Particle effective density and mass during steady-state operation of GDI, PFI, and diesel passenger cars

Highlights

• We measured effective density of five passenger cars meeting LEV III standards.

• Gasoline PFI emissions have greater density and “fractal dimension” than GDI emissions.

• TSI EEPS and SMPS size distributions are compared quantitatively over steady-state cycles.

• Particles larger than 560 nm as measured by a TSI APS contribute substantially to total mass.

• Time-integrated suspended mass agrees well with gravimetric measures well below 1 mg/mi.

Abstract

Particle effective density is an important physical property of vehicle exhaust, and is required for estimating particulate matter (PM) mass emissions using the Integrated Particle Size Distribution (IPSD) method. In this study, we measure particle effective density of five light-duty vehicles with PM emissions below the Low Emission Vehicle (LEV) III PM standards of 1 or 3 mg/mi (0.62 and 1.86 mg/km) using the Differential Mobility Analyzer (DMA) – Centrifugal Particle Mass Analyzer (CPMA) approach. Test vehicles included two gasoline direct injection (GDI) vehicles without particulate filters, and for the first time reported in the literature, two port-fuel injected (PFI) vehicles and a turbocharged direct injection (TDI) light-duty diesel vehicle with a diesel particulate filter (DPF). The particle effective density functions generally resemble previous work on GDI and diesel engines without particulate filters but, for many size ranges, the PFI and TDI vehicles produced emissions with higher particle effective densities than GDI vehicles. Good linear correlation was found between the gravimetric and IPSD methods when applying the new particle effective density functions to size distribution measured by the TSI Engine Exhaust Particle Sizer (EEPS, 5.6–560 nm, R²=0.84); however, the IPSD method underestimated gravimetric mass by 64%. When using a TSI Aerodynamic Particle Sizer (APS, 0.54–2.5 µm) to measure the contribution of larger particles, underestimation bias was virtually eliminated and the correlation improved dramatically (R²=0.96). Even stronger correlation between IPSD and gravimetric methods was achieved when using a Scanning Mobility Particle Sizer (SMPS, 8.7–365 nm) and the APS (R²=0.97). A procedure for correcting EEPS measurements using the SMPS is presented and evaluated.

Introduction

Exposure to particulate matter (PM) is associated with increased cardiopulmonary morbidity and mortality (Pope & Dockery, 2006) and is influenced by mobile source emissions (EPA, 2002, Hill et al., 2009, Lloyd and Cackette, 2001). The California Air Resources Board (ARB) recently adopted the Low Emissions Vehicle (LEV) III regulations which, by 2025, will reduce the light-duty vehicle PM emission standards for the Federal Test Procedure (FTP) from 10 to 3 and ultimately to 1 mg/mi (6.2–0.62 mg/km) (CARB, 2012a). Recently, ARB and U.S. EPA demonstrated measurement of PM emissions below 1 mg/mi (0.62 mg/km) using the existing filter-based gravimetric method (Hu et al., 2014). However, alternative measurement approaches are still of great interest to better understand the characteristics of PM at very low levels.

One alternative method for PM mass measurement is the Integrated Particle Size Distribution (IPSD) method, a phrase first used by Liu et al. (2009) referring to the general method for estimating PM concentrations from particle size distribution and effective density. Defined as mass divided by electrical mobility equivalent volume (Kelly & McMurry, 1992), particle effective density enables rapid conversion between number and mass distributions, or between mobility and aerodynamic diameters, without assuming bulk density or morphology. Several studies have reported particle effective density of emissions from gasoline, natural gas, and diesel engines (e.g. Barone et al., 2011, Bullock and Olfert, 2014, Maricq and Xu, 2004, Olfert et al., 2007, Park et al., 2003, Ristimäki et al., 2002, Van Gulijk et al., 2004, Virtanen et al., 2004, Zelenyuk et al., 2005). Except for one study reporting constant effective density of emissions from a natural gas engine (Bullock & Olfert, 2014), the majority of studies on gasoline and diesel engines show particle effective density decreases as a function of particle size, in general agreement with the power fit model for fractal aerosols using a mass–mobility scaling exponent (Sorensen, 2011). However, to the best of our knowledge, only one study has evaluated gasoline direct injection (GDI) emissions (i.e. Maricq & Xu, 2004), and no previous studies have measured particle effective density from port-fuel injected (PFI) gasoline or light-duty diesel vehicles equipped with a diesel particulate filter (DPF). Therefore, redefining the particle effective density functions over the breadth of current vehicle technologies is needed to properly evaluate IPSD.

