Dispersion Modeling Guidance for Airports Addressing Local Air Quality Health Concerns (2017)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

6This chapter describes how airport emissions disperse, mix with background pollutants and react and/or transform in the atmosphere. It also describes the regulatory framework important to understanding the relative significance of different sources and pollutants. It also discusses the reasons for conducting dispersion modeling, alternative modeling approaches, and when, where, and why different approaches are appropriate. 2.1 Overview of Airport Air Quality Modeling Airport air quality modeling is used to determine the impact of airport emissions on people and the environment. Models are numerical approximations of physical phenomena, and spe- cific air quality models approximate the physical and chemical processes that occur in the atmo- sphere. Given that these are approximations, uncertainties exist both in the inputs and outputs of a model. Accordingly, airport operators should exercise caution in the application and use of models. However, air quality models have played a significant role in understanding the sources of air pollution, and in developing alternate emissions scenarios to reduce air pollution. The air quality modeling process begins with quantifying the mass (e.g., lbs. or kg) of pol- lutant emissions from all sources at the airport to produce an airport emissions inventory. An inventory is useful to compare the emissions from these sources to better understand the relative contributions of each source, and may be an essential element of regulatory reports, planning studies, or sustainability programs. Major emissions sources at airports are aircraft engines, auxiliary power units, ground sup- port equipment, and ground access vehicles. These sources are generally described as mobile sources (e.g., aircraft and passenger automobiles, which involve discrete vehicles emitting pol- lutants); area sources (e.g., painting booth or parking garages, in which many sources emit pol- lutants over a wide area); or point sources (e.g., emergency generators, boilers used to heat water for terminal heating, and other stationary equipment). Quantifying emissions from these many sources over a common time period is the first step in creating an inventory. To understand the actual impact of these emissions, it is necessary to determine the pollutant concentration (mass per unit volume, which may be measured in ppm or mg/m3) at the point where exposure takes place. EPA has set limits on exposure to six common pollutants to protect public health and envi- ronmental welfare against the effects of outdoor air pollution, and these limits are referred to as the National Ambient Air Quality Standards (NAAQS). These pollutants, generally referred to as criteria pollutants, are carbon monoxide (CO), lead (Pb), nitrogen dioxide (NO2), ozone (O3), particulate matter of a size less than 2.5 microns (PM2.5) and of a size less than 10 microns (PM10), C h a p t e r 2 Primer on Airport Air Quality and Dispersion Modeling

primer on airport air Quality and Dispersion Modeling 7 and sulfur dioxide (SO2). It is also important to know the average concentration of criteria pol- lutants over a specified averaging time period, because that is how US air quality regulations are defined. The primary limits are health-based standards geared toward protecting people who are sensitive or at-risk, including asthmatics, children, and elderly people. The secondary limits are designed to prevent impacts to animals, vegetation, and physical structures, and to prevent reduced visibility. Table 2 summarizes the limits and averaging times for the six common pollut- ants. The researchers also refer to EPA’s Green Book, which has extensive information (including tables and maps) on individual areas of the country that are in non-attainment areas for each criteria pollutant (see https://www.epa.gov/green-book). As of September 2016, there were 130 US airports located within areas designated as being in non-attainment or maintenance of the NAAQS for one or more criteria pollutants (see https://www.faa.gov/airports/environmental/ vale/ for current and historical attainment data). Calculating concentrations is very complex and requires computer models to quantify and track the movement of pollutants from the emission source as they spread out due to weather, local terrain, and mixing. Atmospheric motion determines the overall speed and direction with which emissions travel. Atmospheric motion is primarily responsible for the mixing, or disper- sion, that takes place within the ambient atmosphere, creating a plume of pollution. An important factor in how emissions disperse has to do with plume dynamics, or the physi- cal condition of the emissions plume. The temperature of the plume affects its buoyancy. High temperature emissions will rise once released into the surrounding air as a result of the tempera- ture difference. The greater the temperature difference, the greater the buoyancy. Another factor in how emissions disperse is the plume velocity. High velocity emissions lead to shear with the local wind and to turbulence, which causes entrainment of and mixing with the surrounding air. Low velocity emissions will have much less shear and lower entrainment. The direction of the plume is determined by the local wind direction, while mixing is related to small-scale effects like turbulence. Likewise, terrain characteristics and local building structures can affect local pollutant concentrations by affecting wind patterns and generating turbulence. Other emission- specific processes also may have an effect, such as dry and wet deposition. All of these physical processes affect atmospheric dispersion and lead to a three-dimensional, time-dependent con- centration distribution of the pollutants. Adding further complexity, many pollutants undergo chemical reaction and transform in the atmosphere. Understanding the transformation of pollutants by chemical reaction is essential for determining the health impacts of emissions. Also, the pollutants released at the airport (i.e., Pollutant Primary/Secondary Limit Averaging Time Level Carbon Monoxide (CO) Primary 8 hours 9 ppm Primary 1 hour 35 ppm Lead (Pb) Primary & Secondary Rolling 3-month average 0.15 μg/m3 Nitrogen Dioxide (NO2) Primary 1 hour 100 ppb Primary & Secondary Annual 53 ppb Ozone (O3) Primary & Secondary 8 hours 0.070 ppm Particulate Matter (PM2.5) Primary Annual 12.0 μg/m 3 Secondary Annual 15.0 μg/m3 Primary & Secondary 24 hours 35.0 μg/m3 (PM10) Primary & Secondary 24 hours 150 μg/m 3 Sulfur Dioxide (SO2) Primary 1 hour 75 ppb Secondary 3 hours 0.5 ppm Source: https://www.epa.gov/criteria-air-pollutants/naaqs-table (accessed December 15, 2016) Table 2. Pollutants, averaging times, and levels for primary and secondary standards.

