Subscriber access provided by UNIV OF PITTSBURGH
Article
Effects of cold temperature and ethanol content on VOC emissions from light-duty gasoline vehicles Ingrid George, Michael D. Hays, Jason Sandor Herrington, William Preston, Richard Snow, James Faircloth, Barbara Jane George, Thomas Long, and Richard Baldauf Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b04102 • Publication Date (Web): 07 Oct 2015 Downloaded from http://pubs.acs.org on October 13, 2015
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 27
Environmental Science & Technology
3
Effects of cold temperature and ethanol content on VOC emissions from light-duty gasoline vehicles
4 5
Ingrid J. George,1 Michael D. Hays,1,* Jason S. Herrington,1,# William Preston,2 Richard Snow,1 James Faircloth,1 Barbara Jane George,3 Thomas Long1 and Richard W. Baldauf1
1 2
6 7 8 9 10
1
Office of Research and Development, National Risk Management Research Laboratory, United States Environmental Protection Agency, Research Triangle Park, NC 27711. 2
ARCADIS U.S. Inc., Research Triangle Park, NC 27711.
3
Office of Research and Development, National Health and Environmental Effects Research Laboratory, United States Environmental Protection Agency, Research Triangle Park, NC 27711.
11
#
12 13
*Corresponding Author: 109 T.W. Alexander Drive, Durham, NC 27711. Phone: (919) 5413984. Fax: (919) 685-3346. E-mail:
[email protected].
14 15 16 17 18 19 20
Now at Restek Corp., Bellefonte PA 16823.
Keywords: dynamometer, gasoline exhaust, volatile organic compounds, carbonyls, mobile source air toxics, ethanol fuel
TOC/Abstract Art
Fraction of total ozone formation potential
1.0
0.8
0.6
0.4
Unsaturates
0.2
Oxygenates
Saturates BTEX 0.0
21
Aromatics SE0
SE10
SE85
WE0
WE10 WE85
22
1 ACS Paragon Plus Environment
Environmental Science & Technology
Page 2 of 27
23
Abstract
24
Emissions of speciated volatile organic compounds (VOCs), including mobile source air toxics
25
(MSATs), were measured in vehicle exhaust from three light-duty spark ignition vehicles
26
operating on summer and winter grade gasoline (E0) and ethanol blended (E10 and E85) fuels.
27
Vehicle testing was conducted using a three-phase LA92 driving cycle in a temperature-
28
controlled chassis dynamometer at two ambient temperatures (-7 °C and 24 °C). The cold start
29
driving phase and cold ambient temperature increased VOC and MSAT emissions up to several
30
orders of magnitude compared to emissions during other vehicle operation phases and warm
31
ambient temperature testing, respectively. As a result, calculated ozone formation potentials
32
(OFPs) were 7 to 21 times greater for the cold starts during cold temperature tests than
33
comparable warm temperature tests. The use of E85 fuel generally led to substantial reductions
34
in hydrocarbons and increases in oxygenates such as ethanol and acetaldehyde compared to E0
35
and E10 fuels. However, at the same ambient temperature the VOC emissions from the E0 and
36
E10 fuels and OFPs from all fuels were not significantly different. Cold temperature effects on
37
cold start MSAT emissions varied by individual MSAT compound, but were consistent over a
38
range of modern spark ignition vehicles.
39 40
2 ACS Paragon Plus Environment
Page 3 of 27
41 42
Environmental Science & Technology
INTRODUCTION Ethanol is the most widely used renewable transportation biofuel in the United States.
43
Consumption has increased from approximately 1% to approximately 10% of the U.S.
44
transportation fuel supply in the past two decades.1 Currently, over 97% of gasoline sold in the
45
U.S. contains ethanol. Most is blended at 10% by volume (also known as E10) and is used as an
46
oxygenated additive. A higher ethanol/gasoline blend containing between 51 to 83% ethanol
47
(commonly known as E85) by volume is available for use by flexible fuel vehicles (FFVs).
48
Ethanol consumption has been encouraged by the renewable fuel standards mandated by the
49
earlier U.S. Energy Policy Act (EPAct) of 2005 and more recently by the U.S. Energy
50
Independence and Security Act (EISA) of 2007. Both laws were passed to improve U.S. energy
51
security and independence and reduce greenhouse gas emissions from the transportation sector.2,3
52
The European Union has also enacted similar legislation to promote biofuel consumption in
53
Europe by setting a target of 10% energy used in the transportation sector from renewable fuels
54
by 2020.4
55
Vehicular emissions are a major source of atmospheric pollutants, including CO, CO2,
56
particulate matter (PM), NOx and volatile organic compounds (VOCs), including a number of
57
mobile source air toxics (MSATs) which are known or suspected to cause adverse health effects.
58
Mobile source emissions react in the atmosphere, leading to ground-level ozone and secondary
59
organic aerosol (SOA) formation, components of photochemical smog.5 The increased future
60
usage of ethanol as a transportation fuel as mandated by RFS2 has generated an urgent need to
61
accurately predict the ethanol fuel effects on mobile source emissions and to ensure that there is
62
no further degradation in air quality due to the transition. In response, numerous recent
63
dynamometer studies have characterized fuel effects on exhaust emissions of modern light-duty
3 ACS Paragon Plus Environment
Environmental Science & Technology
Page 4 of 27
64
(LD) spark ignition vehicles operating on various ethanol gasoline blends.6-19 Conflicting results
65
have been observed in the literature on fuel effects of ethanol addition to gasoline with respect to
66
non-methane hydrocarbons (NMHC), NOx, CO, and a few MSATs (benzene and 1,3-butadiene),
67
partly due to the confounding inter-study variability in fuel properties (e.g. aromatic content,
68
Reid Vapor Pressure (RVP)) with ethanol blending. Differences in driving cycles, fuel blend
69
preparation (i.e. splash vs. match blending) and vehicle characteristics such as the emission
70
control technologies, age, models, engine calibrations of the vehicles may also help explain
71
conflicting emissions data. However, there is widespread agreement that ethanol blends increase
72
ethanol and acetaldehyde in exhaust emissions, with ethanol photochemically transforming to
73
acetaldehyde, an important MSAT in the atmosphere. Several modeling studies have examined the health and air quality impacts of increased
74 75
future usage of E85 blends in the U.S. and northern Europe. These studies generally suggest that
76
moving to E85 may lead to small changes in ozone concentrations within a few ppb in the U.S.20-
77
24
78
conditions (up to ~39 ppbv increase) compared to summertime (~7 ppbv increase) in a high NOx
79
area. For wintertime driving conditions, these modeling studies either relied on extremely sparse
80
emissions data of speciated VOCs and MSATs in LD vehicle exhaust (using ethanol blends or
81
gasoline) at cold temperature or assumed no change in emission rates at cold temperatures.
