Assessment of the Impact of CO, NO x and PM 10 on Air Quality during Road Construction and Operation Phases

: The road sector is one of the main sources of air emissions in the atmosphere during both construction and operation. The objective of the present paper is a comprehensive evaluation of the impact on air quality during the two main phases of life cycle of roads. In this case study of a motorway project, the emissions of the primary pollutants, CO, NOx, and PM 10 are estimated, and the results showed that (i) CO and NOx pollutants released during both phases are comparable, while the emissions of PM 10 are more signiﬁcant in the construction phase; (ii) 85% of PM 10 in construction is due to storage, transit on unpaved road, and crushing; (iii) the portals of the tunnel are the sites where there are higher concentrations of pollutants in operation; and (iv) the CO concentrations estimated by the dispersion model are strongly inﬂuenced by the topography. for the abatement of particulate in the crushing plants. In operation, roadside vegetation barriers can help to the pollution dispersion. ﬁndings of this the stakeholders, road designers, controllers, and enterprises, to develop activity-speciﬁc (in construction) and site-speciﬁc (in construction and operation) control strategies and mitigation measures set up properly monitoring plans in order to reduce the emissions and make road construction and operation more and sustainable. The ﬁndings are related to the motorway project and therefore merit further assessment and validation in future research studies about other types of roads, locations, and tra ﬃ c conditions.


Introduction
The road sector is one of the main sources of air emissions in the atmosphere during both construction and operation [1]. According to the European Commission, in 2018, the 24% of global CO 2 emissions due to fuel combustion derived from the transportation sector. In particular, road transportation of passengers and goods (by car, truck, bus, or two-wheeler) was by far the main sector responsible, accounting for nearly three-quarters of emissions (72%). These figures extensively justify the growing interest towards the methods and practices in road construction, maintenance, rehabilitation, and operation aimed at reducing the air emissions. Furthermore, the increasing demand for freight and passenger transport, especially in developed countries, calls for an increasing attention to issues relating to air pollution. After all, the negative health effects of air pollutants, and particularly the particular matter, are well known and widely demonstrated by several studies [2,3], together with the economic implication deriving from the related increased premature mortality [4]. The exposure to air pollution is recognized as the major global risk factor for premature morbidity and mortality and has been linked to cardiovascular diseases [5], hypertension [6], and atherosclerosis [7]. Not negligible effects are also produced in the local vegetation [8].
All that considered, the assessment of the concentration of these substances in the air and the adoption of proper control strategies and mitigation measures are fundamental steps in achieving the environmental and social sustainability of road transportation.
During construction, the site activities such as material production and transportation, equipment and machineries, and excavation and earthmoving release into the atmosphere a large amount of pollutants [9,10]. Off-road machines, which are diesel-powered, are a significant font of direct air emissions such as carbon oxide (CO), nitrogen oxides (NO x ), and particular matter (PM). The U.S. Environmental Protection Agency (EPA) estimates that off-road diesel equipment is the third largest source of NO x emissions from mobile fonts (14.5%) and the second largest source of PM (24.3%). Significant sources of PM emission are also the transits of trucks on unpaved roads, the aggregate crushing, and the storage of excavated materials [10].
During operation, the air pollution is due to vehicles exhaust gases. The pollution levels, especially in transportation microenvironments (i.e., near roadway or inside the tunnels), are higher than ambient concentrations [11,12], but the higher concentration, in case of major roads, is exhausted within few hundred meters.
In the literature, several studies have treated the topic of air pollution, but they answer the questions related to the emission and concentration of air pollutants separately for construction and operational phases of the roads. For construction, a lot of research has been dedicated to the effort of creating emission inventories for the off-road equipment [13][14][15]. Some authors [16,17] apply the portable emission measurement system to estimate the non-road machinery emission. Pirjola et al. [18] compared the exhaust emissions from a tractor in real-world and laboratory conditions. They found that emission factors were higher on-road compared to laboratory conditions and highest particle emissions were observed on-road during accelerations and engine braking. Other studies studied the influence of construction sites on air quality and focused on the estimation of emissions from road construction [9,[19][20][21]. Mulenski et al. [22] investigated on PM 2,5 and PM 10 emission factors to assess emission inventories for the main construction activities.
Regarding the operation, different approaches have been pursued in the urban and rural contexts. At the urban scale, Munõz Miguel et al. [23] developed an air pollution model to evaluate the possible air pollution decrease in Madrid as the consequence of a hypothetical road pricing. They find that road pricing has significant effects on decreasing of air pollution during the peak operating and on shift from private car to public transport.
Wanga et al. [24] proposed a study in which the traffic-related emissions at intersections have been analyzed. Results of the study show that the average concentration of the pollutants is higher than the background site, and the level of traffic and the meteorological conditions also affects it.
In urban and rural roads, numerous studies have focused on the real-time air quality monitoring [25,26] and on the modeling of the traffic-related pollutants' dispersion [27,28].
In the rural context, some studies have emphasized the importance of taking into account the effect of road grade in calculation of the emission, especially when duty-heavy vehicles are considered [29][30][31].
The combined assessment of the impact on air quality due to the construction and operation of road is scarcely represented in literature. In this view, the main contribution of this paper is precisely the assessment, at design stage, of the impact on air quality of the construction activities and traffic of a new motorway. The purpose is, firstly, the comparison of the impact on air quality of two important phases of the road life cycle. This is because, for main roads, such as the motorway, especially when made in complex contexts (hill or mountains), the construction phase can have a long duration due to the construction of demanding components, such as tunnels and bridges, and therefore the impact on air quality, although temporary, can produce undesirable and inacceptable effects on communities.
The second main objective of this study, which is part of wider research, is the assessment of the admissibility of the air emissions and the identification of the hotspots. In this way, it is possible to provide accurately indications in the development of effective strategies for pollution control and air quality management at roadsides for the aim of making the pollutants concentration tolerable and complying with the regulations. Previous papers reporting the results on the on-going research focused on the detailed analysis of the amount of emissions deriving from construction of the motorway [9] and on the study of the dispersion of concentration of PM 10 emissions nearby the worksites [10]. The present paper takes into account also the traffic-related emissions during the operational phase for a comprehensive evaluation of the air emissions of a motorway and in order to quantify the real impact of the two road life steps on the air quality.

