performance review report prr 2016

PERFORMANCE REVIEW COMMISSION

PERFORMANCE REVIEW REPORT An assessment of Air Traffic Management in Europe during the calendar year 2016

PRR 2016

Draft Final Report

For consultation with stakeholders

(17 March – 07 April 2017)

CAVEAT

Some of the data are still provisional.

They will be updated, where necessary, prior to the publication of the final PRR 2016.

NOTICE

The PRC has made every effort to ensure that the information and analysis contained in this document are as accurate and complete as possible. Only information from quoted sources has been used and information relating to named parties has been checked with the parties concerned. Despite these precautions, should you find any errors or inconsistencies we would be grateful if you could please bring them to the PRU’s attention. The PRU’s e-mail address is [email protected].

2017 COPYRIGHT NOTICE AND DISCLAIMER

© European Organisation for the Safety of Air Navigation (EUROCONTROL)

This document is published by the Performance Review Commission in the interest of the exchange of information.

It may be copied in whole or in part providing that the copyright notice and disclaimer are included. The information contained in this document may not be modified without prior written permission from the Performance Review Commission.

The views expressed herein do not necessarily reflect the official views or policy of EUROCONTROL, which makes no warranty, either implied or express, for the information contained in this document, neither does it assume any legal liability or responsibility for the accuracy, completeness or usefulness of this information.

Printed by EUROCONTROL, 96 rue de la Fusée, B-1130 Brussels, Belgium. The PRC’s website address is http://www.eurocontrol.int/european-ans-performance-review.

The PRU’s e-mail address is [email protected].

mailto:[email protected]

http://www.eurocontrol.int/european-ans-performance-review

DOCUMENT IDENTIFICATION SHEET

DOCUMENT DESCRIPTION

Document Title

Performance Review Commission

Performance Review Report covering the calendar year 2016 (PRR 2016)

PROGRAMME REFERENCE INDEX: EDITION: EDITION DATE:

PRC Performance Review Report Draft report 16-MAR-2017

SUMMARY

This report of the Performance Review Commission analyses the performance of the European Air Traffic Management System in 2016 under the Key Performance Areas of Safety, Capacity, Environment and Cost-efficiency.

Keywords

Air Traffic Management Performance Measurement

Performance Indicators ATM ANS

CONTACT:

Performance Review Unit, EUROCONTROL, 96 Rue de la Fusée,

B-1130 Brussels, Belgium. Tel: +32 2 729 3956, E-Mail: [email protected]


DOCUMENT STATUS AND TYPE

STATUS DISTRIBUTION

Draft General Public

Proposed Issue EUROCONTROL Organisation

Released Issue Restricted

INTERNAL REFERENCE NAME: PRR 2016

mailto:[email protected]


FOREWORD by the PRC Chairman

For almost 20 years, the independent Performance Review Commission (PRC) has been measuring pan-European ATM performance and making recommendations for improvements. The EUROCONTROL performance review scheme, which began in 1998, was a world-first at the time. Since then, elements have been adopted by ICAO and applied by States worldwide including China, Brazil and Singapore.

Closer to home, the European Commission built on the solid body of work produced by the PRC by establishing a performance scheme for the Single European Sky (SES).

The Commission designated the PRC, supported by the Performance Review Unit (PRU), as the first Performance Review Body (PRB) of the Single European Sky. This designation ended on 31 December 2016. Thus, from 2017 onwards, the PRB will be a separate group designated by the European Commission.

To ensure that there are no overlaps between the PRC’s tasks and those of the PRB, the PRC has agreed to a joint proposal made by EUROCONTROL and the European Commission on how the PRC’s future tasks could complement those of the PRB and avoid duplication.

The PRC held a series of meetings with stakeholders in 2016 to listen to their needs and requirements. The purpose was to establish whether the usefulness of the PRC’s main products – the annual Performance Review Report and the annual ATM Cost-effectiveness (ACE) Benchmarking report – could be further improved.

The dialogue with stakeholders has been effective and constructive and the PRC thanks all stakeholders concerned.

The PRC has listened and taken action. From now on, there will be improved PRC reporting. With the PRU in support, the PRC will continue to develop its web presence and publish short quarterly reviews, so that high level performance information is available more quickly. This will also help to slim-down the PRR and ACE reports, as a lot of information will become available online.

I hope that you find this approach, and this new-look PRR, even more useful for your requirements.

Should you wish to contact the PRC, you can find contact details on the inside-back cover of this report.

Pleasant reading!

Ralph Riedle

Chairman

Performance Review Commission

EXECUTIVE SUMMARY

EXECUTIVE SUMMARY PRR 2016

i

EXECU TI VE SUMM ARY 1

This report assesses the performance of Air Navigation Services (ANS) in the EUROCONTROL area for the calendar 2 year 2016 for all key performance areas, except for cost-efficiency, which analyses performance in 2015 as this is 3 the latest year for which actual financial data are available. 4

In 2016, air traffic in the EUROCONTROL area (ESRA08) continued to increase for the third year in 5 a row. On average, the number of controlled flights increased by 2.4% compared to 2015. The 6 main driver of the observed growth in 2016 was the growth in the intra-European low cost traffic 7 segment. As in previous years, passenger numbers grew at a higher rate than traffic (+5.1% vs. 8 2015). 9

In 2016, annual traffic reached the pre-economic crisis level of 2008 and the third quarter in 2016 was the highest 10 on record. Of the 39 Air Navigation Service Providers (ANSPs) included in the analysis, 25 showed an increase in 11 traffic compared to 14 ANSPs which showed a decline in 2016. In absolute terms, ENAIRE (Spain), NATS (UK) and 12 DSNA (France) experienced the highest year on year growth in 2016. DHMI (Turkey), UKSATSE (Ukraine) and 13 ROMATSA (Romania) reported the highest absolute decrease in 2016. 14

The substantial traffic increase in some areas contributed to a decrease in overall service quality. The share of 15 flights arriving within 15 minutes of their scheduled time decreased by 1.6 percent points to reach 81.5% in 2016. 16

Safety is the primary objective of ANS and overall safety levels in the EUROCONTROL area remain 17 high. There was only one reported air traffic accident with direct ANS contribution in 2015, which 18 is the latest year for which validated data are available. 19

With the exception of Unauthorised Penetrations of Airspace (UPAs), the number of all key risk 20 occurrence types (Separation minima infringements (SMIs), Runway incursions (RIs), and ATM Specific 21 Occurrences) decreased in the EUROCONTROL area in 2015, despite the increase in traffic. Overall, there were 22 15 SMIs and 28 UPAs per hundred thousand controlled flight hours in the airspace and 8 RIs per hundred 23 thousand movements at airports reported in 2015. 24

The quality and completeness of safety data reported to EUROCONTROL increased over the past years but with 25 scope for further improvement, particularly in terms of severity classification. Although this has been pointed out 26 by the PRC on several occasions, 16% of the reported occurrences were still not severity classified in 2015. 27

The PRC review of the implementation status of the Acceptable Level of Safety Performance (ALoSP) concept in 28 EUROCONTROL Member States clearly suggested that there is a need for common definitions and guidance 29 material in order to ensure a harmonised approach in the EUROCONTROL area. 30

The PRC’s concern about over conservative capacity planning and the risk of performance 31 deterioration when traffic grows again has been voiced on several occasions. In 2016, total en-32 route ATFM delays increased by 21% compared to 2015 and the share of flights affected by en-33 route ATFM delays increased from 3.9% to 4.8% in 2016. 34

ATC Capacity/Staffing related constraints remained by far the main driver of en-route ATFM 35 delays (55.3%), followed by weather-related constraints (18.3%), ATC disruptions/ industrial actions (12.3%) and 36 Event related constraints (9.1%) which also include delays due to ATC system upgrades. 37

Three quarters of the en-route ATFM delays were generated by four air navigation service providers: DSNA 38 (41.6%), DFS (13.0%), Maastricht (11.4%) and ENAIRE (9%). The vast majority of Area Control Centres (ACCs) 39 performed well in 2016, with notable improvements at Lisbon, Athens, and Zagreb ACCs. The most constraining 40 ACCs in 2016 were Brest, Nicosia, Bordeaux, Brussels, Barcelona, Prestwick, Maastricht UAC, Warsaw, Canarias, 41 Karlsruhe UAC and Marseille. Together, they accounted for 70.1% of all en-route ATFM delays but only 30.1% of 42 total flight hours controlled in the EUROCONTROL area. 43

The reasons for the constraints varied by ACC and were in some cases exacerbated by the higher than expected 44 traffic growth. In view of the number of planned major project implementations over the next years it is important 45 to reiterate the message from last year’s PRR that ANSPs need to effectively coordinate the planning and 46 implementation of all changes to the ATM system that could adversely affect operations with the Network 47 Manager. 48

EXECUTIVE SUMMARY


ii

Horizontal en-route flight efficiency in the EUROCONTROL area decreased slightly from 97.3% to 97.1% in 2016, 1 after a continuous improvement over the past years. 2

The effects of ATC industrial action on specific days in 2016 are clearly visible but the overall impact on system 3 wide flight efficiency remains within 0.03% points. 4

Despite a slight decrease in flight efficiency at system level in 2016, the benefits of Free Route Airspace (FRA) 5 implementation and related reductions in fuel burn, emissions and costs are clearly visible in a number of 6 Member States. On average, flight efficiency is 1.6% points better in Member States where FRA is fully 7 implemented all day, and actual flown trajectories are notably closer to the filed flight plans. 8

Complementary to horizontal flight efficiency, an initial evaluation of vertical en-route flight efficiency in this 9 year’s PRR enabled clear differences on specific airport pairs to be identified. Work is in progress to better 10 quantify the measured inefficiencies in terms of fuel burn and CO2 emissions in the future. 11

Closer civil-military cooperation and coordination is an important enabler to improve capacity and flight efficiency 12 performance. Some areas for further improvement identified in a PRC survey relate to the lack of impact 13 assessment in terms of capacity and route options for restricted/segregated airspace and the absence of clear 14 strategic objectives. 15

The analysis of the top 30 airports in terms of traffic showed that ten airports (Amsterdam, 16 Istanbul Ataturk, London Gatwick, Stockholm Arlanda, Istanbul Sabiha Gökçen, Dublin, Berlin 17 Tegel, Geneva, Lisbon and Warsaw) reported their highest traffic level on record, surpassing 18 the levels observed before the economic crisis starting in 2008. Amsterdam reported a 5.9% 19 increase in traffic in 2016 which made it the airport with the most commercial movements in 20 Europe in 2016. 21

The two Istanbul airports, which reported a remarkable traffic growth over the past years, were affected by the 22 situation in Turkey, resulting in a notable slowdown in traffic growth. Of the top 30 airports, six showed a traffic 23 decrease in 2016 with the highest decrease observed for Brussels airport (-6.5% vs 2015) as a result of the 24 reduced capacity following the terrorist attacks in March 2016. 25

The substantial traffic increase at some airports contributed to higher levels of operational inefficiency and 26 resulted in somewhat higher additional times during descent and in the taxi-out phase compared to 2015. 27

Average airport arrival ATFM delay and additional holding (ASMA) time decreased slightly in 2016 at the top 30 28 airports but were still heavily concentrated among a few airports. Five airports (Istanbul Sabiha Gökçen, Istanbul 29 Ataturk, Amsterdam, London Heathrow, and London Gatwick) accounted for 59% of the airport arrival ATFM delay 30 reported for the top 30 airports. The situation in Istanbul is expected to improve with the opening of the first 31 phase of the new Istanbul Airport which is scheduled for 2017/2018. Airport arrival ATFM performance at 32 Amsterdam and the two London airports (LHR, LGW) was to a large extent affected by weather which required the 33 available capacity to be reduced. 34