82
Notably, Ginnebaugh et al.22 predicted greater increases in ozone levels under wintertime
The objective of this study was to investigate the impact of ethanol blends (E0, E10,
83
E85), ambient cold temperature (-7 vs. 24 °C) and driving cycle on speciated VOC (and MSAT)
84
emissions from three LD spark ignition vehicles meeting Tier 2 emission standards. The vehicles
85
were tested in a temperature controlled chamber housing a chassis dynamometer using the LA92
86
driving cycle simulation. This work presents emission rates of the major ozone precursor VOCs, 4 ACS Paragon Plus Environment
Page 5 of 27
Environmental Science & Technology
87
including ethanol, and MSATs in the vehicle exhaust for the three ethanol blend fuels. VOC
88
MSATs investigated in this work include carbonyls (i.e., formaldehyde, acetaldehyde, acrolein),
89
aromatics (i.e., benzene, toluene, ethylbenzene, xylenes (BTEX), naphthalene, styrene) and other
90
toxic hydrocarbons (n-hexane, 1,3-butadiene). This research considerably improves our
91
understanding of mobile source emissions of speciated VOCs, and particularly ozone precursor
92
and MSAT VOCs, at cold temperatures. The improved emissions information will be used in air
93
quality models to more accurately predict the health and environmental impacts of the growing
94
biofuels consumption under wintertime conditions.
95
EXPERIMENTAL METHODS
96
Dynamometer testing
97
The dynamometer testing conducted in this study has previously been described in
98
detail25 and will only be covered briefly here. Dynamometer testing for this study was conducted
99
at the U.S. EPA dynamometer facilities located in Research Triangle Park, NC. Three 2008
100
model year light-duty Tier 2 bin 5 compliant vehicles were tested during the study: V1) Honda
101
Civic LX, V2) Chevrolet Impala LS, and V3) Chrysler Town & Country. Odometer readings
102
were 26,459 km for V1, 23,785 km for V2 and 78,283 for V3. Both V1 and V2 were equipped
103
with multiport fuel injection systems, and V3 had a sequential fuel injection system. V2 and V3
104
were flexible fuel vehicles (FFVs). Six fuels were obtained from Gage Products Co. (Ferndale,
105
MI, U.S.), including summer-grade (S) and winter-grade (W) ethanol blends of E0, E10 and E85
106
(henceforth designated as SE0, SE10 and SE85 for summer fuels and WE0, WE10 and WE85 for
107
winter-grade fuels). The E85 blends were not tested on V1 as it was not an FFV. Fuels were
108
blended to achieve seasonally-appropriate Reid vapor pressures that were consistent within
109
summer and winter fuel sets. Benzene content in all fuels was kept constant at ~1% v/v. The 5 ACS Paragon Plus Environment
Environmental Science & Technology
Page 6 of 27
110
ethanol volume fractions in the ethanol blended fuels are as follows: SE10 = 8.6%, WE10 =
111
8.7%, SE85 = 84.0%, and WE85 = 77.7%. Fuel properties for all test fuels are listed in Table S1
112
in the Supporting Information.
113
Vehicle testing was conducted on a 48 in. roll electric Burke E. Porter model 4100
114
chassis dynamometer (Burke E. Porter Machinery Co., Grand Rapids, MI, U.S.). The
115
dynamometer was housed inside a temperature controlled test chamber, which was held at -7 or
116
24 °C during vehicle preconditioning, soaking and testing with winter-grade and summer-grade
117
test fuels, respectively. Vehicle testing procedures followed the Code of Federal Regulations
118
(CFR), Title 40, Part 86. Emissions testing was conducted over a three-phase LA92 cycle, with
119
the first 300 s representing cold start Driving Phase 1, followed by 1135 s hot stabilized Driving
120
Phase 2. After a 600 s engine off segment, the entire 1435 s LA92 cycle was repeated as Driving
121
Phase 3. Each set of test conditions in the study matrix (see Supporting Information Table S2)
122
was tested in triplicate (or greater) on sequential test days. The vehicle preconditioning
123
procedure included driving two LA92 cycles under the new test conditions and holding the
124
vehicle in the climate chamber at the set ambient temperature overnight before testing. Further
125
details on the fuel change procedures are found in Hays et al.25 and the Supporting Information.
