Professional Documents
Culture Documents
Background
The Airbus A380 can carry the most passengers and is also the heaviest commercial passenger aircraft to have entered service. The entry into service of the aircraft required careful management by the airports, airlines and aviation authorities that would handle the aircraft in order to minimize the disruption caused to the air transportation system. Airports were required to upgrade runways, taxiways and gates to deal with both the size and passenger volume of the A380. The authorities responsible for aviation safety were particularly concerned about the danger posed by the wake of such a high gross weight aircraft and initially took a conservative approach to wake separation standards which could potentially have reduced the passenger capacity at airports served by the A380. This study will examine the tradeo between the additional passenger capacity of the A380 and the additional wake separation that the aircraft requires.
1.1
The Aircraft
The A380 entered into service in October 2006 with Singapore Airlines after an 18 month delay due to a variety of technical issues during development of the aircraft. At the time of writing (November 2009) only 20 A380s have entered service [1], with a production rate of approximately 1 aircraft a month for 2008 and 2009 [2] potentially increasing to 20 deliveries in 2010 [3]. This means that the impact of the A380 on the air transportation system to date has been gradual and sparsely distributed. However as more of these aircraft enter service in the coming years the eect of their operations will become more signicant at the major international hubs where they operate. For the purposes of this study the seating capacity of the A380 is an important variable. However there is considerable variability in this value even amongst the 20 aircraft delivered
5
" 6"
Alexander Donaldson
2.
1.
0.
&
/.
.
. (
#
"
$
!
%
) %
#$ & '
!" '
& *
+ ,
+ -
Figure 1: Current (November 2009) A380 Orders and Deliveries by Airline [1] to date from a minimum of 450 seats in the aircraft operated by Qantas up to 525 seats in the aircraft recently delivered to Air France. This variation in seating capacity makes a signicant dierence in the runway passenger throughput, therefore the capacity model will be run at these high and low bounds.
1.2
In addition to an unprecedented passenger capacity for a commercial airliner, the size and weight of the A380 brought with it the likelihood of an exceptionally powerful wake. During development of the aircraft and the ight test program the International Civil Aviation Organization (ICAO) recommended a very conservative separation criteria of 10 n.m. for all aircraft following the A380 (Table 1a) on approach, unless that aircraft was another A380 in which case there was no wake separation requirement (the A380 could follow any aircraft including another A380 without any wake separation requirement). Just before the 2 of 15
Alexander Donaldson
entry into service of the A380 the ICAO draft guidance was revised based on the results of an extensive wake vortex measurement campaign carried out by Airbus. The nal ICAO guidance was to simply add two nautical miles to the separation required behind a Heavy aircraft (Table 1b. This change in guidance had important implications for the throughput achieved by the A380 as will be demonstrated in this paper. The United Kingdom CAA largely mirrors the ICAO guidance with modications to t its own weight categorization scheme (Table 1c). It is interesting to note that the UK CAA found it necessary to revise upwards the ICAO separation guidelines with regard to the A380 as a following aircraft, based upon operational experience with the aircraft in busy UK terminal airspace environments [4]. The United States FAA is taking a more conservative approach to A380 operation requiring separations (Table 1d) greater than the ICAO recommendation particularly with regard to Medium and Light aircraft Following the A380. Table 1: Dierent Approach Separation Standards (R denotes Radar separation minimum applies - 2.5 n.m. for JFK and LHR)
(a) Initial ICAO Guidance (b) Nov. 2009 ICAO Guidance
A380 H M L
A380 R R R R
Following H M 10 10 4 5 R R R R
S 10 6 5 R
A380 H M L
Following A380 H M R 6 7 R 4 5 R R R R R R
S 8 6 5 R
Leading
Leading
A380 H UM LM S L
A380 4 4 R R R R
H 6 4 R R R R
Following UM LM 7 7 5 5 3 4 R R R R R R
S 7 6 4 3 3 R
L 8 7 6 5 4 R
A380 H B757 M L
A380 6 R R R R
Following H B757 6 8 4 5 4 4 R R R R
M 8 5 4 R R
L 10 6 5 4 R
1.3
Leading
Example Airports
The impact of the A380 on three major airports will be examined in this paper. These airports have been chosen because they have dierent modes of operation and mixes of aircraft sizes, yet they are all expected to receive a signicant number of A380 operations in the coming years. 3 of 15
Leading
Alexander Donaldson
1.3.1
London Heathrow
London Heathrow will be used as the baseline airport in this study for several reasons: Large number of expected A380 operations. Extensive operational data readily available. Arrivals and departures are always segregated. Operation at close to runway capacity throughout the day.