London Heathrow, Istanbul Ataturk and Istanbul Sabiha Gökçen show all up with a continuously high arrival 35 throughput close to the peak declared arrival capacity. Although this maximises the use of capacity, the high 36 intensity operation close to maximum capacity can result in high delays and possibly cancellations when there is a 37 mismatch between scheduled demand and the capacity that can be made available. 38

The group of smaller Greek airports reported in last year’s report continued to generate high ATFM delays in 39 2016. The issue appears to be linked to scheduling and variability. It needs to be addressed proactively in order to 40 avoid a repetition of high delays also in summer 2017. The PRC will be monitoring the situation which has 41 persisted now for several years. 42

Whereas A-CDM implementation is considered to be an enabler to improve situation awareness and performance, 43 it is important to ensure that the available information is used to improve local processes. A-CDM can also help to 44 improve the data quality which is presently an issue for the measurement of ATC pre-departure delays. 45

Vertical flight efficiency in climbs and descents at the top 30 airports has been added as a new metric in this year’s 46 report. On average, inefficiencies were more than 6 times higher in descent than in climb with notable differences 47 by airport. 48

EXECUTIVE SUMMARY


iii

In 2015, which is the latest year for which actual financial data are available, the en-route ANS 1 unit costs of the Pan-European system amounted to 49.2 €2009 per service unit (TSU). This is 2 -2.4% lower than in 2014 since in 2015 the number of TSUs rose faster (+3.9%) than en-route 3 ANS costs (+1.5%). En-route unit costs are expected to reduce by -1.6% p.a. over the 2015-4 2019 period and reach a value of 46.1 €2009. If these plans materialise, the en-route unit costs 5

in 2019 will be some -23% lower than in 2009, implying substantial cost-efficiency improvements during this 10 6 year period. 7

In 2015, European terminal ANS unit costs amounted to 171.6 €2009 per terminal service unit (TNSU) and are 8 expected to decrease by -2.1% p.a. until 2019. This performance improvement reflects the fact that total terminal 9 ANS costs are planned to reduce by -0.7% p.a. while TNSUs are expected to increase by +1.4% p.a. between 2015 10 and 2019. 11

Detailed ANSPs benchmarking analysis indicates that in 2015 gate-to-gate ATM/CNS provision costs increased by 12 +1.7% and amounted to some €8.2 Billion at Pan-European system level. At the same time traffic, expressed in 13 terms of composite flight hours, increased at a slightly higher rate (+1.8%). As a result, gate-to-gate unit ATM/CNS 14 provision costs remained fairly constant in 2015 (-0.1% vs 2014). 15

In order to also consider the service quality provided by ANSPs, the gate-to-gate economic performance combines 16 ATM/CNS provision costs and the cost of ATFM delays. 17

Although unit ATM/CNS provision costs remained constant in 2015, unit economic costs increased by +4.2% to 18 reach €505 per composite flight-hour reflecting a substantial increase in the unit costs of ATFM delays (+38.7% vs. 19 2014). 20

In fact, the trend of decreasing ATFM delays observed in previous years stopped in 2013, when a new cycle 21 characterised by higher delays started. 22

The analysis provided in the operational en-route ANS performance chapter of this report indicates that this trend 23 continued in 2016 since en-route ATFM delays were +20.9% higher than in 2015. 24

This implies that in 2016, the unit costs of delays will be significantly higher than in 2015 and will negatively affect 25 ANSPs economic cost-effectiveness. 26 27

PRC Recommendations 2016 28

PRC recommendations will be included in the final report. 29

T A B L E O F C O N T E N T S

EXECUTIVE SUMMARY .................................................................................................................. I

PRC RECOMMENDATIONS 2016 ................................................................................................................ III

1 INTRODUCTION AND CONTEXT ............................................................................................. 1

1.1 ABOUT THIS REPORT ..................................................................................................................... 1

1.2 EUROPEAN AIR TRANSPORT KEY INDICES ........................................................................................... 3

2 SAFETY ................................................................................................................................. 7

2.1 INTRODUCTION ............................................................................................................................ 7

2.2 ACCIDENTS.................................................................................................................................. 7

2.3 INCIDENTS .................................................................................................................................. 9

2.4 REPORTING AND INVESTIGATION ...................................................................................................11

2.5 ACCEPTABLE LEVEL OF SAFETY PERFORMANCE (ALOSP) ...................................................................12

2.6 CONCLUSIONS ...........................................................................................................................13

3 OPERATIONAL EN-ROUTE ANS PERFORMANCE .................................................................... 14

3.1 INTRODUCTION ..........................................................................................................................14

3.2 TRAFFIC EVOLUTION ...................................................................................................................15

3.3 ANS-RELATED FLIGHT EFFICIENCY CONSTRAINTS (EN-ROUTE) .............................................................16

3.4 CIVIL MILITARY COOPERATION & COORDINATION .............................................................................32

3.5 CONCLUSIONS ...........................................................................................................................34

4 OPERATIONAL ANS PERFORMANCE AT AIRPORTS ............................................................... 35

4.1 INTRODUCTION ..........................................................................................................................35

4.2 TRAFFIC EVOLUTION AT THE TOP 30 EUROPEAN AIRPORTS .................................................................37

4.3 CAPACITY MANAGEMENT (AIRPORTS).............................................................................................38

4.4 ANS-RELATED FLIGHT EFFICIENCY CONSTRAINTS AT AND AROUND AIRPORTS .........................................40

4.5 CONCLUSIONS ...........................................................................................................................47

5 ANS COST-EFFICIENCY (2015) .............................................................................................. 49

5.1 INTRODUCTION ..........................................................................................................................49

5.2 EN-ROUTE ANS COST-EFFICIENCY PERFORMANCE ............................................................................50

5.3 TERMINAL ANS COST-EFFICIENCY PERFORMANCE.............................................................................55

5.4 ANSPS GATE-TO-GATE ECONOMIC PERFORMANCE ...........................................................................58

5.5 CONCLUSIONS ...........................................................................................................................62

L I S T O F F I G U R E S

Figure 1-1: EUROCONTROL States (2016) ................................................................................................ 2 Figure 1-2: Evolution of average daily flights ........................................................................................... 3 Figure 1-3: Year on year change versus 2015 .......................................................................................... 3 Figure 1-4: Traffic growth by Air Navigation Service Provider ................................................................. 3 Figure 1-5: Traffic complexity score (2016) ............................................................................................. 4 Figure 1-6: Traffic seasonality (2016) ....................................................................................................... 4 Figure 1-7: Evolution of European IFR flights (1990-2022) ...................................................................... 5 Figure 1-8: European air traffic indices (2008-2016) ............................................................................... 5 Figure 1-9: Evolution of arrival punctuality.............................................................................................. 5 Figure 1-10: ANS contribution towards departure total departure delays.............................................. 6 Figure 2-1: Accidents in EUROCONTROL area (2011-16P) [TBU] ............................................................. 8 Figure 2-2: Accidents risk distribution (2012-16P) [TBU] ......................................................................... 8 Figure 2-3: Accidents with ATM contribution in the EUROCONTROL area (2011-16P) [TBU] ................. 8 Figure 2-4: Occurrence rates EUROCONTROL area (2016P) [TBU] .......................................................... 9 Figure 2-5: Reported high-risk SMIs (EUROCONTROL) .........................................................................10 Figure 2-6: Reported high-risk UPAs (EUROCONTROL) ..........................................................................10 Figure 2-7: Reported high-risk RIs (EUROCONTROL) .............................................................................10 Figure 2-8: Reported high-risk ATM Spec. Occurrences (EUROCONTROL) ............................................10 Figure 2-9: Reported occurrences (2007-2016P) ...................................................................................11 Figure 2-10: Severity not classified or not determined (2007-2016P) [TBU] .........................................11 Figure 2-11: Completeness of AST reported data in 2016(P) [TBU] ......................................................12 Figure 3-1: Traffic variation by ANSP (2016/2015) ................................................................................15 Figure 3-2: Traffic growth by ACC (2016) ...............................................................................................15 Figure 3-3: Average en-route ATFM delay (EUROCONTROL area) .........................................................16 Figure 3-4: En-route ATFM delayed flights and delay per delayed flight (EUROCONTROL area) ..........16 Figure 3-5: Estimated ATC capacity/staffing related impact on airline operations (2016) ....................17 Figure 3-6: Impact of weather related en-route ATFM delays on airline operations (2016).................17 Figure 3-7: Estimated ATC strike related impact on airline operations (2016) ......................................17 Figure 3-8: Estimated special event related impact on airline operations (2016) .................................18 Figure 3-9: Planned major project implementations (2017-2021) ........................................................18 Figure 3-10: Overview of most constraining ACCs (2016) .....................................................................19 Figure 3-11: Brest ACC en-route performance overview (2016) ...........................................................19 Figure 3-12: Bordeaux ACC en-route performance overview (2016) ....................................................20 Figure 3-13: Marseille ACC en-route performance overview (2016) .....................................................21 Figure 3-14: Nicosia ACC en-route performance overview (2016) ........................................................21 Figure 3-15: Brussels ACC en-route performance overview (2016) .......................................................22 Figure 3-16: Barcelona ACC en-route performance overview (2016) ....................................................22 Figure 3-17: Maastricht UAC en-route performance overview (2016) ..................................................23 Figure 3-18: Maastricht UAC traffic evolution (2010-2016) ..................................................................24 Figure 3-19: Warsaw ACC en-route performance overview (2016) .......................................................24 Figure 3-20: Karlsruhe UAC en-route performance overview (2016) ....................................................25 Figure 3-21: Karlsruhe UAC traffic evolution (2010-2016) .....................................................................25 Figure 3-22: Canarias ACC en-route performance overview (2016) ......................................................25 Figure 3-23: Traffic and ATFM delay by weekday – Canarias ACC (2016) ..............................................26 Figure 3-24: ATFM performance (network indicators) ..........................................................................26 Figure 3-25: Horizontal en-route flight efficiency (Pan-European level) ...............................................27 Figure 3-26: Flight efficiency by State (2016) ........................................................................................28 Figure 3-27: Horizontal en-route flight efficiency (actual trajectory) by State (2016) ..........................29 Figure 3-28: Local and network effects on flight efficiency by State (2016) ..........................................29 Figure 3-29: Example distribution of maximum filed flight levels .........................................................30 Figure 3-30: Results for the top 20 airport pairs in terms of total VFI ...................................................31 Figure 3-31: Distribution of maximum filed flight levels for LFBO-LFPO ...............................................31