126
Emissions sampling and analysis
127
Diluted exhaust emissions were sampled from a constant volume sampling dilution tunnel as
128
described in Hays et al.25 Dilution air was preconditioned by passing through a filter and
129
charcoal trap and then mixed with exhaust in the dilution tunnel at 21 °C and with a dilution
130
factor between 15 to 40. A turbine (Spencer Turbine Company, Model 2025-H-SPEC, Windsor,
131
CT, U.S.) was used to pull the dilution flow through the dilution tunnel, and a critical flow
132
venturi (Horiba Instruments, Inc., CVS-48M, Ann Arbor, MI, U.S.) controlled the total flow rate 6 ACS Paragon Plus Environment
Page 7 of 27
Environmental Science & Technology
133
of ~21 m3 min-1. Along with real-time sampling of regulated emissions with continuous
134
emissions monitors, PM and VOC emissions were also sampled in a time-integrated manner
135
from the dilution tunnel. While the speciated PM emissions were discussed elsewhere25, VOC
136
emissions were the focus of this work. VOC measurements consisted of sampling through a 113
137
°C transfer line to a pre-cleaned 6 L stainless steel canister for VOCs, a 2,4-
138
Dinitrophenylhydrazine (DNPH)-coated cartridge (Sigma-Aldrich Corp., St. Louis, MO, U.S.,
139
LpDNPH H30) for carbonyls, and two chilled water impingers in series for ethanol. These
140
samples were integrated separately over each driving phase (Driving Phase 1, 2 and 3). A
141
background sample of the dilution air was taken each test day, and one blank sample was taken
142
for each set of test conditions. All sampling flow rates were controlled by calibrated mass flow
143
controllers. Ethanol samples and DNPH cartridges were refrigerated until analysis. VOCs and
144
ethanol samples were analyzed by gas chromatography (GC) with mass spectrometry (MS) and
145
flame ionization detector (FID). DNPH cartridges were extracted with acetonitrile, and extracts
146
were analyzed by high performance liquid chromatography with ultraviolet detection. Further
147
details regarding analytical methods are given in the supporting information.
148
Statistical analysis
149
SAS 9.3 (SAS Institute Inc. 2011. SAS/ STAT 9.3 User’s Guide. Cary, NC) was used for the
150
statistical analysis. Linear mixed models of emission rate concentrations were fit separately for
151
each compound to assess relationships with the covariates of driving phase, temperature, and fuel
152
nested within temperature (since fuels were season-specific). These models of emission rate
153
concentrations used log-transformed measurements, favored by tests for normality, restricted to
154
concentration measurements above the detection limit. The models accounted for the repeated
155
measure structure of testing each vehicle in all three driving phases and for the fixed effect of the 7 ACS Paragon Plus Environment
Environmental Science & Technology
156
vehicle. The p-values for the tests of the differences of least square means are given in Table S7
157
with the significance level set to p < 0.05.
158
RESULTS
159
VOC emissions by driving phase
160
Page 8 of 27
VOC speciation analysis of the vehicle exhaust included a total of 141 target compounds,
161
of which 120 were detected in at least one diluted exhaust sample during the study. Emission
162
rates for all test conditions for all individual VOCs can be found in Table S6 in the Supporting
163
Information along with results from statistical testing (Table S7). Figure 1 displays average
164
emissions profiles for SE0 (24 °C) tests separated by driving phase for the most highly emitted
165
VOCs, which included major components of unburned gasoline fuel (e.g., 2,2,4-
166
trimethylpentane, isopentane) and products of partial combustion (e.g., propylene, aromatics).
167
The most abundant mobile source air toxics (MSATs), denoted with an asterisk, were benzene,
168
toluene, xylenes, acetaldehyde, formaldehyde, and 1,3-butadiene. Speciated VOC emission rates
169
in Driving Phase 1 cold start were at least an order of magnitude higher compared to Driving
170
Phases 2 and 3 for SE0 tests, where the products of partial combustion appeared to make a
171
greater contribution to total emissions in Driving Phase 1 than the unburned gasoline
172
components.
173
The sum of speciated VOC (ΣVOCs) emission rates (ERs) is shown by test condition and
174
driving phase in Figure 2 (top panel), along with percent contributions of ethanol (middle panel)
175
and sum of MSAT compounds (bottom panel) to ΣVOCs. Figure 2 clearly indicates that driving
176
phase and temperature were the most important factors studied in this work impacting VOC
177
exhaust emissions for all three LD vehicles studied. As noted earlier, the total VOC emissions 8 ACS Paragon Plus Environment
Page 9 of 27
Environmental Science & Technology
178
were substantially higher in Driving Phase 1 cold start than Driving Phases 2 and 3 by one to
179
three orders of magnitude. These differences in ΣVOCs with driving phase were statistically
180
significant (p < 0.05), as they were for nearly all VOCs detected (see Table S7). ΣVOCs
181
emission rates measured during the Driving Phase 1 tests at -7 °C were approximately an order
182
of magnitude higher than for cold starts at 24 °C. The effect of ethanol fuel blends on ΣVOCs
183
was less clear. Ethanol was the highest emitted VOC measured overall during the study.
184
However, the majority of ethanol emissions (> 93%) were from Driving Phase 1 of WE85 fuel
185
tests for V2 and V3, during which ethanol dominated VOC emissions (Figure 2, middle panel).
186
Ethanol also made up the majority of Driving Phase 2 emissions during these two WE85 tests,
187
although the emission rates were much lower than in Driving Phase 1. MSATs represented 16%
188
of total VOC emissions on average as shown in Figure 2 (bottom panel). Speciated MSATs are measured less often during emissions testing than non-methane
189 190
hydrocarbons (NMHC, as determined from total hydrocarbon and methane measurements), so it
191
is generally assumed that MSAT emissions are correlated with NMHC. In this study, R2 values
192
for speciated MSATs and NMHC emission rates are summarized in Table S5 in the Supporting
193
Information. Although the ΣMSATs and some hydrocarbon MSATs (e.g., benzene, 1,3-
194
butadiene) were strongly correlated with NMHC (R2 = 0.59 - 0.97), a weaker correlation was
195
found for the carbonyl MSATs with NMHC measurements (R2 = 0.23 - 0.32). The fuel and
196
temperature effects on individual VOC and MSATs emissions are discussed in more detail
197
below.
198
199
9 ACS Paragon Plus Environment
Environmental Science & Technology
200
201
Page 10 of 27
Driving Phase 1 temperature effects VOC emissions increased dramatically during Driving Phase 1 from 24 °C to -7 °C tests,
202
which was consistently observed for all three vehicles and for all fuels (Figure 2). Statistically
203
significant increases of at least a factor of two for the majority of the individual VOCs,
204
ΣMSATs, ΣVOCs and NMHC with the decrease in test cell temperature were observed (see
205
Table S7). However, there were a few notable exceptions, where changes of less than a factor of
206
two were found for acetone, acrolein, n-butane, and formaldehyde for E0 and E10 tests.