Materials and Methods
The method used in this work for assessing the air emission during construction and operation consists of the following steps: • Definition of the study site • Data collection on construction plan (construction sites, construction activities, duration of activities, equipment, and plants), motorway features (configuration, traffic flows), topography, and meteorology • Calculation of emissions in construction and operation phases • Modeling of critical pollutant dispersion in operation phase • Comparison of impact on air quality in the two phases. Figure 1 illustrates the conceptual framework of the study. The main steps are described in detail in the following paragraphs.

Materials and Methods
The method used in this work for assessing the air emission during construction and operation consists of the following steps:


Definition of the study site  Data collection on construction plan (construction sites, construction activities, duration of activities, equipment, and plants), motorway features (configuration, traffic flows), topography, and meteorology  Calculation of emissions in construction and operation phases  Modeling of critical pollutant dispersion in operation phase  Comparison of impact on air quality in the two phases. Figure 1 illustrates the conceptual framework of the study. The main steps are described in detail in the following paragraphs.

Study Site
The assessment of pollutants emissions in construction and operation in this work was carried out by considering a project of an Italian motorway, three lanes dual carriageway, belonging to the A category of Italian road standards [32]. The motorway has a length of 18 km, distributed into 5 tunnels (13.3 km), 5 bridges (1.4 km), and fills and embankments (3.2 km). The scheme of the motorway under investigation is shown in Figure 2, which also reports the distribution of the worksites.

Study Site
The assessment of pollutants emissions in construction and operation in this work was carried out by considering a project of an Italian motorway, three lanes dual carriageway, belonging to the A category of Italian road standards [32]. The motorway has a length of 18 km, distributed into 5 tunnels (13.3 km), 5 bridges (1.4 km), and fills and embankments (3.2 km). The scheme of the motorway under investigation is shown in Figure 2, which also reports the distribution of the worksites.

Materials and Methods
The method used in this work for assessing the air emission during construction and operation consists of the following steps:


Definition of the study site  Data collection on construction plan (construction sites, construction activities, duration of activities, equipment, and plants), motorway features (configuration, traffic flows), topography, and meteorology  Calculation of emissions in construction and operation phases  Modeling of critical pollutant dispersion in operation phase  Comparison of impact on air quality in the two phases. Figure 1 illustrates the conceptual framework of the study. The main steps are described in detail in the following paragraphs.

