AIRPORT CAPACITY IMBALANCE - Eurocontrol · 2020. 12. 11. · The chance for demand/capacity...

PERFORMANCE REVIEW COMMISSION EUROCONTROL, Rue de la Fusée 96, 1130 Brussels, Belgium

[email protected]

TECHNICAL NOTE

07-10-2020

AIRPORT CAPACITY IMBALANCE

STUDY

PERFORMANCE REVIEW COMMISSION

1

Edition History

The following table records the complete history of the successive editions of the present document.

Edition No. Edition Validity Date Reason

0.1 30-07-2020 Initial draft published for consultation

0.2 19-08-2020 Correction of time period used for the study. Clarification of methodology followed for identification of the runway configurations and their share.

0.3 29-09-2020 Reviewed text based on consultation feedback.

0.4 06-10-2020 Runway identification labels added to airport layouts.

2

Contents I. List of Figures .................................................................................................................................. 5

II. List of Tables ................................................................................................................................... 5

1. PURPOSE OF THE STUDY ................................................................................................................ 6

2. BACKGROUND ................................................................................................................................ 6

2.1. THE INTERDEPENDENCY TRIANGLE .......................................................................................... 7

2.2. CAPACITY ANALYSIS ................................................................................................................... 7

2.3. AIRPORT OPERATIONAL PERFORMANCE.................................................................................. 9

2.4. RUNWAY SYSTEM CONFIGURATION ......................................................................................... 9

3. APPROACH .................................................................................................................................... 11

3.1. AVAILABLE DATA ...................................................................................................................... 11

3.2. ANALYSIS .................................................................................................................................. 12

4. RESULTS ........................................................................................................................................ 17

5. CONCLUSIONS .............................................................................................................................. 24

6. REFERENCES ................................................................................................................................. 26

7. ANNEX I: RESULTS SUMMARY ............................................................................................................ 27

1. EBBR (BRU) – Brussels ............................................................................................................... 28

2. EBCI (CRL) - Charleroi ................................................................................................................ 29

3. EDDB (SXF) - Berlin/ Schoenefeld ............................................................................................. 30

4. EDDC (DRS) – Dresden .............................................................................................................. 31

5. EDDE (ERF) – Erfurt ................................................................................................................... 32

6. EDDF (FRA) - Frankfurt .............................................................................................................. 33

7. EDDG (FMO) - Muenster-Osnabrueck ....................................................................................... 34

8. EDDH (HAM) - Hamburg ............................................................................................................ 35

9. EDDK (CGN) - Cologne-Bonn ..................................................................................................... 36

10. EDDL (DUS) - Dusseldorf ....................................................................................................... 37

11. EDDM (MUC) - Munich ......................................................................................................... 38

12. EDDN (NUE) - Nuremberg ..................................................................................................... 39

13. EDDP (LEJ) - Leipzig-Halle ...................................................................................................... 40

14. EDDR (SCR) - Saarbruecken ................................................................................................... 41

15. EDDS (STR) - Stuttgart ........................................................................................................... 42

16. EDDT (TXL) - Berlin/ Tegel ..................................................................................................... 43

17. EDDV (HAJ) - Hanover ........................................................................................................... 44

18. EDDW (BRE) - Bremen ........................................................................................................... 45

19. EETN (TLL) - Tallinn ................................................................................................................ 46

20. EFHK (HEL) - Helsinki/ Vantaa ............................................................................................... 47

3

21. EGBB (BHX) - Birmingham ..................................................................................................... 48

22. EGCC (MAN) - Manchester .................................................................................................... 49

23. EGGD (BRS) - Bristol .............................................................................................................. 50

24. EGGW (LTN) - London/ Luton................................................................................................ 51

25. EGKK (LGW) - London/ Gatwick ............................................................................................ 52

26. EGLC (LCY) - London/ City ..................................................................................................... 53

27. EGLL (LHR) - London/ Heathrow ........................................................................................... 54

28. EGNT (NCL) - Newcastle ........................................................................................................ 55

29. EGPD (ABZ) - Aberdeen ......................................................................................................... 56

30. EGPF (GLA) - Glasgow ............................................................................................................ 57

31. EGPH (EDI) - Edinburgh ......................................................................................................... 58

32. EGSS (STN) - London/ Stansted ............................................................................................. 59

33. EHAM (AMS) - Amsterdam/ Schiphol ................................................................................... 60

34. EICK (ORK) - Cork ................................................................................................................... 61

35. EIDW (DUB) - Dublin ............................................................................................................. 62

36. EKCH (CPH) - Copenhagen/ Kastrup ...................................................................................... 63

37. ELLX (LUX) - Luxembourg ...................................................................................................... 64

38. ENGM (OSL) - Oslo/ Gardermoen ......................................................................................... 65

39. EPWA (WAW) - Warszawa/ Chopina .................................................................................... 66

40. ESGG (GOT) - Göteborg ......................................................................................................... 67

41. ESSA (ARN) - Stockholm/ Arlanda ......................................................................................... 68

42. EVRA (RIX) - Riga ................................................................................................................... 69

43. EYVI (VNO) - Vilnius ............................................................................................................... 70

44. GCFV (FUE) - Fuerteventura .................................................................................................. 71

45. GCLP (LPA) - Gran Canaria ..................................................................................................... 72

46. GCRR (ACE) - Lanzarote ......................................................................................................... 73

47. GCTS (TFS) - Tenerife Sur/Reina Sofia ................................................................................... 74

48. GCXO (TFN) - Tenerife North ................................................................................................. 75

49. LBSF (SOF) - Sofia .................................................................................................................. 76

50. LCLK (LCA) - Larnaca .............................................................................................................. 77

51. LDZA (ZAG) - Zagreb .............................................................................................................. 78

52. LEAL (ALC) - Alicante ............................................................................................................. 79

53. LEBB (BIO) - Bilbao ................................................................................................................ 80

54. LEBL (BCN) - Barcelona .......................................................................................................... 81

55. LEIB (IBZ) - Ibiza ..................................................................................................................... 82

56. LEMD (MAD) - Madrid/ Barajas ............................................................................................ 83

4

57. LEMG (AGP) - Málaga ............................................................................................................ 84

58. LEPA (PMI) - Palma de Mallorca ............................................................................................ 85

59. LEVC (VLC) - Valencia ............................................................................................................ 86

60. LEZL (SVQ) - Sevilla ................................................................................................................ 87

61. LFBO (TLS) - Toulouse-Blagnac .............................................................................................. 88

62. LFLL (LYS) - Lyon-Saint-Exupéry ............................................................................................. 89

63. LFML (MRS) - Marseille-Provence ......................................................................................... 90

64. LFMN (NCE) - Nice-Côte d’Azur ............................................................................................. 91

65. LFPG (CDG) - Paris-Charles-de-Gaulle ................................................................................... 92

66. LFPO (ORY) - Paris-Orly ......................................................................................................... 93

67. LFSB (BSL) - Bâle-Mulhouse................................................................................................... 94

68. LGAV (ATH) - Athens ............................................................................................................. 95

69. LHBP (BUD) - Budapest/ Ferihegy ......................................................................................... 96

70. LICC (CTA) - Catania ............................................................................................................... 97

71. LIMC (MXP) - Milan/ Malpensa ............................................................................................. 98

72. LIME (BGY) - Bergamo ........................................................................................................... 99

73. LIMF (TRN) - Torino Caselle ................................................................................................. 100

74. LIML (LIN) - Milan/ Linate.................................................................................................... 101

75. LIPE (BLQ) - Bologna ............................................................................................................ 102

76. LIPZ (VCE) - Venice .............................................................................................................. 103

77. LIRA (CIA) - Rome/Ciampino ............................................................................................... 104

78. LIRF (FCO) - Rome/Fiumicino .............................................................................................. 105

79. LIRN (NAP) - Naples ............................................................................................................. 106

80. LJLJ (LJU) - Ljubljana ............................................................................................................ 107

81. LKPR (PRG) - Prague ............................................................................................................ 108

82. LMML (MLA) - Malta ........................................................................................................... 109

83. LOWW (VIE) - Vienna .......................................................................................................... 110

84. LPFR (FAO) - Faro ................................................................................................................ 111

85. LPPR (OPO) - Porto .............................................................................................................. 112

86. LPPT (LIS) - Lisbon ............................................................................................................... 113

87. LROP (OTP) - Bucharest/ Otopeni ....................................................................................... 114

88. LSGG (GVA) - Geneva .......................................................................................................... 115

89. LSZH (ZRH) - Zürich .............................................................................................................. 116

90. LZIB (BTS) - Bratislava .......................................................................................................... 117

5

I. List of Figures

Figure 1: The interdependency triangle .................................................................................................. 7

Figure 2: Airport capacity planning phases. ............................................................................................ 8

Figure 3: Steps of the analysis. .............................................................................................................. 12

Figure 4: Difference in peak service rate and share for representative runway configurations .......... 18

Figure 5: Airport Capacity Resilience for the 90 airports analysed in this study .................................. 20

Figure 6: Runway configurations’ impact on peak service rate and performance ............................... 22

Figure 7: Runway configurations’ impact on peak service rate and performance ............................... 23

II. List of Tables

Table 1: Identification of runway system configuration for each time interval. .................................. 13

Table 2: Calculation of peak throughput. ............................................................................................. 14

Table 3: Representative runway system configurations and peak service rates. ................................. 14

6

1. PURPOSE OF THE STUDY

Available runways at a given airport, together with their relative disposition, dimensions,

navigational aids, ATC procedures, environmental constraints and local meteorological

conditions (wind rouse, visibility,…) can result in very different operating conditions, with a

direct impact on the capacity of the runway system and the operational performance.

This study intends to evaluate the potential imbalance between these operating conditions

in terms of capacity and performance, through the identification of runway configurations

and the analysis of the difference in capacity and performance depending on these

configurations, associated with the probability of each configuration.

There are therefore 3 main elements to investigate:

- Capacity per runway configuration.

- Operational performance associated to the runway configuration.

- Runway configuration probability.

2. BACKGROUND

The airport capacity declaration and associated scheduling process entails high complexity as

it is influenced by many different factors.

Airport Airside Capacity refers to the ability of the airport runway/taxiway/apron system to

handle a given demand of flights within a specified time period, incurring an acceptable level

of delay (to be determined by the airport stakeholders). It is defined by the International Civil

Aviation Organisation (ICAO) as the ‘number of movements per unit of time that can be

accepted during different meteorological conditions’ [1], whilst Airport Council International

defines it in terms of ‘maximum aircraft movements per hour assuming average delay of no

more than four minutes, or such other number of delay minutes as the airport may set’ [3].

Airport capacity is a combination of the available infrastructures, the existing ATM/ANS

systems and the capabilities of human actors. Investments in the runway system

infrastructure are usually the most expensive, so the capacity at the apron, taxiway system

and terminal should always be adapted to get the most out of the runway system, being the

determining factor for overall capacity. Additionally, new technologies and optimization in

aircraft spacing and sequencing could result in an airport capacity growth when accounting

for the operational side.

Among all factors affecting the runway system capacity, the runway layout and configuration

is considered as the most relevant. While some airport layouts allow for very similar operation

conditions in one runway configuration or another, resulting in equal or similar capacities for

all possible runway configurations, other layouts do not offer that “symmetric capacity” as

the different configurations might have very different operating conditions, dependencies

and limitations.

7

2.1. THE INTERDEPENDENCY TRIANGLE

Airport capacity, demand and delay are three elements that influence each other (see Figure

1). Capacity refers to the theoretical traffic density the airport can absorb while the demand

corresponds to the airline scheduled operations in correspondence to the declared capacity.

Both elements are linked by the delay which results from the disproportion between the

demand and the available capacity. If one of them changes, the other 2 elements might be

affected. The operational problem to be addressed by planners and operational decision

makers is how to strike a balance between these elements, e.g. how much delay is acceptable

to accommodate more demand.

Figure 1: The interdependency triangle

In case of a change in the runway configuration that reduces the capacity, different scenarios

are possible:

- If the change is unforeseen, the demand is unlikely to be adapted fast enough to this

reduction in capacity. In this case demand (temporarily) exceeds the new (reduced)

capacity and there will be an impact on the airport performance and related delays.

- If the change is unforeseen but the demand remains below the new (reduced)

capacity, the airport system could cope without an impact on performance.

- If the change is foreseen (e.g. night runway configurations due to noise abatement),

demand should be adapted to account for the available runway system capacity. In

this scenario demand will be managed to the available capacity (e.g. limited number

of slots). The chance for demand/capacity imbalance is kept to a minimum but still can

exist.

Hence, assuming no change in demand, a change in runway configuration might have an

impact on the airport performance.

2.2. CAPACITY ANALYSIS

One of the main difficulties found while developing the present study deals with the

calculation of the airport capacity, as there is no universal method for this calculation.

Although all methods might take into account similar factors, there are different approaches.

Among the elements used in such studies are: structural layout (runway, taxiway, apron,

gates, terminals, local airspace); ATC procedures, environmental impact and economic

factors.

8

The analysis and methodologies will also depend on the time horizon for which the capacity

calculation is being made, serving different purposes in the airport capacity plans, as shown

in Figure 2

Figure 2: Airport capacity planning phases.

An example of capacity calculation methodology would follow these steps:

1. Historical throughput data (track performance & areas to prioritize)

2. Tables & analytical models (high level starting point)

3. Simulation models (fast time and real time)

4. Pareto frontiers (optimum solution)

Nevertheless, as previously mentioned, not all airports do regular capacity studies or follow

the same steps. At the moment the only common information repository including airport

capacity is Eurocontrol’s Airport Corner1 [4] where airports can report their capacities on a

voluntary basis. Ideally, this information is provided per runway configuration. However,

many airports do not provide any information in this regard. Even when the information is

available, the differences between methodologies and recurrence in the capacity studies for

airports across Europe make very difficult to have a consistent capacities database.

Peak Service Rate

When demand is above the operational capacity, the delivered throughput (that is, the

number of movements in a certain period of time) will be an indication of the capacity itself.

This is the principle behind the peak service rate: defined as the percentile 99 of the

throughput per hour, it is a proxy for the airport operational capacity, that is, the highest

sustainable throughput the airport can achieve under optimum conditions, assuming there is

sufficient demand, like during periods of congestion.

The percentile 99 is based on the cumulative distribution of the movements per hour over a

sample time period.

Note that at airports which are never or rarely congested, the peak service rate might be

significantly lower than the peak airport operational capacity. This proxy will be valid only

when the period considered includes enough hours where the demand was in fact above the

1 Airport Corner is an airport-focused data repository developed by Eurocontrol with operational information provided by the airports (as a result of local coordination between Airport Operator and ATC)

OPERATIONALPLANNEDSTRATEGIC

-X years M-18 to D-8 D-7 to D

MACRO-STRATEGIC• Long term assessment• Infrastructure capability

STRATEGIC• Mid term assessment• Airline schedule coordination

meeting

PRE-TACTICAL• Short term assessment• Detailed capacity plan

9

capacity. If an airport is underutilised, the peak service rate will simply detect the peak

demand, but not the capacity.

At airports that are rarely congested, the observed maximum throughput can provide a better

indication of the operational capacity.

Using the peak service rate as proxy allows for a consistent analysis of operational capacity

across all airports based on historical data.

2.3. AIRPORT OPERATIONAL PERFORMANCE

As result of airport demand/capacity balancing, the traffic demand will be ‘pushed’ through the capacity bottlenecks (constraints) at the various planning stages and flight phases. Depending on the phase at which the capacity constraint is known, the balancing will be done in different ways. The runway configuration defines one of these capacity constraints.

Until the day before operations: Some runway configurations are already known (or preselected for certain times of the day) long enough in advance for the schedule to be adapted to them and therefore it should result in no impact on performance.