207
Formaldehyde emissions decreased with the cold temperature for both E0 and E10, consistent
208
with the more fuel rich combustion conditions needed during colder temperature/cold start
209
operation. Comparisons of VOC emission profiles for 24 °C and -7 °C tests for a given vehicle
210
and ethanol blend indicated that the VOC emission profiles changed to some extent (R2 ~ 0.63 –
211
0.92) with test cell temperature.
212
This change can be seen more clearly in Figure 3, displaying the relative changes in VOC
213
emissions for each individual VOC grouped by compound class and averaged for each ethanol
214
blend. As a reference, NMHC emissions increased in Driving Phase 1 on average by 9 - 10 times
215
for all fuels, while ΣVOCs increased by ~7, 9 and 23 times for E0, E10 and E85 fuels,
216
respectively. Aromatics (ARs) aside from BTEX (i.e. benzene, toluene, ethylbenzene, xylenes)
217
and unsaturated (UNSAT) compounds had the greatest range of relative changes, with the
218
highest relative increases up to a few hundred times, particularly for E85 tests. Relative changes
219
with temperature for BTEX, saturated compounds (SATs) and oxygenates were up to ~40 times
220
for Driving Phase 1. The smallest changes in emissions were observed for oxygenates, but the
221
temperature effect was more pronounced during E85 tests for these compounds.
10 ACS Paragon Plus Environment
Page 11 of 27
222
Environmental Science & Technology
These results are consistent with emissions of both unburned fuel components and
223
products of partial combustion for both gasoline and ethanol rising due to the fuel rich
224
combustion conditions needed during cold temperature cold start operations, but the degree of
225
the effect may vary by individual VOC and the fuel used. It is probable that the combustion
226
conditions may need to be even more fuel rich with E85 use during cold temperature cold starts
227
than with the lower ethanol blends due to the higher heat of vaporization of E85. This could
228
explain the intensified temperature effects on the emissions with increasing ethanol content in
229
the fuels as seen in Figure 3. Engine calibration, an important factor that influences exhaust
230
emissions at cold start, was not studied in depth in this work.13
231
Driving Phase 1 fuel effects
232
The impact of E10 and E85 compared to E0 fuel blends are shown as average relative changes in
233
individual VOC emissions by compound class and temperature in Figure 4. Whereas cold
234
temperature effects generally led to significant increases in most VOCs (Figure 3), the
235
directionality, magnitude, and degree of statistical significance of the changes in VOC emissions
236
due to use of ethanol-blended fuels was highly dependent on whether or not the VOCs originated
237
from the fuel components. Most compounds did not show significant changes in emissions from
238
E0 to E10 for either summer or winter blends (Table S7). However, emissions of many
239
hydrocarbons were significantly impacted when E0 and E10 blends were compared to E85 under
240
the same temperature. The emissions of compounds that were most strongly linearly related to
241
ethanol emissions generally increased with use of ethanol blend fuels, particularly for E85
242
blends, including acetaldehyde, acetylene and other carbonyls (Oxygenates) and C2
243
hydrocarbons. In addition, the emissions of the majority of the VOCs associated with the
11 ACS Paragon Plus Environment
Environmental Science & Technology
244
petroleum fuel component and its combustion byproducts decreased consistently as the
245
petroleum fuel fraction decreased.
246
Page 12 of 27
Because the change in the ambient temperature from 24 to -7 °C led to an increase in
247
VOC emissions generally, the reductions in gasoline related VOC emissions due to the ethanol
248
blend fuel effect were partially counteracted at cold temperatures. For example, Driving Phase 1
249
benzene emissions changed from an ~ 30% reduction for SE85 to an ~ 40% increase for WE85
250
compared to respective E0 blends. Note that the changes in benzene emissions corresponded to
251
much higher p-values (p = 0.01 - 0.44) as seen in Table S7 for the E85 blend fuel comparisons
252
compared to the other monoaromatic compounds (p < 0.0001), likely due to the fixed benzene
253
content (~ 1%) in all fuels. Therefore, the benzene emissions were less reflective of changes in
254
BTEX in this study, but may indirectly indicate the increase in fuel consumption and/or fuel
255
enrichment during combustion. While non-methane hydrocarbon emissions generally decreased
256
from E0 to other ethanol blends (statistically significantly for fuel comparisons with E85 blends),
257
ΣMSATs and ΣVOCs increased for some tests using E85 at 24 °C and both E10 and E85 fuels at
258
-7 °C, mainly due to dramatic increases in acetaldehyde and ethanol.
259
Ozone formation potentials
260
The VOC emission rates determined from this study were weighted by their relative
261
ozone reactivities and summed for each test condition to determine total ozone formation
262
potentials (ΣOFPs) in units of g O3/km based on the Maximum Incremental Reactivity (MIR)
263
scale as outlined by Carter.26 The MIR scale is used to determine the maximum potential impact
264
of an incremental increase in VOC mass emissions on ground level ozone formation, typically
265
assuming relatively high NOx conditions (i.e., VOC-limited conditions for ozone formation). It
12 ACS Paragon Plus Environment
Page 13 of 27
Environmental Science & Technology
266
should be noted this paper does not address NOx impacts and that real-world air quality impacts
267
will depend on ratios of hydrocarbons to NOx. Best-estimate MIR values (in g O3/g VOC) used
268
convert emission rates to OFPs for all tests are listed in Table S6 for all individual VOCs. The
269
relative contribution of each VOC compound class to the ΣOFPs during Driving Phase 1 for each
270
fuel averaged over the test vehicles is summarized for each test fuel in Figure 5 along with ΣOFP
271
values. OFP values are given for each test condition and statistical results are in the Supporting
272
Information (Table S6 and S7, respectively). Driving Phase 2 and 3 ΣOFPs were only a few
273
percent of the Driving Phase 1 values, so the focus of this discussion is on Driving Phase 1 only.
