Qingyao Hu1, Cheng Huang 1, Liping Qiao1, Yingge Ma1, Qiang Yang2, Wei Tang2, Min Zhou1, Shuhui Zhu1, Shengrong Lou1, Shikang Tao1, Yun Chen3, Li Li 1

State Environmental Protection Key Laboratory of Cause and Prevention of Urban Air Pollution Complex, Shanghai Academy of Environmental Sciences, Shanghai 200233, China
Hangzhou Institute of Environment Sciences, Hangzhou 310014, China
Hangzhou Motor Vehicle Exhaust Pollution Management Office, Hangzhou 310014, China


Received: September 1, 2018
Revised: March 11, 2019
Accepted: May 20, 2019

Download Citation: ||https://doi.org/10.4209/aaqr.2018.07.0272  

  • Download: PDF


Cite this article:


Hu, Q., Huang, C., Qiao, L., Ma, Y., Yang, Q., Tang, W., Zhou, M., Zhu, S., Lou, S., Tao, S., Chen, Y. and Li, L. (2019). Speciated PM Composition and Gas and Particle Emission Factors for Diesel Construction Machinery in China. Aerosol Air Qual. Res. 19: 1820-1833. https://doi.org/10.4209/aaqr.2018.07.0272


HIGHLIGHTS

  • RDE test were performed on 9 diesel machineries under real-world conditions.
  • Gaseous and particulate emission factors of the tested machineries were determined.
  • The OC, EC, and WSIs, elements, OM of PM from the tested machineries were analyzed.
 

ABSTRACT


On-board emission measurements were performed on nine diesel construction machines operated under real-world conditions in China. The emission factors (EFs) for NOx, CO, the total hydrocarbons, PM, and the particle number concentration were determined under actual operating conditions, i.e., during idling, moving, and working. To investigate the chemical composition of PM from diesel machinery, organic carbon (OC), containing particulate organic matter (POM); elemental carbon (EC); water-soluble ions (WSIs); and elements were also analyzed herein. OC was the most abundant component (contributing 39.3 ± 11% of the mass), followed by EC (37.7 ± 13%). POM species, including n-alkanes, hopanes, polycyclic aromatic hydrocarbons (PAHs), and n-fatty acids, contributed approximately 2.4–6.4% of the total PM mass. Compounds with 3 or 4 aromatic rings, including pyrene, phenanthrene, and fluoranthene, dominated the particulate PAHs, accounting for 62% of the total mass. These results are consistent with those of previous studies and can help to minimize uncertainties in emission inventories and source apportionment for non-road machinery.


Keywords: Non-road machinery; Emission factors; PM components; PAH emissions; Diesel emissions.


INTRODUCTION


Emissions from non-road machinery have attracted increasing attention due to their increasing contributions to emission inventories over recent decades (McDonald et al., 2015; Campbell et al., 2018a, b). Nitrogen oxides (NOx) and particulate matter (PM) are among the major pollutants in diesel exhaust. NOx is a key precursor to ozone and nitrate in PM2.5 in the atmospheric boundary layer (Seinfeld and Pandis, 2006), while PM contains abundant organic carbon (OC), elemental carbon (EC), and intermediate volatility organic compounds (iVOCs), inorganic components that greatly impact visibility (Pitchford et al., 2007; Li et al., 2017; Chow et al., 2018), human health (IARC, 2012; Mohankumara and Senthilkumar, 2017), and global warming (Ayhan, 2009; Smith et al., 2009).

The amount of non-road machinery has increased considerably over the last decade in China. The number of non-road construction machines in China reached 7.2 million by the end of 2017, a 67% increase over the year 2012 (MEP, 2018). However, Chinese emission standards for non-road diesel machinery currently lag two stages behind those for on-road diesel vehicles; in 2015, Stage 3 emission standards were implemented for non-road diesel machinery, while on-road diesel vehicles were subject to Stage 5 standards. This delay in emission standard implementation has stimulated higher emissions from non-road diesel machines in China. In urban areas, non-road machinery consists primarily of construction machinery; construction machinery in turn is dominated by excavators, loaders, and forklifts, which account for 26.8%, 27.3%, and 41.8%, respectively, of the total number of construction machines (MEP, 2018).

