A fuel-based approach for emission factor development for highway paving construction equipment in China

ABSTRACT

The objective of this paper is to develop and demonstrate a fuel-based approach for emissions factor estimation for highway paving construction equipment in China for better accuracy. A highway construction site in Chengdu was selected for this study with NO emissions being characterized and demonstrated. Four commonly used paving equipment, i.e., three rollers and one paver were selected in this study. A portable emission measurement system (PEMS) was developed and used for emission measurements of selected equipment during real–world highway construction duties. Three duty modes were defined to characterize the NO emissions, i.e., idling, moving, and working. In order to develop a representative emission factor for these highway construction equipment, composite emission factors were estimated using modal emission rates and the corresponding modal durations in the process of typical construction duties. Depending on duty mode and equipment type, NO emission rate ranged from 2.6–63.7mg/s and 6.0–55.6g/kg–fuel with the fuel consumption ranging from 0.31–4.52 g/s correspondingly. The NO composite emission factor was estimated to be 9–41mg/s with the single-drum roller being the highest and double-drum roller being the lowest and 6–30g/kg-fuel with the pneumatic tire roller being the highest while the double-drum roller being the lowest. For the paver, both time-based and fuel consumption-based NO composite emission rates are higher than all of the rollers with 56mg/s and 30g/kg-fuel, respectively. In terms of time–based quantity, the working mode contributes more than the other modes with idling being the least for both emissions and fuel consumption. In contrast, the fuel-based emission rate appears to have less variability in emissions. Thus, in order to estimate emission factors for emission inventory development, the fuel-based emission factor may be selected for better accuracy.

Implications: The fuel-based composite emissions factors will be less variable and more accurate than time-based emission factors. As a consequence, emissions inventory developed using this approach will be more accurate and practical.

Introduction

The population of construction equipment has dramatically increased as the Chinese economy boosted in the past several decades. The total population of industrial equipment, mainly consisting of construction equipment (approximately 65%), is approximately 6 million with an annual increase of 7–22% during the period 2004–2013 (China Industrial Engineering Machinery Almanac [CIEMA], 2014). Emissions from these construction equipment have become a significant source to air pollution in China. For example, NO_X emissions from construction equipment accounted for 13.5% of the total NO_X emissions in Pearl River Delta (Zhang et al., 2010).

In China, since the construction of the highway system started in 1984, the total length of highway/expressway/interstate reached more than 110,000 km in 2014 and will continue to increase. For example, it was expected that another total of 50,000 km of highway would be added to the highway system in 2015 (Ministry of Transportation of the People’s Republic of China [MOT-PRC], 2014). There are usually two types of equipment used for highway construction. One type is for base/subbase construction, such as excavators, bulldozer, and backhoes; the other is for paving construction, such as rollers and pavers. Since equipment used for highway construction is mainly diesel fueled, NOx and PM emissions from this equipment will play a key role in contributing to air pollution in China. As the ozone concentrations in many cities of China continue to increase, ozone has imposed more of a health threat than the PM_2.5 (UPI News, 2015; Xue et al., 2014). Furthermore, NOx emissions from the transportation sector will continue to increase if emission control measures are not efficiently taken and implemented (Zhao et al., 2013). Thus, as an important precursor to O₃ formation, an accurate estimate of NOx emissions from this construction equipment is essential for air quality management and improvement, especially as the foggy haze air pollution has swept across the entire country in recent years (Dong et al., 2016; Wang et al., 2015; Ye et al., 2016)

In general, emissions inventories for off-road equipment are developed using emission factors models such as NONROAD (EPA, 2005) and OFFROAD (California Air Resources Board [CARB], 2007), in which engine power-based emissions factors and corresponding engine activities are essential. However, Kean et al. (2000) found that the emission inventory developed using the engine power-based approach versus the one using the fuel-based approach could differ by a factor of 2.3 for NO_x. As reported in the literature (Abolhasani et al., 2008; Frey, Kim, et al., 2008), a fuel-based emission factor appears to be less variable than the time-based emission factor. Thus, for a more accurate estimate of an emission inventory from off-road equipment, the fuel-based approach may be used.