On the day of operations: If there is the need to change to a less favourable runway configuration than the one initially planned, the demand on the runways is managed differently for arrivals and departures: ‐ For arrivals: A combination of arrival regulations (leading to arrival ATFM delay)

together with queuing in the Arrival Sequencing and Metering Area (leading to ASMA additional time). As regulations will only hold the flights before their take-off, they are not able to arrange the arrival flow with immediate effect. Therefore when there is an unexpected change of configuration, the adjustment of the arrival flow is done through holdings and vectoring in the approach.

‐ For departures: A combination of pre-departure delays (flights held at the stand/parking position, with or without departure regulations and leading to ATC pre-departure delay), and queuing at the runway (leading to additional taxi-out time). The balance between these two measures will also depend on other factors like the need to free parking stands.

In summary, a change in runway configuration creates (potentially) an airport capacity resilience issue which has an impact on the airport performance. This impact on performance can be studied by observing the following indicators: ‐ Additional taxi-out time ‐ Additional ASMA time ‐ ATC pre-departure delay ‐ Arrival ATFM delay

2.4. RUNWAY SYSTEM CONFIGURATION

The preferable runway direction is related to the wind conditions (direction and speed) but

the decision on runway configuration (OPS) in use also depends highly on other factors:

10

- Operational Safety. Aircraft lift is by design influenced and benefitted from

aircraft departing and landing into the wind.

- Demand. Arrival and departure demand play a key role in configurations selection,

especially in high demand situation when high capacity configurations are preferred

to serve incoming traffic.

- Air Navigation Systems. Existence, performance and operational status of the

navigational aids (both ground-based and space based) and instrument approach

procedures serving each runway, as different levels of accessibility derived from

asymmetries in instrument approach performance may significantly influence an

airport’s runway configuration.

- Meteorology (ceiling & visibility + wind gusts). Besides wind speed and direction,

other meteorological conditions are of great importance for runway system

configuration, such as visibility and cloud ceiling and wind gusts that can cause serious

harm to aircraft on its vicinity.

- Noise abatement procedures. Noise abatement procedures are used at most major

airports in order to reduce noise impact on neighbour communities and are normally

active at night and early morning period.

- Inertia (controllers’ preference). Air traffic controllers tend to prefer certain runway

system configurations or to remain in a same configuration in order to avoid changes,

so it has an important factor on configuration selection. These habits or Subtle

Navigation Factors (SNF) have been identified in SESAR projects using Machine

Learning Techniques.

- Time of the day (curfews). The time of the day influences the staff availability and the

range of possible configurations that can be selected.

- Coordination (TMA/airport). Flows in and out the airport need to be coordinated,

especially in multi airport Terminal Manoeuvring Areas.

- Other factors: Unavailability of runways (works in progress, maintenance, snow

removal…) or other systems (runway lighting system and MET system status).

Regulatory authorities may further restrict the use of the runway system to so-called

preferential runway system (PRS). Typically, PRS refers to a subset of the total set of runway

system configurations defined by specific conditions.

Research studies are mainly focused on configuration selection process prediction. These

investigations are based on two types of models: prescriptive & descriptive:

- Prescriptive models: Look for an optimal solution (accounting different factors)

- Descriptive models: Conduct historical data analysis

Examples on configuration prediction models include a data-driven model using discrete

choice modelling framework which computes configuration prediction in every next 15 min

interval, extended to 3h probabilistic forecast. Case studies performed in LGA and SFO

11

airports in USA reveal an accuracy of ≈80% [5] Another example is a decision-tree based

model to predict airport acceptance rate used as a decision support tool in Ground Delay

Programs (GDPs) [6].

3. APPROACH

Given the lack of consistent capacity and runway configuration information, the analysis will

use a data driven descriptive model that focuses on the available data in the Airport Operator

Data Flow (APDF) managed by the PRU. Furthermore, this data source is currently used for

the computation of the required indicators for the performance monitoring, so using the

same source for all areas of the study ensures the alignment between them.

3.1. AVAILABLE DATA

The Airport Operator Data Flow is established for 90 airports (status as April 2020) and it

includes, amongst other extensive data for every flight, the runway time (that is, take off time

for departures and landing time for arrivals) for every movement, the type of movement

(arrival or departure) and the runway used.

The data is provided monthly by the airport operators and integrated in a common database

after data quality checks.

This data allows for an approach that addresses the key elements of this study:

- Identification of runway configuration: based on the type of movement, time at the

runway and runway used.

- Runway configuration probability: understood as the percentage of time each

configuration is in use at each airport during the period being analysed.

- Capacity per runway configuration: understood as the peak service rate of each

configuration and calculated as the 99th percentile of the throughput in 1 hour

intervals with such runway configuration.

- Performance indicators for each movement and associated with the corresponding

runway configuration. The analysis of the impact on arrival ATFM delay will be dealt

with in the next version of this study. Due to data quality issues in the provision of pre-

departure delay information, the analysis of ATC pre-departure delay across most of

the 90 airports is not possible at the moment. Therefore the performance indicators

analysed per runway configuration will be Additional ASMA time and Additional taxi-

out time.

These 90 airports present a wide variety of typologies of airport layout (Annex I: Results

summary includes the results with layouts of each of the 90 airports analysed for further

information):

1 runway

2 runways intersecting

2 runways not parallel

12

2 runways parallel and closely spaced

2 runways parallel and independent

3 or more runways

The study is conducted for the calendar year 2019, covering the time window between 7h

and 22h local time for each day.

3.2. ANALYSIS

The APDF data allows for a post-ops data driven analysis including the runway use. The

objective is to identify the runway system configurations in use based on the historical data,

taking into account that in theory each airport with N runways has 6N possible configurations

(assuming each runway can be operated for arrivals, departures or both and in either

direction) (for example an airport with 2 landing strips could be operated in 36 different

ways).

In parallel, as a proxy for capacity, the peak service rate (that is, the percentile 99 of the hourly

throughput) will be calculated for each of these runway system configurations.

Figure 3: Steps of the analysis.

For each airport the analysis involves the following steps:

1. Establish 15 min time intervals that will be used for identification of the runway system

configuration (c.f. Table 1)

2. Identification, in each time interval, of active runways and type of movement (ARR/DEP)

at each runway (c.f. Table 1), resulting in a detected runway configuration for that 15 min.

time interval.

When for a time interval only a type of movement (e.g. arrivals) has been observed, that

configuration will be completed with a runway for the other type of movement (departures

in this example). That runway will be assigned by analysing the adjacent intervals, or if

necessary, assuming the runway is used in mixed mode. This is generally observed at small

airports with very low demand.

3. Identification of sustained runway configurations for each 15 min. rolling hour, by

checking, for each 15 min time interval (i), configurations in the following 45 min [(ii), (iii),

1. APDF in 15 min time intervals

3. Runway configuration for

each interval. Probability

2. Count ARR&DEP per

runway

5. Calculate imbalance

and resilience

4. Calculate peak service rate and

performance for each runway configuration

13

(iv)]. If the configuration observed in (i) is also observed in at least 2 of the next 3 intervals

[(ii), (iii), (iv)], the configuration is considered sustained and valid (condition TRUE in Table

2) and associated to the rolling hour starting with time interval (i).

Table 1: Identification of runway system configuration for each time interval.

The reason to allow 1 in 4 time intervals to correspond to another configuration is to be

able to accommodate some unexpected movement that does not exactly fit with the

runway configuration (in the example in Table 2, the use of runway 25R for a departure of

a Heavy aircraft, although that was not the standard runway for departures in that hour).

The calculation of the time share for each configuration is done on the basis of these rolling

hours with condition TRUE. To identify typical configurations, a minimum share of 3% of

the analysed time has been considered (that is, if a configuration is not active more than

3% of the period of the study, it is not considered representative). The time window

considered is 7 to 22h local time to discard night operations where the demand is normally

too low to consider the peak service rate as a proxy for capacity and the configurations

that are mainly related to environmental constrains.

ARR:25R-DEP:25L

ARR:02-DEP:07R

AIRPORT CALLSIGN TIME UTC LOCAL HOUR PERIOD ARR or DEP RUNWAY

LEBL AFL2512 29-06-2018 20:46 22 4 ARR 25R

LEBL VLG20GK 29-06-2018 20:47 22 4 ARR 25R

LEBL VLG91QL 29-06-2018 20:49 22 4 ARR 25R

LEBL VLG18JC 29-06-2018 20:50 22 4 ARR 25R

LEBL VLG66YE 29-06-2018 20:52 22 4 ARR 25R

LEBL ABG8326 29-06-2018 20:53 22 4 DEP 25L

LEBL KLM81J 29-06-2018 20:53 22 4 ARR 25R

LEBL WZZ1075 29-06-2018 20:55 22 4 DEP 25L

LEBL RYR6369 29-06-2018 20:55 22 4 ARR 25R

LEBL RYR15HD 29-06-2018 20:56 22 4 DEP 25L

LEBL RYR74BH 29-06-2018 20:56 22 4 ARR 25R

LEBL VLG18MC 29-06-2018 20:58 22 4 ARR 25R

LEBL ORO901 29-06-2018 21:03 23 1 DEP 07R

LEBL BCS8049 29-06-2018 21:05 23 1 DEP 07R

LEBL IBK59V 29-06-2018 21:07 23 1 ARR 02

LEBL VLG39LM 29-06-2018 21:09 23 1 ARR 02

LEBL QTR142 29-06-2018 21:09 23 1 DEP 07R

LEBL WZZ152 29-06-2018 21:11 23 1 ARR 02

LEBL IBK7VY 29-06-2018 21:12 23 1 ARR 02

LEBL UAE188 29-06-2018 21:13 23 1 DEP 07R

LEBL AFR144X 29-06-2018 21:14 23 1 ARR 02

LEBL IBK3ZR 29-06-2018 21:16 23 1 ARR 02

AIRPORT CALLSIGN TIME UTC LOCAL HOUR PERIOD ARR or DEP RUNWAY

LEBL AFL2512 29-06-2018 20:46 22 4 ARR 25R

LEBL VLG20GK 29-06-2018 20:47 22 4 ARR 25R

LEBL VLG91QL 29-06-2018 20:49 22 4 ARR 25R

LEBL VLG18JC 29-06-2018 20:50 22 4 ARR 25R

LEBL VLG66YE 29-06-2018 20:52 22 4 ARR 25R

LEBL ABG8326 29-06-2018 20:53 22 4 DEP 25L

LEBL KLM81J 29-06-2018 20:53 22 4 ARR 25R

LEBL WZZ1075 29-06-2018 20:55 22 4 DEP 25L

LEBL RYR6369 29-06-2018 20:55 22 4 ARR 25R

LEBL RYR15HD 29-06-2018 20:56 22 4 DEP 25L

LEBL RYR74BH 29-06-2018 20:56 22 4 ARR 25R

LEBL VLG18MC 29-06-2018 20:58 22 4 ARR 25R

LEBL ORO901 29-06-2018 21:03 23 1 DEP 07R

LEBL BCS8049 29-06-2018 21:05 23 1 DEP 07R

LEBL IBK59V 29-06-2018 21:07 23 1 ARR 02

LEBL VLG39LM 29-06-2018 21:09 23 1 ARR 02

LEBL QTR142 29-06-2018 21:09 23 1 DEP 07R

LEBL WZZ152 29-06-2018 21:11 23 1 ARR 02

LEBL IBK7VY 29-06-2018 21:12 23 1 ARR 02

LEBL UAE188 29-06-2018 21:13 23 1 DEP 07R

LEBL AFR144X 29-06-2018 21:14 23 1 ARR 02

LEBL IBK3ZR 29-06-2018 21:16 23 1 ARR 02

TIME INTERVAL n

TIME INTERVAL n+1

14

Table 2: Calculation of peak throughput.

4. The analysis of the peak service rate requires first the calculation of the throughput for all

rolling hours. Taking then only the hours with valid configurations (that is, rolling hours

where a configuration is sustained in 3 of 4 consecutive time intervals), we calculate the

percentile 99 (peak service rate) for each configuration with a representative time share

(>3%) (see 3.2 Step 3). The throughput analysis can be split in arrivals, departures or total

(c.f. Table 3).

Table 3: Representative runway system configurations and peak service rates.

5. Once a runway configuration has been assigned to certain interval, each flight in that

interval can be associated with a configuration. As the performance indicators are

calculated per flight, this association will be used for the analysis of performance per

runway configuration, resulting in different additional ASMA and taxi-out times per runway

configuration.

Now that all the components of the study have been analysed, the following indicators can

be calculated:

3.2.1. Airport Capacity Resilience

AIRPORT TIME_SLICE CONFIGURATION FOUR_SLICE ARR_HOURLY_THROUGHPUTDEP_HOURLY_THROUGHPUTTOTAL_HOURLY_THROUGHPUT

LEBL 29-06-2018 20:30 ARR:25R - DEP:25L TRUE 34 31 65



LEBL 29-06-2018 21:15 ARR:25R - DEP:25L - DEP:25R FALSE 28 32 60





LEBL 29-06-2018 22:30 ARR:25R - DEP:25L FALSE 21 18 39

LEBL 29-06-2018 22:45 ARR:25R - DEP:25L FALSE 13 16 29

LEBL 30-06-2018 06:00 DEP:07R FALSE 8 19 27

LEBL 30-06-2018 06:15 ARR:02 - DEP:07R TRUE 11 26 37

LEBL 30-06-2018 06:30 ARR:02 - DEP:07R FALSE 9 34 43

LEBL 30-06-2018 06:45 ARR:02 - DEP:07R FALSE 13 37 50



AIRPORT CONFIGURATION PERC99_ARR_HOURLY_THROUGHPUTPERC99_DEP_HOURLY_THROUGHPUTPERC99_TOTAL_HOURLY_THROUGHPUT

LEBL ARR:25R - DEP:25L 37 39 70

LEBL ARR:07L - DEP:07R 38 38 70

LEBL ARR:02 - DEP:07R 29 31 56

LEBL ARR:25L - DEP:25L 26 28 49

15

When the probability of a certain runway configuration and the corresponding peak service

rate have been established, the capacity resilience for a given configuration, conf i, will be

calculated as:

Configuration Capacity Resilienceconf i (%) = 1 − (Probabilityconf i ∗ Capacity Reductionconf i)

Where:

Probabilityconf i = share of time intervals with runway system configuration Conf i

Capacity Reductionconf i =Reference Capacity − P99 TotalConf i

Reference Capacity

Reference Capacity = maxi

(P99 TotalConf i)

Conf i ∶ all those configurations with a probability > 3%

Finally, the airport capacity resilience:

Airport Capacity Resilience = mini

(Configuration Capacity Resilienceconf i)

Understanding that the airport resilience will be the lowest of the resilience for the different

configurations.

When the Airport Capacity Resilience is 1, it means that there is no imbalance in terms of

capacity that can be delivered due to runway configurations (either because all

configurations offer the same capacity or because the probability of a runway configuration

with lower capacity is too low).

3.2.2. Impact on the peak service rate

Taking the previously calculated difference between the reference capacity and the peak

service rate of conf i: (Reference Capacity − P99 TotalConf i), calculated for each of the n

representative configurations, and the probability of such reduction (given by the probability

of the configuration i), the impact on the peak service rate, expressed in absolute number of

movements, can be calculated as the weighted average:

Impact on Peak Service Rate = ∑(Probabilityconf i ∗ (Reference Capacity − P99 TotalConf i))

𝑛

𝑖=1

This indicator informs in absolute terms of how many movements per hour are lost (assuming

enough demand) when taking all runway configurations into account, with respect to an

hypothetical 100% share of the most favourable runway configuration in terms of peak

service rate.

