CHAPTER 3: Airport Air Quality Background

This chapter provides an overview of the issues and components related to understanding airport air quality contributions.

3.1 AIRPORT SOURCE CHARACTERISTICS

Airport emissions sources include those involving the combustion of fossil fuels and various evaporative sources, such as solvent usage and fuel evaporation. Emissions from mobile combustion equipment (e.g., aircraft) are generally considered the main emission sources at airports, but other sources also can contribute significant emissions as well. Emission characteristics depend on several factors that include, but are not limited to, the type of source (i.e., aircraft, GSE, vehicles, stationary combustion, or evaporative), equipment power setting, fuel type, and pollution control technologies implemented.

3.1.1 Source Types and Pollutants

Although other sources of emissions exist at airports, mobile sources are often the largest sources of emissions. Aircraft, as well as their auxiliary power units (APUs), ground access vehicles (GAVs), and GSE make up the bulk of emissions from mobile sources, although GSE can be both mobile and stationary. Stationary equipment sources include waste incinerators, boilers for producing heat and hot water, and power plants. When airports propose projects that require construction work (e.g., runway modifications, new terminal buildings, etc.), the emissions from construction equipment and associated activities must be accounted for as part of the project even though the emissions are temporary in nature.

Like combustion-related sources (i.e., emissions from vehicle and equipment exhaust), evaporative emission sources also must be considered. Maintenance activities, fuel storage operations, solvent usage, painting, and other activities can result in the release of volatilized compounds.

Table 3-1 provides a summary of the types of pollutants that potentially can be generated by the different sources at an airport. Although aircraft have been grouped separately, the types of pollutants emitted from aircraft are similar to those from other combustion sources. Table 3-1 also provides a subset of pollutants that tend to be of primary interest with regard to health concerns. This is an indication of the pollutants that are receiving the most research focus. Although the other criteria pollutants (including the secondary formation of ozone) continue to be a concern, the current focus of research studies is largely on emissions of PM (including UFPs), as well as HAPs.

including HAPs tend to be higher at low power conditions while NO_x emission factors tend to be higher at higher power settings (i.e., using fuel-based emission factors such as gram of pollutant per kg of fuel burned). For GSE, emission factors are generally based on power settings in horsepower (e.g., gram of pollutant per horsepower per hour of equipment usage. Emissions from aircraft and GSE are typically calculated using FAA’s Aviation Environmental Design Tool (AEDT, see https://aedt.faa.gov/). See Section 3.7 for more information on emissions modeling with AEDT. Modes for GAVs include transit, startup, and idling. Passenger cars and trucks and airport delivery and maintenance vehicles would mostly be operated in the transit mode, during which emissions are assessed based on miles traveled. Vehicles such as airport shuttle buses and taxis may spend significant time in idling mode, during which emissions are assessed based on the amount of time spent idling, though many airports limit the amount of idling time allowed. All of these vehicles would have a startup, with startup emissions assessed at the beginning of each trip. Emission or emission factors from GAVs are typically estimated using the latest version of EPAs Motor Vehicle Emission Simulator (MOVES, see https://www.epa.gov/moves) model. With this model, users can either obtain emissions or emission factors for each relevant operating mode for GAVs or, using model defaults or user-specific input based on the operation patterns of vehicles of interest, calculate emissions or emission factors that capture typical operation for the respective vehicle type accounting for all expected modes. Emission factors and emissions from GSE can also be estimated with MOVES.

In addition to mobile equipment, stationary source equipment, including power plants and incinerators, also have different modes of operation. However, emissions from such sources are typically assessed assuming constant, average emission factors.

3.1.3 Fuel Types

Several types of fuels are used at airports. Jet A is used by jet and turboprop engines while AvGas is used by piston-engine aircraft. Diesel has typically been used for GSE but electric equipment has increasingly been used for ground equipment, while gasoline and alternative fuels are used for some types of GSE. GAVs can be fueled by gasoline, diesel, electric, and compressed natural gas (CNG).