Furthermore, when using the engine-power based emission factors for inventory development, the engine activities such as load factor, engine power, and duration of engine usage will be needed (EPA, 2005; CARB, 2007). However, this information usually is not easy to access, especially the load factor, and will vary by equipment type, job duty, and other factors. In contrast, information regarding the overall fuel consumption is easier to obtain. This makes the fuel-based emission factors more favorable than the engine power-based ones in meeting the challenge of improving the accuracy of emission inventory development.

Compared to the developed countries such as the United States, emissions from mobile sources are a relatively new issue in China that has not been given adequate attention in China until the vehicle population boosted in recent years (Yao et al., 2011). There are not many studies of emissions from mobile sources (both on-road vehicles and off-road equipment) available in the literature (Fu et al., 2012). In contrast to on-road vehicles, real-world measurements of emissions from off-road equipment are even more limited. Real-world measurements of emissions from excavators and loaders by Fu et al. (2012) comprise one of the few studies that are available, where measurements were taken under quasi-duty settings (the equipment was not working on a real construction project but was instructed to mimic the duties). A total of 20 equipment items in compliance with the national emissions standards Stage I and II were tested. Since diesel engines in compliance with the national emissions standards Stage III and IV have phased in quickly, emission testing on the newer diesel engines is also desperately needed.

Given the lack of the real-world emissions measurement data, emissions factors used for inventory development in China are mainly from other countries or from the NONROAD and OFFROAD models, which might not necessarily represent the real-world situation (Zhang et al., 2010). Thus, there is a huge data gap in this regard, and a need to develop emissions measurements-related data sets for air quality management and improvement.

In summary, the key objectives of this study are to develop a fuel-based approach for emission factor estimation for highway construction equipment in China and to develop emissions measurements data sets for highway construction equipment. Several 2012 model year paving equipment items such as rollers and pavers are used as examples in this study. Rollers and pavers are two types of equipment most used in highway paving construction and thus are the focus of this study. A highway construction site, where rollers and pavers are heavily used, was selected to conduct real-world measurements of NO emissions for the selected equipment and to demonstrate the developed fuel-based approach.

Methodology

This section mainly describes (1) a portable emission measurement system (PEMS) that was developed and used for real-world measurements of emissions; (2) experimental design for real-world measurements; and (3) fuel-based approach for emission factor estimation. The details are given in the following.

Portable emissions measurement system

As shown in Figure 1, the key components of the PEMS developed for this study include a five-gas analyzer, a sensor array, a wireless data logger, sampling accessories, a computer, and a power supply unit. The five-gas analyzer (Senwit, SV-5Q) uses nondispersive infrared (NDIR) to measure hydrocarbons (HC), CO, and CO₂, and electrochemical cells to measure NO and O₂; the sensor array consists of a revolutions-per-minute (RPM) sensor (Yuanben, MEVS-03) to measure the engine speed, a manifold absolute pressure (MAP) sensor (BYP-3000) to measure the manifold air pressure, and a thermocouple (BYP-T210) to measure the intake air temperature (IAT); the wireless data logger was used to collect data from the gas analyzer and sensor array, and to transmit the data wirelessly to the computer for further analysis. Because of the limited space that could be used for sitting the PEMS for off-road equipment, a wireless data receiver exhibits advantages over a wired system, in which the movement of the equipment is always limited. The sampling accessories include a sampling probe, sampling tubing, sensors, probe, and others; the computer connects to a data receiver to collect measurement data and perform further data analysis; the power supply unit uses a mini generator to power the gas analyzer, sensor array, and data logger.

Figure 1. The schematics of a portable emission measurement system.