16

3.2.3. Impact on the performance indicators (Additional ASMA and taxi-out times)

In terms of additional ASMA and taxi-out times (c.f. 3.2 ANALYSIS: 5 abovePoint 5), the best

performance corresponds to the lowest additional times (that is, less queuing). Therefore in

this case the reference performance is the minimum, as follows:

Reference Add. ASMA time = mini

(Add. ASMA timeConf i)

And in a similar way to the impact on the peak service rate, the total impact on the additional

ASMA time will be the weighted average of the impact for each configuration:

Impact on Add. ASMA time

= ∑(Probabilityconf i ∗ (Add. ASMA timeConf i − Reference Add. ASMA time))

𝑛

𝑖=1

The same calculation applies for the additional taxi-out time.

17

4. RESULTS

This section provides an overview of the results after applying the described methodology to

the 90 airports under study Annex I: RESULTS SUMMARY provides details for each individual

airport. After applying the described methodology to the 90 airports under study, there are

several aspects to be evaluated. This section provides an overview of the results, see Annex

I: Results summary for further information on each airport.

4.1. Identification of runway configuration and probability

The approach shows good coverage in terms of identifying a valid runway configuration for

each time interval. The minimum share of 3% of the analysed time proves to be a reasonable

threshold to discard non-representative configurations, covering more than 95% of the

operations in 2019 for 73 out of 90 airports. Only four airports had less than 90 % of the

operation covered by these representative runway configurations: Hannover (EDDV; 89%

coverage), Helsinki (EFHK; 83%), Rome Fiumicino (LIRF; 86%) and Amsterdam (EHAM; 73%).

In cases like Amsterdam (EHAM) where more available runways result in numerous

configuration possibilities, the 15 minutes intervals might be too short to detect all runways

in use, or the 3% threshold too restrictive (as many runway configurations also mean reduced

shares for each one of them).

When comparing the identified runway configurations for each airport with the information

registered in the Airport Corner, the methodology proves to find not only the main runway

configurations, but also other operating modes that are not recognised in the Airport Corner

but are actually in use at these airports.

4.2. Capacity per runway configuration

Annex I: Results summary presents the detailed results for each airport including the

identified runway configurations and corresponding peak service rates for each airport,

together with the calculated resilience and impact. As mentioned in previous sections, the

main limitation for the identification of the capacity for each configuration is the absence of

enough demand, which might be the case for many of the smaller airports and even some of

the bigger ones.

Figure 4 illustrates the results for the top 30 airports in Europe (from which data is available

for 27) in terms of peak service rate and probability for each runway configuration. Each

bubble represents a runway configuration with the corresponding peak service rate (vertical

axis) and share of utilization (size of the bubble).

18

Figure 4: Difference in peak service rate and share for representative runway configurations

Within these 27 busier airports, 15 have no significant deviations in this peak service rate for

the main configurations, that is, the peak throughput does not seem that affected by the

runway configurations (the difference is 2 movements or less). On the other hand, there are

some cases like Oslo (ENGM), Helsinki (EFHK), Stockholm (ESSA) and Amsterdam (EHAM)

where the dispersion is much higher, and cases like Manchester, where not only there is a

significant deviation, but also an even distribution of the share amongst the configurations.

The proposed indicators (resilience and impact on peak service rate) will integrate these

different aspects (deviations and share) to be able to assess the imbalance of the operation

as a whole.

The comparison of the observed peak service rate and the maximum total throughput with

the capacities declared in the Airport Corner, for those airports where information is

available, yields mixed results.

At airports like Heathrow (EGLL), Gatwick (EGKK), Frankfurt (EDDF), Munich (EDDM),

Dusseldorf (EDDL), Dublin (EIDW), Warsaw (EPWA), Lisbon (LPPT), Zurich (LSZH) or Athens

(LGAV), the calculated peak service rate during 2019 is very close to the declared total

capacity for each runway configuration.

On the other hand, at other airports like Stockholm (ESSA), Helsinki (EFHK), Copenhagen

(EKCH) Rome (LIRF), Charles de Gaulle (LFPG) or Orly (LFPO), the peak service rate, or

051015202530

Fran

kfu

rt (

FRA

)

Am

ster

dam

(A

MS)

Par

is (

CD

G)

Lon

do

n (

LHR

)

Mad

rid

(M

AD

)

Mu

nic

h (

MU

C)

Bar

celo

na

(BC

N)

Ro

me

(FC

O)

Lon

do

n (

LGW

)

Vie

nn

a (V

IE)

Zuri

ch (

ZRH

)

Co

pen

hag

en (

CP

H)

Osl

o (

OSL

)

Du

blin

(D

UB

)

Mila

n (

MX

P)

Sto

ckh

olm

(A

RN

)

Bru

ssel

s (B

RU

)

Du

esse

ldo

rf (

DU

S)

Par

is O

rly

(OR

Y)

Lisb

on

(LI

S)

Ath

ens

(ATH

)

Pal

ma

(PM

I)

Man

ches

ter

(MA

N)

Lon

do

n (

STN

)

Hel

sin

ki (

HEL

)

War

saw

(W

AW

)

Ber

lin T

ege

l (TX

L)

0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Mo

vem

ents

/ho

ur

Peak Service Rate Total movements (arr + dep) at the top 30 airports in 2019per runway configuration

Bubble size: share of time of a given RWY configuration in use

19

percentile 99 is too low compared to the declared capacities. In these cases, the maximum

delivered throughput comes closer and might be a better guess for the capacity.

Nevertheless, as mentioned in the section 2.2 CAPACITY ANALYSIS, differences between

methodologies and recurrence in the airport capacity studies in Europe also make difficult a

consistent comparison.

4.3. Airport capacity resilience

Figure 5 represents the Airport Capacity Resilience results obtained for the analysed airports

following the described methodology. As highlighted in the chart, only 1 airport, Bratislava

(LZIB) has a resilience below 80%, and only other three airports, Gran Canaria (GCLP), Málaga

(LEMG) and Nice (LFMN) show a resilience value below 90%.

Although this study does not allow to decipher if the imbalance in the peak service rate is due

to lack of capacity or lack of demand, especially for the less busy airports like these four, the

analysis at airport level (See Annex I: Results summary for further information) allows to

better understand the imbalance at each particular airport, unveiling for example cases in

which the most common configuration is a single runway use (even when multiple runways

are available) that has lower peak service rate with an associated detrimental impact on

performance.

20

Figure 5: Airport Capacity Resilience for the 90 airports analysed in this study

The airport capacity resilience is a relative indicator, that is, the magnitude of the problem is

measured with respect to the reference capacity at the airport. Therefore, an imbalance of 4

movements at Amsterdam (EHAM; reference capacity=109) will have a lower reduction of

50%

60%

70%

80%

90%

100%EB

BR

(B

RU

) -

Bru

sse

ls

EBC

I (C

RL)

- C

har

lero

i

EDD

B (

SXF)

- B

erlin

Sch

oe

nef

eld

EDD

C (

DR

S) -

Dre

sden

EDD

E (E

RF)

- E

rfu

rt

EDD

F (F

RA

) -

Fran

kfu

rt

EDD

G (

FMO

) -

Mu

en

ster

EDD

H (

HA

M)

- H

amb

urg

EDD

K (

CG

N)

- C

olo

gne-

Bo

nn

EDD

L (D

US)

- D

uss

eld

orf

EDD

M (

MU

C)

- M

un

ich

EDD

N (

NU

E) -

Nu

rem

ber

g

EDD

P (

LEJ)

- L

eip

zig-

Hal

le

EDD

R (

SCR

) -

Saar

bru

ecke

n

EDD

S (S

TR)

- St

utt

gart

EDD

T (T

XL)

- B

erl

in T

egel

EDD

V (

HA

J) -

Ha

no

ver

EDD

W (

BR

E) -

Bre

me

n

EETN

(T

LL)

- Ta

llin

n

EFH

K (

HEL

) -

Hel

sin

ki

EGB

B (

BH

X)

- B

irm

ingh

am

EGC

C (

MA

N)

- M

anch

este

r

EGG

D (

BR

S) -

Bri

sto

l

EGG

W (

LTN

) -

Lon

do

n L

uto

n

EGKK

(LG

W)

- Lo

nd

on

Gat

wic

k

EGLC

(LC

Y) -

Lo

nd

on

Cit

y

EGLL

(LH

R)

- Lo

nd

on

Hea

thro

w

EGN

T (N

CL)

- N

ewca

stle

EGPD

(A

BZ)

- A

be

rde

en

EGPF

(G

LA)

- G

lasg

ow

Airport Capacity Resilience

50%

60%

70%

80%

90%

100%

EGPH

(ED

I) -

Ed

inb

urg

h

EGSS

(ST

N)

- Lo

nd

on

Sta

nst

ed

EHA

M (

AM

S) -

Am

ster

dam

EIC

K (

OR

K)

- C

ork

EID

W (

DU

B)

- D

ub

lin

EKC

H (

CP

H)

- C

op

en

hag

en

ELLX

(LU

X)

- Lu

xem

bo

urg

ENG

M (

OSL

) -

Osl

o

EPW

A (

WA

W)

- W

arsz

awa

ESG

G (

GO

T) -

Gö

teb

org

ESSA

(A

RN

) -

Sto

ckh

olm

EVR

A (

RIX

) -

Rig

a

EYV

I (V

NO

) -

Viln

ius

GC

FV (

FUE

) -

Fue

rte

ven

tura

GC

LP (

LPA

) -

Gra

n C

anar

ia

GC

RR

(A

CE)

- L

anza

rote

GC

TS

(TFS

) -

Ten

erif

e Su

r

GC

XO

(T

FN)

- Te

ner

ife

No

rth

LBSF

(SO

F) -

So

fia

LCLK

(LC

A)

- La

rnac

a

LDZA

(ZA

G)

- Za

gre

b

LEA

L (A

LC)

- A

lican

te

LEB

B (

BIO

) -

Bilb

ao

LEB

L (B

CN

) -

Bar

celo

na

LEIB

(IB

Z) -

Ibiz

a

LEM

D (

MA

D)

- M

adri

d

LEM

G (

AG

P)

- M

álag

a

LEP

A (

PM

I) -

Pa

lma

de

Mal

lorc

a

LEV

C (

VLC

) -

Va

len

cia

LEZL

(SV

Q)

- Se

villa

50%

60%

70%

80%

90%

100%

LFB

O (

TLS)

- T

ou

lou

se

LFLL

(LY

S) -

Lyo

n

LFM

L (M

RS)

- M

arse

ille

LFM

N (

NC

E) -

Nic

e

LFP

G (

CD

G)

‐ C

har

les

de

…

LFP

O (

OR

Y) -

Par

is O

rly

LFSB

(B

SL)

- B

âle

-Mu

lho

use

LGA

V (

AT

H)

- A

the

ns

LHB

P (

BU

D)

- B

ud

apes

t

LIC

C (

CTA

) -

Cat

ania

LIM

C (

MX

P)

‐ M

ilan

/…

LIM

E (B

GY)

- B

erg

amo

LIM

F (T

RN

) -

To

rin

o C

ase

lle

LIM

L (L

IN)

- M

ilan

/ Li

nat

e

LIP

E (B

LQ)

- B

olo

gna

LIP

Z (V

CE)

- V

enic

e

LIR

A (

CIA

) -

Ro

me

Cia

mp

ino

LIR

F (F

CO

) -

Ro

me

Fiu

mic

ino

LIR

N (

NA

P)

- N

aple

s

LJLJ

(LJ

U)

- Lj

ub

ljan

a

LKP

R (

PR

G)

- P

ragu

e

LMM

L (M

LA)

- M

alta

LOW

W (

VIE

) -

Vie

nn

a

LPFR

(FA

O)

- Fa

ro

LPP

R (

OP

O)

- P

ort

o

LPP

T (L

IS)

- Li

sbo

n

LRO

P (

OTP

) -

Bu

char

est

LSG

G (

GV

A)

- G

ene

va

LSZH

(ZR

H)

- Zü

rich

LZIB

(B

TS)

- B

rati

slav

a

21

resilience than the same imbalance at Bratislava (LZIB; reference capacity=18). The following

indicators (impact on peak service rate and performance) provide an indication of the issue

in absolute terms.

4.4. Impact on peak service rate and performance

The results of the calculated impact on both peak service rate and performance indicators derived from the use of different runway configurations are presented in Figure 6 and Figure 7. Regarding the impact on the peak service rate, 29 of the analysed airports show total capacity symmetry (impact on peak service rate is zero); at 36 airports the impact is less than 1 movement per hour, and 10 more airports have an impact between 1 and 2 movements. From the remaining 15 airports, the highest impacts are observed at Málaga (LEMG), followed by Nice (LFMN), Oslo (ENGM), Gran Canaria (GCLP), Helsinki (EFHK), Dusseldorf (EDDL) and Bratislava (LZIB). Most of these airports that show the biggest impact on peak service rate in absolute terms, also had the lowest resilience results. Additionally, when observing the impact on performance for these airports, it is noticeable that three of them, Oslo (ENGM), Málaga (LEMG) and Helsinki (EFHK), also suffer important combined impacts on the performance indicators of above 1 minute per flight. As explained before, an unexpected change of configuration can worsen performance in the

form of higher delays and holdings. In this study two performance indicators related to the

management of the arrival and departure flows are analysed separately for the identified

runway configurations (c.f. 3.2.3). Nevertheless, the runway configuration also has a great

influence in the taxi-out and approach routes, where it might perfectly be that one

configuration is more prone to suffer bottlenecks on the taxiway system or conflicting

approaches. In that case, even with equal or similar peak service rates, the performance of

two runway configurations might be very different. This is the case for airports like

Saarbruecken (EDDR), Porto (LPPR), Budapest (LHBP), Leipzig (EDDP), Vienna (LOWW), Zurich

(LSZH) and even Frankfurt (EDDF), all of them with an impact on delays above 1 minute per

flight (additional ASMA and taxi-out combined).

The detailed analysis airport by airport (see Annex I: Results summary) shows as well the peak service rates for arrival and departures for each runway configuration, together with the maximum hourly throughput observed in 2019. The results for arrivals and departures also evidence that some configurations are preferential for departure peaks and others for arrival peaks (e.g. Amsterdam)

22

Figure 6: Runway configurations’ impact on peak service rate and performance

0 2 4 6 8 10 12

GCLP

GCFV

EYVI

EVRA

ESSA

ESGG

EPWA

ENGM

ELLX

EKCH

EIDW

EICK

EHAM

EGSS

EGPH

EGPF

EGPD

EGNT

EGLL

EGLC

EGKK

EGGW

EGGD

EGCC

EGBB

EFHK

EETN

EDDW

EDDV

EDDT

EDDS

EDDR

EDDP

EDDN

EDDM

EDDL

EDDK

EDDH

EDDG

EDDF

EDDE

EDDC

EDDB

EBCI

EBBR

(movements/hour)

Impact Peak Service Rate

00.511.522.5

GCLP

GCFV

EYVI

EVRA

ESSA

ESGG

EPWA

ENGM

ELLX

EKCH

EIDW

EICK

EHAM

EGSS

EGPH

EGPF

EGPD

EGNT

EGLL

EGLC

EGKK

EGGW

EGGD

EGCC

EGBB

EFHK

EETN

EDDW

EDDV

EDDT

EDDS

EDDR

EDDP

EDDN

EDDM

EDDL

EDDK

EDDH

EDDG

EDDF

EDDE

EDDC

EDDB

EBCI

EBBR

(min/flight)

Impact Add. Taxi-Out Time Impact Add. ASMA Time

Impact of runway configurations usage on operational performance

23

Figure 7: Runway configurations’ impact on peak service rate and performance

0 2 4 6 8 10 12

LZIB

LSZH

LSGG

LROP

LPPT

LPPR

LPFR

LOWW

LMML

LKPR

LJLJ

LIRN

LIRF

LIRA

LIPZ

LIPE

LIML

LIMF

LIME

LIMC

LICC

LHBP

LGAV

LFSB

LFPO

LFPG

LFMN

LFML

LFLL

LFBO

LEZL

LEVC

LEPA

LEMG

LEMD

LEIB

LEBL

LEBB

LEAL

LDZA

LCLK

LBSF

GCXO

GCTS

GCRR

(movements/hour)

Impact Peak Service Rate

00.511.522.5

LZIB

LSZH

LSGG

LROP

LPPT

LPPR

LPFR

LOWW

LMML

LKPR

LJLJ

LIRN

LIRF

LIRA

LIPZ

LIPE

LIML

LIMF

LIME

LIMC

LICC

LHBP

LGAV

LFSB

LFPO

LFPG

LFMN

LFML

LFLL

LFBO

LEZL

LEVC

LEPA

LEMG

LEMD

LEIB

LEBL

LEBB

LEAL

LDZA

LCLK

LBSF

GCXO

GCTS

GCRR

(min/flight)

Impact Add. Taxi-Out Time Impact Add. ASMA Time

Impact of runway configurations usage on operational performance

24

5. CONCLUSIONS

This study presents a data-driven analysis of the airport capacity imbalance from a conceptual

definition to its implementation. It analyses the configurations of 90 European airports, based

on one year of data (2019) covering airport movements between 7h and 22h local time, and

the risk of capacity reduction associated to the change from one configuration to another,

together with the impact on throughput and performance. The main conclusions are

summarised as follows:

The followed data-driven approach provides a common framework for airport

configuration determination which ensures that the real airport operation is taken into

account rather than the declared or planned one. This allows an easy-to-maintain

methodology for the identification of the runway system configuration in use at each

airport at each time.

