Free Flight En Route Metrics Mike Bennett The CNA Corporation

advertisement
Free Flight En Route Metrics
Mike Bennett
The CNA Corporation
The Free Flight Metrics Team
•
FAA
–
•
CNA Corporation
–
•
Dale Peterman, Dave Bartlett
DI
–
•
Joe Post, Mike Bennett, James Bonn, Dan Howell, Ashish
Khatta, Dan Murphy, Tony Rubiera
JTA
–
•
Dave Knorr, Ed Meyer, Antoine Charles, Esther Hernandez, Ed
Jennings
Ed Freeman
Other Support
–
NEXTOR, Northrup/Grumman, MITRE/CAASD, RPI, Aerospace
January 29, 2004
NEXTOR Moving Metrics Workshop
What we do
•
Free Flight Tools
URET
– TMA
– CPDLC
– CDM
•
Estimate potential benefits
pool
•
Future benefits projection
–
–
–
•
January 29, 2004
Investment Analyses
OMB Exhibit 300
Post-implementation
measurement of impact
NEXTOR Moving Metrics Workshop
En Route Metrics
Tie projected benefits to observable metrics
Observed Modeled
Excess distance (compared to great circle)
¾ Primary metric for en route
Flight times
Wind-adjusted
Excess distance and flight time by phase of
flight
“Lines data”
Flight Plan Amendments
Distance savings from amendments
En Route Throughput
“Hoses data”
Delay
Ground, Airborne
January 29, 2004
NEXTOR Moving Metrics Workshop
What about Wind-Optimal?
•
Wind-optimal is the most efficient trajectory
Computationally intensive
– Availability of wind data
– Moving target
–
•
Are great circle routes a good proxy for windoptimal?
.
Compare:
Actual Route
Wind-Optimal
Great Circle
(Exclude within 50
miles of airports)
For all flights on
two sample days
January 29, 2004
Source: J. Bonn
NEXTOR Moving Metrics Workshop
Actual vs. Idealized Trajectories
Distance
12
3.0
2.5
2.0
1.5
1.0
Route
.5
0.0
Great Circle
20020211
DATE
20020410
Excess distance over G.C. (nmi)
Excess time over Wind-Optimal (min)
Time
10
8
6
4
Route
2
0
Wind Optimal
20020211
20020410
DATE
When considering improvements to actual routes,
In the mean, Great Circles are a good proxy for wind-optimal
Potential Benefits Pool: 370,000 nmi per day
Is all of that pool recoverable?
January 29, 2004
NEXTOR Moving Metrics Workshop
Benefits Pool with Conflicts
•
Use FACET to identify conflicts and provide
geometry and aircraft speeds
Potential
Conflict
For sample day,
0.47 conflicts / flight
•
Original
Flight Path
θ
Revised
Flight
Path
Numerically solve for minimum conflict cost
Buffer
Cost of Conflict
Pool Reduction
Adjusted Pool
5 nmi
1.4 nmi
6%
310K nmi/day ($700M/yr)
10 nmi
3.6 nmi
16%
345K nmi/day ($790M/yr)
Source: D. Howell, J. Bonn
January 29, 2004
NEXTOR Moving Metrics Workshop
Average excess distance per flight (nmi)
Excess distance and traffic load
A Framework to approach En Route Improvements
More Efficient Routes More Capacity or less
More Directs
(URET,PARR,D2)
(RNP, airspace redesign)
conflicts (RVSM, URET,
Data link)
10
Opportunity
Regime
Congestion
Route Structure
Regime
Regime
0
0
20
40
60
80
100
Percent of maximum center traffic
January 29, 2004
NEXTOR Moving Metrics Workshop
Distance Saved from Lateral Amendments
•
As URET is deployed, we track
Number of flight plan amendments
– Distance savings from lateral amendments
–
•
Periodically update benefits estimates
–
Free Flight Reports, OMB Exhibit 300
Distance Saved from Lateral Amendments*
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
Aug-02 Oct-02 Dec-02 Feb-03 Apr-03 Jun-03 Aug-03 Oct-03 Dec-03
Direct Amendments
Other Amendments
Distance Saved Per Day (nmi)
Number of Amendmets per Day
URET Amendments
20000
15000
10000
ZID, ZME
ZKC,ZOB,ZAU,ZDC
ZJX,ZFW,ZMP
5000
0
10
20
30
40
50
# of months since IDU
Source: D. Murphy
January 29, 2004
*weighted average
NEXTOR Moving Metrics Workshop
Excess distance vs. traffic load by center
Important to establish site-specific baselines
ZOA
- has higher traffic levels
- handles a higher proportion of arrivals and departures than ZAB
Average excess
distance per flight
(n. mi.)
