Analyzing & Implementing Delayed Deceleration Approaches

Analyzing & Implementing Delayed Deceleration Approaches Tom G. Reynolds, Emily Clemons & Lanie Sandberg R. John Hansman & Jacquie Thomas 12 th USA/Europe ATM Research & Development Seminar, Seattle, WA June 26-30, 2017 DISTRIBUTION STATEMENT A. Approved for public release: distribution unlimited. 2017 Massachusetts Institute of Technology

Aircraft Operations Environment Assessment Initial scoping study to identify & evaluate operational techniques to reduce fuel burn and environmental impacts in the near/midterm with minimal implementation barriers* Objectives: Identification, evaluation and prioritization of 60+ options Detailed analysis of benefits & barriers of promising options Cruise Altitude and Speed Optimization Departure (D): 11 mitigations ORIGIN AIRPORT Cruise (C): 14 mitigations Surface (S): 16 mitigations Miscellaneous (M): 11 mitigations Approach (A) & Landing (L) 9 mitigations DESTINATION AIRPORT Delayed Deceleration Approaches Promote deployment of best practice operations Socialization with stakeholders Integration strategies for current & future operations DDA - 2 *Marais, K., et al., Evaluation of Potential Near-term Operational Changes to Mitigate Environmental Impacts of Aviation, Journal of Aerospace Engineering, Vol. 227, No. 8.

Delayed Deceleration Approach (DDA) Concept Keep aircraft clean for longer on approach when appropriate without impacting terminal area entry or final approach stabilization criteria Between these speed gates, opportunity for encouraging more efficient approach speed profiles Dirty configuration Terminal area entry speed Clean configuration Delayed Decel. => Low Power/ Low Drag Typical Conventional Sample flap 1 Sample flap 2 Final approach speed Runway 230-250 kts IAS Airspeed Airspeed 160-180 kts IAS 30 NM Distance to touchdown Distance to Touchdown 10 NM DDA - 3

Outline Delayed Deceleration Approach (DDA) Concept DDA fuel burn and emissions reduction potential Analyzing speed profiles and barriers at US airports Noise analysis Implementing DDA via RNAV procedures Conclusions & Recommendations DDA - 4

DDA Benefits Potential European A320 Flight Data Recorder Analysis (similar results for B757 & B777) A320 performance profiles (NM) (NM) Lowest fuel burn flights (green profiles) associated with delayed deceleration 30-50% fuel burn reduction potential from DDAs from 10,000 ft to touchdown (NM) (NM) DDA - 5

System-Wide DDA Benefit Potential Estimated system-wide fuel saving benefits potential of increased DDA utilization Aircraft Weight Class Example Aircraft Types DDA Saving per Approach Approx # Flights per Day Fuel Reduction Benefit Pool (gal/yr) $ Annual Savings per 1% Inc. in DDA Use* DDA Fuel Saving Relative to Average (lbs) 400 350 300 250 200 150 100 50 0 y = 0.7465x + 89.752 R² = 0.847 Based on FDR analysis RJ Small NB A320 B757 Large NB Two Engine WB B777 0 50 100 150 200 250 300 350 400 Maximum Gross Take-Off Weight (metric tons) Four Engine WB RJ Small NB Large NB Two Engine WB Four Engine WB CRJ ERJ A320 B737 120 7500 49m $1.5m 146 14,400 115m $3.5m B757 183 1,800 18m $0.5m A330 B777 A340 B747 276 3,900 59m $1.8m 375 2,400 49m $1.5m Totals 30,000 290m $8.8m DDA - 6 * FAA investment analysis recommended fuel price *FAA investment analysis recommended fuel price of $3.02/gal

Fuel Burn Correlations with Other Observable Metrics Need to find correlations between fuel burn and states observable in US radar track data Correlation analysis identified Airspeed and Time with Flap1 as main drivers of fuel burn Flap1 speed 180-210 kts for most large aircraft Time flown below 180kts used as proxy for fuel burn for large weight category aircraft DDA - 8

Quantifying Speed Profiles at US Airports Analyzed speed profiles at range of US airports Capacity-constrained standalone airports (ATL, LAX, BOS, CLT) New York metroplex airports (EWR, JFK, LGA) Washington DC metroplex airports (DCA, IAD, BWI) Capacity-unconstrained standalone airports (STL, RIC, DFW) Wind Data DDA Analysis Metrics Radar Archives Airspeed Estimation Fuel Efficiency: Time Flown Below Throughput Efficiency: Minimum Approach Spacing DDA Opportunity Evaluation 9 months of radar archives from 2011 and 2015 Ground speed converted to airspeed using NARR wind data DDA - 9

ATL Approach Speed Analysis Airspeeds <= 210KIAS <= 180KIAS ATL Large aircraft arrivals from ZDC, 4 sample days 20 NM DDA - 10

ATL Approach Speed Analysis Time flown below 180 kts correlated most strongly with fuel burn for large aircraft type Arrivals from ZDC JAN SEP 2011 Cumulative curves of time flown below 180 kts used as key airport speed metric DDA - 11