Jet A is denser and has a higher energy content than gasoline, but also results in greater carbon (e.g., CO₂) output on a per unit of energy basis. This does not, however, directly equate to differences in emissions of CO and hydrocarbons/VOCs because the emissions rates for those pollutants are also influenced by other factors, including combustion technologies and pollution controls. Aircraft typically have higher sulfur oxide (SO_x) emissions on an energy output basis than do motor vehicles due to the higher sulfur content in Jet A than in motor vehicle gasoline or diesel fuel.

Due to the continued use of lead in AvGas, GA airports have come under scrutiny for their lead contributions to local air quality. Historically, human exposures to lead have occurred through the use of lead in paints and automobile fuels (i.e., the use of tetraethyl lead in fuels to reduce engine “knocking”). Although these uses have largely been phased out, lead continues to be actively used in AvGas. Most GA aircraft with piston engines use AvGas. The EPA estimates that 70% of the lead entering the air annually is from piston-engine aircraft (see https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P1015S8N.pdf).

Diesel fuel and unleaded gasoline are the primary fuel types used to power GSE and GAVs. These fuels have different characteristics that contribute to different pollutant emissions, but the age, size, and use category of GAVs and GSE can make more of a difference to pollutant emissions

than fuel type. With an increase in electricity fueling GSE at airports as well as an increasing proportion of GAVs, emissions from these GSE and vehicles occur at the site of the plant producing the electricity (which is usually not on site at the airport) rather than at the airport, leading to a reduction in emissions at the airport. Many airports look to electrification as a means for obtaining potentially significant emission reductions at the airport and in the local communities and some airports have installed charging stations that support using electric GSE and GAVs. Airport buses and shuttles as well as GSE and GAVs may also use alternative fuels such as CNG.

3.1.4 Pollution Control Technologies

Pollution control technologies (or pollution controls, for short) typically refer to some device or equipment that helps to reduce pollutant emissions. In the case of GAVs, GSE, and aircraft, these technologies are typically included in the design of these mobile sources in order to meet the appropriate emission standards that apply to the source. Emissions reductions from aircraft engines are generally achieved through new combustor designs. In contrast, ground mobile equipment such as GSE and GAVs typically use catalytic material (i.e., as part of a catalytic converter in a gasoline engine) located in the exhaust system to convert pollutants such as CO and unburned hydrocarbons (including HAPs) to CO₂ and water. Controlling emissions from diesel engines is generally accomplished using aftertreatment technologies, which also often include the use of catalysts. Stationary sources (e.g., incinerators, power plants, etc.) also may use catalysts but they typically employ controls such as scrubbers and baghouses to convert or filter out pollutants depending on the size and design of the equipment/systems. These stationary source controls are used to meet emission limitations or operational procedures that are required in the operating permit for the emission source. Section 3.6 discusses other types of measures and actions that may be used to reduce emissions from airport sources.

3.2 SOURCE EMISSIONS CONTRIBUTIONS

Airports differ in many ways, including in activities, geography, and infrastructure. Nevertheless, it is generally accepted that the approximate rank of emissions source contributions (from highest to lowest emitters) is as follows:

• Aircraft in the landing and takeoff (LTO) phase;
• Road vehicles on airport landside roads and on the road network around the airport;
• Ground support equipment (GSE);
• Airport ground access vehicles (GAVs);
• Aircraft auxiliary power units (APUs);
• Airport heating and boiler plants;
• Evaporative losses (e.g., fuel storage, maintenance, etc.); and
• Airport fire training exercises (with a concern for per- and polyfluoroalkyl substances (PFAS) that might be released from certain firefighting foams).

It should be noted that this is a general rank and that it varies by pollutant type and airport. For example, depending on the extent of the roadway network that is included in an airport air quality study, road vehicle emissions could be significantly greater than emissions from aircraft.