Study Site
The assessment of pollutants emissions in construction and operation in this work was carried out by considering a project of an Italian motorway, three lanes dual carriageway, belonging to the A category of Italian road standards [32]. The motorway has a length of 18 km, distributed into 5 tunnels (13.3 km), 5 bridges (1.4 km), and fills and embankments (3.2 km). The scheme of the motorway under investigation is shown in Figure 2, which also reports the distribution of the worksites.

Data Collection
The primary step for the calculation of the emissions during both construction and operational phases consists in the collection of the following data: • Relevant characteristics (position, extension, function) of 25 work areas distinguished in construction sites (CS), in which the storage of materials and the production of the concrete and the elements for tunnels lining take place, and in technical areas (TA), where all equipment and plants needed for construction of bridges and tunnels are located.

•
The type of off-road machines and the equipment used in each worksite (see Table 1). The number and type of equipment depend on the construction activities taking place in the worksite. Table 1. Off-road machines and equipment in construction sites [9]. • The duration of construction of the motorway: 8 years.

•
The traffic data. During operation, the amount of emissions depends mainly on the number and type of vehicles circulating. The motorway under examination, which has a length of 18 km, can be subdivided in two sections for the presence of two interchanges located at 4.5 km from the beginning and at the end. The day and night peak hour flow (maximum hourly flow during the day time is 6:00 a.m.-10:00 p.m. and night time is 10:00 p.m.-6:00 a.m.) and the percentage of heavy traffic (vehicle weight more than 3 tons) for the two sections are reported in Table 2. • The data on the actual air quality gathered at five monitoring points along the site of the future motorway.

Modelling of Pollutants Emission
The calculation of the emissions was carried out separately for the construction and operation phase. For construction, based on the gathered data, the first step of the study was the calculation of the emissions in each worksite originating from the activities that took place there.
To do this, the emission factors for all the running equipment and construction activities were calculated.
For equipment and trucks, the emission factors provided by [33,34] and reported in Table 3 were used. In addition, in the worksites, the following activities were considered: Transits of trucks on unpaved road • Crushing of aggregates.
These activities produce a large amount of fine particulate. The models used for the calculation of the PM 10 emission factors are reported in Table 4. Major details of the calculations are reported in [9]. Table 4. PM 10 emission factors.

