.
• ROG – NOx ambient surrogate irradiation with added propyleneglycol with 9 µg/m3 (NH4)2SO4 seed aerosol at 25% RH.
• ROG – NOx ambient surrogate irradiation with added ethyleneglycol with 7 µg/m3 NH4HSO4 seed aerosol at 30% RH.
Note: Aerosol and humidity added to only one reactor becauseof limited aerosol generation and humidification capacity.
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 46
Results of Glycol Dark Decay Experiment
Time after first sample (hours)
Con
cent
ratio
n (p
pm)
35% RH10 µg/m3
(NH4)2SO4Seed Aerosol
0.0
0.1
0.2
0 2 4 6
Dry,No Aerosol
0.0
0.1
0.2
0.3
0 2 4 6
Ethylene Glycol
Propylene Glycol
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 47
Propylene Glycol Ethylene GlycolDry, no Aerosol
25% RH, 9 µg/m3 (NH4)2SO4 30% RH, 7 µg/m3 NH4HSO4
Irradiation time (minutes) `
0.00
0.05
0.10
0.15
0.20
0.25
0 60 120 180 240 300
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360
Test Data
Base Data
Test Model
Base Model
0.00
0.05
0.10
0.15
0.20
0.25
0 60 120 180 240 300 360
0.00
0.05
0.10
0.15
0.20
0 60 120 180 240 300 360
Effects of Aerosol and Humidity on GlycolReactivity Experiments
Model calculationsused adjustedaromatics basemechanism for bestfits.
No base data inadded aerosol runsbecause of limitedhumidification andaerosol generationcapacity.
∆([O
3]-[N
O])
(ppm
)
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 48
Glycol Decay Rates in Availability Runs:Comparison with Literature k(OH) Values
Propylene Glycol vs m-Xylene Ethylene Glycol vs. m-Xylene
ln([m-Xylene]0/[m-Xylene]t)
ln([G
lyco
l] 0/[G
lyco
l] t)
0.0
0.2
0.4
0.6
0.8
0.0 0.2 0.4 0.6 0.8
Humidified Seed Aerosol Run Literature kOH Line
0.0
0.2
0.4
0.6
0.8
0.0 0.3 0.6 0.9 1.2
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 49
Glycol Availability Experiments:Preliminary Conclusions
• No clear effect on glycol consumption rate or ozone reactivity forhumidity up to 35% and (NH4)2SO4 or NH4HSO4 seed aerosol upto 10 µg/m3.
• But there still may be a measurable effect at higher humidity oraerosol concentration, with a different type of aerosol
• Upgrades are being made to the chamber facility to facilitateexperiments at higher RH, aerosol levels.
• But experiments that measure increases in aerosol mass whenexposed to gas-phase VOCs may give a more sensitivemeasure of VOC uptake on aerosols
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 50
Evaluation of Methods to EstimateComplex Hydrocarbon Solvent Reactivities
• MIRs for complex hydrocarbon solvents currently estimatedusing a “binning” procedure based on correlations betweencarbon numbers, type distributions, and MIRs
• Bin MIRs evaluated by comparison with MIRs calculated usingdetailed compositional data for a wide variety of solvents• Agree within ±25% except for bins with light cycloalkanes
• An alternative “spreadsheet” method developed for derivingestimated compositions for solvents with limited data• Separates compositional and reactivity estimates. Permits
derivations in reactivities for other scales besides MIR.• Agrees with calculations using detailed compositional data
within ±10% in most cases• Can be used as a basis for updating hydrocarbon bin
reactivities when reactivity scale is changed or updated.