UK AIP
AD ELEV 83FT
LONDON HEATHROW
EGLL
VAR 2.1W - 2009
HEATHROW
IAA/ IRR
N
RVP North ILS GP
092M
MEHT 66
116 117
(34) (33)
103
(20)
PAPI (3)
3901m x 50m
AB13
MLS M-HER Ch 522
AB12 SNAPA
Link 57
Tw
AB11 A11
Taxiway A
Twy C Twy D
PAPI (3)
MEHT 73
27 R
A4 A3 AY4
Tw
yA
T
09 L
A12 A9W
Taxiway A Twy H
A1 A2
Link 22
A13
A9E
Link 11 Link 12
A8
Link 13
A7
Taxiway A
A6
Twy J
AY5
A5
PLUTO
Twy L
Link 23
yA
105
(22)
Link 58
RABIT
yB
23
COBRA
Tw
Link 56
Taxiway B
Taxiway B
TITAN
B
AY3
Link 21
126
(43)
Link 55
Twy G
Control Tower
362
(143)
226
Tw y
Radar
Terminal 1
r3 Pie
Under Construction
Lin
DINGO
C2
D2
E2
Pier 4A F2
xi r4 Ta Pie Under Construction
yJ wa
L2
wy
Maintenance Area 1
M
Taxiway D
xiw Ta H ay
Twy L
Pie
(138)
221
Taxiway E
r5
Taxiway F
Twy B
Terminal 5A
Taxiway C
Taxiway B
Taxiway A
Terminal 5B
ro
Link 54
Terminal 3
Ta
Ta x
Link 36
Eu
iw
ay
29
Lin
Taxiway B
Taxiway B Twy R
Twy A
Link 34
Link 33
Link 32
Under Construction
Ta x
D1
Twy D
E1
F1
iw
HANLI
Twy Y
DASSO
N11
100 105
(22) (17)
VIKAS C1 OSTER Link xiw Taxiway B Link 53 ay B 53 Link HORKA 51 wa Taxiway A Link 52 yA Disused N8 N10 PAPI (3) NB8 NB10
ay
Link 35
Pier 6
Twy A
Twy
101
(18)
SB3 S5 S4
Taxiway S
43
PAPI (3)
MEHT 64
272M 119
(36)
S3
ILS GP S1N
Twy S
125
(42)
Lin
Cargo Apron
27 L
Ta xiw
Taxi
Lin k
ay
Cargo Apron
way
GUND (Geoid Undulation) = The height of the Geoid (MSL) above the Reference Elipsoid (WGS 84) at the stated position. BEARINGS ARE MAGNETIC ELEVATIONS AND HEIGHTS ARE IN FEET
Car Park
ELEVATIONS IN FEET AMSL HEIGHTS IN FEET ABOVE AD
(279)
pie r
Ta xiw ay G
T5C
Pie
Ta xiw ay F
L1
P
NEVIS ETTIV
y Tw U
MORRA
r7
Link 27
Taxiway Y
Twy Y
Tw P y
Under Construction
Taxiway M
xiw ay R
Link 2
Link 28
Twy A
Link
28
Ta
Twy Q
Twy C
Taxi
N2E NB2E
N7
N6
N5W
N5E
N4W
N4E
N1 NB1
092M
MEHT 67
09 R
ay xiw
SB1
Rwy 27L Thr Elev 77 512753.83N 0002602.68W (GUND Elevation 151)
Taxiway Z
Royal Suite
Link 44
Link 41
W
42
S1S
COM ATIS TWR 128.075, 113.750, 115.100 (Arrival) 121.935 (Departure) 118.700, 118.500, 124.475 121.975 (GM Planning) 121.900, 121.700, 121.850 (GMC) 121.600 LIGHTING THR 09L 09R THR 27L 27R HI Green with HI W bars. HI Green with HI W bars. HI bi-d colour coded C/L. TDZ 900m. HI bi-d white edge (first 300m Red). End lights red. HI bi-d colour coded C/L. TDZ 901m. HI bi-d white edge. End lights red. HI bi-d colour coded C/L. TDZ 899m. HI bi-d white edge (first 300m Red). End lights red. HI bi-d colour coded C/L. TDZ 901m. HI bi-d white edge. End lights red. Green C/L and Red stop bars with selective switching on all taxiway routes. HEATHROW INFO HEATHROW TOWER HEATHROW DELIVERY HEATHROW GROUND HEATHROW FIRE
Terminal 4
Highest Elev in TDZ 78 512753.67N 0002651.99W (GUND Elevation 151)
RVP South
xiw Ta
ay
362 (279)
100 500 0 0 100 500 200 300 1000 400 500m
RUNWAY/TAXIWAY/APRON PHYSICAL CHARACTERISTICS APRON / RWY / TWY RWY 09L/27R RWY 09R/27L Aprons Taxiways SURFACE Grooved Asphalt Grooved Asphalt ConcreteAsphalt Concrete/Asphalt BEARING STRENGTH 83/F/A/W/T 83/F/A/W/T -
1500ft
I-BB & I-LL 109.50D (Ch 32X) IBB/ ILL 512749.56N 0002730.77W 92'
