Methodology to Improve AEDT Quantification of Aircraft Taxi/Idle Emissions (2016)

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.

APPENDIX A TASK 3 WORKING PAPER: Literature Review and Review of EDMS/AEDT Modeling Inputs Michael Kenney KB Environmental Sciences, Inc. St Petersburg, FL 33702

A-i Table of Contents Section Page List of Tables .......................................................................................................................... A-ii List of Figures ......................................................................................................................... A-ii Initialisms and Acronyms ...................................................................................................... A-iii 1. Introduction ................................................................................................................ A-1 2. Problem Statement Objectives ................................................................................... A-1 3. Results of the Literature and EDMS/AEDT Reviews .............................................. A-1 3.1 Relevant Literature and Research .................................................................. A-2 3.1.1 Aircraft Performance Characteristics ............................................... A-2 3.1.1.1 Taxi Process/Time in Mode ................................................ A-2 3.1.1.2 Airfield/Aircraft Operational Procedures that Reduce Emissions ............................................................................ A-3 3.1.1.3 Flight Data Recorder Data .................................................. A-4 3.1.2 Aircraft Engine Emissions .............................................................. A-9 3.1.2.1 Idle Thrust Setting (Fuel Flow Rate): ICAO Versus Actual Operations ............................................................. A-12 3.1.2.2 Inter-Engine Variability in Unburned Hydrocarbon Emissions .......................................................................... A-12 3.1.2.3 Near-Idle Unburned Hydrocarbon Emissions Sensitivity to Fuel Flow Rate and Ambient Temperature................... A-14 3.1.2.4 Hydrocarbon Speciation into Individual Compounds ...... A-15 3.1.2.5 Single (Reduced) Engine Taxi Considerations ................. A-16 3.1.3 EDMS/AEDT Performance and Development .............................. A-16 3.1.3.1 Model Architecture ........................................................... A-16 3.1.3.2 Development Programs/Timeframes ................................ A-19 3.1.3.3 Model Accuracy/Sensitivity Tests .................................... A-19 3.1.4 Regulatory Framework .................................................................. A-20 3.1.4.1 National Environmental Policy Requirements.................. A-20 3.1.4.2 Clean Air Act .................................................................... A-21 3.2 EDMS/AEDT Review .................................................................................. A-21 3.2.1 Input .................................................................................................... A-22 3.2.2 Algorithms .......................................................................................... A-23 3.2.3 Databases ............................................................................................ A-24 3.2.4 Output ................................................................................................. A-24 3.2.5 AEDT ................................................................................................. A-24 4. Issues in Need of Resolution .................................................................................... A-25 5. Go Forward Plan ...................................................................................................... A-26 Annotated Bibliography ........................................................................................................ A-28

A-ii List of Tables Number Title Page 1 Ground Speeds for Taxiing Operations .............................................................. 5 2 Engine Use – Taxi Out ........................................................................................ 6 3 Engine Use – Taxi In .......................................................................................... 6 4 Engine Operating Parameters – Taxi Out ........................................................... 7 5 Engine Operating Parameters – Taxi In .............................................................. 7 6 Engine Operating Parameters – Accelerating Aircraft ....................................... 8 7 Maximum and Minimum Modeling Parameters ................................................. 23 8 Summary of EDMS/AEDT Shortcomings Related to Taxi/Idle Emissions ....... 27 List of Figures Number Illustrates Page 1 ICAO Emission Indices for CO and NOx for a CFM56-7B24 Engine............... 10 2 Total NOx, CO, and HC Emitted by a Single CFM56-7B24 Engine ................. 11 3 ICAO 7 Percent HC Indices for Three Common Aircraft Engines .................... 13 4 Dependence of the CO, HCHO, and THC Emission Indices on Fuel Flow Rate for the CFM56-7B24 .................................................................................. 14 5 Normalized Emission Indices as a Fraction of Ambient Temperature ............... 15 6 EDMS Architecture ............................................................................................ 17 7 EDMS Taxi/Delay Datasets ................................................................................ 18 8 AEDT Architecture ............................................................................................. 18

A-iii Initalisms and Acronyms AAFEX Alternative Aviation Fuel Emissions Experiment AAM Aircraft Acoustic Module ACRP Airport Cooperative Research Program AEDT Aviation Environmental Design Tool AEE Office of Environment and Energy AEM Aircraft Emissions Module APE Aerospace Particulate Emissions APEX Aircraft Particulate Emissions Experiments APM Aircraft Performance Module APU Auxiliary Power Unit ASPM Aviation Policy's Aviation System Performance Metrics ATC Air Traffic Control AWP Amplified work plan BTS Bureau of Transportation Statistics C2H4 Ethene CAA Clean Air Act CO Carbon monoxide ECAC European Civil Aviation Conference EDMS Emissions and Dispersion Modeling System EPA U.S. Environmental Protection Agency FAA Federal Aviation Administration FDR Flight Data Recorder FID Flame Ionization Detection FSC Fuel sulfur content GSE Ground support equipment GUI Graphical Users Interface HAPs Hazardous air pollutants HC Hydrocarbons HCHO Formaldehyde ICAO International Civil Aviation Organization INM Integrated Noise Model ISA International Standard Atmosphere LTO Landing-Takeoff Cycle MAGENTA Model for Assessing Global Exposure to the Noise of Transport Aircraft NAAQS National Ambient Air Quality Standard NEPA National Environmental Policy Act NIRS Noise Integrated Routing System NMHC Nonmethane hydrocarbons NOx Nitrogen oxides PMFO Fuel organics particulate matter PMNV Nonvolatile particulate matter PMSO Volatile sulfates particulate matter ROG Reactive organic compounds RPM Revolutions per minute SAGE System for Assessing Aviation's Global Emissions SO2 Sulfur dioxide SOx Sulfur oxides THC Total hydrocarbons

A-iv TOG Total organic gases UHC Unburned hydrocarbons US United States VOC Volatile organic compounds

A-1 1. Introduction Consistent with the Amplified Work Plan (AWP) approved by the Project Panel, this Task 3 White Paper outlines the scope and preliminary results for the Task 1 Literature Review and Task 2 Emissions Dispersion Modeling System (EDMS)/Aviation Environmental Design Tool (AEDT) Review elements of the Airport Cooperative Research Program (ACRP) 02-45 Research Project. For both tasks, an overview of the collected literature and information are provided and discussed in the context of how the collected material will aid the completion of subsequent research tasks. An annotated bibliography is also provided after Section 5 (Plan Going Forward) of this Working Paper. 2. Problem Statement Objectives There is a need to improve the current method of estimating air pollutant and pollutant precursor emissions from aircraft engines during the taxi/idle mode of operation. More specifically, there is a need to evaluate whether the emissions indices (i.e., factors) that are used to derive emissions inventories and used by EDMS and AEDT appropriately reflect real world emission rates. To address this need, the following two objectives were developed by the Project Panel for this ACRP 02-45 Research Project: • Develop a prioritized list of potential improvements to EDMS and AEDT that will increase the predictive accuracy of these tools to estimate commercial jet aircraft emissions during the taxi/idle phase of operation; and • Prepare documentation that highlights and describes high priority improvements that should be accomplished in the near term. The initial two tasks to meet the above objectives were to review and summarize relevant scientific and industry literature, published guidance, and pertinent research on the combined topics of aircraft taxi/idle emissions and EDMS/AEDT (i.e., the Literature Review) and to review the relevant assumptions, algorithms, database coverage, and outputs of both EDMS and AEDT (the EDMS/AEDT Review) The results of these tasks are presented in this Working Paper with the intent that the information will serve as the foundation from which the ACRP 02-45 project will proceed. 3. Results of the Literature and EDMS/AEDT Reviews Because the corresponding answers are necessary to develop a prioritized list of potential improvements to EDMS/AEDT, the Research Team was mindful of the following questions during the course of the literature and EDMS/AEDT reviews: • What are the factors that affect an aircraft departing an airport from the time the aircraft leaves the airport’s gate until the aircraft reaches a runway (i.e., taxi out)? • What are the factors that affect an aircraft arriving at an airport from the time the aircraft exits a runway until the aircraft reaches a gate (i.e., taxi in)? • What thrust settings do pilots use for aircraft engines during the taxi process? • Are there readily available air pollutant/pollutant precursor emission indices for engine thrust settings other than those already in the EDMS/AEDT databases? • Do the current databases, input variables, algorithms and sub-models of EDMS/AEDT provide a reasonable estimate of taxi-related emissions? The following summarizes the Research Team’s knowledge and findings from the Task 1 Literature Review and Task 2 EDMS/AEDT Review.

A-2 3.1 Relevant Literature and Research The literature review focused on the following four categories of information: • Aircraft Performance Characteristics – including aircraft taxi/idle times-in-modes under alternative airfield conditions, single-engine taxi procedures, flight data recorder (FDR) data, etc.; • Aircraft Engine Emissions – including International Civil Aviation Organization (ICAO) reference fuel flows and emission indices, aircraft engine emissions and ambient measurements, etc.; • EDMS/AEDT Performance and Development – including model architecture, development programs and timeframes, model accuracy and sensitivity tests, etc.; and • Regulatory Framework – including the Federal Aviation Administration’s (FAA’s) rules/regulations that relate to the taxi process and the requirements of the National Environmental Policy Act (NEPA), the federal Clean Air Act (CAA), and other state/local requirements as they pertain to airport emissions and ambient air quality. The literature and research review results are discussed below. The results were also compiled in an annotated bibliography that is provided at the end of this Working Paper. Within this text, references to the literature included in the bibliography are enclosed inside brackets ( […] ). 3.1.1 Aircraft Performance Characteristics Aspects of an aircraft’s performance while in the taxi mode that are relevant to emissions include factors which affect the amount of time an aircraft spends on taxiways and in hold areas, and airfield/aircraft operational procedures that reduce aircraft engine emissions. These aspects are discussed in the following subsections. A source of real-world aircraft performance data (i.e., FDRs) is also discussed. 3.1.1.1 Taxi Process/Time in Mode Without considering environmental influences (e.g., temperature), an aircraft’s total taxi-related emission load depends on aircraft type, the amount of time the aircraft spends in taxi, the duration of aircraft delay and the power settings (i.e., thrust) of each of the aircraft’s engines. Several of the literature sources describe the aircraft taxi out and taxi in process. From these sources, the following taxi process descriptions were developed [FAA, 2013b; Khadilkar et al, 2012; Bhadra et al, 2011; Page, 2013; Grinspun, 2002; Boeing, 2002]: • Taxi out – The taxi out process begins in one of two ways–1) an aircraft’s engine(s) is started at a gate and a pilot begins taxiing using the aircraft’s own power or, 2) for nose-in gates, a pushback tug is used to back the aircraft out of the gate, the engines are started and the aircraft begins taxiing after it is disconnected from the pushback tug. These initial activities occur in what is referred to as the nonmovement area of an airport. The movement of aircraft in the nonmovement area is the responsibility of the pilot, the airport operator or airport management. When exiting the nonmovement area, Air Traffic Control (ATC) issues an aircraft’s pilot a route that is to be followed to the departure end of an airport’s runway (i.e., the movement area). Along this route, the pilot may be instructed to “hold short” at any point (e.g., a runway other than the departure runway, a taxiway). Depending on the departure demand rate at an airport, pilots may also be instructed to hold short of the departure runway in a “departure queue”. Key influences on the taxi out process include the taxi path, the taxi distance (i.e., as the taxi distance increases, the chances of having more holds increases because of the likelihood of a taxiing aircraft having to negotiate intersections and other taxiing aircraft), the weather (which could include the time spent to deice an aircraft), and required separation distances between departures. Generally, the taxi out process is comprised of the following five components:

A-3 − Pushback − Unimpeded taxi − Route delay − Runway queue delay − Deicing (when applicable) • Taxi in – The taxi in process begins when an aircraft exits a runway and a pilot begins taxiing along an assigned taxi route. Along this route, a pilot may also be instructed to hold short at any point due to factors such as airfield congestion or the weather. An aircraft may also be delayed in the nonmovement area as pilots are instructed to hold short of an assigned gate that is occupied by another aircraft. The taxi in process is comprised of three components: − Unimpeded taxi − Route delay − Gate hold delay Throughout the taxi out and taxi in processes, pilot’s will turn an aircraft, decelerate and accelerate an aircraft’s taxi speed, move at a constant speed, brake, and stop with the thrust settings for the aircraft’s engine(s) varying according to the process component. For commercial service jets, aircraft manufacturers recommend taxi procedures and speeds in Flight Crew Training Manuals. A manual prepared for the Boeing 737-300/400/500 series aircraft states the following regarding taxi speeds, thrust and braking: • “To begin taxi, release brakes, smoothly increase thrust to minimum required for airplane to roll forward, then reduce thrust to idle.” • “Normal taxi speed is approximately 20 knots, adjusted for conditions. On long straight taxi routes, speeds up to 30 knots are acceptable…” • “Because of additional operational procedural requirements and crew workload, taxiing out for flight with an engine shut down is not recommended.” Taxi-out and taxi-in delays are metrics reported by the FAA’s Office of Aviation Policy’s Aviation System Performance Metrics (ASPM). The taxi delays are computed as the difference between the taxi- out (or in) duration and the unimpeded taxi time. For the purpose of the ASPM, the unimpeded taxi time is estimated by regression equations (i.e., the times are not directly observed from surveillance data). 3.1.1.2 Airfield/Aircraft Operational Procedures that Reduce Emissions As a method of reducing fuel-burn, airlines promote the practice of taxiing with less than all of an aircraft’s engines operating. Although this practice is generally referred to as “single engine taxiing”, it can involve more than one of an aircraft’s engines operating. There are three categories of considerations for pilots making a decision to single-engine taxi. The categories are crew workload, aircraft systems implications and breakaway thrust levels. These categories are briefly described below [UK Dept. of Transport, 2012]: • Crew workload – So that airfield congestion is avoided, an aircraft taxiing with fewer engines must be able to taxi at the same speed that would be possible with all engines operating. In order to do so, additional systems checks requiring the attention of the flight crew may be necessary. • Systems implications – Taxiing with fewer engines can provide less or degraded power to some systems on an aircraft which would require the operation of an aircraft’s auxiliary power unit (APU). Even if all systems have sufficient power, use of an APU may be necessary in case there is a systems failure.

