Policy Analysis pubs.acs.org/est
Spatial Differences and Costs of Emissions at U.S. Airport Hubs Matthew J. Nahlik Graduate Student Civil, Environmental and Sustainable Engineering, Arizona State University, Tempe, Arizona 85281, United States
Mikhail V. Chester,* Assistant Professor Civil, Environmental and Sustainable Engineering, Arizona State University, Tempe, Arizona 85281, United States
Megan S. Ryerson, Assistant Professor Department of City and Regional Planning University of Pennsylvania, Philadelphia, Pennsylvania 19104, United States
Andrew M. Fraser Graduate Student Civil, Environmental and Sustainable Engineering, Arizona State University, Tempe, Arizona 85281, United States ABSTRACT: As local governments plan to expand airport infrastructure and build air service, monetized estimates of damages from air pollution are important for balancing environmental impacts. While it is well-known that aircraft emissions near airports directly affect nearby populations, it is less clear how the airport-specific aircraft operations and impacts result in monetized damages to human health and the environment. We model aircraft and ground support equipment emissions at major U.S. airports and estimate the monetized human health and environmental damages of near airport (within 60 miles) emissions. County-specific unit damage costs for PM, SOx, NOx, and VOCs and damage valuations for CO and CO2 are used along with aircraft emissions estimations at airports to determine impacts. We find that near-airport emissions at major U.S. airports caused a total of $1.9 billion in damages in 2013, with airports contributing between $720 thousand and $190 million each. These damages vary by airport from $1 to $9 per seat per one-way flight and costs per passenger are often greater than airport charges levied on airlines for infrastructure use. As the U.S. aviation system grows, it is possible to minimize human and environmental costs by shifting aircraft technologies and expanding service into airports where fewer impacts are likely to occur.
■
INTRODUCTION The airline industry is forecast to grow in the United States (U.S.) and around the world, and such growth can have tremendous economic benefits for communities.1 Scholars have linked the number of destinations served by an airport as well as the quantity of air service present at an airport positively to regional employment, wages, and job growth.2−7 To allow air service to expand and meet surging demand, 17 of the top 35 airports (by passenger volume) expanded (or began planning to expand) their airfields from 2000 to 20138 and the Federal Aviation Administration (FAA) is investing in procedures and plans to increase flight operations at major airports. 9 Concurrently, airports have seen significant growth near facilities6 and communities are becoming increasingly aware of the human health and environmental impacts of air emissions. Airport planners have long estimated environmental impacts to balance these impacts with economic benefits of a © XXXX American Chemical Society
decision; however, these estimated impacts focus on pollutant levels rather than the specific damage caused by pollutants.10 As airports plan for expansions and growth of operations, monetized estimates of damages from airport pollution must include location-specific costs of air pollutants and greenhouse gases to better understand the human and environmental risks that result from airport activities.11 Local air quality is assessed by airports and the FAA in environmental impact assessments of new airfield capacity or to provide air quality inputs to State Implementation Plans (state plans that account for the air quality related to transportation facilities). During the planning process for changes to airfield Received: September 15, 2015 Revised: February 10, 2016 Accepted: March 23, 2016
A
DOI: 10.1021/acs.est.5b04491 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Policy Analysis
Environmental Science & Technology
efforts have produced valuable insight into the monetized costs (often external) of air travel.25−27 Airport damage studies have focused largely on the human health impacts of only a few emissions of particular concern, namely the criteria pollutants (especially primary and secondary PM2.5), and in recent history GHG emissions.25,26,28,29 Efforts to monetize such pollutants generally are to support studies of the increased morbidity and mortality for residents nearby airports.22,30−32 Translating human health and environmental impacts more broadly (beyond morbidity and mortality) to monetary costs has become an important area of study as damages expressed in currency form can directly inform planning and trade-off analysis for airport plans. Monetized damages of exposure to air pollution are often based locationspecific impact assessment of human health to local air pollution and distributions of the costs of treatment.26,27 They often focus on a subset of pollutants.22,27,33 Several methods exist for estimating and valuing the impacts of emissions from aircraft.19 Yet none identified consider the costs of environmental impacts in addition to human health impacts, few consider the costs of exposure to all criteria air pollutants, and few consider the social cost of GHG emissions. The recent emergence of location-specific monetized damages for human health, environmental, and climate change impacts offer new opportunity for understanding the consequences of infrastructure planning.34 The monetized damages of air pollution exposure vary by location and cost estimates should account for the marginal impacts of air pollution based on background air quality in addition to meteorological conditions and population characteristics. To this end estimates of the monetized costs of air pollution at major U.S. airports are developed with flight operations data, an emissions model, and county-specific unit costs of pollution from the Air Pollution Emission Experiments and Policy analysis (henceforth referred to as APEEP) model. This research addresses two major questions: (1) What pollutants are emitted from commercial aircraft and ground support equipment at each airport and what are the associated monetized damages of this pollution to human health and the environment? (2) Which airports in the U.S. cause the largest damages and which pollutants are the leading cause of damages?