The PEMS reports second-by-second engine and concentrations of exhaust gases data. The exhaust flow rate was estimated based on the engine parameters (i.e., MAP, IAT, and RPM) and exhaust gas concentrations using a mass balance approach well developed and used in the literature (Vojtisek-Lom and Cobb, 1997; Vojtisek-Lom Allsop, 2001; Frey et al., 2003; Zhang, 2006; Pang, 2007; Abolhasani et al., 2008). Gas analyzers and sensors used in this study were all commercially available, with comparable accuracy and precision to widely used PEMS such as HORIBA OBS2000 (Oliver et al., 2009), Montana 2000 (Frey, Zhang, and Rouphail, 2008) with a precision of ±3% ppm for NO, 0.01% for CO and CO₂, and 1 ppm for HC, respectively. In order to further verify the accuracy and precision of the developed PEMS, the system was compared to a HORIBA OBS2000 on an engine dyno. The R² of the scatter plot of the measured time-based emission rates of these two PEMS is 0.98 for NO and 0.99 for CO₂, respectively. The NO emissions from diesel engines are of most concern for air pollution compared to other gaseous emissions. In addition, although NO will convert to NO₂ quickly in the atmosphere and the NOx can be approximated by multiplying a coefficient from NO emissions, the ratio of NO to NO₂ in the mix of NO_x varies by the combustion and ambient conditions. Furthermore, NO_x emissions in the tailpipe include both NO and NO₂. Quantification of NO_x emissions solely based on NO measurement might be slightly underestimated. In order for interstudies comparisons where similar PEMS was used, NO emission data are directly analyzed and characterized.

Experimental design

This section describes the experimental design for the real-world measurements of emissions and activities for highway construction equipment, including duty cycles/modes definition; PEMS calibration, installation, and deinstallation; and data error checking and postprocessing. The details are given in the following.

Duty mode definition

Real-world emissions from construction equipment vary by different work duties (Abolhasani et al., 2008; Zhu et al., 2011; Fu et al., 2012; Sanhu et al., 2015). In order to obtain a more representative emission factor for emission inventory development, a composite emission factor, which accounts for different work duties during a normal usage, is preferred. Three duty modes are defined for this study, namely, idling, moving, and working. The idling mode refers to when the construction equipment is not doing anything while the engine is on; the moving mode refers to activities when the equipment is moving from one location to another during the work transfer without completing a work task; and the working mode refers to the rolling, paving, and other activities rather than idling and moving.

Although engine parameters such as MAP and RPM can be used for defining duty mode (Abolhasani et al., 2008) due to their significant differences in values between different duties (e.g., stable and small values of RPM and MAP are usually observed when the equipment is idling; the RPM and MAP are much higher when the equipment is performing duties), these parameters are internal variables to the engine and are not easy to obtain, especially when extensive activity data collection on off-road equipment are needed to derive time-weighted emission rates. Thus, only three duty modes are defined in this study.

During the measurements, in order to measure the emissions for each of the duty mode, the driver was advised to follow the typical real-world job sequence of idling, moving, and working without disrupting the duties. Each of the three modes usually lasts 10–20 minutes and mode changes are tagged. In addition, a video camera was used to record all the activities during a normal duty of an equipment item. These data were then used to estimate the time percentage of each mode and thus the composite emission factor. The details regarding the estimation of composite emission factors are given later in this paper.

PEMS calibration, installation, and deinstallation

In order to assure that the PEMS is under working conditions, the PEMS was calibrated before and after each measurement. The calibration gas with known concentration used in this study is 8.02% for CO, 20.06% for CO₂, 1603 ppm for C₃H₈, and 2950 ppm for NO. The ambient air was used as zero-air during measurements.

The PEMS was installed onto the selected equipment with all components being properly connected and securely fastened. The position where the PEMS was installed varies by equipment depending the available space of each equipment item for installation. Cushions were placed underneath the PEMS to reduce the vibration of the equipment to the PEMS. Usually, it takes 1–1.5 hr to install the PEMS and 45 min for deinstallation. In addition, a blower was used to clean the dust off the PEMS when the PEMS was deinstalled after each measurement.

Data error checking and postprocessing

Data error checking and postprocessing are essential for any real-world measurements. For measurements using the PEMS, the typical errors include: gas analyzer “freeze,” unusual data values, missing values, and others. In addition, since the PMES consists of a variety of measuring units, such as gas analyzer, sensors, global positioning system (GPS), and others, data from these units might not be recorded and reported with the same time stamp due to different response time of the measuring units and measuring location (Zhang and Frey, 2008). Data collected from these units should be checked for errors and synchronized before being used for further data analysis.

