Environmental Optimization of Aircraft Departures: Fuel Burn, Emissions, and Noise (2013)

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.

64 B-1. Introduction The purpose of this task was to identify representative noise- sensitive airports where one or more of the various types of NAPs have been implemented and develop a testing protocol to determine the impact of NAPs on noise, fuel burn, emissions, and capacity. The goals of this case study analysis are to • Generate a large amount of results data from which the most optimal procedures will be determined by comparing results to the existing NAPs. • Present results without qualification, instead providing information on the tradeoffs between all modeled variables. • Model single-event environmental effects and quantify potential effects in capacity. Note: The case studies will not be able to address ground operations (taxiing) because only single-departure events will be modeled. B-2. Case Study Airport Selections This section describes the process of analyzing and select- ing airports for the case study analysis performed under Task 4. B-2.1. NAP Information The Boeing Commercial Airplane Company maintains a Noise and Emissions Regulations database (“Boeing NER database”) with 643 of the world’s airports which have some type of environmental restrictions on aircraft operations. The Boeing NER database shows basic airport information (i.e. where the airport is located, the IATA airport code, etc.) and provides details of the associated environmental restrictions. Note that in many cases these restrictions are voluntary. Boeing makes this airport database publicly avail- able at http://www.boeing.com/commercial/noise/list.html. As the first step in this task, we transferred the entire Boeing NER database to a Microsoft Excel spreadsheet (“extraction spreadsheet”) so we could manipulate the data more easily. B-2.2. Flight Track Information The AEDT airports database contains flight track informa- tion for a number of airports. The flight track data for these airports generally comes from the International Commercial Aviation Organization’s Committee on Aviation Environ- mental Protection (ICAO/CAEP) MAGENTA (Model for Assessing Global Noise Exposure to the Noise of Transport Aircraft) project, which estimates airport noise around the world. The track data was extracted from the AEDT airports database using a SQL query. Although flight track data may exist in the AEDT database, the quality of that data was not documented at this point in the process; only the existence of the data was of interest. There were 220 airports extracted from the AEDT airports database which have flight track information. However, not all of the airports from the AEDT airports database were in the Boeing NER database, so we noted on the Boeing data- base spreadsheet which airports in the Boeing NER database also existed in the AEDT flight track database. Since some of the airports had slightly different names between the two databases, the IATA airport code was used to ensure airports with slightly different names between the two databases were the same airport. There were 172 airports that were both in the Boeing NER database and that had tracks in the AEDT airport database. B-2.3. Operational Information For each of the United States airports in the Boeing NER database with a Noise Abatement Procedure (NAP) or a pref- erential runway program, we extracted the total operations A P P E N D I X B Testing Protocol for Case Studies