capacity−the airport master planning process or the National Environmental Policy Act process through which an Environmental Impact Statement (EIS) is prepared−airports (with the support of the FAA) model local pollution to disclose any changes in environmental pollutants due to the proposed infrastructure changes. Furthermore, the Environmental Protection Agency (EPA) regulates that airports, as a part of a state’s transportation system, meet air quality conformity requirements specific to each local pollutant (and thus attain air quality standards); airport air quality assessments become an input into the State Implementation Plans which influence state transportation funding and environmental priorities.12 As of 2005, 37 of the 50 busiest airports in the U.S. by passenger movements are in regions that are considered nonattainment areas for ozone by the Environmental Protection Agency.13 Prior to 2007 air quality assessments included local pollutants only; this changed in in 2007 when the Supreme Court ruled that greenhouse gases be classified as criteria pollutants to be regulated by the EPA.14 This ruling combined with the EPA’s positive endangerment finding for greenhouse gases15 and the FAA’s subsequent guidance on how airports should account for greenhouse gas (GHG) emissions16 cemented that they be included with local pollutants in airport environmental assessments. Air quality assessments of aircraft operations are based on the activities that occur between ground level and the mixing height (about 3000 ft). While the cruising phase of a flight contributes to environmental and human health degradation it is below the mixing height that local pollutants are of direct and significant concern to human health.11,17,18 There are many activities at airports that create air emissions including aircraft, ground support equipment (GSE), auxiliary power units (APU), vehicles (airport, personal, and freight), stationary power plants, and construction equipment. For aircraft, airports will assess pollution based on the flight stages that occur below 3000 ft (engine warm up, taxi out, take off, climb out, approach, and taxi in). These flight stages are described as the landingtakeoff (LTO) cycle and capture an estimated 10% of NOX, SOX, and PM emissions and 30% of CO and HC emissions that occur during the flight (the remaining percentages being associated with the cruise phase).19 In the U.S., the FAA’s Emissions and Dispersion Modeling System (EDMS) has been the state-of-the-art tool for estimating emissions at airports. As of May 2015 EDMS became part of the FAA’s Aviation Environmental Design Tool (AEDT). Airports and researchers have used EDMS either directly or in concert with other models to estimate the dispersion of pollutants ultimately enabling the assessment of human health exposure, morbidity, and mortality effects.19−22 While a small fraction of total flight emissions, this pollution has significant potential for local human health and environmental impacts.23 In the past two decades, the FAA, aviation organizations, and researchers have made significant advances toward understanding and mitigating the impacts from pollution caused by airports. Data on flight operations, aircraft emissions, pollution dispersion, and human and environmental impacts have become prevalent. These air pollution studies are well documented in guidebooks and synthesis studies that recommend best practices for mitigating pollution at airports.24 In addition to a rich body of literature that focuses on quantifying pollution at airports, there have been many efforts to monetize the impacts of air pollution at airports. These
■
MATERIALS AND METHODS In this study, we estimate near-airport emissions from aircraft landings, takeoffs, surface movements, and ground support equipment at major U.S. airports as well as calculate the associated monetized damages. We create emission profiles (which include: carbon monoxide (CO), carbon dioxide (CO2), particulate matter smaller than 2.5 μm (PM2.5), sulfur oxides (SOx), nitrogen oxides (NOx), and volatile organic compounds (VOC)) by modeling the landing and takeoff (LTO) cycle for 26 aircraft categorized into six characteristic aircraft groups. The APEEP model34 is used with existing research on CO damages and the social cost of CO2 (SCC) to estimate marginal damage costs of aircraft emissions and calculate annual damages associated with airport flight activity. For the purpose of this research, the word “damages” describes the valuation of harmful effects from all near-airport emissions and is presented in 2013 U.S. dollars. Nationwide Aircraft Movement and Emissions. We create profiles of aircraft movements at each airport and assess the emission from aircraft LTO cycles at 70 airports of national importance in the continental 48 states. The FAA designates, in B
DOI: 10.1021/acs.est.5b04491 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Policy Analysis
Environmental Science & Technology
Table 1. Aircraft Used for Emissions Modelinga
their Aviation System Performance Metrics (ASPM) database for planning purposes, 77 airports across the U.S. as the most critical to the system. Planning these airports and their related airspaces is highly complex.35 Starting with the 77 airports, airports that do not reside within the continental 48 states (the APEEP unit damage cost data set is only for the continental U.S.) or that do not have commercial flights are removed. Three airports are added based on large service volume. The final list includes 70 airports that each handled over 1.5 million passengers in 2013 and facilitated 87% of commercial passenger enplanements for the entire U.S. The selected 70 airports are comprised of 58 large and medium commercial hubs that provided 97% of commercial enplanements for hubs of that size and 12 small hubs that handled 24% of all U.S. enplanements at small hubs in 2013. For each of the airports, a profile of aircraft movements and resulting emissions based on the annual LTO cycles performed by each type of aircraft is created. We model emissions from aircraft during LTO using the Federal Aviation Administration’s (FAA) Emissions and Dispersion Modeling System (EDMS). EDMS estimates nearairport emissions (including those from aircraft as well as emissions from ground support equipment) for each LTO cycle based on inputs such as type of aircraft, engines used, loading capacity, and weather conditions.36 To capture the emissions (and their impact) at an airport, we simulate a year of scheduled flights at each airport, using schedules collected from the Bureau of Transportation Statistics (BTS) T100 database. Table T100 is an annual summary of airline-reported statistics about domestic and international travel operated by U.S. and international air carriers including departure airport, destination