The runway configurations obtained via this data-driven methodology coincide or are very

similar to those published in the Airport Corner for the majority of the airports. That means

that the data-driven detected configurations with a minimum 3% probability are in the

majority of cases the same as the ones declared by the airports.

Regarding the identification of the capacity for each runway configuration, the peak service

rate is in some cases significantly lower than the declared capacities in the Airport Corner

That may be an indication that these airports still have some buffer for traffic increase,

while the others, where the peak service rate is close to the declared capacity, may be

already operating at their maximum capacity.

Resilience values are calculated based on the reference capacity (maximum peak service

rate of all available configurations at the airport), accounting for real operation aspects of

the year of data analysed (2019). This implies that the resilience obtained corresponds to

the time period analysed; whenever operation in the airport changes, resilience values will

also change to adapt to the new situation. Hence, the airport resilience is a dynamic

indicator and its evolution can be monitored.

The daily time-window covered in the analysed goes between 7h to 22h local time, which

ensures curfews (like for noise abatement procedures) are not taken into account in the

analysis since they are normally active in the in-between period. However, in some cases

specific noise abatement configurations might be activated outside of this time window,

with its consequent impact on the resilience calculation, as night configuration imply less

staff and normally also less demand in the airport.

Resilience results suggest there is no drastic decrease in capacity (in relative terms) related

to a runway configuration change at most airports. Whenever there is a significant

decrease in capacity, it occurs for a configuration with a very low probability, as such

limiting the impact on the airport’s resilience.

Most of the 90 airports analysed show very small changes in capacity for the most common

runway configurations, as derived from the Impact on peak service rate. There are some

of them, though, that show significantly less throughput and higher delays.

25

The impact on the additional taxi-out and ASMA times yields very interesting results and

evidences, at some airports, an important imbalance in terms of performance for different

runway configurations. These results highlight configurations that, beyond a potential

capacity issue, have intrinsic bottlenecks in both approach and taxi-out procedures.

The environmental constrains, the runway closures due to works in progress and many

other factors like implementation of new procedures have a clear impact in the identified

runway configurations and performance. The high level results are to be completed by the

analysis per airport and when important imbalances are detected, further breakdown per

month and/or hours of the day can bring additional insight.

In summary, this data driven analysis allows to study the imbalance of the operation

associated to the runway configurations use and its impact on performance, helping in the

identification of operational constraints at airports. This can significantly contribute to the

fine tuning of the process of assessing capacity limits specially delivered by slot coordinators

at the congested airports.

26

6. REFERENCES

[1] ICAO. Doc 9883 “Manual on Global Performance of the Air Navigation System”, 1st

Edition (2009)

[2] Malawko, Jan: “Airport capacity imbalance metrics”. Presentation to the

Performance Review Commission (2018)

[3] ACI. “Guide to Airport Performance Measures”, (2012).

[4] EUROCONTROL Airport Corner: https://www.eurocontrol.int/tool/airport-corner.

[5] Avery, Jacob, and Hamsa Balakrishnan. "Predicting airport runway configuration: A

discrete-choice modeling approach" (2015).

[6] Jones, James C., et al. "Predicting & quantifying risk in airport capacity profile

selection for air traffic management" 14th USA/Europe Air Traffic Management

Research and Development Seminar (ATM2017), Seattle, USA. (2017).

[7] U.S./Europe comparison of air traffic management-related operational performance

for 2015: https://www.eurocontrol.int/publication/useurope-comparison-air-traffic-

management-related-operational-performance-2015.

https://goto.agency-be.eurocontrol.int/tool/,DanaInfo=www.eurocontrol.int,SSL+airport-corner

https://goto.agency-be.eurocontrol.int/publication/,DanaInfo=www.eurocontrol.int,SSL+useurope-comparison-air-traffic-management-related-operational-performance-2015

https://goto.agency-be.eurocontrol.int/publication/,DanaInfo=www.eurocontrol.int,SSL+useurope-comparison-air-traffic-management-related-operational-performance-2015

27

7. ANNEX I: RESULTS SUMMARY

This annex compiles the results obtained for each of the 90 airports analysed, including the following

indicators:

Annual traffic 2019

Number of configurations

Impact Add. Taxi-Out Time (min/dep)

Impact Add. ASMA Time (min/arr)

Impact PSR2 (mov/hour)

Airport Resilience

A graph displaying these results for each configuration:

Configuration identification with active runways for arrivals (ARR)

and departures (DEP)

Probability

Peak Service Rate (P99) Total movements

Maximum Throughput Total movements

Peak Service Rate (P99) Arrivals

Maximum Throughput Arrivals

Peak Service Rate (P99) Departures

Maximum Throughput Departures

Average Additional ASMA time per arrival

Average Additional Taxi-out time per departure

The airport layouts are also included for a better understanding of the configurations and

identification of potential interdependencies in the use of runways.

This comprehensive overview allows to evaluate the main operating modes and differences in the

resulting performance.

2 PSR: Peak service rate

28

1. EBBR (BRU) – Brussels

The analysis of the operation during 2019 at Brussels shows 5 representative configurations, 4 of them

in segregated mode. The most common configuration does not show the highest peak service rate but

it does show the best arrival rate and low additional times compared to the rest. Configuration

ARR:01-DEP:07R (7%) causes the highest delays in both ASMA and taxi-out, with a lower peak service

rate.

Airport ICAO Airport Name

Annual traffic 2019

Number of configura-tions



Impact PSR (mov/h)

Airport Resilience

EBBR Brussels 229281 5 0.37 0.40 2.30 97.12%

58% 21% 7% 6% 3%

31

42

3538 37

41

46

39 40 3942

24

3533

31

50

31

36 3633

5760

55

59

54

65 6562

65

58

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0

10

20

30

40

50

60

70

ARR:25L - ARR:25R -DEP:25R

ARR:25L - DEP:25R ARR:01 - DEP:07R ARR:07L - DEP:07R ARR:19 - DEP:25R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EBBR (BRU) - Brussels

Add TXOT Add ASMA P99 DEP MAX DEP P99 ARR MAX ARR P99 TOTAL MAX TOTAL

29

2. EBCI (CRL) - Charleroi

Charleroi is a single runway airport with only two possible operating modes and a clear runway

direction preference. In general terms the performance in both runway configurations is very similar,

and only additional taxi-out when using RWY06 shows an increase compared to when using the

preferential runway direction (RWY24).


Annual traffic 2019


Impact Add. Taxi-Out Time


Impact Peak Service Rate (mov/h)

Airport Resilience

EBCI Charleroi 54763 2 0.03 0.01 0.87 96.23%

87% 13%

13 14

19

17

13 13

18

2322

23

32

28

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0

5

10

15

20

25

30

35

ARR:24 - DEP:24 ARR:06 - DEP:06

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EBCI (CRL) - Charleroi


30

3. EDDB (SXF) - Berlin/ Schoenefeld

Berlin Schoenefeld operates only one runway (07L/25R) in both directions mixed mode, with almost

identical result in performance.


Annual traffic 2019





Airport Resilience

EDDB Berlin/ Schoenefeld

90231 2 0.02 0.06 0.00 100.00%

71% 29%

14 14

20

18

14 14

19 19

25 25

31

29

0.0

0.5

1.0

1.5

2.0

2.5

0

5

10

15

20

25

30

35

ARR:25R - DEP:25R ARR:07L - DEP:07L

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EDDB (SXF) - Berlin/ Schoenefeld


31

4. EDDC (DRS) – Dresden

Dresden is a single runway airport, operated in mixed mode in both directions, being RWY22 the

preferential. Performance in terms of peak service rate is identical, and while RWY04 results in longer

additional ASMA times, RWY 22 shows higher additional taxi-out times.


Annual traffic 2019





Airport Resilience

EDDC Dresden 20524 2 0.06 0.03 0.00 100.00%

78%22%

6 6

7 7

6 6

7

8

9 9

13

10

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0

2

4

6

8

10

12

14

ARR:22 - DEP:22 ARR:04 - DEP:04

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EDDC (DRS) - Dresden


32

5. EDDE (ERF) – Erfurt

Erfurt operates one runway in mixed mode in both directions, being RWY28 the preferential. The peak

service rate for RWY28 is 2 movements higher than for RWY10, but the maximum throughput is

actually higher for RWY10, which leads to think the imbalance in peak service rate is driven by demand

and not capacity. In terms of performance, it is interesting to observe how the preferential runway

use results in higher additional times.


Annual traffic 2019





Airport Resilience

EDDE Erfurt 4607 2 0.22 0.25 0.53 97.33%

73% 27%

11

9

1213

10 10

1213

20

18

24

26

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0

5

10

15

20

25

30

ARR:28 - DEP:28 ARR:10 - DEP:10A

dd

itio

nal

tim

es (m

inu

tes/

fligh

t)

Mo

vem

ents

/ho

ur

EDDE (ERF) - Erfurt


33

6. EDDF (FRA) - Frankfurt

Frankfurt, the busiest airport in Europe, operates 4 runways, 3 of them in both directions and 1 (18/36)

only as RWY18. This results in many different configuration possibilities, from which the study

identified only 3 as representative, covering 97% of the operation. Out of these 3 configurations, 2 are

identical to those declared in the Airport Corner in 2019, and one is very similar. Peak service rate for

the three configurations is almost the same, resulting in very high resilience and low impact on the

PSR. Regarding performance, configuration ARR:25L-ARR:25R-DEP:18-DEP:25C, used 22% of the time,

shows the considerably worse additional times, especially in the taxi-out phase, with more than 5

min/dep.


Annual traffic 2019





Airport Resilience

EDDF Frankfurt 513866 3 0.66 0.33 0.58 99.68%

40% 35% 22%

6258 61

68 65 6765 6561

72 7166

110 109 109

121116 119

0.0

1.0

2.0

3.0

4.0

5.0

6.0

0

20

40

60

80

100

120

140

ARR:07L - ARR:07R - DEP:07C -DEP:18

ARR:25C - ARR:25L - ARR:25R -DEP:18 - DEP:25C

ARR:25L - ARR:25R - DEP:18 -DEP:25C

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EDDF (FRA) - Frankfurt


34

7. EDDG (FMO) - Muenster-Osnabrueck

Muenster has one runway (07/25) that can only be operated in mixed mode in both directions. PSR is

identical for both configurations and performance is very similar, with a small impact on the additional

ASMA time when using RWY07 (30% share). The big dispersion between the maximum observed

throughput and the PSR signals the airport might be underutilised and therefore the PSR would not

be a good proxy for capacity.


Annual traffic 2019





Airport Resilience

EDDG Muenster-Osnabrueck 18604 2 0.01 0.03 0.00 100.00%

70% 30%

7 7

1011

15 15

2322

19 19

28

26

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0

5

10

15

20

25

30


dd

itio

nal

tim

es (m

inu

tes/

fligh

t)

Mo

vem

ents

/ho

ur

EDDG (FMO) - Muenster-Osnabrueck


35

8. EDDH (HAM) - Hamburg

Hamburg has 2 crossing runways, resulting in 6 identified runway configurations. The PSR is similar for

all configurations (1 or 2 movements difference, so high resilience results), even when using only one

of the runways in mixed mode (ARR:23-DEP:23 or ARR:33-DEP:33). However, single runway

configurations show a clear impact on performance, with much higher additional times. The most

commonly used configuration offers the best results in performance and throughput.


Annual traffic 2019





Airport Resilience

EDDH Hamburg 149221 6 0.36 0.41 0.39 99.52%

46% 19% 13% 10% 5% 5%

22 22 22 22 21 21

2826 25 25 25 2423 22 22 22 21 21

28 27 27 27

23 23

40 39 40 4038 38

4846

4345

41 41

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

10

20

30

40

50

60

ARR:23 - DEP:33 ARR:15 - DEP:23 ARR:23 - DEP:23 ARR:05 - DEP:33 ARR:33 - DEP:33 ARR:15 - DEP:05A

dd

itio

nal

tim

es (m

inu

tes/

fligh

t)

Mo

vem

en

ts/h

ou

r

EDDH (HAM) - Hamburg


36

9. EDDK (CGN) - Cologne-Bonn

Köln has 3 runways, but for almost 80% of the time, the airport operates only one (14L/32R) in single

runway mixed mode, as the other 2 runways are much shorter and might not offer the same

navigational aids. The maximum throughput and PSR is reached though, when using also RWY 24 for

arrivals in addition to 14L in mixed mode. In terms of performance, there is an imbalance in the

additional times between configurations, and different impact depending on the indicator.


Annual traffic 2019





Airport Resilience

EDDK Cologne-Bonn 140929 3 0.03 0.19 2.00 95.90%

41% 38% 12%

15 1516

19 19 19

16 1618

20 2021

2728

30

3334

35

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0

5

10

15

20

25

30

35

40

ARR:14L - DEP:14L ARR:32R - DEP:32R ARR:14L - ARR:24 - DEP:14L

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EDDK (CGN) - Cologne-Bonn


37

10. EDDL (DUS) - Dusseldorf

Dusseldorf has two parallel runways that are sometimes used in segregated mode around 30 % of the

time. However 57% of the operation is handled in a single runway mixed mode, which results in much

lower PSR and maximum throughput, and also less efficient taxi-out times. Operation of both RWY

simultaneously is restricted to 56 hours per week due to court ruling, which results in a resilience that

is consequently lower than at other airports, and also a higher impact on the additional taxi-out times.

It is worth mentioning that works north of runway 05L/23R started in Summer 2019 impacting the

normal operation, as arrivals could not use this runway.


Annual traffic 2019





Airport Resilience

EDDL Dusseldorf 225541 5 0.41 0.15 6.37 91.30%

48% 25% 9% 5% 5%

28

35

28

3229

3438

31

37

31

25

32

25

31

2629

37

27

3228

45

54

44

55

4751

57

46

59

48

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0

10

20

30

40

50

60

70

ARR:23L - DEP:23L ARR:23R - DEP:23L ARR:05R - DEP:05R ARR:05L - DEP:05R ARR:05R - DEP:05R -DEP:23L

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EDDL (DUS) - Dusseldorf


38

11. EDDM (MUC) - Munich

Munich offers a perfect symmetrical layout with 2 parallel independent runways, both of them used

in mixed mode, with RWYs 26L/R used 61% of the time. The symmetry also applies to the results in

terms of PSR and maximum throughputs leading to high resilience. The observed performance,

although similar, is slightly better when using RWYs 08L/R


Annual traffic 2019





Airport Resilience

EDDM Munich 414222 2 0.23 0.11 0.37 99.59%

61% 37%

58 59

67 66

56 56

64 62

90 89

97 96

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0

20

40

60

80

100

120

ARR:26L - ARR:26R - DEP:26L - DEP:26R ARR:08L - ARR:08R - DEP:08L - DEP:08R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EDDM (MUC) - Munich


39

12. EDDN (NUE) - Nuremberg

Nuremberg is a single runway airport, used in mixed mode on both directions and with a big dispersion

between the maximum observed throughput and the PSR, which signals the airport might be

underutilised and therefore the PSR not be a good indication for capacity. In terms of performance,

there are almost no differences between runway configurations.