To evaluate emissions from non-road machinery, Wang et al. (2016) compiled a non-road machinery emission inventory for 2012 in China. However, this emission inventory contains relatively large uncertainties due to limited data on real-world emission factors (EFs) for non-road machines (Wang et al., 2016). Thus, additional measurements of non-road machinery emissions have been conducted in the past decade. Abolhasani et al. (2008) performed real-world measurements of emissions from diesel excavators using a Portable Emission Measurement System (PEMS) in 2008. Since then, this method has been applied to other types of construction equipment and to the evaluation of emission reductions due to fuel selection and new engine technologies (Frey et al., 2008; Frey et al., 2010; Pirjola et al., 2017; Zavala et al., 2017). Recent localized measurements of construction machinery emissions conducted in China have found that EFs are generally higher for Chinese machines than for American machines (Fu et al., 2012; Li et al., 2016). Particulate and gaseous species, including OC, EC, water-soluble ions (WSIs), and elements, have come under scrutiny (in addition to criteria pollutants, such as NOx) for their importance in source apportionment, human health, and climate change (Lindgren et al., 2011; Cui et al., 2017). However, the characteristics of particulate and gaseous emissions, and particularly organic compound emissions, from on-road diesel construction machinery remain largely unknown.

In this study, EFs were determined for nine diesel construction machines (viz., four excavators, three wheel loaders, one bulldozer, and one roller) via on-board measurements under real-world operating conditions. Measurements were performed during typical operating modes, including idling, moving, and working. The PM chemical composition was determined via OC, EC, WSI, elemental, and particulate organic matter (POM) analysis. The results can be used to quantify real-world emissions and develop emission inventories for construction machinery in China. 


EXPERIMENTAL METHODS



Test Machines

In this study, emission measurements were carried out on nine non-road machines, viz., four excavators (EX), three wheel loaders (WL), one bulldozer (BD), and one roller (RL). These test machines represent the majority of the construction machinery types in China, accounting for 26.9%, 27.3%, 1.1%, and 2.1%, respectively, of the total fleet (MEP, 2018). These machines featured rated power outputs between 83 and 162 kW, which is representative of the power distribution of construction machinery in the Yangtze River Delta region in China (Lu et al., 2017), and model years between 2004 and 2016. The machine technical specifications, such as emission standards, engine displacement, and rated power, are presented in Table 1. The test machines were rent from several construction companies, and the emission measurements were performed at a large construction site in Hangzhou, China during December 2016. In this study, four excavators (EX), three wheel loaders (WL), and one roller (RL) were fueled with low-sulfur diesel (Fuel Sample 1) which were uniformly supplied by the same gas station at the same time. The bulldozer (BD) was fueled with low sulfur diesel (Fuel Sample 2), which was supplied by the fuel truck. The test fuel properties are presented in Table 2.


Table 1. Test construction machine details and operating modes. 