A-4 • Breakaway thrust levels – The thrust level necessary to start moving or to move a heavy aircraft cannot be such that it creates a jet blast risk to other aircraft. The increased thrust of the operating engine may also increase the potential for debris to be picked up from the ground. Other considerations include the surface condition of the taxipath (e.g., it could be slippery), excessive taxiway slopes, congested maneuvering areas, and the need for full power for runway crossings. Notably, although promoted by most airlines, single-engine taxiing is performed only at the discretion of an aircraft’s pilot and, all engines typically need to be running for three to five minutes prior to an aircraft’s departure and for up to five minutes after arriving to cool down. Recognizing that the reduced use of an aircraft’s engines on the ground could result in significant fuel savings, at least two manufacturers/joint ventures are developing methods to taxi aircraft on the ground with no engines operating. Honeywell, and the aerospace firm Safran, have been working together to develop an electric motor that would power an aircraft’s main wheels that would be powered by an aircraft’s APU. The Honeywell/Safran product is known as the Electric Green Taxiing System (EGTS) and it is reported by the joint venture that the product will be available in the year 20171 [AOPA 2011; Franc24, 2013]. Wheeltug, a system being developed by a subsidiary of Borealis Exploration Limited, involves an electric motor in the hub of the nose wheel.2 This motor would also be powered by an aircraft’s APU. As recently reported by Wheeltug, they expect to begin installing their system onto aircraft belonging to 11 airlines, including six flagship carriers, in the near term.3 An airlines acceptance of an electric taxi system could depend on the cost of the system and whether or not the taxi-related fuel savings offsets the increased fuel consumption to carry the systems added weight during flight. A study performed to evaluate the operational and environmental benefits of electric taxi excluded heavy aircraft (i.e., aircraft heavier than an Airbus A321) because the analysts expected that the additional weight of the motors and resultant fuel consumption during a long cruise would more than offset the reduction in fuel usage during these aircraft’s ground movements. And, while the analysts concluded that the electric motors would make a valuable contribution to reducing emissions, they report that the greatest benefit would result from the systems being installed on aircraft that connect large airports that are located close to each other but have long taxi times [Wollenheit et. al., 2013]. At congested airports, ATC may use procedures that also result in emission savings. These procedures are often referred to as virtual queuing procedures because they control the rate at which aircraft are pushed backed from a gate on departure (i.e., departure management) without an aircraft losing its place in an airport’s departure queue. Assuming another aircraft does not need a gate, keeping a departing aircraft at a gate, as opposed to the aircraft waiting in the movement area, saves fuel and reduces emissions [Bhadra et al, 2011; Simaiakis et al, 2010; Baik, UNK]. 3.1.1.3 Flight Data Recorder Data The best source of aircraft performance data are FDRs which are also known as “black boxes”. FDRs record any instructions (i.e., thrust settings) that are sent to any electronic system on an aircraft. Unfortunately, due to the proprietary nature of the data on FDRs, the data is not readily available for research purposes. However, Research Team members have FDR data from which summary results can be used for the purpose of the Project. Team members also have access to the raw data that has been collected as part of several field campaigns, including Aircraft Particulate Emissions Experiments (APEX), JETS-APEX2, APEX3, ACRP 02-03A, and the Alternative Aviation Fuel Experiment (AAFEX) 1 Electric Green Taxiing System overview - www.greentaxiing.com/overview.html 2 Wheeltug - www.wheeltug.gi/ 3 Royal Aeronautical Society, Toulouse, 13 September 2013. Delivered by Isaiah Cox, WheelTug Chief Executive Officer

A-5 projects. Notably, much of the data has been published in peer-reviewed literature [e.g., Yelvington et al. 2007] or as an ACRP report [Herndon et al. 2012]. ACRP’s 11-02 Task 8 Research Project, Enhanced Modeling of Aircraft Taxiway Noise (Page et al. 2009) examined FDR information from a major European airline and the airline’s affiliate regional carriers from which statistical generalities were developed regarding aircraft taxi operations [Page et al, 2013b]. The FDR data included one year of operational data from “gate to runway to air to runway to gate”, for a multitude of international airport pairs. Although from a European airline, the data is reported to have included information for some United States (U.S.) airports and was considered generally applicable to U.S. airports. The 11-02 research project assessed the taxi out and taxi in components separately. Operations at a gate while engines were spooling down were not included in the arrival taxi segment and stationary segments were defined as those with reported ground speed less than one knot. Data from Tables 1 through 4 of the Volume 1 (Scoping) report for ACRP’s 11-02 are provided below in Tables 1 through 6 of this Working Paper. Notably, there are currently no RJ100 aircraft and only one RJ85 aircraft registered to a domestic airline. This fact will be considered by the Research Team in the evaluation/use of this available FDR information. Table 1 Ground Speeds for Taxiing Operations Aircraft Departures Arrivals Average Speed (knots) Standard Deviation (knots) Average Speed (knots) Standard Deviation (knots) A319 9.26 3.34 11.72 3.27 A320 9.10 2.92 11.08 3.27 A321 9.39 3.31 11.28 4.67 A330 10.05 3.32 13.07 3.21 A340 9.26 2.98 9.88 2.92 B757 8.87 2.28 13.23 2.68 B767 11.13 3.13 12.65 2.60 B777 8.97 3.18 11.45 2.26 RJ100 9.14 3.57 14.10 4.44 RJ85 8.23 3.08 14.67 4.77 Source: Web-Only Document 9: ACRP Project 11-08 Task 8, June 2009

A-6 Table 2 Engine Use - Taxi Out Aircraft Percent of Time Number of Engines Were Operating One Engine Two Engines Three Engines Four Engines Average Standard Deviation Average Standard Deviation Average Standard Deviation Average Standard Deviation Stationary A319 2.4 5.47 97.4 5.68 -- -- -- -- A320 3.3 7.62 96.6 7.85 -- -- -- -- A321 2.1 3.56 97.8 3.85 -- -- -- -- A330 9.6 14.93 90.3 15.12 -- -- -- -- A340 1.2 7.25 7.1 15.86 3.2 5.7 88.2 21.51 B757 5.1 5.58 94.7 5.86 -- -- -- -- B767 17.1 15.24 82.7 15.38 -- -- -- -- B777 7.3 14.01 92.5 14.14 -- -- -- -- RJ100 0.0 -- 0.0 -- 0.0 -- 100.0 0.00 RJ85 0.0 -- 0.0 -- 0.0 -- 100.0 0.00 Moving A319 11.2 19.54 88.6 19.41 -- -- -- -- A320 9.3 18.39 90.6 18.27 -- -- -- -- A321 8.7 16.10 91.1 15.99 -- -- -- -- A330 11.3 14.72 88.5 14.75 -- -- -- -- A340 1.8 2.98 7.2 8.74 1.4 1.88 89.2 10.83 B757 2.8 4.51 97.1 4.76 -- -- -- -- B767 6.6 9.73 93.3 9.90 -- -- -- -- B777 10.5 10.87 89.4 10.92 -- -- -- -- RJ100 0.0 -- 0.0 -- 0.0 -- 100.0 -- RJ85 0.0 -- 0.0 -- 0.0 -- 100.0 -- Source: Web-Only Document 9: ACRP Project 11-08 Task 8, June 2009 Note: -- = Not applicable Table 3 Engine Use - Taxi In Aircraft Percent of Time Number of Engines Were Operating One Engine Two Engines Three Engines Four Engines Average Standard Deviation Average Standard Deviation Average Standard Deviation Average Standard Deviation Stationary A319 3.0 15.13 98.9 15.21 -- -- -- -- A320 1.1 7.63 98.8 7.77 -- -- -- -- A321 0.0 0.00 100.0 0.00 -- -- -- -- A330 4.4 19.42 95.5 19.50 -- -- -- -- A340 1.0 8.60 1.7 10.36 1.2 10.7 95.9 19.35 B757 1.2 10.87 98.7 10.87 -- -- -- -- B767 0.6 5.39 99.3 5.44 -- -- -- -- B777 0.0 -- 100.0 0.00 -- -- -- -- RJ100 0.0 -- 0.0 -- 0.0 -- 100.0 -- RJ85 0.0 -- 0.0 -- 0.0 -- 100.0 -- Moving A319 0.4 2.85 99.5 2.99 -- -- -- -- A320 1.1 6.28 98.7 6.46 -- -- -- -- A321 0.0 0.53 99.9 0.62 -- -- -- -- A330 1.7 11.23 98.2 11.36 -- -- -- -- A340 0.5 3.05 1.7 10.09 0.1 0.68 97.5 11.04 B757 0.4 3.60 99.5 3.72 -- -- -- --

A-7 B767 0.5 7.36 99.4 7.36 -- -- -- -- B777 0.0 0.18 99.9 0.25 -- -- -- -- RJ100 0.0 -- 0.0 -- 0.0 -- 100.0 -- RJ85 0.0 -- 0.0 -- 0.0 -- 100.0 -- Source: Web-Only Document 9: ACRP Project 11-08 Task 8, June 2009 Note: -- = Not applicable Table 4 Engine Operating Parameters - Taxi Out Aircraft Average N1 average Standard Deviation N1average Average Percent Thrust Stnd Dev Percent Thrust Average EMS Thrust Stnd Dev EMS Thrust Average EMS Enhanced Standard Dev EMS Enhanced Stationary A319 19.42 1.41 8.41 1.18 1975.24 278.42 1975.24 278.42 A320 19.16 1.29 7.45 1.34 2011.77 361.01 2011.77 361.01 A321 20.06 1.55 6.15 1.15 1843.61 345.29 1843.61 345.29 A330 21.16 3.08 N/A N/A 3845.13 2484.18 N/A N/A A340 19.50 4.73 N/A N/A 2210.70 1027.83 N/A N/A B757 20.47 1.51 2.68 0.67 1077.75 268.58 1077.75 268.58 B767 23.78 2.66 5.74 1.24 3565.83 772.65 N/A N/A B777 19.89 2.30 4.86 0.86 5615.41 989.00 N/A N/A RJ100 22.67 1.80 21.70 1.99 1518.85 139.30 N/A N/A RJ85 22.32 1.59 21.44 1.71 1500.55 120.04 N/A N/A Moving A319 19.56 3.34 9.20 1.92 2162.66 451.69 2162.86 451.69 A320 19.71 3.29 8.22 1.85 2220.48 498.40 2220.48 498.40 A321 20.32 3.13 6.89 1.43 2066.97 429.37 2066.97 429.37 A330 22.28 3.30 N/A N/A 4261.80 2792.32 N/A N/A A340 20.45 4.08 N/A N/A 2407.10 1008.95 N/A N/A B757 23.28 2.20 3.73 1.01 1500.41 405.07 1500.41 405.07 B767 26.14 1.71 6.58 1.14 4085.75 708.94 N/A N/A B777 20.08 1.94 5.16 0.77 5960.23 884.92 N/A N/A RJ100 25.49 2.45 24.61 2.47 1722.56 172.80 N/A N/A RJ85 24.59 2.11 23.85 2.38 1669.50 166.76 N/A N/A Source: Web-Only Document 9: ACRP Project 11-08 Task 8, June 2009 N/A = Not available (Source states that some A330/A340 departure values were erroneous in the FDR database. Table 5 Engine Operating Parameters - Taxi In Aircraft Average N1 average Standard Deviation N1average Average Percent Thrust Stnd Dev Percent Thrust Average EMS Thrust StndDev EMS Thrust Average EMS Enhanced Standard Dev EMS Enhanced Stationary A319 17.37 3.98 8.62 2.20 2026.47 516.45 2026.47 516.45 A320 17.51 3.04 7.72 1.89 2086.66 510.60 2083.66 510.60 A321 18.21 3.82 6.78 1.97 2034.77 591.40 2034.77 591.40 A330 21.51 4.25 5.80 4.46 3947.56 3035.05 2815.96 3582.02 A340 19.20 3.93 6.38 2.85 2590.74 839.84 1516.45 1520.39 B757 19.64 2.50 1.39 0.87 560.33 347.74 560.33 347.74 B767 26.79 1.56 6.09 4.47 3781.28 2777.72 N/A N/A B777 21.41 0.91 5.46 0.43 6312.84 496.08 N/A N/A RJ100 17.51 6.52 16.54 6.41 1157.97 449.01 N/A N/A RJ85 18.03 5.69 17.06 5.45 1194.22 381.76 N/A N/A

A-8 Moving A319 19.94 1.05 9.89 0.70 2323.04 164.33 2323.04 164.33 A320 19.62 1.38 8.70 1.32 2350.02 355.09 2350.02 355.09 A321 20.72 1.55 7.62 0.57 2284.95 169.66 2284.95 169.66 A330 23.15 2.23 6.56 4.26 4459.99 2892.61 2886.91 3742.11 A340 20.03 2.55 7.04 2.54 2862.06 555.44 1672.96 1585.15 B757 22.29 1.79 3.34 0.61 1341.46 245.01 1341.46 245.01 B767 27.17 2.13 6.65 0.89 4130.46 549.77 N/A N/A B777 21.53 0.45 5.48 0.31 6336.52 359.74 N/A N/A RJ100 23.84 6.19 22.85 6.16 1599.68 431.08 N/A N/A RJ85 23.44 6.67 22.55 6.63 1578.58 463.89 N/A N/A Source: Web-Only Document 9: ACRP Project 11-08 Task 8, June 2009 N/A = Not available (The source reports that some A330/A340 departures values were erroneous in the FDR database. Table 6 Engine Operating Parameters – Accelerating Aircraft Arrival/ Departure Aircraft Avg Accel Time (s) Avg N1 average Avg N1 Max Avg Avg Percent Thrust Avg max Percent Thrust Avg. Max Long Accel (g) # Events Burst Arrival A319 5.00 14.88 15.58 8.01 8.50 0.29 2 A320 8.33 18.40 20.45 7.21 8.44 0.03 6 A321 7.00 18.56 18.71 6.88 6.94 0.03 5 A330 7.31 21.89 23.80 5.82 6.27 0.01 13 A340 9.50 22.77 25.88 8.07 9.69 0.03 10 B757 6.68 18.85 20.50 N/A N/A 0.02 95 B767 8.00 26.10 27.31 4.61 11.88 0.01 25 B777 9.27 21.04 22.10 5.23 5.73 0.01 41 Departure A319 6.63 28.24 36.07 13.74 21.24 0.20 92 A320 7.07 27.93 35.79 13.74 18.68 0.20 121 A321 6.07 28.80 35.45 12.74 16.69 0.18 61 A330 8.00 40.07 54.75 10.44 16.58 0.15 95 A340 7.50 29.52 41.41 8.57 14.04 0.14 34 B757 7.27 35.58 44.34 10.29 14.30 0.15 75 B767 8.17 41.16 57.39 14.19 28.82 0.18 71 B777 8.17 31.14 40.96 9.45 13.02 0.15 101 Gentle Arrival A319 104.41 17.85 26.12 8.71 14.17 0.02 17 A320 115.47 15.32 23.89 6.64 11.30 0,03 43 A321 54.76 22.08 27.72 7.89 11.41 0.03 21 A330 32.20 24.32 29.84 8.76 11.49 0.02 25 A340 29.44 23.28 29.09 8.50 11.40 0.02 18 B757 12.41 19.70 21.94 N/A N/A 0.02 106 B767 41.84 21.94 27.73 4.55 16.35 0,02 98 B777 17.50 21.19 22.66 5.24 5.89 0.01 54 Departure A319 71.53 23.07 27.90 10.87 14.49 0.02 334 A320 70.02 22.46 26.77 9.42 12.22 0.03 547 A321 70.63 23.39 28.18 8.22 11.53 0.03 248 A330 59.65 27.27 32.93 6.46 9.18 0.02 103 A340 48.94 23.38 31.10 5.22 8.68 0.03 17 B757 67.99 27.38 33.84 5.14 8.95 0.04 296 B767 74.26 27.64 29.86 6.84 22.62 0.02 291