Annual traffic 2019





Airport Resilience

EDDN Nuremberg 49207 2 0.02 0.01 1.94 92.80%

65% 35%

13 14

19

32

1214

22 2324

27

40

43

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0

5

10

15

20

25

30

35

40

45

50

ARR:28 - DEP:28 ARR:10 - DEP:10

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EDDN (NUE) - Nuremberg


40

13. EDDP (LEJ) - Leipzig-Halle

Leipzig has 2 parallel runways and not that many movements for that infrastructure. The analysis

indicates that for 82% of the time, the airport is operated in a single runway mixed mode, and only

10% in segregated mode. In fact, significant lower throughput is observed for the segregated modes

and also significant worse additional taxi-out times for conf. ARR:08L-DEP:08R.


Annual traffic 2019





Airport Resilience

EDDP Leipzig-Halle 75413 6 1.02 0.27 1.31 96.68%

32% 26% 21% 6% 4% 3%

9

1110

87

10

16

1213

9

7

11

9

1110

810 10

11

1514

9

12

10

17

19 19

12

15

17

26

22

25

13

15

20

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0

5

10

15

20

25

30

ARR:26R -DEP:26R

ARR:26L -DEP:26L

ARR:08R -DEP:08R

ARR:08L -DEP:08R

ARR:26L -DEP:26R

ARR:08L -DEP:08L

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EDDP (LEJ) - Leipzig-Halle


41

14. EDDR (SCR) - Saarbruecken

Saarbruecken is a single runway airport, operated in mixed mode in both directions, being RWY27 the

preferential. Performance in terms of peak service rate is identical, but the performance for the

preferential runway in terms of additional times is considerably worse, resulting in a high impact on

both additional taxi-out and ASMA times.


Annual traffic 2019





Airport Resilience

EDDR Saarbruecken 7978 2 0.50 0.59 0.00 100.00%

73%26%

4 4

6

55

4

6

5

7 7

8

7

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0

1

2

3

4

5

6

7

8

9

ARR:27 - DEP:27 ARR:09 - DEP:09

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EDDR (SCR) - Saarbruecken


42

15. EDDS (STR) - Stuttgart

Stuttgart operates only one runway (07/25) in both directions mixed mode, with almost identical

result in performance. There is however a significant difference between the PSR and the maximum

throughput, which questions the validity of PSR as indication for capacity


Annual traffic 2019





Airport Resilience

EDDS Stuttgart 132609 2 0.01 0.07 0.62 98.40%

62% 37%

23 24

2931

22 23

2830

38 39

4749

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

10

20

30

40

50

60

ARR:25 - DEP:25 ARR:07 - DEP:07

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EDDS (STR) - Stuttgart


43

16. EDDT (TXL) - Berlin/ Tegel

Berlin Tegel has two parallel dependent runways operated in segregated mode, with very similar

results in terms of throughput and performance. The only impact is observed in the additional ASMA

times, higher for conf. ARR:08L-DEP:08R.


Annual traffic 2019





Airport Resilience

EDDT Berlin/ Tegel 191779 2 0.00 0.06 0.00 100.00%

71% 27%

24 24

29 2825 25

29 30

44 44

50 51

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0

10

20

30

40

50

60

ARR:26R - DEP:26L ARR:08L - DEP:08R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EDDT (TXL) - Berlin/ Tegel


44

17. EDDV (HAJ) - Hanover

Like Leipzig (EDDP), Hanover airport has not that many movements and despite having two parallel

runways, normally only one runway in mixed mode is operated (81% of the time). PSR shows only 1

or 2 movements difference for the most common configurations, and in fact it is lowest for the

segregated conf ARR:09L-DEP:09R (8% share). There is some imbalance in performance, especially

concerning the additional taxi-out times when using RWY 09L.


Annual traffic 2019





Airport Resilience

EDDV Hanover 64581 5 0.21 0.05 1.24 97.82%

39% 30% 9% 8% 3%

12 1113

79

17

15

13

8

10

12 12 12

109

17 1716

11

9

2122

20

16 16

26

29

22

1617

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

5

10

15

20

25

30

35

ARR:27R - DEP:27R ARR:27L - DEP:27L ARR:09R - DEP:09R ARR:09L - DEP:09R ARR:09L - DEP:09L

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EDDV (HAJ) - Hanover


45

18. EDDW (BRE) - Bremen

Bremen operates only one runway (09/27) in both directions mixed mode, with identical result in

performance. In terms of PSR the impact is very low (only 1 movement difference) but the big

difference between the maximum observed throughput and the PSR signals the airport might be

underutilised which makes the PSR not a good proxy for capacity


Annual traffic 2019





Airport Resilience

EDDW Bremen 29375 2 0.00 0.00 0.33 98.27%

67% 33%

10 10

19

1311 11

1715

1918

35

24

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0

5

10

15

20

25

30

35

40

ARR:27 - DEP:27 ARR:09 - DEP:09

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EDDW (BRE) - Bremen


46

19. EETN (TLL) - Tallinn

Tallinn is a one runway airport with a layout that clearly favours one runway direction utilization. In

fact, RWY26 is used (in mixed mode) 95% of the time in 2019. The other runway configuration is not

used more than 3% of the time (with some time intervals being the transition between configurations

or missing runway information) so it is not considered representative and there is no imbalance to be

analysed.


Annual traffic 2019





Airport Resilience

EETN Tallinn 43904 1 0.00 0.00 0.00 100.00%

95%

9

16

9

14 14

21

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0

5

10

15

20

25

ARR:26 - DEP:26

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EETN (TLL) - Tallinn


47

20. EFHK (HEL) - Helsinki/ Vantaa

Helsinki, with 3 runways (2 parallel plus one crossing) is operated in 6 representative configurations,

mostly in segregated mode. The results show significant differences in both throughput and

performance, where some configurations are clearly indicated for departure peaks and others for

arrival peaks. The 3 most common configurations are used in the departure peaks and result in lower

total throughput, higher additional taxi-out times and lower additional ASMA times. The two

configurations for the arrival peaks allow for higher total throughput but naturally increased additional

ASMA times. The most efficient configuration seems to be ARR:15-DEP22L (even if those are crossing

runways), when the number of departures and arrivals is balanced, but it is in use only 8% of the time.


Annual traffic 2019





Airport Resilience

EFHK Helsinki/ Vantaa 194634 6 1.00 0.26 6.77 95.84%

25% 24% 16% 8% 6% 5%

36 3733

37

1921

40 4139

41

22 22

29

22 21

38

4649

3633

24

40

5053

48 4844

5753

5760

58

50

62

57 58

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0

10

20

30

40

50

60

70

ARR:22L -DEP:22R

ARR:04L -DEP:04R

ARR:15 - DEP:22R ARR:15 - DEP:22L ARR:04L -ARR:04R -DEP:04R

ARR:22L -ARR:22R -DEP:22R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EFHK (HEL) - Helsinki/ Vantaa


48

21. EGBB (BHX) - Birmingham

Birmingham only has one runway that can be operated in both directions in mixed mode. The

performance for both runway directions use is very similar in throughput and additional taxi-out time,

and with higher additional ASMA for the most frequent configuration (RWY33). However, the analysis

also identifies as representative a configuration that implies opposite use of the runway: ARR:15-

DEP:33 (3% share). Normally this opposite use is observed of the periods of transition between runway

configurations, but the share is always much lower in that case. Another surprising aspect is that the

total throughput is almost the same for this configuration, and only additional taxi out time suffers a

considerable increase. This particular configuration would require further detailed analysis.


Annual traffic 2019





Airport Resilience

EGBB Birmingham 107768 3 0.07 0.32 0.49 98.42%

49% 46% 3%

18 1819

2122 22

16 16 16

2021

20

2928 28

3635 35

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0

5

10

15

20

25

30

35

40

ARR:33 - DEP:33 ARR:15 - DEP:15 ARR:15 - DEP:33

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EGBB (BHX) - Birmingham


49

22. EGCC (MAN) - Manchester

Manchester, with 2 parallel dependent runways is operated in 4 configurations: 2 segregated (63%

total share) and 2 using only one runway in mixed mode (35% total share). The PSR is clearly lower for

the single runway use, and although performance is not much worse, maybe a segregated use in those

periods (if possible) could have reduced the delays.


Annual traffic 2019





Airport Resilience

EGCC Manchester 202935 4 1.01 0.21 3.82 94.32%

50% 29% 13% 6%

30

24

29

22

35 36 35

2526

2225

20

3128 29

23

51

41

49

38

57

63

59

40

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0

10

20

30

40

50

60

70

ARR:23R - DEP:23L ARR:23R - DEP:23R ARR:05R - DEP:05L ARR:05L - DEP:05L

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EGCC (MAN) - Manchester


50

23. EGGD (BRS) - Bristol

Bristol is a single runway airport, operated in mixed mode in both directions, being RWY27 the

preferential. Performance in terms of peak service rate differs in only 1 movement, and only additional

taxi-out times show worse performance for the non-preferential runway use.


Annual traffic 2019





Airport Resilience

EGGD Bristol 66393 2 0.21 0.00 0.69 96.55%

69% 31%

12 12

1615

12 12

15 15

1920

2425

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0

5

10

15

20

25

30

ARR:27 - DEP:27 ARR:09 - DEP:09

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EGGD (BRS) - Bristol


51

24. EGGW (LTN) - London/ Luton

Luton has only one runway (08/26) that operates in both directions mixed mode, with identical results

in terms of PSR. The preferential runway (RWY26; 71% share) also has the lower additional times. In

general terms, no important imbalance is observed at this airport.


Annual traffic 2019





Airport Resilience

EGGW London/ Luton 140958 2 0.13 0.16 0.00 100.00%

71% 29%

21 21

26 26

19 19

23 22

34 34

40 39

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0

5

10

15

20

25

30

35

40

45

ARR:26 - DEP:26 ARR:08 - DEP:08

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EGGW (LTN) - London/ Luton


52

25. EGKK (LGW) - London/ Gatwick

Gatwick is the busiest airport in Europe that is always operated with one runway in single mode. Given

the saturation, the PSR is a very good indication for capacity, and shows perfect symmetry in the

operation in terms of throughput. Additional taxi-out times however are impacted by the runway use,

being more than a minute higher for the most common runway configuration (RWY26L).


Annual traffic 2019





Airport Resilience

EGKK London/ Gatwick 284916 2 0.93 0.03 0.00 100.00%

69% 31%

33 34

39 38

28 28

3331

55 5558 59

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0

10

20

30

40

50

60

70

ARR:26L - DEP:26L ARR:08R - DEP:08R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EGKK (LGW) - London/ Gatwick


53

26. EGLC (LCY) - London/ City

London City, with only one runway operated in both directions in mixed mode, shows nearly

symmetric PSR. Additional taxi-out time though, are considerably higher for most common runway in

use.


Annual traffic 2019





Airport Resilience

EGLC London/ City 84208 2 0.93 0.00 0.31 99.03%

69% 31%

18 18

2221

17 17

2019

3231

3534

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0

5

10

15

20

25

30

35

40

ARR:27 - DEP:27 ARR:09 - DEP:09

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EGLC (LCY) - London/ City


54

27. EGLL (LHR) - London/ Heathrow

London Heathrow has two parallel runways operated always in segregated mode. The most common

runway direction is 27 alternating the arrivals and departures on 27R and 27L. While there is a perfect

symmetry in the PSR, performance results show very similar additional times, with additional taxi-out

times a little bit more impacted by the change of runway configuration.


Annual traffic 2019





Airport Resilience

EGLL London/ Heathrow 478081 3 0.45 0.16 0.00 100.00%

36% 35% 25%

50 49 4953 54 54

45 44 4548 48 47

91 91 91

98 98 97

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0

20

40

60

80

100

120

ARR:27L - DEP:27R ARR:27R - DEP:27L ARR:09L - DEP:09R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EGLL (LHR) - London/ Heathrow


55

28. EGNT (NCL) - Newcastle

Newcastle is a single runway airport, operated in mixed mode in both directions, being RWY25 the

preferential. Performance in terms of peak service rate differs in only 1 movement, and only additional

taxi-out times show worse performance for the non-preferential runway use. The big difference

between the maximum observed throughput and the PSR for the preferential runway use signals the

airport might be underutilised which makes the PSR not a good proxy for capacity


Annual traffic 2019





Airport Resilience

EGNT Newcastle 43438 2 0.08 0.03 0.72 95.74%

72% 27%

1011

17

14

10 10

20

1516

17

25

21

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0

5

10

15

20

25

30

ARR:25 - DEP:25 ARR:07 - DEP:07

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EGNT (NCL) - Newcastle


56

29. EGPD (ABZ) - Aberdeen

Aberdeen is a single runway airport (plus three helipads), operated in mixed mode in both directions,

with a very even use of both runway directions (55%-45%). Performance in terms of peak service rate

is identical for both configurations, but runway 34 shows longer additional times, mainly in the taxi-

out phase.


Annual traffic 2019





Airport Resilience

EGPD Aberdeen 51233 2 0.23 0.13 0.00 100.00%

55% 45%

10 10

1312

10 10

15 15

17 17

2223

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

5

10

15

20

25

ARR:16 - DEP:16 ARR:34 - DEP:34

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EGPD (ABZ) - Aberdeen


57

30. EGPF (GLA) - Glasgow

Glasgow only has one runway that can be operated in both directions in mixed mode. The

performance for both runway directions use is similar in throughput (2 movements less for the

secondary runway use) showing higher taxi-out times for RWY 23 and higher additional ASMA times

for RWY 05. Nevertheless there is still a significant difference between the maximum throughput and

the PSR that might suggest the percentile 99 is not the best proxy for capacity.


Annual traffic 2019





Airport Resilience

EGPF Glasgow 84345 2 0.16 0.04 0.54 97.86%

73% 27%

1615

22 22

14 14

1918

25

23

3130

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0

5

10

15

20

25

30

35

ARR:23 - DEP:23 ARR:05 - DEP:05

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EGPF (GLA) - Glasgow


58

31. EGPH (EDI) - Edinburgh

Edinburgh is a single runway airport, operated in mixed mode in both directions, with almost identical

performance in terms of peak service rate. Additional times are slightly higher for the non-preferential

runway use.


Annual traffic 2019





Airport Resilience

EGPH Edinburgh 131457 2 0.08 0.20 0.00 100.00%

68% 32%

1920

24 24

18 18

21 21

32 32

3635

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0

5

10

15

20

25

30

35

40

ARR:24 - DEP:24 ARR:06 - DEP:06

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EGPH (EDI) - Edinburgh


59

32. EGSS (STN) - London/ Stansted

Stansted is one of the top 30 airports in Europe, and it is a single runway airport. The runway is

operated in both directions in mixed mode, being RWY22 the most commonly used (71% share)

Performance in terms of throughput is completely symmetric, and additional times are very similar,

although the additional taxi out times are slightly higher for the preferential runway.