Fuel-based approach for emission factor estimation

The general procedures of developing fuel-based emission factors include:

1. Mass emission rates recalculation. After the collected data are checked and postprocessed, mass emission rates are recalculated using a mass balance-based approach as reported in the literature (Zhang, 2012).

2. Modal emissions analysis. Emissions rates for idling, moving, and working mode were summarized using the corresponding tagged data during the measurements. Both mass emission rates per time and per unit fuel consumption were estimated in this study.

3. Composite emissions rates development. The composite emission rates were calculated using the modal emission rates and its corresponding time percentage as shown in eq 1:

(1)

where CER_NO is the composite NO emission rate, g/sec, or g/kg-fuel; ER_{NO, I} the NO emission rate for duty mode i, g/sec or g/kg-fuel; Wt_i the time percentage during a normal task of an equipment, %; and i the duty mode, that is, idling, moving, and working. A time-based emission rate can be converted to a fuel-based emission rate by dividing the emissions by the corresponding fuel consumption, where both emissions and fuel consumption are on a time-resolved basis. The fuel-based and time-based composite emission rates are both developed and compared in this study.

A statistical two-sample t-test was used to evaluate the significance of differences in emissions data. The details of the t-test are given elsewhere (Casella and Berger, 2002).

Results and discussions

The key materials presented in this section include (1) real-world emissions measurements of highway paving construction equipment in this study and (2) mass emissions rates of highway paving construction equipment.

Real-world emissions measurement of highway paving construction equipment

Selection of highway construction equipment for in-use emissions measurement has been challenging due to accessibility and safety issues, and usually has to be done based on availability. For this reason, a highway construction site in Qingbaijiang in the proximity of Chengdu was selected for this study, where the second around-the-city expressway was under construction. Four paving construction equipment items available and frequently used at this site were chosen as examples for this study, including three road rollers and one paver, as shown in Table 1.

Table 1. Highway construction equipment selected for the study.

Equipment type	Manufacturer	Equipment model	Engine model	Engine size (L)	Power (kw)	After-treatment*	Emissions standards^d
Single-drum roller (2012)	Xuzhou^a	XS262J	SC8D1902G2	8.3	140	EGR	III
Double-drum roller (2012)	Xuzhou^a	XP301	SC8D180G2B1	8.3	132	EGR, SCR	III
Pneumatic tire roller (2012)	Hummer^b	HD130	BF4M2012C	4.7	98	EGR	III
Paver (2012)	Sichuan^c	MT9000C	6CTA83C215	8.3	160	EGR	III

^aThe Xuzhou Engineering Machinery Corporation of Jiangshu.

^bHummer, Inc., USA.

^cMeishan Construction Equipment, Inc., of Sichuan.

^dThe China National Emissions Standards Stage for Diesel Engines.

*EGR, exhaust gas recirculation; SCR, selective catalytic converter.

Figure 2 shows the installation of the PEMS to the selected equipment for emissions measurements. The installation of sensors and the placement of the PEMS are both very challenging.

Figure 2. Installation of PEMS to the selected equipment for emissions measurements.

The measurements were taken for a week in August 2015 at the preselected highway construction site while the selected equipment items were working on a project. The measurements were made on the equipment following the general project duty order, that is, paver, single-drum roller, double-drum roller, and pneumatic tire roller. Second-by-second engine and emissions were collected on each of these equipment for 8–9 hr continuously. Collected data were checked for errors and postprocessed before being used for analysis. Sensors collecting engine data generally perform better than the gas analyzer with less erratic rates. Among all identified errors in the collected data, the majority is the gas analyzer “freeze.” Usually exhaust gas concentrations change or fluctuate over time even when the true gas concentration is a constant. When the data do not change over 3 consecutive seconds, the gas analyzer likely freezes. Therefore, the corresponding records were all eliminated. Overall, the cleaned data accounts for approximately 50% of the total collected data and thus there are at least 4 hr of second-by-second data that can be used for data analysis for each of the selected equipment. As mentioned in the literature (Frey, Zhang, and Rouphail, 2008), 3 hr or more of second-by-second data is regarded as adequate to capture the majority variability in emissions.

