Integrated Analysis of Airportal Capacity and Environmental ...

Integrated Analysis of Airportal Capacity and
Environmental Constraints
Shahab Hasan, Principal Investigator
George Hart, Research Fellow
Rosa Oseguera-Lohr,
NASA Technical Monitor
Presented at the JPDO Airports Working Group NextGen Airport Capacity Workshop
February 17, 2009

Task Description
• Objective:
– Identify and rank the key factors limiting the achievement of the
capacity goals of NextGen
– Identify gaps in available tools for conducting system-level trade and
benefit studies
– Results will help prioritize NASA’s research to enable NextGen
• Outcomes:
1. Define a set of critical airports and the most significant limitations to
airport capacity at those airports
2. Define the relevant environmental constraints that limit Airportal
capacity growth
3. Catalog the available tool set for analysis and modeling, assess gaps
4. Estimate the point of diminishing returns for hub capacity increases
vs. shifting to metroplex operations
P A G E 2

Notional Mock-up of Key Results
.
.
.
.
.
.
.
.
.
P A G E 3

Overview of Subtasks and Schedule
1. Develop Set
of Scenarios
2. Develop Set of
Metrics
3. Develop List of
Critical Airports
4. Analyze Airportal Capacity
Constraints
Runway
Constraints
5. Analyze Airportal Environmental
Constraints
Noise
Constraints
Taxiway
Constraints
Gates
Constraints
Emissions
Constraints
7. Analyze Shift to
Metroplex
Operations
6. Catalog and Assess Tool Set
Time
Feb.
2008
Kickoff
July
2008
Report, Version 1
Jan.
2009
Report, Version 2
June
2009
Report, Version 3
Dec.
2009
Final Report
P A G E 4

Subtasks Completed
• Subtask 1: Develop Set of Scenarios
– 2007, 2015, and 2025 demand
– Capacity per JPDO NextGen deployment schedule
• Subtask 2: Develop Set of Metrics
– Primary metric for this study is “projected throughput”
– Recommended rich set of metrics for use by NASA Airportal project
• Subtask 3: Develop Set of Critical Airports
– Set of 310 airports which cover 98.6 percent of all air carrier
operations and 99.8 percent of air carrier enplanements
• Subtask 6: Catalog and Assess Tool Set
– Final comprehensive report and database tool
P A G E 5

Subtasks In Progress
• Subtask 4: Analyze Airportal Capacity Constraints
– Developed the runway capacities and have done initial constraint
analysis
– Completed initial gate constraint modeling and analysis
– Started the detailed taxiway modeling of a few airports
• Subtask 5: Analyze Airportal Environmental Constraints
– 2025 fleet forecast augmented to use FAA’s Continuous Lower
Energy, Emissions and Noise (CLEEN) and N+1 Subsonic Fixed
Wing projections with a 2016 insertion
• Subtask 7: Analyze Shift to Metroplex Operations
– Started to brainstorm technical approach but have not begun analysis
P A G E 6

Subtask 4.1: Analyze Airportal Capacity Constraints (Runways)
Runway Capacity Analysis at 310 Critical Airports
• We assume no change to the airport capacities at the smaller 200
airports
– Likely cost prohibitive to deploy NextGen technologies so widely
• For the 110 larger airports, their capacities can be increased by
– New runways
– NextGen technologies
• One primary airport runway configuration for each meteorological
operating condition
• Airport runway configurations based on analysis of FACT2 and FAA
configurations, airport diagrams, capacity data, procedure charts,
and knowledge from prior tasks
P A G E 7

Subtask 4.1: Analyze Airportal Capacity Constraints (Runways)
Projected Throughput Based on Runway Capacity
• Only the runway capacities are considered
– There is no other constraint assumed
• Capacities are allocated along the Pareto curve to give the best
balance between the arrival and departure capacities
• Extra demand is removed from the schedule (trimmed) according
to rules enforced by our model for demand/capacity (D/C) ratios
– The process is done for every epoch of 15 minutes
• The rules for D/C ratios
– Cannot exceed 1.2 for every 15-minute epoch
– Cannot exceed 0.9 for every rolling hour
– Same ratios as used for JPDO studies
– Allows for short-term spikes but also the need for recovery
– Targets a relatively high level of service (low delay)
P A G E 8