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 51
Comparison of CARB Bin MIRs withMIRs Calculated Using Compositional Data
-75%
-50%
-25%
0%
25%
50%
75%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Bin Number
Bin
Ass
ignm
ent
Rel
aitv
e to
Det
aile
d C
alcu
latio
n
Analyzed Solvent Bin Average
Bins 1, 3-5 are mixtureswith cycloalkanes in thelowest boiling point range
No data for bins 18-20
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 52
Comparison ofSpreadsheet Estimated MIRs with
MIRs Calculated Using Compositional Data
-20%-15%-10%-5%0%5%
10%15%20%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Bin Number
Spr
eads
heet
Cal
cula
tion
Rel
aitv
e to
Det
aile
d C
alcu
latio
n
Analyzed Solvent Average
Note that previous plothad ± 75% range
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 53
Further Development of aDirect Reactivity Measurement Method
• VOCs affect O3 directly through their own reactions or indirectlythrough the effects of their reactions on radicals and NOx.
• A measurement of direct reactivity would reduce uncertainties inmechanism evaluations and provide a reactivity screening tool
• A direct reactivity measurement method was developed in aprevious CARB project but was not suitable for coatings VOCs• Required GC analysis, so not suitable for complex mixtures
or low volatility compounds
• For this project, a total carbon measurement was interfaced tothe system to eliminate the need for GC analysis.
• Problems encountered. Absolute direct results not consistentwith model predictions; but better agreement with relative results
• Resources for this task exhausted before it could be completed.
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 54
Direct Reactivity MeasurementsRelative to n-Octane
0.0 0.5 1.0 1.5
Mineral Spirits
n-C14
n-C12
n-Octane
Propane (x 3)
Direct Reactivity Relative to n-Octane
Model Experimental Error bars show range of variability of experiments
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 55
Overall Conclusions of Chamber Studies• Chamber data for Texanol®, butyl carbitol, and primarily alkane
petroleum distillates are consistent with SAPRC-99 predictions.
• Chamber data for Aromatics-100 consistent with SAPRC-99 forMIR conditions, but O3 inhibition at low NOx underpredicted.
• Reactivities of at least some synthetic hydrocarbon mixturesmay be underpredicted by up to a factor of 2.
• Glycol reactivities underpredicted by ~30% in some experiments,but unclear whether adjustments are appropriate.
• New mechanism developed for benzyl alcohol that simulateschamber data about as well as mechanisms for other aromatics
• Relative secondary PM impacts: benzyl alcohol >> butyl carbitol> petroleum distillates. No measurable PM impacts for others.
• No evidence that humidity and aerosol affects glycol availabilityat the relatively low aerosol loadings and humidities examined
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 56
Recommendations• Aromatics mechanisms need to be improved to further reduce
uncertainties in reactivity assessments (e.g., glycols)
• Extrapolation of current mechanisms to higher aromatics, suchas Aromatics 200, still highly uncertain
• Direct reactivity measurements needed to reduce uncertaintiesfor some VOCs, particularly mixtures of branched alkanes.
• A modified base case experiment that gives better correlationsbetween chamber and atmospheric reactivity would be useful
• No compelling need to change current bin assignments, exceptperhaps for those with light cycloalkanes and synthetic mixtures.But new procedure will be needed when reactivity scale updated
• Well-characterized environmental chamber data needed todevelop predictive secondary PM models. Work needed onbackground PM characterization in chambers
W. P. L. Carter 04/11/2005 Atmospheric Impacts of Coatings VOCs 57
Acknowledgements• Funded by CARB and SCAQMD
• CARB, SCAQMD staff and RRAC members: Helpful discussions
• Andrew Jaques of the ACC, Bob Hinrics of Citco, and ArleanMedeiros of ExxonMobil: Helpful discussions and providinghydrocarbon mixture compositional data
• David Morgott and Rodney Boatman of Eastman Kodak: Helpfuldiscussions regarding Texanol®.
• Albert Censullo of California Polytechnic State University:Provided compositional data prior to completion of his report
• Irina Malkina, Kurt Bumiller, Chen Song, Bethany Warren,Dennis Fitz, Charles Bufalino, and John Pisano of CE-CERT:Carried out or assisted with experiments
• David Cocker and Chen Song of CE-CERT: Assistance with PMdata and analysis