Estimating Emissions from Sources of Air Pollution

6.1 Development of an Emissions Inventory

6.1.1 Overview Introduction

An emissions inventory is a compilation of emissions information related to the sources of one or more air quality problems in a location of interest that normally resides in some form of database structure. It should be noted that especially with respect to global warming gases, that emission sinks (i.e. negative emission sources) can also be an important consideration as well. Emission inventories are a mandatory component for the development of an effective air quality management process as discussed in Chapter 3. They are typically used to support the analysis of the air quality impacts of sources, to support trend analysis for air pollution reduction programs, and to support policy and regulatory analyses of air quality management efforts. While the bottom line for an emissions inventory is, of course, the quantification of emissions into the atmosphere, the intended uses of the inventory, such as air quality modeling and regulatory analysis, require the inclusion of additional information in the inventory other than just the emissions information. Thus, the design and development of an inventory is critical to producing an effective air quality management process. The Intergovernmental Panel on Climate Change has produced a series of documents on developing emission inventories for greenhouse gases ( The principles outlined in their “General Guidance and Reporting” document are applicable to most other air quality problems and provide a good deal of insight into the emissions inventory process. It is recommended as a good source of information on emission inventory development.

In general, a good emissions inventory meets the following guidelines. These guidelines were adapted from, “2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 1” (

Transparency: There is sufficient and clear documentation such that individuals or groups other than the inventory compilers can understand how the inventory was compiled.

Completeness: Estimates are reported for all relevant categories of sources and sinks, and gases.

Consistency: Estimates for different inventory years, gases, and categories are made in such a way that differences in the results reflect real differences in emissions.

Comparability: The inventory is reported in a way that allows it to be compared with other similar inventories.

Accuracy: The inventory contains neither over- nor under-estimates so far as can be judged.

A properly developed emissions inventory will: identify the sources of emissions that are creating the air pollution problems of interest, will allow the projection of the impacts of emission control scenarios, will support air quality modeling where needed, and will facilitate policy and regulatory analysis. To meet these conditions, it is best to use some form of large linked database format; however, useable emission inventories have been developed using Excel spreadsheets or similar database structures.

It is noteworthy that air quality problems viewed as a single issue often involve the emission of several pollutants. For example, global warming relates to the emission of four major pollutant groups and several more minor emission groups. Ozone problems are created by the presence of hydrocarbons, nitrogen oxides, and carbon monoxide in the atmosphere at the same time, and particulate problems relate to the direct emission of particulate matter, but also result from the emission of nitrogen oxides, sulfur oxides, and hydrocarbons. Thus, the emission inventories must include all of the important pollutants that relate to a problem in order to properly address the problem. Geographic Boundaries of an Inventory

As noted above, all emissions that contribute to a particular air quality problem should be included in the inventory. The geographic extent of the inventory is an equally important consideration as well. In the case of carbon monoxide, the sources tend to be local, and it is rare that an inventory needs to extend more than eight kilometers from the locations of concern for carbon monoxide. On the other hand, ozone and much of the mass of particulates are formed in the atmosphere over time and take many kilometers downwind to reach their maximum level. Thus, the contributors to an ozone or particulate problems can be one eighty kilometers or more upwind of the problem. The acid rain related inventories in the United States require inclusion of all of the United States east of the Mississippi river. Some sources of the acid rain problem are two thousand kilometers from the locations of greatest impact.

In the case of global warming, all global warming related gases going into the atmosphere contribute to the problem regardless of the location. The decision on what geographic range to include in an inventory for global warming becomes more of a political decision as opposed to a scientific decision. The national territory is a common boundary that is selected. However, in cases where a locality such as a city or county wants to address its contribution to the problem then the political boundaries of that locality are typically chosen.

There is no easy way to determine the geographic boundary for an emission inventory. Air quality modeling is the best tool to use to estimate if a source will have a meaningful contribution to a problem. However, in some cases, it is not practical to do modeling in advance of the development of an emissions inventory. In this case it is best to err on the side of including too much as compared to including too little. There can be obvious geographic features that suggest an inventory boundary. For example, Los Angeles in the United States is surrounded on three sides by mountains and the ocean on the forth side. The mountains are a logical choice for defining the geographic extent of the Los Angeles ozone and particulate inventory. Basic Inventory Design

In order to support air quality modeling, the emissions inventory should include information on the types and amounts of pollutants being emitted, the times of day and year that the emissions are released, the location of the emission point(s), the velocity(s) and temperature(s) of pollutant releases, and the height(s) of pollutant releases. To support trend analysis, the same information is needed as described for air quality modeling, but it must be compiled for different timeframes, typically for months or years. To support policy and regulatory analysis, data relating the emissions to energy use and energy source, product flow to and from other businesses, job relationships, and control costs are very useful components of the database. Table 6.1.1-1 summarizes the data needs depending upon the inventory use.
6.1.1-1 Summary of Data Needs for Various Inventory Uses

Clearly, all of the desired data is not always available for input into an inventory. However, it is valuable in conceptualizing the design of an inventory database at the beginning of the emission inventory development process and to make provisions for the easy addition of information to the inventory database as the air quality management program becomes more sophisticated and more data is obtained.