274
The general trends in ΣOFPs for each temperature were E10 < E0 < E85. However, the
275
fuel effects on ΣOFPs were not statistically significant (see Table S7). Unsaturated compounds
276
(particularly ethylene, propylene and isobutene/1-butene) were responsible for the majority of
277
the ozone reactivity for E0 and E10 blends, whereas oxygenates (mostly ethanol and
278
acetaldehyde) contributed the most to the OFPs for E85 blends. Cold ambient temperature led to
279
statistically significant increases in OFPs between 7 to 21 times for the three ethanol blends. The
280
highest ΣOFP value was calculated for WE85 emissions at 18.1 g O3/km. The influence of
281
ethanol fuel and cold temperature on relative contributions of compound classes and changes in
282
OFP values in this study are in agreement with previous dynamometer studies.7, 11, 27 However,
283
the OFP values in this work are substantially higher for both ambient temperatures when
284
compared with Driving Phase 1 values from other studies despite excluding CO in our
285
calculation, possibly due to the more extensive VOC speciation of both hydrocarbons and
286
oxygenates included in the calculations here, or the difference in absolute emissions for different
287
driving cycles and vehicles.
13 ACS Paragon Plus Environment
Environmental Science & Technology
Page 14 of 27
288
DISCUSSION
289
Comparison with literature
290
From this work, it is clear that cold ambient temperature has considerable impact on VOC
291
emissions during cold starts for light-duty vehicles operating on gasoline/ethanol blended fuels.
292
This research suggests that the magnitude of the cold temperature effects on MSATs is highly
293
dependent on the fuel composition (see Figure 3). To determine whether this trend is generally
294
robust and to gain a better understanding of the current state of the research, cold temperature-
295
induced changes in MSAT emissions at cold starts from the literature were compared with this
296
work as summarized in Figure 6. Seven studies were included that measured and reported MSAT
297
emission changes from “warm” (20 to 24 °C) to “cold” (-7 to -10 °C) temperatures for light-duty
298
vehicles of model year ≥ 1998 with three-way catalysts (mostly Tier 2, Euro4 and 5a emission
299
standards) operating on a range of ethanol blends from E0 to E85.7, 9, 11, 27-30 This study is the first
300
of its kind to our knowledge to report cold temperature cold start changes for all twelve MSATs
301
listed in Figure 6 for Tier 2 or similar vehicles, where half the MSATs have no previously
302
reported cold temperature ERs. As shown in Figure 6, the pattern of cold temperature increases
303
by ethanol blend for individual MSATs in the literature are remarkably consistent with this work
304
apart from the large variations in the literature values for benzene in the range of -0.4 to +174
305
times (full scale version of Figure 6 shown in the Supporting Information). In fact, average
306
values by fuel from this study and the literature were linearly related (slope = 1.1, R2 = 0.88)
307
when compared pairwise if benzene values were excluded, suggesting that the relative trends in
308
cold temperature effects on MSATs (apart from benzene) by ethanol-blended fuel from this
309
study may be relevant to other spark ignition vehicles of similar model years.
14 ACS Paragon Plus Environment
Page 15 of 27
Environmental Science & Technology
310
SPECIATE is a U.S. EPA speciated VOC and PM emission profile database for
311
numerous emission sources used to create emission inventories and used in source receptor and
312
air quality models. Emission profiles from this work have been added to SPECIATE Version
313
4.431 (profiles #8904 - 8927). VOC emission profiles from this work brought a substantial
314
increase in the number of E85 profiles in the SPECIATE database including the first cold
315
temperature E85 profile. The emission profiles from this work were compared to 250
316
gasoline/ethanol blend exhaust speciated VOC profiles. The relative contribution of MSATs to
317
the emission profiles from other SPECIATE profiles are shown with those from this work in the
318
Supporting Information (Figure S2). Although generally comparable trends in MSAT %
319
contributions are observed between this work and previous emission profiles, the values from
320
previous studies span several orders of magnitude for each MSAT due to a number of inter-study
321
differences that may impact VOC emissions.
322
Air quality implications of cold temperature effects
323
In this study, driving tests for LD vehicles using E0, E10 and E85 simulating wintertime
324
ambient conditions (-7 °C) led to dramatic increases in cold start MSAT emissions (Figure 6)
325
and ozone formation potentials (Figure 5) by nearly an order of magnitude greater than emissions
326
during standard testing conditions (24 °C). As noted earlier, air quality modeling studies have
327
suggested that the overall impact of increased E85 use on ground-level ozone formation is on a
328
much smaller scale, which appears to be highly dependent on NOx emission changes and less
329
sensitive to VOC emission changes than assumed by the OFP calculation.20-24 Another reason for
330
the minimal predicted effects of E85 use is that the increase in oxygenated ozone precursor
331
emissions with E85 use are counterbalanced by a reduction in hydrocarbon ozone precursors.
15 ACS Paragon Plus Environment
Environmental Science & Technology
332
Nevertheless, this study suggests that the ozone formation potential of E85 emissions should
333
increase at colder temperatures, which these air quality studies have not yet taken into account.
334
Another factor that moderates the environmental impact of E85 in air quality models is that
Page 16 of 27
335
direct acetaldehyde emissions from mobile sources contribute a small fraction to overall ambient
336
acetaldehyde levels, especially in summer due to relatively greater photochemical acetaldehyde
337
production.32 In winter, direct acetaldehyde emissions become more important in urban areas
338
compared to summer, but ozone production is less important in wintertime generally due to
339
reduced photochemical activity. Thus, on a regional scale, the fuel effect of E85 on ground-level
340
ozone production may lead to negligible modifications to ozone levels. Near road emissions of
341
MSATs such as acetaldehyde may be important, particularly at colder temperatures where
342
increased MSAT emissions and less photochemically produced acetaldehyde are expected.