The variation in these source contributions may be illustrated by considering the emissions generated by Airport XYZ (a fictitious airport) presented in Table 3-2. This example illustrates that many different airport sources contribute to local air quality, and that the relative contribution

is dependent upon the pollutant of interest. In this example, aircraft are the most significant source of emissions of NO_x and SO_x, but, for some pollutants such as CO, they can produce fewer emissions than GSE and roadway vehicles. In fact, the off-airport roadway emissions can be significantly higher than aircraft emissions depending on the roadway extent included in the analysis. The emissions contributions by source type for each airport are dependent on the airport and road network layout, equipment types, and operations of the airport and can be very different from this example inventory. However, in general, aircraft, GSE, and roadway vehicles tend to be the largest sources of on airport emissions.

3.3 AIRPORT OPERATIONS

Airport operations are the activities (e.g., usage) of a source such as aircraft, GSE, or boiler. In general, the greater the usage, the greater the magnitude of emissions. However, many factors influence the spatial and temporal impacts of the emissions on air quality. For example, the distribution and transport of pollutants at an airport are determined by the airport layout and the operations schedule. Airports usually have a schedule that reflects a “peak day” and “peak hour” of operations (e.g., Thursdays between 5:00 p.m. and 6:00 p.m.). Emissions from other mobile sources and GSE also would likely peak (i.e., maximized usage occurs) around the same time. Wind direction determines both which runway is used, and which area is downwind of the airport at a given time.

Airport emissions tend to decrease with improvements in the design and/or efficiency of sources and to increase with increased usage due to airport growth. This is the challenge for airport operators (and more broadly the aviation industry): that increases in overall emissions due to airport growth tend to exceed emissions reductions due to operational or technological improvements. An emissions inventory, usually completed on an annual basis, is used to track the amount of emissions from each source category over time, taking into consideration both airport growth and operational improvements. In contrast, air quality assessments must be performed with more detailed information taking into account appropriate temporal conditions (e.g., time of day, concentration averaging periods, etc.) to properly determine pollutant concentrations that can be compared to health benchmarks (e.g., NAAQS). Those assessments also take into account other factors, including meteorology and spatial information (i.e., source and receptor locations and geography). All of the factors must be taken into account accurately when assessing air quality impacts and trends.

3.4 GEOGRAPHY

Physical geography can play a significant role in both airport operations and local air pollutant dispersion. Mountain ranges not only require a specific aircraft approach procedure but can define localized weather patterns and channel air sheds to form distinct wind patterns. The emissions from airports in valleys would not tend to disperse as rapidly as emissions at airports in open terrain that experience no major geographical hindrance to dispersion.

For example, Los Angeles and Los Angeles International Airport (LAX) sit in a bowl ringed by mountains to the north and east that trap pollutants in an urban basin such that in warm weather, a cool sea breeze is drawn onshore at ground level creating a temperature inversion that prevents pollutants from dispersing. This can result in the formation of photochemical smog. Similarly, Mexico City’s MEX Airport is situated at over 7,000 feet above mean sea level in a basin constrained by mountains with intense solar radiation; these characteristics combine to cause air quality problems involving both primary and secondary pollutants. Even with relatively flat terrain, changes in land use (e.g., urban sprawl) also may appreciably affect the surrounding meteorology because changes in the local surface energy budget (e.g., urban heat island effect) can impact diurnal air temperatures and wind patterns, thereby affecting the dispersion of pollutants.

3.5 METEOROLOGY

The daily and seasonal meteorological components that affect local concentrations of pollutants include wind direction, wind speed, mixing depth, ambient temperature, relative humidity, and solar insolation (i.e., solar energy received on a surface). Prevailing wind direction and other meteorological conditions are particularly important in the dispersion of the air pollutants emitted by airport sources. Below the mixing height (nominally about 3,000 feet above ground level), dispersion occurs based on the turbulent strength of the atmosphere (largely defined by the diurnal heating and cooling cycle) and mean wind characteristics.

Winds are of particular significance in that they determine the direction in which airport emissions will move and the area over which they will disperse. Wind patterns often demonstrate correlations with seasonality—for example, wind direction may be predominantly northwest in the winter and predominantly southwest in the summer, as is the case at New York’s John F. Kennedy International Airport (JFK). Similarly, predominant wind speeds may show seasonal trends. Periods of very low or nil wind may cause stagnation near the point of emission, leading to localized pollution episodes (increased concentrations). Varying wind patterns also arise due to the interaction of air flows with local topography and differential heating effects. The local wind patterns are considered in the choice of runway orientations at any given airport.