airport, aircraft type used, number of passengers, payload weight, and flight purpose.37 Across the entire set of flights between U.S. airports present in the T100 database in 2013, 88% of annual aircraft landings and takeoffs at these airports occurred on 48 different aircraft types (type specifying an aircraft model and engine type). We find thatby generalizing that related aircraft models with the same engine types are highly similar26 aircraft types well represent the top 48 aircraft types that are in operation in general across all U.S. airports (Table 1). For example, when consolidating 48 aircraft types into 26, an older model of the Boeing 737 with the same engine type as the 737−300 will be classified as a 737−300. We simulate all flights at a single airport and assign one of the 26 aircraft types in Table 1 to each flight given aircraft type and engine type. To evaluate an “average” flight on a general aircraft type, we further define six characteristic aircraft groups designated, based on body types and propulsion systems, into the following groups: short-range regional jets, short-range turboprops, medium-range small jetliners, medium-range large jetliners, long-range small jetliners, and long-range large jetliners (also shown in Table 1). This binning approach is similar to previous airline studies where the number of seats available was used to determine aircraft types.38 For example, the short-range regional jet group includes three aircraft models: Bombardier CRJ-200, Embraer ERJ-145, and Bombardier CRJ-700 (shown in Table 1). For each group of aircraft, we calculate a weighted-average emission profile based on the distribution of LTOs performed by each of the group’s aircraft. Total annual emissions at each airport are then estimated from the LTO emission profiles by pairing them with the annual aircraft group-specific LTO cycles at each airport. This methodology is similar to the FAA Aircraft Engine Emissions Database Model which has been shown to produce
group
aircraft
manufacturer
share of LTOs
engine
Short-Range Regional Jet Short-Range Twin-Engine Regional Jet (50 Average Seats) CRJ−200 Bombardier 44% CF34−3B1 ERJ−145 Embraer 36% AE3007A1/1 CRJ−700 Bombardier 19% CF34−8C5 Short-Range Turboprop Short- to Medium-Range Twin-Engine Turboprop (50 Average Seats) Dash−8−400 Bombardier 32% PW150A EMB−120 Embraer 20% PW118 Brasilia Dash−8−100 Bombardier 19% PW120A Dash−8−300 Bombardier 15% PW123 340B Saab 12% CT7−9B ATR−72 ATR 2% PW124B Medium-Range Small Jetliner Short- to Medium-Range Narrow-Body Twin-Engine Jetliner (140 Average Seats) 737−800 Boeing 38% CFM56−7B27 A320−100 Airbus 25% CFM56−5B5/P DC9MD−80 McDonnell 12% JT8D−7 Douglas ERJ−175 Embraer 8% CF34−8E5 CRJ−900 Bombardier 6% CF34−8C5 737−300 Boeing 6% CFM56−3−B1 717−200 Boeing 5% BR700−715A1 Medium-Range Large Jetliner Medium-Range Narrow-Body Twin-Engine Jetliner (180 Average Seats) 757−200 Boeing 79% RB211−535E4 A321 Airbus 21% CFM56−5B3/P Long-Range Small Jetliner Medium- to Long-Range Wide-Body Twin-Engine Jetliner (230 Average Seats) 767−300 Boeing 44% PW4060 A330−200 Airbus 23% PW4156 A300−600 Airbus 20% PW4156 767−200 Boeing 11% PW4056 A310−200 Airbus 3% JT9D-7R4D Long-Range Large Jetliner Long-Range Wide-Body Quad-Engine Jetliner (350 Average Seats) 747−400 Boeing 84% CF6−80C2B1F A380−800 Airbus 9% Trent 895 A340−200 Airbus 7% CFM56−5C4 a
The six aircraft groups are shown with distinguishing characteristics. Twenty-six total aircraft are listed with the engine type used for modeling emissions and their share of landing and takeoff cycles (LTOs) within the group.
accurate results (within 6%) when compared to higher resolution emissions inventory models.19 Combining emissions with county-specific marginal damage costs enables the estimation of the total damages from air pollution associated with U.S. airports. To capture the variability in emissions for similar aircraft types within our six groups, we simulate the LTO cycle emissions of 26 major aircraft types using EDMS. We configure EDMS to run in Performance Based mode for the 26 aircraft for each of the airports. For each airport, the 26 aircraft are specified and are assigned default LTO stage settings to capture facility-specific variations in operations with the exception of airport specific taxi times which are based on available performance data.39 A LTO cycle is defined with eight stages C
DOI: 10.1021/acs.est.5b04491 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Policy Analysis
Environmental Science & Technology
worldwide environmental emissions inventorying protocols− outlined by the authors include: Scope 1, emissions over which the airport has direct control (such as those from airport ground service vehicles); Scope 2, those that are indirectly related to airport actions (such as the purchase of electricity); and Scope 3, those that are indirect and optional to include because they are under the ownership of another entity (such as emissions from airlines and concessionaries).29 GHG emissions from aircraft are “Scope 3” in this context; however, when airports perform environmental assessments they are required to account for all pollutants from aircraft up to 3000 feet. While the airport might not have direct control over such emissions, any change to the airporta change in service or in infrastructurewill necessarily influence aircraft emissions. Estimating Damages from Emissions. APEEP provides ground-level damage valuations for local air pollutants that are specific to every county in the contiguous 48 states. County damages are assigned to each airport.34,49,50 The integrated assessment model uses 2005 baseline emissions from the National Emissions Inventory to estimate how pollutants released in one county cause impacts within the county of release as well as other counties to which they are transported. In addition to a pollutant fate and transport model, APEEP uses epidemiological dose−response models and local population data to estimate the effect of each pollutant on human morbidity and mortality, as well as other environmental impacts. APEEP attributes the damage from secondary pollutants back to the precursor emission, such as the cost of secondary particulate matter. Environmental degradation and human health impacts (using a $6 million 2000 USD value of a statistical life (VSL)) are valued for every county in the continental U.S. When adjusted to 2013 USD the value of statistical life is $8.1 million, closer to more recent estimates, and an uncertainty analysis is developed for a range of VSL.51 The APEEP costs are over 90% comprised of impacts to human health, with the remaining damages caused by visibility loss, reduced agricultural yields, reduced timber yields, accelerated depreciation of man-made materials (due to acid rain), and lost recreation usage due to impaired forest