343
An ongoing source of uncertainty in modeling the air quality and health impacts of mobile
344
sources is the temperature dependence of mobile source emissions and the lack of up to date
345
speciated emissions data for modern vehicles at cold temperatures.33 This study and our previous
346
work on emissions of heavy-duty trucks using biodiesel blends34 have attempted to fill some
347
gaps to better understand cold temperature effects of exhaust emissions, particularly MSATs and
348
VOCs, from modern vehicles operating on biofuels. However, newer Tier 2 vehicles that are
349
required to be compliant with the 2007 MSAT Rule35 are expected to have lower MSAT
350
emissions at cold temperatures than the MY 2008 test vehicles studied in this work. We have
351
only performed testing at two temperatures, so the mathematical relationship between emissions
352
and temperature cannot be ascertained. A few studies have measured emissions at multiple
353
temperatures,28, 36, 37 and have found that VOC/MSAT emissions increase nonlinearly (in some
354
cases exponentially) with decreasing temperature during cold starts. Clearly, more studies are 16 ACS Paragon Plus Environment
Page 17 of 27
Environmental Science & Technology
355
needed to understand how emissions are influenced by ambient conditions, which is essential to
356
accurate assessment of the environmental impacts of the transportation sector.
357
ACKNOWLEDGMENTS Authors would like to thank Eastern Research Group, Inc. for analytical support, and
358 359
Joseph McDonald, Rich Cook and Deborah Luecken from U.S. EPA for helpful discussions. The
360
views expressed in this article are those of the authors and do not necessarily represent the views
361
or policies of the U.S. Environmental Protection Agency.
362
Supporting Information
363
A description of VOC analysis details is included in the supporting information along
364
with tables of fuel properties, fuel change procedures, test matrix, linear comparisons of VOCs,
365
VOC emission rates, and table of statistical results. Figures include a full scale version of Figure
366
6 and a comparison of data with SPECIATE and literature values. Note that Tables S6 and S7 are
367
presented in a separate Microsoft Excel file as tabbed sheets. This information is available free of
368
charge at http//pubs.acs.org/.
369
References
370
(1) Renewable Fuels Association Website, www.ethanolrfa.org, (accessed May 2015).
371
(2) Energy policy act of 2005. Public Law 109-58, 2005;
372
http://www.gpo.gov/fdsys/pkg/BILLS-109hr6enr/pdf/BILLS-109hr6enr.pdf (accessed May
373
2015).
374
(3) Energy independence and security act of 2007. Public Law 110-140, 2007;
375
http://www.gpo.gov/fdsys/pkg/BILLS-110hr6enr/pdf/BILLS-110hr6enr.pdf (accessed May
376
2015).
17 ACS Paragon Plus Environment
Environmental Science & Technology
377
(4) Directive 2009/28/EC of the European Parliament and of the Council of 23 April 2009 on
378
the promotion of the use of energy from renewable sources and amending and subsequently
379
repealing Directives 2001/77/EC and 2003/30/EC. Directive 2009/28/EC, 2009.
380
(5) U.S. Clean Air Act. Public Law 88-260 (and subsequent amendments),
381 382
Page 18 of 27
http://www.epa.gov/air/caa/ (accessed May 2015). (6) Aakko-Saksa, P. T.; Rantanen-Kolehmainen, L.; Skyttä, E. Ethanol, isobutanol, and
383
biohydrocarbons as gasoline components in relation to gaseous emissions and particulate matter.
384
Environ. Sci. Technol. 2014, 48 (17), 10489-10496.
385
(7) Clairotte, M.; Adam, T. W.; Zardini, A. A.; Manfredi, U.; Martini, G.; Krasenbrink, A.;
386
Vicet, A.; Tournié, E.; Astorga, C. Effects of low temperature on the cold start gaseous
387
emissions from light duty vehicles fuelled by ethanol-blended gasoline. Applied Energy 2013,
388
102, 44-54.
389
(8) Durbin, T. D.; Miller, J.W.; Younglove, T.; Huai, T.; Cocker, K. Effects of fuel ethanol
390
content and volatility on regulated and unregulated exhaust emissions for the latest technology
391
gasoline vehicles. Environ. Sci. Technol. 2007, 41 (11), 4059-4064.
392
(9) Flex fuel vehicles (FFVs) VOC/PM cold temperature characterization when operating on
393
ethanol (E10, E70, E85); EPA-HQ-OAR-2005-0036-1164; Southwest Research Institute: San
394
Antonio, TX, 2007.
395
(10) EPAct/V2/E-89 final report assessing the effect of five gasoline properties on exhaust
396
emissions from light-duty vehicles certified to Tier 2 standards: Analysis of data from EPAct
397
Phase 3; EPA-420-R-13-002; U.S. Environmental Protection Agency: Ann Arbor, MI 2011;
398
http://www.epa.gov/otaq/models/moves/epact.htm.
399
(11) Graham, L. A.; Belisle, S. L.; Baas, C. Emissions from light duty gasoline vehicles
400
operating on low blend ethanol gasoline and E85. Atmos. Environ. 2008, 42 (19), 4498-4516.
401
(12) Hochhauser, A. M.; Schleyer, C. H.; Summary of research on the use of intermediate
402 403
ethanol blends in on-road vehicles. Energy & Fuels 2014, 28 (5), 3236-3247. (13) Hubbard, C. P; Anderson, J. E.; Wallington, T. J. Ethanol and air quality: Influence of
404
fuel ethanol content on emissions and fuel economy of flexible fuel vehicles. Environ. Sci.
405
Technol. 2014, 48 (1), 861-867.
