1,383
Views
13
CrossRef citations to date
0
Altmetric
Technical Papers

Evaluation of solid particle number and black carbon for very low particulate matter emissions standards in light-duty vehicles

&
Pages 677-693 | Received 05 Jul 2016, Accepted 16 Nov 2016, Published online: 27 Mar 2017

ABSTRACT

To reliably measure at the low particulate matter (PM) levels needed to meet California’s Low Emission Vehicle (LEV III) 3- and 1-mg/mile particulate matter (PM) standards, various approaches other than gravimetric measurement have been suggested for testing purposes. In this work, a feasibility study of solid particle number (SPN, d50 = 23 nm) and black carbon (BC) as alternatives to gravimetric PM mass was conducted, based on the relationship of these two metrics to gravimetric PM mass, as well as the variability of each of these metrics. More than 150 Federal Test Procedure (FTP-75) or Supplemental Federal Test Procedure (US06) tests were conducted on 46 light-duty vehicles, including port-fuel-injected and direct-injected gasoline vehicles, as well as several light-duty diesel vehicles equipped with diesel particle filters (LDD/DPF). For FTP tests, emission variability of gravimetric PM mass was found to be slightly less than that of either SPN or BC, whereas the opposite was observed for US06 tests. Emission variability of PM mass for LDD/DPF was higher than that of both SPN and BC, primarily because of higher PM mass measurement uncertainties (background and precision) near or below 0.1 mg/mile. While strong correlations were observed from both SPN and BC to PM mass, the slopes are dependent on engine technologies and driving cycles, and the proportionality between the metrics can vary over the course of the test. Replacement of the LEV III PM mass emission standard with one other measurement metric may imperil the effectiveness of emission reduction, as a correlation-based relationship may evolve over future technologies for meeting stringent greenhouse standards.

Implications: Solid particle number and black carbon were suggested in place of PM mass for the California LEV III 1-mg/mile FTP standard. Their equivalence, proportionality, and emission variability in comparison to PM mass, based on a large light-duty vehicle fleet examined, are dependent on engine technologies and driving cycles. Such empirical derived correlations exhibit the limitation of using these metrics for enforcement and certification standards as vehicle combustion and after-treatment technologies advance.

Introduction

Research has demonstrated that chronic exposure to ambient particulate matter (PM) is associated with increased cardiopulmonary morbidity and mortality (Pope and Dockery, Citation2006; Brook et al., Citation2010). Meanwhile, PM from mobile sources is directly linked to adverse health outcomes, contributes to climate change, is detrimental to visibility, and is a major source in urban or near-road environments (Lloyd and Cackette, Citation2001; EPA, Citation2002; Hill et al., Citation2009; May et al., Citation2014). The California Air Resources Board (CARB) implemented Low Emission Vehicle (LEV) III standards for light duty passenger vehicles (LDVs) as part of the Advanced Clean Cars (ACC) program in January 2012. LEV III reduces the legacy 10-mg/mile PM standards over the Federal Test Procedure (FTP-75) (CARB, Citation2012a) to 3 mg/mile and 1 mg/mile, beginning with model years (MY) 2017 and 2025, respectively. LEV III standards also include separate PM standards applicable to the Supplemental FTP (US06, 10 mg/mile in 2017 and 6 mg/mile in 2019) test cycle, as well as higher interim in-use compliance emission limits (twice the LEV III certification standards).

With the adoption of the LEV III standards, CARB began confirming the feasibility of measuring PM emissions below the 1-mg/mile standard using the existing filter-based gravimetric test method (Hu et al., Citation2014; Sardar et al., Citation2016). This effort focused on quantifying uncertainties in gravimetric analysis and determining measurement precision of PM sampling, relative to vehicle repeatability and lab reproducibility. Meanwhile, Giechaskiel et al. (Citation2012) and Bushkuhl et al. (Citation2013) have suggested the approach of an “equivalent to PM” method for PM emission standards in the United States with several other measurement approaches (Bushkuhl et al. Citation2013; Giechaskiel et al., Citation2012). In general, these alternatives utilize real-time instrumentation that measures one or more non-mass characteristics of PM, which are then converted to PM mass with an empirically derived conversion factor. The most widely investigated approaches include the use of black carbon (BC), total or solid particle numbers (PN or SPN), and integrating particle mass using particle number, size, and effective density (integrated particle size distribution, or IPSD) (Khalek, Citation2006; Liu et al., Citation2009; Khalek et al., Citation2010; Maricq et al., Citation2011; Mamakos et al., Citation2012).

The Particle Measurement Programme (PMP) under the United Nations Economic Commission for Europe (UNECE) measures solid particles greater than 23 nm (d50 = 23 nm) for the Euro 5/6 standards (Andersson et al., Citation2007; Martini et al., Citation2009; Andersson et al., Citation2010; Giechaskiel et al., Citation2012). Volatile and semi-volatile PM in diluted exhaust is removed by either an evaporation tube, a thermal denuder, or a catalytic stripper, and the soot-dominant refractory PM2.5 is measured and reported in numbers with a particle counter. This approach is used to minimize the high measurement variability as a result of nucleation of semi-volatile PM with different dilution patterns. While SPN can be measured quickly and relatively simply, its correlation to adverse health effects is less investigated than the effects of PM mass on health. Removal of semi-volatile organic compounds excludes the ability to assess and determine the impact of potential secondary organic aerosol (SOA) formation from vehicle emissions on ambient air quality, as primary emissions continue to decrease. Moreover, criteria of particle generation and calibration materials for the PMP instrument calibration method (primary method) are still under development and implementation. Routine PMP instrument calibration primarily relies on verification and adjustment to simultaneous measurements by a reference or “golden” solid particle counter. Several studies also have reported that sub-23-nm solid particles can make up a significant fraction of total solid particle emissions and may vary substantially among different engine technologies (Zheng et al., Citation2011; Mamakos et al., Citation2013). The PMP Working Group is currently investigating the feasibility of a lower cutoff size (41st PMP Working Group Meeting, Citation2016).

BC is often of great interest because it comprises a significant fraction of PM emissions and can be a strong surrogate for tailpipe PM mass. BC is emitted as a result of incomplete combustion, is a radiative agent for climate forcing, and shows a strong correlation to adverse public health (Menon et al., Citation2002; Gauderman et al., Citation2004; Intergovernmental Panel on Climate Change [IPCC], Citation2007; Ramanathan and Carmichael, Citation2008; Mordukhovich et al., Citation2009). Bushkuhl et al. (Citation2013) reported success in reducing PM mass emission variability by apportioning PM mass, accumulated over a whole driving cycle, with real-time BC measurement for two GDI base vehicles at three testing configurations. These results indicated a relationship of BC to PM mass of 94%, ranging up to 40 mg/mile. However, this relationship may not be applicable at very low emission levels, such as those needed to meet the 3- or 1-mg/mile standards.