NYC Metroplex Approach Speed Analysis EWR <= 210KIAS <= 180KIAS LGA <= 210KIAS <= 180KIAS JFK <= 210KIAS <= 180KIAS LGA Earlier decelerations generally observed under IMC compared to VMC LGA under IMC has earliest decelerations EWR JFK 100 Percentage of All Landings 80 60 Arrivals from ZDC JAN SEP 2011 40 LGA VMC n = 10,530 LGA IMC n = 1,600 JFK VMC n = 9,070 20 JFK IMC n = 1,048 EWR VMC n = 4,435 15 mi EWR IMC n = 775 0 Arrivals from ZDC, 8 sample days 1 3 5 7 9 11 13 15 Time Flown Below 180 kt (min) DDA - 12

Airport Approach Speed Profile Comparison Time Flown Below 180 kts for 50% of Flights (mins) 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 VMC IMC Weighted DCA LGA EWR BOS IAD JFK BWI LAX ATL RIC STL Constrained metroplex Arrivals from ZDC JAN SEP 2011 Unconstrained standalone DDA - 13

Analyzing Drivers of Speed Profiles Need to understand causes of differing speed profiles to identify opportunities for increased DDA operations If primary drivers to encourage greater DDA usage are easily modifiable (e.g., ATC training or procedures) => good target airports More difficult to increase DDA-type procedures at airports where primary drivers are elements such as airspace or airport constraints Created decision trees to find combinations of independent variables correlated with time flown below 180kts Independent Variables Weather (VMC or IMC) Hourly Airport Acceptance Rate (AAR) Total arrival demand (15 min bins) Airport configuration Airline DDA - 14

Classification Tree Approach Technique identified key drivers of approach speed profiles Wx=VMC ATL Example Condition 1. Weather conditions (VMC vs. IMC) Config=West flow 2. Dominant operator 4 3. Airport configuration AAR >= 107.5 Airline=Delta 4. Higher capacities (AAR) ArrDem >= 18.5 AAR >= 118.5 4 1 Relative importance varies by airport Dominant carrier influence suggests airline procedural effect 3 AAR >= 113 AAR >= 114.5 4 3 2 1 2 Time Flown Below 180kts (s) 1: t < 122 2: 122 <= t < 169 3: 169 <= t < 267 4: t >= 267 DDA - 15 AAR = Airport Acceptance Rate (per hr) ArrDem = Arrival Demand (per 15 mins)

DDA Noise Impacts Airframe and engine noise both important during approach & landing Airframe noise Engine noise Landing gear noise Noise hot spots Empirical & modeling studies conducted to understand DDA noise impacts DDA - 17

BOS Noise Measurement Campaign Measurement Period: 11/13/15 1/25/16 Radar Flight Track Data Airline Standard Operating Procedures, Flap Schedules, etc. DDA Opportunity Analysis Airport Selection For DDA Feasibility Noise Monitor Site Selection Noise Measurement Campaign Noise & Aircraft Track Analysis Research Outputs: Noise modeling validations Noise as f(speed, configuration, approach procedures, aircraft type) Time Flown Below 180 kts for 50% of Flights (min) 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 VMC IMC Weighted DCA LGA EWR BOS IAD JFK BWI LAX ATL RIC STL 50 100 150 200 250 300 350 400 450 500 Runway 22L/22R Arrivals Jan-Mar and Jun-Aug 2014 5 nmi BOS 50 100 150 200 250 300 350 400 450 500 1000 900 800 700 600 500 400 300 200 100 0 No. of Radar Surveillance Hits DDA - 18

Noise Measurement Results A320 (n=135) Monitor A A320 (n=269) Monitor B A320 (n=365) Monitor C Linear fit slope: -0.01 Linear fit slope: -0.03 Linear fit slope: -0.06 B737 (n=163) B737 (n=295) B737 (n=352) Linear fit slope: 0.00 Linear fit slope: -0.02 Linear fit slope: -0.05 E190 (n=58) E190 (n=112) E190 (n=153) Linear fit slope: -0.03 Linear fit slope: -0.02 Linear fit slope: -0.01 Slight negative trends indicate marginal decrease in noise with increasing airspeed on average Similar results for Lmax 10-15 dba variability in data Not removed when correction for energy change rate applied Attributed to atmospheric differences DDA - 19

DDA - 20 TASOPT, BADA4 & ANOPP Model Integration

Modeling of Enhanced DDA Operations Modeled range of approaches with different aircraft types & profiles A320 Airframe Only Total Airframe & Engine Engine Only Overall, negligible effect of DDA speed profiles on noise impacts on ground DDA - 21

Outline DDA fuel burn and emissions reduction potential Analyzing speed profiles and barriers at US airports Noise analysis Implementing DDA via RNAV procedures Feasibility of RNAV procedures for DDA implementation Design considerations for RNAV DDAs Assessing RNAV Procedure Targets Conclusions & Recommendations DDA - 22