HEATHROW
CHANGE: AREAS UNDER CONSTRUCTION ADDED/AMENDED/REMOVED. TWY A EXTENDED (TO THE SOUTH). HOLDS N3/NB3/N2W & TWY N REMOVED. LINK 28/HOLD N2E/NB2E/NEW ACCESS TO RWY 27L/REPORTING POINT L2 ADDED. AERO INFO DATE 17 AUG 09
AMDT 11/09
Figure 2: London Heathrow Airport Layout[5] In November 2009 London Heathrow accommodated 4 daily Heathrow operations (2 ights to Singapore, one to Dubai and one to Sydney), this gure will rise substantially as Airbus delivers more aircraft across the world given that Heathrow is a major international hub. Heathrow is also likely be the base of operations for the 18 A380s to be operated by British Airways and Virgin Atlantic. In addition to the volume volume of future A380 operations Heathrow is a useful baseline airport given the wealth of data available about operations at the airport as well as the simple operational modes of its runways. Heathrow always operates one runway for arrivals and one runway for departures (to minimize the noise impact on communities under the approach path [6]). Heathrow also operates close to its runway capacity for most of the day due to slot controls at the airport. These factors 4 of 15
Alexander Donaldson
together mean that a simple runway capacity model described in Section 2.1 should yield an accurate estimate of arrival capacity at Heathrow. 1.3.2 New York JFK
An American airport was included in the study in order to examine the eects of the more stringent separation requirements impose by the FAA as well as the eect of the lower aircraft size seen on average in the U.S. New York JFK (JFK) is likely to be one of the top U.S airports in terms of A380 operations1 . JFK is also an interesting contrast to London Heathrow because its runways are frequently operated in a mixed-mode conguration, with arrivals and departures sharing the same runway.
NE-2, 17 DEC 2009 to 14 JAN 2010
D
(JFK)
ELEV 13
AUX FIRE
STATION
392 X 2 6
NE 1 7.7
GND CON
CLNC DEL
DEP 1 5.1
40^40N
1 9.1 281.5
123.9 281.5
SW 1 5.4
135.05 348.6
E
ELEV 12
2 4.1^
22R
KENNEDY TOWER
121.9 348.6
FB
FB
C
ZA
YA
ELEV 13
CH
C
H
YA
ZA
Y
CB
CD
READBACK OF AL RUNWAY
CAUTION: BE ALERT TO
NEW YORK /
UA
U
B
G
Y
ELEV 13
TA
A 197
1 351 X 150
CONTROL
AL-610 (FA )
SD
ARRIVAL TERMINAL
U.S CUSTOMS
ELEV 12
INTERNATIONAL
SC
31L
KA
A A A
LA
SB
314.1^ KC
SA
NC
NA
CAT 2
MA
STATION
A
P
FIRE
B
M
HOLD
73^49W
EMAS
Q
QB
H
P
14572 X 150
KB
04 .1^
Q
QD PB
QC
PA
405 X 2 7
Q
P
PD
K1
QG
QF
QH
PC
PE
0.0^E
4L
VAR 13.3 ^W
134.0^
JANUARY 20 5
AIRPORT DIAGRAM
ELEV 12
40^37N
73^46W
NB
FIELD ELEV
4R
04 .1^
13
TOWER
09351
AIRPORT DIAGRAM
40^38N
ELEV
12
73^47W
1.3.3
Dubai International
Dubai International Airport (DXB) was chosen as the third airport in this study due to the potentially unmatched future level of A380 operations asa result of Emirates Airlines
Los Angeles International (LAX) my handle more A380s however it was not used in this study due to the complications imposed on A380 operations by its closely spaced parallel runways and the uncertainty surrounding the resolution of these issues.
