george mason university transportation lab air transportation system limitations, constraints and...
TRANSCRIPT
George Mason UniversityTransportation Lab
Air Transportation System Limitations, Constraints and Trends
George L. DonohueMarch 19-20, Wye Woods Conference Center© George Donohue 2002
2 George Mason University
Demand has grown Faster than National Infrastructure
Relative Growth in Transportation Modes
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
1960 1965 1970 1975 1980 1985 1990 1994
Year
Mile
s /
Mile
s in
19
60
AirCarrierBus
Truck
Car
Source: DOT Statistics
3 George Mason University
Initial Observationsand an Hypothesis
FACTS: Airspace above Airport Runway Thresholds (Operational
Capacity) is a Limited, Nationally Allocateable Commodity National Airport and Airspace Management Infrastructure
growth has seriously lagged behind Growth in Air Transportation Demand
Utilization of this Capacity Commodity is Constrained by Airline Schedule Conflicts, Delay Tolerance, FAA Ground Delay Programs and Aircraft Safety (i.e. Aircraft Spacing)
HYPOTHESIS: A DoT Supervised Auction System may be Required to
Efficiently allocate Airport Capacity within Delay and Safety constraints
4 George Mason University
Incentives for Operational Improvements and Modernization
Key Decision Points DP 1 NATCA Contract Negotiations and
Controller Mass Retirement Threat (Avg. Age=50 + Service=25) ~2007
DP 2 Termination of Slot Controls - 2007 DP 3 Sector Congestion and limits of Radio
Frequency Spectrum Availability ~ 2010 Transition Barriers
- Ground Based Infrastructure L----M----H- Airborne Equipment L----M----H - Labor Issues L----M----H- Regulation L----M----H- Required Culture Change L----M----H- Communication Bandwidth L----M----H- LACK OF INCENTIVES TO CHANGE !!!!
5 George Mason University
Outline
Limitations on Air Transportation Capacity
Safety, Capacity and Delay System Network EffectsFuture Security EffectsObservationsFuture Vision
6 George Mason University
Operational Capacity is a Limited Commodity
CMAX = 2 C AR MAX S i (XG)i Ri {Airports}
–K AK(t) {Airspace Management Intervention}
S = f ( Safety, ATC , Wake Vortex, etc.) ~ 0.6
AK(t) = (A/CREQUEST – A/CACCEPT) ~ [ 0 to >1,000] AK(t) = f ( GDP:Weather, Sector Workload Constraints )
C AR MAX ~ 64 Arrivals/Hour (set by Runway Occupancy Time)
Ri = Number of Runways at ith Airport
XGi = Airport Configuration Factor at ith Airport i = 1 to N, where N is approximately 60 Airports K = 1 to M, where M is typically much less than 100 Sectors
7 George Mason University
Regional Distribution of Airport Infrastructure is Uneven
TABLE 1 Regional Air Transportation Capacity Fraction For (57) Major AirportsNUMBER Estimated % Avg 8 yr TAF
HUB # A/C TURN Number Ops/Hr Cap97/ Growth 1997 ENP OPERATIONS
REGION R/W POINTS MODEL 1997 CapMAX Rate % X10E6 2012 1997
NORTH EAST 14 420 348 294 84 9 54 1,950,000 1,645,786
PACIFIC SOUTHWEST 9 262 403 298 74 10 43 2,205,000 1,670,280
PACIFIC NORTHWEST 22 353 693 455 66 8 62 3,364,000 2,549,603
NOTHERN MIDWEST 42 773 1090 684 63 32 99 5,522,000 4,040,088
ATLANTIC COAST 13 269 438 241 55 8 31 1,701,000 1,347,458
CENTRAL MIDWEST 12 205 237 131 55 3 19 1,496,000 1,114,207
WEST 22 415 758 405 53 9 62 3,180,000 2,270,307
SOUTHEAST 21 424 776 391 50 -2 54 2,704,000 2,190,557
FLORIDA & LATIN AM 14 322 602 287 48 18 48 2,114,000 1,608,673
SOUTH SOUTHWEST 27 380 892 433 48 16 59 3,468,000 2,424,105
TOTAL 196 3823 6239 3620 58 11 532 27,704,000 20,861,064
% NATIONAL TOTAL 89 78 77
Donohue and Shaver, TRB 2000
8 George Mason University
Airport Diseconomies of Scale
Airport Runway Diminishing Returns
0.00
0.20
0.40
0.60
0.80
1.00
1.20
0 2 4 6 8
RUNWAYS
XG
for
Typ
ical
Air
port
s
XG FactorPower (XG Factor)
9 George Mason University
Non-Linear Network Characteristics
NAS is a Highly Non-Linear, Adaptive System Controller-in-the-Loop AOC-in-the-Loop Independent Network
Schedules Stochastic In Nature May exhibit Chaotic
Behavior under Some Conditions
Additive Improvements DO NOT result in Additive Increases in NAS Capacity ie. pFAST, Runways, etc.