Assessing Feasibility of RNAV DDAs Biggest uncertainty in effective implementation of DDA: track distance to touchdown PBN (RNAV/RNP) approach procedures reduce uncertainty Simulated approaches with: Same programmed lateral/vertical route as published RNAV arrivals Different speed constraints Different aircraft weights Utilized Lincoln FMS analysis capability Pegasus FMS flight code Integrated to B757-200 simulation Latitude (deg) 34.1 34 33.9 33.8 33.7 ATL DIRTY RNAV Arrival to Runway 26R Late decel. profile Early decel. profile ATL 160K BAMBU 10NM 150K 250K 200K 250K BAMMM 27NM ZELOW 20NM HAARY 40NM 230K 175K 33.6-84.4-84.2-84 -83.8 Longitude (deg) DIRTY 49NM 200K 185K 200K DDA - 23

Simulated B757 ATL RNAV DDA Arrivals Fuel Burn Inside TRACON 40 NM to 50 ft AGL (lbs) 1800 1600 1400 1200 1000 800 600 400 200 0 Aircraft Weight Light Medium Heavy Scenario 1: Late decel. Scenario 2: Medium decel. Scenario 3: Early decel. Early deceleration shown to use 54% more fuel in TRACON compared to late Late deceleration fuel burn shows the least sensitivity to aircraft weight Deceleration Profile Late Medium Early TRACON Duration 12 mins 14 mins 15 mins TRACON Fuel Use 1030 lbs 1347 lbs 1588 lbs TRACON Fuel Use Relative to Lowest - +31% +54% DDA - 24

Design Considerations for RNAV DDAs For RNAV DDA procedures to be practically useful, need to minimally modify existing procedures and be flyable by different types Based on BADA 4 analysis Evaluated speed envelope for range of representative profiles and aircraft types Significant differences in deceleration profiles observed between types Need to account for aircraft deceleration capabilities in RNAV DDA design DDA - 25

Comparison of Fuel Saving with Different Speed Profiles for Different Types Modeled fuel burn on BOS ROBUC2 RNAV arrival with four different speed profiles Fuel Burn ROBUC to Touchdown (kg) 1200 1000 800 600 400 200 Empirical early ROBUC2 published B757 latest A/c type latest Based on BADA 4 analysis 23% Ref-1% -15% 19% Ref 0% -23% 53% Ref -8% -8% 35% Ref -24% -34% 0 A320 B737 B757 B777 Significant difference between B757 latest and a/c type latest Trade-off between procedure simplicity and fuel saving DDA - 26

Assessment of Current RNAV Arrival Speed Targets Compared speed profiles of range of current RNAV arrivals compliant with As published speed targets Latest speed profile flyable by B757 reference aircraft Indicated Airspeed (kts) Airport/RNAV proc/rwy/speed profile BOS/ROBUC/4R/as published BOS/ROBUC/4R/B757 latest BOS/JFUND/15R/as published BOS/JFUND/15R/B757 latest CLT/FILPZ/36L/as published CLT/FILPZ/36L/B757 latest CLT/FILPZ/18R/as published CLT/FILPZ/18R/B757 latest RNAV procedure speed targets Distance to Touchdown (NM) DDA - 27

Assessment of Current RNAV Arrival Speed Targets Area between lines indicates how close published procedure speed targets are to B757-optimized speed profile Area comparison methodology proposed as a procedure efficiency screening tool Decreaseing RNAV Fuel Efficiency Increasing RNAV Fuel Efficiency Area Between RNAV Speed Target Compliant & B757 Latest Speed Profile Curves (NM.kts) 2000 1800 1600 1400 1200 1000 800 600 400 200 0 DDA - 28

Conclusions DDA is a promising concept for reducing approach fuel & emissions Radar analysis and classification tree approach provides insight into airports with highest benefit potential and lowest barriers Noise analysis showed negligible effects of DDA on noise impacts on ground RNAV arrival procedures offer a promising implementation path Tools presented for identifying RNAV arrival procedure candidates for modified speed targets to gain DDA benefits DDA - 30

Recommended Next Steps Further promote DDA concept to relevant stakeholders ATC facilities, Airlines/flight crews, Procedure designers Use proposed methodologies and tools to undertake: Wider screening of existing procedures Identify candidates for re-design process Redesign, analyze and deploy appropriate modified procedures to realize DDA benefits in operational system Assess human factors implications of DDA on ATC and flight crews Explore how new automation (e.g., TSS) could be leveraged to promote DDA DDA - 31 TSS = Terminal Sequencing and Spacing

Acknowledgments & Disclaimer Many thanks for Chris Dorbian, Aniel Jardines, Stephen Merlin & James Hileman of the FAA Office of Environment & Energy for supporting this work. DISTRIBUTION STATEMENT A. Approved for public release: distribution unlimited. This material is based upon work supported by the Federal Aviation Administration under Air Force Contract No. FA8721-05-C-0002 and/or FA8702-15-D-0001. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the FAA. 2017 Massachusetts Institute of Technology. Delivered to the U.S. Government with Unlimited Rights, as defined in DFARS Part 252.227-7013 or 7014 (Feb 2014). Notwithstanding any copyright notice, U.S. Government rights in this work are defined by DFARS 252.227-7013 or DFARS 252.227-7014 as detailed above. Use of this work other than as specifically authorized by the U.S. Government may violate any copyrights that exist in this work. DDA - 32