3.6 MITIGATION MEASURES FOR AIRPORT SOURCE EMISSIONS

Typically, it is the airport operator that leads the preparation and delivery of an airport air quality management plan, which includes a measurement program, air quality assessments, and various mitigation activities. However, many emission sources at an airport, including the two most significant sources—aircraft and access road traffic (as well as GSE in many cases)—are not within the direct control of the airport operator. Therefore, any airport mitigation plan needs to be developed in collaboration with airport tenants in order to properly account for all potential sources of emissions and reductions.

A range of mitigation options is available at airports to reduce local air quality pollutants. Mitigation options are typically categorized by emissions source type. However, mitigation options also can be considered according to the type of measure that is being implemented (see

Table 3-3). Note that not all of these measures are available to airport operators as they do not have direct control over some of these sources.

Table 3-3. Categorization of air pollutant emissions mitigation measures.

In the United States, the FAA runs the Voluntary Airport Low Emission (VALE) Program. Any airport in a nonattainment area is eligible to take part in the program, which provides airport operators with a legal mechanism to raise funds through their Passenger Facility Charge (PFC) and provides funding for the financing of certain air quality pollutant mitigation initiatives (Airport Improvement Program funds) such as low emission vehicles, refueling and recharging facilities, and gate electrification. The FAA also created the Zero-Emissions Airport Vehicles and Infrastructure Pilot Program in 2012, which provides funds for the purchasing of zero-emissions vehicles and the supporting infrastructure at airports.

In addition to the nonattainment status of an area, airport emissions reduction programs may be triggered by findings of significant impacts of non-criteria pollutants (e.g., HAPs). Airports may be incentivized to participate in existing voluntary programs or to initiate their own programs as a result of increased public interest in better understanding airport contributions to local air quality and increased scrutiny from the public on health concerns.

3.7 AIRPORT EMISSIONS AND DISPERSION MODELING CAPABILITIES

To assess potential health impacts from airports, impacts of pollutant emissions on local and regional air quality must be quantified. Emissions and atmospheric dispersion models are used to supplement measurements of emissions and air quality, which can be costly and may not be representative (e.g., for certain locations or time periods). The following sections provide overviews of the current state-of-the-art capabilities in these areas, as well as their limitations.

3.7.1 Emissions Modeling

The first step in any air quality modeling work is to quantify emissions. Airport source emission modeling is similar to that done for other industries since many of the sources are the same (e.g., GAVs are the same sources as those found on highways and boilers/incinerators are similar to those found in industrial applications). For modeling emissions, there are two key categories of data:

• Emission factors and
• Activity information.

Emission factors are generally in the form of mass amount of a pollutant emitted per some activity or period of time. For example, grams per mile and grams per second are common units for an emission factor. These factors are specific to each pollutant and vary according to the characteristics of a source including but not limited to the following:

• Type of equipment,
• Age of equipment,
• Emissions control technology,
• Fuel type, and
• Power setting.

Although some emission factors may be static (e.g., available in a data table), others may need to be developed using models that account for these characteristics. Once an emission factor is available, it can be applied (e.g., multiplied) to the corresponding activity data to calculate emissions. The activity data represents some measure of use or operation of the source (e.g., hours of usage).

Currently, the state-of-the-art emissions modeling capability for airports is represented by the FAA’s modeling tool known as AEDT. AEDT was designed to replace and expand the capabilities of FAA’s Emissions and Dispersion Modeling System (EDMS), and older reports sometimes refer to both models together as “EDMS/AEDT.” At this writing, the most recent version of AEDT, version 3f, was released on December 15, 2023. The sources modeled in AEDT are categorized as follows:

• Aircraft (including engine testing),
• Auxiliary power units (APUs),
• Ground support equipment (GSE), and
• Stationary sources (Boilers/space heaters, deicing areas, emergency generators, fuel tanks, incinerators, sand/salt piles, solvent degreasers, surface coating/painting, and training fires.)