Table 2. Test fuel properties


Emission Measurement System

An on-board emission measurement system was constructed for this study. The system consists of a portable emission measurement system (PEMS), particulate measurement, and collection system. The PEMS system (SEMTECH-ECOSTAR; Sensors, Inc., USA), which was used to measure gaseous species, featured a non-dispersive infrared (NDIR) analyzer for CO and CO2 measurements, a chemiluminescence detector (CLD) for NO and NO2 measurements, and a flame ionization detector (FID) for total hydrocarbon (THC) measurements. The system was zeroed with pure nitrogen and calibrated with standard gases before each test. The dilution system (FPS 4000; Dekati, Finland) was connected to an Engine Exhaust Particle Sizer spectrometer (EEPS 3090; TSI Inc., USA) and a four-channel PM sampling system; the EEPS 3090 was used to simultaneously measure the instantaneous PM and particle number (PN) concentrations from 5.6 to 560 nm with 32 particle-size channels. During the measurements, the tailpipe exhaust was vented into a mass flow measurement device (SEMTECH EFM; Sensors, Inc., USA) to measure the exhaust flow rate using pitot tube technology. Two exhaust samples were extracted from the mass flow measurement device. One was sent into the PEMS system, and the other was sent into the dilution system. Both sampling lines were heated to 190°C before entering the instruments. The dilution was performed using HEPA-filtered compressed air in two stages. The primary dilution was carried out with a diffusion-type perforated tube. In this tube, dilution air was introduced through small pores along the transport line to minimize losses inside the probe. The second dilution stage involved an ejector type diluter, which pulled the primary diluter sample flow. It is necessary to note that the dilution system used in this study is based on a constant dilution method. The advantage of this method is its output volume is big enough to reach 100 L min–1, which can meet the requirements for synchronous sampling of at least four channels of filters (10 L min–1 for each channel) and other on-line measurement instruments. Data of one set or two parallel sets of PM concentration and chemical compositions, including OC, EC, WSIs, elements, and organic matter (OM), can be obtained for one measurement. A similar method has been applied to the emission measurement of non-road machinery by Cui et al. (2017). It should be noted that proportional dilution system is more suitable to on-board emission measurement of mobile source in principle. In contrast to the proportional dilution system, which would control the dilution rate by proportionally changing the exhaust flow rate, the constant dilution system will lead to uneven sampling ratio when exhaust flow has large fluctuation. The application of constant dilution system will cause some uncertainty, underestimating the various PM compositions emission factors at high engine loads (usually with high exhaust flow) of the machines (Zheng, 2016). However, the existing proportional dilution system also have some fatal weaknesses: The flow rate of the sample gas after dilution is less than 20 L min–1, which is impossible to collect enough particles in a typical test cycle (1–2 hours) for speciated PM composition analysis; the sampling probe is likely to be clogged by soot, which would significantly influence the dilution ratio control precision (Zheng et al., 2017).

Construction machines usually possess intricate structures and limited space for installation of the measurement system. Therefore, the instruments were housed on a test platform. During the measurements, the platform was towed behind the test machine using a towing hook. Fig. 1 shows a schematic of the measurement system.


Fig. 1. Schematic of the measurement system.
Fig. 1. Schematic of the measurement system.


Test Procedures

Previous studies have indicated that emissions from construction machines are strongly related to the operating mode (Abolhasani et al., 2008; Frey et al., 2008; Fu et al., 2012; Cui et al., 2017; Zavala et al., 2017). To reflect the emission characteristics of machinery operated under real-world operating conditions, the measurements were performed during actual construction tasks. The excavator and wheel loader operation modes include idling, moving,

and working. The average operational time proportion of idling, moving, and working condition for the excavators accounted for 21 ± 4%, 9 ± 11%, and 70 ± 14%, respectively. The tree conditions for the wheel loaders accounted for 20 ± 12%, 16 ± 3%, and 64 ± 11%, respectively. The bulldozer and roller operation modes include idling and working operating modes; idling accounted for 43% and 34% for the two machines, respectively, while working accounted for 57% and 64%. The test machinery operating modes are presented in Table 1