Some research has suggested reconstructing PM mass with either its chemical or its physical properties (Kweon et al., Citation2002; Zielinska et al., Citation2004; Schauer et al., Citation2008; Liu et al., Citation2009). A sum of total chemical speciation showed good agreement to PM mass (within 20%), but this approach is unlikely to hold for very low PM emissions due to the inherent artifacts of organic carbon collected on quartz fiber filters. In addition, chemical species are more likely to fall below detection limits as the sample air volumes (up to 2 m3) are predetermined by sampling procedures described in 40 CFR Part 1066 (Citation2015). Others have calculated PM mass by integrating 1-Hz measurements of particle numbers and sizes, multiplied by their corresponding particle density factors (IPSD; Liu et al., Citation2009; Li et al., Citation2014; Xue et al., Citation2015; Quiros et al., Citation2015). Quiros et al. showed a moderate improvement in measurement repeatability with IPSD but found PM mass emissions are underestimated, likely due to the limited range of particle sizes measured by instrumentation, inaccuracy in the size response, and predefined particle density functions (Quiros et al., Citation2015).

A primary objective of the LEV III PM standards is to ensure future vehicles continue to have very low PM emissions while meeting increasingly stringent criteria pollutant and greenhouse gas standards. By testing a large fleet of LDVs that includes vehicle technologies that potentially meet a 1-mg/mile PM standard, we are able to evaluate the feasibility of using either SPN or BC as an alternative to PM mass, based on their relationship to PM mass. This includes an assessment of the BC fraction of the PM emissions in this vehicle fleet, and an analysis of the variability of all three metrics. Finally, to gain a better understanding of the measurement variability of these metrics in emission testing, we compare normalized cumulative SPN and BC emission profiles for various engine technologies and test cycles.

Methods

Emission testing vehicle, cycle, and fuels

Emission testing was conducted in three light-duty test cells at the CARB Haagen-Smit Laboratory (HSL) in El Monte, CA, as shown schematically in . Each test cell is equipped with a 48-inch single-roll electric chassis dynamometer, a constant-volume sampler (CVS), and one or more PM2.5 sampling systems that meet requirements defined by 40 CFR Part 1065/1066. A more detailed description of the test cells, typical test procedures, and the test cell reproducibility can be found in Hu et al. (Citation2014).

Figure 1. Schematic of testing setup, showing vehicle, CVS tunnel, and PM instrumentation.

Figure 1. Schematic of testing setup, showing vehicle, CVS tunnel, and PM instrumentation.

Testing was conducted over both FTP-75 and US06 cycles at typical CVS flow rates of 350 and 750 standard cubic feet per minute (scfm, or 9,910 and 21,237 standard liters per minute, within 2%), respectively. An FTP-75 test is intended to represent an urban driving pattern with a cold start phase, a stabilized phase, and a hot start phase, labeled phases 1–3, respectively. The FTP-75 cycle was originally based on real-world driving and its acceleration was limited to accommodate the vehicle dynamometers in the 1970s. The US06 cycle is comprised of a single phase and is completed after a warmup cycle. It was intended to represent more aggressive, high-speed, and high-acceleration driving. The US06 cycle has an average speed of 48.4 miles per hour (mph), a top speed of 80.3 mph, and a maximum acceleration of 8.46 mph/sec. In comparison, the FTP-75 cycle has an average speed of 21.2 mph, a top speed of 56.7 mph, and a maximum acceleration of 3.3 mph/sec (40 CFR Part 86, Citation1996).

In total, 46 vehicles were tested over several test programs, including 22 with port fuel injection (PFI), 20 with gasoline direct injection (GDI), and 4 light- or medium-duty diesels equipped with diesel particulate filters (LDD/DPF or MDD/DPF). Test fuels for the gasoline vehicles included California Phase III certification-grade or commercial-grade gasoline containing 10% ethanol, with two exceptions: One vehicle (2011 Grand Caravan) was tested with certification grade E-85 fuel, and another vehicle (2011 Buick Regal) was tested with EPA indolene certification gasoline that contained no ethanol. All of the LDD/DPF vehicles were tested with California commercial-grade diesel containing less than 15 ppmw sulfur.

Instrumentation and measurements

PM mass was collected with an AVL SPC 472 Smart Sampler (Anstalt für Verbrennungskraftmaschinen, Germany), and/or a collocated Horiba Quad Sampler (Kyoto, Japan), using 47-mm Teflon filters (2 µm, Whatman, USA) at a targeted nominal flow rate of 2.3 scfm (65 lpm) and sample air temperature at 47 ± 5ºC. Gravimetric analysis for PM samples is conducted in a temperature- and dew-point-controlled cleanroom, with an automated filter weighing system (FWS; MTL, Minneapolis, MN) equipped with a seven-decimal microbalance (Mettler Toledo, Inc., Columbus, OH). The PM gravimetric analyses meet the requirements in 40 CFR Part 1066, and details of gravimetric analysis and filter media handling can be found elsewhere (Zhang and McMahon, Citation2012; Hu et al., Citation2014; Sardar et al., Citation2016).

SPN and BC were also drawn from the CVS tunnel, near where the PM mass samples were drawn. SPN was measured with a PMP-compliant Horiba Solid Particle Counting System (Horiba MEXA-2000 SPCS), which consists of a volatile particle removal (VPR) unit and a condensation particle counter (CPC). BC was quantified with an AVL Micro Soot Sensor (MSS 483; Graz, Austria), which measures BC with photoacoustic spectroscopy at 808 nm. SPN and BC measurements were recorded at 1 Hz and/or 2 Hz frequencies, respectively.

PM mass tunnel blanks (filtered dilution air sampled at the end of the CVS) were collected regularly and were measured at approximately 0.1 mg/mile. The equipped multi-PM sampling systems were evaluated for their equivalency (at 95% confidence level) and yielded a PM mass measurement precision of 11.1% at a PM mass loading at or above 10 µg (~0.1 mg/mile; Sardar et al. Citation2016). Thus, this measurement precision is used as a criterion for equivalency assessment and emission about 0.1 mg/mile is the focus of comparison. Earlier work has measured tunnel blank values of 2 × 109 particles/km for SPN and 0.15 mg/mile for BC (CARB, Citation2012b; Kamboures et al., Citation2013). To simplify the comparisons between different PM metrics and to be consistent with companion papers (Quiros et al., Citation2015; Kamboures et al., Citation2015; Sardar et al., Citation2016), tunnel blank subtraction was not conducted for this evaluation.

Data collection and reduction

A data-processing program was developed using Igor Pro 6.3.4 (Wavemetrics, Inc.) to integrate raw real-time data from both the AVL MSS and the Horiba MEXA-2000 SPCS. The processing program incorporates additional test information, including CVS flow rate, vehicle speed, and test parameters. The total SPN count is calculated for each phase of a test, as

(1)

where SPNtot is the total solid particle count for the test cycle, SPN(t) is the particle count at time t, Q(t) is the CVS flow rate, and PCRF refers to the particle concentration reduction factor, a calibration factor that includes internal dilution and mean particle losses, which was set by the instrument manufacturer and is compliant with PMP protocol.

The total BC mass for a test phase is calculated as

(2)

where BCtot is the total black carbon for the test cycle and BC(t) is the real-time black carbon mass corrected with the corresponding dilution ratio at time t.