Annual traffic 2019





Airport Resilience

EGSS London/ Stansted 198511 2 0.33 0.04 0.00 100.00%

71% 29%

29 29

36 36

25 25

30 29

47 47

5351

0.0

1.0

2.0

3.0

4.0

5.0

6.0

0

10

20

30

40

50

60

ARR:22 - DEP:22 ARR:04 - DEP:04

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EGSS (STN) - London/ Stansted


60

33. EHAM (AMS) - Amsterdam/ Schiphol

Amsterdam has 6 runways (more than any other airport in Europe) translating into many different

possible configurations. In addition, the 15 minutes intervals might not be long enough for all active

runways to be used in the interval. Nevertheless, the methodology detects all the 8 runway

configurations declared in the Airport Corner, and 4 additional ones. The usage of the configurations

is quite distributed, not having a clear preferential one. Although the PSR of total movements is quite

similar for all configurations, looking at the PSR for arrivals or departures it can be easily deducted

that the configurations are used either for arrival peaks or departure peaks. In terms of performance,

there are significant differences in both additional ASMA and additional taxi-out times amongst most

of the configurations, with configuration ARR:36C;ARR:36R-DEP:36L showing the worst results (more

than double compared to other configurations).


Annual traffic 2019





Airport Resilience

EHAM Amsterdam/ Schiphol 509185 12 1.00 0.96 2.84 99.22%

12% 12% 8% 6% 6% 5% 4% 4% 4% 4% 4% 3%

40

78

40 40

79

42 42

75

46

77 7673

45

80

4441

82

44 43

76

46

78 77 7571

39

73 73

42

70 71

41

73

38 38 39

73

42

77 75

43

71 72

43

75

4138

43

102106 107 104

109104 104 106 107 108

103 103108

111 111 109112

107 107 106111 110

104 104

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

0

20

40

60

80

100

120

ARR:18C -ARR:18R -DEP:18L

ARR:18R -DEP:18L -

DEP:24

ARR:06 -ARR:36R -DEP:36L

ARR:18C -ARR:18R -

DEP:24

ARR:18R -DEP:18C -DEP:18L

ARR:36C -ARR:36R -DEP:36L

ARR:18R -ARR:22 -DEP:24

ARR:36R -DEP:36C -DEP:36L

ARR:27 -ARR:36C -DEP:36L

ARR:06 -DEP:36C -DEP:36L

ARR:27 -DEP:24 -DEP:36L

ARR:06 -DEP:09 -DEP:36L

Ad

dit

ion

al t

ime

s (m

inu

tes/

fligh

t)

Mo

vem

ents

/ho

ur

EHAM (AMS) - Amsterdam/ Schiphol


61

34. EICK (ORK) - Cork

Cork has two crossing runways but it only operates the longest one in mixed mode, with a very even

use of both runway directions (54%-45%). Performance in terms of peak service rate is almost identical

for both configurations, but RWY16 shows longer additional times, both in the approach and the taxi-

out phase.


Annual traffic 2019





Airport Resilience

EICK Cork 27002 2 0.10 0.26 0.52 97.90%

54% 45%

13 13

2018

13 13

1817

2425

3533

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0

5

10

15

20

25

30

35

40

ARR:16 - DEP:16 ARR:34 - DEP:34

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EICK (ORK) - Cork


62

35. EIDW (DUB) - Dublin

Dublin has two runways but operates only one of them in both directions mixed mode, with RWY28

as preferential (78% share). The PSR is almost identical for both configurations, but there is a clear

impact on performance when using RWY10, both in additional taxi-out times (bottleneck in the

departure queue for RWY10) and in ASMA times.


Annual traffic 2019





Airport Resilience

EIDW Dublin 238044 2 0.42 0.47 0.20 99.59%

78% 20%

29 30

37 36

25 25

2927

49 48

5250

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

0

10

20

30

40

50

60

ARR:28 - DEP:28 ARR:10 - DEP:10

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

EIDW (DUB) - Dublin


63

36. EKCH (CPH) - Copenhagen/ Kastrup

Copenhagen has 2 parallel runways and one crossing, but the crossing runway 12/30 is rarely used.

The two representative configurations use the parallel runways in segregated mode, with identical

total PSR. The preferential runway use ARR:22L-DEP:22R (73% share) shows slightly higher additional

times, both in the approach and the taxi-out phase.


Annual traffic 2019





Airport Resilience

EKCH Copenhagen/ Kastrup 263434 2 0.15 0.25 0.00 100.00%

73% 21%

36 37

43 43

3533

4239

61 61

6966

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

10

20

30

40

50

60

70

80

ARR:22L - DEP:22R ARR:04L - DEP:04R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EKCH (CPH) - Copenhagen/ Kastrup


64

37. ELLX (LUX) - Luxembourg

Luxembourg is a single runway airport, operated in mixed mode in both directions, being RWY24 the

most commonly used (68% share). Performance in terms of peak service rate differs in only 1

movement, and only additional taxi-out times show slightly worse performance for the preferential

runway use. The big difference between the maximum observed throughput and the PSR for the

preferential runway use signals the airport might be underutilised which makes the PSR not a good

proxy for capacity.


Annual traffic 2019





Airport Resilience

ELLX Luxembourg 76300 2 0.20 0.01 0.68 97.50%

68% 32%

1516

2122

15 15

2019

2627

3634

0.0

0.5

1.0

1.5

2.0

2.5

0

5

10

15

20

25

30

35

40


dd

itio

nal

tim

es (m

inu

tes/

fligh

t)

Mo

vem

ents

/ho

ur

ELLX (LUX) - Luxembourg


65

38. ENGM (OSL) - Oslo/ Gardermoen

Oslo has two parallel independent runways and operates them in many different ways, resulting in 8

configurations. About 42% of the time the runways are used in mixed mode independent, resulting in

the highest throughput and lower delays in the taxi-out and approach phases, but the other

configurations, using the runways as segregated, show lower PSR and especially much higher

additional taxi-out times (up to 7 minutes higher than other configurations for conf ARR:01R-DEP:01L)

This results in the highest impact on the additional taxi-out times of all 90 airports analysed. For about

9% of the time the airport also operates only RWY01L/19R (as single runway mixed mode), reducing

the PSR by more than 30 movements with respect to the reference capacity. Although this could be

due to a lack of demand, these configurations also observe worse additional taxi-out times than the

independent use of both runways.


Annual traffic 2019





Airport Resilience

ENGM Oslo/ Gardermoen 251872 8 2.13 0.23 6.95 97.49%

23% 19% 18% 11% 8% 6% 5% 3%

40 4037 37

34

26

34

25

4744

41 43

3632

39

27

36 34 33 3430

22

3024

39 3835 35 34

26

3328

71 6965 64

57

41

55

38

7975 76

71

61

46

57

43

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

0

10

20

30

40

50

60

70

80

90

ARR:01L -ARR:01R -DEP:01L -DEP:01R

ARR:19L -ARR:19R -DEP:19L -DEP:19R

ARR:01R -DEP:01L

ARR:19R -DEP:19L

ARR:19R -DEP:19L -DEP:19R

ARR:19R -DEP:19R

ARR:01L -DEP:01L -DEP:01R

ARR:01L -DEP:01L

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

ENGM (OSL) - Oslo/ Gardermoen


66

39. EPWA (WAW) - Warszawa/ Chopina

Warsaw has 2 crossing runways operated in 5 different configurations. The high saturation level

ensures the PSR is a good proxy for capacity. The configuration that shows the best PSR (ARR:33-

DEP:29-DEP:33) is used only 3% of the time. In 2019, for about 40% of the time only RWY15/33 was in

use in mixed mode due to works on the other runway and although the impact on PSR was not very

high, the additional times were significantly longer, especially when using RWY15.


Annual traffic 2019





Airport Resilience

EPWA Warszawa/ Chopina 194160 5 0.70 0.61 3.40 97.42%

34% 29% 22% 11% 3%

27 27 28 2730

32 3134

3234

27 26 2724

26

32 3230

2628

42 41 4240

4546 47 47

42

48

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0

10

20

30

40

50

60

ARR:33 - DEP:29 ARR:33 - DEP:33 ARR:11 - DEP:15 ARR:15 - DEP:15 ARR:33 - DEP:29 -DEP:33

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EPWA (WAW) - Warszawa/ Chopina


67

40. ESGG (GOT) - Göteborg

Göteborg is a single runway airport, operated in mixed mode in both directions, being RWY21 the

most commonly used (70% share). Performance in terms of peak service rate differs in only 1

movement, and additional times show slightly worse performance for the non-preferential runway

use.


Annual traffic 2019





Airport Resilience

ESGG Göteborg 69265 2 0.08 0.07 0.70 96.50%

70% 30%

12 12

17

15

12 12

16 16

1920

2524

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0

5

10

15

20

25

30

ARR:21 - DEP:21 ARR:03 - DEP:03

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

ESGG (GOT) - Göteborg


68

41. ESSA (ARN) - Stockholm/ Arlanda

Stockholm has three runways, two of them parallel. The analysis yields a maximum of two runways

are used at a time, always in segregated mode. Furthermore, for a 4% of the time only RWY01R is

used, in mixed mode, and this results in a significantly reduced PSR with consequent much higher

additional times. Even when using two runways there is a considerable lower PSR for configuration

ARR:26-DEP19L. Nevertheless, this configuration shows the lowest additional taxi-out times and

relatively low additional ASMA times, indicating that this configuration might be simply used in periods

with lower demand. In general the difference between the maximum throughput and the PSR signals

that congestion at Stockholm might not be enough to consider the PSR a good proxy for capacity.


Annual traffic 2019





Airport Resilience

ESSA Stockholm/ Arlanda 233007 7 0.42 0.67 3.67 98.19%

24% 18% 18% 12% 9% 7% 4%

37 3835

32 32 3228

42 42 41

3537 37

3134 35

3230 30

32

24

38 3836

3235

37

25

6062

58

53

58 57

47

6770

6361

6361

48

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0

10

20

30

40

50

60

70

80

ARR:19L -DEP:19R

ARR:01R -DEP:01L

ARR:26 -DEP:19R

ARR:26 -DEP:19L

ARR:01L -DEP:08

ARR:19R -DEP:08

ARR:01R -DEP:01R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

ESSA (ARN) - Stockholm/ Arlanda


69

42. EVRA (RIX) - Riga

Riga is a single runway airport, operated in mixed mode in both directions, with almost identical

performance in terms of peak service rate. Additional times are slightly higher for the non-preferential

runway use.


Annual traffic 2019





Airport Resilience

EVRA Riga 86646 2 0.03 0.03 0.00 100.00%

.

63% 34%

20

18

27

24

17 17

21 21

24 24

30

28

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

5

10

15

20

25

30

35

ARR:18 - DEP:18 ARR:36 - DEP:36

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EVRA (RIX) - Riga


70

43. EYVI (VNO) - Vilnius

Vilnius, with only one runway operated in both directions in mixed mode, shows very similar PSR (only

one movement difference) and additional ASMA times. Additional taxi-out time though, are higher for

RWY01.


Annual traffic 2019





Airport Resilience

EYVI Vilnius 46775 2 0.13 0.02 0.66 95.59%

66% 31%

8 8

16

109 9

1312

1415

1920

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0

5

10

15

20

25

ARR:19 - DEP:19 ARR:01 - DEP:01

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

EYVI (VNO) - Vilnius


71

44. GCFV (FUE) - Fuerteventura

Fuerteventura is a one runway airport with conditions that clearly favour one runway direction

utilization. In fact, RWY01 is used (in mixed mode) 97% of the time in 2019. The other runway

configuration is not used more than 3% of the time (with some time intervals being the transition

between configurations or missing runway information) so it is not considered representative and

there is no imbalance to be analysed.


Annual traffic 2019





Airport Resilience

GCFV Fuerteventura 46356 1 0.00 0.00 0.00 100.00%

97%

12

16

12

16

21

28

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0

5

10

15

20

25

30

ARR:01 - DEP:01

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

GCFV (FUE) - Fuerteventura


72

45. GCLP (LPA) - Gran Canaria

Gran Canaria has two dependent parallel runways. For 56% of the time, only one of these two runways

are in use, in mixed mode, showing a much lower throughput than the combined use: conf ARR:03L-

DEP:03L;DEP:03R. This configuration shows the highest throughput but also the highest additional

taxi-out times, which might indicate is used for demand peaks.


Annual traffic 2019





Airport Resilience

GCLP Gran Canaria 124152 3 0.34 0.16 6.77 84.15%

53% 40% 3%

17

23

14

20

27

1416

21

17

21

27

19

28

40

27

36

45

29

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

5

10

15

20

25

30

35

40

45

50

ARR:03L - DEP:03L ARR:03L - DEP:03L - DEP:03R ARR:03R - DEP:03R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

GCLP (LPA) - Gran Canaria


73

46. GCRR (ACE) - Lanzarote

Lanzarote is a one runway airport with conditions that clearly favour one runway direction utilization.

In fact, RWY03 is used (in mixed mode) 97% of the time in 2019. The other runway configuration is

not used more than 3% of the time (with some time intervals being the transition between

configurations or missing runway information) so it is not considered representative and there is no

imbalance to be analysed.


Annual traffic 2019





Airport Resilience

GCRR Lanzarote 59130 1 0.00 0.00 0.00 100.00%

97%

12

18

13

16

22

26

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0

5

10

15

20

25

30

ARR:03 - DEP:03

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

GCRR (ACE) - Lanzarote


74

47. GCTS (TFS) - Tenerife Sur/Reina Sofia

Tenerife Sur has only one runway operated in both directions in mixed mode and conditions that

clearly favour one runway direction utilization (RWY07:95% share). The results shows very similar PSR

(only one movement difference) but additional times are higher for RWY25, although with low impact

due to the share (5%).


Annual traffic 2019





Airport Resilience

GCTS Tenerife Sur/Reina Sofia 68636 2 0.03 0.01 0.95 96.72%

95%5%

1617

20 20

1517

19 19

2829

3332

0.0

0.5

1.0

1.5

2.0

2.5

0

5

10

15

20

25

30

35

ARR:07 - DEP:07 ARR:25 - DEP:25

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

GCTS (TFS) - Tenerife Sur/Reina Sofia


75

48. GCXO (TFN) - Tenerife North

Tenerife North is a single runway airport operated in mixed mode in both directions, being RWY30 the

most commonly used (79% share). The PSR is identical for both configurations but additional times

are slightly higher for the non-preferential runway use.


Annual traffic 2019





Airport Resilience

GCXO Tenerife North 72597 2 0.01 0.04 0.00 100.00%

79% 21%

13 13

1516

12 12

15 15

21 21

2625

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0

5

10

15

20

25

30

ARR:30 - DEP:30 ARR:12 - DEP:12

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

GCXO (TFN) - Tenerife North


76

49. LBSF (SOF) - Sofia

Sofia operates only one runway (09/27) in both directions mixed mode, with almost identical result in

PSR. In terms of performance a slight increase of additional taxi-out times can be observed when using

RWY27 (64% share).


Annual traffic 2019





Airport Resilience

LBSF Sofia 60266 2 0.12 0.01 0.00 100.00%

64% 35%

10 10

1413

1110

1514

17 17

23

21

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0

5

10

15

20

25

ARR:27 - DEP:27 ARR:09 - DEP:09

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LBSF (SOF) - Sofia


77

50. LCLK (LCA) - Larnaca

The data provided by Larnaca airport does not allow for the calculation of the performance indicators

(additional times) and the runway information is missing for 39% of the time, so the analysis in this

case cannot be performed.


Annual traffic 2019





Airport Resilience

LCLK Larnaca 60657 3 0.81 97.78%

44% 39% 14%

1112

10

13

21

1312 12

9

14

16

13

1920

17

2324

20

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0

5

10

15

20

25

30

ARR:22 - DEP:22 ARR:NA - DEP:NA ARR:04 - DEP:04

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LCLK (LCA) - Larnaca


78

51. LDZA (ZAG) - Zagreb

Zagreb is a single runway airport operated in mixed mode in both directions, being RWY05R the most

commonly used in 2019 (71% share). The PSR is identical for both configurations but additional times

are slightly higher for the non-preferential runway use.