Figure 3-32: Distribution of maximum filed flight levels for EGLL-EHAM ..............................................31 Figure 3-33: Identified improvement areas for civil/military cooperation and coordination ...............33 Figure 4-1: ANS-related operational performance at airports (overview) ............................................36 Figure 4-2: Traffic variation at the top 30 European airports (2016/2015) ...........................................37 Figure 4-3: European airports coordination level (>20.000 movements/year) .....................................38 Figure 4-4: Capacity utilisation at top 30 European airports .................................................................38 Figure 4-5: Arrival throughput at the top 30 airports ............................................................................39 Figure 4-6: Evolution of arrival throughput at the top 30 airports (2016) .............................................39 Figure 4-7: ANS-related inefficiencies on the arrival flow at the top 30 airports in 2016 .....................40 Figure 4-8: Arrival ATFM delayed arrivals at the top 30 airports (2016) ...............................................41 Figure 4-9: Five most contributing airports in 2016 (Arrival ATFM delay/ ASMA add. time) ................41 Figure 4-10: ATFM slot adherence at airport (2016) .............................................................................42 Figure 4-11: ANS-related inefficiencies on the departure flow at the top 30 airports in 2016 .............43 Figure 4-12: Five most contributing airports in 2016 (taxi-out add. time) ............................................43 Figure 4-13: ATC Pre-departure delay reporting at the top 30 airports ................................................44 Figure 4-14: Average time flown level per flight at the top 30 airports ................................................45 Figure 4-15: Median CDO/CCO altitude at the top 30 airports..............................................................45 Figure 4-16: Monthly average time flown level per flight to/from EHAM .............................................46 Figure 4-17: Monthly median CDO/CCO altitude to/from EHAM ..........................................................46 Figure 4-18: Vertical trajectories of Amsterdam (EHAM/AMS) arrivals ................................................47 Figure 4-19: Horizontal trajectories of Amsterdam (EHAM/AMS) arrivals ............................................47 Figure 5-1: SES and non-SES States ........................................................................................................50 Figure 5-2: Reconciliation between RP1 and RP2 en-route ANS costs for SES States (€2009) ..............51 Figure 5-3: Real en-route unit costs per SU for EUROCONTROL Area (€2009) ........................................51 Figure 5-4: Breakdown on en-route ANS costs by nature ......................................................................52 Figure 5-5: Breakdown of changes in en-route costs (2014-2015, (€2009)) ............................................52 Figure 5-6: 2015 Real en-route ANS costs per TSU by charging zone (€2009) .........................................53 Figure 5-7: Pan-European en-route cost-efficiency outlook 2016-2019 (in €2009) .................................54 Figure 5-8: Geographical scope of terminal ANS cost-efficiency analysis .............................................55 Figure 5-9: Changes in the reporting of terminal ANS data for SES States between 2010 and 2015 ....55 Figure 5-10: Comparison of 2015 terminal ANS unit costs by TCZ ........................................................56 Figure 5-11: Distribution of terminal ANS costs and TNSUs by TCZ in 2015 ..........................................57 Figure 5-12: Real terminal ANS costs per TNSU, total costs (€2009) and TNSUs .....................................57 Figure 5-13: Breakdown of gate-to-gate ATM/CNS provision costs 2015 (€2015) [TBU] ......................58 Figure 5-14: Changes in economic cost-effectiveness, 2010-2015 (€2015) [TBU] ................................59 Figure 5-15: Economic gate-to-gate cost-effectiveness indicator, 2015 [TBU] .....................................60 Figure 5-16: ANSPs contribution to ATFM delays increase at Pan-European system level in 2015 [TBU]

.......................................................................................................................................................61 Figure 5-17: Breakdown of changes in cost-effectiveness, 2014-2015 (€2015) [TBU] ..........................61

Chapter 1: Introduction

PRR 2016- Chapter 1: Introduction 1

1 Introduction and context 1

1.1 About this report 2

Air Navigation Services (ANS) are essential for the safety, efficiency and sustainability of civil and 3 military aviation, and to meet wider economic, social and environmental policy objectives. 4

The purpose of the independent Performance Review Commission (PRC) is “to ensure the effective 5 management of the European Air Traffic Management system through a strong, transparent and 6 independent performance review”, per Article 1 of its Terms of Reference [Ref. 1]. More information 7 about the PRC is given on the inside cover page of this report. 8

This Performance Review Report (PRR 2016) has been produced by the PRC with its supporting unit 9 the Performance Review Unit (PRU). Its goal is to provide policy makers and ANS stakeholders with 10 objective information and independent advice concerning the performance of European ANS in 2016, 11 based on analysis, consultation and information provided by relevant parties. It also gives some 12 information on other PRC activities in 2016. 13

As in previous years, stakeholders are given an opportunity to comment on PRR 2016 before it is 14 finalised. The PRC will sent the draft final Report to stakeholders, and will post it on the 15 EUROCONTROL internet site, for consultation and comment from 17 March – 07 April 2017. 16

On the basis of PRR 2016, the PRC will provide independent advice on ANS performance and propose 17 recommendations to the EUROCONTROL States. 18

The PRC’s recommendations can be found in the Executive Summary. 19

20

1.1.1 Further PRC work 21

In addition to the PRR which provides a holistic view of Pan-European ANS performance across all key 22 performance areas, the PRC work consists of tasks complementary to those of the Performance 23 Review Body of the Single European Sky performance scheme. They include: 24

- production of annual ATM cost-effectiveness (ACE) Benchmarking reports which present yearly 25 factual data and analysis on cost-effectiveness and productivity for Air Navigation Service 26 Providers (ANSPs) in Europe; 27

- involvement in international benchmarking studies to foster discussions on how to improve the 28 air navigation system for the benefit of all users and to support the International Civil Aviation 29 Organization (ICAO) in establishing common principles and related guidance material for ANS 30 performance benchmarking; 31

- provision of in-depth analysis and independent ad-hoc studies on ATM performance either on 32 the PRC’s own initiative or at the request of interested parties; 33

- basic R&D into the development of performance measurement; 34

- investigation of how performance could be best described/measured in the long-term; 35

- development of possible future performance indicators and metrics; and, 36

- identification of future improvements in performance. 37

In order to allow easier access and to make information available more quickly, the PRC has 38 developed its online reporting tools. 39

More information on the PRC quarterly online ANS performance review as well as information on 40 studies, performance methodologies and data for monitoring ANS performance in the 41 EUROCONTROL area is available online at: http://www.ansperformance.eu/prcq. 42

43

http://ansperformance.eu/references/acronym/ans.html

http://ansperformance.eu/references/acronym/ace.html

http://www.ansperformance.eu/prcq

PRR 2016 - Chapter 1: Introduction

2

1.1.2 Report scope and structure 1

Unless otherwise indicated, PRR 2 2016 relates to the calendar 3 year 2016 and refers to ANS 4 performance in the airspace 5 controlled by the 41 Member 6 States of EUROCONTROL (see 7 Figure 1-1), here referred to as 8 “EUROCONTROL area”. 9

In 2016, EUROCONTROL signed 10 agreements with Israel and 11 Morocco with a view to fully 12 integrating the two States into 13 the agency’s working structures 14 and also to include them in 15 future performance reviews. 16

17

18

PRR 2016 addresses the Key 19 Performance Areas: Capacity, Cost Effectiveness, Efficiency, Environmental sustainability and Safety. 20

It is organised in five chapters: 21

Chapter 1- Introduction and context: General context including a high level review of air traffic demand and punctuality trends in the EUROCONTROL area.

Chapter 2 – Safety: Review of Safety ANS performance in terms of accidents, ATM-related incidents and the level of safety occurrence reporting in the EUROCONTROL area.

Chapter 3 - En-route ANS performance: Review of operational en-route ANS performance (ATFM delays, en-route flight efficiency), including a detailed review of the most constraining ACCs in 2016.

Chapter 4 - ANS performance @ airports: Review of the operational ANS Performance of the top 30 airports in terms of traffic in 2016.

Chapter 5 - ANS Cost-efficiency: Analysis of ANS cost-efficiency performance in 2015 (the latest year for which actual financial data were available) and performance outlook, where possible.

22

Although there is no dedicated Environmental chapter in this year’s PRR, the PRC acknowledges that 23 sustainable development is an important political, economic and societal issue and the aviation 24 industry has a responsibility to minimise its global and local environmental impact. 25

In PRR 2016, the environmental component of ANS performance is addressed indirectly in Chapters 3 26 and 4 as it is closely linked to operational performance (ANS-related inefficiencies in terms of fuel 27 and CO2 emissions). 28

The environmental impact of ANS performance can generally be divided into the impact on: (1) 29 global climate, (2) local air quality, and (3) noise at airports. The PRC is presently evaluating 30 possibilities how to better address the ANS-related contribution towards environmental 31 sustainability in future publications. 32

Figure 1-1: EUROCONTROL States (2016)

TR

UA

FR

FI

ES

SE

IT

DE

PL

NO

RO

GB

BG

IE

GRPT

AT HU

RS

CZ

LT

LV

GE

BA

EE

CH

HR

SK

NLBE

MD

SI

AL

AM

DK

MKME

CY

LU

MT

MC

MAIL

EUROCONTROL 2016

Comprehensive Agreement States

http://ansperformance.eu/references/definition/eurocontrol_area.html


3

Figure 1-4: Traffic growth by Air Navigation Service Provider

Lower Airspace

Traffic growth vs. 2015

<= -2.5%

-2.5% - 0%

0% - 2.5%

2.5% - 5%

> 5%

1.2 European air transport key indices 1

On average, air traffic in the EUROCONTROL area (ESRA08) continued to increase for the third year in 2 a row in 2016 and reached the pre-3 economic crisis level of 2008. 4

At system level, air traffic increased by 5 2.4% which corresponds to an additional 6 681 flights per day on average. 7

The observed growth corresponds to the 8 baseline forecast scenario (+2.4%) 9 predicted for the ESRA08 area in the 10 STATFOR 7-year forecast - Feb. 2016 11 [Ref. 2]. 12

Figure 1-3 shows the change compared to 13 2015 in terms of flight type, traffic 14 segment, flight distance and flight hours. 15

The main driver of the observed 2.4% traffic growth in 2016 was the growth in the intra-European 16 low cost traffic segment. 17

18

Figure 1-3: Year on year change versus 2015 19

Flight hours (+2.6% vs 2015) and distance (+3.2%) grew at a higher rate than flights in the 20 EUROCONTROL area which suggests an increase in average flight distance and also in average speed 21 in 2016. 22

Peak traffic load continued to rise at a higher rate than average traffic in 2016 and the 3rd quarter in 23 2016 was the highest on record. 24

September 9th 2016 was the peak day in 25 2016 with 34,024 flights. It was also the 26 2nd highest on record (27 June 2008). 27

The highest growth compared to 2015 28 was observed in Portugal (+10.5%), 29 Ireland (7.5%), Spain (+7.5%) and Poland 30 (+7.3%). 31

The most notable traffic decreases in 32 2016 were in Ukraine (-9.0%), Moldova 33 (-8.3%), Armenia (-7.8%) and Albania 34 (-7.8%). 35

Traffic growth at Area Control Centre 36 (ACC) level is analysed in more detail in 37 Chapter 4. 38

3.3%

-0.3% -5.2%

-400

-200

0

200

400

600

800

Pan

-Eu

rop

ean

Fro

m/t

o E

uro

pe

Ove

rflig

hts

Ave

rage

dai

ly f

ligh

ts

change vs. 2015 by flight type

IFR flights

10.0 M (+2.4% )

Flight hours 15.3 M (+2.6% )

Flight distance 10 867 M (+3.2% )

Avg. flight duration 91.5 min (+0.2%)

Avg. speed712 Km/h (+0.6%)

Avg. flight length 1,085 km (+0.8%)

1.6%

7.4%

-15.6%

-0.8%

0.2% 2.3%

-400

-200

0

200

400

600

800

Trad

. Sc

hed

ule

d

Low

-Co

st

Ch

arte

r

Bu

sin

ess

Avi

atio

n

Car

go

Oth

er (

incl

. mili

tary

)

Ave

rage

dai

ly f

ligh

ts

change vs. 2015 by traffic segment change vs. 2015

Source: STATFOR

Figure 1-2: Evolution of average daily flights

3.9%

5.0%

0.1%

-6.4%

0.8%3.1%

-2.7%

-0.8%1.7%

1.5%2.4%

22 000

24 000

26 000

28 000

30 000

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

Evolution of average daily flights (EUROCONTROL area)

http://ansperformance.eu/references/definition/ESRA_2008_Area.html

http://ansperformance.eu/references/acronym/acc.html

http://ansperformance.eu/references/acronym/acc.html


4

Although the relationship between “traffic complexity” and ANS performance in general is not 1 straightforward, complexity is generally a factor to be taken into account when analysing ANS 2 performance. High density can lead to a better utilisation of resources but a high structural 3 complexity entails higher ATCO workload and potentially less traffic. 4

The annual complexity score 5 shown in Figure 1-5 combines 6 traffic density (concentration of 7 traffic in space and time) and the 8 intensity of potential interactions 9 between traffic (structural 10 complexity). 11