Subtask 4.1: Analyze Airportal Capacity Constraints (Runways)
Projected Throughput Results: Overview
• Significant flight reductions (unsatisfied demand) at the busiest
hub airports
• Flight reductions in later years grow at some major airports, as
demand increases at a rate faster than some of the NextGen
improvements
• However, flight reductions at many airports decrease or are
eliminated in future years, as capacity enhancement from new
runways and NextGen improvements outpace demand growth
(e.g., ORD in 2025)
P A G E 9

Subtask 4.1: Analyze Airportal Capacity Constraints (Runways)
Projected Throughput Results: Top 10 Airports
5000
The percentage = projected throughput / unconstrained throughput
4500
4000
3500
3000
2500
2000
1500
95%
84%
78%
100%
97%
100%
100%
100%
100%
97%
83%
70%
90%
67%
99%
100%
100%
100%
100%
99%
99%
98%
93%
87%
Trimmed Flights
Untrimmed Flights
90%
91%
74%
87%
100%
100%
1000
500
0
ATL - 2007
ATL - 2015
ATL - 2025
ORD - 2007
ORD - 2015
ORD - 2025
DFW - 2007
DFW - 2015
DFW - 2025
LAX - 2007
LAX - 2015
LAX - 2025
LAS - 2007
LAS - 2015
LAS - 2025
DEN - 2007
DEN - 2015
DEN - 2025
IAH - 2007
IAH - 2015
IAH - 2025
PHX - 2007
PHX - 2015
PHX - 2025
PHL - 2007
PHL - 2015
PHL - 2025
CLT - 2007
CLT - 2015
CLT - 2025
P A G E 10

Subtask 4.1: Analyze Airportal Capacity Constraints (Runways)
Projected Throughput Results: Comparison
The percentage = projected throughput / unconstrained throughput
180000
160000
Trimmed Flights
Untrimmed Flights
99%
140000
98%
120000
99%
95%
Operations
100000
80000
99%
97%
88%
60000
40000
98%
95%
20000
0
OEP35-2007
LMI110-2007
LMI310-2007
OEP35-2015
LMI110-2015
LMI310-2015
OEP35-2025
LMI110-2025
LMI310-2025
P A G E 11

Subtask 4.2: Analyze Airportal Capacity Constraints (Taxiways)
Approach to Developing Taxiway Capacity Model
• Divided the airport surface into
two regions: terminal area, and
outer taxiways
• Focused on the outer taxiways;
for the majority of the airports in
this study, congestion in the
terminal area is not a concern
• Flag airports with characteristics
that typically lead to terminal area
congestion (narrow alleyways,
aircraft pushback into taxiways,
etc.) and adjust expected taxi
delays appropriately
P A G E 12

Subtask 4.2: Analyze Airportal Capacity Constraints (Taxiways)
Approach to Developing Taxiway Capacity Model
• Used off the shelf simulation software (Rockwell’s Arena)
• Focused only on bottlenecks formed by active runway/taxiway
interaction
• Runways and runway crossings modeled as resources
• Taxiing aircraft can be preempted from runway crossings by
landing or departing (high priority) aircraft
• Delays take place as aircraft wait for runway crossings to
become available
• Delays can be calculated by comparing high demand taxi times
to those of very low demand
P A G E 13

Subtask 4.2: Analyze Airportal Capacity Constraints (Taxiways)
Taxiway Capacity Model Example: PHX
P A G E 14

Subtask 4.2: Analyze Airportal Capacity Constraints (Taxiways)
Taxiway Capacity Model: Progress to date
• Taxiway capacity models have been built for:
– SAN (1 active runway without active runway crossings)
– LGA (2 non-parallel active runways with active runway crossings)
– PHX (2 parallel active runways with active runway crossings)
– SEA (3 parallel active runways with active runway crossings)
• Analysis for these models has not yet been conducted
• Next up: ATL, DFW, ORD
P A G E 15

Subtask 4.3: Analyze Airportal Capacity Constraints (Gates)
Determining Airports’ Gate Capacity
• An airport’s gate capacity is determined by many
complex interrelated factors
• No centralized source for data on gate capacity
– We collected gate count data for the top 310 airports from a
wide variety of sources
– No data on which gates can serve which types of aircraft
• The data that are available are unreliable and follow
many different definitions for a “Gate”
– We focused on gates with passenger bridges
• Adding additional space for ground-loading aircraft is relatively easy
compared with adding additional passenger bridge-equipped gates
• Airlines increasingly prefer to use passenger bridges
P A G E 16