Sampling and Analysis

PM was sampled using four separate filter trains. Two Teflon filters (47 mm, TE 38; Whatman, UK) and two quartz filters (47 mm, QM-A; Whatman, UK) were used during each measurement period; one filter was placed in each of the four channels. The flow rate through each channel was 12 L min–1. The dilution ratios were between 16:1 and 20:1. A measurement period of ~45 min was used to ensure sufficient particle loading on the filters. To determine PM mass, the Teflon filters were analyzed gravimetrically after equilibration for 24 h at 50 ± 5% relative humidity and 20 ± 1°C. After gravimetric analysis, one Teflon filter was extracted and analyzed for water-soluble ions using an ion chromatograph (940 Professional IC Vario; Metrohm, Switzerland); elemental content was analyzed on the other Teflon filter using an energy-dispersive X-ray fluorescence spectrometer (Epsilon 5 EDXRF; PANalytical, Finland). OC and EC were determined on one of the quartz filters using a thermal/optical carbon analyzer (Model 2001; Desert Research Institute) and the IMPROVE-A protocol (Chow et al., 2007). OM was estimated by multiplying the measured OC by a factor of 1.2 (Turpin and Lim, 2001). The other quartz filter was analyzed for solvent-extractable organic compounds (SEOC), including n-alkanes, hopanes, particulate polycyclic aromatic hydrocarbons (PAHs), and n-fatty acids, using gas chromatographymass spectrometry (GC-MS; Agilent, USA); the analytical procedure has been described previously (Feng et al., 2007). Briefly, one fourth of the quartz filter was spiked with a surrogate mixture consisting of tetracosane-d50, chrysene-d12, anthracene-d10, perylene-d12, and heptadecanoic-d33 acid and then ultrasonically extracted in three 60 mL dichloromethane/methanol (2:1, V/V) aliquots at room temperature. The combined extract was filtered and reacted with freshly prepared diazomethane, which would not react with PAH and alkanes, to promote free organic acid esterification. The total extract was then analyzed via GC-MS (6890/5975 GC/MSD; Agilent, USA). Hexamethyl benzene was added prior to GC-MS analysis as an n-alkanol internal standard and to verify the recovery of alkanes, PAHs, and fatty acids. As shown in Table 4, 21 n-alkanes, 10 hopanes, 21 particulate PAHs (3–5-ringed), and 21 n-fatty acids were analyzed quantitatively by GC-MS; alkanes, PAHs, and fatty acids were quantified using deuterated internal standards with chemical characteristics and retention times similar to those of the target analytes. The MS analysis was performed at 70 eV in electron impact mode over a mass range of 45–550 m/z. The GC, which features an HP-5MS capillary column (30 m × 0.25 mm × 0.25 µm), was operated with helium carrier gas at a flow rate of 1.0 mL min–1. The temperature procedures used for non-polar and polar organic matter are same. The procedures are as follows: The oven temperature was set initially at 40°C and continued with 4 minutes of isothermal holding time, then increased to 120°C at 10°C min–1 and held for 2 minutes, and finally increased to 300°C at 20°C min–1 and held for 20 minutes. Analytes were quantified via linear regressions from five-point calibration curves between authentic-standard-to-internal-standard concentrations ratios and the corresponding peak area ratios.

The external system audits and inter-laboratory comparisons of analytical equipment were conducted to keep the QA in high level. The QC procedure in the process of sample collection and analysis were followed: The flow control valves of the particle sampling system were calibrated by a mass flow meter (Mass Flow Meter 41403; TSI Inc., USA) before each test; the quartz filter was preheated to reduce the interference, and the blank samples and other quality control samples were also included in this study to minimize the artifact effect result of OC/EC, WSIs, and OM. Full-scale calibration of the flame ionization detector and verification of trace O2 concentrations (< 100 ppmv) in the He analysis atmosphere were carried out before every OC/EC analysis (Chow et al., 2007). The X-ray energy was calibrated before every elemental content analysis; the calibration verification and background determination were conducted regularly. When OM was analyzed by GC-MS, a complete analysis of blank and duplicate samples was performed for each batch of samples. If only the analysis results of the blank and duplicate samples met the requirements, the formal sample analysis would be performed. The analysis results show that the recovery rates of the organic matter of the samples ranged from 93.2% to 98.3%.


Data Processing

Fuel-based gaseous, particulate matter, and particle number EFs (6.04–523 nm) were calculated using the carbon balance method, as shown in the following equations:

  

where EFgaseous i is the fuel-based emission factor (g kg1) of gaseous pollutant i, EFPM and EFPN are the fuel-based emission factors of PM (g kg1) and PN (# kg1), ∆mi is the measured mass concentration of species i in the exhaust gas (mg m3), ∆PMj and ∆PNj are the measured particulate matter concentration (g m3) and particle number concentration (# m3) of the particle size channel j, and xc is the mass fraction of carbon (%) in the diesel, which is set to be 86% in this study according to Table 2. CO2, CO, PMc, and THC are the background-corrected carbon concentrations of CO2, CO, carbonaceous PM, and THC (gC m3).

Fuel consumption (FC) rate were calculated by carbonaceous emissions concentration and exhaust gas flow rate, as shown in the following equation:

 

where FC is the fuel consumption rate (kg h–1), mex is the exhaust gas flow rate (kg h–1), and ρex is the density of the exhaust gas (kg m3). 