In the EUROCONTROL area the 12 complexity score increased further 13 in 2016 and reached 6.9 minutes 14 of potential interactions with other 15 aircraft per flight hour in the 16 airspace. 17

As can be expected, the highest 18 complexity scores are observed in 19 the core area with scores notably 20 higher than the EUROCONTROL 21 area average. In Figure 1-5, the 22 complexity score is shown as an annual average and, subject to the level of seasonality in the area, 23 the complexity score may be notably higher during peak months. More information on the 24 methodology and more granular complexity data are available online at www.ansperformance.eu. 25

Traffic variability can also affect performance if not addressed with appropriate measures. It can be 26 characterised as temporal (seasonal, daily, hourly) and spatial (location of traffic in an airspace) 27 variability. Figure 1-6 provides an indication of the seasonality by comparing the peak week to the 28 average week in 2016. 29

High seasonality is traditionally observed for the classical holiday destinations in the South. 30

If traffic is highly variable and there is limited flexibility to adjust the capacity provision according to 31 actual traffic demand, the result may be poor service quality or an underutilisation of resources. 32

If addressed proactively, traffic variability can be mitigated or resolved by utilising previous 33 experience. If demand is higher at 34 weekends than during weekdays, 35 then it is possible to roster staffing 36 levels to suit. 37

Similarly, if demand is higher 38 during certain periods, for example 39 July and August, then it is possible 40 to make more operational staff 41 available by reducing ancillary 42 tasks performed by ATCOs during 43 the peak period. 44

Hence, traffic variability and 45 complexity is therefore a factor 46 that needs to be carefully 47 managed as it may have an impact 48 on productivity, cost-efficiency, 49 and the service quality provided by 50 air navigation service providers. 51

Figure 1-5: Traffic complexity score (2016)

Lower Airspace

Traffic complexity score 2016

<= 2

2 - 4

4 - 6

6 - 8

> 8

Figure 1-6: Traffic seasonality (2016)

Lower Airspace

Traffic seasonality 2016 (peak week vs avg. week)

<= 1.15

1.15 - 1.25

1.25 - 1.35

1.35 - 1.45

> 1.45

http://www.ansperformance.eu/


5

Figure 1-7: Evolution of European IFR flights (1990-2022)

-6%

-4%

-2%

0%

2%

4%

6%

8%

10%

5

6

7

8

9

10

11

12

13

19

90

19

92

19

94

19

96

19

98

20

00

20

02

20

04

20

06

20

08

20

10

20

12

20

14

20

16

20

18

20

20

20

22

% a

nn

ual

gro

wth

(b

ars)

IFR

flig

hts

(m

illio

n)

STATFOR (Feb. 2017)7-year forecast

IFR Flights in 2016: 10.0 M (+2.4%)

Source : EUROCONTROL/STATFOR

Feb. 2008 forecast Feb. 2011forecast

Figure 1-8: European air traffic indices (2008-2016)

Figure 1-9: Evolution of arrival punctuality

90

95

100

105

110

115

120

125

130

135

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

Inde

x 1

00

= 2

00

8

Sources: ACI; STATFOR (ESRA2008); CRCO

+ 2.6% Flight hours

+ 2.4% IFR flights

+ 1.4% Avg. weight (MTOW)

+4.2% En-route Service Units (CRCO area)

+5.1% Passengers (ACI)

+ 3.2% Distance

change vs. 2015 (%)

80.5%

70%

75%

80%

85%

90%

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

% o

f fl

igh

ts

Source: CODA

Share of arrivals within 15 min of scheduled time

80.5% of arrival were punctual (-1.6% pt. vs 2015)

Figure 1-7 shows the evolution of European IFR flights (ESRA08)1 since 1990 together with selected 1 traffic forecasts2. 2

The (Feb. 2017) STATFOR 7-3 year forecast [Ref. 3] has 4 been revised upwards and 5 predicts European flights 6 (ESRA08) to grow by 2.8% in 7 2017 (Low: 1.4%; High 8 4.1%). 9

The average annual growth 10 rate (AAGR) between 2015 11 and 2023 is forecast to be at 12 1.9% (Low: 0.5%; High 13 3.4%). 14

Despite the stagnation 15 following the economic crisis, air traffic demand in Europe is expected to reach 11.6 million flights by 16 2023 which is 14% more than in 2016. 17

Figure 1-8 shows the 18 evolution of European air 19 traffic indices3 between 20 2008, the year (with the 21 highest recorded traffic 22 levels before the start of the 23 economic crisis) and 2016. 24

The trend already observed 25 over the past years 26 continued also in 2016. 27 Average distance and take-28 off weight grew at a higher 29 rate than the number of 30 flights leading also to a 31 higher growth of en-route 32 service units4. 33

The high passenger load 34 factors reported over the 35 past years also continued in 36 2016 and passenger numbers 37 continued to outpace the 38 growth in flights. 39

The continued traffic growth 40 over the past three years 41 contributed to a decline of 42 service quality. 43

The share of arrivals within 44 15 minutes of scheduled 45

1

European Statistical Reference Area defined by the EUROCONTROL Statistics and Forecast Service (STATFOR). 2

STATFOR 2008 forecast (before the economic crisis), STATFOR 2011 forecast (before the start of the SES performance scheme), and the latest available STATFOR Feb. 2017 forecast.

3 Note that the individual indices can refer to slightly different geographical areas.

4 Used for charging purposes based on aircraft weight factor and distance factor.

http://ansperformance.eu/references/definition/ESRA_2008_Area.html

http://ansperformance.eu/references/acronym/statfor.html


6

Figure 1-10: ANS contribution towards departure total departure delays

5.1% ATFM en-route 6.8%

7.8% ANS-related (airport) 6.8%

2.3% ATFM (weather) 2.4%

Source: CODA

36.2%

45.7%

Departure delay (2015)

10.2 min per

departure

Reactionary delay

Turn around (airline, airport, etc.)

46.0%

35.2%

Departure delay (2016)

11.2 min per

departure

time decreased for the third consecutive year. In 2016, 80.5% of arrivals were punctual, a decrease of 1 1.6% points compared to 2015. 2

Average departure delay 3 per flight increased from 4 10.2 minutes to 11.2 5 minutes per departure in 6 2016. 7

Reactionary delay 8 originating from previous 9 flight legs continued to 10 be the main delay cause 11 followed by turn around 12 delays. 13

The network sensitivity5 to primary delays increased from 0.84 to 0.85 leading to an increase in 14 reactionary delays in relative terms in 2016. 15

The ANS contribution increased due to en-route traffic flow measures and ATFM weather related 16 delays in 2016 but decreased for airport ANS related performance. A thorough analysis of non-ANS 17 related delay causes is beyond the scope of this report. A more detailed analysis of departure delays 18 reported by airlines is available from the Central Office for Delay Analysis (CODA)6. 19

After this outline of key air transport trends in the EUROCONTROL area, the following chapters will 20 provide a detailed analysis of ANS performance in the areas of Safety (Chapter 2), Operational ANS 21 en-route performance (Chapter 3), ANS performance at airports (Chapter 4) and ANS Cost-efficiency 22 (Chapter 5). 23

24

5

Reactionary delay for each minute of primary delay. 6

The Central Office for Delay Analysis (CODA) publishes detailed monthly, quarterly, and annual reports on more delay categories (see http://www.eurocontrol.int/coda).

http://ansperformance.eu/references/acronym/coda.html

Chapter 2: Safety

PRR 2016 - Chapter 2: Safety 7

2 Safety 1

Note that 2016 preliminary (P) data will become available in April 2017.

The chapter will be updated accordingly before the final version of the report is published.

2 SYSTEM TREND (AST REPORTING) 2015 2016(P) Trend % change

Accidents and incidents

Total number of reported Accidents with ATM Contribution 1

Total number of reported Serious Incidents

Total number of reported ATM incidents 23 654

Occurrences not severity classified 11%

Separation Minima Infringements (SMI)

Total number reported 2 338

Share of Severity A+B 10.3%

Runway incursions (RI)



Unauthorised penetration of airspace (UPA)



ATM Specific Occurrences


Share of Severity AA+A+B 2.7% 3

2.1 Introduction 4

This chapter reviews the Air Navigation Services (ANS) safety performance of the EUROCONTROL 5 Member States between 2007 and 2016 (note that 2016 data is only preliminary). 6

Sections 2.2 and 2.3 in this Chapter show the trends in ANS-related accidents and incidents in the 7 EUROCONTROL area. Section 2.4 provides an analysis of the current status of safety data reporting 8 and investigation in EUROCONTROL Member States while Section 2.5 addresses acceptable Levels of 9 Safety Performance (ALoSP). 10

The review of ANS safety performance in this chapter is based on accident and incidents data 11 reported to EUROCONTROL via the Annual Summary Template (AST) reporting mechanism and 12 complemented with additional sources of information when necessary. 13

Since 1997, the PRC has used data from the AST reporting mechanism for the analysis of accidents 14 and incidents. Complementary to the AST data, from 2013 to 2016, the PRC has also analysed safety 15 data using the European Central Repository (ECR) safety occurrence database, on a trial basis. 16 However, in this year’s report, the review of ANS safety performance is again entirely based on data 17 reported via the AST reporting mechanism as it is presently considered to be complete as it covers all 18 Member States. 19

2.2 Accidents 20

Safety is clearly the primary objective of ANS. However, not all accidents can be prevented by ANS 21 and there are a number of accidents without ANS involvement. 22

Figure 2-1 shows the total number of air traffic accidents in the EUROCONTROL area between 2011 23 and 2016, based on AST data submitted by the EUROCONTROL Member States. The data was cross 24 checked and supplemented with the available information from the ICAO Accident/Incident Data 25 Reporting (ADREP). 26

The analysis covers accidents involving aircraft above 2,250 kg maximum take-off mass (MTOM), 27 irrespective of whether the ATM domain contributed to the event or not. 28

http://ansperformance.eu/references/acronym/ast.html

PRR 2016 - Chapter 2: Safety

8

In 2015, there were 91 accidents 1 in the EUROCONTROL area of 2 which 18 were fatal. Similarly to 3 2014 this represents 4 approximately the 20% of total 5 accidents. 6

The majority of ANS-related 7 accidents between 2014 and 8 2016 were related to ‘Collisions 9 on the ground between aircraft 10 and vehicle/person/obstruction’ 11 and Controlled Flight into Terrain 12 (CFIT). 13

Unfortunately, almost three 14 quarters of the reported 15 accidents were put in the 16 category ‘Other’ hence the real 17 picture might be different if these 18 were coded differently. 19

To improve situation in the 20 future, the EUROCONTROL 21 DPS/SSR Safety Analysis Team will 22 provide further support to 23 Member States in order to 24 improve the quality of accident 25 coding in the national databases. 26

2.2.1 Air traffic accidents with ATM Contribution 27

There was only one reported 28 accident with direct ATM 29 contribution in 2015 which is the 30 same number as in 2014. 31

Due to the slight increase in total 32 accidents in 2015, the share of 33 accidents with ATM contribution 34 (direct or indirect) decreased 35 from 1.2% to 1.1% in 2015. 36

The accident with direct ATM 37 contribution in 2015 was a non-38 fatal ground collision. 39

In 2016 (based on preliminary 40 data) there were X reported 41 accidents with direct7 or indirect8 42 ATM contribution. 43

7 Where at least one ATM event or item was judged to be DIRECTLY in the causal chain of events leading to an

accident or incident. Without that ATM event, it is considered that the occurrence would not have happened. 8 Where no ATM event or item was judged to be DIRECTLY in the causal chain of events leading to an accident or

incident, but where at least one ATM event potentially increased the level of risk or played a role in the emergence of the occurrence encountered by the aircraft. Without such ATM event, it is considered that the accident or incident might still have happened.