EXAMPLE OF AIR TRANSPORTATION SYSTEM NON-LINEARITY
0
50
100
150
200
250
300
350
400
450
500
0 100 200 300 400 500
TOTAL AIRPORT CAPACITY (OPS/HR)
SYST
EM C
APAC
ITY
(OPS
/HR)
LINEAR SYSTEM
NON-LINEAR SYSTEM
3 Airport Network at 100 operations each + 50 operations increase
10 George Mason University
Airline Schedule has a Strong Effect on Network Performance – Model Prediction to 20% Airport Capacity Increase
DPAT Simulation, benchmark capacity, airports ranked by delay extent, with sector
0
0.5
1
1.5
2
2.5
3
Ori
gin
alL
GA
EW
RP
HL
BO
SM
CO
JFK
PH
XO
RD
SL
CT
PA
IAD
MIA
DE
NL
AS
DC
AL
AX
DT
WA
TL
MS
PC
LT
CV
GS
FO
SE
AS
TL
DF
WIA
HP
ITS
AN
BW
IM
EM
original1 2 3 4 5 6 7 8 9 1011 12131415161718192021222324252627282930
(min
)
Average Arrival Delay( queue on runway)
Linear System Response
MITRE DPAT MODEL
11 George Mason University
Outline
Limitations on Air Transportation Capacity
Safety, Capacity and Delay System Network EffectsFuture Security EffectsObservationsFuture Vision
12 George Mason University
Capacity vs. Delay Penalty
5
10
15
20
25
Arrival_Rate5
10
15
20
25
Departure_Rate
0
200
400
600
Delay
5
10
15
20
25
Arrival_Rate
[3] “ACE 1999 Plan,” CD-ROM. Federal Aviation Administration – Office of system capacity.
IAD
Arrivals
Depart
Delay
13 George Mason University
NY LaGuardia: A non-Hub Maximum Capacity Airport
1 Arrival Runway 1 Departure Runway 45 Arrivals/Hr (Max) 80 Seconds Between Arrivals 11.3 minute Average Delay 77 Delays/1000 Operations 40 min./Delay
TOTAL SCHEDULED OPERATIONS AND CURRENT OPTIMUM RATE BOUNDARIES
0
4
8
12
16
20
24
28
32
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Schedule Facility Est. Model Est.
14 George Mason University
New York LaGuardia Airport Arrival- Departure Spacing VMC
0
10
20
30
40
50
60
0 10 20 30 40 50 60
Departures per Hour
Arr
iva
ls p
er
Ho
ur
ASPM - Apr 2000 - Visual Approaches
ASPM - Oct 2000 - Visual Approaches
Calculated VMC Capacity
Optimum Rate (LGA)
40,40
Each dot represents one hour of actual traffic
during April or October 2000
45 Arr./Hr/RW@ 80 sec separation
DoT/FAA
15 George Mason University
Atlanta: A Maximum Capacity Fortress Hub Airport
2 Runways – Arrivals 2 Runways – Departures 50 Arrivals/Hr/RW – Max 72 Seconds Between Arrivals 8.5 minutes Average Delay 36 Delays/1000 Operations 38 min./delay
TOTAL SCHEDULED OPERATIONS AND CURRENT OPTIMUM RATE BOUNDARIES
0
10
20
30
40
50
60
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Schedule Facility Est. Model Est.