RESULTS AND DISCUSSION



Gaseous and Particulate Matter EFs

The fuel-based EFs for NOx, CO, THC, PM, and PN from each type of machinery are summarized along with EFs from the literature in Table 3. The NOx, CO, and THC EFs measured herein for the excavator agree well with results from excavators with the same emission standards (Stages 0–2 or Tiers 0–2) in previous studies (Abolhasani et al., 2008; Frey et al., 2010; Fu et al., 2012). The wheel loader NOx, CO, and THC EFs herein are similar to those measured by another study in China (Fu et al., 2012). The NOx EFs for the bulldozer and roller were higher than those in previous studies (Frey et al., 2010; Li et al., 2016). For various types of machines, NOx EFs showed consistent decreases under enhanced emission standards, as shown in studies by Cao et al. (2016) and Zavala et al. (2017) in the U.S. and Mexico. The PM EFs were relatively varied. The average PM EFs from the excavators, wheel loaders, bulldozer, and roller were within the range of measurements from previous studies (Abolhasani et al., 2008; Frey et al., 2010; Fu et al., 2012; Cao et al., 2017). PN EFs for the test machines, with the exception of the bulldozer, were above 1015 # kg1, similar to results from heavy-duty diesel vehicles without after-treatment devices (Hallquist et al., 2013; Huang et al., 2013; Ježek et al., 2015; Pirjola et al., 2016).


Table 3. Gaseous and particulate emission factors from this and previous studies.

Fig. 2 shows fuel consumption (FC) rates and fuel-based EFs for each type of machine under different operating modes. Test machine FC rates were higher under moving and working conditions than while idling. Under working conditions in particular, the FC rates for the excavators, wheel loaders, bulldozer, and roller were 4.3, 2.1, 1.7, and 5.0 times higher than those while idling. However, the NOx, CO, and THC EFs for the excavators, wheel loaders, and bulldozer were generally higher while idling than under moving and working conditions, which is consistent with previous studies (Abolhasani et al., 2008; Fu et al., 2012; Li et al., 2016; Zavala et al., 2017). Unlike the machinery above, the roller had higher EFs (for all species except THC) under working conditions than while idling. The comparison of fuel-based NOx, CO, and THC EFs under different operation modes are associated with pollutant concentrations and FC. As we know, NOx concentration would increase with the operation load rate rise, but when the increase rate of NOx concentration is less than the FC increase rate, the fuel-based NOx EF will decrease; otherwise, the NOx EF will rise (Moussa et al., 2016). PM EFs did not change significantly with the operating condition. For the excavators and loaders, there were no consistent changes in PM and PN emissions under varying operating conditions. For example, under idling conditions, the machines usually had lower PM EFs and higher PN EFs, indicating smaller particle sizes at that time; particles were larger under moving and working conditions.


Fig. 2. (a) Average fuel consumption rates. Fuel-based EFs of (b) NO (dark shading), NO2 (light shading), and NOx; (c) CO; (d) THC; (e) PM; and (f) PN for each machine type under different operating modes. Error bars indicate standard deviation. Fig. 2. (a) Average fuel consumption rates. Fuel-based EFs of (b) NO (dark shading), NO2 (light shading), and NOx; (c) CO; (d) THC; (e) PM; and (f) PN for each machine type under different operating modes. Error bars indicate standard deviation.


Speciated PM Emissions

The speciated PM was collected and analyzed in all of the modes, and the analyzed results are representing the whole operation process. The fractional mass contributions of main category of various PM chemical components are shown in Fig. 3(a); the detailed fractions of speciated PM are summarized in Table 4. The PM filter for EX_2 was contaminated during the sampling process, so PM data is not available for this machine. OM and EC were the main PM components in the diesel construction machinery exhaust. OM accounted for 41 ± 17%, 52.3 ± 15%, 50.9%, and 46.6% of the total PM mass for the excavator, loader, bulldozer, and roller, respectively, while EC contributed 47.2 ± 8%, 34.5 ± 15%, 19.6%, and 37.1%.


