Capacity and delays at Newark Airport,

theoretical and experimental approaches

Thomas Morisset

Objectives

• Review and apply essential notions studied in class:– Theoretical capacity– Actual capacity (based on maximum throughput)– Delay estimates based on queuing theory– Actual delays

• Compare and understand the differences between theory and reality.

Newark Airport

• 443,000 movements• 36 million passengers• 13th busiest airport in the United States• 22 minutes mean arrival delay, congested• EWR-JFK-LGA metroplex: – 1,276,000 movements– 110,000,000 passengers– busiest multi-airport system in the world

figures for 2007

Experimental approach to capacity [1]

26.7 mov per 15 min

Experimental approach to capacity [2]

24 mov per 15 min

Experimental approach to capacity [3]

22 mov per 15 min

27262524232221201918171615141312111098765432100.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

9.0%

VFRIFR

movements realized per 15-minute intervals

freq

uenc

y in

200

7 98%

per

centi

le

Experimental approach to capacity [4]

22 mov per 15 min

• Capacity varies depending on runway configuration, weather conditions:

Experimental approach to capacity [5]

Utilization 3rd runway % time % movts mov/15min (98 %ile)

With 11/29 36.6% 39.0% 23

Without 11/29 62.8% 60.4% 21

Weather condition % time % movts mov/15min (98 %ile)

VFR 83.4% 84.3% 22

IFR 16.6% 15.7% 21

• Run the MACAD* capacity simulation program• Treat the simple case of 2 close parallel

runways (4R|4L or 22L|22R)• Data needed:– aircraft mix: ASPM– IFR separation standards: FAA AC’s– VFR separation estimates**– approaching speeds, runway occupancy times

Theoretical approach to capacity

* Odoni, Stamatopoulos ** Trani

• Run the MACAD* capacity simulation program• Treat the simple case of 2 close parallel

runways (4R|4L or 22L|22R)• Data needed:– aircraft mix: computed from ASPM– IFR separation standards: FAA AC’s– VFR separation estimates**– approaching speeds, runway occupancy times

Theoretical approach to capacity

* Odoni, Stamatopoulos ** Trani

RESULTS:

- IFR theoretical capacity: 18 movements per 15 minutes

- VFR theoretical capacity: 20 movements per 15 minutes

Theoretical capacities are slightly lower than capacities yielded by theoretical approaches

Method / Sourcearrivals+departures in 15

minutes capacity

98th percentile 22

Pareto envelope (>100 obs) 22

Scheduled demand 23

FAA benchmark report 21-23

Macad Model (VFR) 20

All capacities

Experimental approach to delays [1]

7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 220

5

10

15

20

25

30

35

40

Arrival delay at EWR in 2007

hourly delayAverage

Hours of the day

min

utes

Experimental approach to delays [2]

-60 -40 -20 0 20 40 60 80 100 120 140 160 1800.00%

0.50%

1.00%

1.50%

2.00%

2.50%

3.00%

3.50%

4.00%

4.50%

5.00%

EWR arrival VFR 2007

arrival delays (1-minute intervals)

freq

uenc

y

Experimental approach to delays [3]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

-10.0

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

-10.0

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

arrival delay departure delay +taxi out delay schedule Actual

hours

num

ber o

f mov

emen

ts

min

utes

Experimental approach to delays [4]

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 230.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

100.0%

Runway configurations

Rest

29, 4R / 4L

11, 4R / 4L

4R / 4L

11, 22L / 22R

22L / 22R, 29

22L / 22R

Experimental approach to delays [5]

Performance of different runway configurations during peak hour (4pm – 8pm)

Runway configuration All 22L | 22R 11, 22L | 22R 4R | 4L 4R, 11 |

4L4R, 29 |

4L

Frequency 100% 30.5% 27.8% 14.4% 13.9% 8.0%

Arrival delay 31.2 43.9 14.5 48.0 18.4 25.4

Departure delay 27.1 35.1 16.8 37.5 20.0 23.4

+ Taxi out delay 20.8 17.7 20.6 19.9 28.2 23.2

• Run a queuing model program: DELAYS*• For different levels of capacity• For a particular day of demand (Feb 4 2007), NOT average demand

Theoretical approach to delays

* Odoni, Kivestu

Conclusions

• Capacity:– theoretical and experimental approach within 15% of each

other, a good match– can assist with the choice of runway configuration,

comparing airports, etc.• Delays:– theoretical estimates much lower than reality– delays cannot be explained by the limited capacity of 1

single airport– must consider the entire network of airport to include

delay propagation