18 ACS Paragon Plus Environment
Page 19 of 27
Environmental Science & Technology
406
(14) Karavalakis, G.; Durbin, T. D.; Shrivastava, M.; Zheng, Z.; Villela, M.; Jung, H. Impacts
407
of ethanol fuel level on emissions of regulated and unregulated pollutants from a fleet of gasoline
408
light-duty vehicles. Fuel 2012, 93, 549-558.
409
(15) Karavalakis, G.; Short, D.; Russell, R. L.; Jung, H.; Johnson, K. C.; Asa-Awuku, A.;
410
Durbin, T. D. Assessing the impacts of ethanol and isobutanol on gaseous and particulate
411
emissions from flexible fuel vehicles. Environ. Sci. Technol. 2014, 48 (23), 14016-14024.
412
(16) Montero, L.; Duane, M.; Manfredi, U.; Astorga, C.; Martini, G.; Carriero, M.;
413
Krasenbrink, A.; Larsen, B. R. Hydrocarbon emission fingerprints from contemporary
414
vehicle/engine technologies with conventional and new fuels. Atmos. Environ. 2010, 44 (18),
415
2167-2175.
416 417 418 419 420
(17) Yao, Y.; Tsai, J.; Wang, I. Emissions of gaseous pollutant from motorcycle powered by ethanol–gasoline blend. Applied Energy 2013, 102, 93-100. (18) Yanowitz, J.; McCormick, R. L. Effect of E85 on tailpipe emissions from light-duty vehicles. J. & Air Waste Manage. Assoc. 2009, 59 (2), 172-182. (19) Effects of ethanol and volatility parameters on exhaust emissions; CRC Report E-67;
421
Coordinating Research Council: Alpharetta, GA, 2006;
422
http://www.crcao.org/publications/emissions/index.html, (accessed June 2015).
423
(20) Cook, R.; Phillips, S.; Houyoux, M.; Dolwick, P.; Mason, R.; Yanca, C.; Zawacki, M.;
424
Davidson, K.; Michaels, H.; Harvey, C.; Somers, J.; Luecken, D. Air quality impacts of
425
increased use of ethanol under the United States’ Energy Independence and Security Act. Atmos.
426
Environ. 2011, 45 (40), 7714-7724.
427
(21) Fridell, E.; Haeger-Eugensson, M.; Moldanova, J.; Forsberg, B.; Sjöberg, K. A
428
modelling study of the impact on air quality and health due to the emissions from E85 and petrol
429
fuelled cars in Sweden. Atmos. Environ. 2014, 82, 1-8.
430
(22) Ginnebaugh, D. L.; Liang, J.; Jacobson, M. Z. Examining the temperature dependence
431
of ethanol (E85) versus gasoline emissions on air pollution with a largely-explicit chemical
432
mechanism Atmos. Environ. 2010, 44 (9), 1192-1199.
433 434
(23) Jacobson, M. Z. Effects of ethanol (E85) versus gasoline vehicles on cancer and mortality in the United States. Environ. Sci. Technol. 2007, 41 (11), 4150-4157.
19 ACS Paragon Plus Environment
Environmental Science & Technology
435
Page 20 of 27
(24) Nopmongcol, U.; Griffin, W. M.; Yarwood, G.; Dunker, A. M.; MacLean, H. L.;
436
Mansell, G.; Grant, J. Impact of dedicated E85 vehicle use on ozone and particulate matter in the
437
US. Atmos. Environ. 2011, 45 (39), 7330-7340.
438
(25) Hays, M. D.; Preston, W.; George, B. J.; Schmid, J.; Baldauf, R.; Snow, R.; Robinson, J.
439
R.; Long, T.; Faircloth, J. Carbonaceous aerosols emitted from light-duty vehicles operating on
440
gasoline and ethanol fuel blends. Environ. Sci. Technol. 2013, 47 (24), 14502-14509.
441
(26) Carter, W. P. L. Estimation of ozone reactivities for volatile organic compound
442
speciation profiles in the SPECIATE 4.2 database. , 2013,
443
http://www.cert.ucr.edu/~carter/emitdb/Speciate-Reactivity.pdf/ (accessed June 2015).
444
(27) Suarez-Bertoa, R.; Zardini, A. A.; Keuken, H.; Astorga, C. Impact of ethanol containing
445
gasoline blends on emissions from a flex-fuel vehicle tested over the Worldwide Harmonized
446
Light duty Test Cycle (WLTC). Fuel 2015, 143, 173-182.
447
(28) VOC/PM cold temperature characterization and interior climate control emissions/fuel
448
economy impact; Draft Final Report Vol. I; Southwest Research Institute: San Antonio, TX,
449
2005.
450 451 452
(29) Karlsson, H.; Gåsste, J.; Åsman, P. Regulated and non-regulated emissions from Euro 4 alternative fuel vehicles. SAE paper 2008-01-1770, 2008, DOI 10.4271/2008-01-1770. (30) Westerholm, R.; Ahlvik, P.; Karlsson, H. L. An exhaust characterisation study based on
453
regulated and unregulated tailpipe and evaporative emissions from bi-fuel and flex-fuel light-
454
duty passenger cars fuelled by petrol (E5), bioethanol (E70, E85) and biogas tested at ambient
455
temperatures of +22 °C and -7 °C; Final Report; Swedish Road Administration: Stockholm,
456
Sweden, 2008.
457 458 459
(31) SPECIATE Website; http://www.epa.gov/ttn/chief/software/speciate/index.html (accessed April 2015). (32) Luecken, D. J.; Hutzell, W. T.; Strum, M. L.; Pouliot, G. A. Regional sources of
460
atmospheric formaldehyde and acetaldehyde, and implications for atmospheric modeling. Atmos.
461
Environ. 2012, 47, 477-490.