Each real-time instrument was controlled by a stand-alone laptop, and data collection was initiated several minutes before each test cycle. Although the laptop clocks were regularly synchronized, there was occasional drift in the time stamps. To align the real-time data files with one another, the data-processing program locates the first significant data spike, which is assumed to be the beginning of the test. The difference in time stamps between the real-time data and emission testing were recorded, and any difference of more than a few minutes was investigated and the test was invalidated if necessary. This time alignment approach is consistent with earlier work, but here it is automated as a batch process (Franco García, Citation2014; Rasdorf et al., Citation2010).

For quality control, the program generates a summary chart for each test, which includes a plot of the real-time data superimposed on the actual vehicle speed. These charts enable rapid evaluation of any problems in the automated data processing. The most common anomaly arose when an errant measurement spike triggered a false test start and caused a time alignment problem; any such issues were either corrected or invalidated. The emissions per phase were then used to calculate emissions per mile, using the standard equations for each test cycle. A number of the automated results were cross-checked with manual data processing, and all cross-checked results were within 2% of manual calculations.

Results and discussion

Test results and a list of the corresponding vehicles and number of tests performed are presented in , including 127 FTP-75 and 87 US06 tests. Due to occasional data recovery failure, instrument availability, and experimental design for each test program, not all tests have collocated SPN, BC, and PM mass measurements. Depending on the test program, some vehicles were tested repeatedly, while others were only tested once over the FTP-75 and/or US06.

Table 1. Vehicles tested, including mileage, fuel injection type, emission standard used for certification, engine displacement and cylinder number, transmission type, presence of turbocharger, number of FTP-75 or US06 tests, and average emissions results and standard deviations for PM mass, SPN, and BC (no diesel particulate filter regeneration).

Linear regression and coefficient of variation (CoV) are used for measurement method comparison because random errors in measurements, background contribution, vehicle emissions variability, and lab reproducibility are usually in normal distribution. The linear relationship allows determination of the design margins needed and/or available to meet emission standards, assessment of effectiveness and planning of emissions reduction to meet the health-based ambient PM2.5 standard in the United States, and calibration of measurement instrument (such as PCFR), if emissions were determined with an “equivalent to PM mass” method.

Emissions variability

CoVs for SPN, BC, and PM mass were calculated for all vehicles with three or more repeat tests. The average emissions and CoV for each vehicle are shown in , separated by test cycle and engine technology.

Figure 2. Emissions variability of PM mass, SPN, and BC by vehicle for (a) FTP-75 and (b) US06 tests. Data and vehicles are shaded in .

Figure 2. Emissions variability of PM mass, SPN, and BC by vehicle for (a) FTP-75 and (b) US06 tests. Data and vehicles are shaded in Table 1.

shows that when comparing emissions variability for individual vehicles, there is no clear trend that one metric is better than another for either test cycle. The result agrees with prior findings that much of the variability is due to vehicle and driver (Kamboures et al., Citation2013; Hu et al., Citation2014; Kamboures et al., Citation2015; Quiros et al., Citation2015). As previously mentioned, the range of PM mass emissions are distinct for different engine technologies. GDI vehicles have higher emissions and lower CoVs than PFI vehicles, as expected since both measurement uncertainty and CoVs tend to decrease at higher emission levels. The average PM mass emission rate for LDD/DPF was 0.11 mg/mile, near or below the reported tunnel blank level of 0.1 mg/mile (Hu et al., Citation2014; Quiros et al., Citation2015; Sardar et al., Citation2016). Very low emitting vehicles have high CoVs for PM mass, most notably LDD/DPF vehicles, essentially because CoV is magnified when the signal (emission) is very close to the background noise (random contribution from dilution air and decrease in measurement precision).

To illustrate the overall emissions variability of SPN, BC, and PM mass, the pooled standard deviations were compared to the pooled average of repeat tests over a large number of vehicles. The results are shown in , separated by engine technologies and test cycles. For FTP-75 tests, the PM mass emission rates for PFI vehicles are mostly below 1 mg/mile, whereas the average PM mass emission rate for GDI is 5.67 mg/mile, within the range of what has been reported before (Maricq et al., Citation2011; Chan et al., Citation2012; Maricq et al., Citation2013). The average US06 emission rates of PM mass for PFI and GDI vehicles were 1.84 and 6.82 mg/mile, respectively, all below the interim LEV III 2017 US06 certification standard of 10 mg/mile. These results also provide an estimation of improvement needed in emission reduction (average PM mass emission rates) and design margins (pooled standard deviations) for the fleet of each tested vehicle technology to meet the 1-mg/mile emission standard.

Table 2. Summary of repeatability for gravimetric PM mass, SPN, and equivalent BC.

For FTP-75 tests, pooled emission variability of PM mass is slightly less than that of either SPN or BC, when defined as a percentage of the pooled standard deviation to the average emissions. For PFI vehicles, variability of PM mass is 20.9%, compared to 27.7% and 23.3%, respectively, for SPN and BC. For GDI vehicles, the emissions variability is 5.7%, 13.3%, and 9.2%, respectively, for PM mass, SPN, and BC. The emissions variability for the US06 is higher than that of the FTP-75 for each of these three metrics. Also, US06 emissions variability of both PM mass and BC are significantly higher than the variability of SPN, in contrast to FTP-75 results. This is consistent with previous work (Chan et al., Citation2014; Giechaskiel et al., Citation2014), and may be caused by a combination of additional vehicle and driver variability for the aggressive, high-acceleration US06 cycle.

Relationships between SPN and PM mass

shows the relationship of SPN to PM mass for FTP-75 and US06 test cycles, presented for all vehicles combined, GDI vehicles only, and PFI vehicles only. Figures are presented in log scale to better demonstrate the distribution of data below 1 mg/mile. Results are presented separately for the two test cycles to highlight the distinct differences in scatter and slope between the two cycles. Because of limited data for LDD/DPF vehicles, no linear fit is presented. Vehicle emissions at that level have sufficient margin to meet the 1-mg/mile standard, despite the higher uncertainties in PM mass measurement and the lack of linearity between measurement methods. The analysis of these correlations is conducted based on engine technologies, not the range of PM emission rates, because data showed distinct ranges for different engine technologies. The overall fit presented will vary depending on the proportion of different engine technologies in vehicle pool, and may be of limited use.

Figure 3. SPN (particles/mile) vs. gravimetric PM mass (mg/mile) for the entire available data set of (a) FTP-75 tests and (b) US06 tests, using weighted average of three phases.

Figure 3. SPN (particles/mile) vs. gravimetric PM mass (mg/mile) for the entire available data set of (a) FTP-75 tests and (b) US06 tests, using weighted average of three phases.

shows that the overall slope of SPN to PM mass in FTP-75 test cycles is 1.5 × 1012 particles/mg, with an R2 of 0.86, which is dominated by that of GDI vehicles (also 1.5 × 1012 particles/mg), whereas PFI vehicles show a higher slope of 2.3 × 1012 particles/mg. No significant difference (within measurement precision of 11%, and intercept less than each measurement detection limit) in the slope or R2 was found either with or without zero-intercept for both GDI and PFI vehicles in the linear regression analysis (same for BC and PM mass). While the slopes for the US06 are in a narrow range, 1.5 × 1012 to 1.6 × 1012 particles/mg, the data are more scattered (R2 = 0.37–0.58) compared to the FTP-75 (R2 = 0.83–0.88), as shown in . It is notable that the scatter does not appear to be larger for emissions below 1 mg/mile, especially for US06 testing.

