Air Emission Inventory

Emission inventory (EI) is a comprehensive listing of pollutants from all sources in a geo- graphical area during a period of time. The development of EI database is very important to air quality management and helps us to find out main sources. In general, to develop a reliable EI for one specific area, the following steps are involved: (i) list the types of sources, (ii) determine the types of air pollutant emission from each of the listed sources, (iii) find out the emission factor (EF) for each of the concerned pollutants, (iv) determine the number and size of specific sources in the area, and (v) multiply the appropriate numbers from step (iii) and (iv) to obtain the total emissions and then sum up the similar emissions to obtain the total for the area. With emission inventory, we can define the main sources in the study area and have the best solution for clean air action. Besides, the result from EIs can be used to study in other sectors related to human health, economic loss from air pollution, etc. This chapter will describe the method for development of an air emission inventory for developing and developed countries.


Introduction
In general, air pollutants are defined as substances in the atmosphere, resulting from the activity of humans or from natural processes, causing harm to humans, plants, or animal lives, or damage to man-made materials and structures, or changes in the weather and climate, or interfering with the comfortable enjoyment of life or other human activities (Elsom, 1987).
There are six common air pollutants (criteria pollutants) which are commonly found in air anywhere. These main pollutants should be included in the emission inventory (EI): TSP (PM 10 and PM 2.5 ), NOx (the sum of NO and NO 2 ), SO 2 , and CO. Other pollutants as VOC and CO 2 will be considered based on the characteristic of the study area and the aim of EI.

Defined sources
Normally, air emission sources are divided into three main categories: area source, point source, and line source. Each source includes many subcategories based on its distributed characteristic, which was shown in detail in Figure 1.

Point sources
Point sources are stationary sources that are typically large manufacturing or production plants that include confined, chimney or stack emission points. Within a given point source, there may be several units of combustion, such as several boilers. The point sources can be hospital incinerators, hospital boilers, hotel boilers, crematorium, and industries.

Area sources
Area sources are defined as sources that are too small and/or too numerous to be considered as point sources (US EPA, 1996). In aggregate, these sources may contribute a significant proportion of the total airshed emissions. Area sources include a wide range of sources, such as residential cooking facilities, rice straw/biomass burning, gas stations, construction activities, printing facilities, auto repair facilities, paint spraying facilities, and so on. They are presented as an area on the map.

Line sources (mobile sources)
Line sources, also considered as mobile sources, are vehicles, engines, and equipment that generate air pollution and that move or can be moved from place to place. Mobile sources are classified under two smaller sources: on-road and non-road sources. On-road sources include vehicles used on roads for transportation of passengers or freight, such as passenger cars, trucks, and motorcycles that may be fueled with gasoline, diesel fuel, or alternative fuels, such as alcohol or natural gas. Non-road sources include gasoline and diesel powered vehicles, engines, and equipment used for construction, agriculture, transportation, recreation, and many other purposes. For example, non-road sources include marine vessels, aircrafts, railways, tractors, etc. They are presented as line shape on the map.

Point sources
The aim of point source emission inventory is to give a comprehensive table sheet of sources, their characteristics, and annual emissions. A source which is large enough and contributes a huge amount of emissions can be considered as a point source, such as the hospital boilers, crematoriums, factory's chimneys, hotel boilers, etc. So the first step is to identify all sources in the study area and classify them into categories based on product characterization and combustion process. The list of these sources and emission data can be provided from local Province Officer and Department of Industry. For estimating emission, further process information from each source type needs to be provided directly from questionnaire or inspection. These data are specific for each category and also used for defining activity data, air emission process, and for input of dispersion model. The data normally require: factory type, size, geographical position (coordinates), number of employees, type of industrial activity: what is produced, combustion activities characterized by firing configuration, operating conditions, fuel specification, fuel consumption, raw material used in industrial/manufacturing process, storage of raw material, equipment age, height of stacks (and if available: inner diameter, volume flow in m 3 /operating hour), hours of operation per year/season, air pollution abatement equipment, and continuous emission monitoring.

Example questionnaire for point source is shown in Appendix A
The emissions for each source category are calculated using Eq. (1) and emission factors. Due to the lack of data monitoring from factories, emission factors are adopted. They can be taken from these following sources, appropriate for production type and process:  [2]. The information can be accessed online. The emission factor, which is taken from this guide, should be explained and clearly stated in the report.