16 George Mason University
Atlanta AirportArrival-Departure Spacing VMC
0
20
40
60
80
100
120
0 20 40 60 80 100 120
Departures per Hour
Arr
iva
ls p
er
Ho
ur
ASPM - April 2000 - Visual Approaches
Calculated VMC Capacity
Optimum Rate (ATL)100,100
Each dot represents one hour of actual traffic
during April 2000
50 Arr/Hr/Rw
@72 sec
Separation
DoT/FAA
17 George Mason University
Major US Airport Congestion
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
DEN
BWI
JFK
PIT
CLT
DFW
DTW
IAD
EWR
PHL
SFO
LGA
MSP
SEA
ORD
STL
ATL
LAXA
IRP
OR
T
DEMAND / CAPACITY RATIO
J. D. Welch and R.T. Lloyd, ATM 2001
Queuing Delays Grow Rapidly
18 George Mason University
Aircraft Arrival Rate:Distance-Time Relationship
0
10
20
30
40
50
60
70
80
0 1 2 3 4 5 6 7
DISTANCE ( NMi)
AR
RIV
AL
S /
RW
/ H
R 120 Knots
130 Knots
140 Knots
Spacing
(sec)
60
90
180
120
72
19 George Mason University
LGA Aircraft Inter-Arrival Time Distribution
LGA Arrival Seperation Histogram
-5
0
5
10
15
20
25
30
35
0 50 100 150 200 250 300 350
Aircraft Inter-Arrival Time (seconds)
Nu
mb
er o
f A
ircra
ft i
n 2
0 S
ec. B
ins
LGA VMC
96 Sec. WV Separation Expected Value
μ = 134 sec., σ = 73 sec.
70 sec
SAFETYCAPACITY
20 George Mason University
Possible Relationship Between Safety and Capacity: ATM Technology Effect
Hypothesis: SAFETY-CAPACITY SUBSTITUTION CURVES
-
20.00
40.00
60.00
80.00
100.00
120.00
140.00
0.00 2.00 4.00 6.00 8.00 10.00 12.00
MILLION DEPARTURES/HULL LOSS
U.S
. MIL
LIO
N D
EPA
RT
UR
ES/
YE
AR
15 HULL LOSS/YR5 HULL LOSS/YR2 HULL LOSS/YRS-C @ CURRENT TECH.S-C @ REDUCED SEP.
US
21 George Mason University
Outline
Limitations on Air Transportation Capacity
Safety, Capacity and Delay System Network EffectsFuture Security EffectsObservationsFuture Vision
22 George Mason University
The Semi-Regulated Market Does Not Act to Minimize Delay: LGA Air 21 Impact
Source: William DeCota, Port Authority of New York
LaGuardia Airport
0
2040
60
80
100120
140
160180
200
06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Time of Day
Historic Movements AIR-21 Induced Svc.