65 (both arrivals and departures) by seat class category from the FAA Enhanced Traffic Management System (ETMS) data- base of operations for 2006. These extracted operations data were added to the extraction spreadsheet. The operational data from 2006 for each of the airports (including detailed fleet mix) were separated into departures and arrivals accord- ing to their seat class following the ICAO seat class definition, as listed in Table B-1. The seat class data were used to help determine which of the airports we should further examine while maintaining a balance between general aviation air- ports (primarily served by aircraft in the lower seat classes), regional airports (primarily served by aircraft in the middle seat classes), and international and hub airports (served by aircraft in the top seat classes). We also separated departure operations by distance traveled; this may be useful since distance flown is a reasonable surro- gate for aircraft weight, which in turn directly influences aircraft performance. Aircraft weight as a function of distance traveled will be determined using the AEDT concept of stage lengths, where the weight of the aircraft increases in discrete increments at 500 or 1000 nautical miles (NM) great circle trip distance increments. Table B-2 below shows the AEDT stage lengths. B-2.4. Air Quality Information Information on the status (nonattainment, maintenance, or attainment) of local air quality for criteria pollutants was col- lected for the airports in the extraction spreadsheet. This data was derived from a list of all U.S. commercial airports main- tained by the FAA’s Voluntary Airport Low Emissions Program (VALE). VALE aids commercial airports in reducing airport emissions by using Airport Improvement Program (AIP) funds and Passenger Facility Charges (PFC) to finance improvements. The VALE “List of U.S. Commercial Service Airports and their Nonattainment and Maintenance Status” is available at: http://www.faa.gov/airports/environmental/vale/. B-2.5. Selection of Airports Next, we noted the airports which met these criteria: depar- ture NAP information on the Boeing NER website, AEDT track information, and located in the United States. There were 81 airports that met all three criteria, each of which was examined in more detail by returning to the Boeing NER web- site and reviewing the available information. In this intermediate down-selection, we removed airports from further consideration if they had a NAP which could not be well modeled by single operations, was vaguely defined, or was not significantly different from the NAP at another air- port. In general, airports with relatively few operations were dropped from further consideration since the effectiveness of NAP operations would be relatively low. After an individual review of the 81 airports, 20 of the airports were considered for further analysis. From that list, 9 airports were selected for analysis and are listed in Table B-3. Although specifications on the NAP and other information of interest (airport type, air quality concerns, annual opera- tions, and proposed baseline aircraft type) are presented, they have been generalized (“sanitized”) in light of sensi- tivities the airports may have regarding presentation of such information. B-3. Testing Protocol The Testing Protocol defines the process for analyzing the case study airports to compute noise, emissions, fuel burn, and capacity. The goal of the case studies analysis is to provide detailed data on the tradeoffs between these variables for dif- ferent types of NAPs. This approach is necessary because it is unlikely that an airport would remove an existing NAP, but it is feasible that an airport could optimize its existing NAP to provide an emissions, fuel burn, or capacity benefit at the expense of changes in noise exposure. That is, since an exist- ing NAP is already optimal for noise exposure in the eyes of both airport and community, any changes to a NAP would be undesirable, unless supporting data can be used to show the benefits of making changes. The Testing Protocol specifies a parametric optimiza- tion process in which many combinations of ground tracks and profiles will be modeled for each airport. These tracks will Seat class Number of passenger seats 1 <20 2 20-50 3 51-99 4 100-150 5 151-210 6 211-300 7 301-400 8 401-500 9 501-600 10 600+ Table B-1. Seat class definitions. Stage Length Minimum Range (NM) Maximum Range (NM) 1 0 499 2 500 999 3 1,000 1,499 4 1,500 2,499 5 2,500 3,499 6 3,500 4,499 7 4,500 5,499 8 5,500 6,499 9 6,500 11,000 Table B-2. Stage length definitions.