8 Dispersion Modeling Guidance for airports addressing Local air Quality health Concerns source emissions) mix with the pollutants that are already in the atmosphere (i.e., background emissions) forming a more complex mixture. To truly understand the health and environmental impacts of airport emissions, it is necessary to understand and track all of these factors. Air dispersion models are used to track the movement and transformation of pollutants over time in the atmosphere. They are composed of a sequence of mathematical equations that require information about the physical setting, emissions sources and their temperature, veloc- ity, direction and spatial location, background pollutant concentrations, weather conditions, surface characteristics of the airport, and locations of receptors (specified points at which mea- surements are taken to gauge the exposure to pollutants of the people and environments at those locations). Even though air dispersion models are simplified representations of reality, they are very complex and are based on the current best scientific understanding of the factors that influence the movement and transformation of pollutants in the atmosphere. The various air dispersion mod- els differ in the basic dispersion assumptions they make, as well as how they represent emission sources. Rather than representing an aircraft taking off as continuously rising, for example, an air dispersion model might represent the aircraft as a series of area sources at increasing height, which allows for simplified computation. Another model might represent the aircraft as a series of emissions puffs (non-continuous distributions of concentrations) occurring at increasing height along the take-off path. After quantifying emissions from the specified sources over a common time period, the mod- els compute the impact of local meteorology, or how weather and other atmospheric conditions cause the emissions to migrate or disperse into the environment. Wind speed and direction, ambient temperature stratification, and surface heating and cooling are the most significant meteorological factors that cause emissions to disperse. Local geographical features can disrupt the effects of meteorology, and these are accounted for in dispersion models. These features could include airport buildings and other structures on the airfield, as well as nearby hills, which make up complex terrain. Downwash (the effect of the turbulent wake in the lee of a building) is a term used to represent the potential effects of a building on the dispersion of emissions from a source. Downwash is considered for sources characterized as point, line, or area sources. The height and proximity of a point source to a structure can be used to determine the significance of downwash. It bears repeating that to evaluate health-related impacts, it is important to consider both the pollutants in the source emissions and emissions chemistry; that is, how the pollutants chemi- cally transform once they have been emitted. Current research indicates that, with regard to airport emissions, the human health effects of PM and NOx are generally the most significant, given that high NOx concentrations can lead to high O3 concentrations, and O3 is an important secondary pollutant that affects human health. At most airports, aircraft are the largest source of NOx emissions. Aircraft-generated pollutants generally transform in three different zones: 1. Immediately after exiting the combustor within the engine, 2. Downstream from the engine in the hot exhaust plume, and 3. After emissions have cooled and mixed with the ambient atmosphere. At the aircraft engine exit, hot combustion gases mix with ambient air to quickly cool the gas stream. Some gases, like heavy hydrocarbons, can condense under these conditions to form aerosol particles. In the exhaust plume, as emissions continue to cool, some molecules undergo chemical reactions and produce other molecules that can also condense into particles. Similarly, gaseous and particle emissions from cars, trucks and ground vehicles that have exhaust pipes, catalytic converters or particle traps, and mufflers will transform in the exhaust