Subtask 4.3: Analyze Airportal Capacity Constraints (Gates)
Gate Capacity: Important Assumptions
• All airlines have equal access to all gates, regardless
of current gate-access leasing agreements
• Due to the shortage of airport-specific information, we
have also assumed that all gates can service all
aircraft types
– This ignores aircraft spacing requirements
– Unfortunately there is no information available on what types
of aircraft a given gate at a given airport can accommodate
– We developed a supplementary tool to track growth in the
maximum required terminal frontage for each airport (more
on this later)
P A G E 17

Subtask 4.3: Analyze Airportal Capacity Constraints (Gates)
Developing the Gate Capacity Model
• Basic steps of the model
1. Use FAA-provided schedules to estimate the number of aircraft on
the ground at the beginning of the day for each airport
2. Use this reference point and the same schedules to track the
number of aircraft at the gates throughout a 24 hour period
3. When gate capacity is exceeded, trim arrivals (and departures)
from the schedule
4. Adjust the estimate for the number of aircraft on the ground at the
beginning of the day to reflect schedule trimming
5. Repeat steps 3 and 4 until the reduction in operations is
proportional to the reduction in the number of aircraft on the ground
at the beginning of the day
P A G E 18

Subtask 4.3: Analyze Airportal Capacity Constraints (Gates)
Model Execution: Terminal Frontage Requirements
• The model also calculates and compares current and future
terminal frontage requirements to help identify aircraft spacing
problems
– We categorized aircraft according to the FAA's gate-group
categorization system
– Space requirements were defined as each category's maximum
wingspan plus the FAA's suggested wingtip-to-wingtip gate spacing
– We cycle through the arrival and departure schedules to calculate
the change in the terminal frontage requirement throughout the day
– We can get an estimate of each airport’s growth in this requirement
by comparing the 2007, 2015, and 2025 maxima
P A G E 19

Subtask 4.3: Analyze Airportal Capacity Constraints (Gates)
Model Execution Example: JFK Int’l Airport (JFK)
Reference Point
200
180
Aircraft at the gates throughout a 24 hour period (JFK - 2015)
160
Adjusted
Reference Point
Aircraft at Gates
140
120
100
80
Gate Capacity
Trimmed
Untrimmed
Gate Limit
60
40
Net Departures
Net Arrivals
20
0
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 90 93
Epoch
Note: Epoch 0 corresponds to 9:00 AM GMT
P A G E 20

Subtask 4.3: Analyze Airportal Capacity Constraints (Gates)
Preliminary Results for a Sample of Large Airports
1800
1600
92%
80%
Arrivals Trimmed
Arrivals Accepted
1400
100%
1200
100%
95%
69%
100%
1000
90%
78%
77%
800
600
400
200
89%
96%
95% 100%
100%
99%
100%
100%
92% 85%
91%
100%
100%
100%
100%
100%
100%
82%
86%
98%
0
ATL-2007
ATL-2015
ATL-2025
BOS-2007
BOS-2015
BOS-2025
EWR-2007
EWR-2015
EWR-2025
IAD-2007
IAD-2015
IAD-2025
JFK-2007
JFK-2015
JFK-2025
LAX-2007
LAX-2015
LAX-2025
MSP-2007
MSP-2015
MSP-2025
ORD-2007
ORD-2015
ORD-2025
PHX-2007
PHX-2015
PHX-2025
SAN-2007
SAN-2015
SAN-2025
Airport-Year
P A G E 21
Arrivals

Remaining Work
• We are on schedule
• Subtask 4: Analyze Airportal Capacity Constraints
– Complete the constraint analysis under the hypothetical scenarios
when the ROT or MIT are not binding, respectively
– Update gate constraint analyses
– Continue the detailed taxiway modeling of airports
• Subtask 5: Analyze Airportal Environmental Constraints
– Airport modeling and environmental constraint modeling
• Subtask 7: Analyze Shift to Metroplex Operations
– Outline the approach and start the modeling
• Results synthesis
– Compile and identify the primary and secondary airportal constraints
P A G E 22

Backup Material
P A G E 23

Subtask 4:
Analyze Airportal Capacity Constraints
• Along with the analysis of environmental constraints, this is
(half of) the heart of the work
• We will rely on the approach we have developed and refined over
several years of NASA and JPDO work which estimates a metric
called “projected throughput”
• Use of this metric allows for an analytical “common currency” to
evaluate the Airportal constraints
– Runways
– Taxiways
– Gates
– (as well as the environmental constraints)
P A G E 24