Fig. 3. (a) Fractional contributions of various species to total PM mass and (b) fractional contributions of organic matter species to total organic mass.Fig. 3. (a) Fractional contributions of various species to total PM mass and (b) fractional contributions of organic matter species to total organic mass.


Table 4. Fractional chemical and organic matter of PM composition for each test machine (%).
Table 4. Fractional chemical and organic matter of PM composition for each test machine (%).

The contributions of WSIs to PM mass were relatively low, with an average of 1.9 ± 1%. WSIs from the test machines consisted primarily of NO3, which accounted for 1.3 ± 1% of the total PM mass, followed by SO42–. These results are consistent with Ma et al. (2018). However, SO42– was the dominant contributor to PM mass, far outweighing NO3, in diesel vehicle and machinery exhaust in previous studies in China (Zhang et al., 2015; Cui et al., 2017); these differences may be attributed to the diesel fuel sulfur content in these studies, which reached up to 500 ppm and 1300 ppm, respectively. The quality of diesel fuel has improved in recent years; diesel sulfur content is currently controlled to < 50 ppm, which generally reduces the fraction of SO42– in PM emissions in diesel exhaust.

The elemental component fractions varied between the different types of machinery. Elements constituted 4.9 ± 2%, 2.6 ± 0.9%, and 1.8% of the excavator, wheel loader, and roller emissions, respectively; Si was the dominant elemental component, accounting for 3.5 ± 1.9%, 1.2 ± 1.1%, and 1.1% of the total PM mass. The elemental fraction was much higher for the bulldozer than for the other machine types, reaching 18.5% of the total PM mass. The major elements emitted from the bulldozer were Si, Ca, Al, and Fe, which accounted for 5.1%, 5.1%, 2.4%, and 1.8%, respectively, of the total PM mass. These elements are generally derived primarily from diesel fuel impurities (Wang et al., 2003). Thus, the inferior quality of the diesel fuel used in the test bulldozer may explain the high elemental emissions. The PM SO42– fraction emitted from the test bulldozer, which measured up to 1.2%, was much higher than the SO42– fractions emitted from the other machinery types (0.1–0.3%).

The fractional contributions of various OM species to the total PM organic mass are shown in Fig. 3(b). Altogether, 21 n-alkanes, 10 hopanes, 21 particulate PAHs, and 21 n-fatty acids were quantified in the PM samples (see Table 4). OM species contributed approximately 2.4–6.4% of the total PM mass. n-Alkanes were the dominant organic component, accounting for 2.26 ± 0.2%, 4.48 ± 2%, 1.87%, and 3.71% of the total PM mass from the excavators, wheel loaders, bulldozer, and roller, respectively; n-fatty acids were the second most abundant category, contributing 0.65 ± 0.4%, 0.42 ± 0.4%, 1.47%, and 0.71% of the total PM mass. The average PM mass fractions of hopanes and particulate PAHs were 0.049 ± 0.03% and 0.078 ± 0.02%, respectively. Cui et al. (2017) reported n-alkane and particulate PAH fractions reaching 5.14% and 0.098%, slightly higher than the proportions found herein; this may be attributed to the use of high-sulfur fuel in that study. The n-alkanes found herein consisted primarily of C18–C23 compounds, except for those emitted from the bulldozer, which contained more high-carbon components (C20–C24), as shown in Table 4. Hopanes can be used as organic tracers for oil combustion processes. 17a(H)21ß(H)-hopane (C30H) and 17a(H)21ß(H)-30-norhopane (C29H) were the most abundant hopanes, contributing 0.016 ± 0.01% and 0.009 ± 0.004% of the total PM mass, respectively, which is similar to the hopane composition of PM emissions from diesel vehicles and ships (Schauer et al., 1999; Sippula et al., 2014; Cui et al., 2017).