20
More complex
route structure
15
ZOA
10
ZAB
Greater
susceptibility to
capacity-related
effects
5
0
0
20
40
60
80
100
Percent of maximum center traffic
January 29, 2004
NEXTOR Moving Metrics Workshop
Efficiency by Phase of Flight
εi
40 n.mi.
60 n.mi.
d2
d1
Dep to 40
100 n.mi.
ε2
ε1
Mid-point
Origin
Airport
100 n.mi.
60 n.mi.
40 n.mi.
Great Circle Route
Actual Flight Track
40 to Arr
di
100 to 40
Destination
Airport
16
14
12
10
8
6
4
2
0
200 to 100
–
Mid (per 100 mi)
–
Algorithm developed and coded at Free Flight
ATALAB generated archive for all flights since 1998
Subset available in ASPM
100 to 200
–
40 to 100
•
Break up flight into segments
Track excess distance, flight time, degrees turned
Excess Distance (n. mi)
•
Phase of Flight
January 29, 2004
NEXTOR Moving Metrics Workshop
En Route Throughput
•
•
Construct throughput lines (“hoses”) that capture
major traffic flows
Measure throughput over lines
–
•
Also track crossing time and position by flight
Algorithm developed by Free Flight and OEP
Coded at Free Flight
– ATALAB generated archive for all flights since 1998
–
7
1
19
20
11
3
14
8
12
2
4
9
18
13
5
6
10
17
16
15
January 29, 2004
NEXTOR Moving Metrics Workshop
En Route Throughput and Departure Delay
Look at impact of holiday flights in ZMA
En Route Throughput
Total flights over
lines
500
1
19
7
20
11
3
14
8
12
4
2
18
9
13
5
Dec. 26
400
Dec. 12
300
200
100
0
6
10
6
17
7
8
16
Airport Operations
1400
1200
1000
800
600
400
200
0
Dec. 12
Dec. 26
FLL
MCO
MIA
PBI
TPA
+17%
-2%
-5%
+38%
-5%
January 29, 2004
Average Departure Delay
(min)
Daily Airport Operations
15
9 10 11 12 13 14 15 16 17 18
Local Hour
Average Departure Delay
40
Dec. 12
30
Dec. 26
20
10
0
FLL
MCO
MIA
PBI
TPA
+304%
+180%
+72%
+254%
+167%
NEXTOR Moving Metrics Workshop
Need for Better En Route Models
•
En Route problems manifest themselves in
several ways
–
•
•
Excess distance, departure delay, MIT, Ground stops
Difficult to separate en route problems from
terminal effects
Current queuing models have shortcomings
–
Don’t deal well with all constraints
•
–
No modeling of airspace performance when
demand < capacity
•
–
TRACON capacity
No “Opportunity” regime
Trajectories are non-adaptive
•
•
January 29, 2004
Tactical (Local congestion, weather)
Strategic (TFM)
NEXTOR Moving Metrics Workshop
Here’s what we’d like to see a model do…
January 29, 2004
NEXTOR Moving Metrics Workshop
Modeling Airspace Performance
Avg. Excess Distance
Excess Distance vs. Sector Load
3
If sector load < capacity,
2.5
2
1.5
Adjust dwell time stochastically
1
0.5
0
0
20
40
60
80
100
120
140
Percent Load (1 Min Bins)
Modeled aircraft approaches
new sector
If sector load > capacity,
• Allow to enter and adjust dwell time, OR
• Delay
If delay is excessive, adjust trajectory—
HOW???