A-9 B777 68.46 22.89 25.80 5.98 7.16 0.02 343 Source: Web-Only Document 9: ACRP Project 11-08 Task 8, June 2009 N/A = Not available (The source reports that some A330 and A340 Departures values were erroneous in the FDR database and were removed). Table 1 lists the ground speed of the departing and arriving aircraft for which FDR data were reviewed. As shown, during taxi out the large jet taxi speed ranged from 8.87 to 11.13 knots (10.21 to 12.81 miles per hour) and during taxi in the speed of these aircraft ranged from 9.88 to 13.23 knots (11.37 to 15.2 miles per hour). Additionally, while the speed of the RJs was similar during taxi out (an average of 8.69 knots (10 miles per hour)) the taxi in speed of these aircraft was higher than the large jets (14.39 knots (16.56 miles per hour)). Tables 2 and 3 provide the percentage of time that each aircraft operated with all (or less than all) engines operating during taxi out and taxi in, respectively. As shown, during taxi out the A319 aircraft had both engines operating 97.4 percent of the time while the B767 only had both engines operating 82.7 percent of the time. Of note also is that the RJs had all four engines operating at all times during the taxi out process. Tables 4 and 5 report engine operating parameters for the taxi out and in process when the aircraft were both stationary (i.e., moving less than one knot) and moving. The parameters include the average rotational speed of the engine spool, the average engine thrust, the average absolute thrust measurement and the recommended average thrust to ensure the most efficient operation at the time of the measurement. As shown, during stationary departure operations, the engines on the A319 had an average rotational speed of 19.42 revolutions per minute (RPM)18 plus or minus 1.41 percent and the engines operated at 8.41 percent thrust (plus or minus 1.18 percent). During the measurement period, the absolute thrust of the engines was 1,975.24 kilonewtons (plus or minus 278.42 kilonewtons) and a different thrust was not recommended. Table 6 provides the engine operating parameters for the aircraft during acceleration. As shown, these data are available for both taxi out and taxi in and for two distinct operation types: 1) short bursts of acceleration (i.e., 15 seconds or less) and 2) gentle, longer and slower accelerations. The documentation for ACRP’s 11-02 research project also provides the following aircraft taxi speed information for which the source was FDR information: • The taxi study conducted for the ACRP 11-02 indicates that typical ground speeds range from 9 to 16 knots with a standard deviation of 3 to 5 knots [Page et al, 2013]. • Analysis of data published by the University of Madrid indicates that most measured aircraft move at a constant speed ranging from 15 to 23 knots with an average taxi speed of 19.8 knots [Page et al, 2013b]. • The ACRP 02-27 project for which results were published in January of 2013 indicates a reference ground taxi speed of 16 knots [Page et al, 2013b]. Other studies are available that use FDR information [e.g., Khadilkar et al, 2012] but the reported values are not relevant to the ACRP 02-45 Research Project. 3.1.2 Aircraft Engine Emissions The ICAO aircraft LTO cycle comprises four modes: idle/taxi, take-off, and climb-out approach. Total emissions of a particular pollutant (e.g., carbon monoxide) per engine per LTO mode are calculated by the product of the fuel-based emission index (grams of pollutant emitted per kilogram of fuel burned), the 18 This value may instead represent a percentage of thrust. Additional research is needed to confirm which unit is appropriate when interpreting the values in Tables 4 and 5.

A-10 fuel flow rate (kilogram of fuel per second), and the total time in mode (in seconds). For example, CO emissions from an engine during the idle/taxi component of an LTO are calculated by the following equation: CO EmissionsIDLE (g) = CO EIIDLE (g/kg) × FFRIDLE (kg/s) × TIMIDLE (s) where: EI = emissions index FFR = fuel flow rate TIM = time-in-mode g = grams kg = kilograms s = seconds Total CO emissions from this engine during an entire LTO are then calculated by adding the emissions from each of the four LTO modes. The aircraft emission indices in the current EDMS database that are used to create emission inventories are based on ICAO certification tests performed at four thrust levels corresponding to the four components of a LTO cycle: idle/taxi (7 percent thrust), take-off (100 percent thrust), climb-out (85 percent thrust) and approach (30 percent thrust). Each thrust setting corresponds to an engine-specific fuel flow rate (with variations determined, among other variables, by “bleed flow” used for auxiliary uses such as air conditioning). For example, using the ICAO default fuel flow rates, time-in-modes, and emission indices, each CFM56-7B24 emits 3,741 grams of CO during the idle/taxi mode: (22 g CO per kg of fuel) × (0.109 kg fuel / second) × (1560 seconds idling). As discussed further below, a general trend common to all engines is that CO and hydrocarbon (HC) emission indices are highest at low thrust settings (because CO and HC result from incomplete combustion of fuel) while nitrogen oxide (NOx) emission indices are highest at high thrust settings (because NOx is mainly created from the high temperature oxidation of atmospheric nitrogen). Figure 1 depicts the CO and NOx emission indices at each of the four components of a LTO cycle for a CFM56- 7B24 engine, one of the most common engines in the commercial fleet. Figure 1. ICAO Emission indices for CO and NOx for the CFM56-7B24 engine. HC emission indices show the same general trend as CO – highest during idle/taxi (low thrust, low fuel flow rate) and lowest at take-off (high thrust and fuel flow rate).

A-11 Because the emission indices, fuel flow rates, and time-in-modes all vary with LTO phase, all three must be considered when calculating total emissions per LTO as per above equation (total LTO emissions being calculated using the appropriate emissions index, fuel flow rates, and time in mode for each of the operational modes). Figure 2 shows the total LTO emissions of CO, NOx, and HC from a CFM56-7B24 engine using the default ICAO values for fuel flow rate, emission indices, and time-in-mode. Evident from Figure 2 is that the idle/taxi mode is responsible for the largest portion of CO and HC emissions by far – because both the emission indices and time-in-mode are highest during idle/taxi, even though the fuel flow rate is lowest. Conversely, most NOx emissions result from the take-off and climb-out phases since both the fuel flow rate and NOx emission indices are highest at high fuel flow rates, even though the total time spent during these two modes is small compared to the idle/taxi mode. The idle/taxi mode accounts for approximately 15 percent of total NOx emissions. Figure 2. Total emissions of NOx, CO, and HC emitted by a single CFM56-7B24 engine during a LTO cycle, calculated using the default time-in-modes and ICAO fuel flow rates and emission indices. The vast majority of CO and HC emissions occur during the taxi/idle mode. These numbers all vary in actual use, however, so using these default numbers can lead to inaccuracies in emission inventories. Time-in-mode values can vary significantly from the default values, and airport- specific values based on actual data and/or the Delay and Sequence module can be input into EDMS. The only way to deviate from the default flue flow rates and emission indices is to create “custom aircraft” (discussed further in Section 3.2.1 of this Working Paper). As initially discussed in Section 3.1.1 and repeated below, actual thrust settings and fuel flow rates can actually vary significantly from 7 percent thrust for the idle/taxi phase. The emission indices for CO and HC depend on the fuel flow rate, and are much greater at lower thrust settings (e.g., 4 percent versus 7 percent thrust). Additionally, CO and HC emission indices are very sensitive to the ambient temperature (i.e., low temperatures lead to increased

A-12 CO and HC emission indices). The rest of this section describes our understanding of emissions from the idle/taxi phase with an emphasis on HC and CO emissions. 3.1.2.1 Idle Thrust Setting (Fuel Flow Rate): ICAO Versus Actual Operation Although unburned HC (UHC) / volatile organic compound (VOC) emissions are currently based on certification tests performed at 7 percent thrust, there is strong evidence that this single thrust value does not accurately reflect true operating conditions. This evidence is based both on FDR data, which directly records fuel flow rate, and comparison of pollutant emission indices from advected plumes of in-use aircraft to emissions data collected at known thrust and fuel flow rate settings. Evidence to date (partially listed below) indicates that while stationary or moving at a constant speed, aircraft operate at thrust values closer to ~4 percent thrust (“ground idle”) and accelerations / turning are associated with higher values that at times exceed 7 percent thrust. • FDR data - Examination of FDR is the most straightforward method to assess what thrust settings / fuel flow rates are actually used in day-to-day practice. For example, FDR data from an A320 during the idle/taxi phase show that during most of the idle/taxi period, the engines operated at a fuel flow rate of ~0.10 kilograms/second, punctuated by occasional “bursts” of 0.14 to 0.17 kilograms/second. The value of 0.10 kilograms/second is 20 percent lower than the ICAO 7 percent fuel flow rate of 0.12 kilograms/second (Figure V-4 of ACRP report 63, project 02- 03a), and the higher values are 17 to 40 percent higher. Similarly, FDR data from a CFM56-7BX engine in Turkey [Turgut et al, 2013] show that fuel flow rate during the idle phase was usually 0.09 kilograms/second with occasional bursts up to 0.17 kilograms/second, compared to the ICAO 7 percent value of 0.11 kilograms/second. Additional corroborating evidence can be found in the work of Patterson et al [1999], Khadilkar and Balakrishnan [2012], and Nikoleris et al. [2011] among others. • Comparison of Staged to In-Use Emission Indices - During several aircraft emissions measurements studies (e.g., the Aircraft Particulate Emissions eXperiments – APEX), emissions were characterized in two different experimental set-ups: 1) with a stationary aircraft operating at exactly known parameters (e.g., fuel flow rate), and 2) measuring diluted advected plumes downwind of the actual in-use aircraft. The first method provides well defined relationships between pollutant emission indices (e.g., for CO, NOx, individual VOCs, etc) that can be used during the second method to infer the actual fuel flow rate. For example, NOx emission indices decrease with increasing thrust value. During the JETS-APEX2 study, [Wood et al, 2008) it was observed that NOx emission indices from idling B737 aircraft at Oakland International Airport were usually lower than the NOx emission indices for 7 percent thrust operation observed during staged tests. The observed emission indices were more consistent with operation at 4 percent thrust with occasional bursts at approximately 15 percent thrust, in agreement with the FDR findings described above. Similar results have been presented by several research teams [Herndon, 2009; Mazaheri, 2009; Schäfer, 2003]. 3.1.2.2 Inter-Engine Variability in Unburned Hydrocarbon Emissions Differences in the 7 percent ICAO HC emission indices between different types of engines (e.g., CFM56 vs. RB211) can easily exceed a factor of 10. As shown, Figure 3 compares the ICAO HC emission indices for three different engines: a V2527 (commonly used in the Airbus A320), a CFM56-7B24 (commonly used in Boeing B737), and the CF6 (commonly used on McDonnell Douglas DC-10).

A-13 Figure 3. ICAO 7 percent HC emission indices for three common aircraft engines.

A-14 3.1.2.3 Near-Idle Unburned Hydrocarbon Emissions Sensitivity to Fuel Flow Rate and Ambient Temperature. Thrust settings lower than the ICAO certification value (7 percent thrust) and temperatures lower than the certification temperature (15 degrees Celcius) lead to increases in CO and HC emission indices. This was observed during the first APEX project [Yelvington et al, 2007] and studied in great detail during ACRP 02-03a, Measurement of Gaseous HAP Emissions from Idling Aircraft as a Function of Engine and Ambient Conditions. These two effects are described separately in the two following sections: • Fuel Flow Rate - As described earlier, evidence to date suggests that 7 percent thrust overestimates the true thrust levels used most of the time by idling aircraft. True “ground” idle appears to be lower than 7 percent thrust (approximately 4 percent), and accelerations result in thrusts that exceed 7 percent. Although these differences in thrusts are associated with seemingly small changes in fuel flow rates, the effects on HC and CO emissions are large. Figure 4 shows the fuel flow rate effect on the emission indices of three pollutants for the CFM56-7B24 engine using data from ACRP 02-03a: CO, formaldehyde (HCHO) and total HC. Although 0.09 kilograms/second is only 14 percent lower than 0.105 kilograms/second, it approximately doubles the emission index for all three pollutants. Note that this does not mean that the emission rate (in grams per second) is twice as high – the decrease in the fuel flow rate partially offsets the higher emission index. Rather, the increase in emission rate is approximately a factor of approximately 1.7. Figure 4. Dependence of the carbon monoxide, formaldehyde, and total HC emission indices on fuel flow rate for the CFM56-7B24. (From Figures A01-E8b and A01-E8d of ACRP Report 63, Project 02- 03a.)

A-15 • Ambient Temperature - The ICAO “reference temperature” for engine certification is 15 degrees Celsius. Ambient temperatures at airports, of course, span a large range of values. Figure 5 summarizes the relative emission indices for HCHO and ethene (C2H4) – two important and representative components of total HC – as a function of ambient temperature, using data from the ACRP 02-03a and APEX1 projects. The relative increase of both pollutants is approximately a factor of two at temperatures just below freezing (i.e., the true HCHO and C2H4 emission indices are twice as high at approximately -2° C as they are at 15 degrees Celsius). Conversely, the emission indices at 27 degrees Celcius (80 degrees Fahrenheit) are approximately half the 15 degrees Celsius values. Figure 5. Normalized emission indices as a function of ambient temperature (from Figure IV-3 of ACRP report 63, Project 02-03a). Compared to the ICAO certification emission indices (at 15 °C), actual HC emission indices are doubled at ~-2 degrees Celsius and halved at 27 degrees Celsius. These relationships between HC emission index and ambient temperatures and fuel flow rates held true for the limited but representative sample of engines studied during the ACRP 02-03a and APEX campaigns. These two parameters (i.e., temperature and fuel flow rate) both act simultaneously – the true HC emission index for an engine that is idling both at very cold temperatures and at sub-7 percent thrust is determined both by the “temperature effect” and by the “fuel flow rate effect”. For example, for a CFM56-7B22 engine, operation at 0.09 kilograms/second (versus the ICAO value of 0.105 kilograms/second) and operation at an ambient temperature of 2 degrees Celsius (versus the ICAO 15 degrees Celsius value) results in a factor of three increase in HCHO emission index. 3.1.2.4 Hydrocarbon Speciation into Individual Compounds There are ICAO certification emission indices for CO, NOx, sulfur dioxide (SO2,) total HC, and smoke number, but not for individual VOCs or hazardous air pollutants (HAPs). Similarly, EDMS/AEDT generate total HC emissions but not speciated VOC emissions. Nevertheless, given the importance of