OC and EC Emissions

On average, the OC and EC measured in the test machine emissions herein contributed 39.3 ± 11% and 37.7 ± 13% of the total PM mass, as shown in Fig. 4. The average OC/EC ratio was 1.24 ± 0.7, which is similar to the average excavator OC/EC ratio (1.18) measured by Cui et al. (2017), lower than the ratios measured for diesel fork lifts (FL; 2.71) and generators (GE; 2.73) by Chow et al. (2011), and higher than the ratios measured in diesel truck exhaust in China (0.43–0.57; Zhang et al., 2015; Wu et al., 2016). Previous studies have indicated that engine load affects OC and EC emissions considerably (Liu et al., 2005; Wu et al., 2016; Cui et al., 2017); OC/EC ratios are much higher under low engine load, such as during idling, than under high engine load, as OC is generated predominantly at low temperatures in fuel-rich zones. The diesel trucks evaluated by Zhang et al. (2015) and Wu et al. (2016) were operated at high speed (high load), resulting in relatively low OC/EC ratios compared with construction machinery in this and previous studies (Chow et al., 2011; Cui et al., 2017).


Fig. 4. OC and EC mass fractions and OC/EC ratios for emissions from different types of machinery in this study and in previous studies on construction machinery and diesel trucks in China.Fig. 4. OC and EC mass fractions and OC/EC ratios for emissions from different types of machinery in this study and in previous studies on construction machinery and diesel trucks in China. 


PAH Emissions

The fuel-based EFs of the speciated PAHs were classified by the number of benzene rings, as shown in Fig. 5(a). The average fuel-based EF for the test machinery was 382 ± 264 g kg-fuel–1. Compounds with 3 or 4 benzene rings were dominant, accounting for 34.9% and 52.3%, respectively, of the total particulate PAHs. Pyrene, phenanthrene, and fluoranthene were the most abundant particulate PAH compounds emitted from the test machines, constituting 37.1%, 12.6%, and 12.2%, respectively, of the total PAHs. 


Fig. 5. Fuel-based EFs for speciated PAHs classified by (a) the number of benzene rings and (b) the speciated components, as compared with previous studies in China.Fig. 5. Fuel-based EFs for speciated PAHs classified by (a) the number of benzene rings and (b) the speciated components, as compared with previous studies in China.

Other important compounds included benzo[b+k]fluoranthene (BbkF), benzo[ghi]perylene (BghiP), chrysene (Chr), benzo[e]pyrene (BeP), and benzo[ghi]fluoranthene (BghiF). The distributions of individual particulate PAH species, as well as compounds with given numbers of rings, were consistent with previous studies on diesel excavators (Cui et al., 2017) and trucks (Cui et al., 2017; Zheng et al., 2017) in China, as shown in Fig. 5.

Molecular diagnostic ratios (MDRs), which consist of ratios of two individual PAHs, have been used as organic markers for various anthropogenic sources (Katsoyiannis et al., 2011) and can also be used in primary organic aerosol source apportionment (Shrivastava et al., 2007). MDRs consisting of common PAHs, e.g., fluoranthene/(pyrene + fluoranthene) [Flu/(Pyr + Flu)], anthracene/(phenanthrene + anthracene) [Ant/(Phe + Ant)], benzo[a]anthracene/(chrysene + benzo[a]anthracene) [BaA/(Chr + BaA)], and Benzo[a]pyrene/Benzo[ghi]perylene [BaP/BghiP], have been used to estimate source contributions from diesel exhaust, as shown in Table 5. The overall Flu/(Pyr + Flu) ratio in this study was 0.25 ± 0.07, similar to values reported by Shah et al. (2005) and Pakbin et al. (2009), but lower than those reported by Schauer et al. (1999), Cui et al. (2017), and Zheng et al. (2017). Katsoyiannis et al. (2011) suggested that Flu/(Pyr + Flu) MDRs between 0.40 and 0.50 represent pyrogenic sources (e.g., fuel combustion). Yunker et al. (2002) and Ravindra et al. (2008) found that Flu/(Pyr + Flu) MDRs of 0.39 ± 0.1 and > 0.5, respectively, could be used to distinguish diesel combustion sources. Our measurements indicate much lower thresholds than those recommended by previous studies, which may introduce substantial uncertainties in the use of these ratios to infer diesel combustion sources. The average Ant/(Phe + Ant) and BaA/(Chr + BaA) ratios in this study were 0.12 ± 0.02 and 0.36 ± 0.02, which are both similar to the recommended MDR lower limits (0.1 and 0.35) used to identify fuel combustion sources in Katsoyiannis et al. (2011). The MDR results herein are similar to measurements in Zheng et al. (2017) but lower than those reported by Pakbin et al. (2009), Liu et al. (2015), and Cui et al. (2017). The overall BaP/BghiP ratio in this study was 0.63 ± 0.5 on average, comparable to the MDR (> 0.6) suggested for traffic emission sources by Katsoyiannis et al. (2011) and the range (0.5–0.6) suggested by Ravindra et al. (2008). The organic markers emitted from diesel machinery and vehicles varied widely with vehicle type, fuel quality, operating mode, and control technology, as shown in Table 5. In addition, the markers recommended in previous studies cannot be used to distinguish on-road emissions from non-road diesel exhaust. Therefore, we suggest that organic markers be applied carefully during diesel emissions source apportionment until additional real-world PAH emissions data can be obtained for on- and non-road diesel vehicles.