1
5 of 15
09351
SE
38
NEW YORK /
KD
CB
C
13L
VA
73^48W
134.1^
10 0 X 150
CA
WA
840 X 20
DA
EA
2 4.1^
CE
73^45W
FA
31R
314.1^
FB
2L
40^39N
EMAS
(JFK)
MB
TB
CB
13R
Alexander Donaldson
(based at DXB) large order for 58 of the type (Figure 1). The airport also adds a third distinct separation standard by applying the ICAO recommendations without modication. Like JFK, Dubai International has a pair of parallel runways assumed to be operating independently with mixed arrivals and departures for the purposes of this study.
I
CHANGES: Twyr.
INC., 1007. ALL RIGHTS RESERVED.
8 JEPPESEN SANDER%M,
Methodology
This study compares the passenger throughput and arrival capacity of the three study airports for the three cases of moving heavy operations to a 525 seat A380, a 450 seat A380 and a 418 seat 747-400. The A380 cases include the appropriate higher wake separation criteria required by that type of aircraft, while the 747-400 case reects a simple upgauging of the heavy category aircraft with no additional separation required. The second part of the study looks only at LHR and compares the eect of applying the four dierent described in Table 1 to the LHR operations to examine the behavior of these criteria for a common set of operations.
2.1
Simple queueing theory will be used to model runway capacity of the airports being studied, using the model described in de Neufville and Odoni [8]. The time separation (in seconds) 6 of 15
Alexander Donaldson
between a lead aircraft in weight category i and a following aircraft in category j can be found from Equation 1. Tij = max 3600 r + sij r vj vi sij vj , od2 + max(oi , od1 ) for vi > vj (1) Tij = max 3600 , od2 + max(oi , od1 ) for vi vj
The time required per arriving passenger (in seconds) can then be calculated for dierent sequences of arriving aircraft as shown in Equation 2. Tpax,ij = Tij ci (2)
Given Tij and Tpax,ij it is possible to calculate the airport arrival capacity using the matrix of likelihoods of any given pair of arrivals (pij ). Cops = Cpax = 3600(nr )
K i=1 K i=1 K j=1 (pij K j=1 (pij
Tij ) Tpax,ij )
(3) (4)
3600(nr )
2.2
Model Inputs
The required inputs for the runway capacity model were collected from a wide variety of data sources with reasonable assumptions being made where data was unavailable. The model inputs are summarized in Table 2 and where appropriate are further explained in this section. 2.2.1 Final Approach Path Length
The nal approach path length (r) is the distance over which air trac control can no longer separate aircraft based on speed since the aircraft are preparing to land. At London Heathrow this distance is a minimum of 4 n.m.[9] and at JFK it is 5n.m[10]. The approach path length for Dubai International could not be found and was therefore assumed to be the same as used at JFK so as not to introduce an unnecessary additional variable.