Figure 2-1: Accidents in EUROCONTROL area (2011-16P) [TBU]

Figure 2-2: Accidents risk distribution (2012-16P) [TBU]

12 17 16 17 18

65

71

49

6873

16%19%

25%20% 20%

0102030405060708090

100

2011 2012 2013 2014 2015 2016 (P)

Fatal accidents Non fatal accidents % of Fatal Accidents

Total air traffic accidents - fixed wing, weight >2250kg MTOM) (EUROCONTROL area)

91 air traffic accidents (+ 6 vs 2014)

20% fatal accidents (+- 0% vs 2014)

0.9%

48.2%

13.6%

36.4%

0.9%

0% 10% 20% 30% 40% 50%

Collisions btn. airborne a/c andvehicle/another a/c on the ground

Collisions on the ground between a/cand vehicle/person/obstruction(s)

Collisions on the ground betweenaircraft

CFIT

Mid-Air Collisions

Relative Risk Importance (%)

Risk Distribution in the EUROCONTROL area (2011-2015)

Figure 2-3: Accidents with ATM contribution in the EUROCONTROL area

(2011-16P) [TBU]

1 0 1 0 3 0 1 1

3

2 2

5.1%

2.7%

1.5%

3.2%3.4%

1.2% 1.1%

0

1

2

3

4

5

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016(P)

Accidents with indirect ATM contribution

Accidents with direct ATM contribution

% of accidents with direct or indirect ATM contribution in total accidents

Accidents with ATM contribution - fixed wing, weight >2250kg MTOM)

(EUROCONTROL area)

1 accident with direct ATM contribution (+/- 0 vs 2014)

1.1% of total accidents (- 0.1% vs 2014)


9

2.3 Incidents 1

This section provides a review of ATM-related incidents, reported through the EUROCONTROL AST 2 reporting mechanism. The PRC has made use of, with gratitude, the data provided by the 3 EUROCONTROL DPS/SSR Unit and EUROCONTROL Safety Regulation Commission (SRC) Annual and 4 intermediate Reports [Ref. 4]. As opposed to accidents analysis, there is no Maximum Take Off 5 Weight (MTOW) limit (2,250 kg) for the ATM-related incidents. 6

The analysis concentrates on the several key risk occurrence types, namely: separation minima 7 infringements (SMIs), runway incursions (RIs), airspace infringements (AIs)/unauthorised 8 penetrations of airspace (UPAs), and ATM Specific Occurrences (ATM-S). 9

Overall, based on the AST reports submitted by 40 EUROCONTROL Member States, there was a 4.4% 10 increase in the total number of incidents reported in comparison with 2016. 11

Table 2-1 shows the EUROCONTROL area overall occurrence rates (as reported by all 40 reporting 12 States) for SMI, RI and UPAs in 2015. 13

Table 2-1: Occurrence rates (SMI, RI, UPA) in the EUROCONTROL area (2015) 14

2015 Rate of SMIs (per 100,000 flight hours)

Rate of RIs (per 10,000 movements)

Rate of UPAs (per 100,000 flight hours)

EUROCONTROL Area 15 0.8 28

Figure 2-4 shows the underlying distribution of occurrence rates of all 40 reporting EUROCONTROL 15 Member States for three categories of occurrences SMI, RI and UPAs compared to the 16 EUROCONTROL area overall rate. 17

18 Figure 2-4: Occurrence rates EUROCONTROL area (2016P) [TBU] 19

In 2015, the EUROCONTROL area SMI rate was approximately 15 SMI per 100 000 flight hours with 20 only a few States having a very high SMI occurrence rate (4 States are above the 90th percentile). 21

A similar picture can be observed for RIs and UAPs. The distribution is skewed with a small number of 22 States with high occurrence rates compared to the rest of the States. 23

At EUROCONTROL level, there was less than 1 reported RI per 10,000 movements in 2015. For UAPs, 24 the occurrence rate was approximately 28 reported UPAs per 100,000 flight hours in 2015. 25

However, similarly to the rate of SMIs, the rate of UPAs shows substantial differences among 26 Member States; and few States have extremely high UPA rates (4 States are above 90th percentile). 27

The next four figures illustrate the trends of SMI, RI, UPAs, and ATM-S occurrences in the period 28 2007-2016 (preliminary), detailing the evolution of the number of reporting States, the total number 29 of occurrences reported per each category and especially the evolution of risk-bearing (Severity AA/A 30 and Severity B) occurrences in each figure. 31

0 10 20 30 40 50 60SMIs per 100 000 flight hours

Separation Minima Infringement(SMI) distribution in EUROCONTROL Member States (2015)

EUROCONTROL area rate

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0RIs per 10 000 movements

Runway incursion (RI) distribution in EUROCONTROL Member States (2015)

0 20 40 60 80 100 120 140 160 180UPAs per 100 000 flight hours

Unauthorised penetration of airspace (UPA) distribution in EUROCONTROL Member States (2015)

0

10

20

30

40

50

60

70

Scal

e

Maximum value

Upper Quartile (75th percentile)

Median(50th percentile)

Lower Quartile (25th percentile)

Minimum value


10

Figure 2-5: Reported high-risk SMIs (EUROCONTROL)

Despite an increase in traffic, the number of reported risk bearing SMIs (Severity A+B) decreased in 2015 from 273 to 241.

Overall, 10% of all occurrences reported in 2015 were categorised as risk bearing occurrences which is 2% less than in 2014.

2016P figures to be included when available

Figure 2-6: Reported high-risk UPAs (EUROCONTROL)

The number of risk bearing UPA occurrences (Severity A+B) increased notably from 60 to 88 in 2015.

As a result, the share of risk bearing UPA occurrences in the total reported UPAs increased from 1% to 2% in 2015.


Figure 2-7: Reported high-risk RIs (EUROCONTROL)

The reported risk bearing runway incursions (Severity A+B) decreased from 100 to 94 in 2015.

At the same time, the share of risk bearing runway incursions remained at 7% of the total reported RI occurrences in 2015.


Figure 2-8: Reported high-risk ATM Spec. Occurrences

(EUROCONTROL)

The total number of risk bearing ATM specific occurrences decreased from 463 to 453 in 2015 (-2.2%).

At the same time, the total number of reported ATM Specific Occurrences increased notably which was mainly due to changes in data submission in one Member State.


2007 2008 2009 2010 2011 2012 2013 2014 20152016

(P)

N° of States reporting 28 29 30 31 33 36 36 40 40

Total n° reported 1 567 1 711 1 418 1 402 1 571 1 796 2 097 2 359 2 338

Severity B 295 236 141 178 217 258 232 250 221

Severity A 70 56 27 16 35 33 30 23 20

% : Proportion of Severity A+B 23% 17% 12% 14% 16% 16% 12% 12% 10%

70 5627 16 35 33 30 23 20

295

236

141 178

217258

232 250221

23%

17%

12%14%

16% 16%

12%12%

10%

0

100

200

300

400

Nu

mb

er o

f o

ccu

rren

ces

Separation Minima Infringements

Severity B Severity A % : Proportion of Severity A+B

Source: EUROCONTROL

2007 2008 2009 2010 2011 2012 2013 2014 20152016

(P)


Total n° reported 2 416 2 797 3 336 3 381 4 742 5 010 3 435 4 325 4 392

Severity B 49 52 53 79 68 50 37 51 76

Severity A 5 3 6 4 12 10 1 9 12

% : Proportion of Severity A+B 2% 2% 2% 2% 2% 1% 1% 1% 2%

53

64 12 10 1 9 12

49 5253

79 68

50

37

51

76

2%

2%2%

2%

2%

1% 1%

1%

2%

0

20

40

60

80

100

Nu

mb

er o

f O

ccu

rren

ces

Unauthorised Penetration of Airspace


Source: EUROCONTROL

2007 2008 2009 2010 2011 2012 2013 2014 20152016(P)


Total n° reported 885 926 1 094 1 385 1 399 1 234 1 421 1 442 1 397

Severity B 44 40 36 77 62 37 61 74 83

Severity A 12 14 15 22 23 12 14 26 11

% : Proportion of SeverityA+B

6% 6% 5% 7% 6% 4% 5% 7% 7%

12 14 15 22 2312 14

2611

44 40 36

7762

37

61

7483

6%6%

5%

7%

6%

4%

5%

7% 7%

0

20

40

60

80

100

120

Nu

mb

er

of

Occ

urr

ence

s

Runway Incursions


Source: EUROCONTROL

50 49 3490 90 93

141206

398

809 799

588392 352 348

0

100

200

300

400

500

600

700

800

900

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016(P)

Nu

mb

er

of

Occ

urr

ence

s

ATM Specific Occurrences

Severity B

Severity A

Severity AA

Source: EUROCONTROL


11

2.4 Reporting and Investigation 1

This section provides a review of the quality and completeness of ATM safety occurrences 2 (operational and ATM specific occurrences) reported through the AST reporting mechanism, as 3 updated in September 2016 (where applicable). 4

2.4.1 Total number of reported occurrences 5

The final 2015 data were received 6 from 40 EUROCONTROL Member 7 States. 8

The number of reported 9 occurrences increased by 4.5% in 10 2015. 11

[2016 trend to be included based on 12 preliminary results] 13

14

. 15

Nevertheless, the available data 16 does not allow conclusions to be 17 drawn if the observed year-on-year 18 change represents a genuine safety 19 performance variation or if it is due 20 to different reporting levels. 21

2.4.2 Unclassified or undetermined occurrences 22

Figure 2-10 shows the number of ATM-related incidents not severity classified or with severity 23 classification not determined (Severity D) for different occurrences categories. The analysis is based 24 on the data submitted via AST in April 2016, covering the reporting year 2015 (final) and 2016 25 (preliminary). 26

In 2015, 11% of reported 27 occurrences were still not 28 severity classified. If the 29 occurrences where the 30 severity is ‘not determined’ 31 are added (i.e. insufficient 32 data provided to fully assess 33 the severity), the percentage 34 rises to just above 16%. 35

Considering each type of 36 occurrence separately (not 37 just SMIs, RIs and UPAs), the 38 percentage varies between 2% 39 and 13%. If the occurrences 40 where the severity is “not determined” (i.e. some data provided but not enough to fully assess the 41 severity) are also included, the range increases to 3% and 23% of total number of reported 42 occurrences in each occurrence category. 43

A decrease of the percentage of occurrences not severity classified for all types of occurrences was 44 reported in 2014 but UPAs and RIs increased again in 2015. Although the overall trend is promising, 45 considering the fact that the application of the severity classification based on the Risk Analysis Tool 46 (RAT) methodology to the reporting of occurrences is a key safety performance indicator of the 47 Single European Sky (SES) Performance Scheme, further actions are needed to ensure the gap is 48 closed. 49

Figure 2-9: Reported occurrences (2007-2016P)

Figure 2-10: Severity not classified or not determined (2007-2016P) [TBU]

0

500

1 000

1 500

2 000

2 500

3 000

0%

5%

10%

15%

20%

25%

30%

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15 To

tal n

um

ber

of

occ

ure

nce

s n

ot

seve

rity

cla

ssif

ied

% b

y ca

tego

ry

Occurrences NOT Severity Classified

SMI RI UPASource: EUROCONTROL


12

Although the number of unclassified or not determined incidents is still higher than in 2006/7, there 1 has been a notable improvement. 2

As already pointed out in several previous reports, the situation needs to be monitored as the quality 3 and completeness of safety data can impact the outcome of the analysis at European and national 4 level, the sustainability of the human reporting system9 and can also have other potential 5 downstream repercussions such as the inadequate prevention of similar incidents or inadequate 6 sharing and dissemination of lessons learned. 7

2.4.3 Completeness of safety data 8

Figure 2-11 shows the typical fields that are either left blank or marked Unknown in the AST, 9 submitted by the EUROCONTROL Member States. 10

ATM contribution = direct; indirect; none

Airspace Class = Class of airspace: A,B,C,D,E

Flight Rules = IFR or VFR

Phase of Flight = taxi, take-off, climb to cruise, cruising, approach

Traffic of Flight = General Air Traffic, Commercial, Military

Type operation = GAT or OAT

Figure 2-11: Completeness of AST reported data in 2016(P) [TBU] 11

It is of concern that a large share of the data required to populate a number of fields is still missing. 12 This lack of completeness of AST data hampers comprehensive safety analysis at European level. 13

14

15

2.5 Acceptable Level of Safety Performance (ALoSP) 16

In last year’s PRR (June 2016) [Ref. 5], the PRC raised the concern that the definition and guidance on 17 the development of the Acceptable Level of Safety Performance (ALoSP) concept (as defined by 18 ICAO) is currently not available in Europe. 19

As the ICAO requirements for ALoSP leave room for interpretation in choosing the best way to 20 implement the concept, the EUROCONTROL Member States could demonstrate leadership in filling 21 such a gap by developing a harmonised approach. A common approach to measuring and managing 22 safety performance will ultimately ensure a harmonised implementation of State Safety Programmes 23 (SSPs) and facilitate the exchange of safety information in the future. 24

Due to the importance of this issue, the Provisional Council (PC) of EUROCONTROL, at its 45th Session 25 (June 2016) therefore requested the PRC to review the implementation status of the ALoSP and to 26 report back to the PC/47 (June 2017). 27

28

29

9 When ATCOs or pilots provide safety reports, if feedback is not provided it can have an adverse impact on

the motivation to report.