Area sources
Area sources are similar to point sources, including small sources with insignificant emissions, evenly distributed in a wide range area. The aggregation of all these sources contributes a considerable emission to air quality. For distributing on the map, emissions from these sources are estimated by each grid cell.
In the beginning, a list of all sources and their information are collected from related local Province Office or other related Departments, as well as the data of inhabitants, the number of households, and the density rate in each ward. These data are used for calculating emission from common sources distributed evenly, such as household cooking, using a top-down approach by which the total emissions for the whole region are calculated, and then they are distributed by each grid cell according to the number of households and the density rate. Sources with the same characteristic are classified under a category, and an appropriate emission estimation method is chosen depending on the source characteristic and the data available. The total emissions of one grid include the emission from household activities and from other sources.
Source categories are different depending on the feature of the study area, and some common categories for cities in developing countries are: biomass (straw) burning, solid waste open burning, household cooking, smoke for meat and fish curing, gas stations, construction, concrete mixing, garages, commercial/consumer solvent used, and graphic arts.
Normally, within one source category (k), combustion technology and fuel type used are similar so that generalized emission factor and fuel consumption rate (or number of source category k) can be used for estimating emission as in Eq. (1). So the main data collected from survey is fuel consumptions, fuel type, and source type that are included in questionnaires for area sources shown in Appendix A. Though several source categories such as photocopies and gas station emissions are not from combustion but from evaporative VOCs, they are also calculated by using this equation with appropriated emission factors.
where E k.i = total emission from pollutant i in source category k in selected grid cell, A k = activity rate: that is, the amount of fuel burnt of source category k in the selected grid cell or surrogate data, and EF k,i = appropriate average emission factor for pollutant i for source category k.
Emission factors for these sources are taken from USEPA, EMEP/EEA Emission Inventory Guidebook or from studies in the local university.

Line sources (mobile sources)
Line (mobile) sources include on-road sources and non-road sources. On-road sources are the emission from traffic which is distributed all around the area, directly affect human health. It is also the main contributor for this source type due to the huge number of vehicles. Non-road sources are emission from harbors, airports, and railways.
The emission from mobile sources can be estimated based on the fuel consumption in the study area and the average emission factor, with Eq. (2). Vehicle categories to be considered are motorcycles, passenger cars, light duty vehicles, heavy duty vehicles, and busses/coaches in which fuel includes gasoline, diesel, liquid petroleum gas (LPG), and natural gas if available. This method is a top-down approach which is suitable for estimating emission in a huge area. However, top-down approach is high level of uncertainty.
where E: total emissions (include emission from each pollutant from each vehicle category and fuel type), FC i : fuel consumption (for each vehicle category and each fuel type), and EF: fuel consumption-specific emission factor (for each pollutant and vehicle category and fuel).
Air Emission Inventory http://dx.doi.org/10.5772/intechopen.79928 For a more detailed approach, subdivided emission into hot emission, cold-start emission, and extra emission as in Eq. (3) can be applied. Generally, the total emissions are the sum of hot emissions (the emissions when a vehicle is running, with the engine thermally stabilized), cold emissions (the emissions when a vehicle is running, during engine warm-up), and evaporative emissions (the emissions from fuel due to the difference between the temperature of engine and ambient temperature). The equation for total emissions is shown below: where E total : total emissions of pollutant i,   Emissions from tyre and brake wear also considered estimating since they contribute a significant amount of particulate matter (PM), especially those from motorcycles. These are considered as extra emissions and causes from friction between the tyre and the road surface when the vehicle is running. There are many calculation methods for this estimation since it is not complicated and requires simply data, for example, the estimation using Tier 1 emission factors from CORINAIR requires the data of the kilometers the vehicle which has annually traveled in the study area, as in Eq. (7). Emission factors can be found in emission factor handbooks and research papers such as CORINAIR, EMEP, AP-42, etc. EFs of pollutants depend not only on the generation processes but also on pollution control systems that have been applied. Most of the traffic emission models are intergraded emission factor database with clear reference, and it can also be modified by the user in accordance with the study area.
In general, for urban traffic emission estimation, an emission model is adopted since it can handle a lot of complicated data and consider different fleet compositions and driving conditions in different streets and integrated emission factor database. The advantages of emission models are the time saved, the availability of many estimating options, and the reduction of the calculating uncertainty. There are several emission models available, like COPERT, MOBILE-5, MOVES, IVE, MOBILEV, EMISENS, etc., but not all emission models are appropriate for calculating traffic emission in single street levels with different traffic situations. Most models were developed for calculations on national or regional level and use, therefore, more aggregated emission factors. An emission model used for an emission inventory on an urban scale with high spatial resolution should fulfill at least the following criteria: highly differentiated emission factor database, a broad selection of street types and traffic situations, an internal module to calculate cold start emissions, a representative and appropriate fleet composition by subsegments for each vehicle category, appropriate default fuel quality parameters, a module to calculate fuel consumption caused by air condition, access to default values in case adoptions are necessary, a user-friendly interface, a user-friendly data input tool, and a comprehensive description of the model.
Input data for emission models usually require the following information which is collected from surveys, previous researches, analysis and observation from the study area: defining main street network/defining road categories, length and width of road, number of lanes, speed limit for each road type/category, define vehicle type/category, average daily traffic, vehicle fleet/composition, vehicle registration database, and vehicle engine type, capacity, emission stage, fuel use, weight, and vehicle traffic curve.

