LMI
Airline Responses to NAS
Capacity Constraints
Peter Kostiuk
Logistics Management Institute
703-917-7427
pkostiuk@lmi.org
National Airspace System Resource Allocation: Economics
and Equity
March 19-20, 2002
LMI
Objectives
• Quantify the magnitude of the demandcapacity shortfall
• Assess the effectiveness and feasibility of
possible solutions
• Estimate the true industry economic losses
at stake if we fail to increase airport
capacity
LMI
Analysis Approach
• Compare baseline travel demand forecast to one
that directly includes airport capacity constraints
– Quantify the “performance gap” between the
constrained and unconstrained forecasts
• Focus on system performance on good weather
days at the top 64 airports
• Assess the impacts of alternative policies on delay,
throughput, costs, and fares
LMI
Problem: Future Demand Exceeds
Capacity
Average Flight Delay (minutes)
45.0
40.0
Do Nothing Scenario
35.0
30.0
25.0
20.0
15.0
10.0
5.0
0.0
2000
2005
2010
Average airport delay per flight at the top 64 airports. Estimates do not
include downstream delay effects.
LMI
Possible Responses
• Reduce or Reallocate Demand
–
–
–
–
–
–
Higher fares
Schedule smoothing
More direct (point-to-point) service
Night-time operations
New hub airports
Slot-limit airports (by lottery, mandate, etc.)
• Increase Capacity
– Build more runways
– Use larger aircraft
– Introduce new ATM technologies
LMI
Who Can Affect What
Airlines
– Higher fares
– Schedule smoothing
– More direct (point-topoint) service
– Night-time operations
– New hub airports
– Use larger aircraft
– Introduce new ATM
technologies
Airports / Gov’t.
NASA/FAA
– Develop and
– Schedule smoothing
implement new
– More direct (point-to-point)
ATM technologies
service
– Night-time operations
– New hub airports
– Slot-limit airports (by lottery,
mandate, etc.)
– Build more runways
– Use larger aircraft
– Introduce new ATM
technologies
LMI
Analysis Requirements
• Require a model of NAS operations that
estimates delay and throughput under
different capacity and demand scenarios
• Require an economic model of the airlines
– Airline cost model
– Air travel demand model to capture changes in
demand in response to fare changes
• Connect the two models
LMI
Analysis Overview
INPUTS
• Unconstrained Demand
• Airport Capacity
• Observed Airport Delay Tolerance
Level
SCENARIO
• Schedule Smoothing
• Nighttime Operations
• Point to Point Service
• Larger Aircraft
• New Hubs
• New Technology
OUTPUTS
• RPMs
• Fares
• Costs
• Aircraft Fleet
• Employment
LMINET Operations Model
CONSTRAI NED DEMAND
•
Baseline
•
Scenarios
Air Carrier
Investment Model
LMI
Modeling the National Airspace System
With the LMINET Operations Model
Current
Schedule
TAF
Traffic Distribution Algorithm
Policy Option
User Input
LMI Airport
Capacity Model
Future Schedule
LMI Airport
Delay Model
Schedule Revision
Algorithm
Flight Delay
LMI
Air Carrier Investment ModelIntegrating Demand With Airline Costs
Estimated
demand coefficients
Per
Percapita
capitaincome
income
Population
Population
Unemployment
Unemployment
Fare
FareYield
Yield
Demand assumptions
Per
Percapita
capitaincome
income
Population
Population
Unemployment
Unemployment
Fare
Fareyield
yield
Total
Totaloperating
operating
revenues
revenues
RPM
RPMtime
timeseries
series
Employment
Employment
Fare
Fareequation
equation
ASM
ASMtime
timeseries
series
Fuel
Fuelcosts
costs
Flight
Flightpersonnel
personnel
labor
laborcosts
costs
Technology assumptions
Productivity
Productivity
Unit
Unitcosts
costs
Year
of
Year ofintroduction
introduction
Initial
Initialpenetration
penetration
Terminal
Terminalpenetration
penetration
Maintenance
Maintenance
costs
costs
Flight
Flightequipment
equipment
capital
capitalcosts
costs
Ground
Groundproperty
property
and
andequipment
equipmentcosts
costs
Indirect
Indirectcosts
costs
Adjusted Operating
Profit Margin
Aircraft
Aircraftfleet
fleet
Total
Totaloperating
operating
costs
costs
LMI
Forecasts With Flight Delay Constraints
• Define limits on acceptable flight delays
(increases in schedule time)
• When an airport reaches that limit, no more flights
will be allowed during that hour
• Delay maximum will be set for each airport based
on current operations or a system-wide average
• Estimate system throughput under the different
policies
• Estimate change in fare yields to match the lower
throughput
LMI
Average Delay for Constrained and
Unconstrained Forecasts
Delay per Flight (minutes)
45.0
40.0
35.0
Unconstrained
Constrained
30.0
25.0
20.0
15.0
10.0
5.0
0.0
2000
2005
2010
LMI