0%

20%

40%

60%

80%

100%

ATM

Co

ntr

ibu

tio

n

Air

spac

e C

lass

Flig

ht

Ru

les

Ph

ase

of

Flig

ht

Typ

e o

f Fl

igh

t

Typ

e o

fO

per

atio

ns

Completness of the AST reported data (2015P)

% Empty +Unknown

% Empty

% Unknown

Source: EUROCONTROL

http://ansperformance.eu/references/acronym/alosp.html


13

2.5.1 ALoSP implementation survey 1

The objectives of the ALoSP implementation survey were as follows: 2

review the current level of ALoSP implementation in EUROCONTROL Member States; 3

present the different approaches used in EUROCONTROL Member States; 4

identify common problems faced in the implementation of ALoSP; 5

identify and share existing best practices so that others can also adopt and implement such 6 an approach; and 7

present a proposal for a way forward in order to support the implementation of ALoSP in 8 EUROCONTROL Member States in a harmonised way. 9

In summary, the purpose of the ALoSP survey was to achieve a deeper and more comprehensive 10 understanding of the ALoSP concept and its implementation in EUROCONTROL Member States, in 11 terms of concept definition, scope, and implementation challenges. 12

The findings will be included in the final version of PRR 2016 in April when the results of the survey 13 will be available. 14

2.6 Conclusions 15

The Safety conclusion will be included in the final version when the preliminary 2016 data and the 16 results from the ALoSP survey are available. 17

Chapter 3: Operational En-route ANS Performance

PRR 2016 - Chapter 3: Operational En-route ANS Performance 14

3 Operational en-route ANS Performance 1

SYSTEM TRENDS 2016 Trend change vs. 2015

IFR flights controlled 10.0M +2.4%

Capacity

En-route ATFM delayed flights 4.8% +0.9 %pt.

Average en-route ATFM delay per flight (min.) 0.86 +0.13 min

Total en-route ATFM delay (min.) 8.7M +20.9%

Environment/ Efficiency

Average horizontal en-route efficiency (flight plan) 95.4% -0.1%pt

Average horizontal en-route efficiency (actual) 97.1% -0.2%pt.

3.1 Introduction 2

Despite the slowdown following the economic crisis in 2008, European air traffic is forecast to reach 3 14.4 million flights by 2035, which is 50% more than in 2012 [Ref.7]. As the airspace is finite, there is 4 a need to increase the operational efficiency of the air navigation system to be able to accommodate 5 future traffic demand, including new airspace user groups such as Remotely Piloted Aircraft Systems 6 (RPAS). 7

The ICAO Global Air Navigation Plan (GANP) and the European ATM Master Plan both aim at 8 improving the air navigation system through a harmonised set of ATM enhancements which provide 9 operational improvements and which make use of existing avionics capabilities. 10

Continuous review helps to monitor the impact of enhancement initiatives on performance over time 11 in order to better understand progress and success of the initiatives and to highlight problems in the 12 current system. 13

This chapter reviews operational en-route ANS performance in the EUROCONTROL area in 2016. 14

Section 3.2 describes the main changes in air traffic demand by air traffic service provider in 2016 15 before Section 3.3 analyses ANS-related flight efficiency constraints on airspace users’ flight 16 trajectories, including en-route ATFM delays and horizontal and vertical flight efficiency. Civil military 17 cooperation and coordination is addressed in Section 3.4. 18

The performance indicators used for the analysis in this chapter, expected benefits and supporting 19 initiatives are shown in Table 3-1. 20

21

Table 3-1: Operational en-route ANS performance (Overview) 22

En-route ANS performance

Expected benefits • Reduce delay and fuel burn (CO2 emissions)

• Improve route network design;

• Improved route availability (CDRs);

• Improved airspace utilisation (civil/military coordination);

Related indicators in this chapter

• En-route ATFM delays;

• Horizontal en-route flight efficiency;

• Vertical en-route flight efficiency

Supporting projects/ initiatives

• Free route airspace (FRA)

• Route network design improvements

• Flexible use of airspace (FUA)

• Enhanced flow performance through network operational planning

23

PRR 2016 - Chapter 3: Operational En-route ANS Performance

15

Figure 3-2: Traffic growth by ACC (2016)

CAN

ANK

LIS

SCO

SHA

TAMSTO

MAD

BRE

ROM

BOD

DNI

WARLON

BAR

BUC

MAA

MAR

ATH+MAK

KIE

MAL

MAL

NIC

BOR

SEV

BRI

ODE

OSL-STV

COP

SOF

LVO

TBI

ZAG

RIG

VIL

BEOMIL

BUD

TAL

PAD

WIE

KAR PRA

KAR

SAJ

MUN

REI

TIR

PAR

CHIBRAT

ZURYER

GEN

SKO

LJU

PAR

PAR

KOS

LAN

BREM

AMS

BRU

DUB

LON TC

PAL

Lower Airspace

Lower Airspace

Traffic growth vs. 2015

<= -2.5%

-2.5% - 0%

0% - 2.5%

2.5% - 5%

> 5%

IST

Lower Airspace

3.2 Traffic evolution 1

The 2.4% traffic increase in the EUROCONTROL area in 2016 was not homogenous throughout the 2 network. Of the 39 ANSPs included in the analysis, 25 showed an increase in traffic compared to 14 3 ANSPs which showed a traffic decline. 4

Figure 3-1 shows the number of average daily flights by ANSP in 2016 at the bottom and the change 5 compared to 2015 in absolute (blue bars) and relative (red dots) terms at the top. The figure is sorted 6 according to the absolute change compared to the previous year. 7

8 Figure 3-1: Traffic variation by ANSP (2016/2015) 9

In absolute terms, ENAIRE (Spain), NATS (UK), and DSNA (France) experienced the highest year on 10 year growth in 2016. DHMI (Turkey), UKSATSE (Ukraine) and ROMATSA (Romania) reported the 11 highest absolute decrease in 2016. 12

The traffic growth by Area 13 Control Centres (ACCs) in 14 Figure 3-2 confirms the 15 contrasted picture already 16 observed at ANSP level in 17 Figure 3-1. 18

ACCs with growth rates above 19 10% in 2016 were Palma, 20 Lisbon, Canarias, and Dublin 21 ACC. 22

It is remarkable that 35 of the 23 63 ACCs reported their highest 24 traffic levels on record in 25 2016, surpassing the 26 previously highest levels 27 dating back before the start of 28 the economic crisis in 2008. 29

30

01 0002 0003 0004 0005 0006 0007 0008 0009 000

10 000

ENA

IRE

(Sp

ain

)

NA

TS-C

on

tin

enta

l (U

K)

DSN

A (

Fran

ce)

MU

AC

(M

aast

rich

t)

DFS

(G

erm

any)

NA

V-C

on

tin

enta

l (P

ort

uga

l)

PA

NSA

(P

ola

nd

)

AN

S C

R (

Cze

ch R

epu

blic

)

IAA

(Ir

elan

d)

ENA

V (

Ital

y)

Hu

nga

roC

on

tro

l-EC

(H

un

gary

)

LPS

(Slo

vaki

a)

LVN

L (N

eth

erla

nd

s)

Skyg

uid

e (S

wit

zerl

and

)

SMA

TSA

(Se

rbia

an

d…

LFV

(Sw

eden

)

NA

VIA

IR (

Den

mar

k)

MA

TS (

Mal

ta)

Slo

ven

ia C

on

tro

l (Sl

ove

nia

)

EAN

S (E

sto

nia

)

AN

A L

UX

(Lu

xem

bo

urg

)

Saka

ero

nav

igat

sia

(Geo

rgia

)

Oro

Nav

igac

ija (

Lith

uan

ia)

DC

AC

Cyp

rus

Bel

goco

ntr

ol (

Bel

giu

m)

Cro

atia

Co

ntr

ol (

Cro

atia

)

Au

stro

Co

ntr

ol (

Au

stri

a)

LGS

(Lat

via)

Fin

avia

(Fi

nla

nd

)

AR

MA

TS (

Arm

enia

)

Mo

ldA

TSA

(M

old

ova

)

Avi

no

r (N

orw

ay)

M-N

AV

(FY

RO

M)

BU

LATS

A (

Bu

lgar

ia)

HC

AA

(G

recc

e)

Alb

con

tro

l (A

lban

ia)

RO

MA

TSA

(R

om

ania

)

UkS

ATS

E (U

krai

ne)

DH

MI (

Turk

ey)

Avg

. dai

ly f

ligh

ts (

20

16

)

7.5

%

5.5

%

4.2

%

4.3

%

2.3

%

10

.5%

7.2

%

6.6

%

7.5

%

2.6

%

3.9

% 6.1

%

4.3

%

1.5

%

2.1

%

1.6

%

1.5

%

7.1

%

2.6

%

2.8

% 6.4

%

3.3

%

1.3

%

0.7

%

0.3

%

0.3

%

0.1

%

0.4

%

0.1

%

-7.9

%

-8.3

%

-1.0

%

-4.6

% -1.5

%

-2.0

%

-7.8

%

-2.5

%

-9.0

%

-1.6

%

-10%

-5%

0%

5%

10%

15%

20%

25%

30%

35%

40%

-100

-50

0

50

100

150

200

250

300

350

400

450

Ch

ange

vs.

pre

vio

us

year

(%

)

Ch

ange

vs.

20

15

(ab

solu

te)

change vs. 2015

Source: NM; PRC analysis

Average daily flights (2016)


16

3.3 ANS-related flight efficiency constraints (en-route) 1

This section evaluates ANS-related flight efficiency constraints on airspace users’ flight trajectories. It 2 addresses several performance areas including efficiency (time, fuel), predictability, and 3 environmental sustainability (emissions, noise). 4

3.3.1 En-route ATFM delays 5

Please note that software release 20.0 of the Network Manager on 04 April 2016 introduced a change to improve the accuracy of the ATFM delay calculation for operational purposes which resulted in an estimated overall reduction of 11.8% of delay compared to the old methodology. More information on the change is available online at www.ansperformance.eu.

