Benefits Analysis of a Runway Balancing Decision-Support Tool Adan Vela 27 October 2015 Sponsor: Mike Huffman, FAA Terminal Flight Data Manager (TFDM) Distribution Statement A. Approved for public release; distribution is unlimited. This work is sponsored by the Federal Aviation Administration under Air Force Contract #FA8721-05-C-0002. Opinions, interpretations, recommendations and conclusions are those of the author and are not necessarily endorsed by the United States Government.
Terminal Flight Data Manager (TFDM) Overcoming Airport Inefficiencies Surface Efficiency Challenges Potential TFDM Solutions Inflexible manual data Controller information overload Surface delays Unnecessary fuel burn Increasing emissions Schedule uncertainty Under-utilized capacity Electronic flight strips Consolidated displays queue metering (virtual queues) Runway balancing Aircraft sequence optimization Airport configuration optimization TFDM designed to deliver NextGen decision support capabilities for surface & terminal operations Runway Balancing - 2
Fixes Dallas/Fort Worth International Airport DFW Procedures FERRA SLOTT NOBLY TRISS CEOLA PODDE DFW SOLDO CLARE TRACON Radius: ~35NM Runway Balancing - 3
Unbalanced Operations Dallas/Fort Worth International Airport Arrivals 18L s 17R Arrivals 18L Time [minutes] Demand Demand 17R Time [minutes] Multiple independent departure runways can lead to demand imbalances Runway Balancing - 4
Runway : Current Operations Runway assignments primarily depend on aircraft departure fixes Air traffic control adjusts runway assignment rules in response to visual observation of the current demand 18L 17R NELYN JASPA ARDIA DARTZ Runway Balancing - 5
Runway Balancing: Proposed TFDM Capability System automates and displays departure-demand forecasts under multiple runway assignment rule-sets Tower control selects optimal assignment rule-set System automatically reassigns aircraft departure runways and estimates future demand and delays Split 30 Min Forecast Demand 18L 17R East/West 16 27 East/West+ARDIA 20 23 East+JASPA/West 12 31 16 27 NELYN JASPA 18L 17R ARDIA DARTZ Update runway assignment rules 20 23 18L 17R NELYN JASPA ARDIA DARTZ Runway Balancing - 6
TFDM Benefits Assessment Provide the FAA investment analysis process with benefit estimates for runway balancing over the 2015-2035 lifetime SFO SEA LAS LAX SAN SLC PHX DEN DFW MSP BOS LGA DTW EWR JFK ORD PHL MDW IAD BWI DCA CLT ATL Analysis Airports DEN, DFW, DTW, IAH, LAX, MCO, MIA, MSP, PHX, SLC Selection Process Airport commonly operates multiple departure runways, and the airport taxiways and runway configuration allow for independent runway queues. Full TFDM analysis airports Runway balancing analysis airports IAH MCO FLL MIA FY2010 FY2012 FY2014 FY2016 Sep 2010 Investment Analysis Readiness Decision Aug 2012 Approval of Revised Scope, Goals and Schedule Mar 2014 Initial Investment Decision Mar 2016 Final Investment Decision Runway Balancing - 7
Benefits Analysis Methodology DFW-2035 Airport Configuration Schedule DFW-2030 DFW-2025 DFW-2020 DFW-2015 Airport Simulation Model Operational Settings: Traditional Runway TFDM Runway Balancing Fully Flexible s Taxi-Time Throughput Airports DEN, DFW, DTW, IAH, LAX, MCO, MIA, MSP, PHX, SLC Schedules include: airline, gate assignment, push-back time, departure procedure, aircraft type, arrival time, arrival runway. 2015-2035 Lifecycle Benefit Runway Balancing - 8
Input: Aircraft Demand Schedule FAA System-Wide Analyses Capability Tool (SWAC): fast-time NAS simulator that forecasts airport demand and converts to realistic schedule Airport Simulation Model SWAC Generated Schedule at DFW Average Number of Daily s 1400 1200 1000 800 600 400 200 0 2015 2020 2025 2030 2035 Year Runway Balancing - 9 Demand At Analysis Airports LAX DFW DEN MIA IAH PHX LAS MSP DTW MCO SLC Flight Number Destination Aircraft Pushback Time AAL1635 PHX MD83 12:40 AAL212 MIA B763 12:40 AAL300 EWR MD83 12:40 NKS718 LGA A320 12:44 EGF2757 DSM E145 12:45 EGF3739 BPT E145 12:45 AAL1108 LGA B738 12:45 AAL122 SFO B763 12:45 N441MB FRG HA4T 12:48 AAL1633 MCO MD82 12:50 ASQ2501 BTR CRJ2 12:50 AAL1537 RDU MD83 12:55 UAL315 DEN A319 12:58 AAL1337 FLL B738 13:00 EGF3712 OKC E135 13:00 EGF2782 BHM E145 13:00
Airport Simulation Model Airport Simulation Model Airport Configuration Taxi-Time Schedule Pushback Taxi to Queue Delays Arrival Schedule Throughput Represents an agent-based discrete-event queueing model Runway Balancing - 10
Traditional Runway Radar data Airport Configuration Runway, Airline, Aircraft type Pushback Taxi to Queue Flight Plans Modeling Trajectory Clustering Data Fusion Random Forest Runway Simulation Airline Aircraft Type Procedure Runway Runway Runway Balancing - 11