primer on airport air Quality and Dispersion Modeling 9 plume after mixing with the ambient atmosphere. Most of the aviation-related PM that reaches airport communities comprises particles released during ground operations, landings, and take- offs. Models vary in the level of detail with which they treat atmospheric chemistry. Once the air quality models have computed an emissions inventory and evaluated the effects of meteorology, plume dynamics, and emissions chemistry, they determine the pollutant con- centration at defined receptor sites, including locations of expected maximum concentration, locations where employees are present, and locations where the general public is commonly present. The results show the degree to which airport employees, passengers, citizens living nearby, and the local community are subject to airport emissions impacts. EDMS, developed by FAA and required for air quality analyses of aviation sources, is EPA’s preferred airport air quality model. EDMS was recently incorporated into FAA’s AEDT, which integrates noise and emissions models and helps assess their interdependencies. The core com- ponents of EDMS prepare airport emissions inventories. Dispersion capabilities are added to process the EDMS inventory and determine pollutant concentrations at specified receptor sites. The standard dispersion model preferred by EPA for use with EDMS is AERMOD. Other dispersion models also can be used with EDMS, depending on the needs of the particular study. The four dispersion models emphasized in this guidebook are: • AERMOD: AERMOD is a steady-state Gaussian plume dispersion model that was developed and is maintained by EPA. The term steady state means that the local meteorological con- ditions are not changing with time and approximate the flow field. Gaussian refers to the shape of the concentration profile within the plume (specifically that it reflects a Gaussian or “normal” distribution). This model incorporates air dispersion based on the boundary layer turbulence structure and scaling concepts and includes treatment of both surface and elevated sources, as well as both simple and complex terrain. It is non-proprietary and is EPA’s preferred regulatory dispersion model for near field (< 50 km) applications. It predicts the dispersion of both primary gas and aerosol emissions and includes chemistry for the conver- sion of NOx to NO2 and decay of SO2, dry and wet deposition, plume buoyancy, and complex terrain. EDMS accounts for emissions from aircraft, auxiliary power units (APU), ground support equipment (GSE), and stationary sources, which are dispersed as predicted by AERMOD to produce pollutant concentrations. The EDMS/AERMOD combination is used for the vast majority of airport air quality analyses performed in the United States. • CALPUFF: The California Puff (CALPUFF) model is a non-proprietary, non-steady-state Lagrangian Gaussian puff model maintained and distributed for no cost by Exponent. The term non-steady state means that the local meteorological conditions can change with time, and Lagrangian refers to following or tracking a puff or parcel of contaminants in space and time. EPA has identified CALPUFF as a preferred model for assessing the impacts of long- range transport of pollutants (greater than 50 km). Long-range transport is usually assessed when primary pollutants from an elevated source are transported to downwind distances and, when they chemically react with other pollutants, form secondary pollutants that affect human health. CALPUFF uses overlapping puffs to represent a continuous plume (see Figure 1). Along with the dispersion of primary gas-phase and aerosol species, plume dynamics, and wet and dry deposition, CALPUFF includes particle formation of nitrates and sulfates (from NOx and SO2, respectively) and anthropogenic secondary organic aerosols. CALPUFF applications typically assess long-range transport to distances as far as 300 km from large point sources such as power plants. • SCICHEM: The Second-Order Integrated Puff Model with Chemistry (SCICHEM) is a non-steady-state Lagrangian puff dispersion model based on the Second-Order Integrated Puff Model (SCIPUFF). The term second-order refers to the model’s “turbulence closure”