7 of 15
Alexander Donaldson
Table 2: Summary of Inputs to the runway capacity model LHR Approach Length (r) Buer time (bi ) Number of Runways(nr ) Separation (sij ) Approach Velocity (vi ) Aircraft Mix (pij ) Arrival Occupancy Time (oa ) Departure set-up time for (od1 ) Departures roll time (od2 ) Passenger Capacity (ci ) 2.2.2 4 n.m. UK CAA[9] JFK 5 n.m. FAA[10] 10 seconds (Assumed) [8] 1 segregated UK CAA (Table 1c) 2 mixed independent US FAA (Table 1d) ICAO (Table 1b) DBX 5 n.m. Assumed
Based only on Boeing aircraft in schedule 2008 Flight Timetable (4 week sample) 2008 ETMS Database (4 week sample) 2004 Annual Operations
Based on data from LHR Study Not Needed Not Needed 45 s (Assumed) [8] 60 s (Assumed) [8]
Approach Velocity
Aircraft approach velocities vary signicantly depending on the weight of the individual arriving aircraft, because of this it is a dicult process to estimate actual approach velocities. Boeing provides a document [11] with reference approach speeds which was used to estimate speeds for this study. The Boeing speeds were applied to all Boeing aircraft (by sub-type e.g. 737-800) in the operational data sets and then averaged over the weight categories in use at each of the study airports. Aircraft that were not manufactured by Boeing or McDonald Douglas were not included in the approach speed calculation. This method provides a reasonable estimate for the approach speeds given the large number of Boeing aircraft represented at the study airports and the wide range of weights of these aircraft.
8 of 15
Alexander Donaldson
2.2.3
Aircraft Mix
The probability (pi )of any given arrival being from a certain weight category is assumed to be equal to the proportion of all arrivals that are from that weight category. For LHR and JFK the proportion of aircraft from each weight category was estimated through examination of 4 weeks of arrivals information from 2008. The rst week in February, May, August and November were chosen to provide a mix of dierent travel seasons while avoiding the holiday period. For DXB only aggregate data from 2004 was available. Given the already high proportion of heavy jets in this data no additional scaling was performed to adjust the mix of aircraft to 2008 levels. Once the probability vectors were compiled the probability of observing a pair of arrivals i followed by j is simply obtained by multiplying together the two probabilities pi and pj . The measured values of pi are shown in Figure 5. It is important to note the proportion of heavy aircraft at each airport, with DXB having signicantly more heavy arrivals than JFK and LHR. A380 operations were simulated by moving a percentage of the heavy operations at each airport to A380 operations. The model was run for up to half of the heavy operation at each airport being converted to A380s.
H UM LM S L H B757 M L H M S
B757
UM
LM
(a) At LHR
(b) At JFK
(c) At DXB
2.2.4
A study conducted in 2005 at LHR [12] measured the runway occupancy times for 170 arrivals at London Heathrow for a wide range of dierent aircraft types and was conducted during good visibility for a dry runway. The results of this study were averaged across the weight categories appropriate for LHR, JFK and DXB. Given that runway occupancy
9 of 15
Alexander Donaldson
time measurements for JFK and DXB were not readily available this data provides realistic estimate of those times and was therefore used across all the airports in this study. The actual runway occupancy time may vary based on actual taxiway geometry and how expeditiously pilots vacate the runway. 2.2.5 Departure Runway Occupancy Time
Data could not be found for the runway occupancy time of departing aircraft, therefore the estimates given in de Neufville and Odoni [8] were used. These
3
3.1
Results
Impact at Dierent Airports
The runway capacity model was run using the inputs described in Section 2.2, for both low (450 seats) and typical (525 seats) aircraft passenger capacities as well as a baseline 747-400 with 418 seats. The results of this analysis are compared in Figure 6 and in detail for each airport in Figure 7. The results show that runway passenger throughput does increase in all cases however in the lower capacity A380 case the gains are marginal. Any gains in passenger capacity come at a cost in terms of operations per hour. The comparison with the 747-400 shows that in all cases a move to 747-400s (which have no addition separation requirements) would increase the passenger throughput more than any of the A380 congurations modeled. This implies that if passenger throughput were the only motivation for customers of the A380 then a high capacity Heavy aircraft would better suit their needs. Fortunately for Airbus the A380 has other economic, environmental and passenger comfort benets over other Heavy aircraft.currently on the market
10 of 15
Alexander Donaldson
9000 Airport Airravl Passenger Capacity (per hour) 8500 8000 7500 7000 6500 6000 5500 50000 2 4 6 8 10 Hourly Arrivals (All runways) 12
14
Figure 6: Comparison of the eect of the A380 on operations at LHR, JFK and DXB for 450 seat and 525 seat variants of the A380 and a 416 seat 747-400
11 of 15
Alexander Donaldson
62
6800 6750 6700 6650 66000 1 4 5 2 3 Hourly A380 Arrivals (All runways)
(b) At JFK
35.5 35.0 Arrivals per Hour 34.5 34.0 33.5 33.0 632.5
956
Figure 7: Impact of A380 Operations on Arrival Capacity in terms of operations (black) and passengers (colored - light: 450 seat A380, dark: 525 seat A380)
12 of 15
Alexander Donaldson
3.2
The comparison of the eect of dierent separation criteria Figure 8 shows considerable variability in the eect of the dierent criteria on a common set of operational data. It is particularly clear why the ICAO interim guidance was revised just before the A380 entered service - the interim guidance would have caused a substantial loss in passenger throughput (dierence between red and gray line in Figure 8). Also of note is the fact that the conservative FAA guidance leads to a reduction in throughput for the low density A380 conguration when applied to Heathrow. These results highlight the importance for regulatory agencies for nding the right balance between ensuring safety and improving the eciency of the air transportation system.