Inventory timeframes must be considered in the design of an emissions inventory. Ozone formed in the troposphere is produced by complex atmospheric interactions between certain common pollutants in the presence of sunlight. Warm temperatures increase the production of ozone. Thus, almost all ozone problems occur during the summer months when there is more sunlight and higher temperatures. On the other hand, the atmosphere is normally the most stagnant during the winter months allowing pollutants to accumulate to higher concentrations. Thus, particulate levels, which include a significant amount of directly emitted particulate matter, are typically more severe during the winter months. Emission rates also change based on the time of year. For example, VOC evaporative emissions are greater during the summer when temperatures are higher. Thus, it is common to maintain an emissions inventory that relates to the summer months and a second emission inventory that relates to the winter months. Of course, in the case of global air pollution problems such as global warming and stratospheric ozone depletion where emissions stay in the atmosphere for years, the time of year of release is unimportant. In this case, an annual inventory is all that is needed. In deciding if an inventory should be divided by the time of day, or day of the week, or by season of the year, or maintained by the total year, it is important to recall that it is possible to take an emission inventory that is divided by hours or other small timeframe and compute emissions for larger timeframes, but it is often not possible to convert from a long time-frame inventory to a short time-frame inventory. Thus, the timeframe that an inventory is defined for is an important consideration depending upon the pollutants of interest.

For the case of Ozone non-attainment problems and PM2.5 related non-attainment problems, an hourly inventory is the most useful. This is because the Ozone and PM2.5 formation rates vary significantly from hour to hour. For the case of particulate modeling in the troposphere, particulate matter that is formed from compounds in the atmosphere is often an important component of the problem. These particles are referred to as secondary particulate matter in that these particulates are not directly emitted but formed in a secondary process in the atmosphere. Directly emitted particulate matter is referred to as primary particulates. The rate of formation of particulate matter and ozone varies by hour throughout the day due to changing atmospheric and emission conditions and by location depending upon the relative amounts of ammonia, hydrocarbons, and nitrogen oxides. Thus, an inventory designed for use in addressing ozone and particulate matter problems needs to be set up hour-by-hour and location-by-location if at all possible.

The location where emissions occur in a region is easy to define in the case of smoke stacks and other similar point sources. In this case, UTM coordinates or latitude and longitude are used to identify the location of the source and this information is recorded in the emissions inventory database in association with the emissions information. For the case of fugitive emissions, area sources, and mobile sources the emissions are distributed over the region of interest. This distribution is normally not consistent but varies from location to location in the region. In order to allow for complex modeling of a region, the region is typically divided into a grid. The number of grids in a region can vary. Smaller grids support detailed modeling analysis but require considerably more data in order to determine the emissions associated with each grid. Common grid sizes are five, ten, and fifty kilometer squares. The smaller grids are preferable to support complex modeling efforts, and it is possible to place the same emission numbers in adjacent grids if adequate data is initially unavailable to divide the information into the smaller grid sizes. Emission Estimation Methods

The methods used to estimate emissions can vary widely. In some cases, specific measurements are made of emission sources under the full range of conditions that the source operates. This method is typically applied to large stationary sources such as an electric power plant. In some cases, large emission sources are equipped with full time emission monitoring equipment, which allows the accurate assessment of emission rates by the time of day or time of year.

A second and more common method is the use of emission factors. In this case, emissions are estimated based on some predetermined process factor such as the emission rate per unit heat input or per hour of operation or per volume of material used. The use of emission factors is based on the assumption that the emission rate of a source is linearly dependent upon some input process factor. This linearity assumption is valid in the case of sulfur emissions based on the amount of sulfur in a fuel being used. However, in the case of the emission of nitrogen oxide, which is generally formed from nitrogen in the air being fed to a combustion process, the emission rate may not be linear with the heat input rate since other factors influence the nitrogen oxide formation rate. The same can hold true for the formation of carbon monoxide and volatile organic compounds. Thus, care must be taken in the use of nitrogen oxide, carbon monoxide, and volatile organic compound emission factors as well as others. Still, the use of emission factors is the most common approach to estimating emissions from many sources. Clearly, this approach depends upon the availability of reliable emission factors that are applicable to the location of interest.

There are a number of sources of emission factors. Two common sources are the U.S. EPA emission inventory database and the IPCC emission Factor database. These can be found online at

and at

Finally, sophisticated emission models are often used to estimate emissions from complex sources. This approach is common in the case of mobile sources which can be the major contributor to an air quality problem. Mobile sources are typically too numerous to be individually measured and emissions vary widely depending upon operational and maintenance considerations and the type of mobile source. Emission models, which are discussed in subsequent sections, are available for on-road and off-road mobile sources.