A number of studies have reported a positive relationship between SPN and PM mass. The slopes ranged from 1 × 1012 (heavy duty, HD) to 4 × 1012 (LD) particles/mg and had varying scatter (Khalek, Citation2006; Maricq et al., Citation2011; Giechaskiel et al., Citation2012; Chan et al., Citation2014) for emissions up to 40 mg/mile. The varying slope is due to the wide range of test vehicle fleet and test cycle (Euro vs. U.S.), emission standards that test vehicles were certified at, vehicle classifications (LD vs. HD), combustion technologies (air/fuel ratios), after-treatment technologies (particulate filter for engines), and fuels (diesel vs. gasoline). Giechaskiel et al. also attributed the disappearing correlation below 3.2 mg/mile (2 mg/km) for LD to the positive adsorption of volatile organics onto filter media (Teflon-coated glass fiber filter, minimum 20 µg). However, the use of Teflon filters in 40 CFR Part 1066 greatly reduces PM mass measurement uncertainty (typically 2–3 µg) and thus improves the correlation.

These wide ranges of slope occur also because a particle number to mass correlation is not a general phenomenon in aerosols. When particle numbers are constant, PM mass greatly depends on the measured size, density, and shape; whereas when PM mass is constant, aerosol nucleation (for total particle number) and agglomeration are the most important mechanisms for both solid and total particle number counting (Hinds, Citation1999; Lipsky et al. Citation2002; Chang et al. Citation2004). A constant slope between SPN and PM mass requires relatively constant and stable particle size distributions (spread and size range), count median diameter (CMD), and chemical compositions of emissions from various test vehicles and technologies. SPN measurement uses the PMP protocol, which shifts the CMD after removing nucleates and semi-volatile compounds, but does not change the aerosol formation mechanism in primary emissions. Particle agglomeration and nucleation occur when dilution air mixes with exhaust. Depending on the PM chemical composition (volatile vs. non-volatile) and abundance in the exhaust, dilution air ratio, and dilution mixing pattern (Reynolds numbers), a pseudo-stabilized particle size distribution is reached before PM measurements are conducted. However, it is unlikely that the chemical composition of PM emissions and emission rates stay unchanged with more advanced combustion and after-treatment technologies, as well as the increasing oxygenated fraction (ethanol) in fuels to meet the low carbon fuel standard.

CARB had previously reported a slope of 2.5 × 1012 particles/mg in 2012 for a combined fleet of 2009 and 2010 model year PFIs and GDIs (CARB, Citation2012b) with different test vehicles. It is much higher than the 1.5 × 1012 particles/mg determined in this study, while these earlier test vehicles are virtually a subgroup of those in this study. These results suggest SPN and PM mass correlations may vary with different engine technologies, and could continue to evolve as combustion and aftertreatment technologies advance.

Correlation between BC and PM mass

The relationships between BC and PM mass for FTP-75 and US06 cycles are shown in . For FTP-75 tests as shown in , BC to PM mass ratios show an overall ratio of 0.89, but once again with a distinction between engine types (0.76 for PFI and 0.90 for GDI). The correlation between BC and PM mass for US06 tests shows a similar trend, as shown in , except with a significantly lower BC/PM mass ratio (~0.7). The high R2 of BC/PM mass relationship for US06 tests is in contrast to that of SPN/PM correlations (R2 <0.58).

Figure 4. BC (mg/mile) vs. gravimetric PM mass (mg/mile) for all available (a) FTP-75 and (b) US06 tests.

Figure 4. BC (mg/mile) vs. gravimetric PM mass (mg/mile) for all available (a) FTP-75 and (b) US06 tests.

Although the BC/PM mass ratio of 0.89 for overall vehicles in this study is close to the 0.94 reported by Bushkuhl et al. (Citation2013), it is much higher than the EC/PM mass ratio of 0.7 that CARB reported in 2012 (BC and EC was about 1:1, CARB, Citation2012b). The BC/PM mass ratio for GDI vehicles in this study is 0.90, also higher than 0.70 reported in 2012. The highly correlated R2 of 0.97 was found in this study for overall vehicles (excluding LDD/DPF), whereas Bushkuhl et al. reported a notable increase in scattering between BC and PM mass below 5 mg/mile. This variability of BC/PM ratios is in agreement with Nam et al. (Citation2008), suggesting that the relationship between these two metrics depends on the LDV fleet in a test program—and may change as vehicle technology evolves. The change in relationship may also occur over less aggressive test cycles as well, with some vehicle technologies. Therefore, although high correlations were determined over the FTP-75 test cycle, applying a constant BC/PM mass ratio to determine PM mass should be considered with caution.

Patterns of real-time SPN and BC emissions

SPN and BC are often used to evaluate the effectiveness of emission reduction strategies, or as surrogates for PM during combustion engine designs. This effort is thought to be the first to include a large data set of real-time emissions characteristics over different U.S. test cycles, using different light-duty technologies in California, with their relationship to PM mass. The real-time emission patterns in a driving cycle of these metrics are especially important for understanding the potential of using these metrics for evaluating emission standards, as well as sources of emission variability.

To evaluate potential differences between different vehicles and different engine technologies, five vehicles were selected to represent varying patterns of emissions. Vehicle selection was based on over 60 BC and SPN measurements for seven PFIs, three GDIs, and three LDD/DPFs by reviewing (1) availability of concurrent PM mass measurement results, (2) representativeness of BC and SPN accumulation patterns of real-time data through a test cycle, (3) representativeness of BC and SPN contributions from each phase to total emissions, and (4) range of PM mass emission rate for each technology category, to meet the 1 mg/mile standard.

The selected vehicles were typical for each technology, and the selected test represents typical or median patterns in a test cycle for that vehicle. These selected vehicles include three PFIs (2011 Nissan Altima, 2009 Toyota Camry, and 2012 Honda Civic), one GDI (2008 Mini Cooper), and one LDD/DPF (2012 Volkswagen Passat TDI). The PFI vehicles were selected to approximately represent three different emission patterns that were observed, while the GDI and LDD vehicles were each broadly representative of their engine technology.

The normalized cumulative real-time SPN and BC emission profiles of the selected vehicles are shown in over the FTP-75 test cycle. This data analysis is focused on the FTP-75 partly because of the abundance of collocated SPN and BC measurements for comparison, and also because of the historical reliance on the FTP-75 emission standards as the principal emission standard for vehicles. shows that there is no consistent SPN or BC emissions pattern for various LDV technologies, although all of these vehicles emit PM during the warm-up/cold-start phase and under acceleration, as expected. Compared to the GDI vehicle, which reaches 60% of its total SPN and ~73% of its BC emissions by the end of phase 1, PFI vehicle emissions can be either higher (e.g., Civic and Altima at ~90% for both SPN and BC) or lower (Camry at ~60% for SPN and 50% for BC) for the same period. The Camry’s normalized cumulative SPN emissions show a noticeable increase in phase 3, which is more than for the other gasoline vehicles.