health.32 The unit costs are used to estimate damages from PM2.5, SOx, NOx, and VOC attributable to each airport on a dollar-per-ton basis. The resulting airport damages are for operations in 2013 while APEEP damages are based on 2005 populations. Uncertainty due to population, meteorology, and air quality changes between 2005 and 2013 may exist. Marginal damages vary in the APEEP model by height of emission release. The APEEP model provides the cost of emissions from mobile and point sources at ground level in addition to damage costs for point sources up to 250 m and point sources from 250 to 500 m.34,49,50 Because not all aircraft emissions occur at ground-level, best-fit power functions are created to estimate how APEEP unit damage costs will change for emissions released above ground-level. Taxi In, Ground Support Equipment, Auxiliary Power Unit, Startup, and Taxi Out occur at ground level.36 The remaining three stages are assessed at an average altitude (460m for Approach, 150m for Takeoff, and 610m for the Climb Out) based on height definitions from the FAA for each stage.36 Stage-specific altitudes are used as input to the county and pollutant-specific power function to find altitude-specific unit damages for PM2.5, SOx, NOx, and VOCs. The result is that as the altitude of the aircraft increases, the damages from emissions decrease. It is acknowledged that uncertainty may exist in the damage
of aircraft or ground support equipment operation that occur below 910m (3000 ft), which is considered the mixing height or maximum altitude where pollutants can evenly mix in the air and affect local air quality.40 These eight stages are (1) Approach, (2) Taxi In, (3) Ground Support Equipment, (4) Auxiliary Power Unit, (5) Startup, (6) Taxi Out, (7) Takeoff, (8) and Climb Out. Emissions of PM2.5, SOx, NOx, VOCs, CO, and CO2 are estimated with EDMS for the LTO cycle. The estimation of stage-specific emissions enables us to assess the impacts of emissions released at various heights.33 We model 1000 LTO cycles for each aircraft at each airport in EDMS and then using BTS T100 profiles we scale emissions based on annual performance of each prototypical category. We note that EDMS does not calculate CO2 emissions from APUs so this process is excluded. Given the challenge of accurately estimating emissions from ground transportation including passenger ingress and egress (by personal vehicle or public transit) and freight activities, these activities are also excluded. Our analysis of the impacts of near airport pollutants include CO2 emissions (whose impacts occur on global scales) in addition to local air pollutants because airports (i) are legally required to include GHG emissions in planning and (ii) can influence CO2 emissions during LTO cycles. On the latter point, airports can influence GHG emission from flights during the LTO cycle through the assignments of departure and arrival routes, the design of runway geometries, and the performance of air traffic controllers.41−44 On the former point, the 2007 Supreme Court case, Massachusetts v Environmental Protection Agency (MA v EPA), resulted in the classification of GHGs as criteria pollutants to be regulated by the EPA.14 In 2009, the EPA made a positive endangerment finding for GHG emissions, thus concluding that they endanger public health and welfare.45 These events led to the requirement for airports and the FAA that GHGs must be accounted for in all environmental assessments (including master plans to Environmental Impact Statements). How airports should account for such pollutants, however, is another consideration. In 2012 the FAA issued guidance to airports that air quality assessments are to include GHG emissions and that the accounting should be consistent with the current approach of accounting for pollutants from aircraft operations only up to the mixing height (FAA, 2012). When evaluating criteria pollutants, emissions that occur above 3000 feet, the mixing height, are considered to be beyond the purview of the airport and also to not significantly affect human health.46,47 In the FAA’s 2012 guidance to airports regarding the accounting of GHG emissions, they state: “the maximum altitude for any analyses for an airport NEPA action would be the landing take-off cycle emissions up to the local mixing height, which is consistent with the current approach and EPA guidance with regard to local air quality evaluations.”16 As such, criteria pollutants and GHG emissions below 3000 feet are included in the damage assessment. There are conflicting guidelines on how airports should account for GHG emissions. While airports must include the pollutant in master plans and Environmental Impact Statements up to 3000 feet, GHG inventory guidelines state that airports should consider the pollutant to be Scope 3, the result of airline activity.48 Kim, Waitz, Vigilante, and Bassarab48 suggest that airports should estimate GHGs from an entire flight and attribute the pollutants to a set of emissions owners including the airport, the airlines, and the traveling public. The best practices for inventorying airport GHGs−consistent with D
DOI: 10.1021/acs.est.5b04491 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Policy Analysis
Environmental Science & Technology
marginal damage costs of any county considered in this study. This effect is likely due to the large population in the Los Angeles area and the meteorological and geographical features, as illustrated by Los Angeles’ history with high levels of air pollution and its current status as a nonattainment area for both ozone and PM2.5.55 Los Angeles International, along with John F. Kennedy International and LaGuardia in New York City, Chicago O’Hare, and Minneapolis St. Paul are located in the four counties with the highest unit damage costs for emissions. As the marginal cost for each pollutant varies based on location, the possibility exists for airports that handle nearly identical aircraft movements to cause different total damages to nearby communities. Consider that, for example, Minneapolis St. Paul International Airport (MSP) in Minnesota and Phoenix Sky Harbor International Airport (PHX) in Arizona both facilitated about 190 000 LTO cycles annually (520 cycles per day) and possessed similar emission profiles. Despite the comparable mass of emissions, damages from PHX approached $43 million while annual damages from MSP totaled $78 million dollars. This occurred due to substantially higher damage costs for PM2.5, SOx, NOx, and VOCs emitted in Hennepin County, Minnesota compared to Maricopa County, Arizona. The total damages caused by Los Angeles International Airport (LAX) were 84% greater than Hartsfield-Jackson International Airport (ATL) in Atlanta, despite ATL handling 70% more air traffic (440 000 LTO cycles annually at ATL and 260 000 