A-16 HAPs from a regulatory perspective, a discussion of the differences between HC and individual VOCs is warranted. The term “total HC” is actually a misnomer since the measurement technique for quantifying “HC” – flame ionization detection (FID) – does not detect all HC compounds. Most importantly, FID is most sensitive to large HC compounds (present in unburned fuel) and less sensitive to compounds with carbon-oxygen bonds (which are usually produced by incomplete combustion). The FID also does not detect HCHO. EPA’s SPECIATE module (EPA 2008) can be used to derive the speciation of HC emissions from aircraft (based on data collected in the past decade). 3.1.2.5 Single (Reduced) Engine Taxi Considerations As stated in Section 3.1.1 (Aircraft Performance Characteristics), airlines promote the practice of single-engine taxiing to reduce fuel burn. However, while shutting down one or more engines during taxi out would on the surface appear to reduce emissions, the engines that are operating may need to operate at a higher thrust to maneuver an aircraft. This practice can also be counterproductive if a departing aircraft has to wait at the end of a runway while an engine(s) is warmed up (typically between three to five minutes) [Kumar et al, 2008]. One study [Kim et al, 2008] reports that where only one engine (out of two) was used to taxi, the increase in the power setting for the running engine ranged between 1.5 to three percent. Similar results were observed for the power settings of aircraft with more than two engines (i.e., the thrust level for the running engines was increased to compensate for one or more engines being off). 3.1.3 EDMS/AEDT Performance and Development This section of this Working Paper summarizes the Research Team’s knowledge and findings on the subjects of the architecture of the EDMS and AEDT computer models, the historical and future development programs for the models, and steps that have been taken to insure the accuracy of the models. 3.1.3.1 Model Architecture Figure 6 (Figure 1-2 of the EDMS User’s Manual) [FAA, 2013] illustrates the interaction of the various components and modules of the EDMS model that are used to process both emission inventories and to perform dispersion analysis. The manual also provides the following description of the model’s components and modules: • The back-end of the inventory and dispersion analysis functions is the databases that contain system data and user-created sources. The front-end of the model is the Graphical User Interface (GUI). • Users of EDMS enter data through the GUI. • Between the GUI and the databases the model contains the set of classes and functions that represent each emissions source and dispersion object along with the source/object associated properties. • The external interfaces to the EDMS include AERMAP (Version 12345), AERMET (Version 12345), AERMOD (Version 11103) and MOBILE (Version 6.2). • EDMS contains an Aircraft Performance Module (APM) and an Aircraft Emissions Module (AEM) that are common to components in AEDT. • EDMS’s view modules permit users to view output, receptor concentrations and system data that are stored in the databases. They also allow users to view a graphical representation of the various sources in an airport-specific input file. • EDMS also incorporates certain utilities for importing and exporting some types of data.

A-17 The EDMS provides three options for modeling aircraft-related taxi operations: 1) default taxi/delay times, 2) user-specified taxi times for each aircraft and 3) delay and sequence modeling. User-specified taxi times can be based on ICAO default values of 26 minutes or based on measurements from FAA’s ASPM 19 and Bureau of Transportation Statistics (BTS) Airline On-time Statistics databases.20 The EDMS Delay and Sequence Module simulates each aircraft’s ground movements using user’s input for an aircraft operations schedule, the assigned aircraft speed on taxiways, the overall capacity of the airport, and the airfield layout associated with runways, aprons, and taxiways. The module then estimates the time it takes each individual aircraft to taxi between apron and runway endpoints, based on airport-specified taxipaths. Notably, the use of sequence modeling is required when performing dispersion analysis using EDMS. For discussion purposes, Figure 7 depicts the aircraft taxi/delay datasets used by the EDMS. Figure 6. Figure 1-2 of the EDMS User’s Manual which illustrates the architecture of the model. 19 The Aviation System Performance Metrics (ASPM) online access system provides detailed data on flights to and from the ASPM airports (currently 77); and all flights by the ASPM carriers (currently 22), including flights by those carriers to international and domestic non-ASPM airports. All instrument flight rules (IFR) traffic and some visual flight rules (VFR) traffic are included. ASPM also includes airport weather, runway configuration, and arrival and departure rates. This combination of data provides a robust picture of air traffic activity for these airports and air carriers. 20 BTS reports taxi information for 16 U.S. air carriers that have at least one percent of total domestic scheduled-service passenger revenue, as well as two other carriers that report their schedule information voluntarily.

A-18 Figure 7. EDMS Taxi/Delay Datasets The architecture of the AEDT is depicted in Figure 8. The AEDT system is anticipated to incorporate the functionality of four noise and emissions modeling applications: 1) the Integrated Noise Model (INM), 2) the EDMS, 3) the Model for Assessing Global Exposure to the Noise of Transport Aircraft (MAGENTA), and 4) the System for assessing Aviation’s Global Emissions (SAGE). As shown, the third party components that are identified for EDMS--AERMAP, AERMET, and AERMOD—will also be components of AEDT (the FAA considered the release of EDMS Version 5.1 as a transition to AEDT) [Iovinelli et al, 2009]. Figure 8. Figure 2 of AEDT-AD-01 AEDT Architecture.

A-19 3.1.3.2 Development Programs/Timeframes The current version of the EDMS was released in August of this year (Version 5.1.4.1). The current version of the AEDT was released in June of 2012 (Version 2a (AEDT2a), Service Pack 1). AEDT2a replaces the Noise Integrated Routing System (NIRS) model, a model used to evaluate the potential environmental impacts of air traffic airspace and procedure actions. The FAA intends to release AEDT2b in 2014. AEDT2b will replace (i.e., sunset) both the FAA’s Integrated Noise Model (INM) and EDMS [Cointin, 2011]. 3.1.3.3 Model Accuracy/Sensitivity Tests There are typically three steps to producing models to predict real-world conditions: calibration, verification, and validation: • Calibration – Models are calibrated by adjusting available parameters to adjust how a model operates and simulates a process. • Verification – Tests are run to verify that a model is operating as it is expected to. • Validation - This step involves comparing the output from a model to historical data for a study area. As stated above, the current version of the EDMS was released in August of this year (Version 5.1.4.1). The FAA’s Office of Environment and Energy (FAA/AEE) and the Environmental Measurement and Modeling Division at the U.S. Department of Transportation’s John A. Volpe National Transportation Systems Center (i.e., the Volpe Center) conducted a validation study for a version of the EDMS that was released in October of 2002 (Version 4.1). The validation was performed to evaluate the addition of EPA’s AERMOD dispersion model. At the time, AERMOD had been validated for stationary sources but not for the varied sources found at an airport (particularly aircraft). Field measurements of CO were obtained at 25 sample positions over several days at a major international airport within the U.S. Airport operational activity for aircraft, ground support equipment (GSE), stationary sources, and motor vehicles on airport roadways were input to the EDMS. The comparisons between measured and modeled results were intended for use in an uncertainty assessment of the EDMS [Wayson et al, UNK]. In 2006, the verification of the EDMS’s ability to predict concentrations of CO was the subject of a thesis prepared by a student at the University of Central Florida [Martin, 2006]. This verification study was performed on Version 4.21 of the model. Two separate modeling exercises were performed (one using general airport information and the other using very detailed data) and the results of each were compared to measured data. The aircraft-related input values including taxi and queue times and taxiway assignments. For each hour of the study, the known arrival taxi and queue times were averaged and the average was applied to arriving aircraft with unknown taxi/queue times. The departure taxi and queue times were also averaged and applied to the departing aircraft with unknown taxi and queue times. The modeling exercises also included input for GSE, stationary sources, mobile lounges, and motor vehicles on airport roadways and in parking lots. The findings of the study were that measured concentrations of CO were overall higher than the model’s predictions. Notably, because this study did not attempt to isolate the various sources of CO for comparison to the measured levels, a single source or process was not identified as being the primary cause of the under prediction. FAA/AEE, the Volpe Center and staff of the Massachusetts Institute of Technology Department of Aeronautics & Astronautics are collaborating to assess a suite of tools that includes AEDT. As part of the development of the tools, tests are being performed to evaluate the sensitivity of the output from the models to uncertainties in model input and assumptions. There are four elements to the assessment program: 1) parametric sensitivity and uncertainty analyses, 2) comparisons to gold standard data, 3) expert reviews and 4) capability demonstrations/sample problems. In 2009, at the Eighth USA/Europe Air Traffic Management Research and Development Seminar, the preliminary results of the parametric sensitivity and uncertainty analyses were presented. Three main modules within AEDT were assessed—

A-20 the APM, the AEM, and the Aircraft Acoustic Module (AAM). Although the study evaluated aircraft emissions of CO2, NOx, CO, SOx, H2O and UHC below 3,000 feet, it does not appear that emissions resulting from ground level taxi operations were included in the assessment [Noel et al, 2009]. 3.1.4 Regulatory Framework The regulations that are most relevant to the 02-45 Research Project are those that mandate the maximum level of pollutants and pollutant precursors that can be emitted from the engine(s) on an aircraft. For aircraft manufactured in the U.S., the U.S. Environmental Protection Agency (EPA) establishes these rates for commercial jet engines. 14 CFR Part 34, Fuel Venting and Exhaust Emission Requirements for Turbine Engine Powered Airplanes and 40 CFR Part 87, Control of Air Pollution from Aircraft and Aircraft Engines; Emission Standards and Test Procedures, specify exhaust emission rates for new and in-use aircraft that vary depending on the date that an engine is manufactured. These regulations also prescribe that the test procedures used to demonstrate whether an engine meets the standards be performed using equipment and procedures specified in ICAO’s Annex 16 [ICAO, 2008]. EPA’s initial regulations for gaseous exhaust emissions from aircraft were promulgated in 1973. In 1997, the agency aligned the emissions standards with those established by ICAO (generally referred to as the CAEP/2 standards). In 2005, the agency promulgated more stringent NOx emissions standards for newly-certified engines. These standards brought the U.S. allowable emission rates closer to the ICAO rates that were effective in 2004 (referred to as the CAEP/4 or Tier 4 standards). Since that time ICAO has adopted two additional standards focused on the reduction of NOx (referred to as the CAEP/6 and CAEP/8 standards). In 2012, the EPA modified 40 CFR Part 87 such that engine models certified on or after July 18, 2012 had to meet the CAEP/6 standards (also referred to as the Tier 6 standards) and engines certified on or after January 1, 2014 must meet the CAEP/8 standards (also referred to as the Tier 8 standards) [EPA, 2012, 2012b, 2013]. 3.1.4.1 National Environmental Policy Act Requirements Enacted in 1970, the NEPA, was created as a result of public concerns about the human impact on the environment. The Act ensures that environmental factors are weighted equally in all the factors used by a federal agency in their decision making process. There are three levels of analysis that an improvement subject to the NEPA may undergo: 1) a Categorical Exclusion, 2) an Environmental Assessment and Finding of No Significant Impact, or 3) preparation of an Environmental Impact Statement. Airport-related projects subject to the NEPA that would involve aircraft taxi emissions include new airports and additional or extended taxiways to access existing runways. 3.1.4.2 Clean Air Act One of the purposes of Title I, Air Pollution Prevention and Control, of the CAA, is the protection and enhancement of the quality of the nation’s air resources so as to promote the public health and welfare. To this end, the CAA requires EPA to establish National Ambient Air Quality Standards (NAAQS). Mitigation measures may also be required under the CAA if it cannot be demonstrated that emissions from a proposed airport activity (i.e., aircraft taxiing on a new or expanded taxiway) would not cause or contribute to a new violation of any of the NAAQS. For the purpose of identifying activities/actions which would result in an increase in emissions that would be clearly de minimis, the EPA established rates which are referred to as the “de minimis levels”. These rates vary depending on whether an area is designated as “non-attainment” or “maintenance” for any of the NAAQS and the air pollutant for which the area has the designation.

A-21 3.2 EDMS/AEDT Review The Research Team conducted a review of the modeling inputs, assumptions, algorithms, database coverage and outputs required by EDMS and AEDT. As previously stated, the current version of AEDT (2a) replaces the NIRS model, a model used to evaluate the potential environmental impacts of air traffic airspace and procedure actions. As also stated, when released by the FAA, Version 2b of the AEDT will sunset the EDMS model. Notably, while AEDT2a does provide fuel burn and emissions data, use of the model is only intended for actions for which the study area is larger than the immediate vicinity of an airport, those that incorporate more than one airport, and/or those that include actions above 3,000 feet above ground level [Marks, 2012]. EDMS was specifically designed by the FAA to estimate emissions of CO, VOC, total organic gases (TOG), nonmethane hydrocarbons, (NMHC), reactive organic compounds (ROG), NOx, and sulfur oxides (SOx). EDMS also provides estimates of particulate matter equal to or less than 10 micrometers (PM10) and particulate matter equal to or less than 2.5 micrometers (PM2.5). However, because aircraft-related emissions of PM are being addressed by studies being performed by others (i.e., E-31, the SAE International aircraft exhaust emissions measurement committee and the European Aviation Safety Agency), emissions of this pollutant are not being addressed in the ACRP 02-45 project. Through the use of AERMOD, EDMS can also be used to perform atmospheric dispersion modeling to determine ambient (i.e., “outdoor”) pollutant concentrations of CO, NHMC, VOC, TOG, NOx, and SOx (PM also). 3.2.1 Input For the purpose of estimating/dispersing aircraft-related taxi and queue delay emissions, users have three options for EDMS input: • Use the ICAO/EPA default taxi/delay times (7 minutes for taxi-in and 19 minutes for taxi out), • Specify taxi/queue times for each aircraft of interest, or • Allow EDMS to calculate the times by invoking the model’s “delay and sequence” option. Estimating aircraft taxi-related emissions by using the EDMS sequencer requires users to specify additional airport layout information, including at least one gate, taxiway, taxipath, and runway. The following describes the required delay and sequence modeling input for an airport’s gates, taxiways, and taxipaths and discusses how a user’s input could affect the resultant level of taxi-related emissions: • Gates – The location of a gate(s) as specified by a user can affect taxi-related emissions because the model calculates the distance an aircraft traverses from a gate to a runway. • Taxiway – The coordinates of a taxiway identify a series of areas sources through which an aircraft travels to/from a gate and to/from a runway. Unless airport-specific data are entered, EDMS uses a default taxi speed for aircraft on taxiways of 15 knots (17.26 miles per hour). • Taxipath – Taxipaths are defined separately for departing and arriving aircraft and the sequence model determining the time-location coordinates of an aircraft as the aircraft moves along the assigned path. Other EDMS study elements that are used by the sequencer to more accurately estimate taxi-related emissions are:

A-22 • User specified operational profiles – These profiles indicate the relative activity at an airport by the quarter-hour, day-of-the-week, and by the month. • Airport schedule data – Users can “attach” a schedule file to the EDMS that contains scheduled pushback and landing times for every aircraft. If not provided, the sequence modeler will derive a schedule based on the annual operations and input (or default) operational profiles. • Configurations – EDMS uses configurations to assign aircraft to a runway based on weather conditions (wind direction, wind speed, hour of the day, ceiling, visibility, and temperature). Other user input that can affect aircraft taxi-related emissions include: • Reference values for temperature, pressure, and relative humidity can be changed by a user for each airport and scenario combination. • The airport’s elevation which redefines the altitude from which the reference thermodynamic conditions are lapsed can be changed. • By creating custom aircraft, users can set their own values for the reference idle emissions indices 𝐸𝐸𝐶𝐶𝐶𝐶, 𝐸𝐸𝐻𝐻𝐶𝐶 , and 𝐸𝐸𝑁𝑁𝐶𝐶𝑥𝑥, as well as reference fuel flow rate 𝑓𝑓𝑟𝑟𝐼𝐼𝐼𝐼 0 . All of the overrides and improved fidelities mentioned above are available for both emission inventories and dispersion analysis. However, dispersion analysis requires a minimum level of fidelity. The time-in- mode basis must be performance-based (though this does not require additional inputs affecting taxi). Taxi time must be sequence-based, requiring airfield configuration and layout definitions. Meteorological data must be hourly, requiring the full set of preprocessed reference data. Using configuration input, the sequence module of EDMS assigns an appropriate runway configuration for each hour of a year based on the meteorological data for each hour. The module then calculates an airport’s capacity for each hour. This capacity data and the calculated airport demand information (which is based on default or user input operational profiles or a schedule) are “fed” to a delay/queuing model— WWLMINET—which calculates an airport’s throughput. Modeling of taxi delay/queuing is performed only for departing aircraft. The departing aircraft are also assumed to form queues only along the taxiways that are assigned to a runway. 3.2.2 Algorithms An algorithm is a detailed sequence of actions that accomplish a task. EDMS’s emissions processing algorithms for the taxi mode are detailed in the model’s Technical Manual [FAA, 2009]. In general, EDMS develops a schedule of aircraft operations then “simulates” each aircraft flight using the calculations in the APM assuming the weather conditions specified by the model user to occur at the same time as the flight. Resulting emissions are then computed by the model using the AEM. The smallest unit of emissions calculations in EDMS is the trajectory segment. A trajectory segment represents some portion of an aircraft’s trajectory, characterized by temporal and spatial ranges, and select associated aspects of environmental and operational conditions. Emissions calculations for a trajectory segment utilize some of these characteristics, along with aircraft properties, to estimate the amounts of pollutants created over the course of the segment. The characteristics of a trajectory segment considered in calculating its emissions are duration, fuel flow rate, temperature, pressure, relative humidity, Mach number, operating mode, and altitude relationship to mixing height (that is, whether or not the aircraft is below the mixing height). For the purpose of emissions calculations, EDMS represents the entire taxi portion of an aircraft operation as a single trajectory segment. That is, taxi is characterized using fixed values for fuel flow rate, weather,

A-23 Mach number, operating mode, and altitude relationship to mixing height, through the duration of the taxi mode These characteristics, combined with emissions parameters specific to the associated aircraft, are used to determine taxi emissions. Some of these characteristics are always the same, whereas others are determined in manners that depend on the selected levels of fidelity. For example, Mach number is always zero, operating mode is always “idle” (though this only relevant to PM calculations), and the aircraft is always assumed to be below the mixing height (also only relevant to PM). Certain emissions parameters are also consistently employed (CO2 factor of 3,155, water factor of 1,237), and others are always used for jets (conversion factors of 1.0 for NMHC, 1.156234049 for TOG, and 0.9947855 for VOC, as well as the entire spectrum of mass ratios for speciated hydrocarbons). The value used for relative humidity is taken directly from the reference weather conditions associated with the air operation. The values used for pressure and temperature are also taken from reference conditions when using ICAO/USEPA performance, but are lapsed to the runway end elevation when using the European Civil Aviation Conference (ECAC) Doc29 performance (this is the only effect flight performance fidelity has on taxi emissions). 3.2.3 Databases EDMS has a database of emission indices (e.g., kilograms of pollutant per kilogram of fuel) and fuel flow rates (e.g., kilograms of fuel per second) for a variety of aircraft/engine combinations and representative operating conditions. These emission indices and fuel flow rates coupled with the amount of time an aircraft spends within the operating mode and the number of engines per aircraft, provides the basis for the emissions estimates. CO, HC, NOx - The calculation of aircraft taxi CO, HC, and NOx emissions requires species-specific emissions indices 𝐸𝐸𝐶𝐶𝐶𝐶, 𝐸𝐸𝐻𝐻𝐶𝐶 , and 𝐸𝐸𝑁𝑁𝐶𝐶𝑥𝑥 for idle engine operation at reference conditions. The calculation of taxi fuel flow rate requires a reference fuel flow rate 𝑓𝑓𝑟𝑟𝐼𝐼𝐼𝐼 0 gathered under the same conditions. These quantities come primarily from the ICAO Aircraft Engine Emissions Databank [ICAO, 2013], though some come directly from engine manufacturers or from the EPA’s AP-42 Volume II Section 1. They are based on measurements taken from engines running in their test bed, at conditions outlined in a certification standard put forth by ICAO [ICAO, 2008]. EDMS stores the emissions indices and fuel flow rates in the database table ENG_EMIS.DBF for standard system aircraft, and in USER_AIR.DBF for user-defined aircraft (AEDT currently stores them in its FLEET database, table FLT_ENGINES). The associated field names, which are consistent between EDMS and AEDT, are CO_REI_ID, HC_REI_ID, and NOX_REI_ID. The maximum values, and minimum non-zero values, specified for each of these parameters in the EDMS and AEDT system databases, are tabulated below. For both models, the lowest value for each of these modeling parameters for commercial jet engines is zero. Such values arise when no measurement is available, or when the measured value was small enough to round to zero. In the case where a value is zero, there is no indication (i.e., a note) of which was the case. Notably, however, the emissions modeling process in both EDMS and AEDT overrides zero values with a value of 0.0001. Table 7 Maximum and Minimum Modeling Parameters Modeling parameter AEDT min AEDT max EDMS min EDMS max 𝐸𝐸𝐶𝐶𝐶𝐶 5.74 1294 5.74 897 𝐸𝐸𝐻𝐻𝐶𝐶 0.002 302.36309 0.08 280.73 𝐸𝐸𝑁𝑁𝐶𝐶𝑥𝑥 0.39 8.53181 0.45 7.35 𝑓𝑓𝑟𝑟𝐼𝐼𝐼𝐼 0 0.00098 0.421 0.00102 0.38

A-24 SOx - The calculation of sulfur oxides emissions requires fuel sulfur content (FSC) and sulfur conversion efficiency (𝜺𝜺). EDMS and AEDT use hard-coded conservative values of 0.068 percent for FSC and 5 percent for 𝜺𝜺 when modeling sulfur according to FOA3a. For EDMS, this is whenever the airport associated with the operation is in the U.S., whereas for AEDT it is determined by a setting in the application configuration file. When not modeling sulfur according to FOA3a, EDMS reads FSC from AC_MAIN.DBF, and 𝜺𝜺 from SCENARIO.DBF, both of which are populated by the user in defining their study. CO2 and Water Emissions Modeling Data - The calculation of CO2 and water emissions requires species-specific emissions indices 𝑬𝑬𝑪𝑪𝑶𝑶𝟐𝟐 and 𝑬𝑬𝑯𝑯𝟐𝟐𝑶𝑶. Reviews of available fuel composition data and a first-principles analysis on an assumption of complete HC burn lead to the development of such values based on average composition [Hadaller et al, 1989 and 1993]. For EDMS, the values of these indices are hard-coded into the software as 𝑬𝑬𝑪𝑪𝑶𝑶𝟐𝟐 = 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 and 𝑬𝑬𝑯𝑯𝟐𝟐𝑶𝑶 = 𝟑𝟑𝟐𝟐𝟑𝟑𝟏𝟏. For AEDT, the values are specified in the application configuration file, and can be modified by the user. TOG, VOCs, and NMHC Emissions Modeling Data - The calculation of TOG, VOC, and NMHC emissions requires species-specific emissions indices that depend on the emissions index for HC and species-specific conversion factors 𝑪𝑪𝑻𝑻𝑶𝑶𝑻𝑻, 𝑪𝑪𝑽𝑽𝑶𝑶𝑪𝑪, and 𝑪𝑪𝑵𝑵𝑵𝑵𝑯𝑯𝑪𝑪. The determination of these factors is based on an estimated value of 0.0052145 for ethane content in TOG, and an assumption that no methane is produced. For EDMS, the values of these indices for jets are hard-coded into the software at 𝑪𝑪𝑻𝑻𝑶𝑶𝑻𝑻 = 𝟑𝟑.𝟑𝟑𝟑𝟑𝟏𝟏𝟑𝟑, 𝑪𝑪𝑽𝑽𝑶𝑶𝑪𝑪 = 𝟎𝟎.𝟎𝟎𝟎𝟎𝟑𝟑𝟐𝟐𝟑𝟑𝟎𝟎𝟑𝟑, and 𝑪𝑪𝑵𝑵𝑵𝑵𝑯𝑯𝑪𝑪 = 𝟑𝟑. For AEDT, the values are specified in the application configuration file. Speciated Organic Gases Emissions Modeling Data- The calculation of speciated organic gases’ emissions requires species-specific emissions indices that depend on the emissions index for HC and species-specific mass fractions. In EDMS, these mass fractions are specified in the database table MASSFRAC.DBF, with values ranging from 0.00002 to 0.15461. For AEDT, the values must be specified by the user in the application configuration file. 3.2.4 Output The EDMS generates an emissions inventory for CO2, CO, total hydrocarbons (THC), NMHC, VOC, TOG, NOx, SOx, and 394 speciated organic gases. Of note, emissions of CO2 are calculated only for aircraft and THC is calculated only for aircraft and APUs. Total fuel consumption is also calculated only for aircraft and is provided separately for taxi out and taxi in. The results of an inventory can be viewed in either summarized or detailed reports. For aircraft, the emission inventory results can be viewed for all aircraft in a “run” (e.g., total CO emissions for all aircraft operating in all modes within an input file) or by aircraft type/engine combination and mode (e.g., total CO for all MD-88 aircraft operating in each of the following modes: startup, taxi-out, takeoff, climb-out, approach, and taxi-in). As previously stated, using AERMOD, EDMS also generates pollutant concentration estimates for CO, NHMC, VOC, TOG, NOx, and SOx (PM2.5 and PM10 also). 3.2.5 AEDT The FAA’s AEDT is currently a work in progress and is being developed to replace all of the FAA’s existing regulatory and policy environmental models including SAGE, MAGENTA, IRS, INM, and the EDMS - essentially in that order. The first public version of AEDT, version 2a, was released in March of 2012. This version was created to replace NIRS for regulatory purposes and as such includes functionality focused on the needs of users performing regional noise analyses in the context of airspace re-design efforts. While it does include the

A-25 ability to calculate fuel burn and emissions, it has no provisions for the modeling of aircraft taxi operations. AEDT2a provides totals for aircraft fuel burn and for aircraft emissions of CO, THC, TOG, VOC, NMHC, NOx, CO2, H2O, and SOx (the nonvolatile component of PM (PMNV), PM from sulfur (PMSO), PM from unburned fuel organics (PMFO),, and PM2.5 also). Emission totals can be aggregated in four ways: 1) for each mode, 2) for each flight, 3) for each mode on each flight, or 4) for each performance segment of each flight. Modes include taxi out, takeoff ground roll, airborne departure for three altitude ranges (below 1,000 feet, below mixing height, and below 10,000 feet), airborne above 10,000 feet, and airborne approach for the same three altitude ranges. The emissions result capabilities of AEDT2a are closely aligned with those of EDMS, but there are some important differences. AEDT2a does not calculate any emissions or fuel burn contributions from non- aircraft sources (e.g., GSE, APUs, stationary sources, or vehicular sources), nor does it include aircraft startup or taxi. AEDT2a also does not support emissions dispersion modeling, or provide emissions results for speciated organic gases (or PM10). Conversely, the H2O emissions provided by AEDT2a are not provided in EDMS results (emissions of PMNV, PMSO, and PMFO are also not provided). The next public version of AEDT, version 2b, is currently under development and is scheduled for completion in 2014. When released, AEDT2b will replace the INM and EDMS for regulatory purposes - and as such will meet or exceed the capabilities of those two tools. It will therefore include the ability to calculate fuel burn, emissions, and potentially aircraft taxi-related noise - depending on available resources, funding, development priorities, and the output and schedule of the ACRP 02-27 Research Project for which the focus is taxi noise modeling. The implementation of taxi modeling functionality has just begun within the last few months for AEDT2b and so far the model only has simple time-in-mode capability (as in EDMS) for taxi emission calculations which are intended to serve the needs of global emissions inventories users of AEDT at the Volpe center. While initial planning has taken place regarding potential improvements to aircraft taxi modeling, firm decisions have yet to be made with respect to the scope of such improvements in the context of AEDT2b. Some of the primary reasons for this are schedule and resource constraints that are affecting AEDT developments. Due to these constraints, the focus for the next 12 months or so of AEDT2b development will be on supporting the functionality that currently exists in the legacy EDMS tool for taxi modeling. As a result, and for the purpose of the ACRP 02-45 Research Project, it is currently assumed that the AEDT capability will match the EDMS capability as described above in all areas, including allowable inputs, assumptions, database coverage, and outputs. 4. Issues in Need of Resolution Based on the findings of Tasks 1 and 2 discussed above, several additional issues, or research “gaps”, in need of resolution are now evident and will be addressed during the subsequent tasks of this research project. These issues are described as follows along with some preliminary recommendations on how they will be addressed: • ACRP 02-03a Findings - First, ACRP 02-03a focused mainly on the CFM56 family of engines and though these results for a few other engines fit the same data, a large number of engine types were not characterized. Secondly, ACRP 02-03a focused on gas-phase HAP emissions, largely comprising individual VOCs like HCHO and benzene. However, aircraft engines are certified for NOx, CO, and HC and EDMS/AEDT outputs emissions for these three gas-phase pollutants - but not for individual HAPs. Finally, the relationship between NOx, HC, and CO emissions on fuel flow rate and ambient temperature was not analyzed in great detail by ACRP 02-03a.