Figure 5. Normalized cumulative (a) total SPN and (b) total BC for five selected vehicles over FTP-75, for tests that were selected to be typical for each vehicle.

Figure 5. Normalized cumulative (a) total SPN and (b) total BC for five selected vehicles over FTP-75, for tests that were selected to be typical for each vehicle.

When comparing the GDI to LDD/DPF vehicles, the accumulation rates of SPN and BC both show a steady increase in phases 2 and 3, except that the LDD/DPF vehicle has a step increase of SPN in end of phase 3, and its SPN only shows less than a 5% rise in phase 2. In phases 1 and 2, the normalized cumulative BC is higher for the GDI vehicle than for the LDD/DPF vehicle, but this is the opposite for SPN. This observation indicates that SPN and BC are distinct measurements that cannot be interchangeable, and emphasizes the importance of understanding the magnitude of differences for various LDV technologies (Giechaskiel et al., Citation2012), as discussed later.

Relationship between modal SPN and BC for engine technologies

The relationship between SPN and BC emissions over a test cycle can be evaluated by comparing the difference between normalized cumulative SPN and BC. This method has an advantage over comparing ratios of the actual measurements because it minimizes bias when a peak is not quantifiable, or when a ratio was amplified by a small denominator. Here, a metric to compare the relative emissions is defined as

(3)

where the sums of SPN(t) and BC(t) are the cumulative SPN and BC emissions at time t (in seconds), and SPNtot and BCtot are the total SPN and BC at the end of test cycles. This is essentially the difference between the accumulated SPN and accumulated BC in . CVS flow and mileage do not affect the results because σ is the difference between these two ratios at any given second. When σ > 0, it suggests that accumulated emissions of SPN is greater than accumulated BC at time t, and vice versa when σ < 0. Since σ is the difference between the normalized metrics in a test cycle, the start and end values are always zero. If the two metrics were interchangeable and correlated (i.e., had a linear relationship with 10% precision), σ should be within ±0.10 at any time throughout the test cycle and at the end of each test phase (Sardar et al., Citation2016). This analysis can also explain the different BC and SPN emission variabilities (CoVs), or other measurement methods, observed for the same test vehicle. Unfortunately, quantitative real-time PM mass measurement is not available at this low emission level.

presents the σ values for the five selected vehicles, using test results in . It is evident that these five vehicles have distinct differences in their SPN and BC emissions characteristics, in both the values and magnitude, and not only overall but during different phases of a cycle. The GDI (Mini Cooper) has a negative σ of –0.28 in the beginning of the test and slowly returns to zero by the end of the FTP-75 test cycle, which suggests that most of the BC is emitted during the cold start, while SPN is emitted throughout the test cycle. In contrast, the first two SPN peaks for the VW Passat are consistently contributing more to total SPN than BC, which results in a positive σ of up to 0.35 in phase 1. The BC emissions for this LDD/DPF vehicle are little more than the background noise, and slowly accumulate to converge by the end of phase 3.

Figure 6. Relative accumulation of SPN and BC over FTP-75, for the five selected vehicles.

Figure 6. Relative accumulation of SPN and BC over FTP-75, for the five selected vehicles.

The σ patterns for the three PFI vehicles are also distinct from one another. The σ value for the Nissan Altima (~0.9 mg/mile) is very small, with only minor variations (<0.05) in both directions throughout the test cycle. For the Honda Civic, although normalized cumulative BC is higher than SPN (σ as low as –0.15) during the first 200 seconds, σ is ±0.05 for the rest of the FTP-75 cycle. However, for the Toyota Camry, the pattern is similar but opposite to that of the GDI vehicle, in that SPN dominates the early portion of the test while BC dominates over phase 3.

The emissions over a standard three-phase FTP-75 test are calculated using a weighted average; hence, the differences in σ values for different phases could cause difficulties in using one metric as a stand-in for another. In addition, the uncertainty in alternating measurement metrics for emissions rate determination also depends on the emissions rates/levels (source strength). For example, the σ values for both the Mini Cooper and the VW Passat in are still significant at the end of both phase 1 and phase 2. Using SPN as an alternative for BC emission rate will result in an underestimation of BC for the Mini Cooper because the actual BC emission rate is 2.33 mg/mile, whereas it is negligible for the Passat because the BC emissions are near the measurement background noise (0.15 mg/mile).

Results in show the limitation of alternate usage between SPN and BC. Studies have shown that not all solid particles are made up of soot (Giechaskiel et al, Citation2010). Moreover, SPN and BC measure different components of PM. Even if all of the counted SPN is BC, BC mass depends on the SPN number and size distributions—which can vary over tests, vehicles, and engine types. The differences between SPN and BC over the course of the test, and particularly after cold start, suggest that the chemistry of emitted PM is not consistent across different vehicles and engine technologies. And finally, a different conversion factor might be necessary in photoacoustic spectroscopy for test cycles or emissions that are rich in organic carbon (e.g., US06), to correct for the bias introduced by the organic carbon coating on aerosols that amplifies the measured absorption cross section (Bueno et al., Citation2011; Kamboures et al., Citation2013), and to better understand the relationship between SPN and BC, or BC and PM mass.

SPN and BC repeatability in phases 2 and 4

Most routine FTP-75 testing at CARB is completed as a three-phase test. For the Nissan Altima only, a number of FTP-75 tests were concurrently conducted with a four-phase cycle (i.e., two consecutive urban dynamometer driving schedules [UDDS]), where phase 4 is an immediate repeat of phase 2 after phase 3. Thus, emissions of phases 2 and 4 in the FTP-75 cycle are generally assumed to be identical, as they are the same driving pattern run with the same vehicle, driver, pretest preparation, measurement instrumentation calibration, and vehicle conditions, while the vehicle is warm and stabilized. However, a four-phase cycle requires longer testing time (i.e., testing cost) and thus is seldom used.

Real-time SPN and BC measurements of four-phase FTP-75 tests allow examining the contribution of vehicle stability to emissions variability with a comparison of the emissions between phases 2 and 4. Overall, the emissions of the Altima were typical for a PFI vehicle at around 1.0 mg/mile, with a CoV of 0.21 (). The normalized cumulative BC and SPN emission profiles for the Altima are shown in . For most tests, phase 4 emissions were higher than in phase 2, often significantly. For SPN, the average emissions over phase 2 was 7.7% (±6.4%), and 20.3% (±20.1%) for phase 4. The average BC emissions for phase 2 was 13.8% (±9.4%) of the total BC, and 21.2 (±18.7%) for phase 4. Results of PM mass emissions for test 1 and 2 also showed that emission in phase 4 is significantly higher than in phase 2 (15% vs. 6% for test 1 and 32% vs. 1% for test 2).