Activity Model Reference
Transits onunpaved road a and b = coefficients [33] Topsoil excavation E = 3.42 [kg/km] [33] Storage  For the calculation of the traffic-related emissions in operation, reference was made to the database provided by ISPRA-SINAnet [34] for the motorways.
In the definition of the emissive flows, the following hypotheses have been adopted: • the emissions are referred to in the 2015 Italian fleet, consisting of 39.7% of diesel and 60.3% of gasoline vehicles • the emissions are estimated considering the 2030 as reference year. According to PIARC [36], corrective factors are adopted in forecasting the future emissions. These factors consider that vehicle legislation has enforced more stringent emission rates and that vehicle technology has rapidly advanced, resulting in lower emissions.
The corrective factors applied to calculate the emissions from 2015 to 2030 are reported in Table 5. Considering the distribution of the type of vehicles in the fleet, the emissive factor derived and applied are those reported in Table 6. In order to quantify the emissions released in correspondence of the tunnel portals, a box model (perfectly mixed phase) is applied [37]. The model, based on the principle of mass conservation, is able to estimate, in stationary and conservative conditions, the concentrations of the different pollutants that are generated inside the tunnel.
The assumptions underlying the box model are the following: The equations for estimating the concentration inside the tunnel are the following: Inlet flow = u wind ·W·H·C 0 + q·W·L Output flow = u wind ·W·H·C The mass flow of a generic pollutant coming out of the portal is equal to the sum of the flow rate emitted inside Q and the flow rate determined by the incoming air having a concentration C 0 .
It is assumed that all the emissions generated inside the tunnel are released in correspondence of the portals.
Therefore, the characterization of the source has been performed as follows: • Estimate of the emission inside the tunnel taking into account its average length and the traffic; • Definition of a virtual point source at the tunnel portal, diameter 6 m, height equal to half of the height of the tunnel, speed 1 m/s and temperature equal to 15 • C, at which is assigned the emission of the entire tunnel section.
The calculated emissions of each section of the motorway are summarized in Table 7.    It is possible to observe that in both cases, CO is the highest emitted pollutant in all worksites and in all motorway sections. This result complies with the existing literature [19,20,21]. Considering the average value of the pollutants, CO is about twice higher than NOx and about four times higher than PM10 in construction stage, while the prominence increases in operation, because the traffic produces a concentration of CO which is 2.6 times higher than NOx and 49 times higher than PM10. The average value of NOx in motorway construction is about double of that in operation. This is because during construction diesel powered vehicles are operated, which are the main cause of NOx emission. On the contrary, gasoline vehicles (main components in vehicular fleet) produce high emissions of CO and low emissions of NOx.
In operation, the maximum values of CO and NOx emissions are related to tunnel 3, and this is due to the fact that it is the longest tunnel in the motorway and the pollutants' concentration at the portals is therefore the highest. For tunnel 4, it is also possible to observe emissions higher than the average of the other sections.
It is worthy to note that, except the aforementioned concentration peaks in operation, the construction phase produces in the worksites an amount of pollutants, which are averagely higher than the ones related to traffic in the corresponding section of the motorway. This outcome emphasizes the need for accurate estimates and effective control procedures of air pollution in road construction.
PM10 is a significant pollutant in the construction phase, although with a lower emission value with respect to the other pollutants. In operation, the PM10 emitted is one order of magnitude less than the one estimated in construction.
85% of the PM10 emissions in the construction phase is due to, on average, activities such as storage, crushing, and transit on unpaved roads, according to literature results [19,22], and about 15% is due to equipment and trucks operation ( Figure 5). It is possible to observe that in both cases, CO is the highest emitted pollutant in all worksites and in all motorway sections. This result complies with the existing literature [19][20][21]. Considering the average value of the pollutants, CO is about twice higher than NO x and about four times higher than PM 10 in construction stage, while the prominence increases in operation, because the traffic produces a concentration of CO which is 2.6 times higher than NO x and 49 times higher than PM 10 . The average value of NO x in motorway construction is about double of that in operation. This is because during construction diesel powered vehicles are operated, which are the main cause of NO x emission. On the contrary, gasoline vehicles (main components in vehicular fleet) produce high emissions of CO and low emissions of NO x .
In operation, the maximum values of CO and NO x emissions are related to tunnel 3, and this is due to the fact that it is the longest tunnel in the motorway and the pollutants' concentration at the portals is therefore the highest. For tunnel 4, it is also possible to observe emissions higher than the average of the other sections.
It is worthy to note that, except the aforementioned concentration peaks in operation, the construction phase produces in the worksites an amount of pollutants, which are averagely higher than the ones related to traffic in the corresponding section of the motorway. This outcome Sustainability 2020, 12, 10549 9 of 14 emphasizes the need for accurate estimates and effective control procedures of air pollution in road construction. PM 10 is a significant pollutant in the construction phase, although with a lower emission value with respect to the other pollutants. In operation, the PM 10 emitted is one order of magnitude less than the one estimated in construction.
85% of the PM 10 emissions in the construction phase is due to, on average, activities such as storage, crushing, and transit on unpaved roads, according to literature results [19,22], and about 15% is due to equipment and trucks operation ( Figure 5). The amount of PM10 emissions released by the equipment and trucks in construction is comparable to those released by the traffic in operational phase. The identification of the most emissive activities allows for the planning of effective strategies for control and mitigation measures.
Regarding the traffic-related emissions, in addition to the pollutants CO, NOx, and PM10, other relevant pollutants, such as benzene (C6H6) and volatile organic compounds (VOC), have also been considered. Figure 6 shows the emissions of these pollutants along the motorway separately for daytime (6 am-10 pm) and night-time (10 pm-6 am). Apart from the specific values depending on types of pollutant, the trend of variation along the motorway is similar, as is to be expected, because the traffic and the motorway configuration play a crucial role in the amount of emissions. As it is obvious, night emissions are notably reduced with respect to the daytime ones, since the traffic at night-time is about the 19% of the daily traffic. The amount of PM 10 emissions released by the equipment and trucks in construction is comparable to those released by the traffic in operational phase. The identification of the most emissive activities allows for the planning of effective strategies for control and mitigation measures.
Regarding the traffic-related emissions, in addition to the pollutants CO, NO x , and PM 10 , other relevant pollutants, such as benzene (C 6 H 6 ) and volatile organic compounds (VOC), have also been considered. Figure 6 shows the emissions of these pollutants along the motorway separately for daytime (6 am-10 pm) and night-time (10 pm-6 am). Apart from the specific values depending on types of pollutant, the trend of variation along the motorway is similar, as is to be expected, because the traffic and the motorway configuration play a crucial role in the amount of emissions. As it is obvious, night emissions are notably reduced with respect to the daytime ones, since the traffic at night-time is about the 19% of the daily traffic. The amount of PM10 emissions released by the equipment and trucks in construction is comparable to those released by the traffic in operational phase. The identification of the most emissive activities allows for the planning of effective strategies for control and mitigation measures.
Regarding the traffic-related emissions, in addition to the pollutants CO, NOx, and PM10, other relevant pollutants, such as benzene (C6H6) and volatile organic compounds (VOC), have also been considered. Figure 6 shows the emissions of these pollutants along the motorway separately for daytime (6 am-10 pm) and night-time (10 pm-6 am). Apart from the specific values depending on types of pollutant, the trend of variation along the motorway is similar, as is to be expected, because the traffic and the motorway configuration play a crucial role in the amount of emissions. As it is obvious, night emissions are notably reduced with respect to the daytime ones, since the traffic at night-time is about the 19% of the daily traffic.  Figure 7 illustrates the share of CO emissions for each type of motorway segment. It is possible to note that tunnels, which constitutes about the 70% of the motorway alignment, mainly contribute (72%) to the CO emissions, in particular the tunnel portals, which are the hotspots for this pollutant. Cuttings contribute for 17%, viaducts for 8%, and embankment for 3%. A similar trend is recorded also for NOx and PM10.