The underlying data sets in AEDT were obtained from various sources and are generally considered the best publicly available emission factors and activity information on a national level (i.e., for general use at all U.S. airports). However, it is recommended that specific equipment and activity information be obtained for each airport to be modeled whenever possible to improve the accuracy of emissions inventories. It should also be noted that although EDMS modeled emission factors for GAVs, AEDT3f does not; instead, GAV emissions must be modeled separately using the EPA’s MOVES model. The resulting MOVES-based emissions representing roadway, parking, and construction emissions can then be imported into AEDT for inclusion in the emissions dispersion modeling performed by AEDT.

Although AEDT is considered state of the art, there are still various areas for improvement, some of which are currently under research (e.g., through ACRP, FAA, etc.). Users need to be mindful that uncertainties exist with the underlying modeling data and methods. To a certain extent, these uncertainties can be decreased by collecting airport-specific activity information (e.g., aircraft operations, aircraft fleet information, GSE hours of usage, etc.). With the conservative nature of the model, a common tactic has been to model worst (or near-worst) cases and compare the resulting emissions inventories to regulatory limits such as the General Conformity de minimis levels. As such, if the worst case produces lower results than regulatory limits, then a more accurately modeled scenario would also be below the limits. This tactic can serve as both a screening approach as well as (in some cases) a means of allaying concerns over worst-case scenarios.

Emissions from GAVs operating at an airport or in the surrounding area are typically modeled using EPA’s MOVES model. The current version of this model, MOVES4, released in 2023, can be executed at varying levels of detail to characterize the vehicle population, fuels used, meteorology, and vehicle operation. At the simplest level. a user can model a default vehicle fleet for the default fuel and meteorological conditions of the county where the airport is located, or the user can provide specific information on the makeup of the vehicle fleet by model year, vehicle type, and operating characteristics (e.g., speed profile) as well as user input fuel characteristics and meteorological conditions. Often, some combination of the MOVES defaults and user input would be used. The model can also be used to either calculate emission factors (such as g/mile for vehicle travel, g/hour for idling, g/vehicle start for startups) or emissions, depending on options selected by the user. The set of pollutants that can be modeled with MOVES is extensive, including all criteria pollutants, HAPs, greenhouse gases, and others.

3.7.2 Dispersion Modeling

As the name implies, dispersion modeling refers to the process of predicting the dispersion of pollutants in the atmosphere once they have been released from a source. There are different scales of assessments—for airports, local-scale (e.g., within a local community) and larger, regional scales may apply. The larger the scale (and, generally, the more time involved for dispersion), the greater the dispersion generally resulting in lower concentrations experienced by the public for directly released pollutants. However, in each scale, secondarily formed pollutants (e.g., through atmospheric chemistry) also can impact local populations. Ozone and PM species are examples of such secondary pollutants.

Much of the local-scale modeling is conducted through the use of Gaussian models. The EPA’s AERMOD modeling system (see https://www.epa.gov/scram/air-quality-dispersion-modeling-preferred-and-recommended-models) is based on a Gaussian methodology and is the regulatory workhorse model used for most local air quality assessments. AERMOD represents the state of

the art in the current scientific understanding of the dispersive nature of the atmosphere. In contrast, regional-scale modeling requires the use of grid-based photochemical models such as the EPA’s Community Multiscale Air Quality (CMAQ, see https://www.epa.gov/cmaq) modeling system. Grid models are useful since they can appropriately model atmospheric chemistry and the influence of background concentrations, whereas Gaussian models are limited in that regard. Some chemistry such as that involving nitrogen dioxide (NO₂) can be modeled through simplified methods in Gaussian models, but ozone and secondary PM formation require photochemical grid models.

Most regulatory studies at airports (e.g., NEPA-related studies) have only required the development of emissions inventories. However, dispersion modeling is necessary to better understand potential health impacts since emissions inventories do not provide a direct correlation between emission sources and pollutant concentrations experienced by the public.