Figure 7. Normalized cumulative (a) total SPN and (b) total BC for the Nissan Altima of four-phase FTP-75 repeat tests (two UDDS cycles). These plots are normalized to 100% at the end of phase 3. Note that PM mass in phases 2 and 4 rise 6% and 15%, respectively, for Test 1, and 1 and 32%, respectively, for Test 2. Phase 4 is a phase 2 test cycle immediately after phase 3.

Figure 7. Normalized cumulative (a) total SPN and (b) total BC for the Nissan Altima of four-phase FTP-75 repeat tests (two UDDS cycles). These plots are normalized to 100% at the end of phase 3. Note that PM mass in phases 2 and 4 rise 6% and 15%, respectively, for Test 1, and 1 and 32%, respectively, for Test 2. Phase 4 is a phase 2 test cycle immediately after phase 3.

Different SPN and BC emissions between phases 2 and 4 were examined by superposing the real-time measurements on vehicle speed traces, as shown in . The actual speed of the vehicle showed very little variation between the two phases. Most emissions occur during acceleration, as expected, but periods of acceleration in phase 4 were more likely to result in measurable tailpipe emissions than in phase 2, for both SPN and BC. On occasion, deceleration also resulted in a small emissions peak, for example, the first deceleration of phase 4 in SPN (). Because of very limited availability of four-phase real-time PM data, it is unclear whether the higher emissions in phase 4 are unique to this vehicle, or a typical occurrence of certain engine technologies. The equivalence of PM, BC, and SPN emissions between phase 2 and 4 needs further investigation.

Figure 8. Comparison of (a) SPN and (b) BC phases 2 and 4, for Nissan Altima in a four-phase FTP-75 test (two UDDS cycles).

Figure 8. Comparison of (a) SPN and (b) BC phases 2 and 4, for Nissan Altima in a four-phase FTP-75 test (two UDDS cycles).

Conclusion

With a large data set of 46 LDVs tested, this study (as well as results from companion papers) clearly shows that each proposed measurement metric is measuring a different component of particulate emissions and they are not interchangeable for PM emission standards. The emissions variability of SPN, BC, and PM mass attributes to the interference, artifacts, and uncertainties in measurement principles; modal emissions characteristics and profiles of the measurement metrics; and varying repeatability of emissions from the vehicle over the speed trace. Other sources of limitation include dynamic measurement ranges in instrumentation to quantify emissions from transient processes, limited range of metric measured (upper or lower size boundaries), and the lack of calibration for measurement test methods with a known standard or procedures.

Our observations suggest there is significant difficulty in using any one of these metrics as a simple alternative for PM mass at this time. The correlations between a surrogate metric and PM mass vary over different engine families, engine technologies, and driving cycles. While a more complex conversion may be possible, for example, if PFI engines result in different correlations than GDI engines, large variations would suggest that these correlations will almost certainly change as vehicle technology continues to evolve. Correlations and variability could also change as manufacturers calibrate engine power trains for more stringent emission standards.

However, the correlations among these measurement metrics are directional; that is, low PM mass emissions (to 0.1 mg/mile) consistently correspond to low SPN and BC emissions. Consequently, SPN and BC measurements can provide semi-quantitative approaches to evaluate PM emissions on a real-time basis. In conjunction with an on-board diagnostic system (OBD), these metrics may be particularly useful to evaluate new emission controls, to inform control strategy calibration or engine development work, or to provide near immediate feedback of PM emissions during laboratory testing programs.

Disclaimer

The statements and opinions expressed in this paper are solely the authors’ and do not represent the official position of CARB. The mention of trade names, products, and organizations does not constitute endorsement or recommendation for use. CARB is a department of the California Environmental Protection Agency. CARB’s mission is to promote and protect public health, welfare, and ecological resources through effective reduction of air pollutants while recognizing and considering effects on the economy. CARB oversees all air pollution control efforts in California to attain and maintain health-based air quality standards.

Acknowledgment

The authors thank the CARB management and staff members who conducted emissions testing. The authors also thank David Eiges, Satya Sardar, Shiyan Chen, Pippin Mader, Bruce Frodin, Shiou-mei Huang, David Quiros, Michael Kamboures, and Mang Zhang for consultation on data validation, as well as Michael McCarthy and Alberto Ayala for providing policy direction and careful review of this paper.

Additional information

Notes on contributors

M.-C. Oliver Chang

M.-C. Oliver Chang is the manager and J. Erin Shields is a staff member of the Aerosol Analysis and Methods Evaluation Section of the California Air Resources Board.

J. Erin Shields

M.-C. Oliver Chang is the manager and J. Erin Shields is a staff member of the Aerosol Analysis and Methods Evaluation Section of the California Air Resources Board.