Motivation for Using Projected Throughput as
a Metric (“schedule trimming”)
• “Unconstrained demand” (e.g., the FAA’s Terminal Area Forecast)
represents the public’s desire for air transportation regardless of
whether sufficient future NAS capacity will exist
• Without sufficient capacity, future flight schedules would incur
unrealistically large delays if all the flights in the unconstrained
schedule were actually flown
– Traditional valuation methods based on delay reduction may not be
appropriate for far-term states of the system
– While it’s easy to mathematically compute the DOC and passenger value
of time benefits for a hypothetical case of reducing delay from 500
minutes to 300 minutes, what is the interpretation?
P A G E 25

Motivation for Using Projected Throughput as
a Metric (“schedule trimming”)
• At some point, capacity constraints would lead schedule-conscious
airlines to cease adding additional flights to the schedule, even though
the demand would exist; to represent this, we analytically remove
flights from future flight schedules according to specified airport delay
tolerances or demand/capacity ratios and sector capacities
– This can also reflect service provider policies such as slot controls
• We call this consolidated capacity metric “projected throughput”
which estimates the number of flights that could be scheduled and
flown for a given level of system capacity
• Airline coping strategies such as using larger aircraft, shifting flights to
other times of the day or other airports, etc., are all possible but all
have their own ramifications and costs
• Projected throughput thus measures the effect of insufficient capacity
and estimates the benefit of additional system capacity
P A G E 26

Process for Going from Unconstrained Demand to
Projected Throughput
• To represent this behavior, we analytically remove flights from the
future flight schedule according to specified airport demand/capcity
ratios and sector capacities
Trim the flight with
the worst score;
Update score for the
rest
Airport Capacities;
Sector Capacities
No
Unconstrained
Demand Schedule
Demand and Capacity
Compared;
Delays Calculated
D/C Ratio Okay?
Sectors Below MAP?
Yes
Objective:
Trim the flights to
satisfy the D/C rules
and ATC sector
capacities while
maintaining maximum
system throughput
Projected Throughput
Schedule
P A G E 27

Subtask 4.2: Analyze Airportal Capacity Constraints (Taxiways)
Airport Taxiway Constraints Analysis
• Approach 1: Use queuing theory to construct a taxi model:
– Most previous studies have focused on runways-not applicable here
– We analyzed average taxi-in time as a function of demand for 70
airports:
• Generally linear functions, and often with near-zero slope
• Taxi time is a flat function of demand if the arrivals do not cross an active
runway active runway/taxiway interaction is the main driver of extra taxi
time
– The linear relationship of the empirical functions cannot be explained
by the standard queuing models:
• Model suggests exponential growth while observed behavior is linear
• Taxi time cannot be successfully modeled with queuing theory
P A G E 28

Subtask 4.2: Analyze Airportal Capacity Constraints (Taxiways)
Approach 2: Develop a physics-based model
• Approach 2: Build physics-based model of taxiway operations
at airports:
– Use simulation to determine taxi-time as a function of demand
– Model a small subset of airports from the OEP35 that together
represent the majority of common airport layouts
– Identify problematic airport layout/demand level combinations
– Extend findings to similar airports
• Challenges:
– Limited data available, especially pathway data
– Taxiway interaction with other airport elements (gates, apron areas,
service equipment) complicates analysis
– Building accurate physics-based simulations of taxiway
operations is highly labor intensive and time consuming
P A G E 29

Subtask 4.3: Analyze Airportal Capacity Constraints (Gates)
Preliminary Results: Overview
Percentage of Airports that are Gate Constrained
60%
50%
2007
2015
2025
Percentage
40%
30%
20%
10%
0%
1 2 3 4 All
Priority Group
• More priority 1 airports are gate-constrained than any other priority group
• The percentage of airports that are gate-constrained stays relatively stable from
2007 to 2025 in low priority airports
• There is a dramatic increase in the percentage of gate-constrained airports in
priority 1, from 21% in 2007 to 25% in 2015, to 53% in 2025
P A G E 30

Subtask 4.3: Analyze Airportal Capacity Constraints (Gates)
Preliminary Results: Overview
100%
90%
80%
70%
Acceptance Rate for Gate Constrained Airports
2007
2015
2025
Percentage
60%
50%
40%
30%
20%
10%
0%
1 2 3 4 All
Priority Group
• Most priority 1 airports have acceptance rates in the 90-100% range
– This corresponds to some small amount of gate-related delays at peak operating periods
• The average acceptance rate for gate-constrained priority 1 airports steadily
declines over time
• Demand will surpass capacity at a greater number of priority 1 airports by a greater
amount in the future
P A G E 31