Percentage Growth in RPMs
Congestion Reduces Growth From the
FAA Forecast
70.0%
60.0%
2005
50.0%
2010
40.0%
30.0%
20.0%
10.0%
0.0%
FAA
Forecast
Constrained Constrained
High Seats Low Seats
LMI
Operational Concepts under
Capacity Constraints
• Accommodate growth by increasing fares and
rationing demand in the face of scarce capacity
• Smooth out the schedules
• Establish new hub airports to mitigate congestion
at existing hubs
• Increase direct service to avoid congested hubs
• Increase nighttime operations
• Use larger aircraft
LMI
Schedule Smoothing
• Re-direct excessive demand to ‘less desirable’ time
(smooth out the peaks and valleys associated with bank
operations)
• Assumes airlines attempt to maintain their schedules as
much as possible on a per-airport basis
– Maintain current hub structure and fleets
• Flights were moved a maximum of one hour from their
originally scheduled time
• Effectiveness depends upon airport’s existing demand
pattern
LMI
Nighttime Operations
• We assume that airlines will only offer nighttime
flights that cover their direct operating costs
• There is disutility to travelers from flying at night
• Effectiveness depends upon
– No curfew or nighttime noise restrictions
– Routes feasibility
– Passenger willingness, price elasticity
LMI
Direct Service
• Redistributes demand spatially, not
temporally
• Effectiveness depends upon
– Market opportunity for point-to-point flights in
non-hub airports
LMI
Larger Aircraft
• Airlines phase in larger aircraft to compensate for
slot shortages
• Desirable from airports’ perspective (e.g., SFO)
• Not necessarily desirable from airlines’
perspective because freed slots will be used by
existing and emerging competitors
• TAF projections include small increase in average
seat size; this scenario increases growth 1%
beyond that
LMI
New Hubs
• Preserves current hub-and-spoke strategy
• Select candidate airports based on current
status and potential for additional transfer
traffic
• No additional infrastructure investments
assumed
LMI
Increase in RPMs Over Constrained
Forecast
RPM Increase (billions)
70
2005
60
2010
50
40
30
20
10
0
Direct
New Hub
Smooth
Night
Larger AC Combined
FAA assumptions for growth in seats per departure.
LMI
RPM Forecast With Schedule Changes
1,200
Unconstrained
RPMs (billions)
Constrained
1,000
Smooth
Night
800
600
400
200
0
2005
No growth in aircraft seats per departure
2010
LMI
Schedule Smoothing Effectiveness
60
50
40
Effective
At SLC
30
20
Capacity
10
Baseline
Smoothed
0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
120
100
Ineffective
At EWR
80
60
40
Capacity
Baseline
Smoothed
20
0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
LMI
Decrease in RPMs (billions)
Lost Industry Output
160
140
120
100
80
60
40
20
0
2005
2010
High Seats/Dep
Low Seats/Dep
LMI
Value of Lost RPMs
20
$ Billions
2005
15
2010
10
5
0
High Seats/Dep
Low Seats/Dep
Does not include the cost of decreased utilization
from increased schedule time.
LMI
Comments
• Benefits of the policies examined are limited
– Results are conservative since they do not include the
costs of the strategies
• Flight delays continue to increase under all of the
policies
– Rise to 10-11 minutes per flight in 2010
• Can any of these strategies be implemented?
– Passenger acceptance
– Airline operations impacts
LMI
Yield Increase (cents)
Congestion Impact on Fare Yields
2
1.5
High Seats/Dep
Constant Seats/Dep
1
0.5
0
2005
Compared to Unconstrained Forecast
2010
LMI
Additional Economic Impacts
• Airline operating costs will rise significantly
– But fares will increase even faster
• Airlines will not need to buy as many new aircraft
– By 2010, US airlines will require about 600
fewer aircraft
• Airlines will not need as many new employees
– 84,000 fewer workers in 2010
LMI
Conclusions
• Current capacity enhancement plans are inadequate
• Airline strategies make limited impact but have
significant issues and obstacles to implementation
• Airline strategies do not keep air transportation on
an active growth path
• Aggressive technology development required to
enable growth
LMI
Knowledge That We Need
• Estimates of how much of the capacity shortfall is
attributable to:
– Misallocated resources such as runway slots
– Systemic shortage of infrastructure
• Better understanding of what an air transportation
system that can accommodate 2X growth would
look like
• New look at how air transportation investments are
financed
– Who pays for what part?
– Balance among concrete, technology, aircraft operations