Total en-route ATFM delays, for the EUROCONTROL area, increased by +20.9% in 2016 which 6 corresponds to 0.86 minutes (51 seconds) of en-route ATFM delay per flight (0.73 in 2015). 7

8 Figure 3-3: Average en-route ATFM delay (EUROCONTROL area) 9

According to the delay classifications, as reported by the local flow management positions (FMPs), 10 Capacity/Staffing related issues remain by far the main driver of en-route ATFM delays (55.3%), 11 followed by weather related delays (18.3%), ATC disruptions/ industrial actions (12.3%), and Event 12 related delays (9.1%) which also include delays due to ATC system upgrades. 13

14 Figure 3-4: En-route ATFM delayed flights and delay per delayed flight (EUROCONTROL area) 15

Following the increase observed already for the past two years, the number of flights affected by 16 ATFM en-route delays in the EUROCONTROL area increased further in 2016 from 3.9% to 4.8%. At the 17 same time, the delay per delayed flight decreased from 18.8 minutes to 18.0 minutes in 2016. 18

1.43

2.03

0.53 0.610.73

0.86

0.0

0.5

1.0

1.5

2.0

2.5

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

ATFM

de

lay

per

flig

ht

(min

ute

s)

En-route Airport

Source: PRU Analysis

Evolution of en-route/ airport ATFM delay per flight

(EUROCONTROL area)

Evolution of en-route ATFM delay (EUROCONTROL area)

56.4% of total ATFM delay (+5.0% pt. vs. 2015)

8.7 M min of en-route ATFM delay (+20.9%) 867 M Euro est. en-route ATFM delay costs (+20.9%)

4.8 M min (55.3%) ATC capacity and staffing related (+9.4%) 1.6 M min (18.3%) en-route weather related (+55.4 %)

1.1 M min (12.3%) ATC disruption/ strike related (+42.9%) 0.8 M min (9.1%) en-route special event related (+10.2 %)

0.86 minutes en-route ATFM delay per flight (+0.13 vs. 2015)

0.0 1.0 2.0 3.0 4.0

ATC Capacity (ERT)

ATC Staffing (ERT)

Weather (ERT)

ATC Disruptions (ERT)

Events (ERT)

Reroutings (ERT)

Disruptions (ERT)

En-route ATFM delays (million minutes)

Total en-route ATFM delay by reported cause

(EUROCONTROL area)

2015 result

3.9

%4

.8%

2.0

%2

.3%

0.5

%0

.8%

0.2

%0

.3%

0.1

%0

.1%

0.5

%0

.8%

0.1

%0

.2%

0.4

%0

.5%

18.015.4 16.0

39.0

14.7

20.7 21.317.0

0

10

20

30

40

50

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

4.5%

5.0%

20

12

20

13

20

14

20

15

20

16

20

12

20

13

20

14

20

15

20

16

20

12

20

13

20

14

20

15

20

16

20

12

20

13

20

14

20

15

20

16

20

12

20

13

20

14

20

15

20

16

20

12

20

13

20

14

20

15

20

16

20

12

20

13

20

14

20

15

20

16

20

12

20

13

20

14

20

15

20

16

En-route ATFMdelay

ER Capacity(ATC)

ER Staffing(ATC)

ER Disruptions(ATC)

ER Reroutings ER Weather ER Disruptions ER Events

ATF

M d

elay

per

del

ayed

flig

ht

(min

)

Shar

e o

f en

-ro

ute

ATF

M d

elay

ed f

ligh

ts (

%)

Source: PRC Analysis; Network Manager

Evolution of en-route ATFM delayed flights and average delay per delayed fight

(EUROCONTROL area)

18.0 minutes delay per en-route delayed flight4.8% of en-route ATFM delayed flights (+0.9% points vs 2015)

3.2% of airport ATFM delayed flights (+0.2% points vs 2015)

ATFM delay per en-route delayed flight (Eurocontrol area)En-route

ATFM delayed

4.8%

Airport ATFM

delayed3.2%

No ATFM delay92.0%

2016


http://ansperformance.eu/references/definition/atfm_delay.html

http://ansperformance.eu/references/definition/atfm_delay_codes.html


17

ATC capacity/staffing related en-route ATFM delays accounted for more than half of all en-route 1 ATFM delays. In 2016, 3.3% of the flights were delayed due to ATC capacity or staffing related ATFM 2 regulations, an increase of 0.5% on 2015. 3

4 Figure 3-5: Estimated ATC capacity/staffing related impact on airline operations (2016) 5

Figure 3-6 shows the impact of weather related en-route ATFM delays on airline operations. In 2016, 6 weather related en-route ATFM delays accounted for 18.3% of all en-route ATFM delays delaying 7 1.9% of the flights. More than half of the weather related delay in 2016 was concentrated in DFS and 8 Maastricht UAC. 9

10 Figure 3-6: Impact of weather related en-route ATFM delays on airline operations (2016) 11

ATC disruptions accounted for 12.3% of all en-route delays, almost entirely attributable to DSNA. 12

13 Figure 3-7: Estimated ATC strike related impact on airline operations (2016) 14

0

200

400

600

800

1000

1200

1400

JAN

FEB

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SEP

OC

T

NO

V

DEC

Impact of ATC capacity/staffing related ATFM delays on airline operations (2016)

ATC capacity

ATC staffing

16.3 minutes delay per delayed flight

4.79 M min of en-route ATFM delay (+410k vs. 2015)

3.3% of flights delayed (+0.5% vs. 2015)

479 million est. delay costs (+41 vs. 2015)

ATC capacity/staffing-related ATFM delay by month ('000 min)

Share of total capacity/staffing-related delay by service provider (%)

55.3% of total en-route ATFM delay 361 days with capacity-related ATFM delay (+5 vs. 2015)

35.8%

12.2% 11.8% 9.9%6.6% 5.5%

0%

10%

20%

30%

40%

DSN

A

ENA

IRE

DFS

MU

AC

NA

TS(C

on

tin

enta

l) PA

NSA

0100200300400500600700800

JAN

FEB

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SEP

OC

T

NO

V

DEC

Impact of weather-related en-route ATFM delays on airline operations (2016)


1 104 days of en-route ATFM delay (+394 vs. 2015)



Weather-related ATFM delay by month ('000 min)

Share of total weather-related delay by service provider (%)

18.3% of total en-route ATFM delay 175 days with weather related ATFM delay (+18 vs. 2015)

26.9%24.4%

19.5%

4.4%

13.1%

0%

10%

20%

30%

DFS

Maa

stri

cht

DSN

A

NA

TS(C

on

tin

enta

l) EN

AIR

E

Oth

er

0

50

100

150

200

20-0

3-2

01

6

21-0

3-2

016

31-0

3-2

016

26-0

1-2

016

28-0

6-2

016

23-0

6-2

016

19-0

5-2

016

26-0

5-2

016

14-0

6-2

016

28-0

4-2

016

05-0

7-2

016

02-0

6-2

016

15-0

9-2

016

Oth

er

Estimated ATC strike related impact on airline operations (2016)

≈ 13 000 est. flight cancellations

942k min of en-route ATFM delay (+419 vs. 2015)



ATC strike related ATFM delay by day ('000 min)

Share of total ATC strike related delay by service provider (%)

12.3% of total en-route ATFM delay


26 days with strike related en-route ATFM delay (+4 vs. 2015)

Brest

Paris

Marseille

Bordeaux

99.6%

0.3%0%

20%

40%

60%

80%

100%

DSNA (France) ENAV (Italy) Other


18

Although only 0.3% of the flights were affected by ATFM delays due to ATC industrial action, the 1 average delay per delayed flight (due to ATC industrial action) of 37.8 minutes caused substantial 2 disruption in the network. Moreover the estimated number of cancellations due to ATC industrial 3 action was 13 000 flights in 2016. 4

The share of special event related delay was 9.1% in 2016 and 0.5% of the flights were impacted 5 with an average delay per delayed flight of 17.1 minutes. Almost 70% of the delay was due to the 6 ERATO implementation in French ACCs. 7

8

Figure 3-8: Estimated special event related impact on airline operations (2016) 9

New or upgrades of ATM systems are planned in a large number of States over the coming years.

The ERATO implementation in France over the past two years showed the substantial impact that airspace and/or equipment changes can have on the network.

As voiced already in PRR2015, it is vital that ANSPs effectively coordinate the planning and implementation of all changes to the ATM system that could adversely affect operations with the Network Manager.

Figure 3-9: Planned major project implementations (2017-2021)

Whilst such changes are inevitable, and indeed desirable, airspace users need to be assured that all 10 appropriate measures have been taken to reduce disruption, and that there will be an operational 11 benefit to the users following the implementation. 12

Most constraining ACCs in 2016 13

While capacity constraints can occur from time to time, air navigation services should not generate 14 high delays on a regular basis. Figure 3-10 shows the most constraining10 ACCs in 2016 by ANSP. 15

In 2016, the most constraining ACCs in 2016 accounted for 69.8% of all en-route ATFM delays and 16 26.3% of total flight hours controlled in Europe. Compared to 2015, Lisbon, Athens, and Zagreb ACCs 17 notably improved their performance and are therefore no longer among the most constraining ACCs. 18

10 The selection threshold was set at more than 30 days with significant en-route ATFM delay (>1 min per flight).

0

50

100

150

200

250

JAN

FEB

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SEP

OC

T

NO

V

DEC

Impact of special event-related en-route ATFM delays on airline operations (2016)


786k min of en-route ATFM delay (+72 vs. 2015)



Special event-related ATFM delay by month('000 min)

Share of total special event-related delay by service provider (%)

9.1% of total en-route ATFM delay 193 days with special event related ATFM delay (+40 vs. 2015)

Brest

Bordeaux

Prestwick Langen

68.2%

21.7%

8.7%1.5%

0%

20%

40%

60%

80%

DSN

A

NA

TS(C

on

tin

enta

l)

DFS

Oth

er


19

Brussels, Bordeaux, Prestwick, Maastricht, Karlsruhe, Warsaw and Marseille ACCs are new among the 1 most constraining ACCs in 2016. 2

In 2016, DSNA (France) generated 41.6% of all en-route ATFM delays in the EUROCONTROL area with three ACCs among the most constraining ACCs (Brest ACC, Bordeaux ACC and Marseille AC). Overall, 5.8% of all flights crossing airspace controlled by DSNA experienced en-route ATFM delay with an average delay per delayed flight of 20.4 minutes.