Modelling of CO Dispersion
With specific reference to CO, which is the highest emitted pollutant, and to the sections of the motorway from tunnel 3 to tunnel 5 characterized by the highest value of the emissions (see Figures  2 and 4), the modelling of the dispersion is carried out using the CALPUFF model [38]. This model simulates the atmospheric pollution dispersion, taking into account the topography and is based on meteorological data (wind, temperature, pressure) processed with CALMET [39]. Table 8 summarizes the characteristics of the road sections included in the modelling and simulated as an area source (embankments and viaducts), while Table 9 reports the characteristics of the tunnel portals considered as point sources.

Viaducts 8%
Embankments 3%  Figure 7 illustrates the share of CO emissions for each type of motorway segment. It is possible to note that tunnels, which constitutes about the 70% of the motorway alignment, mainly contribute (72%) to the CO emissions, in particular the tunnel portals, which are the hotspots for this pollutant. Cuttings contribute for 17%, viaducts for 8%, and embankment for 3%. A similar trend is recorded also for NO x and PM 10 .  Figure 7 illustrates the share of CO emissions for each type of motorway segment. It is possible to note that tunnels, which constitutes about the 70% of the motorway alignment, mainly contribute (72%) to the CO emissions, in particular the tunnel portals, which are the hotspots for this pollutant. Cuttings contribute for 17%, viaducts for 8%, and embankment for 3%. A similar trend is recorded also for NOx and PM10.

Modelling of CO Dispersion
With specific reference to CO, which is the highest emitted pollutant, and to the sections of the motorway from tunnel 3 to tunnel 5 characterized by the highest value of the emissions (see Figures  2 and 4), the modelling of the dispersion is carried out using the CALPUFF model [38]. This model simulates the atmospheric pollution dispersion, taking into account the topography and is based on meteorological data (wind, temperature, pressure) processed with CALMET [39]. Table 8 summarizes the characteristics of the road sections included in the modelling and simulated as an area source (embankments and viaducts), while Table 9 reports the characteristics of the tunnel portals considered as point sources.