66 cover variations of existing NAPs, as well as the most-direct ground track from takeoff to a departure fix. Capacity will be modeled by determining runway throughput for each airport. NAPs only have an impact on airport capacity in as much as they enable or result in changes in the speed of departing air- craft, the length of the common path that is shared by succes- sive departures, or the time between departures and arrivals. By comparing each of the result sets against the existing NAPs, we will determine the most optimal solutions. It is important to note that although many of the studies in the literature review employed detailed computer models which optimized ground tracks and profiles, such a model is beyond the scope of this project. Figure B-1 illustrates the 7-step process detailed in the Test- ing Protocol. B-3.1. Airport Data Collection The purpose of Step 1 is to review, for each airport, the airport and airspace layout, fleet mix, and existing NAPs. In addition, we will collect data to feed the AEDT model and GIS analysis. A base map was created for each airport. 1. Collect general airport data: runway layout and configu- rations (assume optimum configuration), and average weather (assume visual conditions). 2. Collect and review navigational charts and departure fix locations. 3. Compare published procedures, including NAPs, to the AEDT airports database. Review AEDT ground tracks and profiles per runway and terminal airspace. Determine Airport Type of Airport Existing Departure NAP (1) Air Quality Concerns(2) Approx. Annual Operations(3) Baseline Aircraft Type(3) APRT1 CargoHub - Airport goes to single direction operation at night - Departure turns based on distance from airport Ozone, PM2.5 100,000 – 200,000 A300 APRT2 Hub,Coastal - RNAV NAP procedures Ozone, CO 300,000 – 400,000 757-200 APRT3 Hub - Community close to airport Ozone, CO, PM2.5 300,000 – 400,000 MD-88 APRT4 Hub - Fanning NAP CO, SO2 400,000 – 500,000 DC9-30 APRT5 Hub,Coastal - Multiple turn restrictions on departure Ozone, CO, PM2.5 300,000 – 400,000 747-400 APRT6 Hub - Departure heading gate (distance-based turns) Ozone, CO, PM10 500,000 – 600,000 737-700 APRT7 GeneralAviation - Distance-based turns Ozone, CO, PM10, PM2.5, NO2 < 50,000 Gulfstream GIIB (Noise Stage 2) APRT8 Regional - Heading restriction based on altitude Ozone < 50,000 CRJ-200 APRT9 Regional - Altitude-based headings Ozone 100,000 – 150,000 EMB-145 Sources: (1) Boeing NER Database 2009; (2) VALE Airport Status List 2009; (3) ETMS 2006 Table B-3. Proposed case study airports. Figure B-1. Overall process. 1. Airport Data Collection 2. Develop Ground Tracks 3. Develop Profiles 4. Environ- mental Modeling 5. Capacity Modeling 7. Tradeoff Assessments 6. Analyze Results Feedback loop to Step 2 (if needed)

67 quality and appropriateness of AEDT tracks and profiles for modeling NAP effects. 4. Identify NAP procedures as modeled in the AEDT airports database (see Table B-3 for each airport’s defined NAPs). Determine if additional modeling is necessary to capture existing NAP. 5. Use ETMS operational data to determine fleet mix, num- ber of operations, and departure stage lengths. Classify aircraft by size (small, large, B757, or heavy). Set up AEDT to model the correct baseline aircraft type and baseline stage length (see Table B-3). 6. Collect geographic data into a GIS database. Will include airport runways and boundary, local features (roads, bodies of water), and census population data. Create a base map for each airport. Land use data will not be collected. B-3.2. Develop Ground Tracks The purpose of Step 2 is to develop, for each airport, a set of ground tracks which represent the existing NAP, a track flying directly to the departure fix, and multiple intermediate tracks which fill the airspace between them. 1. Define the NAP ground track in AEDT. In some cases, the AEDT airport database will suffice. In other cases, an NAP ground track may need to be created based on published procedures. 2. Define the most-direct possible ground track which uses the same runway as the NAP and reaches the same departure fix, in the shortest possible length/time. This ground track (series of X-Y points) will be built in GIS then imported to AEDT. For each aircraft type the minimum safe distance from takeoff before a turn can be initiated, and the mini- mum radius of turn, will be determined. 3. Define a set of alternate ground tracks (each a series of X-Y points) which fill a range of possible trajectories bounded by the NAP and most-direct ground track. This will include: tracks making turns off of initial runway heading at a variety of distances from takeoff; tracks which follow an NAP to a variety of positions then turn to the departure fix. The set of alternate ground tracks will be spaced such that the farthest distance between each track is no more than 0.5 NM. See Figures B-2 and B-3. 4. Once all ground tracks are defined in GIS, convert to AEDT format. Input to AEDT databases. See flow chart in Figure B-4. Each gap is ≤ 0.5 NM Most-direct Runway Fix NAP ground track Figure B-2. Illustration of NAP turn restriction ground tracks. Each earlier turn spaced at 0.5 NM Most-direct Runway Fix NAP ground track NAP turn point Figure B-3. Illustration of DME turn restriction ground tracks.