462
(33) Cook, R.; Touma, J. S.; Fernandez, A.; Brzezinski, D.; Bailey, C.; Scarbro, C.;
463
Thurman, J.; Strum, M.; Ensley, D., Baldauf, R. Impact of underestimating the effects of cold
464
temperature on motor vehicle start emissions of air toxics in the United States. J. Air & Waste
465
Manage. Assoc. 2007, 57 (12), 1469–1479. 20 ACS Paragon Plus Environment
Page 21 of 27
466
Environmental Science & Technology
(34) George, I. J.; Hays, M. D.; Snow, R.; Faircloth, J.; George, B. J.; Long, T.; Baldauf, R.
467
W. Cold temperature and biodiesel fuel effects on speciated emissions of volatile organic
468
compounds from diesel trucks. Environ. Sci. Technol. 2014, 48 (24), 14782−14789.
469 470
(35) Control of hazardous air pollutants from mobile sources. EPA–HQ–OAR–2005–0036, 2007; http://www.gpo.gov/fdsys/pkg/FR-2007-02-26/pdf/E7-2667.pdf.
471
(36) Ludykar, D.; Westerholm, R.; Almén, J. Cold start emissions at +22, -7 and -20°C
472
ambient temperatures from a three-way catalyst (TWC) car: regulated and unregulated exhaust
473
components. Sci. Total Environ. 1999, 235 (1-3), 65-69.
474
(37) Knapp, K. T.; Stump, F. D.; Tejada, S. B. The effect of ethanol fuel on the emissions of
475
vehicles over a wide range of temperatures. J. Air & Waste Manage. Assoc. 1998, 48 (7), 646-
476
653.
477 478 479 480 481 482
21 ACS Paragon Plus Environment
Environmental Science & Technology
Page 22 of 27
483 484 485
Figure 1. Average emission profiles by driving phase for all vehicles for SE0. Error bars
486
represent 1 standard deviation of mean values. Mobile source air toxics are labeled with an
487
asterisk.
488
22 ACS Paragon Plus Environment
Page 23 of 27
Environmental Science & Technology
V1
% Ethanol
ΣVOCs ER (g/km)
Temp.= 10
24
V2 -7
24
V3 -7
24
-7
1
0.1
0.01
100 80 60 40 20
% MSATs
0 100 80 60
Driving Phase 1 Driving Phase 2 Driving Phase 3
40
489
V1 V1 -SE -S 0 V1 E1 0 V 1 -W E -W 0 E V2 10 S V2 E -S 0 V2 E1 -S 0 V2 E8 5 V 2 -W E -W 0 V2 E -W 1 0 E V3 85 V3 -SE -S 0 V3 E1 -S 0 V3 E8 5 V 3 -W E -W 0 V3 E -W 1 0 E8 5
20
490 491
Figure 2. Average sum of VOCs (ΣVOCs, top panel) emission rates (ER), ethanol (middle
492
panel) and mobile source air toxics (MSATs, bottom panel) fraction (in %) of ΣVOCs for all test
493
conditions by driving phase. Error bars represent 1 standard deviation of mean values.
494
23 ACS Paragon Plus Environment
Environmental Science & Technology
Page 24 of 27
(ER-7C - ER24C) / ER24C
100 E0 E10 E85
80
300
200
100
60 0 AR
BTEX
UNSAT
SAT Oxygenates
40
20
0
495
AR
BTEX
UNSAT
SAT
Oxygenates
496
Figure 3. Relative changes in individual VOC emissions in Driving Phase 1 from 24 to -7 °C
497
averaged by fuel and grouped by compound class: aromatics except BTEX (AR), BTEX,
498
unsaturated (UNSAT) and saturated (SAT) hydrocarbons, and oxygenates (Oxygenates). Inset
499
shows full scale.
500
501
502
503
24 ACS Paragon Plus Environment
Page 25 of 27
Environmental Science & Technology
(ERE10,E85 - ERE0) / ERE0
5
SE10 WE10 SE85 WE85
4 3 2 1 0 -1
504
AR
BTEX
UNSAT
SAT
Oxygenates
505
Figure 4. Relative changes in individual VOC emissions in Driving Phase 1 from E0 averaged
506
by temperature and fuel and grouped by compound class: aromatics except BTEX (AR), BTEX,
507
unsaturated (UNSAT) and saturated (SAT) hydrocarbons, and oxygenates (Oxygenates) . Four
508
points are off scale for Oxygenates (SE85= 33.1, 44.3; WE85= 8.2, 94.0).
509
510
511
25 ACS Paragon Plus Environment
Environmental Science & Technology
Relative contribution of each VOC compound class to total OFP
ΣOFP (g O3/km) = 1.0
0.91
0.71
1.1
7.9
7.7
Page 26 of 27
18.1
0.8
UNSAT SAT Oxygenates BTEX AR
0.6
0.4
0.2
0.0 SE0
SE10
SE85
WE0
WE10
WE85
512 513
Figure 5. Relative contribution of each compound class (unsaturated (UNSAT) and saturated
514
(SAT) hydrocarbons, oxygenates (Oxygenates), aromatics except BTEX (AR) and BTEX)) to
515
Phase 1 total ozone formation potentials (ΣOFP) for each fuel. Average ΣOFP values are listed
516
above the respective columns.
517
518
519
520
521
26 ACS Paragon Plus Environment
Page 27 of 27
Environmental Science & Technology
(ERcold - ERwarm) / ERwarm
60 This work - E0 This work - E10 This work - E85 Lit. - E0-E5 Lit. - E10-20 Lit. - E75-85
50 40 30 20 10
1,
3-
522
n-
Bu
ta
di
en e H e Fo xa rm ne al de hy Ac de et al de hy de Ac ro le in Be n ze Et hy ne lb en ze ne To lu en m e ,p -X yl en e oXy le N ne ap ht ha le ne St yr en e M SA Ts
0
523
Figure 6. Relative changes in cold start mobile source air toxic (MSAT) emissions at cold
524
temperature for this work and the literature7, 9, 11, 27-30. Several data points for benzene and
525
naphthalene are off scale with a full scale version shown in the Supporting Information (Figure
526
S1).
27 ACS Paragon Plus Environment