A-26 It is recommended that the NOx, CO, and HC data from ACRP Project 02-03a could be further analyzed to help address these shortcomings listed above. • Taxi/Idle Sub-Phases - Given the scope of the dependence of true HC emission rates on actual fuel flow rates and ambient temperatures, it is apparent that the current method used by EDMS to compile HC and CO emissions (i.e., fuel flow rate and emission indices based on the ICAO 7percent value) can lead to inaccurate emission inventories. Recommendations to address this shortcoming of EDMS/AEDT are included as elements of later tasks of this project and two possible suggestions are briefly described below: − Multiple Sub-Phases - Based on the evidence that the idle/taxi phase cannot be well described by a single fuel flow rate and concomitant emission indices (especially for CO and HC), the idle/taxi phase could be sub-divided into multiple sub-phases. This approach was used by Stettler et al. [2011] based on data presented in Patterson et al [2009]. In that study, the idle/taxi phase was divided into the following components (w/ mean time-in-mode and percent thrust in parentheses): landing roll (46 seconds, 4 – 7 percent), reverse thrust (15 seconds, 30 percent), taxi in (371 seconds, 4 – 7 percent), taxiway in acceleration (10 seconds, 7 – 17 percent), taxi out (780 seconds, 4 – 7 percent), taxiway out acceleration (10 – 20 seconds, 7 – 17 percent), and hold (341 seconds, 4 – 7 percent). The true values for each sub-phase vary among airports and are affected by numerous parameters including aircraft type, time-of-day, weather conditions, etc. − Corrected Emission Indicies - For each sub-phase, a “corrected” emission index appropriate for the actual fuel flow rate and ambient temperature could be used. For example, ACRP 02-03a presented a model by which HAP emission indices for 7 percent thrust can be multiplied by a correction factor that is determined by the actual fuel flow rate, the 7 percent fuel flow rate, and the ambient temperature, as follows: EI HAP “corrected” = EI HC(7 percent, 15 degrees Celcius) × (Fuel flow rate correction factor) × (ambient temperature correction factor) Although this approach is geared toward individual gas-phase HAP compounds and cannot be blindly applied to CO or HC emissions (both of which show the same overall trends as individual HAP compounds), it is likely that an analogous model applicable to CO and HC could be developed. 5. Go Forward Plan At the end of December, 2013 or first of January 2014, a web-based meeting will be held with both the Project Panel and Research Team. During this meeting, the Panel/Team will discuss the findings of the Task 1 literature and Task 2 EDMS/AEDT reviews presented in this Working Paper and the next two tasks that will be performed in January and February of 2014. The next two tasks involve the following: • Task 4 – Analyze Engine Performance Data - KBE Team members have FDR data from which summary results can be used for the purpose of this ACRP Reseach Project. Team members also have access to the raw data that has been collected as part of several field campaigns, including APEX, JETS-APEX2, APE3, ACRP 02-03A, and the AAFEX projects. Notably, much of the data has been published in peer-reviewed literature (e.g., Yelvington et al. 2007) or

A-27 as an ACRP report [Herndon et al. 2012]. By leveraging a broader ASDE-X data set covering taxi operations, enhancements can be better applied within EDMS/AEDT through improved knowledge as to when to apply them. Better knowledge of the prevalence and nature of accelerations and decelerations, hold times, and other operational details are crucial to making good use of the knowledge of how those situations impact taxi fuel flow and emissions. • Task 5 – Evaluate Model Inaccuracies – This task represents one of the most important elements of the ACRP 02-45 Research Approach as it will help pave the way forward to improving how EDMS/AEDT quantifies aircraft taxi/idle emissions. Although additional inaccuracies with EDMS/AEDT taxi/idle emissions estimates are likely to be identified during the course of Task 4, the Task 1 and 2 Literature and EDMS reviews confirmed what the Research Team believed are major faults with EDMS/AEDT taxi/idle emissions estimates. These shortcomings (presented in the Work Plan) and a general description of their impact on taxi/idle emissions are summarized below: Table 8 Summary of EDMS/AEDT Shortcomings Related to Taxi/Idle Emissions Shortcoming Impact EDMS/AEDT assumptions regarding the duration of taxi/idle modes are not representative of actual conditions. Differences in actual emissions and EDMS/AEDT emissions are (at least) directly proportional to differences in the duration of taxi/idle modes. EDMS/AEDT assumes fixed fuel flow rates during taxi/idle based on 7 percent thrust. Actual fuel flow rates vary considerably and emission indices are a function of fuel flow rate. EDMS/AEDT uses one emission index value for each pollutant Actual emission indices are complex functions of fuel flow rate, ambient temperature, and other factors. EDMS/AEDT does not account for variations in operational practice including tug assisted single-engine (i.e., reduced engine) or electric taxiing. Discrepancies between actual and EDMS/AEDT assumed operating patterns directly impact the accuracy of model estimates. The purpose of Task 5 is to put the relative importance and implications of these inaccuracies in EDMS/AEDT related to emission estimates into perspective both in terms of airport emission inventories as well as potential impacts on local and regional air quality planning. Work under Task 5 will begin with development of a spreadsheet model that can be used to quantify the impacts of taxi/idle related inaccuracies in EDMS/AEDT identified during the course of Tasks 1, 2 and 4. The primary focus will be on THC, CO, and NOx, and, to the extent possible, HAPs where suitable data are available. Given that fuel usage rates will also be directly calculated, impacts on CO2 emissions will also be reported using standard carbon content factors applicable to jet fuel. The spreadsheet model will then be used in combination with appropriate data and assumptions including the actual operating data obtained under Task 4 to analyze the potential impact that the flaws in EDMS/AEDT have on estimated taxi/idle emissions at the individual aircraft level.

A-28 76 Annotated Bibliography — Literature Search ACRP 02-45: Methodology to Improve EDMS/AEDT Quantification of Aircraft Taxi/Idle Emissions Author(s) or Regulatory Agency Year Title Citation Annotation Aircraft Owners and Pilots Association 2011 Taxi Green in 2016 Aircraft Owners and Pilots Association. n.p. Web. October 2, 2013. Describes a system being developed by Honeywell and Safran that would allow aircraft to taxi to and from gates using electric motors powered by auxiliary power units. Alternative Emissions Methodology Task Group, CAEP7 2006 Results from a Number of Surveys of Power Settings Used During Taxi Operations CAEP7-WG3-AEMTG-WP7-08 The results of a number of surveys that were carried out to investigate the actual power settings used during taxi operations. Anderson B.E., Chen, G., Blake, D. 2006 Hydrocarbon Emissions from a Modern Commercial Airliner Atmospheric Environment 2006: 40(19): 3601 - 3612 Emissions from the RB211-535-E4 engine during the EXCAVATE project. Numerous speciated hydrocarbons measured. Idle thrust is listed as “4 – 7 percent”. Baik, H., Sherali, H., Trani, A. UNK Time-Dependent Network Assignments Strategy for Taxiway Routing at Airports Transportation Research Board, Paper No. 02-3660 A time-dependent network assignment strategy is proposed for efficiently handling aircraft taxiway operations at airports. Balakrishna, P., Ganesan, R., Sherry, L. 2010 Accuracy of Reinforcement Learning Algorithms for Predicting Aircraft Taxi-out Times: A Case-Study of Tampa Bay Departures Transportation Research Part C 18 (2010) 950-962 Presents the results of a taxi-out time case study performed at Tampa International Airport. Investigates the accuracy of taxi out time prediction using a nonparametric reinforcement learning (RL based method. Beyersdorf, A., Thornhill K., Winstead, E., Ziemba, L., Blake, D., Timko, M., Anderson, B. 2012 Power-Dependent Speciation of Volatile Organic Compounds in Aircraft Exhaust Atmospheric Environment 61 (2012): 275-82. APEX3 measurements of speciated VOCs. Found that the universal VOC scaling didn’t apply at high powers - the X / C2H4 ratios varied. Quantified speciated VOC EIs at 4 percent and 7 percent. Bhadra, D., Knorr, D., Levy, B. 2011 Benefits of Virtual Queuing at Congested Airports using ASDE-X: A Case Study of JFK Airport Presented at Ninth USA/Europe Air Traffic Management Research and Development Seminar, 2011 Discusses the potential for delay management by using departure data recorded by the ASDE-X system at JFK Airport before runway reconstruction in 2010 began. Lays out concepts, data, metrics, and a framework to estimate benefits from virtual queuing, a departure management system that allows an aircraft to maintain a rolling spot in the queue without physically joining a queue. Buttress, Jenna and Kevin Morris 2005 An Estimation of the Total NOx Emissions Resulting from Aircraft Engine Ground Running at Heathrow Airport British Airways Technical Paper ENV/KMM/1127/14.18 Results of a study performed to evaluate the levels of NOx emissions during three categories of ground operations--check starts, runs at no more than ground idle, and runs at powers greater than ground idle

A-29 Author(s) or Regulatory Agency Year Title Citation Annotation Clare, B., Richards, A. UNK Optimization of Taxiway Routing and Runway Scheduling University of Bristol Describes an optimization method for the combined issues of airport taxiway routing and runway scheduling. Clewlow, R., Simaiakis, I., Balakrishnan, H. UNK Impact of Arrivals on Departure Taxi Operations at Airports American Institute of Aeronautics and Astronautics Through an analysis of departures at JFK International and Boston Logon International, the effect of arrivals in delaying departure operations, is evaluated. Cointin, R. 2011 Aviation Environmental Design Tool: Interdependencies of Aircraft Noise, Emissions and Fuel Burn Accessed at http://www.fican.org/pdf/Roadm ap2011/2011_1020_Cointin_AE DT_Briefing_Noise_Workshop. pdf. November 26, 2013 Topics included were 1) why AEDT, 2) What is AEDT, 3) AEDT Timelines, and 4) Uncertainty Quantification Collins, B. 1982 Estimation of Aircraft Fuel Consumption Journal of Aircraft 19, 969-975 Describes an algorithm for estimating the fuel consumption of commercial aircraft from path profile data. Deonandan, I., Balakrishnan, H. UNK Evaluation of Strategies for Reducing Taxi-out Emissions at Airports American Institute of Aeronautics and Astronautics Evaluates the effects of single engine taxiing and operational tow-outs Diana, Tony. 2013 An Application of Survival and Frailty Analysis to the Study of Taxi-out Time: A Case of New York Kennedy Airport Air Transport Management 26 (2013) 40-43 Evaluates how selected operational factors affect the duration of aircraft taxi-out times at John F. Kennedy Airport. DuBois, D., Paynter, G. 2006 Fuel Flow Method 2 for Estimating Aircraft Emissions Society of Automotive Engineers Paper: 01-1987 Presents derivation, updates, and clarifications of the fuel flow method methodology known as "Fuel Flow Method 2". European Civil Aviation Conference 2005 Report on Standard Method of Computing Noise Contours around Civil Airports, Volume 1: Applications Guide, 3rd Edition ECAC.CEAC Doc 29 Guidance for the best practice methodology for aircraft noise contour modelling. European Civil Aviation Conference 2005 Report on Standard Method of Computing Noise Contours around Civil Airports, Volume 2: Technical Guide, 3rd Edition ECAC.CEAC Doc 29 The technical companion to the Applications Guide. Federal Aviation Administration 2013 Emissions and Dispersion Modeling System (EDMS) User's Manual FAA-AEE-07-01 Detailed information on the functionality of the EDMS model. Federal Aviation Administration Unknown Runway Safety A Best Practices Guide to Operations and Communications Federal Aviation Administration, n.p. Web, October 2, 2013. Describes the common tasks that pilots should incorporate in to their taxi procedures.

A-30 Author(s) or Regulatory Agency Year Title Citation Annotation Federal Aviation Administration 2013b Order JO7110.65U Air Traffic Organization Policy, Order JO7110.65U, Air Traffic Control, August 22, 2013 Prescribes air traffic control taxi and ground movement control procedures. Federal Aviation Administration 2009 Emissions and Dispersion Modeling System (EDMS) Version 5 Technical Manual FAA-AEE-07-07 Technical support documentation for Version 5 of the EDMS. Federal Aviation Administration 2013 Aeronautical Information Manual FAA, 2013 The official guide to basic flight information and Air Traffic Control Procedures Federal Aviation Administration 2007 Environmental Desk Reference for Airport Actions Federal Aviation Administration. Office of Airports. Environmental Desk Reference for Airport Actions. October, 2007. http://www.faa.gov/airports/envi ronmental/environmental_desk_r ef/. Regulatory guidance document. The Desk Reference summarizes applicable special purpose laws in one location for convenience and quick reference. Its function is to help FAA integrate the compliance of NEPA and special purpose laws. Federal Aviation Administration 1997 Air Quality Procedures for Civilian Airports & Air Force Bases Federal Aviation Administration. Office of Environment & Energy. Air Quality Procedures for Civilian Airports & Air Force Bases, prepared by EEA Inc. and CSSI, Inc., April 1997. Regulatory guidance document. Procedures for the preparation of air quality assessments for proposed Federal actions are required for compliance with the National Environmental Policy Act, the Clean Air Act and other environment-related regulations and directives. Federal Aviation Administration 2004 Air Quality Procedures for Civilian Airports & Air Force Bases - Addendum Federal Aviation Administration. Office of Environment & Energy. Air Quality Procedures for Civilian Airports & Air Force Bases, Addendum, 2004. Regulatory guidance document. Addendum to the 1997 version. Procedures for the preparation of air quality assessments for proposed Federal actions are required for compliance with the National Environmental Policy Act, the Clean Air Act and other environment-related regulations and directives. Federal Aviation Administration 2012 Vision 100 - Century of Aviation Reauthorization Act Public Law 108–176 (enacted December 12, 2003) as Amended through Public Law 112–95 (enacted February 14, 2012) One part of the Vision 100—Century of Aviation Reauthorization Act was to establish a legal framework to reduce emissions from airport vehicles, ground support equipment (GSE) and infrastructure at commercial service airports in air quality nonattainment and maintenance areas. The Act defines a program and procedures for determining emission reduction credits for voluntary early reduction measures that could be counted towards transportation conformity determinations or new source review requirements at airports.

A-31 Author(s) or Regulatory Agency Year Title Citation Annotation Federal Aviation Administration 2006 Environmental Impacts: Policies and Procedures. Federal Aviation Administration. Order 1050.1E. Effective March 20, 2006. Agency Order. This order updates the FAA agency-wide policies and procedures for compliance with the National Environmental Policy Act (NEPA) and implementing regulations issued by the Council on Environmental Quality (40 CFR parts 1500-1508). Order 1050.1E cancels Order 1050.1D. Federal Aviation Administration 2006 National Environmental Policy Act (NEPA) Implementing Instructions for Airport Actions Federal Aviation Administration. Order 5050.4B. Effective April 28, 2006. Agency Order. Order 5050.4B supplements FAA Order 1050.1E, “Environmental Impacts: Policies and Procedures.” Order 5050.4B substantially updates and revises Order 5050.4A, “Airports Environmental Handbook.” ARP’s issuance of Order 5050.4B cancels Order 5050.4A. Federal Aviation Administration 2007 Voluntary Airport Low Emission (VALE) Program Technical Report Federal Aviation Administration. Office of Airports. Voluntary Airport Low Emission (VALE) Program Technical Report, Version 7. December 2010. Regulatory guidance document. FAA technical guidance to support VALE applications. Version 7 is the latest of the series of reports. Federal Aviation Administration 2002 General Conformity Guidance for Airports, Questions and Answers Federal Aviation Administration and US Environmental Protection Agency. General Conformity Guidance for Airports, Questions and Answers. September 25, 2002. Regulatory guidance document. Joint FAA and USEPA technical guidance based on a stakeholders group to address airport air quality improvements - focusing on NOx reductions. Fleuti, E. et al 2009 Air Quality Assessment Sensitivities - Zurich Airport Case Study Fleuti et al. 2009 The case study is based on Zurich airport's 2008 activity data and examined the emissions variability of the LTO cycle based on the sophistication of the data inputs. The inventory analysis begins with ICAO default methods and successively adds increasing details including ambient temperature impacts, altitude impacts and airline specific thrust settings. The model used for the study is LASPORT version 2.0 (aka LASAT for airports) developed by Janicke Consulting (one of the co-authors); it is not clear if the model's methods for ambient, altitude and trust impacts on emissions are published in the public domain. If the methods are not public, the resulting impacts on the inventories by level of detail is still a useful point of comparison. Pollutants covered are NOx, HC, CO, PM, and CO2.