68 B-3.3. Develop Profiles The purpose of Step 3 is to define NADP-1 and NADP-2 profiles for each aircraft type. Since the current AEDT alpha version does not yet have a user interface, INM ver- sion 7.0b will be used to edit and create profiles using pro- cedure steps. 1. For each aircraft type, select the AEDT default Standard, NADP-1 and NADP-2 profile for the appropriate stage length(s). Note that the AEDT default profiles are the same as those in INM version 7.0b (see Figure B-5). 2. Review the AEDT default NADP-1 and NADP-2 profiles compared to the ICAO 2007 NADP survey (see Appen- dix C). Based on this review, take one of the following actions (listed in order of preference): • Use AEDT default NADP’s, if both exist for given aircraft type. • Modify the default NADP’s, using the ICAO 2007 NADP survey data and using the INM interface and guidance from the INM User’s Manual. See Figure B-6. • If no default NADP’s exist in AEDT, build ICAO 2007 NADP procedure steps using the INM interface and guidance from the INM User’s Manual. See Figure B-6. • For business jets, use National Business Aviation Associa- tion (NBAA) Close-In Departure Procedure (available at: http://www.nbaa.org/ops/environment/quiet-flying). 3. If necessary, extend profile above the AEDT cutoff altitude of 10,000 feet AGL. 4. Input all modified/new NADP profiles to AEDT fleet database. B-3.4. Environmental Modeling The purpose of Step 4 is to execute the FAA’s AEDT to compute noise, fuel burn, and emissions for every possible AEDT Airports Database Convert Vector Tracks to Point Tracks Select NAP Representative tracks Create alternative NAP tracks Create direct track Input to AEDT flight track database Figure B-4. Track creation process. Figure B-5. Example of AEDT standard and NADP altitude profiles. INM standard profile Copy to new profile Build new profile (modify procedure steps) Review procedure steps in INM DBF files Copy database records to AEDT format Input into AEDT fleet database Figure B-6. NADP creation process.

69 combination of ground tracks and profiles (including NADPs and standard profiles). Each unique combination will be referred to as a “case.” 1. Organize all inputs (ground tracks and profiles) for all air- ports within the AEDT databases. Perform quality control of all inputs. 2. Define departure operations for the correct aircraft type for each case. 3. Set up AEDT Configuration files (separate files for stan- dard profile fleet database, close-in and distant NADP fleet databases). 4. Final quality control and bug check. 5. Run AEDT results processor module; query results data- base to confirm completion. B-3.5. Capacity Modeling The purpose of Step 5 is to perform an analysis of runway throughput for each airport to determine the effects of NAPs on capacity (see Figure B-7). See Appendices C and D for details on capacity impacts. 1. AEDT will output 4-dimensional trajectories in the perfor- mance database (i.e., speed, ground track position, and alti- tude of aircraft for each segment of departure flight path). 2. Generate and process sequences for each runway separately. 3. Determine baseline capacity for each runway separately, then weight each runway and sum the total airport capac- ity. Assume no runway interactions for simplicity. 4. Create baseline capacity curve (Pareto frontier) showing arrivals per hour vs. departures per hour. See example in Figure B-8. 5. Determine effect of using noise abatement profiles: Model the effect of NADP-1 and NADP-2 (which affect departure speeds) on capacity for one given runway. Then weight this new capacity against the rest of the airport runways. Create alternate capacity curves. 6. Determine effect of using noise abatement ground tracks: Model ground tracks, which affect length of common AEDT trajectory information Generate and process sequences Determine baseline capacity Baseline capacity curve Model effect of NADPs Compile all modeling results Model effect of ground track NAPs Figure B-7. Capacity modeling process. Figure B-8. Example of calculated runway capacity Pareto frontier.