at LAX). Technological Disparities in Damage Costs. Trade-offs in damages exist between competing aircraft technologies. Figure 2 shows that, on average, long-range small-bodied jetliners cause $1.26 fewer damages per seat for each flight than long-range large-bodied jetliners. Similarly, turboprop aircraft cause $0.88 fewer damages per seat than short-range jets. For the medium-range distance market, larger aircraft often cause fewer damages per seat than the smaller aircraft. For example, medium-range large jets in Los Angeles cause $0.92−1.30/seat per flight fewer damages than small jets. In 2009 the Government Accountability Office reported that some airlines switched from regional jets to turboprops because turboprops offered better fuel efficiency and lower costs during the fuel price spike of 2007−2008.56 Despite turboprop aircraft being more fuel efficient than jet aircraft,57 damages per seat on a turboprop are highly variable across all airports compared with jet aircraft because turboprops emit VOCs at a higher rate, and VOCs carry a greater marginal damage cost in particular communities due to their human health impacts. Substituting one aircraft for another at some airports could more efficiently move greater numbers of passengers while causing similar damages. Replacing medium-range small aircraft with large aircraft could reduce damages while preserving (or potentially increasing) the number of seats available at most of the top 28 airports (ending with Chicago Midway International) in Figure 2. For the bottom 42 airports, the large- and small-bodied medium-range aircraft cause nearly identical damages per-passenger and no recommendations can be made about aircraft substitutions. Operational Disparities in Damages. To evaluate the operational disparities across the airportsthe concept that different airports have different damage costs overall because they handle vastly different quantities of flightsdamages are normalized by the quantity of aircraft and passengers handled at each airport. Together the airports served nearly 640 million passengers in 2013 with a high level of dispersion in the
estimates above 500 m. Where airports are located across two counties, average unit damage cost factors from both counties are used. Because APEEP assesses only four of the six emissions in the inventory, other valuation sources are used to estimate unit damages of CO and CO2 emissions. Previous research summarizes the cost ranges for CO and CO2 and these are used to assess the damages associated with these pollutants. The average damage cost of CO in the U.S. is estimated at $640/Mg in the year 2000, with high and low values of $1,300/ Mg and $1.2/Mg, respectively.52 The monetization of aircraft CO2 emission damages, from the SCC, are based on strategies for mitigation as a lower cost boundary or the most expensive strategies for adaptation. The average SCC with a 3% discount rate is reported as $34/Mg in the U.S., while the lowest average cost with a 5% discount rate and the 95th percentile cost with a 3% discount rate are reported as $11/Mg and $97/Mg in the year 2013 by the U.S. EPA.53 Results are computed using an average cost of CO and CO2, and an uncertainty assessment is developed to capture the low and high values. After damage valuations are calculated from these sources, costs are adjusted to year 2013 USD with the Bureau of Labor Statistics’ Consumer Price Index.54
■
RESULTS Emissions at the 70 airports caused a total of $1.9 billion in domestic damages in 2013, with airport-specific contributions ranging from between $720 thousand and $190 million (Figure 1). The climate impacts from 13 Tg of CO2 emissions are felt both locally and globally, but they are attributed to the airport where they are released. The damages from all other emissions are experienced locally, in the origin county where the pollutants were first emitted or in neighboring counties, and attributed to the airport that facilitated the aircraft movements. Damages from NOx and SOx showed the greatest variation to percent contribution of total damages, because the unit damage costs vary from nearly zero up to $46,000/Mg for NOx or $8,700−$470,000/Mg for SOx. Emissions of PM2.5 and SOx were the most costly on average per unit of emission, at $230,000/Mg and $57,000/Mg respectively, yet only caused one-third of damages on average because the mass of pollutants emitted were relatively low. Figure 1 shows the pollutant composition and total monetary damages for each of the examined airports. Toward decomposing the results in Figure 1, the damages from a single flight and of one seat on each aircraft type are first explored (Figure 2). Following this, the normalized damages based on passengers and aircraft moved are explored. Spatial Disparities in Damages. The location of pollutant release coupled with airport-specific operation conditions (here, the time spent in each LTO stage) can have significant effects on the damages associated with flights (Figure 2). For example, a long-range jet can cause $3,100 in damages in New York as compared to $500 in damages in New Orleans, a 520% difference. While the damages are smaller for smaller aircraft types, the differences across airports are still stark. For shortrange turboprops, per flight damages range from $320 in John F. Kennedy International to $24 in Los Angeles-Ontario. These differences are due in part to the differences in time spent in each LTO stage, but are more significantly affected by the airport-specific marginal damage costs which are driven by population, existing pollution levels, and meteorological conditions. Los Angeles County, California has the highest E
DOI: 10.1021/acs.est.5b04491 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Policy Analysis
Environmental Science & Technology
Figure 1. Daily Airport Traffic and Annual Damages by Emission. Annual damages are rank-ordered and itemized by emission in a stacked column per airport. Points on a secondary axis show daily passengers served and daily aircraft landing and takeoff cycles (LTO).
Figure 2. Damages per Flight and per Seat from One LTO by Aircraft and Airport. The damages for each of the six aircraft groups performing one LTO cycle are shown (top), in addition to the same damages normalized over the number of available seats in the aircraft (bottom). These are not additive at each airport; each point is damages from an individual aircraft.
F
DOI: 10.1021/acs.est.5b04491 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Policy Analysis
Environmental Science & Technology
Figure 3. Damages Averaged by Passenger Movements and Payload Weight. Annual damages were divided by passenger movements and by aircraft payload weight to compare airport damages independent of aircraft movements. Aircraft landing fees charged at each airport are also shown.