Mass emissions and fuel consumption rates analysis

After data collected were checked for errors and postprocessed, the cleaned data were used for further analysis.

As shown in Figure 3, there is a large variability in vehicle emissions under real-world situations. Although there are variations in emissions within each duty mode, NO emissions for the working mode are the highest and are the lowest for the idling mode. For the single-drum roller, the moving mode has more variations than the idling mode, while for the paver, the moving mode only exhibits slightly higher overall emissions than the idling mode. The variations within each mode might be due to the work duties changes during the measurements as the equipment is working on a real project, although the driver was instructed to follow the predesigned procedure as discussed earlier.

Figure 3. The time-resolved modal emissions for a single-drum roller.

Since the other two rollers follow patterns similar to that of the single-drum roller as shown earlier, their time-resolved modal emissions patters are not repeated here. The calculated modal emissions and fuel consumption rates are shown in Table 2.

Table 2. Modal NO emissions and fuel rate for selected equipment.

		NO emission rate
Equipment	Duty mode	mg/sec	g/kg fuel	Fuel consumption rate (g/sec)
Single-drum roller	Idling	10.2	29.5	0.60
		(9.8–10.7)	(29.4–29.6)	(0.54–0.66)
	Moving	24.2	20.5	2.03
		(24.0–24.3)	(17.4–23.7)	(1.96–2.10)
	Working	50.8	18.0	4.56
		(50.6–51.1)	(17.9–18.2)	(4.56–4.56)
Double-drum roller	Idling	2.6	6.9	0.62
		(2.5–2.6)	(6.2–7.7)	(0.57–0.66)
	Moving	6.7	7.2	1.47
		(6.5–6.9)	(6.9–7.5)	(1.47–1.48)
	Working	10.0	6.0	2.53
		(10.0–10.1)	(5.9–6.0)	(2.53–2.53)
Pneumatic tire roller	Idling	5.8	30.6	0.31
		(5.7–5.8)	(30.5–30.7)	(0.31–0.32)
	Moving	11.4	28.4	0.66
		(11.4–11.4)	(28.1–28.6)	(0.66–0.67)
	Working	32.7	30.8	1.92
		(32.4–33.1)	(30.8–30.9)	(1.89–1.94)
Paver	Idling	25.5	52.7	0.77
		(24.0–27.0)	(49.9–55.5)	(0.77–0.78)
	Moving	25.4	55.6	0.73
		(25.3–25.5)	(55.0–56.3)	(0.70–0.76)
	Working	63.7	24.1	4.12
		(63.7–63.8)	(24.1–24.1)	(4.11–4.12)

Depending on duty modes and types of equipment, NO emission rate on average ranges from 2.6 mg/sec to 63.7 mg/sec, and the fuel consumption rate on average ranges from 0.3 g/sec to 4.56 g/sec. In general, time-based NO emission and fuel consumption rates are significantly different among duty modes. Both time-based NO emission and fuel consumption rates during the working mode are the highest compared to the other two duty modes with the idling mode being the least. However, for the paver, emission and fuel consumption rates for the idling mode and moving mode are not statistically significant. This is because the engine activities for these two modes for the paver are very similar (as shown in Table 3).

Table 3. Modal time distribution for typical highway construction equipment.

Equipment	Duty mode	Activities	Time perc. (%)
Single-drum roller	Idling	Waiting for the paver	3.6
	Moving	Forward	30.2
	Working	Vibration backward	66.3
Double-drum roller	Idling	Waiting for single-drum roller	1.5
	Moving	Forward	20.9
	Working	Vibration Backward	77.6
Pneumatic tire roller	Idling	Waiting for double-drum roller	9.5
	Moving	forward	41.0
	Working	Vibration backward	49.6
Paver	Idling	Waiting for loader	13.0
	Moving	Loader dumping	7.5
	Working	Mixing and paving	79.6