References

  • 40 CFR Part 86. 1996. Appendix I—Dynamometer Schedules. http://www.ecfr.gov/cgi-bin/retrieveECFR?gp=&SID=fbb38ac0d3a44e15ace72120abce973e&mc=true&n=pt40.19.86&r=PART&ty=HTML#ap40.21.86_11931_686_11999.i (accessed January 20, 2017).
  • 40 CFR Part 1066. 2015. Vehicle Testing Procedures. http://www.ecfr.gov/cgi-bin/text-idx?SID=ba447754d6f766672ab21e5aa4146283&mc=true&node=pt40.33.1066&rgn=div5 (accessed January 20, 2017).
  • 41st PMP Working Group Meeting. 2016. In UN-ECE Particle Measurement Programme, Brussels, Belgium.
  • Andersson, J., B Giechaskiel, R. Munoz-Bueno, E., Sandbach, and P. Dilara. 2007. Particle Measurement Programme (PMP) Light-Duty Inter-Laboratory Correlation Exercise (Ilce_Ld).European Commission Joint Research Centre. http://publications.jrc.ec.europa.eu/repository/bitstream/111111111/429/2/7386%20-%20PMP_LD_final.pdf (accessed February 8, 2017).
  • Andersson, J., A. Mamakos, B. Giechaskiel, M. Carriero, and G. Martini. 2010. Particle Measurement Programme (PMP) Heavy-Duty Inter-Laboratory Correlation Exercise (Ilce_Hd) Final Report. EC-JRC Sci. and Tech. Reports. http://publications.jrc.ec.europa.eu/repository/bitstream/111111111/15075/1/pmp%20hd%20validation%20exercise%20final%20report.pdf (accessed February 8, 2017).
  • Brook, R.D., S. Rajagopalan, C.A. Pope, J.R. Brook, A. Bhatnagar, A.V. Diez-Roux, F. Holguin, Y. Hong, R.V. Luepker, and M.A. Mittleman. 2010. Particulate Matter air pollution and cardiovascular disease, an update to the scientific statement from the American Heart Association. Circulation 121:2331–78. doi:10.1161/CIR.0b013e3181dbece1
  • Bueno, P.A., D.K. Havey, G.W. Mulholland, J.T. Hodges, K.A. Gillis, R.R. Dickerson, and M.R. Zachariah. 2011. Photoacoustic measurements of amplification of the absorption cross section for coated soot aerosols. Aerosol Sci. Technol. 45:1217–30. doi:10.1080/02786826.2011.587477
  • Bushkuhl, J., W. Silvis, J. Szente and M. Maricq. 2013. A new approach for very low particulate mass emissions measurement. SAE Int. J. Engines 6(2):1152–62. doi:10.4271/2013-01-1557
  • CARB. 2012a. Development of particulate matter mass standards for future light-duty vehicles. Sacramento, CA: CARB.
  • CARB. 2012b. LEV III Appendix P Technical support document development of particulate matter mass standards for future light-duty vehicles. Sacramento, CA: CARB.
  • Chan, T.W., E. Meloche, J. Kubsh, and R. Brezny. 2014. Black carbon emissions in gasoline exhaust and a reduction alternative with a gasoline particulate filter. Environ. Sci. Technol. 48:6027–34. doi:10.1021/es501791b
  • Chan, T.W., E. Meloche, J. Kubsh, D. Rosenblatt, R. Brezny, and G. Rideout. 2012. Evaluation of a gasoline particulate filter to reduce particle emissions from a gasoline direct injection vehicle. SAE Int. J. Fuels Lubr. 5:1277–90.doi:10.4271/2012-01-1727
  • Chang, M.-C.O., J.C. Chow, J.G. Watson, P.K. Hopke, S.-M. Yi, and G.C. England. 2004. Measurement of ultrafine particle size distribution from coal-, oil-, and gas-fired stationary combustion sources. J. Air Waste Manage. Assoc. 54:1494–505. doi:10.1080/10473289.2004.10471010
  • Franco García, V. 2014. Evaluation and improvement of road vehicle pollutant emission factors based on instantaneous emissions data processing. http://www.tdx.cat/bitstream/handle/10803/146187/vfranco.pdf;jsessionid=53652F5DE37395585FE3951529DC97C7?sequence=1 (accessed January 2017).
  • Gauderman, W.J., E. Avol, F. Gilliland, H. Vora, D. Thomas, K. Berhane, R. McConnell, N. Kuenzli, F. Lurmann, and E. Rappaport. 2004. The effect of air pollution on lung development from 10 to 18 years of age. N. Engl. J. Med. 351:1057–67. doi:10.1056/NEJMoa040610
  • Giechaskiel, B., R. Chirico, P.F. Decarlo, M. Clairotte, T. Adam, G. Martini, M.F. Heringa, R. Richter, A.S. Prevot, U. Baltensperger, and C. Astorga. 2010. Evaluation of the Particle Measurement Programme (PMP) protocol to remove the vehicles’ exhaust aerosol volatile phase. Sci. Total Environ. 408:5106–16. doi:10.1016/j.scitotenv.2010.07.010
  • Giechaskiel, B., A. Mamakos, J. Andersson, P. Dilara, G. Martini, W. Schindler, and A. Bergmann. 2012. Measurement of automotive nonvolatile particle number emissions within the European legislative framework: A review. Aerosol Sci. Technol. 46:719–49. doi:10.1080/02786826.2012.661103
  • Giechaskiel, B., M. Maricq, L. Ntziachristos, C. Dardiotis, X. Wang, H. Axmann, A. Bergmann, and W. Schindle. 2014. Review of motor vehicle particulate emissions sampling and measurement: From smoke and filter mass to particle number. J. Aerosol Sci. 67:48–86. doi:10.1016/j.jaerosci.2013.09.003
  • Hill, J., S. Polasky, E. Nelson, D. Tilman, H. Huo, L. Ludwig, J. Neumann, H. Zheng, and D. Bonta. 2009. Climate change and health costs of air emissions from biofuels and gasoline. Proc. Natl. Acad. Sci. USA 106:2077–82. doi:10.1073/pnas.0812835106
  • Hinds, W.C. 1999. Aerosol Technology: Properties, Behavior, and Measurement of Airborne Particles, 2nd ed. Hoboken, NJ: Wiley-Interscience.
  • Hu, S., S. Zhang, S. Sardar, S. Chen, I. Dzhema, S.-M. Huang, D. Quiros, H. Sun, C. Laroo, and L.J. Sanchez. 2014. Evaluation of gravimetric method to measure light-duty vehicle particulate matter emissions at levels below one milligram per mile (1 mg/mile). SAE Technical Paper. 2014-01-1571, 2014. doi:10.4271/2014-01-1571
  • Intergovernmental Panel on Climate Change. 2007. Climate change 2007: Synthesis report. Geneva, Switzerland: IPCC Secretariat. https://www.ipcc.ch/publications_and_data/publications_ipcc_fourth_assessment_report_synthesis_report.htm (accessed January 2017).
  • Kamboures. K.A., P.L. Rieger, S. Zhang, S.B. Sardar, M.-C.O. Chang, S.-M. Huang, I. Dzhema, M. Fuentes, M.T. Benjamin, A. Hebert, and A. Ayala. 2015. Evaluation of a method for measuring vehicular PM with a composite filter and a real-time BC instrument. Atmos. Environ. 123:63–71. doi:10.1016/j.atmosenv.2015.10.061
  • Kamboures, M.A., S. Hu, Y. Yu, J. Sandoval, P. Rieger, S.-M. Huang, S. Zhang, I. Dzhema, D. Huo, and A. Ayala. 2013. Black carbon emissions in gasoline vehicle exhaust: A measurement and instrument comparison. J. Air Waste Manage. Assoc. 63:886–901. doi:10.1080/10962247.2013.787130
  • Khalek, I.A. 2006. Diesel particulate measurement research. CRC Project E-66, Phase-2, Southwest Research Institute. https://crcao.org/reports/recentstudies2006/E-66-Phase%202-Final%20Report-IAK-R5.pdf (accessed February 8, 2017).
  • Khalek, I.A., T. Bougher, and J.J. Jetter. 2010. Particle emissions from a 2009 gasoline direct injection engine using different commercially available fuels. SAE Int. J. Fuels Lubr. 3(2):623–37. doi:10.4271/2010-01-2117