70 departure paths, for one given runway. Then weight this new capacity against the rest of the airport runways. Cre- ate several alternate capacity curves (one for each modeled noise abatement track). B-3.6. Analyze Results The purpose of Step 6 is to compile results databases for each airport. 1. Noise Results – Regularly-spaced noise grids will be out- put from AEDT for each case. • Noise contour maps: – SEL 85 dB contour maps (level based on speech and classroom speech interference research which uses Lmax 75 dB; however AEDT does not output Lmax met- ric at this time) • Difference grids will be computed using NMPlot. These grids will show the difference as increase and decrease in noise from the existing NAP. – Population changes within difference areas. • Tables of population within noise contour and contour area (for all cases) and difference compared to existing NAP. 2. Emissions & Fuel Burn Results – AEDT results processor will be used to output tables of all available pollutants, fuel burn, and performance data for each case. • AEDT will output the following pollutants for local air quality: – CO; HC; PM2.5; SOx – AEDT cannot compute ozone directly; instead AEDT will compute pollutants which are precursors to ozone such as NOx and VOC • AEDT will output the following pollutants for green- house gases (GHG): – CO2 • AEDT will output the following performance data: – Fuel burn (FB) and time of flight • We will set a cutoff point for all cases within each air- port. The purpose of the cutoff is to compute emissions and performance across all cases in a consistent manner by comparing the results from equal distances/altitudes flown. – Low Cutoff: Mixing layer 3,000 feet AFE – High Cutoff: Distance to fix (i.e., common point of convergence of ground tracks) • Emissions inventory tables will be developed for each case showing total mass of emissions. Summary tables will be created to show the percent differences from the existing NAPs. 3. Capacity Results – The metrics used will include through- put (number of arrivals and departures per hour on a runway) and time (between consecutive departures). • Compare and contrast all capacity curves. Graphically assess how the Pareto frontier shifts for each noise abate- ment procedure compared to baseline capacity curve. • Tables of throughput and time for each case. • Summary tables of percent differences in throughput and time comparing each case to the existing NAP. 4. Technology Assessment – Determine the source noise reduction theoretically needed to result in no increases in noise. This will be automated using the Noise Source Reduction Optimizer (NSRO). See Figure B-9 below. • Use noise difference grids from Step 6.1 to determine the greatest value of noise increase (for each case). • Compute equivalent reduction of operations required to model reduced source noise level (per ECAC 2005): N = 10DL/10 (N = Number of aircraft operations; DL = Noise reduction in dB) • Run new source noise through AEDT. • Re-compute noise difference grids. • Once NSRO has determined the source noise reduc- tion level, classify the reduction according to technol- ogy goals specified under NextGen, CLEEN, and NASA Output AEDT noise level grids Convert to NMPlot file format Use NMPlot to compute difference grid Compute greatest noise increase Convert delta dB to ops reduction Re-Run AEDT with reduced source noise If no increases, process is complete Return to Box 1 Figure B-9. NSRO process.

71 Fundamental Aeronautics Research Program (i.e., N+1, N+2, N+3). 5. Feedback to Step 2 – Perform a feedback loop, if deemed necessary by the team, to develop and analyze additional ground tracks and/or profiles. The decision will be based on the results analysis. The goal is to allow for further refinement of ground track or profile procedures to bet- ter model the most optimal procedures. For example, ground tracks may be added to the analysis to study the effect of noise on specific locations; or, a variant of NADP-1 may be added to the analysis to compare thrust cutback at different altitudes. • Additional procedures/ground tracks modeled at SFO, BOS. B-3.7. Tradeoff Assessments The purpose of Step 7 is to synthesize all case study results, for each airport and across the set of all airports. This will include additional tables and charts, and a discussion of the accuracy of the results. The most optimal procedures will be selected based on noise level, emissions, fuel burn, and capac- ity. Population impact will not be used as a selection criterion since it may vary locally. 1. Selection of most optimal procedures based on results of Step 6 • Combine ground track and NADP into optimal procedure • Discuss day vs. night procedures • Emissions – pollutants of concern vary by airport • Selection of optimal procedures: – Minimal change in noise levels (population and area) – Greatest FB and emissions reduction – Greatest capacity increase 2. General accuracy of modeling results (as compared to real operations) • Full thrust modeling assumption vs. reduced thrust used in practice – Reference to ICAO (2007) which presented results with and without reduced thrust • Single-event modeling does not consider real-world dispersion of flights around a published procedure • Phase 2 will address capacity and ground operations more completely