Non-road sources
Non-road sources include gasoline and diesel powered vehicles, engines, and equipment used for construction, agriculture, transportation, recreation, and many other purposes. For example, non-road sources include marine vessels, aircrafts, railways, tractors, etc.

Emission from seaport
There are many different approaches to develop an emission inventory for ports. They vary significantly according to the availability of time, money, and effort. For example, a detailed inventory approach requires detailed data on vessels and land-based equipment characteristics and activities, as well as detailed information on port geography and ship paths within the port. This is the best practice for all ports; however, its applications are limited with available resources. Meanwhile, the streamlined inventory approach requires less input data. The methodology can be tailored to the amount of data available. In some cases, the emission inventory can be developed using extrapolation data.
To calculate air emissions in the port, a streamlined air emission inventory approach was used following US EPA guidelines [3].

Ocean going vessels (OGVs)
The available data for some ports in developing countries are not sufficient to develop a detailed inventory by calculating emissions from each vessel. The US EPA guidance [3] proposes a mid-tier approach for calculating air emissions in ports and allows us to calculate emissions from each category using a single representative vessel specification and operating pattern, rather than calculating emissions from every different vessel call. In this case, a detailed or mid-tier approach was selected based on the available data.

Cargo-handling equipment (CHE)
Emissions were calculated for each category using Eq. (10):

Emission from airport
Emissions from this type of sources are often overlooked since it is not a typical source. For estimating emission, the following data are required: • Fuel consumption for each flight phases (landing and taking-off (LTO), cruise phase…) • Emission factor for each flight phases and aircraft type Calculation steps for this approach are as the following: • Obtain the total amount of fuel sold for all aviation (in ktonnes).
• Obtain the total amount of fuel used for domestic aviation (in ktonnes).
• Calculate the amount of fuel used for international aviation by subtracting the domestic aviation (step 2) from the total fuel sold (step 1) (in ktonnes).
• Obtain the total number of landing and taking off carried out per aircraft type for domestic aviation.
• Calculate the fuel use for landing and taking off activities per aircraft type for domestic aviation.
• Calculate the total fuel use for domestic cruise by subtracting the total amount of fuel for LTO activities found in step 6 from the total in step 2.
• Estimate the emissions from domestic LTO activities per aircraft type. The number of landing and taking off for each aircraft type is multiplied by the emission factor related to the particular aircraft type and pollutant.
• Estimate the emission from domestic cruise activities. Use the domestic cruise fuel use and the corresponding emission factor for the most common aircraft type used for domestic cruise activities.
• Calculate the total emission for landing and taking off activities for domestic aviation.
• Calculate the total emission for cruise activities for domestic aviation.
• Do the same for calculating for international aviation.
Data on the aircraft types and the number of flights were collected directly from the airport and from the Department of Transportation (DOT).

Emission from railway
Emissions from railway are mainly from locomotives which generally are one of the three types: diesel, electric, or steam. Like other sources, emissions are the product of fuel combustion with CO, NOx, SO 2 , PM, hydrocarbons, and other heavy metals. According to EMEP/ EEA, the locomotives fall into three categories: shunting locomotives-used for shunting wagons whose power output is typically in the range of 200-2000 kW, railcars-used for short distance rail traction with power output about 150-1000 kW, and line haul locomotives-used for long distance rail traction both for freight and passengers with power output around 400-4000 kW.
The methodology to estimate the emission for railway is from the total fuel type used and related emission factors as in Eq. (12) as follows: where E i : emission of pollutant i for the period concerned in the inventory (kg or g), FC j. m : fuel consumption of fuel type m used by category j for the period and area considered (tonnes), EF i. j. m : emission factor of pollutant i for each unit of fuel type m used by category j (kg/tonnes), m: fuel type (diesel and gas oil), and j: locomotive category (shunting, rail-car, and line haul).