number of passengers moved across airports. The busiest airport is Hartsfield-Jackson Atlanta International which served 45 million passengers in 2013, while Long Island MacArthur Airport served the fewest passengers at 670 000. Dividing the total damage cost at each airport by the total annual passengers served at that airport, damages per passenger varied from $1 to just over $9 on average by airport (Figure 3). Significant variation can be seen across the top eight airports where damages range from $5−9 per passenger. Per-passenger damages range from $1 to $4 for the remaining 62 airports. Similarly, dividing the total damage cost by 1000 lbs of payload weight served at that airport, a range of $4/1000 lbs to $27/ 1000 lbs of weight is calculated. The airports located in New York City and Los Angeles are among those with the highest damage cost per passenger and weight due to the high marginal pollutant costs in each city. Almost half ($890 million or 48%) of the domestic damages are caused by seven airports, but these cities have great variation in both marginal pollutant cost and operations. Two of the busiest airports in the U.S. in 2013, Atlanta HartsfieldJackson and Dallas Fort Worth, created large damages compared to other airports due to significant quantities of emissions. The aircraft traffic at these two airports was large enough to offset the relatively low unit damage costs associated with the Atlanta and Dallas metro areas. Large number of Aircraft LTOs and/or large unit damage costs for emissions result in the top five airports (LAX, JFK, ORD, LGA, and ATL) generating damages in excess of $100 million. Los Angeles County had the highest unit damage costs from air emissions and Los Angeles International was the fourth busiest airport by aircraft movements in 2013, which together caused it to produce the largest damages of any airport ($190 million annually). The damages from Los Angeles International are valued at 11% greater than the second most damaging airport, John F. Kennedy, mostly due to the high cost of sulfur emissions in Los Angeles County. Conversely, five airports each caused less than $2 million in annual damages and four of those five are smaller aviation fields (handling fewer than 50 LTO cycles each day) that are located within 100 miles of and serve three major metropolitan areas: New York, Los Angeles, and
Boston. These airports caused minimal damages because each facilitated a small amount of aircraft traffic in 2013. Comparing the damages per passenger moved and per landing weight can help contextualize the results against other aviation system costs. Typical airline operating costs to cover fuel, crew, and maintenance are on the order of $0.10 per passenger-mile.58 Using this figure, it would cost an airline approximately $150 per passenger to operate a 1500 mile flight halfway across the country, from Phoenix, Arizona to St. Louis, Missouri. Using that cost as a base, adding the $1−9 perpassenger pollutant damages from takeoffs and landings could increase the operating cost by up to 6% if airports require airlines to internalize the environmental and human health damages. A direct cost incurred per-flight by the airlines is the landing fee: the fee charged by weight for an aircraft landing collected to cover the marginal cost that flight imposes on airport maintenance costs, or very directly, the marginal cost imposed by each flight to use the infrastructure. These fees typically range from $1−10 per 1000lbs of aircraft landing weight; specific airport fees are shown in (Figure 3) (fees were collected from the FAA’s Airport Financial Reporting Program, Operating and Financial Summary, FAA Form 5100−127).59 Damages per 1000 lbs of aircraft landing weight (which includes the aircraft base weight along with the passenger and freight payloads) at each airport varied from $4 to almost $27 (Figure 2). The estimated damage cost per payload weight was greater than the existing landing fee at nearly every airport, with the exception of both airports in Washington D.C., BaltimoreWashington International, Lambert St. Louis International, Austin-Bergstrom International, Buffalo Niagara International, T.F. Green Airport, Birmingham-Shuttlesworth International, Portland International, and Ontario International. Damage costs being greater than landing fees provides insight into the economic costs of travel and the potential unaccounted impacts associated with aircraft movements. Results, including an interactive map, are available online at www.transportationlca. org/aviationdamages/. G
DOI: 10.1021/acs.est.5b04491 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Policy Analysis
Environmental Science & Technology
Figure 4. Annual damage cost uncertainty range: An uncertainty range of potential damages from each airport is shown in the same rank order as in Figure 2.
■
■
UNCERTAINTY AND DATA VALIDATION
DISCUSSION The results of this studythat damage profiles from flights are highly spatially variable and dependent on aircraft technology and the quantity of operationshowcase how important understanding damages in a contextualized way is for those making aviation policy and investment decisions. The results and methods can help policy and decision makers develop more informed strategies for how the aviation system can expand while minimizing impacts. Consider that the closing of three major hub airportsthose in Cleveland, Milwaukee, and Pittsburghwas lamented by local and federal policymakers because of the loss in service and connectivity for those populations. However, the results show that these three midsize cities, the home of former airline hubs, have high damages per flight because their communities incur greater damages specifically from NOx, SOx, and VOC emissions. If these airports were to rebuild their airline service, the impact on a per passenger basis would be greater than if airlines were to choose to focus operations elsewhere. There is uncertainty in the results and future work should focus on improving the accuracy of the underlying inputs. Aircraft emission profiles, county-specific damages, climate change mitigation and adaptation costs, and the use of the APEEP model (whose population data and meteorological data may have changed) each introduce uncertainty into the results. Furthermore, we only include CO2 emissions as other greenhouse gas emissions are not modeled with EDMS, and these other greenhouse gases can be significant.61−63 While the uncertainty analysis indicates that the reported average results are likely conservative, future work should focus on reducing this uncertainty so that aviation systems planners have as precise information as possible to work with airlines to reduce impacts. As high-quality monetized damage assessments emerge, there will be new opportunities for better understanding the impacts of transportation systems. Quantifying the monetized damages from high emissions intensity transport hubs can reveal unintended and significant costs, and airports provide an