TFDM Runway Balancing Pushback Taxi to Queue Procedure Runway Runway Runway queues Scheduled Demand -to-runway Mapping Select mapping that minimizes delay imbalance across runways 18L 17R Scheduled Arrival Demand Calculate Available Runway Dep. Capacities 30 min Runway Balancing - 12
Fully Flexible Runway Pushback Taxi to Queue Gate Runway Runway Runway queues Scheduled Demand Calculate Earliest Expected Take-off Time for each Runway Select assignment with the earliest expected take-off time 18L 17R Scheduled Arrival Demand Calculate Available Runway Dep. Capacities Every Aircraft Fully flexible runway assignments (greedy by design) are not operationally feasible due to crossing departure procedures Runway Balancing - 13
Taxi-Time Pushback Taxi to Queue Airport surface radar data Airline, Gate, Runway, Taxi-time Gate E17 Taxi-Time Look-Up Table Modeling Historical Data Taxi-Time [minutes] 25 20 15 10 17R 18L 0 50 100 Percentile Simulation Gate Runway Taxi-Time Taxi-Time Look-Up Table Runway Balancing - 14
Runway Service Queue Modeling Flight Plans P(t<T) Select knee Inter-departure Service Time Pushback Taxi to Queue Wheels-off time Runway Aircraft type Simulation Prior Current Generate tables Spacing times for aircraft on different departure routes [seconds] (L/F = Leader/Follower) L/F Small Large 757 Heavy Small 60 60 60 60 Large 60 60 60 60 757 105 105 90 90 Heavy 120 115 115 95 Spacing times for aircraft on the same departure route [seconds] (L/F = Leader/Follower) L/F Small Large 757 Heavy Small 60 60 60 60 Large 90 90 90 90 757 105 105 90 90 Heavy 120 115 115 95 Wait-time Similar tables are generated for mixed operation runways Runway Balancing - 15
Dallas/Fort Worth International Airport 0.75 miles Current-Day Baseline Mapping 1 16 15 14 13 2 3 4 18L 17R 5 6 7 8 12 11 10 9 Cumulative Delay Reduction Benefits [2015-2035] 9,200 Hours Operations are well-balanced; small differences in runway taxi distances ensure runway balancing remains effective 18L 17R Relative Rate Traditional 47% 53% 1.0 Runway Balancing 48% 52% 1.0 Fully Flexible 46% 54% 1.0 Runway Balancing - 16
DFW Runway Balancing Benefits Avg. Surface Delay Per Aircraft [minutes] Avg. Daily Reduction in Surface Delays [Hours] 14 13 12 11 10 9 8 35 25 15 5 Traditional (Baseline) TFDM Runway Balancing Fully Flexible (Benefit Pool) 2015 2020 2025 2030 2035 Year Traditional (Baseline) TFDM Runway Balancing Fully Flexible (Benefit Pool) -5 2015 2020 2025 2030 2035 Year 20 seconds saving per aircraft. The delay-reduction benefits associated with runway balancing at DFW appears to be relatively small even at 2035 traffic levels Cumulative Benefits 2015-2035: 9,200 Hours Runway Balancing - 17
Orlando International Airport Current-Day Baseline Mapping 1 2 3 Strong airline preferences 1.5 miles Cumulative Delay Reduction Benefits [2015-2035] 0 Hours Low demand and asymmetric taxi times ensures traditional runway assignments remain the most effective operational paradigm 11 10 9 8 36R 35L 7 6 36R 35L 4 5 Relative Rate Traditional 48% 52% 1.0 Runway Balancing 52% 48% 1.0 Fully Flexible 51% 49% 1.0 Runway Balancing - 18
Phoenix Sky Harbor International Airport Arrivals Current-Day Baseline Mapping Arrivals Secondary s Majority s 9 1 2 10 26 25R 8 7 6 3 4 5 Cumulative Delay Reduction Benefits [2015-2035] 776,000 Hours The automated demand prediction capability required to perform runway balancing enables better utilization of the available departure capacity on 26. 25R 26 Relative Rate Traditional 96% 4% 1.0 Runway Balancing 71% 29% 1.2 Fully Flexible 77% 23% 1.2 Runway Balancing - 19
Runway Balancing Benefits Results Cumulative Delay Savings* (Hours) [2015-2035] Relative Rate [2035] MIA 1,028,000 1.2 PHX 776,000 1.2 DEN 212,000 1.1 MSP 176,000 1.1 DTW 27,000 1 SLC 25,000 1 DFW 9,200 1 LAX 0 1 IAH 0 1 MCO 0 1 The airports that benefit the most are those with mixedoperation runways (PHX, MIA, MSP, SLC) and/or underutilized capacity (DEN) Cumulative Delay- Reduction Benefits (Hours) [2015-2035] 1200000 1000000 800000 600000 400000 200000 0 Dedicated Runways Mixed Operation Runways PHX MSP SLC MIA DEN 0 500 1000 1500 Avg. Daily Demand Total Time Savings [2015-2035]: 2,253,200 Hours Total Monetized Fual Savings (Non-discounted): $1,500,000,000 Runway Balancing - 20 *Benefit estimates are sensitive to the taxi-time model
Summary Runway balancing is a potential TFDM tool envisioned to reduce airport surface delays Lincoln Laboratory developed a simulation model to quantify the delay-reduction benefits directly associated with runway balancing Identified that airports with mixed-use runways will benefit significantly from runway balancing The results of the runway balancing benefit analysis will be used in support of an FAA final investment decision in 2016 Runway Balancing - 21