Kelly and McMurry (1992) first measured effective density of laboratory aerosol using a Differential Mobility Analyzer (DMA) and an inertial cascade impactor. Subsequently, several studies used a DMA placed upstream of an Electrical Low Pressure Impactor (ELPI, Ahlvik et al., 1998, Keskinen et al., 1992, Maricq et al., 2000, Maricq and Xu, 2004) providing real-time aerodynamic size distribution for a given DMA set point. A DMA has also been operated as part of a Scanning Mobility Particle Sizer (SMPS) in parallel to, rather than in series with, an ELPI for making density measurements more quickly by fitting size distributions (Ristimäki et al., 2002, Virtanen et al., 2004, Virtanen et al., 2002). A DMA has also been placed upstream of an Aerosol Particle Mass Analyzer (APM, Ehara, Hagwood, & Coakley, 1996) and Centrifugal Particle Mass Analyzer (CPMA, Olfert & Collings, 2005), which both classify according to mass to charge ratio using two rotating concentric cylinders to balance electrostatic and centrifugal forces. McMurry, Wang, Park, and Ehara (2002) used a DMA and APM to measure the effective density of atmospheric aerosol, and the approach has been subsequently applied to characterize engine exhaust particles (e.g. Barone et al., 2011, Park et al., 2003). Later, the CPMA was designed to improve the transfer function of the APM by using slightly different angular velocities for the two rotating cylinders (Olfert and Collings, 2005, Olfert et al., 2006). There are other methods for measuring particle effective density, such as using a DMA and Single Particle Laser Ablation Time-of-flight Mass Spectrometer (SPLAT, Zelenyuk et al., 2005). However, the system measures vacuum aerodynamic diameter and has low (<0.1%) detection efficiencies for particles below 50 nm (Zelenyuk, Yang, Choi, & Imre, 2009) and is not commercially available. The Dekati Mass Monitor (DMM, Lehmann, Niemelä, & Mohr, 2004) uses yet another approach by combining one mobility with six aerodynamic channels measuring size distribution to estimate particle effective density in real time. The DMM is used widely to measure PM mass, but density values are not reported but are used to directly report mass concentration based on a unimodal size distribution, mass median diameter, and geometric standard deviation (GSD), which have all shown to largely deviate from accepted reference methods (Mamakos et al., 2006, Quiros et al., 2014).

In this study, we use the DMA–CPMA method to measure particle effective density from two gasoline direct injected (GDI-1 and GDI-2) and two port fuel injected (PFI and PFI-E85) gasoline vehicles, and one turbo direct injection (TDI) light-duty diesel vehicles on a chassis dynamometer. Because the DMA–CPMA approach requires several minutes to complete each measurement, steady-state testing was conducted. The primary objective of this study was to determine particle effective density functions that can be used to evaluate the capability of IPSD to estimate PM mass emitted from light-duty vehicles meeting the LEV III standards. A secondary objective is to compare size distributions measured by the TSI Engine Exhaust Particle Sizer (EEPS, 5.6–560 nm) and SMPS (8.7–365 nm). The SMPS is regarded as the reference method for measuring size distribution; however, it requires one to two minutes to complete a scan, and therefore the EEPS was developed to measure transient particle size distributions. Therefore, SMPS-to-EEPS ratio is calculated under controlled steady-state conditions, and is used to correct EEPS measurements. Finally, this study also aims to measure the contribution of larger particle sizes using a TSI Aerodynamic Particle Sizer (APS, 0.54–2.5 µm), in order to compare total suspended real-time mass with the filter-based gravimetric standard method.