Overview
In this section, the process to apply the above theory for conducting an air emission inventory Can Tho City (CTC) is described. The emission inventory includes basic information about the dispersion of air pollutants, such as location, sources, and emission rates. The emission inventory assessment is necessary because it helps businesses implement effective policies to control their emissions, while it also helps the administration assess the current emissions and construct air quality management strategies in the future.
Can Tho is the fourth largest city in Vietnam and the largest city in the Mekong Delta. It is noted for its floating market, rice paper-making village, and picturesque rural canals. In 2013, in an area of 140,894.9 ha, the city had 1,232,260 inhabitants. Can Tho city is located on the south bank of the Hau River, a branch of the Mekong River. The city is located at N 9055′08″-10,019′38″; E 105013′38″-105,050′35″. It consists of five urban districts (Thot Not, O Mon, Binh Thuy, Ninh Kieu, and Cai Rang) and four suburban districts (Vinh Thanh, Thoi Lai, Co Do, and Phong Dien). Can Tho City was administratively formed on 1 st January, 2004, when the old province of Can Tho was divided into Hau Giang province and Can Tho City; Can Tho City then directly belongs to the Central Committee (City Profile, 2013).
In general, there are four main types of emission sources: point, area, and on-road, and nonroad mobile sources. Figure 2 shows the locations of emission sources in the study domain. Different symbols represent various kinds of point and area sources including gas stations, garages, construction sites, industrial factories, hospitals, hotels, etc.

Collecting data and calculating
EI was conducted for two districts: Ninh Kieu and Binh Thuy (NKBT) ( Table 1).

Total emission from area sources
The highest area emission source type for all pollutants in NKBT is burning fuel activities from households/residential activities ( Table 2).

Point source
Point sources in CTC also include emissions from industries, hospital incinerators, and hotel boilers. The total emission for this source is presented in Table 3.
Air pollution emissions that are NO X , SO 2 , CO, CO 2 , and VOC (NMVOC) from industrial activities are the largest among the three categories of point source emitters.

.1. Emissions from Hoang Dieu Port
We conducted a survey with an emission declaration table. For ocean going vessels: in general, anchorage areas for OGVs meet the standard, and the activity data are used in calculation including: name of OGVs, arrived day, leaving day, good categories (container, roll-on/ roll-off, package, etc.), gross register tonnage (GRT), dead weight tonnage (DWT), power, etc. Some information included: time and speed in reduced speed zone (RSZ) (transit) mode, in maneuver mode, etc.
For CHE, there are details about the number, type, capacity, and equipment life, allowing the preparation of a basic inventory.
Emissions from Hoang Dieu Port in 2014 are from 22 OGVs and 11 CHE (including cranes and forklift trucks) shown in Table 4.
As you can see, the total emissions are mainly generated from CHE, except for SO 2 from OGVs. This is the result of using fuel with a high content of sulfur. 73.3% of PM 10 , 74.4% of PM 2.5 , and 77.3% of CO 2 come from CHE.

Emissions from Can Tho airport
Can Tho International Airport (CTIA) is located in Binh Thuy district, and emissions from this source can affect people living in CTC. The numbers of flights from CTIA are shown in Table 5.
Emission factors for air plane are referred in the documents:  Emissions from airports can be generated through airplanes taking off and landing. Air emissions from this source are estimated by using the number of flights per year and emission factors related (kg/flight) for each type of airplane.

Total non-road mobile emissions
Emissions of CO, VOC, and CO 2 from the airport are much higher than the marine port within the non-road mobile emission source category ( industrial activities) for NO x and CO 2 . It was found that on-road traffic emitted a significant amount of pollutants into the atmosphere. Also, household activities, which were classified as area sources, played an important role [1].
Point sources are the biggest emitters of NO x and CO 2 ; area sources are the main emitters of SO 2 and PM 10 ; and on-road mobile sources are the main emitters of CO and VOC ( Table 7).

Conclusions
Emission inventory is a basic step of air quality management. There are several approaches of EI as described above, and based on each study area, considered using EI approaches consistent with the available data. High accuracy in EI is proportional to time consuming and costly and inversely. In general, less time and low cost are proportional with inaccuracy. And in general, an EI should be done annually or at least every 5 years to update the air emission status and to observe the trend of major air pollution sources and evaluate the effectiveness of air quality control.    Note: * Only emission from on-road mobile source is for PM, and in Table 7 considers it as PM 10 .