Modelling of CO Dispersion
With specific reference to CO, which is the highest emitted pollutant, and to the sections of the motorway from tunnel 3 to tunnel 5 characterized by the highest value of the emissions (see Figures 2  and 4), the modelling of the dispersion is carried out using the CALPUFF model [38]. This model simulates the atmospheric pollution dispersion, taking into account the topography and is based on meteorological data (wind, temperature, pressure) processed with CALMET [39]. Table 8 summarizes the characteristics of the road sections included in the modelling and simulated as an area source (embankments and viaducts), while Table 9 reports the characteristics of the tunnel portals considered as point sources.  Results of dispersion modelling are reported in the map in Figure 8, where the areas with different values of CO concentration (average of 8 h) are represented with different colors. The values of the concentration are the sum of the background concentration measured in the two monitoring points reported in the maps and the contribution of the future traffic in the motorway. By analyzing the results of the simulation, it is possible to observe how the spatial distribution of concentrations of pollutants on the ground is strongly influenced by the topography of the area. In general, the areas at higher CO concentration are those adjacent to tunnel portals. However, also in these hotspots, the concentration does not determine overcoming of the normative limits (CO limit = 10 mg/m 3 ) in correspondence of the receptors, but on the contrary, the concentration remains well below the limit.  Results of dispersion modelling are reported in the map in Figure 8, where the areas with different values of CO concentration (average of 8 h) are represented with different colors. The values of the concentration are the sum of the background concentration measured in the two monitoring points reported in the maps and the contribution of the future traffic in the motorway. By analyzing the results of the simulation, it is possible to observe how the spatial distribution of concentrations of pollutants on the ground is strongly influenced by the topography of the area. In general, the areas at higher CO concentration are those adjacent to tunnel portals. However, also in these hotspots, the concentration does not determine overcoming of the normative limits (CO limit = 10 mg/m 3 ) in correspondence of the receptors, but on the contrary, the concentration remains well below the limit.

Conclusions
This study deals with the assessment of the impact on air quality of the construction and operation of a motorway. The impact on air quality is assessed in terms of the emissions of the primary pollutants CO, NO x , and PM 10 . The main conclusions of the analyses and evaluations carried out are: 1.
The rate of CO and NO x pollutants released during both phases, construction and operation, are comparable, while the emissions of PM 10 are lower. However, PM 10 is a significant pollutant in the construction phase, being one order of magnitude higher than in operational phase. 2.
PM 10 in the construction phase originated in a percentage of 85% by storage, transit on unpaved road, and aggregate crushing, and in little part (15%) by the exhaust gases emissions of equipment and trucks. 3.
In operation, the portals of the tunnel, throughout which exit all the emissions produced inside the tunnel, are the main hotspots of the motorway. The pollutant concentrations in these points are four times higher than the maximum values estimated in the construction for CO and NO x . Except these points, in most cases, the average emissions of CO and NO x in construction is higher than the ones in the operation phase. 4.
As a consequence of these findings, in road construction, an accurate estimate of air emissions appears essential, as well as the constant monitoring aimed at the comparison of the estimated theoretical values from the equations and the measured ones and the implementation of proper mitigation measures, which can be eventually further tailored based on the monitoring outcomes. The attention to the construction phase should be higher, especially for the roads that require activities' duration to be significantly higher for the presence of demanding components, tunnels and bridges. 5.
The most emissive activities in construction and the tunnel portals in motorway operation are the hotspots, to which the greatest attention must be directed in order to achieve objectives of sustainability in road transportation.
The overall assessment of how the motorway affects air quality in the two main phases of its life cycle has yielded interesting results of more general application than the case study covered in the paper. In fact, the detailed analysis of the construction and operation phases and their implications in terms of pollutants emission allowed highlighting the activities and sections more emissive and therefore preferable targets of mitigation strategies. Effective mitigation measures during construction can be wetting of the ground, coverage of the open-air container of the truck when filled with construction materials, and application of modern technologies for the abatement of particulate emissions in the crushing plants. In operation, roadside vegetation barriers can help to reduce the pollution dispersion.
The findings of this study can assist the stakeholders, road agencies, designers, controllers, and enterprises, to develop activity-specific (in construction) and site-specific (in construction and operation) control strategies and mitigation measures and to set up properly monitoring plans in order to reduce the emissions and make road construction and operation more environmentally and socially sustainable. The findings are related to the motorway project and therefore merit further assessment and validation in future research studies about other types of roads, locations, and traffic conditions.