At the job site of this study, the general procedure for the roadway construction starts with the paver putting down concrete materials, followed by the single-drum roller, double-drum roller, and pneumatic tire roller, in order. Because the surface conditions are different at different stages of the construction, emissions and fuel consumptions rates are different. For rollers, the single-drum consumes most fuel per unit time, while the pneumatic tire roller consumes the least. The single-drum roller also emits more NO emissions than the pneumatic tire roller. Compared to these two rollers, the double-drum roller has less NO emissions. This might be because the double-drum roller is equipped with a NOx after-treatment device. However, the model and rated capacity of these rollers are different, so the impact of the use of the NOx after-treatment device on fuel consumption is confounded by these factors. For example, the fuel consumption for the double-drum roller is much higher than the pneumatic roller but is slightly lower than for the single-drum roller. Of course, since the sample size of this study is relatively small, for future study, a larger sample size of emission testing will be needed in order to fully evaluate the impact of the use of engine exhaust after-treatment devices on both emissions and fuel consumption.

For the rollers, comparing the time-based NO emission rate among different duty modes, NO emission rate for the working mode can be 4–6 and 1.5–3 times as high as that for the idling mode and the moving mode, respectively. However, the differences become a factor of 0.6–1 and 0.8–1.1 if the fuel consumption-based NO emission rate is compared. This implies that the variability in the fuel consumption-based emissions rate is relatively smaller than that in the time-based emissions rate. Thus, for a more accurate estimation of emission inventory, a fuel consumption-based emission factor may be used.

For the paver, comparing the time-based NO emission rate among different duty modes, NO emission rate for the working mode can be approximately 2.5 times as high as that for both the idling mode and the moving mode. However, the differences become a factor of 0.4 if the fuel consumption-based NO emission rate is compared.

In addition, because the emissions and fuel consumption rates changes due to change of duty modes are not proportional, the fuel consumption-based modal emission rate might not follow the same pattern as the time-based modal emission rate, especially for the idling mode. For example, the ratio of the fuel-based emission rate to the time-based emission rate for the idling mode can be 2–5 times, depending on equipment.

In order to develop a composite emission factor, the time distribution for each duty mode during a typical construction job should be estimated. In this study, the modal time distribution was analyzed and estimated using captured videos on site. The results are shown in Table 3.

Depending on equipment, the working mode accounts for more than 50% of the total work time. However, there is still 20.5–50.5% of the time that the equipment is not working on the task. The time distribution for each duty mode along with the modal emission rates are used to estimate the composite emission rates using eq 1. The results are shown in Figure 4.

Figure 4. Composite NO emission rates for selected highway construction equipment.

For rollers, time-based NO composite emission rate ranges from 9 to 41mg/sec, with the single-drum roller being the highest and the double-drum roller being the lowest. The fuel consumption-based NO composite emission rate ranges from 6 to 30 g/kg-fuel, with the pneumatic tire roller being the highest and the double-drum roller being the lowest. The reasons are similar to those discussed earlier in this paper. For the paver, both time-based and fuel consumption-based NO composite emission rates are higher than all of the rollers, with 56 mg/sec and 30 g/kg-fuel, respectively.

Since there are few studies in emissions from rollers and pavers, only construction-related equipment was compared. As shown in Table 4, the fuel-based emission factors developed in this study are slightly lower than what have been reported in the literature (Kean et al., 2000; Frey et al., 2010; Fu et al., 2012) except for OFFROAD. Reasons include that (1) the testing equipment in this study are in compliance with higher emission standards than those tested in the literature; and (2) the comparisons are not side-by-side equipment type-wise. Furthermore, as seen in the literature, even when the same type of equipment was compared, there also exists a huge difference. Because the horsepower for the diesel-fueled pavers and rollers in the OFFROAD is relatively small (less than 20 kW), the emissions rates are smaller compared to others. This implies that emission factors used for inventory development should be regionally based and equipment type specific.

Table 4. Comparisons of fuel-based NO emissions rate by different studies.