January 29, 2004
NEXTOR Moving Metrics Workshop
Modeling Airspace Performance
How to do route adjustment to avoid excessive delay?
Iterative:
Limited set of alternate
full trajectories
Dynamic:
Route around congested area
Also need ability to implement
TFM initiatives if multiple flights
are affected
January 29, 2004
NEXTOR Moving Metrics Workshop
Modeling Airspace Performance
What about terminal delay that can’t be routed around?
Need to deal with airport and TRACON
capacity
If delay is excessive, may need to implement
strategic solution
Hold
Hold on ground,
Then release on
same trajectory?
January 29, 2004
NEXTOR Moving Metrics Workshop
Summary
•
Free Flight uses several en route metrics
Projections of future benefits
– Assessment of deployed tools
–
•
Our approach
Need to understand magnitude of problem (size of pool)
– Tie projected benefits to observable metrics
– Establish site-specific metrics baselines
–
•
Need better en route models
January 29, 2004
NEXTOR Moving Metrics Workshop
January 29, 2004
NEXTOR Moving Metrics Workshop
En Route Throughput and Departure Delay
En Route Throughput
19
7
1
20
3
500
14
8
12
4
2
18
9
13
5
6
10
17
16
15
Total flights over
lines
11
Dec. 26
400
Dec. 12
300
200
100
0
6
7
8
Average Departure
Delay (min)
Number of hourly
departures
Airport Operations
200
150
100
Dec. 26
50
Dec. 12
0
6
7
8
9 10 11 12 13 14 15 16 17 18
Local Hour
Departure Delay
30
Dec. 26
25
Dec. 12
20
15
10
9 10 11 12 13 14 15 16 17 18
Local Hour
5
0
6
7
8
9 10 11 12 13 14 15 16 17 18
Local Hour
MIA, MCO, TPA
January 29, 2004
NEXTOR Moving Metrics Workshop
Impact of Sector Capacity
•
Use line data to look at excess distance for flights
encountering busy sectors
70
60
Encountering a single
busy sector seriously
affects excess distance
Sum of Line XD (nmi)
50
Total Flight Length
40
500 to 1000
30
over 1000
20
under 500
10
0 to 40
80 to 100
40 to 80
over 100
Maximum Sector Load (% of sector capacity)
January 29, 2004
NEXTOR Moving Metrics Workshop
Modeled Sector En Route Daily Delay
High Sectors
2002
Many en route sectors are currently
capacity constrained
January 29, 2004
2012
Capacity constraints in en route
airspace will become more of a
problem in the future
NEXTOR Moving Metrics Workshop
Implementing TRACON capacity
70%
max
60%
95th %ile
50%
Average of 95th
percentile:
29%
40%
30%
20%
10%
0%
ATL
BOS
BWI
CLE
CLT
CVG
DCA
DEN
DFW
DTW
EWR
FLL
IAD
IAH
JFK
LAS
LAX
LGA
MCO
MD
MIA
MSP
ORD
PDX
PHL
PHX
PIT
SAN
SEA
SFO
SLC
STL
TPA
Number of AC in Tracon
(% of hourly AAR)
Tracon Capacity, Major Airports March 7, 2002
Modeled delay with and without Tracon capacity
20000
Total Daily Enroute
Delay (minutes)
Total daily enroute delay
(minutes)
Modeled delay with and without Tracon Capacity
Unlimited
15000
29% of AAR
10000
5000
0
2002
January 29, 2004
2010
2020
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
Unlimited
29% of AAR
Baseline
5%
15%
20%
Increase in Enroute Sector Capacity
30%
NEXTOR Moving Metrics Workshop
Download