8200 Airport Airravl Passenger Capacity (per hour) 8000 7800 7600 7400 7200 7000 6800 66000 2
UK-High US-High ICAO Ini-High ICAO Final-High UK-Low US-Low ICAO Ini-Low ICAO Final-Low
4 6 Hourly A380 Arrivals (All runways) 8 10
Figure 8: Comparison of the eect on operations at LHR of dierent separation criteria (UK CAA, US FAA, ICAO initial and ICAO nal guidance)
13 of 15
Alexander Donaldson
Conclusions
This study has shown that the A380 does not have a detrimental impact on passenger throughput as some members of the aviation industry had fear it would. The increasing numbers of A380 operations at hub airports around the world will however have a signicant impact on the breakdown of their trac by weight category. Major international hubs will operate most eciently if the trac mix is heavily weighted towards Heavy and A380 category aircraft. For airports (such as JFK) that expect to continue operating a substantial number of lighter aircraft as well as several A380 operations careful management of the A380 operation will be required to ensure that the new aircraft has a positive impact on their airport.
14 of 15
Alexander Donaldson
References
[1] Airbus. Orders and deliveries spreadsheet. http://www.airbus.com/fileadmin/ backstage/documents/od/November1_2009.xls, November 2009. [2] Max Kingsley-Jones. Dubai 09: Airbus set to decide on A380 production revamp. http://www.flightglobal.com/articles/2009/11/17/335080/dubai-09airbus-set-to-decide-on-a380-production-revamp.html, November 2009. [3] Max Kingsley-Jones. Airbus slows A380 nal assembly ramp-up. http: //www.flightglobal.com/articles/2009/05/14/326416/airbus-slows-a380final-assembly-ramp-up.html, May 2009. [4] David Kaminski-Morrow. http://www.flightglobal.com/articles/2009/01/26/ 321601/uk-rethinks-a380-wake-separation-from-heavy-jets.html, January 2009. [5] Civil Aviation Authority. London heathrow aerodrome chart. http://www.natsuk.ead-it.com/aip/current/ad/EGLL/EG_AD_2_EGLL_2-1_en.pdf, August 2009. [6] BAA. BAA heathrow website: Mixed mode. http://www.heathrowairport. com/portal/page/Heathrow%5EGeneral%5EOur+business+and+community% 5EFuture+growth%5EMixed+mode/1c8851dcd7423110VgnVCM10000036821c0a___ _/448c6a4c7f1b0010VgnVCM200000357e120a____/. [7] FAA. New York JFK airport diagram. http://www.naco.faa.gov/d-tpp/0913/ 00610AD.PDF, December 2009. [8] R. De Neufville and A.R. Odoni. Airport systems: Planning, design, and management. McGraw-Hill Professional, 2002. [9] Safety Regulation Group. CAP 493 Manual of Air Trac Services Part 1. Civil Aviation Authority, November 2009. [10] Federal Aviation Administration. Order JO 7110.65S Air Trac Control, change 1 edition, February 2008. [11] Boeing. Airport reference code and approach speeds for boeing airplanes. http: //www.boeing.com/commercial/airports/faqs/arcandapproachspeeds.pdf, August 2007. [12] British Airways and BAA Heathrow. Results from two surveys of the use of reverse thrust of aircraft landing at heathrow airport. http://www.dft.gov.uk/adobepdf/ 165217/282786/6_ENV1128.pdf, November 2005.
15 of 15