A-32 Author(s) or Regulatory Agency Year Title Citation Annotation Franc24 2013 Electric Taxiing Unveiled at Paris Air Show France24. n.p. Web. June 18, 2013. An article on the electric taxi system developed by Honeywell and Safran. States that the system was demonstrated on an A320, the companies planned on marketing the system in 2016 with the intent of equipping approximately 2,600 aircraft with their Electric Green Taxiing System (EGTS). Goldberg, B., Chesser, D. 2008 Sitting on the Runway: Current Aircraft Taxi Times Now Exceed Pre-9/11 Experience Bureau of Transportation Statistics, SR-008, May 2008. Figures illustrate average taxi-in/taxi-out times for the period 1995-2007. Also analyzes ground time by size of airport and flight volume. Grinspun,Y., Miller, E. 2002 A Survey-Based Approach to Measure Taxiway Delay and Predictability at Lester B. Pearson International Airport Presented at the 82nd Annual Meeting of the Transportation Research Board, Washington, D.C., 2003. A study of taxiway delay and taxi-in/out times at this airport in Toronto. Hadaller, A. M. Momenthy 1989 The Characteristics of Future Fuels Boeing, D6-54940 Hadaller, A. M. Momenthy 1993 Characteristics of Future Aviation Fuels Report to American Council of ran Energy-Efficient Economy, Washington D.C., 1993 Herndon, S. C., Wood, E., Northway, M., Miake-Lye, R., Thornhill, L., Beyersdorf, A., Anderson, B., Dowlin, R., Dodds, W., Kinghton, W. 2009 Aircraft Hydrocarbon Emissions at Oakland International Airport Environmental Science and Technology 43 (6):1730-1736 Characterization of VOC EIs measured from in-use aircraft during the JETS-APEX2 campaign (at OAK). Showed that even for a single engine type EIs are variable (reflecting changes in ambient temperature and fuel flow rate), but most emission ratios (e.g., propene / formaldehyde) are fairly constant. This supports a “universal” VOC scaling profile – i.e., regardless of total VOC emissions, the portion accounted for by formaldehyde, propene, etc. is constant. Analysis suggests OAK idle VOC emissions are underreported by 16 – 45 percent. Herndon, S., Jayne, J., Lobo, P., Onasch, T., Fleming, G., Hangem, D., Whitefield, P., Miake-Lye, R. 2008 Commercial Aircraft Engine Emissions Characterization of in-Use Aircraft at Hartsfield- Jackson Atlanta International Airport Environmental Science & Technology 42 (2008): 1877-83. Emission indices of in-use aircraft at Hartfield-Jackson Atlanta International Airport inferred from advected plumes. The CO emission index observed in ground idle plumes was greater (up to 100 percent) than predicted by engine certification data for the 7 percent thrust condition, consistent with actual idle operation being at thrusts lower than 7 percent. Plenty of CF34, JT8D, CFM56, PW2037, and CF6 engines observed.

A-33 Author(s) or Regulatory Agency Year Title Citation Annotation Herndon, S., Rogers, T., Dunlea, E., Jayne, J., Miake-Lye, R., Knighton, B. 2006 Hydrocarbon Emissions from In- Use Commercial Aircraft during Airport Operations Environmental Science & Technology 40 (14):4406 - 4413. Advected plumes at Boston Logan International Airport, May 2003. Found that the sum of individual VOCs and projected total VOCs were higher than ICAO 7 percent UHC (consistent with aircraft idling at lower thrusts with higher UHC emissions) Herndon, S., Wood, E., Franklin, J., Miake- Lye, R., Knighton, W.B, Babb, M., Nakahara, A., Reynolds, T., Balakrishnan, H. 2012 Measurement of Gaseous HAP Emissions from as a Function of Engine and Ambient Conditions (ACRP Project 2-03a) ACRP Report 63. In depth analysis of dependence of VOC emissions on engine condition (i.e., fuel flow rate) and ambient temperature. Temperatures ranged from -8 °C to 25 °C, fuel flow rate ranged from ground idle (approximately 3 percent) to 15+ percent (above the “idle” regime). A simple model tool was developed that can be used to “correct” the ICAO UHC EI based on ambient temperature and fuel flow rate. i.e., the temperature and fuel flow rate determine the appropriate factor by which to multiply the ICAO values. This model is the frontrunner model for how to “fix” the EDMS/AEDT default 7 percent values. Engine covered: numerous CFM56- 7B24 and -3B1’s, a V2527 and a PW4090. International Civil Aviation Organization 2013 Engine Exhaust Emissions Databank, Issue 19. ICAO, April 15, 2013. Contains exhaust emissions for those aircraft engines that have entered production. The information is provided by engine manufacturers. International Civil Aviation Organization 2008b International Standards and Recommended Practices, Annex 16, Environmental Protection: Aircraft Engine Emissions ICAO, Annex 16, Vol 2, Montreal, 3rd ed., 2008 Adopted by the Council of ICAO it achieves "the highest practicable degree of uniformity in regulations, standards, procedures and organization in relation to aircraft, personnel, airways, and auxiliary services in all matters in which such uniformity will facilitate and improve air navigation." International Civil Aviation Organization 2008 ICAO Annex 16: Environmental Protection, Volume II - Aircraft Engine Emissions ICAO, Annex 16 2008 Provides all provisions that relate to the environmental aspects of aircraft engine emissions for aircraft engaged in international civil aviation. International Civil Aviation Organization 2011 Airport Air Quality Manual ICAO, Doc 9889, 2011 Discusses the reference emissions/trust settings of each aircraft operational phase, the operational flight cycle (i.e., engine start, taxi to runway, hold, etc.) and approaches to emissions calculations (inventories and dispersion) Iovinelli, Ralph and Mohan Gupta 2009 FAA's Airport Air Quality Model: Aviation Sector's Tool for Analysis of Criteria and Hazardous Pollutants In the U.S. Environmental Protection Agency 18th Annual International Emissions Inventory Conference Proceedings, April 14-17, 2009, Baltimore, Maryland.. Discusses the development of AEDT and the enhancements that were included in Version 5.1 of the EDMS.

A-34 Author(s) or Regulatory Agency Year Title Citation Annotation Jung, Y. 2010 Fuel Consumption and Emissions from Airport Taxi Emissions Presentation at NASA Green Aviation Summit, Washington DC A presentation regarding a method to calculate fuel consumption and emissions of phases of taxi operations. Khadilkar, K., Balakrishnan, H. 2012 Estimation of Aircraft Taxi Fuel Burn Using Flight Data Recorder Archives Transportation Research Part D (2012) 532-537 Creates a model for estimating fuel consumption of a taxiing aircraft using flight data recorder information from operational aircraft. Kim, B., Rachami, J. 2008 Aircraft Emissions Modeling Under Low Power Conditions Report to Observatory of Sustainability in Aviation Presents the results from different analysis levels involving a single aircraft and engine and aggregated fleet levels to illustrate the range of errors in the thrust setting at idle. Kumar, Vivek, Lance Sherry and Terry Thomspon 2008 Analysis of Emissions Inventory for "Single-Engine Taxi-out" Operations In the International Conferences on Research in Air Transportation 3rd Conference Proceedings, June 1-4, 2008, Fairfax, Virginia, edited by Vivek Kumar, 1-6. Examples the sensitivity of emission factors (number of engines, engine efficiency, and fleet mix, taxi-out time) through case study of departure operations at Orlando International Airport and LaGuardia Airport. Legge, J., Levy, B. 2008 Departure Taxi Time Predictions Using ASDE-X Surveillance Data Presented at 26th International Congress of the Aeronautical Sciences, 2008 Analyzes the utility of the Airport Surface Detection Equipment, Model X (ASDE-X) surface surveillance data using data from March 2008 at ATL. Levine, Brian, H. Oliver Gao 2006 Aircraft Taxi-Out Emissions at Congested Hub Airports and the Implications for Aviation Emissions Reductions in the United States Levine, Brian. Presented at the 86th Annual Meeting of the Transportation Research Board, Washington, D.C., 2007. Discusses and describes the aircraft taxi process at a congested hub airport (Newark Liberty Intl) Lobo, P. et al 2007 The Development of Exhaust Speciation Profiles for Commercial Jet Engines Lobo et al. 2007 This study reports the emissions of CO, CO2, NOx, and speciated HC at six thrust settings: 4 percent, 7 percent, 30 percent, 40 percent, 65 percent and 85 percent measured from both engines on four parked 737 aircraft at the Oakland International Airport. The engine types were selected to represent both old and new technologies. Sponsored by CARB, this collaboration between University of Missouri – Rolla, Aerodyne Research and University of California - Riverside forms the basis of the California agency's speciation for commercial aircraft exhaust. Germane to the ACRP study at hand are the emissions data collected by thrust setting. There was some data loses noted in the Executive Summary related to specific HC compounds.

A-35 Author(s) or Regulatory Agency Year Title Citation Annotation Marks, Julie 2012 FAA Order 1050.1E, Change 1 Guidance Memo #4: Guidance on Using AEDT2a to Conduct Environmental Modeling for FAA Air Traffic Airspace and Procedure Actions Prepared by Rebecca Cointin and Steve Urlass, March 21, 2012. Provides guidance on the use of AEDT2a to conduct aircraft noise, fuel burn, and emissions modeling for air traffic airspace and procedure actions under the National Environmental Policy Act. Martin, A. 2006 Verification of FAA's Emissions and Dispersion Modeling System (EDMS) Thesis (M.S.E) -- University of Central Florida, Dept. of Civil and Environmental Engineering, 181 p. Presents the results of a study conducted by the FAA, the Volpe Center, and CSSI to verify EDMS's ability to predict CO concentrations in the vicinity of an airport. Mazaheri, M., Johnson, G., Morawska, L. 2009 Particle and Gaseous Emissions from Commercial Aircraft at Each Stage of the Landing and Takeoff Cycle Environ. Sci. Technol. 43 (2009): 441-46 Australian measurements of in-use aircraft. Found that “idle” thrust was lower than “taxi” thrust and less than the ICAO value, meaning that idle is not at 7 percent thrust. Miller, Bruno, Kenneth Minogue and John-Paul Clarke 2010 Constraints in Aviation Infrastructure and Surface Aircraft Emissions Accessed at http://www. areco. org/AQ% 20Aircraft% 20Surface%20Constrants%20Mi ller. pdf, January 18, 2010. Discusses the growth of aviation emissions from 1995 to 2000 and investigates potential methods to reduce emissions including single-engine taxiing. Nikoleris,T., Gupta, G, Kistler, M. 2011 Detailed Estimation of Fuel Consumption and Emissions During Aircraft Taxi Operations at Dallas/Fort Worth International Airport Transportation Research Part D 16 (2011) 302-308 Estimates fuel consumption and emissions during taxi operations using aircraft position data from actual operations at Dallas-Ft. Worth International Airport. Uses assumptions for the thrust level during each taxi state, fuel flow and emission index values from ICAO's databank. No information 2012 Air Quality Assessment Sensitivities - Zurich Airport Case Study Zurich Airport 2012 A second version of SR1002, updated to include a section (3.7) on "Effect of Emissions on Regional Concentrations." Noel, George, Doug Allaire, Stuart Jacobson, Karen Willcox and Rebecca Cointin 2009 Assessment of the Aviation Environmental Design Tool In the USA/Europe Air Traffic Management Research and Development 8th Seminar, June 29-July 2, 2009, Napa, California, edited by George Noel. Unpaginated document. Presents a summary of an assessment of the AEDT component of the FAA's Tool Suite. Page, J., Bassarab, K., Hobbs, C., Robinson D., Schultz, T., Sharp, B., Usdowski, S., Lucic, P. 2009 Enhanced Modeling of Aircraft Taxiway Noise, Volume 1 Scoping ACRP Web-Only Document 9 Presents the results of a study to determine the best way to model airport noise from aircraft taxi operations and to create a plan for implementation of noise prediction capability in to the INM in the near term and the AEDT in the long term.

A-36 Author(s) or Regulatory Agency Year Title Citation Annotation Page, J., Hobbs, C., Gliebe, P. 2013b Enhanced Modeling of Aircraft Taxiway Noise, Volume 2 ACRP Web-Only Document 9 Summarizes taxi-related data from ACRP Project 11-02 Task 8 (Enhanced Modeling of Aircraft Taxiway Noise - Scoping), a number of surveys of power settings used in normal taxi operations that are being considered by ICAO, and data from the ICAO Best Practices Certification Database (BPDB- IACO/CAEP8) which lists nominal percentage taxi thrusts. Partnership for AiR Transportation Noise and Emissions Reduction (PARTNER) 2009 Aircraft Impacts on Local and Regional Air Quality in the United States PARTNER Project 15 final report Documents the findings of a study to evaluate ways to promote fuel conservation measures and opportunities to reduce air traffic inefficiencies that increase fuel burn and emissions. Roof, C., Hansen, A., Fleming, G., Thrasher, T., Nguyen, A., Hall, C., Dinges, E., Bea, R., Grandi, F., Kim, B., Usdrowski, S., Hollingsworth, P. 2007 Aviation Environmental Design Tool (AEDT) System Architecture Federal Aviation Administration, Document AEDT-AD-01 Describes the "building blocks" that form AEDT's architecture (i.e., EDMS, INM, MAGENTA, SAGE), timelines, and development specifications. Santoni, G., Lee, B., Wood, E., Herndon, S., Miake-Lye, R., Wofsy, S., McManus, J., Nelson, D., Zahniser, M. 2011 Aircraft Emissions of Methane and Nitrous Oxide During the Alternative Aviation Fuel Experiment Environmental Science & Technology (2011): 110720130733026. Emission indices of the greenhouse gases methane and nitrous oxide during the AAFEX project. CH4 EIs are ~500 to 50 mg/kg fuel for the CFM56-2C1 engine at idle (and negative at high power). Schäfer, K., Jahn, C., Sturm, P., Lechner, B., Bacher, M. 2003 Aircraft Emission Measurements by Remote Sensing Methodologies at Airports Atmospheric Environment 37, no. 37 (2003): 5261-71. Remote sensing of EIs from in-use aircraft at a few European airports. Actual observed EIs are higher (CO) and lower (NOx) than ICAO 7 percent values, suggesting actual idle operation is usually at sub-7 percent thrusts. Senzig, D., Fleming, G., Iovinelli, R. 2009 Modeling of Terminal-Area Airplane Fuel Consumption Journal of Aircraft 46 (4) Presents a method of modeling fuel consumption that was developed using data from a major airplane manufacturer. Simaiakis, I., Balakrishnan, H. 2010 Impact of Congestion on Taxi Times, Fuel Burn, and Emissions at Major Airports Transportation Research Record: Journal of the Transportation Research Board, No. 2184, 2010, pp 22-30 Assesses the impact of surface congestion on taxi times, fuel burn, and emissions through analysis of departing traffic from four major U.S. airports