Maximum Hourly Operations Based on Current Airspace & ATC Design
23 George Mason University
Annual and Seasonal Delay Trends(Note Possible Effect of Air 21 on LGA & System)
OPSNET Total System Delays
0
10
20
30
40
50
60
Month
Th
ou
sa
nd
s o
f D
ela
ys
2000
1999
1998
1997
1996
1995
24 George Mason University
Outline
Limitations on Air Transportation Capacity
Safety, Capacity and Delay System Network EffectsFuture Security EffectsObservationsFuture Vision
25 George Mason University
Observations Approximately 10 of the Top US Hub Airports are
Operating close to Maximum Safe Capacity Demand / Capacity Ratio’s Greater than 0.7 lead to Very
Rapid Increase in Arrival and Departure Delays Higher Delays Lead to Loss of Schedule Integrity 25 New Runways Not a Solution
Airline Hub and Spoke Network System Produces a Highly Non-Linear, Connected System Weather, Security or Terminal Delays Propagate System
Wide Airline Schedules are part of the Problem & Solution
ATC Sector Controller Workloads and Weather also Produce Network Choke-Points that Produce Capacity Constraints
26 George Mason University
Observations (cont.) 100% EDS Baggage Screening will either Increase
Delays or Travel Block Times for Commercial Ops Current Regulations on Airlines and Airports do
not provide Incentives for either Safe or Efficient Operations Airlines are over-scheduling Major Airports ATC is spacing Aircraft at the limits of current
technology leading to growing safety concerns Airlines are moving to Smaller aircraft to increase
frequency of operations and profitability, leading to increased congestion and delays
Airlines are resisting modernizing their aircraft with the technology required to decrease spacing and increase capacityIncentives are to be last to equip
27 George Mason University
Outline
Limitations on Air Transportation Capacity
Safety, Capacity and Delay System Network EffectsFuture Security EffectsObservationsFuture Vision
28 George Mason University
Vision: Incentives for Operational Improvements and Modernization
Brief Summary of Vision:
Major Hub Airports will Allocate Slots by DoT Auctions:-Both Strategic, Near Term and Spot Auctions-Peak runway loading will be reduced to Government Established
Safety and Capacity optimized schedules-Aircraft Size will be driven by a combination of airline profits and maximum enplanement opportunities
Business travel will migrate to Travel on Demand via air-taxi or private aircraft ownership and operation
Increased En-route Traffic density will be accommodated by Aircraft Self Separation-Technology-Equipped Flight Corridors
Auctions will provide incentives for aircraft technology insertion and a government contract to provide enhanced benefits
George Donohue
29 George Mason University
Outline
Limitations on Air Transportation Capacity
Safety, Capacity and Delay System Network EffectsFuture Security EffectsObservationsFuture Vision
30 George Mason University
Key Airport System FlowsMIT Queuing Model
ArrivalArrivalPathsPaths
Dept.Dept.PathsPaths
RunwaysRunways TaxiwaysTaxiways RampRamp GatesGates
PaxPaxScreenScreen
Ckd BagCkd BagScreenScreen
Check-InCheck-InIDID
Drop-offDrop-offParkingParking
EntryEntryFixFix
DepartureDeparture FixFix
AirsideAirside GroundsideGroundside
ArrivalsArrivals
DeparturesDepartures
PassengersPassengers
Bags/CargoBags/Cargo
GndGndTransTrans
31 George Mason University
0
100
200
300
400
500
600
700
800
900
1000
6 7 8 9 10 11 12 13 14 15 16 17 18
Time of Day
Nu
mb
er
of
Ch
ec
k-I
n B
ag
s p
er
3-M
inu
te
Inc
rem
en
t Scheduled Demand, No Spread Demand Spread = 30 mins (flat)
Total Check-InBags = 56,516
Baggage: Actual and Spread Demand for 1998 DFW Case (RAND Study)
Dr. R. Shaver
32 George Mason University
Planned Time of Arrival According to Passenger Propensity to Accept Risk
0
15
30
45
60
75
90
105
120
0.1%1.0%10.0%100.0%
Probability that Bag Misses Plane
Pla
nn
ed
Tim
e o
f A
rriv
al
Be
fore
Sc
he
du
led
D
ep
art
ure
--In
clu
des
15 m
inu
te U
np
lan
ned
D
ela
y B
efo
re B
ag
ga
ge C
he
ck-I
n (
min
ute
s)
Machines Deployed = [17:24] Machines Deployed = [19:28] Machines Deployed = [20:29] Machines Deployed = [22:32] Machines Deployed = [24:34] Machines Deployed = [15:22]
“The Passenger’s View”
R. Shaver, RAND