Annual traffic 2019





Airport Resilience

LDZA Zagreb 44307 2 0.03 0.07 0.00 100.00%

71% 29%

10 10

13 13

9 9

1211

14 14

18 18

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0

2

4

6

8

10

12

14

16

18

20


Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LDZA (ZAG) - Zagreb


79

52. LEAL (ALC) - Alicante

Alicante has one runway (10/28) that can only be operated in mixed mode in both directions. PSR is 2

movements higher for RWY10 (67% share) and performance in terms of additional taxi-out times is

very similar, Additional ASMA time when using RWY28 (33% share) are slightly higher.


Annual traffic 2019





Airport Resilience

LEAL Alicante 101210 2 0.01 0.13 0.65 97.89%

67% 33%

1817

23

2018

17

23

20

3129

37

34

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0

5

10

15

20

25

30

35

40

ARR:10 - DEP:10 ARR:28 - DEP:28

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LEAL (ALC) - Alicante


80

53. LEBB (BIO) - Bilbao

Bilbao has two runways but it normally uses only the longest one, operating it in mixed mode in both

directions. RWY30 is in use most of the time (81% share). Performance in terms of peak service rate

differs in only 1 movement, and RWY30 observed higher additional ASMA times while RWY12 shows

higher additional taxi-out times.


Annual traffic 2019





Airport Resilience

LEBB Bilbao 48984 2 0.01 0.07 0.19 98.74%

81%19%

109

1312

9 9

1211

1514

1918

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0

2

4

6

8

10

12

14

16

18

20

ARR:30 - DEP:30 ARR:12 - DEP:12

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LEBB (BIO) - Bilbao


81

54. LEBL (BCN) - Barcelona

Barcelona has two parallel runways and one crossing. The crossing runway is normally used only

sometimes before 10h and after 19h, not reaching the 3% threshold of representative configurations

for the period analysed. The most common use is in segregated mode with a clear preference for

direction 25 (ARR:25R-DEP:25L; 81% share). This configuration shows a PSR one movement lower than

the opposite direction, but higher maximum throughput and lower additional times both in the

approach and in the taxi-out phase.


Annual traffic 2019





Airport Resilience

LEBL Barcelona 344508 2 0.10 0.15 0.81 98.87%

81% 17%

40 3944 42

38 38

44 42

71 72

7975

0.0

1.0

2.0

3.0

4.0

5.0

6.0

0

10

20

30

40

50

60

70

80

90

ARR:25R - DEP:25L ARR:07L - DEP:07RA

dd

itio

nal

tim

es (m

inu

tes/

fligh

t)

Mo

vem

en

ts/h

ou

r

LEBL (BCN) - Barcelona


82

55. LEIB (IBZ) - Ibiza

Ibiza is a single runway airport, operated in mixed mode and using equally both runway directions

(51%-49%). Performance in terms of peak service rate differs in 2 movements, and although RWY06

shows the highest throughputs, it also shows higher additional ASMA and taxi-out times, which leads

to think that it is the configuration that was used when there was higher demand.


Annual traffic 2019





Airport Resilience

LEIB Ibiza 73356 2 0.29 0.15 1.02 96.71%

51% 49%

1617

22 22

1617

1921

2931

33

37

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

5

10

15

20

25

30

35

40

ARR:24 - DEP:24 ARR:06 - DEP:06

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LEIB (IBZ) - Ibiza


83

56. LEMD (MAD) - Madrid/ Barajas

Madrid has four runways, parallel two by two and operated in independent mode. The North

configuration is the preferred one (ARR:32L;ARR:32R-DEP36L;DEP36R; 69% share). The PSR differs

between configurations in one movement, but the significant difference between the maximum

throughput and the PSR signals the capacity might be higher than the percentile 99. In terms of

additional times, the best combined performance is found for the preferred configuration, having the

South configuration the worst additional taxi-out times and the configuration ARR:32L-

DEP36L;DEP36R almost triple the additional ASMA times.


Annual traffic 2019





Airport Resilience

LEMD Madrid/ Barajas 426185 3 0.27 0.24 0.73 99.23%

69% 21% 4%

58 59 6167

7067

49 48

38

53 53

40

88 89 88

104100

93

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0

20

40

60

80

100

120

ARR:32L - ARR:32R - DEP:36L -DEP:36R

ARR:18L - ARR:18R - DEP:14L -DEP:14R

ARR:32L - DEP:36L - DEP:36RA

dd

itio

nal

tim

es (m

inu

tes/

fligh

t)

Mo

vem

en

ts/h

ou

r

LEMD (MAD) - Madrid/ Barajas


84

57. LEMG (AGP) - Málaga

Málaga has two non-parallel runways, but for 81% of the time the airport is operated as a single

runway in mixed mode using only RWY13/31. Being a seasonal airport is clear that for most of the

time the single runway use does not pose a problem, and although for these configurations the

throughput is much lower, it is probably just due to the lack of demand. For the segregated use

configurations, there are two (ARR:12-DEP:13 and ARR:31-DEP30) that have the highest peak service

rates (only one movement difference) but a significant imbalance in terms of additional times.


Annual traffic 2019





Airport Resilience

LEMG Málaga 140721 5 0.93 0.42 10.86 85.10%

58% 23% 11% 3% 3%

2118

27

222526

22

30

24

28

2119

26

212425 25

28

2325

3532

47

38

4643

40

53

41

49

0.0

1.0

2.0

3.0

4.0

5.0

6.0

0

10

20

30

40

50

60

ARR:13 - DEP:13 ARR:31 - DEP:31 ARR:12 - DEP:13 ARR:31 - DEP:30 -DEP:31

ARR:31 - DEP:30

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LEMG (AGP) - Málaga


85

58. LEPA (PMI) - Palma de Mallorca

Palma de Mallorca has two parallel runways most generally used in segregated mode, being direction

24 the preferred one (73% share). The PSR is almost identical for both configurations but performance

in terms of additional times is clearly better for configuration ARR:24L-DEP:24R, especially in the

approach. Being the most seasonal airport in the study, it would be interesting to analyse the PSR only

for the busiest months, which would probably be a better indication of the actual capacity.


Annual traffic 2019





Airport Resilience

LEPA Palma de Mallorca 217096 2 0.05 0.23 0.00 100.00%

73% 24%

35 34

4240

34 3539 38

65 65

73 72

0.0

0.5

1.0

1.5

2.0

2.5

0

10

20

30

40

50

60

70

80

ARR:24L - DEP:24R ARR:06L - DEP:06R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LEPA (PMI) - Palma de Mallorca


86

59. LEVC (VLC) - Valencia

Valencia is a single runway airport, operated in mixed mode and using almost equally both runway

directions (52%-48%). Performance in terms of peak service rate differs in one movement, and

although RWY12 shows the highest throughputs, it also shows higher additional ASMA and taxi-out

times, which leads to think that it is the configuration that was used when there was higher demand.


Annual traffic 2019





Airport Resilience

LEVC Valencia 72464 2 0.13 0.15 0.52 97.39%

52% 48%

12 12

15

17

1112

1415

1920

23

26

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0

5

10

15

20

25

30

ARR:30 - DEP:30 ARR:12 - DEP:12

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LEVC (VLC) - Valencia


87

60. LEZL (SVQ) - Sevilla

Sevilla is a single runway airport, operated in mixed mode in both directions. The peak service rate for

both runway directions differs in one movement, but performance in terms of additional times is

nearly identical, showing no real imbalance.


Annual traffic 2019





Airport Resilience

LEZL Sevilla 58721 2 0.02 0.01 0.62 96.73%

62% 38%

1112

1615

11 11

1615

1819

2423

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0

5

10

15

20

25

30

ARR:27 - DEP:27 ARR:09 - DEP:09

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LEZL (SVQ) - Sevilla


88

61. LFBO (TLS) - Toulouse-Blagnac

Toulouse airport has two parallel runways (plus one helipad) operated in several different ways, from

segregated to single runway use in mixed mode, being the share quite distributed in 7 configurations.

There are several movements difference in throughput between configurations and the highest PSR

value is observed for conf ARR:14R-DEP:14L (29% share).The data provision from Toulouse does not

allow the analysis of the additional times.


Annual traffic 2019





Airport Resilience

LFBO Toulouse-Blagnac 95665 7 1.76 97.29%

35% 29% 9% 7% 6% 6% 5%

15 1513 13 13

11

15

1819

14 1413

1416

1314

1315

1213

1515

19

1416

1415

17

2426

22

25

21 21

25

3029

23

27

2223

28

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0

5

10

15

20

25

30

35

ARR:32L -DEP:32R

ARR:14R -DEP:14L

ARR:32R -DEP:32R

ARR:32L -ARR:32R -DEP:32R

ARR:14R -DEP:14R

ARR:32L -DEP:32L

ARR:14L -DEP:14L

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LFBO (TLS) - Toulouse-Blagnac


89

62. LFLL (LYS) - Lyon-Saint-Exupéry

Lyon has two parallel runways but it normally operates only the longest one in mixed mode, in either

direction. The PSR only differs in one movement, and performance in terms of additional times is very

similar, with only slightly higher additional ASMA times for RWY 35L (59% share). The usage of only

one runway plus the deviation between PSR and maximum throughput signal that congestion might

be too low to consider the PSR a good proxy for capacity.


Annual traffic 2019





Airport Resilience

LFLL Lyon-Saint-Exupéry 116451 2 0.01 0.09 0.39 98.95%

59% 39%

25 25

31 31

26 25

3331

37 36

45 45

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0

5

10

15

20

25

30

35

40

45

50

ARR:35L - DEP:35L ARR:17R - DEP:17R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LFLL (LYS) - Lyon-Saint-Exupéry


90

63. LFML (MRS) - Marseille-Provence

Marseille has two parallel runways but like Lyon it normally operates only the longest one in mixed

mode, in both directions. Performance in terms of PSR is identical with slightly higher additional ASMA

times for RWY 31R (66% share). The data provision from Marseille does not allow for the analysis of

the additional taxi-out times.


Annual traffic 2019





Airport Resilience

LFML Marseille-Provence 102011 2 0.15 0.00 100.00%

66% 32%

17 17

21 21

16 16

2019

27 27

3231

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0

5

10

15

20

25

30

35


Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LFML (MRS) - Marseille-Provence


91

64. LFMN (NCE) - Nice-Côte d’Azur

Nice airport has two parallel runways operated in several different ways, from segregated to single

runway use in mixed mode, with very different results in throughput. However it is important to

consider the high seasonality of Nice, and in fact the most common single runway use (RWY04R; 29%

share) shows the lowest PSR but a significantly higher maximum throughput and the best performance

in terms of additional times, so it is probably a configuration used in low demand. The two

configurations using the direction 22 show much higher additional times in the approach.


Annual traffic 2019





Airport Resilience

LFMN Nice-Côte d’Azur 145645 5 0.33 0.52 8.99 89.15%

51% 29% 8% 6% 4%

26

18 19

27 27

33

2523

37

32

25

18 18

2427

29

2321

2731

47

3235

4851

56

4339

6157

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0

10

20

30

40

50

60

70

ARR:04L - DEP:04R ARR:04R - DEP:04R ARR:22L - DEP:22L ARR:22R - DEP:22L ARR:04L - ARR:04R -DEP:04R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LFMN (NCE) - Nice-Côte d’Azur


92

65. LFPG (CDG) - Paris-Charles-de-Gaulle

Paris Charles de Gaulle has 4 runways, almost parallel that are used as two independent sets of two

runways operating in segregated mode. The longest runways are always used for departures and the

shortest for arrivals. The capacity is nearly symmetric according to the results and the performance in

terms of additional times is also very similar for both configurations.


Annual traffic 2019





Airport Resilience

LFPG Paris-Charles-de-Gaulle 504887 2 0.08 0.03 0.62 99.43%

62% 35%

66 66

75 74

61 6268 68

108 109116 115

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0

20

40

60

80

100

120

140

ARR:26L - ARR:27R - DEP:26R - DEP:27L ARR:08R - ARR:09L - DEP:08L - DEP:09R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LFPG (CDG) - Paris-Charles-de-Gaulle


93

66. LFPO (ORY) - Paris-Orly

Paris Orly has three runways, from which RWY02/20 is operated only in exceptional circumstances.

During the second part of the year the runway 08/26 was completely rebuilt, deeply affecting the

normal operating conditions of the airport and consequently the results of this study. Taking the entire

year into account, the most common use is the segregated modes ARR:26-DEP:24 (41% share) and

ARR:06-DEP:08 (23% share), but of course due to the runway closure the single runway use in mixed

mode also appears as representative. Although the schedule was adapted and reductions were

coordinated during the period of the works, the impact of the single runway use and the works is clear

in terms of additional taxi-out and ASMA times.


Annual traffic 2019





Airport Resilience

LFPO Paris-Orly 221602 4 0.54 0.25 3.18 97.41%

41% 23% 18% 12%

29 3026 26

36

32 31 302931

26 26

35 36

28 28

5254

46 46

62 62

5048

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0

10

20

30

40

50

60

70

ARR:26 - DEP:24 ARR:06 - DEP:08 ARR:24 - DEP:24 ARR:06 - DEP:06

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LFPO (ORY) - Paris-Orly


94

67. LFSB (BSL) - Bâle-Mulhouse

Bâle-Mulhouse has two crossing runways but one is too short so it normally uses only the longest one,

operating it in mixed mode in both directions. RWY15 is used 91% of the time in 2019, showing a PSR

1 movement higher than the opposite direction, but this difference is minimized by the share and the

resilience remains high. Nevertheless, the additional ASMA times are higher for this configuration,

while additional taxi-out times are lower.


Annual traffic 2019





Airport Resilience

LFSB Bâle-Mulhouse 84560 2 0.04 0.52 0.09 99.63%

91% 9%

14 14

28

18

13 13

19

15

2423

36

27

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0

5

10

15

20

25

30

35

40

ARR:15 - DEP:15 ARR:33 - DEP:33

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LFSB (BSL) - Bâle-Mulhouse


95

68. LGAV (ATH) - Athens

Athens has two parallel runways operated in segregated mode, with three representative runway

configurations. The most common configuration ARR:03L-DEP:03R (50% share) shows the highest

throughput (both in PSR and in maximum) but it also has the highest additional taxi-out times, while

showing the best performance in the approach. There is a four movements difference between the

two most common configurations, affecting the resilience and the impact on the PSR.


Annual traffic 2019





Airport Resilience

LGAV Athens 220639 3 0.33 0.14 1.15 97.87%

50% 29% 14%

3229

31

3734 34

3027 28

3431 32

54

50

54

6057 58

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

10

20

30

40

50

60

70

ARR:03L - DEP:03R ARR:21R - DEP:21L ARR:03R - DEP:03L

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LGAV (ATH) - Athens


96

69. LHBP (BUD) - Budapest/ Ferihegy

Budapest airport has two parallel runways, only used simultaneously in segregated mode 65% of the

time (the rest of the time only one runway in mixed mode is in use). The PSR is only two movements

lower for the single runway configurations than for the segregated mode, but the performance in

terms of additional times is much worse for both taxi-out and approach phase when only one runway

is used for both arrivals and departures.


Annual traffic 2019





Airport Resilience

LHBP Budapest/ Ferihegy 122132 5 0.76 0.45 0.57 99.00%

43% 22% 16% 10% 6%

18 18 18 18 17

2122 22

2022

1918 18 17 18

23 23 23

20 21

32 3230 30 31

3937

3234

37

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0

5

10

15

20

25

30

35

40

45

ARR:31R - DEP:31L ARR:13R - DEP:13L ARR:31L - DEP:31L ARR:13R - DEP:13R ARR:31R - DEP:31R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LHBP (BUD) - Budapest/ Ferihegy


97

70. LICC (CTA) - Catania

Catania is a one runway airport with conditions that clearly favour one runway direction utilization. In

fact, RWY01 is used (in mixed mode) 97% of the time in 2019. The other runway configuration is not

used more than 3% of the time (with some time intervals being the transition between configurations

or missing runway information) so it is not considered representative and there is no imbalance to be

analysed.