  • Kweon, C.-B., D.E. Foster, J.J. Schauer, and S. Okada. 2002. Detailed chemical composition and particle size assessment of diesel engine exhaust. SAE Technical Paper No. 2002-01-2670. doi:10.4271/2002-01-2670
  • Li, Y., X. Jian, K. Johnson, T. Durbin, M. Villela, L. Pham, E. Hosseini, D. Short, A. Asa-Awuku, G. Karavalakis, D. Quiros, S. Hu, T. Huai, A. Ayala, and H. Jung. 2014. Determination of suspended exhaust PM mass for light-duty vehicles. SAE Technical Paper No. 2014-01-1594. doi:10.4271/2014-01-1594
  • Lipsky, E., C.O. Stainer, S.N. Pandis, and A.L. Robinson. 2002. Effect of sampling conditions on size distribution of fine particulate matters emitted from a pilot-scale pulverized coal combustor. Engergy & Fuel 16:302–10.doi:10.1021/ef0102014
  • Liu, Z.G., V.N. Vasys, M.E. Dettmann, J.J. Schauer, D.B. Kittelson, and J. Swanson. 2009. Comparison of strategies for the measurement of mass emissions from diesel engines emitting ultra-low levels of particulate matter. Aerosol Sci. Technol. 43:1142–52.doi:10.1080/02786820903219035
  • Lloyd, A.C., and T.A. Cackette. 2001. Diesel engines: Environmental impact and control. J. Air Waste Manage. Assoc. 51:809–47.doi:10.1080/10473289.2001.10464315
  • Mamakos, A., C. Dardiotis, and G. Martini. 2012. Assessment of particle number limits for petrol vehicles. European Commission Joint Research Center, Brussels, Belgium.
  • Mamakos, A., G. Martini, and U. Manfredi. 2013. Assessment of the legislated particle number measurement procedure for a Euro 5 and a Euro 6 compliant diesel passenger cars under regulated and unregulated conditions. J. Aerosol Sci. 55:31–47. doi:10.1016/j.jaerosci.2012.07.012
  • Maricq, M.M., J. Szente, M. Loos, and R. Vogt. 2011. Motor vehicle PM emissions measurement at LEV III levels. SAE Int. J. Engines 4:597–609. doi:10.4271/2011-01-0623
  • Maricq, M.M., J.J. Szente, J. Adams, P. Tennison, and T. Rumpsa. 2013. Influence of mileage accumulation on the particle mass and number emissions of two gasoline direct injection vehicles. Environ. Sci. Technol. 47:11890–6. doi:10.1021/es402686z
  • Martini, G., B. Giechaskiel, and P. Dilara. 2009. Future European emission standards for vehicles: The importance of the UN-ECE Particle Measurement Programme. Biomarkers 14:29–33. doi:10.1080/13547500902965393
  • May, A.A., N.T. Nguyen, A.A. Presto, T.D. Gordon, E.M. Lipsky, M. Karve, A. Gutierrez, W.H. Robertson, M. Zhang, C. Brandow, O. Chang, S. Chen, P. Cicero-Fernandez, L. Dinkins, M. Fuentes, S.-M. Huang, R. Ling, J. Long, C. Maddox, J. Massetti, E. McCauley, A. Miguel, K. Na, R. Ong, Y. Pang, P. Rieger, T. Sax, T. Truong, T. Vo, S. Chattopadhyay, H. Maldonado, M.M. Maricq, and A.L. Robinson. 2014. Gas- and particle-phase primary emissions from in-use, on-road gasoline and diesel vehicles. Atmos. Environ. 88:247–60. doi:10.1016/j.atmosenv.2014.01.046
  • Menon, S., J. Hansen, L. Nazarenko, and Y. Luo. 2002. Climate effects of black carbon aerosols in China and India. Science 297:2250–3. doi:10.1126/science.1075159
  • Mordukhovich, I., E. Wilker, H. Suh, R. Wright, D. Sparrow, P.S. Vokonas, and J. Schwartz. 2009. Black carbon exposure, oxidative stress genes, and blood pressure in a repeated-measures study. Environ. Health Perspect. 117(11):1767–172. doi:10.1289/ehp.0900591
  • Nam, E., C. Fulper, J. Warila, J. Somers, H. Michaels, R. Baldauf, R. Rykowski, and C. Scarbro. 2008. Analysis of particulate matter emissions from light-duty gasoline vehicles in Kansas City. EPA420-R-08-010. https://cfpub.epa.gov/si/si_public_file_download.cfm?p_download_id=499633 (accessed January 26, 2017).
  • Pope, C.A. III, and D.W. Dockery. 2006. Health effects of fine particulate air pollution: Lines that connect. J. Air Waste Manage. Assoc. 56:709–42. doi:10.1080/10473289.2006.10464485
  • Quiros, D.C., S. Zhang, S. Sardar, M. Kamboures, D. Eiges, M. Zhang, H. Jung, M.J. McCarthy, M.-C.O. Chang, and A. Ayala. 2015. Measuring Particulate emissions of light duty passenger vehicles using integrated particle size distribution (IPSD). Environ. Sci. Technol. 49:5618–27. doi:10.1021/acs.est.5b00666
  • Ramanathan, V., and G. Carmichael. 2008. Global and regional climate changes due to black carbon. Nat. Geosci. 1:221–7. doi:10.1038/ngeo156
  • Rasdorf, W., C. Frey, P. Lewis, K. Kim, S.-H. Pang, and S. Abolhassani. 2010. Field procedures for real-world measurements of emissions from diesel construction vehicles. J. Infrastruct. Syst. 16:216–25. doi: 10.1061/(ASCE)IS.1943-555X.0000027
  • Sardar, S., L. Larsen, B. Frodin, W. McMahon, S.-M. Huang, and M.-C.O. Chang. 2016. Evaluation of PM Measurement precision and the quivalency of the single and three filter sampling methods for LEV III FTP standards. SAE Int. J. Engines 9(1):342–54. doi:10.4271/2015-01-9045.
  • Schauer, J., C. Christensen, D. Kittelson, J. Johnson, and W. Watts. 2008. Impact of ambient temperatures and driving conditions on the chemical composition of particulate matter emissions from non-smoking gasoline-powered motor vehicles. Aerosol Sci. Technol. 42:210–23. doi: 10.1080/02786820801958742
  • U.S. Environmental Protection Agency. 2002. Health assessment document for diesel engine exhaust. Washington, DC: U.S. EPA.
  • Xue, J., Y. Li, X. Wang, T.D. Durbin, K.C. Johnson, G. Karavalakis, A. Asa-Awuku, M. Villela, D. Quiros, S. Hu, T. Huai, A. Ayala, and H.S. Jung. 2015 Comparison of vehicle exhaust particle size distributions measured by SMPS and EEPS during steady-state conditions. Aerosol Sci. Technol. 49 (10):984–96. doi:10.1080/02786826.2015.1088146
  • Zhang, S., and W. McMahon. 2012. Particulate emissions for LEV II light-duty gasoline direct injection vehicles. SAE Int. J. Fuels Lubr. 5:637–46. doi:10.4271/2012-01-0442
  • Zheng, Z., K.C. Johnson, Z. Liu, T.D. Durbin, S. Hu, T. Huai, D.B. Kittelson, and H.S. Jung. 2011. Investigation of solid particle number measurement: Existence and nature of sub-23nm particles under PMP methodology. J. Aerosol Sci. 42:883–97. doi:10.1016/j.jaerosci.2011.08.003
  • Zielinska, B., J. Sagebiel, J.D. McDonald, K. Whitney, and D.R. Lawson. 2004. Emission rates and comparative chemical composition from selected in-use diesel and gasoline-fueled vehicles. J. Air Waste Manage. Assoc. 54:1138–50. doi:10.1080/10473289.2004.10470973

Reprints and Corporate Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

To request a reprint or corporate permissions for this article, please click on the relevant link below:

Academic Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

Obtain permissions instantly via Rightslink by clicking on the button below:

If you are unable to obtain permissions via Rightslink, please complete and submit this Permissions form. For more information, please visit our Permissions help page.