10 Dispersion Modeling Guidance for airports addressing Local air Quality health Concerns scheme and to assumptions/hypotheses used in the solution to the equations governing the wind, turbulence, and concentration of the puff transport and dispersion. SCICHEM uses a collection of Gaussian puffs to represent concentration fields. The model is non-proprietary and is designated by EPA as an alternative dispersion model to be used on a case-by- case basis for both short- and long-range (> 50 km) regulatory applications. SCICHEM accounts for puff dynamics, complex terrain, wet and dry deposition, and secondary par- ticle formation. Plumes are represented in three dimensions by numerous puffs that are advected and dispersed independently, reflecting the local meteorology. Puff merging and splitting occurs to reflect the variability inherent in weather. SCICHEM simulates chemi- cal processes in the gas, aerosol and aqueous phases, with chemical transformation taking place in the plumes. Traditionally, SCICHEM has been used to model large point sources, such as power plants. • ADMS-Airport: The Atmospheric Dispersion Modeling System at Airports (ADMS-Airport) is a Gaussian plume dispersion model for aircraft-related sources developed and maintained by Cambridge Environmental Research Consultants (CERC). ADMS-Airport is a propri- etary model, which means users must obtain a license from the developer for its use. ADMS- Airport accounts for chemical reactions for NO, NO2, and O3, as well as the production of sulfate particles from SO2. It can accommodate complex terrain and can account for puffs or plumes, wet and dry deposition, and plume dynamics. It includes emissions sources found at airports, including aircraft, APU, GSE, on-road mobile sources, and airport stationary sources and uses algorithms designed specifically to model dispersion from aircraft engines. The aircraft jet model within ADMS-Airport includes equations for conservation of mass, momentum, heat, and pollutant species. It computes the effect that movement of the aircraft engine has on reducing the effective buoyancy of the exhaust. This calculation is particularly important for evaluating dispersion from the high-momentum, buoyant take-off ground roll from aircraft. ADMS-Airport’s ability to model atmospheric chemistry and its aircraft jet model are important reasons for considering its use. ADMS-Airport has been used to model air quality at London’s Heathrow Airport as part of the UK Department for Transport’s Proj- ect for Sustainable Development of Heathrow and it is one of the models used by the Inter- national Civil Aviation Organization Committee on Aviation Environmental Protection (ICAO CAEP). Where to Obtain Models The four primary models discussed in ACRP Research Report 179 can be obtained using links at the following websites: • AERMOD: https://www3.epa.gov/ttn/scram/dispersion_prefrec.htm#aermod • CALPUFF: https://www3.epa.gov/ttn/scram/dispersion_prefrec.htm#calpuff • SCICHEM: https://sourceforge.net/projects/epri-dispersion/files/SCICHEM/ • ADMS-Airport: http://www.cerc.co.uk/environmental-software/ADMS-Airport- model.html Figure 1 shows a schematic of plume versus puff models. The left panel shows the instanta- neous plume as it is realistically observed, versus the average plume that is modeled by AERMOD and ADMS-Airport. The right panel shows how the plume can be modeled as a sequence of puffs, a method that is especially useful with time-varying winds, as in the case of SCICHEM and CALPUFF.

primer on airport air Quality and Dispersion Modeling 11 2.2 Selecting a Dispersion Model for Airport Air Quality Analysis Selecting the right model to use for a particular study at an airport is important because the data and manpower requirements can vary widely, affecting the cost of conducting the study. Questions a researcher may ask when choosing a model are: • Am I creating an emissions inventory to meet a regulatory requirement for determining pol- lutant concentrations? • Will the requirements for computing pollutant concentrations be driven by regulatory needs or evaluation of pollutant health impacts? • Does the model have the scientific rigor for the pollutants of interest, and is it accepted for practice in the scientific/regulatory modeling community? A decision tree can assist with choosing an air quality model (see Figure 2). In this figure, although AERMOD and ADMS-Airport are the preferred models for primary pollutants, they do include treatment of the conversion from NO to NO2, which could be important at localized scales. CALPUFF has a relatively simpler parameterized scheme for PM2.5, but SCICHEM has a more complex treatment for other secondary pollutants like O3 and PM2.5. Plume Puffs Figure 1. Schematic showing plumes versus puffs.

12 Dispersion Modeling Guidance for airports addressing Local air Quality health Concerns Figure 2. Decision tree for selection of an air quality model.