A-37 Author(s) or Regulatory Agency Year Title Citation Annotation Simaiakis, I., Khadilkar, H., Balakrishnan, H., Reynolds, T.G., Hansman, R.J., Reilly, B., Urlass, S. 2001 Demonstration of Reduced Airport Congestion Through Pushback Rate Control American Institute of Aeronautics and Astronautics Presents the results of field tests to evaluate control strategies to airport congestion at Boston Logan International. The approach determines a suggested rate to meter pushbacks from gates in order to prevent surface congestions and reduce the time that aircraft spend with their engines on while taxiing to the runway. Spicer, C.W., Holdren, M., Riggin, R., Lyon, T. 1994 Chemical-Composition and Photochemical Reactivity of Exhaust from Aircraft Turbine- Engines Annales Geophysicae- Atmospheres Hydrospheres and Space Sciences 12, no. 10-11 (1994): 944-55. One of the first studies to quantify speciated VOCs in aircraft exhaust (CFM56 and TF39). Also looked at photochemical reactivity and showed that carbon mass balance was mostly closed (i.e., sum of speciated VOCs = UHC) Srivastava, Amal 2011 Improving Departure Taxi Time Predictions Using ASDE-X Surveillance Data In the Digital Avionics System Conference IEEE/AIAA 30th Proceedings, October 16-20, 2011, Seattle, Washington, edited by Arnal Srivastav, 2B5-1 - 2B5-14. doi:10.1109/DASC.2011.609598 9 Two models that were developed to predict taxi-out time are presented. The models used data from JFK during the summer of 2010. Results are compared to values from FAA's Enhanced Traffic Management System (ETMS) Stettler, M.E.J., Eastham, S., Barrett, S.R.H. 2011 Air Quality and Public Health Impacts of UK Airports. Part I: Emissions Atmospheric Environment 45, no. 31 (2011): 5415-2 Generated a modified EDMS-based emission inventory for UK airports. For time-in-mode, they subdivided the idle phase into landing roll, reverse thrust, taxi in, taxiway acceleration taxi out, taxiway acceleration, hold. Used.BFFM2 and method of Kim et al 2005 to generate HC and CO EIs. Stettler, M.E.J., Eastham, S., Barrett, S.R.H. 2011 SUPPORTING INFORMATION for Air Quality and Public Health Impacts of UK Airports. Part I: Emissions Atmospheric Environment 45, no. 31 (2011): 5415-2 Contains detailed information on how they (Stettler et al) generated time-in-modes based on # of runways, etc. The Boeing Company 2002 737-300/400/500 Flight Crew Training Manual Document No.FCT 737 CL revision 2. October 31, 2002. Provides information and recommendations for maneuvers and techniques for the Boeing 737-300/400/500 series aircraft Thrasher, T., Nguyen, A., Hall, C. Fleming, G., Roof, C., Balasubramamian, S., Grandi, F., Usdrowski, S., Dinges, E., Burleson, C., Maurice, L., Iovinelli, R. 2007 AEDT Global NOx Demonstration USA/Europe ATM R&D Seminar 2007 Provides results of a demonstration of the capabilities of AEDT and concludes that that the dynamic gate-to-gate aircraft performance element of the model is "successful"

A-38 Author(s) or Regulatory Agency Year Title Citation Annotation Timko, M. T., Herndon, S.C, Wood, E., Onasch, T., Northway, M., Jayne, J., Canagaratna, M., Miake-Lye, R. 2010 Gas Turbine Engine Emissions - Part 1: Volatile Organic Compounds and Nitrogen Oxides J. Eng. Gas Turb. Power 132:doi: 10.1115/1.4000131. Summarizes VOC and NOx emission indices from the JETS/APEX2 and APEX3 projects. Covers following common engine types: CFM56-7B22, CFM56-3B1, RB211- 535E4-B, PW4158, AE3007, CJ6108A. Measured HCHO EI roughly scales with ICAO UHC EI (e.g., RB211 engine have both relatively low UHC EI and low HCHO EI). Timko, M., Herndon, S., de la Rosa Blanco, E., Wood, E., Yu, Z., Miake-Lye, R., Knighton, W., Shafer, L., DeWitt, M., Corporan, E. 2009 Combustion Products of Petroleum Jet Fuel, a Fischer– Tropsch Synthetic Fuel, and a Biomass Fatty Acid Methyl Ester Fuel for a Gas Turbine Engine Combustion Science and Technology 183, no. 10 (2011): 1039-68. Emission measurements from the Alt Fuels campaign. Relevant findings: speciation of VOC emissions somewhat affected by fuel content (aromatic versus oxygenates) Transport Canada 2010 Taxi Check and Procedures Transport Canada. n.p. Web. October 2, 2013. Provides Standard Operating Procedures (SOPs) for taxiing aircraft on departure Transport Canada 2010 Aircraft Icing Operations - Taxi Transport Canada. n.p. Web. October 2, 2013. Provides Standard Operating Procedures (SOPs) for taxiing aircraft during conditions when aircraft are deiced. Turgut, E., Usanmaz, O., Rosen, M. 2013 Empirical model assessment of commercial aircraft emissions according to flight phases International Journal of Energy and Environmental Engineering 4, no. doi:10.1186/2251-6832-4- 15 (2013). Secured FDR data from ten randomly selected B737-800 (CFM56-7B26) flights in Turkey. Fig 7 shows the actual fuel flow rates – looks like idle values are ~0.09 kg/s, with occasional acceleration bursts no more than 0.25 kg/s. They also present their own model for predicting true emission indices as a function of fuel flow rate for NOx, CO, and HC. U.S. DOT Volpe Center 2012 Aviation Environmental Design Tool (AEDT) 2a Technical Manual DOT-VNTSC-FAA-12-09 U.S. Environmental Protection Agency 2012 Control of Air Pollutant from Aircraft and Aircraft Engines; Final Emission Standards and Test Procedures - Summary and Analysis of Comments EPA-420-R-12-011 EPA's responses to comments received on the July 2011 proposal for new NOx emissions standards for aircraft turbofan and turbojet engines with rated thrust greater than 26.7 kilonewtons. U.S. Environmental Protection Agency 2012b EPA Adopts NOx Emission Standards for Aircraft Gas Turbine Engines EPA-420-F-12-027 Overview of the Tier 6 and Tier 8 standards for NOx.

A-39 Author(s) or Regulatory Agency Year Title Citation Annotation U.S. Environmental Protection Agency 2008 Clean Air Act 42 U.S.C. §§ 7401 et seq (2008) The Clean Air Act (CAA) is the Federal law that authorizes EPA to regulate air emissions from stationary and mobile sources, to establish National Ambient Air Quality Standards (NAAQS) to protect public health and public welfare and to regulate emissions of hazardous airborne pollutants and to coordinated all Federal noise pollution control activities. CAA was originally enacted in 1963 with significant amendments enacted in 1970, 1977 and 1990. U.S. Environmental Protection Agency 2013 Control of Air Pollution from Aircraft and Aircraft Engines 40 C.F.R. 87 (2013) Under the authority of 42 U.S.C. §§ 7401 et seq, 40 C.F.R. Part 87 contains the codification of the EPA rules and regulations related to the control of emissions from aircraft and aircraft engines. U.S. Environmental Protection Agency 2013b National Primary and Secondary National Ambient Air Quality Standards 40 C.F.R. 50 (2013) Under the authority of 42 U.S.C. §§ 7401 et seq, 40 C.F.R. Part 50 contains the codification of the EPA rules and regulations related to establishing National Ambient Air Quality Standards (NAAQS) under Section 109 of the CAA . U.S. Environmental Protection Agency 2013c Requirements for Preparation, Adoptions, and Submittal of Implementation Plans 40 C.F.R. 51 (2013) Under the authority of 42 U.S.C. §§ 7401 et seq, 40 C.F.R. Part 51 contains the codification of the EPA rules and regulations related to procedures for developing Implementation Plans to meet the NAAQS. State and Federal Implementation Plans (i.e., SIPs and FIPs, respectively) apply to areas that are not in attainment with the NAAQS. Included are the regulations related to conformity to which airports are subject and the specification of models (e.g., EDMS) required for airport air quality evaluations. U.S. Environmental Protection Agency 2013d Determining Conformity of Federal Actions to State or Federal Implementation Plans 40 C.F.R. 93 (2013) Under the authority of 42 U.S.C. §§ 7401 et seq, 40 C.F.R. Part 93 contains the codification of the EPA rules and regulations related to SIP and FIP determinations of conformity resulting from federal actions. U.S. Environmental Protection Agency 1982 National Environmental Policy Act, as Amended 42 U.S.C. §§ 4371 et seq National Environmental Policy Act (NEPA) is the Federal law establishing the President's Council on Environmental Quality (CEQ). NEPA established procedural requirements for the preparation of environmental assessments (EAs) and environmental impact statements (EISs) with responsibility assigned to the various overseeing federal agencies. NEPA was originally enacted in 1970 and amended in 1975 and 1982.

A-40 Author(s) or Regulatory Agency Year Title Citation Annotation U.S. Environmental Protection Agency 1978 Council on Environmental Quality 40 C.F.R. Chapter V Under the authority of 42 U.S.C. §§ 4371 et seq and § 309 of the CAA, 40 C.F.R. Chapter V contains the codification of the rules and regulations of the Council on Environmental Quality as pertaining to the procedural requirements for the preparation of environmental assessments (EAs) and environmental impact statements (EISs). U.S. Environmental Protection Agency 2004 Guidance on Airport Emission Reduction Credits from Early Measures through Voluntary Airport Low Emission Programs U.S. Environmental Protection Agency. Office of Air Quality Planning and Standards. Guidance on Airport Emission Reduction Credits from Early Measures through Voluntary Airport Low Emission Programs. September 2004. Regulatory guidance document. As permitted by the Vision 100 - Century of Aviation Reauthorization Act, this document provides guidance on emission reduction credits for voluntary early emission reduction programs at airports under the General Conformity and New Source Review (NSR) programs. Unique 2004 Aircraft NOx-Emissions within the Operational LTO Cycle Prepared by Emmanuel Fleuti and Juan Polymeris in cooperation with Swiss Flight Data Monitoring, CH-8058. Zurich. Evaluates flight specific data aimed at defining an operational LTO cycle, deriving operational times in mode, fuel flow and emissions data. United Kingdom Department for Transport 2012 Reducing the Environmental Impacts of Ground Operations and Departing Aircraft: An Industry Code of Practice Prepared by the Departures and Ground Operations Code of Practice Working Group Presents an interim voluntary Code of Practice for aircraft operators shutting down one or more engines during taxi-in operations Waitz, Ian, et al 2008 ACRP Report 9: Summarizing and Interpreting Aircraft Gaseous and Particulate Emissions Data ACRP Report 9 or Waitz et al. 2008 A primer document (compendium) of various field studies including evaluating emissions by thrust setting. Should be cross-checked to make sure a pertinent reference is not omitted from our final literature review. Watterson, J., Walker, C., Eggleston, S. 2004 Revision to the Method of Estimating Emissions from Aircraft in the UK Greenhouse Gas Inventory: Report to Global Atmosphere Division, Defra Report to Global Atmosphere Division, Defra Contains the methodology used by Stettler et al to derive their time-in-modes. Wayson, R., Fleming, G., Garrity, N., Kim, B., MacDonald, J., Lau, M., Draper, J. UNK Validation of FAA's Emissions and Dispersion Modeling System (EDMS): Carbon Monoxide Study United States Department of Transportation/Volpe Center, Paper 69607 A study of carbon monoxide measurements at 25 locations at a major U.S. international airport (airside and landside). The EDMS-predicted concentrations were compared to measured concentrations and a detailed statistical assessment was performed of the AERMOD dispersion algorithm.

A-41 Author(s) or Regulatory Agency Year Title Citation Annotation Wollenheit, R., Muhlhausen, T. 2013 Operational and Environmental Assessment of Electric Taxi Based on Fast-Time Simulation Transportation Research Record: Journal of the Transportation Research Board, No. 2336, pp. 36-42 Presents the findings of a fast-time simulation model at two airports to demonstrate the fuel savings of an electric taxi system Wood, E. C., Yelvington, P.E., Timko, M.T., Herndon, S.C., Miake- Lye, R. 2008 Speciation and Chemical Evolution of Nitrogen Oxides in Aircraft Exhaust near Airports Environmental Science & Technology. March 15, 2008 Comparison of NOx EIs from in-use at OAK to "formal engine tests" shows that idle thrust appears to be approximately 4 percent thrust for non-accelerating aircraft and approximately 15 percent for accelerating aircraft. Wood, E., Herndon, S., Miake-Lye, R., Nelson, D. 2008 ACRP Report 7: Aircraft and Airport-Related Hazardous Air Pollutants: Research Needs and Analysis ACRP Report 7 or Wood et al. 2008 Summarizes the uncertainties associated with emissions from aircraft Yaworksi, M., Dinges, E, and Iovinelli, R. 2011 High-Fidelity Weather Data Makes a Difference Calculating Environmental Consequences with FAA's Aviation Environmental Design Tool In the USA/Europe Air Traffic Management Research and Development 9th Seminar, June 14-17, 2011, Berlin, Germany. Unpaginated document. Examines the use of high-fidelity weather data to model aircraft performance for the purpose of quantifying environmental consequences using the AEDT model. Yelvington, P. E., Herndon, S., Wormhoudt, J., Jayne, J., Miake-Lye, R., Knighton, W., Wey, C. 2007 Chemical Speciation of Hydrocarbon Emissions from a Commercial Aircraft Engine Journal of Propulsion and Power 2007, 23 (5), 912-918 Description of VOC EIs measured during the APEX campaign (from a CFM56-2C1). Showed strong inverse dependence of VOC emissions on ambient temperature and fuel flow rate. Yin, K., Tian, C., Wang, B., Quadrifoglio, L. 2012 Analysis of Taxiway Aircraft Traffic at George Bush Intercontinental Airport, Houston, Texas Transportation Research Record: Journal of the Transportation Research Board, No. 2266, pp 85-94 Assesses the congestion at IAH by analyzing taxi times and flight data during different hours of the day Yoder, Tim 2005 Development of Aircraft Fuel Burn Modeling Techniques with Applications to Global Emissions Modeling and Assessment of the Benefits of Reduced Vertical Separation Minimums Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 50 p. Discusses methods to improve the FAA's System for assessing Aviation's Global Emissions (SAGE) specifically focusing on the way fuel consumption is calculated and improving the algorithms to process weather information. UNK = Unknown 77