Because of the additional factors affecting dispersion, predicted concentrations can have significantly greater uncertainties than emissions inventories. Concentrations are highly dependent on meteorology and the spatial relationship between sources (e.g., aircraft) and receptors (i.e., population). Any uncertainties in these factors—as well as various others such as the surrounding geography, seasonality, source activities, etc.—can drastically affect modeled concentrations. Also, it should be noted that dispersion modeling is only as accurate as the modeled emissions will allow. That is, any uncertainties in the emissions will carry through to the concentrations. Airport air quality studies illustrate the challenges of accurately predicting pollutant concentrations arising from airport emissions. Model users need to understand the potential limitations and uncertainties of these dispersion modeling processes.

It also should be noted that although alternative models exist, AERMOD is a static model generally used to predict concentrations by hour (i.e., average concentration for each hour). So, although AERMOD can provide hour-by-hour concentrations, it is considered a static model due to its Gaussian plume methodology. Finer time-varying models such as those employing Gaussian puffs rather than plumes may allow better correlations of source activities with population exposures, although the importance of this modeling refinement would depend on pollutant and health outcomes (i.e., whether short-term or long-term exposure is under consideration).

3.8 AIR QUALITY MEASUREMENT CAPABILITIES

From the literature review conducted for this reporting, dispersion modeling is used more frequently to characterize air quality contributions from airports than measurements (monitoring). This is in large part due to the costs and resources required to conduct measurements often resulting in limitations on the number of measurement sites and samples that can be supported. Although measurements have further drawbacks of not being source-specific (difficult to assess contributions from specific sources) and have uncertainties in the monitoring equipment/methods and influences from various other factors (e.g., meteorology) that may cause difficulties in obtaining good samples, measurements are generally considered to provide the best information because they represent real-world values.

Additionally, there have been developments of relatively low-cost AQ sensors available on the market. These sensors include indoor and ambient monitoring capability in real time. Many AQ sensors are less than $300, making them accessible to the public. However, low-cost AQ sensors are less reliable and have more instrument noise than established models, especially at fine PM sizes.

Uncertainties in measurements can vary depending on the types of equipment employed. For example, readings from continuous gas analyzers tend to be more accurate than air samples (gaseous or PM) collected and analyzed over an averaging period (e.g., 1 hour, 24 hours, etc.). Although uncertainties exist, if proper measurement protocols are followed, measured concentrations will tend to be more accurate than modeled results, which can involve greater degrees of errors.

A compromise that includes both measurements and modeling is possible, and often times, a preferred approach. For example, limited monitoring can be used to help establish background concentrations and measured data can be used to help validate modeled values. Also, measured meteorological data could be used to support more accurate modeling. Modeling can be used to provide greater spatial coverage and cover greater time periods to establish temporal trends. As stated in EPA’s Guideline on Air Quality Models, 40 CFR Part 51, Appendix W “Air quality measurements can be used in a complementary manner to air quality models, with due regard for the strengths and weaknesses of both analysis techniques, and are particularly useful in assessing the accuracy of model estimates.” When looking to verify model performance, Appendix W states “There are no specific levels of any model performance metric that indicate “acceptable” model performance. The EPA’s preferred approach for providing context about model performance is to compare model performance metrics with similar contemporary applications.”

Generally, measurement methods and equipment are related either to regulatory needs or research at airports. The promulgation of reference and equivalent measurement methods for specific pollutants also prescribes the type of equipment used. Table 3-4 provides a high-level overview of the most common types of measurement equipment by pollutant.

Therefore, only the emissions occurring below this mixing height are included in an airport air quality study.

Although aircraft generally continue climbing well above 3,000 feet, their flight segments above this height are defined as part of the overall cruise mode. Cruise emissions are typically excluded in airport air quality studies because they occur above the mixing height and are considered to have negligible effects on local air quality. In addition, there is no defined, standard power setting for cruise but there are power settings for the LTO modes; and there are no defined emission factors for cruise. However, cruise emissions have the potential for secondary effects on larger scales (e.g., regional, national, and global). These effects may include acid deposition, ozone formation, and secondary PM, and may have detrimental effects to human populations at significant distances from the airport.