Damages can vary due to a number of factors including aircraft emissions, county-specific impacts, and climate change impacts. We characterize the uncertainty in emissions due to these factors to assess whether the reported (average) results are likely conservative or aggressive. To do so, uncertainty in emissions is assessed through a Monte Carlo simulation where ambient temperatures, wind, aerodynamic drag, engine fuel consumption, takeoff weight, and engine emission profiles are varied.60 APEEP standard errors were used to assess the uncertainty in airport damages.49 We calculate a 95% confidence interval for each airport (using mean values and standard errors) and combined with the low and high costs of CO and CO2.49,52,53 Furthermore, we vary the value of statistical life between $2.9 and $12 million.26 By combining the lower ranges, the minimum total damages in the uncertainty assessment were estimated at $920 million (Figure 4). Conversely, annual damages could be as high as $12 billion dollars across the U.S. with higher emission rates, the upper 95% APEEP confidence interval, and the most expensive costs from the SCC. Overall, there is more uncertainty that could lead to higher damage costs than there is less uncertainty that could lead to lower damage costs implying that the average results are conservative. Large standard errors in APEEP for both NOx and VOC (which are precursors to smog formation) on the East coast of the U.S. heavily influenced the upper confidence interval. Of the 17 airports where the maximum uncertainty is at least 500% greater than the mean damages, 15 are located on the East coast. This shows that each additional unit of emission in the more densely populated East coast has the potential to cause greater harm to human health than if it were to be emitted elsewhere in the U.S. Beyond estimating uncertainty associated with the unit damage costs, existing airport emission valuation studies provide opportunities for external validation. H
DOI: 10.1021/acs.est.5b04491 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Policy Analysis
Environmental Science & Technology
(19) Kurniawan, J. S.; Khardi, S. Comparison of methodologies estimating emissions of aircraft pollutants, environmental impact assessment around airports. Environmental Impact Assessment Review 2011, 31 (3), 240−252. (20) Arunachalam, S.; Wang, B.; Davis, N.; Baek, B. H.; Levy, J. I. Effect of chemistry-transport model scale and resolution on population exposure to PM2.5 from aircraft emissions during landing and takeoff. Atmos. Environ. 2011, 45 (19), 3294−3300. (21) Woodmansey, B.; Patterson, J. New Methodology for Modeling Annual Aircraft Emissions at Airports. Journal of Transportation Engineering 1994, 120 (3), 339−357. (22) Levy, J. I.; Woody, M.; Baek, B. H.; Shankar, U.; Arunachalam, S. Current and Future Particulate-Matter-Related Mortality Risks in the United States from Aviation Emissions During Landing and Takeoff. Risk Anal. 2012, 32 (2), 237−249. (23) Simone, N. W.; Stettler, M. E. J.; Barrett, S. R. H. Rapid estimation of global civil aviation emissions with uncertainty quantification. Transportation Research Part D: Transport and Environment 2013, 25, 33−41. (24) Airport Cooperativie Research Program. Guidebook for Preparing Airport Emissions Inventories for State Implemenation Plans; Transportation Research Board: Washington, DC, 2013. (25) Wolfe, P. J.; Yim, S. H. L.; Lee, G.; Ashok, A.; Barrett, S. R. H.; Waitz, I. A. Near-airport distribution of the environmental costs of aviation. Transport Policy 2014, 34, 102−108. (26) Brunelle-Yeung, E.; Masek, T.; Rojo, J. J.; Levy, J. I.; Arunachalam, S.; Miller, S. M.; Barrett, S. R. H.; Kuhn, S. R.; Waitz, I. A. Assessing the impact of aviation environmental policies on public health. Transport Policy 2014, 34, 21−28. (27) Yim, S. H.; Lee, G. L.; Lee, I. H.; Allroggen, F.; Ashok, A.; Caiazzo, F.; Eastham, S. D.; Malina, R.; Barrett, S. R. Global, regional and local health impacts of civil aviation emissions. Environ. Res. Lett. 2015, 10 (3), 034001. (28) Givoni, M.; Rietveld, P. The environmental implications of airlines’ choice of aircraft size. Journal of Air Transport Management 2010, 16 (3), 159−167. (29) Airport Cooperativie Research Program. Guidebook on Preparing Airport Greenhouse Gas Emissions Inventories; Transportation Research Board: Washington, DC, 2009. (30) Pope, C. A.; Ezzati, M.; Dockery, D. W. Fine-Particulate Air Pollution and Life Expectancy in the United States. N. Engl. J. Med. 2009, 360 (4), 376−386. (31) Rehdanz, K.; Maddison, D. Local environmental quality and lifesatisfaction in Germany. Ecological Economics 2008, 64 (4), 787−797. (32) Yu, K. N.; Cheung, Y. P.; Cheung, T.; Henry, R. C. Identifying the impact of large urban airports on local air quality by nonparametric regression. Atmos. Environ. 2004, 38 (27), 4501−4507. (33) Unal, A.; Hu, Y.; Chang, M. E.; Talat Odman, M.; Russell, A. G. Airport related emissions and impacts on air quality: Application to the Atlanta International Airport. Atmos. Environ. 2005, 39 (32), 5787− 5798. (34) Muller, N. Z. Using Index Numbers for Deflation in Environmental Accounting. Environment and Development Economics 2014, 19 (4), 466−486. (35) Federal Aviation Adminstration. Aviation System Perfomance Metrics, 2015. (36) Federal Aviation Administration. Emissions and Dispersion Modeling System (EDMS) 5.1.4; Federal Aviaiton Adminstration Ofice of Environment and Energy: Washington, DC, 2013; p 155. (37) Bureau of Transportation Statistics. Table T100 Domestic Segment (All Carriers), 2013. (38) Kim, D.; Barnhart, C. Flight schedule design for a charter airline. Computers & Operations Research 2007, 34 (6), 1516−1531. (39) Bureau of Transportation Statistics. Airline On-Time Performance Data; U.S. Department of Transportation: Washington, DC, 2013. (40) Beychock, M. R. Atmos. Environ. 2005, 29, 3397. (41) Guo, R.; Zhang, Y.; Wang, Q. Comparison of emerging ground propulsion systems for electrified aircraft taxi operations. Transportation Research Part C: Emerging Technologies 2014, 44, 98−109.
ideal subject for new research. A small number of airports constitute the majority of U.S. air travel and most are located in major metropolitan areas. A rigorous understanding of the monetized damages of all passenger systems will provide us with an opportunity to better plan a system that is highly interdependent.