Table 5. Molecular diagnostic ratios (MDRs) for particulate PAHs from diesel construction machinery and trucks in this and previous studies.


CONCLUSIONS


This study presents the detailed EFs and PM chemical component measurements from an on-board emission measurement system for nine diesel construction machines in China. The NOx, CO, THC, PM, and PN EFs and the fractional contributions of PM components (including OC, EC, WSIs, elements, and OM) were determined for each test machine under real-world conditions, i.e., during idling, moving, and working. The particulate and gaseous EFs were generally consistent with those found in previous studies. Although the fuel-based EFs were slightly lower while the machines were moving and working than while they were idling, the FC rates were much higher—with prominent overall emissions—during the first two conditions. Due to the lag in emission standards, the diesel construction machines displayed high EFs compared with newer machinery (manufactured according to more stringent standards). OC was the most abundant particulate component, contributing 39.3 ± 11% of the total PM mass, followed by EC, which accounted for 37.7 ± 13% on average. The WSIs and elements formed 1.9 ± 1% and 5.3 ± 6% on average of the emissions, respectively, and NO3 was the most abundant WSI, whereas Si dominated the elemental content. Higher-quality fuel decreased the emission of WSIs and elements, with SO42–, Si, Ca, Al, and Fe being emitted in higher quantities from machinery powered by inferior diesel fuel. Organic matter, including n-alkanes, hopanes, particulate PAHs, and n-fatty acids, contributed approximately 2.4–6.4% of the total PM mass. Compounds with 3 or 4 aromatic rings, including pyrene, phenanthrene, and fluoranthene, were dominant, accounting for 62% of the total particulate PAHs. Other major compounds included benzo[b+k]fluoranthene, benzo[ghi]perylene, chrysene, benzo[e]pyrene, and benzo[ghi]fluoranthene. The emissions from the test machines exhibited overall ratios of 0.25 ± 0.07, 0.12 ± 0.02, 0.36 ± 0.02, and 0.63 ± 0.5 for Flu/(Pyr + Flu), Ant/(Phe + Ant), BaA/(Chr + BaA), and BaP/BghiP, respectively. Although the ratio of Flu/(Pyr + Flu) was not consistent with the MDRs determined in previous studies, the other ratios were comparable. More measurements are necessary in order to identify organic markers of non-road diesel machinery.


ACKNOWLEDGEMENT


This work was supported by the National Key R&D Program of China (Grant No. 2016YFC0201501), the National Natural Science Foundation of China (Grant No. 21777101), the Science and Technology Commission of Shanghai Municipality Fund Project (Grant No. 16dz1206704).



Don't forget to share this article 

 

Subscribe to our Newsletter 

Aerosol and Air Quality Research has published over 2,000 peer-reviewed articles. Enter your email address to receive latest updates and research articles to your inbox every second week.

Latest coronavirus research from Aerosol and Air Quality Research

2018 Impact Factor: 2.735

5-Year Impact Factor: 2.827


SCImago Journal & Country Rank

Aerosol and Air Quality Research (AAQR) is an independently-run non-profit journal, promotes submissions of high-quality research, and strives to be one of the leading aerosol and air quality open-access journals in the world.