Studies	Equipment	g/kg fuel
This study (2015)	Single-drum roller	19.21
	Double-drum roller	6.23
	Pneumatic tire roller	29.84
	Paver	30.22
Fu et al. (2012)	Excavator	31.09
	Wheel loader	83.38
Frey et al. (2010)	Excavator	36.88
	Wheel loader	42.56
Kean et al. (2000)	Construction and mining	46.00
OFFROAD*(CARB, 2007)	Paver	3.65
	Roller	3.83

*Emission rates are converted from the average g/hp-hr using the average brake-specific fuel consumption (BSFC) given in the OFFROAD2007 model. In addition, the horsepower for diesel-fueled pavers and rollers in OFFROAD2007 is less than 25 hp.

As discussed earlier, when a time-resolved measure (mass per time) is used to quantify the emissions, the emission rate in the idling mode is relatively small compared to other modes. However, when a fuel-based approach is used, the emission rate during the idling mode is still comparable to other duty modes. The nonworking modes usually are associated with high fuel-based emission rates and thus will have significant impact on composite emission factor estimation. This implies that emissions from highway construction equipment could be significantly reduced if the time spent in nonworking modes could be reduced through applicable operation management such as minimum idling requirement.

Summary

This paper developed and demonstrated a fuel-based approach for emission factor estimation for highway surfacing construction equipment. The developed PEMS was shown to be capable of taking real-world measurements of emissions for off-road equipment such as rollers and pavers. Furthermore, wireless communication between the measuring unit and the computer enhances the PEMS’s applicability and capability for emissions measurement in a real-world situation. The developed approach can be used for extensive data collection of emissions from off-road equipment and estimations of emissions factors for inventory development.

Emissions vary by different duty modes and equipment. Both inter- and intra-equipment variability in emissions should be taken into account for emission inventory development. The fuel-based emission rates usually have less variability than the time-based emission rates. Thus, in order to better estimate the emission inventory for off-road equipment, a fuel-based composite emission rate taking into account different duty modes may be used. In addition, compared to the time-resolved emission factors (e.g., g/hp­hr) used for emission inventory development, the equipment activity data (use hours) usually are difficult to obtain thus a robust estimate of emission inventory is very challenging. However, since it is relatively easy to get the fuel consumption data, a fuel­based emission factor will have benefits for improving emissions estimates.

There exist large uncertainties in emissions from both on-road and nonroad equipment, as reported in the literature (Frey, Zhang, and Rouphail, 2008; Frey et al., 2010). In order to capture as much as possible the variability in emissions from nonroad equipment, emission measurements made with a longer period of time will be needed when a small number of equipment is selected. Although only four equipment items were selected for emissions measurements for this study, the amount of collected second-by-second data is adequately large to cover the majority of the variability in emissions from these equipment items used for highway construction. Thus, the collected data and the developed fuel-based emission factors can be used for emissions inventory updates for the corresponding equipment. It should be noted that an extensive data collection on off-road equipment, even on the equipment tested in this study, is still needed in order to improve the accuracy of the emission inventory development. The developed approach for data collection and analysis in this study can be served as a guideline for the forthcoming efforts in developing emission inventory from nonroad equipment in China.

Acknowledgment

The authors thank Shiying Pu, Wu Cheng, Kailong Chen, and many others for acquiring the testing equipment and facilitate the tests. The authors are thankful to the AMED group of Sichuan University in helping data collection during this tough time.

Funding

This study was sponsored by the Public Environmental Service Project of the Ministry of Environmental Protection of PRC (number 201409012). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the sponsoring parties mentioned.

Additional information

Funding

This study was sponsored by the Public Environmental Service Project of the Ministry of Environmental Protection of PRC (number 201409012). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the sponsoring parties mentioned.

Notes on contributors

Zhen Li

Zhen Li is a graduate student in the Department of Environmental Science and Engineering at Sichuan University. She holds a BE degree in Environmental Engineering from Chongqing University of Arts and Science.

Kaishan Zhang

Kaishan Zhang is a faculty member in the Department of Environmental Science and Engineering at Sichuan University.

Kaili Pang

Kaili Pang is a senior undergraduate student in the Department of Environmental Science and Engineering at Sichuan University.

Baofeng Di

Baofeng Di is a faculty member in the Department of Environmental Science and Engineering at Sichuan University.