Annual traffic 2019





Airport Resilience

LICC Catania 75399 1 0.00 0.00 0.00 100.00%

100%

12

16

12

16

21

27

2.2

2.3

2.3

2.3

2.3

2.3

2.4

2.4

2.4

2.4

0

5

10

15

20

25

30

ARR:08 - DEP:08

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LICC (CTA) - Catania


98

71. LIMC (MXP) - Milan/ Malpensa

Milan Malpensa has two parallel runways operated in segregated mode and almost equally used in

both directions. The PSR is 2 movements higher for the more common configuration ARR:35R-

DEP:35L; 51% share) and this configuration also shows slightly better performance in terms of

additional times, especially in the approach.


Annual traffic 2019





Airport Resilience

LIMC Milan/ Malpensa 233978 2 0.01 0.25 0.90 98.55%

51% 45%

3533

39 39

32 31

3734

6260

69 68

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0

10

20

30

40

50

60

70

80

ARR:35R - DEP:35L ARR:35L - DEP:35R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LIMC (MXP) - Milan/ Malpensa


99

72. LIME (BGY) - Bergamo

Bergamo has two runways but one is a short landing strip, so basically it is operated as a one runway

airport. According to the reported runway use, RWY28 is used (in mixed mode) a 100% of the time in

2019. There is therefore no imbalance to be analysed. During the first half of 2019 the systems at

Bergamo could not register properly the movements on RWY10, and RWY28 was the default value.

Nevertheless, in the second half of the year, although some movements are observed on RWY10, the

share is minimal (less than 1%).


Annual traffic 2019





Airport Resilience

LIME Bergamo 95147 1 0.00 0.00 0.00 100.00%

100%

14

19

14

20

23

30

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0

5

10

15

20

25

30

35

ARR:28 - DEP:28A

dd

itio

nal

tim

es (m

inu

tes/

fligh

t)

Mo

vem

ents

/ho

ur

LIME (BGY) - Bergamo


100

73. LIMF (TRN) - Torino Caselle

Torino is a one runway airport with conditions that clearly favour one runway direction utilization. In

fact, RWY36 is used (in mixed mode) a 100% of the time in 2019. There is therefore no imbalance to

be analysed.


Annual traffic 2019





Airport Resilience

LIMF Torino Caselle 39229 1 0.00 0.00 0.00 100.00%

100%

9

15

9

17

15

23

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0

5

10

15

20

25

ARR:36 - DEP:36

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LIMF (TRN) - Torino Caselle


101

74. LIML (LIN) - Milan/ Linate

Milan Linate operates as a one runway airport and its conditions clearly favour one runway direction

utilization. In fact, RWY36 is used (in mixed mode) a 100% of the time in 2019. There is therefore no

imbalance to be analysed.


Annual traffic 2019





Airport Resilience

LIML Milan/ Linate 84458 1 0.00 0.00 0.00 100.00%

100%

17

24

16

21

29

36

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

5

10

15

20

25

30

35

40

ARR:36 - DEP:36

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LIML (LIN) - Milan/ Linate


102

75. LIPE (BLQ) - Bologna

Bologna is a single runway airport operated in mixed mode in both directions, being RWY12 the most

commonly used in 2019 (76% share). The PSR is almost identical for both configurations but additional

times are slightly higher for the preferential runway use.


Annual traffic 2019





Airport Resilience

LIPE Bologna 77090 2 0.16 0.05 0.00 100.00%

76% 23%

1112

1415

12 12

1516

19 19

2221

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0

5

10

15

20

25

ARR:12 - DEP:12 ARR:30 - DEP:30

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LIPE (BLQ) - Bologna


103

76. LIPZ (VCE) - Venice

Venice is has two runways but one is used mainly as taxiway. Conditions clearly favour one runway

direction utilization, resulting in one configuration, where only RWY04R is used (in mixed mode) a

100% of the time in 2019. There is therefore no imbalance to be analysed.


Annual traffic 2019





Airport Resilience

LIPZ Venice 95266 1 0.00 0.00 0.00 100.00%

100%

16

20

15

20

27

34

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

5

10

15

20

25

30

35

40

ARR:04R - DEP:04R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LIPZ (VCE) - Venice


104

77. LIRA (CIA) - Rome/Ciampino

Rome Ciampino is a single runway airport operated in mixed mode in both directions, with RWY15 in

use most of the time (91% share). The PSR is identical for both configurations but additional ASMA

times are slightly higher for the non-preferential runway use.


Annual traffic 2019





Airport Resilience

LIRA Rome/Ciampino 51154 2 0.00 0.03 0.00 100.00%

91%8%

109

14

1110

9

16

11

16 16

22

18

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0

5

10

15

20

25

ARR:15 - DEP:15 ARR:33 - DEP:33

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LIRA (CIA) - Rome/Ciampino


105

78. LIRF (FCO) - Rome/Fiumicino

Rome Fiumicino has three runways, two of them parallel, and from all operating possibilities the

analysis identifies three representative configurations. The PSR is two movements lower for the least

common configuration, with a low impact on resilience. The additional times at Rome, especially on

the taxi-out phase are very high in general. There is no clear configuration that performs better than

the others for both indicators, one being better for taxi and a different one being better for the

approach.


Annual traffic 2019





Airport Resilience

LIRF Rome/Fiumicino 309783 3 0.12 0.36 0.24 99.68%

60% 14% 12%

37

45

36

4549

40

49

34

48

56

39

52

74 74 72

8581

76

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

0

10

20

30

40

50

60

70

80

90

ARR:16L - ARR:16R - DEP:25 ARR:16L - DEP:25 ARR:34L - ARR:34R - DEP:25A

dd

itio

nal

tim

es (m

inu

tes/

fligh

t)

Mo

vem

ents

/ho

ur

LIRF (FCO) - Rome/Fiumicino


106

79. LIRN (NAP) - Naples

Naples has one runway operated in mixed mode with quite even share of both directions (58%-40%).

Performance in terms of peak service rate differs in two movements, and while RWY24 observed

higher additional taxi-out times, RWY06 shows significantly higher additional ASMA times


Annual traffic 2019





Airport Resilience

LIRN Naples 84387 2 0.16 0.32 1.15 95.73%

58% 40%

1415

17

19

1415

1819

25

27

32 32

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

5

10

15

20

25

30

35

ARR:24 - DEP:24 ARR:06 - DEP:06

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LIRN (NAP) - Naples


107

80. LJLJ (LJU) - Ljubljana

Ljubljana is a one runway airport with conditions that clearly favour one runway direction utilization.

In fact, RWY30 is used (in mixed mode) 98% of the time in 2019. The other runway configuration is

not considered representative and there is no imbalance to be analysed.


Annual traffic 2019





Airport Resilience

LJLJ Ljubljana 27935 1 0.00 0.00 0.00 100.00%

98%

9

13

10

1314

18

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0

2

4

6

8

10

12

14

16

18

20

ARR:30 - DEP:30

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LJLJ (LJU) - Ljubljana


108

81. LKPR (PRG) - Prague

Prague has two crossing runways but it normally only operates one of them at a time, in mixed mode.

The PSR only differs in one movement less for the least common runway (RWY30; 5% share), resulting

in high resilience. Performance in terms of additional times is best for both approach and taxi-out

phase for the RWY06 (24% share) and not for the preferred runway (RWY24; 69% share).


Annual traffic 2019





Airport Resilience

LKPR Prague 150434 3 0.56 0.18 0.05 99.87%

69% 24% 5%

23 23 24

3028 28

24 24 23

2931

26

39 39 38

47

4442

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0

5

10

15

20

25

30

35

40

45

50

ARR:24 - DEP:24 ARR:06 - DEP:06 ARR:30 - DEP:30

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LKPR (PRG) - Prague


109

82. LMML (MLA) - Malta

Malta airport has two converging runways but it normally uses only the longest one, operating it in

mixed mode in both directions. The most commonly used (RWY31; 59%) shows a PSR significantly

higher than the opposite direction, which results in a lower resilience value. Nevertheless, the

additional taxi-out times are also higher for this configuration, while additional ASMA times are very

similar.


Annual traffic 2019





Airport Resilience

LMML Malta 58181 2 0.17 0.03 1.59 91.63%

59% 32%

11

9

15

1211

9

1413

19

14

23

20

0.0

0.5

1.0

1.5

2.0

2.5

0

5

10

15

20

25

ARR:31 - DEP:31 ARR:13 - DEP:13

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LMML (MLA) - Malta


110

83. LOWW (VIE) - Vienna

Vienna airport has two converging runways that are operated in 7 representative configurations

according to the data found in 2019. The PSR for all of the configurations that use the two runways

differs in one or two movements, but in addition there is a configuration in which only RWY29 is in

use (5% share) with a significant reduction of the PSR but most importantly, a drastic increase of the

additional times both in the taxi-out and the approach phase. It is also noticeable the configurations

used in departure or arrival peaks, which show the highest throughputs.


Annual traffic 2019





Airport Resilience

LOWW Vienna 281716 7 0.53 1.29 2.71 98.14%

40% 16% 14% 12% 7% 5% 4%

37 37

4541

2833

4443 43

52

45

3336

50

40 39

28

37

44

2927

43 42

31

41

47

31 32

62 6265 63 64

50

65

7268

7267 67

57

76

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

0

10

20

30

40

50

60

70

80

ARR:34 -DEP:29

ARR:16 -DEP:29

ARR:34 -DEP:29 -DEP:34

ARR:11 -DEP:16

ARR:11 -ARR:16 -DEP:16

ARR:29 -DEP:29

ARR:16 -DEP:16 -DEP:29

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LOWW (VIE) - Vienna


111

84. LPFR (FAO) - Faro

Faro is a single runway airport operated in mixed mode in both directions, being RWY28 the most

commonly used in 2019 (73% share). The PSR for both configurations differs only in one movement

and additional times are very similar, with additional ASMA time slightly higher for the preferential

runway use.


Annual traffic 2019





Airport Resilience

LPFR Faro 60664 2 0.03 0.09 0.73 97.30%

73% 27%

15 15

2021

15 16

20 20

2627

3335

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0

5

10

15

20

25

30

35

40

ARR:28 - DEP:28 ARR:10 - DEP:10

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LPFR (FAO) - Faro


112

85. LPPR (OPO) - Porto

Porto in a single runway airport, used in mixed mode in both directions.RWY35, with a 69% share,

shows the best performance both in throughput and in terms of additional times both in the taxi-out

and approach phase, which are significantly worse for RWY17 (31% share).


Annual traffic 2019





Airport Resilience

LPPR Porto 98816 2 0.50 0.52 0.31 98.70%

69% 31%

14 14

17 17

13 13

18

16

2423

30

27

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0

5

10

15

20

25

30

35

ARR:35 - DEP:35 ARR:17 - DEP:17

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LPPR (OPO) - Porto


113

86. LPPT (LIS) - Lisbon

Lisbon airport has one runway (the former runway 17/35 was reconverted into a taxiway) used in

mixed mode in both directions. The difference in PSR is only one movement less for RWY 21, and

although this configuration shows higher additional taxi-out times, the most common runway

(RWY03; 72% share) shows higher additional ASMA times.


Annual traffic 2019





Airport Resilience

LPPT Lisbon 220938 2 0.16 0.30 0.28 99.35%

72% 28%

24 23

29

26

23 22

25 24

43 42

45 45

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0

5

10

15

20

25

30

35

40

45

50

ARR:03 - DEP:03 ARR:21 - DEP:21

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LPPT (LIS) - Lisbon


114

87. LROP (OTP) - Bucharest/ Otopeni

Bucharest Otopeni has two parallel runways used in segregated mode for 82% of the time. There is up

to four movements difference in PSR between configurations, but it is interesting to observe that the

use of only RWY08L in mixed mode is the second highest in terms of PSR. The single runway

configurations (either RWY08L or RWY08R) show however the longest additional taxi-out times, while

additional ASMA times are higher for two of the segregated configurations.


Annual traffic 2019





Airport Resilience

LROP Bucharest/ Otopeni 122831 5 0.28 0.28 1.98 96.91%

36% 24% 22% 7% 5%

22

1819

1816

29

2325

23 23

1816

1718 18

2120

21 2120

31

2728

29

26

36

33

37

3331

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0

5

10

15

20

25

30

35

40

ARR:08L - DEP:08R ARR:08R - DEP:08L ARR:26L - DEP:26R ARR:08L - DEP:08L ARR:08R - DEP:08R

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

ents

/ho

ur

LROP (OTP) - Bucharest/ Otopeni


115

88. LSGG (GVA) - Geneva

Geneva is a single runway airport operated in mixed mode in both directions. RWY22, used 59% of the

time, has a PSR 1 one movement lower than RWY04, and additional times are higher both in the

approach and the taxi-out phase.


Annual traffic 2019





Airport Resilience

LSGG Geneva 179115 2 0.20 0.34 0.59 98.67%

59% 41%

25 25

30 30

24 2427 28

43 4446

48

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0

10

20

30

40

50

60

ARR:22 - DEP:22 ARR:04 - DEP:04

Ad

dit

ion

al ti

mes

(min

ute

s/fli

ght)

Mo

vem

en

ts/h

ou

r

LSGG (GVA) - Geneva


116

89. LSZH (ZRH) - Zürich

Zürich has three runways, operated most of the time in segregated mode. The most common

configuration ARR:14-DEP:28 (42% share) has the highest throughput and the lowest combined

delays. The least common configuration ARR:34-DEP:32 (5% share) shows a PSR 10 movements below

the reference capacity, and the highest delays in the taxi-out and approach while configuration

ARR:14-DEP:16;DEP:28 (30% share) seems to be used for departure peaks and has high additional taxi-

out times.


Annual traffic 2019





Airport Resilience

LSZH Zürich 269223 4 1.19 0.54 2.06 98.50%

42% 30% 19% 5%

36

41

3431

3943

3835

40

33 3330

43

36 36

31

6462

59

54

7268

63

56

0.0

1.0

2.0

3.0

4.0

5.0

6.0

0

10

20

30

40

50

60

70

80

ARR:14 - DEP:28 ARR:14 - DEP:16 - DEP:28 ARR:28 - DEP:32 ARR:34 - DEP:32A

dd

itio

nal

tim

es (m

inu

tes/

fligh

t)

Mo

vem

ents

/ho

ur

LSZH (ZRH) - Zürich


117

90. LZIB (BTS) - Bratislava

Bratislava has two crossing runways operated in four representative configurations. The configuration

that shows the highest throughput and the best performance in terms of delay in the taxi-out phase

is used only 5% of the time (ARR:22-DEP:13). Although the demand is probably low to consider the

PSR a good proxy for capacity, the two most common configurations, with a combined share of 89%

show a low departure throughput with much higher additional taxi-out times, which might signal

issues like bottlenecks in the taxi-out flows.


Annual traffic 2019





Airport Resilience

LZIB Bratislava 22593 4 0.51 0.07 6.22 72.60%

71% 18% 5% 4%

7 7

9

7

1110

13

9

7 7

10

6

109

12

7

1112

18

11

1415

23

12

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0

5

10

15

20

25

ARR:31 - DEP:04 ARR:31 - DEP:31 ARR:22 - DEP:13 ARR:31 - DEP:13A

dd

itio

nal

tim

es (m

inu

tes/

fligh

t)

Mo

vem

en

ts/h

ou

r

LZIB (BTS) - Bratislava


AIRPORT CAPACITY IMBALANCE - Eurocontrol · 2020. 12. 11. · The chance for demand/capacity...

Documents

Transcript of AIRPORT CAPACITY IMBALANCE - Eurocontrol · 2020. 12. 11. · The chance for demand/capacity...