■
AUTHOR INFORMATION
Corresponding Author
*Phone: 480-965-9779; e-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS The authors thank three anonymous reviewers for their insights and feedback. REFERENCES
(1) ICAO. Global Air Transport Outlook to 2030 and Trends to 2040; International Civil Aviation Organization, 2013. (2) Bilotkach, V. Are airports engines of economic development? A dynamic panel data approach. Urban Studies 2015, 52 (9), 1577−1593. (3) Brueckner, J. K. Airline Traffic and Urban Economic Development. Urban Studies 2003, 40 (8), 1455−1469. (4) Button, K.; Doh, S.; Yuan, J. The role of small airports in economic development. J. Airport Manage. 2010, 4 (2), 125−136. (5) Button, K.; Taylor, S. International air transportation and economic development. Journal of Air Transport Management 2000, 6 (4), 209−222. (6) Green, R. K. Airports and economic development. Real Estate Economics 2007, 35 (1), 91−112. (7) Sheard, N. Airports and urban sectoral employment. Journal of Urban Economics 2014, 80, 133−152. (8) Ryerson, M. S.; Woodburn, A. Build Airport Capacity or Manage Flight Demand? How Regional Planners Can Lead American Aviation Into a New Frontier of Demand Management. Journal of the American Planning Association 2014, 80 (2), 138−152. (9) Joint Planning and Development Office. Concept of Operations for the Next Generation Air Transportation System (V 3.2); Joint Planning and Development Ofice, 2011; p 154. (10) Mahashabde, A.; Wolfe, P.; Ashok, A.; Dorbian, C.; He, Q.; Fan, A.; Lukachko, S.; Mozdzanowska, A.; Wollersheim, C.; Barrett, S. R. H.; et al. Assessing the environmental impacts of aircraft noise and emissions. Progress in Aerospace Sciences 2011, 47 (1), 15−52. (11) Airport Cooperativie Research Program. Kim, B., Nakada, K., Wayson, R., Christie, S. Understanding Airport Air Quality and Public Health Studies Related to Airports; Transportation Research Board: Washington, DC, 2015. (12) Environmental Protection Agency. General Conformity Guidance for Airports: Questions and Answers, 2002. (13) Federal Aviation Administration. Aviation & Emissions: A Primer; Office of Environment and Energy, 2005. (14) Massachusetts et al. v. Environnemental Protection Agency, Supreme Court of the United States. Certiorari to the United States Court of Appeals for the District of Columbia Circuit, 2007. (15) Meltz, R. Federal Agency Actions Following the Supreme Court’s Climate Change Decision in Massachusetts v. EPA: A Chronology; Congressional Research Service, 2014. (16) Federal Aviation Administration. Considering Greenhouse Gases and Climate Under the National Environmental Policy Act (NEPA): Interim Guidence; Washington, DC, 2012. (17) Barrett, S. R. H.; Britter, R. E.; Waitz, I. A. Global Mortality Attributable to Aircraft Cruise Emissions. Environ. Sci. Technol. 2010, 44 (19), 7736−7742. (18) Masiol, M.; Harrison, R. M. Aircraft engine exhaust emissions and other airport-related contributions to ambient air pollution: A review. Atmos. Environ. 2014, 95, 409−455. I
DOI: 10.1021/acs.est.5b04491 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Policy Analysis
Environmental Science & Technology (42) Rice, C. Restricting Use of Reverse Thrust as an Emissions Reduction Strategy. Transp. Res. Rec. 2002, 1788, 124−131. (43) Khadilkar, H.; Balakrishnan, H. Estimation of aircraft taxi fuel burn using flight data recorder archives. Transportation Research Part D: Transport and Environment 2012, 17 (7), 532−537. (44) Nikoleris, T.; Gupta, G.; Kistler, M. Detailed estimation of fuel consumption and emissions during aircraft taxi operations at Dallas/ Fort Worth International Airport. Transportation Research Part D: Transport and Environment 2011, 16 (4), 302−308. (45) Meltz, R., Federal Agency Actions Following the Supreme Court’s Climate Change Decision in Massachusetts v. EPA: A Chronology In Congressional Research Service, 2012. (46) Schrooten, L.; Vlieger, I. D.; Panis, L. I.; Torfs, R. Forecasted Maritime Shipping Emissions for Belgium with an Activity Based Emission Model. In Proceedings of the TAC-Conference, Oxford, UK, 2006. (47) Yang, D.; Kwan, S. H.; Lu, T. An Emission Inventory of Marine Vessels in Shanghai in 2003. Environ. Sci. Technol. 2007, 41 (15), 5183−5190. (48) Kim, B.; Waitz, I.; Vigilante, M.; Bassarab, R. Guidebook on Preparing Airport GHG Emissions Inventories, Airport Cooperative Research Program Report 11, Washington, DC, 2009. (49) Muller, N. Z.; Mendelsohn, R. Measuring the damages of air pollution in the United States. Journal of Environmental Economics and Management 2007, 54 (1), 1−14. (50) Muller, N. Z.; Mendelsohn, R. Efficient Pollution Regulation: Getting the Prices Right. American Economic Review 2009, 99 (5), 1714−1739. (51) Environmental Protection Agency. Guidelines for Preparing Economic Analyses; National Center for Environmental Economics Office of Policy, 2010. (52) Matthews, H. S.; Lave, L. B. Applications of Environmental Valuation for Determining Externality Costs†. Environ. Sci. Technol. 2000, 34 (8), 1390−1395. (53) Interagency Working Group on Social Cost of Carbon. Social Cost of Carbon for Regulatory Impact Analysis-Under Executive Order 12866; United States Government, 2013. (54) Bureau of Labor Statistics. Consumer Price Index (CPI); United States Department of Labor: Washington, DC, 2015. (55) Environmental Protection Agency. The Green Book Nonattainment Areas. (56) Government Accountablility Office. Aircraft Emissions Expected to Grow, but Technological and Operational IMprovements and Government Policies Can. Help Control Emissions, 2009; p 100. (57) Ryerson, M. S.; Hansen, M. The potential of turboprops for reducing aviation fuel consumption. Transportation Research Part D: Transport and Environment 2010, 15 (6), 305−314. (58) Massachusetts Institute of Technology. Airline Data Project In Global Airline Industry Program, 2015. (59) Federal Aviation Administration. Certification Activity Tracking System (CATS), 2014.08 ed. (60) Lee, J. J.; Waitz, I. A.; Kim, B. Y.; Fleming, G. G.; Maurice, L.; Holsclaw, C. A. System for assessing Aviation’s Global Emissions (SAGE), Part 2: Uncertainty assessment. Transportation Research Part D: Transport and Environment 2007, 12 (6), 381−395. (61) Azar, C.; Johansson, D. J. A. Valuing the non-CO2 climate impacts of aviation. Clim. Change 2012, 111 (3), 559−579. (62) Dorbian, C. S.; Wolfe, P. J.; Waitz, I. A. Estimating the climate and air quality benefits of aviation fuel and emissions reductions. Atmos. Environ. 2011, 45 (16), 2750−2759. (63) Deuber, O.; Luderer, G.; Sausen, R. CO2 equivalences for shortlived climate forcers. Clim. Change 2014, 122 (4), 651−664.
J
DOI: 10.1021/acs.est.5b04491 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