Figure 3-10: Overview of most constraining ACCs (2016)

The PRC also note that a considerable amount of en-route delays (48k minutes, circa 5%) have been 3 recorded in France but without assignment to one of the existing ACC / UACs (see Figure 3-10). 4 Instead these delays have been grouped under the label LFDSNA referring to all French ACCs. 5

The PRC understands that this is the result of a trial to improve cooperation and coordination 6 between individual ACCs but de-linking the ATFM delay from specific locations risks losing the ability 7 to identify, and therefore resolve, the root causes of capacity constraints. Even though the 48k 8 minutes allocated to LFDSNA may be due to constraints at a small number of specific capacity 9 bottlenecks, if these bottlenecks are not identifiable, they cannot be resolved, and will continue to 10 constrain airspace users. 11

Whilst efforts to improve cooperation and coordination among ANSPs, with the objective of 12 improving the service provided to airspace users, should be encouraged; it is essential to be able to 13 accurately identify specific capacity constraints and the impact such constraints have on air traffic. 14 15

Brest ACC continued to generate significant delays due to the implementation of the ERATO system, 16 until April 2016. (Original planning for implementation of the ERATO system published in NOP 2014 17 and NOP 2015 envisaged capacity reductions for a limited period of 1-2 months only). 18

Capacity levels increased from April 2016 and July 2016 saw Brest ACC handling the highest monthly 19 traffic levels on record, albeit with high delays (285k minutes). 20

21 Figure 3-11: Brest ACC en-route performance overview (2016) 22

There were 23 days in July 2016 when delays in Brest ACC exceeded 2 minutes per flight. The table 23 below shows, for the three main sector groups: North, South and East, the 5 days with the highest 24

Brest

Bordeaux

Marseille

Other

Karlsruhe Barcelona

Canarias

Prestwick

41.6%

13.0%11.4%

9.0% 8.1%

3.4% 3.3% 2.4% 7.9%

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

DSN

A

DFS

Maa

stri

cht

ENA

IRE

NA

TS(C

on

tin

enta

l)

PA

NSA

Be

lgo

con

tro

l

DC

AC

Cyp

rus

All

Oth

er A

CC

s

Shar

e o

f to

tal e

n-r

ou

te A

TFM

del

ay in

20

16

Warsaw Brussels Nicosia

Not assigned to a geographic

location

Most constraining ACCs by service provider in 2016(ACCs with more than 30 days of average delay >1 min per flight are highlighted in blue)

0

2 000

4 000

6 000

8 000

10 000

JAN

-20

16

FEB

-20

16

MA

R-2

01

6

AP

R-2

01

6

MA

Y-2

01

6

JUN

-20

16

JUL-

20

16

AU

G-2

01

6

SEP

-20

16

OC

T-2

01

6

NO

V-2

01

6

DEC

-20

16

Ave

rage

dai

ly

Monthly en-route ATFM delay and traffic

All other causes

Industrial action 'I'

Weather 'W'

Staffing 'S'

Capacity 'C'

IFR flights

Source: PRU analysis

20.1% of total en-route ATFM delay in 2016

19.7 min delay per delayed flight (- 0.3min)8.9% of flights ATFM delayed (+1.9% vs. 2015)

166 days of en-route ATFM delay >1 min. (+ 39d)

1207 days of generated en-route ATFM delay (+300d) 173.8 million Euro est. delay costs (+43m)

6.3% growth vs. 2015 (Forecast: H 6.6% - B 5.2% - L 3.8%)

7.1 interactions per flight hour (complexity avg: 6.9) 28% higher traffic in peak week (vs. avg. week)

Brest ACC en-route performance overview (2016)

0%

1%

2%

3%

4%

5%

6%

7%

8%

1-5

16

-20

31

-35

46

-50

61

-65

76

-80

91

-95

10

6-1

10

12

1-1

25

13

6-1

40

15

1-1

55

16

6-1

70

18

1-1

85

19

6-2

00

21

1-2

15

22

6-2

30

ho

urs

per

yea

r (%

)

flights per hour

2008 2016

2015

Requ. ref. capacity 2016 (NOP): 219 Planned capacity 2016 (NOP): 200

Evolution of hourly throughput


20

delays in July, the maximum number of sectors opened and the period for which this capacity was 1 provided. 2

Table 3-2: ATFM regulations applied by Brest ACC (July 2016) 3

Date Sector Group

Planned sectors at maximum

capacity (NOP)

Highest number sectors actually

opened

Time of operation at

highest config. (hh:mm)

Period of regulations due to ATC capacity.

(hh:mm)

Delay due ATC capacity

(minutes)

Overlap btw. ATC capacity regulation and deployment of highest cap. on that day

(hh:mm)

(ACC delay per flight)

01/07 North 6 6 3:00 2:20 1106 0 0%

(5.3) South 6 5 5:30 11:20 8843 5:30 49%

East 6 6 8:30 12:17 6596 6:20 52%

02/07 North 6 5 5:30 12:40 7087 5:30 43%

(5.4) South 6 6 2:00 6:00 4426 0:00 0%

East 6 6 6:30 12:50 8261 6:30 43%

05/07 North 6 4 Industrial Action

(4.0)

South 6 3

East 6 4

15/07 North 6 5 5:30 7:40 5333 4:20 57%

(4.5) South 6 5 4:00 8:40 5164 2:00 23%

East 6 6 4:00 9:55 7003 4:00 40%

16/07 North 6 6 1:00 8:00 4593 0:00 0%

(4.3) South 6 5 6:30 11:55 5829 6:10 52%

East 6 6 2:00 15:10 6319 1:40 11%

The delays on Tuesday 5th July were due to industrial action and an associated reduction in the 4 numbers of sectors available. 5

The above table raises two concerns. Firstly, even though the demand levels were high and massive 6 delays were accruing, there was an inability or refusal to open the maximum number of sectors. 7 Secondly, there are significant mismatches between the deployment of maximum capacity and the 8 traffic demand, evidenced by the necessity to apply regulations for lengthy periods when only a 9 limited number of sectors are opened. 10

The Provisional Council, in recommendations from PRR 2014 and PRR 2015, highlighted the need for 11 capacity to be made available during peak traffic periods rather than regulating demand to meet 12 reduced capacity. 13

Bordeaux ACC saw an increase in traffic over 2015 levels (+5.4%) and recorded the highest traffic 14 level on record. Industrial disputes were responsible for delays in every month, from January until 15 July, except February. 16

17 Figure 3-12: Bordeaux ACC en-route performance overview (2016) 18

Delays attributed to ATC Capacity prevailed from May until September peaking in July at almost 94k 19 minutes of delay for 98k flights, approximately 1 minute per flight. 20

Delays were also attributed to adverse en-route weather phenomena from May to September 21 peaking again in July at 31k minutes of delay. 22

0

2 000

4 000

6 000

JAN

-20

16

FEB

-20

16

MA

R-2

01

6

AP

R-2

01

6

MA

Y-2

01

6

JUN

-20

16

JUL-

20

16

AU

G-2

01

6

SEP

-20

16

OC

T-2

01

6

NO

V-2

01

6

DEC

-20

16

Ave

rage

dai

ly


All other causes


Weather 'W'

Staffing 'S'

Capacity 'C'

IFR flights








Bordeaux ACC en-route performance overview (2016)

0%

1%

2%

3%

4%

5%

6%

7%

8%

1-5

16

-20

31

-35

46

-50

61

-65

76

-80

91

-95

10

6-1

10

12

1-1

25

13

6-1

40

15

1-1

55

16

6-1

70

18

1-1

85

19

6-2

00

21

1-2

15

22

6-2

30

ho

urs

per

yea

r (%

)

flights per hour

200820162015




21

November saw the beginning of implementation of the ERATO system (as previously implemented in 1 Brest ACC) with a reduction in capacity. Following the experiences in Brest ACC, the DSNA, the 2 Network Manager, and adjacent ACCs worked together to reduce the impact of the ERATO 3 implementation. Action such as mandatory rerouting and off-loading into adjacent ACCs / ANSPs 4 reduced the traffic demand below normal operational levels. 5

Marseille ACC handled 4.7% more traffic in 2016 than in 2015. Delays attributed to industrial action 6 made up 40.5% of the total delays in Marseille ACC during 2016, 85% of which occurred in March. 7

8 Figure 3-13: Marseille ACC en-route performance overview (2016) 9

Nicosia ACC showed a significant capacity improvement in 2016. July August and September saw 10 higher traffic levels with significantly lower delays than in 2015. However, Nicosia continued to be a 11 bottleneck in the European network and previously published capacity plans were not implemented 12 as had been envisaged. 13

14 Figure 3-14: Nicosia ACC en-route performance overview (2016) 15

The NOP promised the availability of 6 ATC sectors during peak periods but this never materialised. 16 The highest number of sectors opened was 5, although this is an improvement on the maximum of 4 17 sectors, provided during the same time in 2015. 18

Nicosia ACC operated five sectors for a total of 38 hours over 21 separate days in July; 75 hours over 19 28 days in August and 66 hours over 25 days in September. The inability to open the maximum 20 number of sectors during peak traffic periods indicates that staffing needs to be addressed, both in 21

0

2 000

4 000

6 000

JAN

-20

16

FEB

-20

16

MA

R-2

01

6

AP

R-2

016

MA

Y-2

01

6

JUN

-20

16

JUL-

20

16

AU

G-2

01

6

SEP

-201

6

OC

T-2

01

6

NO

V-2

01

6

DEC

-20

16

Ave

rage

dai

ly


All other causes


Weather 'W'

Staffing 'S'

Capacity 'C'

IFR flights



24.9 min delay per delayed flight (+- 0.0min)1.8% of flights ATFM delayed (+1.0% vs. 2015)





Marseille ACC en-route performance overview (2016)

0%

1%

2%

3%

4%

5%

6%

7%

8%

1-5

16

-20

31

-35

46

-50

61

-65

76

-80

91

-95

10

6-1

10

12

1-1

25

13

6-1

40

15

1-1

55

16

6-1

70

18

1-1

85

19

6-2

00

21

1-2

15

22

6-2

30

24

1-2

45

25

6-2

60

27

1-2

75

28

6-2

90

30

1-3

05

ho

urs

per

yea

r (%

)

flights per hour

200820162015



0

2 000

JAN

-20

16

FEB

-20

16

MA

R-2

01

6

AP

R-2

01

6

MA

Y-2

01

6

JUN

-20

16

JUL-

20

16

AU

G-2

01

6

SEP

-20

16

OC

T-2

01

6

NO

V-2

01

6

DEC

-20

16

Ave

rage

dai

ly


All other causes


Weather 'W'

Staffing 'S'

Capacity 'C'

IFR flights



17.1 min delay per delayed flight (- 3.6min)3.7% of flights ATFM delayed (-8.2% vs. 2015)

72 days of en-route ATFM delay >1 min. (- 149d)

142 days of generated en-route ATFM delay (-405d) 20.4 million Euro est. delay costs (-58m)

0.7% growth vs. 2015 (Forecast: H 2.4% - B 0.5% - L -1.3%)


Nicosia ACC en-route performance overview (2016)

0%

5%

10%

15%

20%

25%

ho

urs

per

yea

r (%

)

flights per hour

2008 2016

2015




22

terms of the recruitment of new area controllers and in deploying the existing controllers in a more 1 efficient manner. 2

It is notable that, despite the significant increase in aircraft-carrier-based military flight operations in 3 the eastern Mediterranean in 2016, no ATFM delays in the Nicosia FIR were attributed to military 4 activity. 5

Finally, the PRC notes the growth in traffic to and from Israel and that a significant portion of this 6 traffic will, most likely, seek to fly through the Nicosia FIR. This underlines the necessity of planning 7 and implementing additional capacity, as soon as possible, to meet the traffic demand. 8

Brussels ACC 75% of the ATFM delays from Brussels ACC in 2016 was attributed to staffing reasons, 9 predominantly in the period April to July. Traffic levels remained reasonably stable with a traffic 10 growth of 0.2%. 11

12 Figure 3-15: Brussels ACC en-route performance overview (2016) 13

Barcelona ACC traffic increased dramatically from 2015 levels during 2016 (+8.4%). July and August 14 saw over 98 thousand flights per month, the highest monthly totals in Barcelona on record. The 15 number of days when delay was more than one minute per flight rose from 37 in 2015 to 49 in 2016. 16 Even though August had a slightly higher number of flights, the amount of delay was significantly less 17 than in July (80k compared to 101k minutes). In July delays attributed to capacity were 95% of the 18 total value compared with 90% for the month of August. 19

20 Figure 3-16: Barcelona ACC en-route performance overview (2016) 21

0

2 000

4 000

JAN

-20

16

FEB

-20

16

MA

R-2

01

6

AP

R-2

01

6

MA

Y-2

01

6

JUN

-20

16

JUL-

20

16

AU

G-2

01

6

SEP

-20

16

OC

T-2

01

6

NO

V-2

01

6

DEC

-20

16

Ave

rage

dai

ly


All other causes


Weather 'W'

Staffing 'S'

Capacity 'C'

IFR flights








Brussels ACC en-route performance overview (2016)

0%1%2%3%4%5%6%7%8%9%

10%

ho

urs

per

yea

r (%

)

flights per hour

2008 2016

2015



0

2 000

4 000

JAN

-20

16

FEB

-20

16

MA

R-2

01

6

AP

R-2

01

6

MA

Y-2

01

6

JUN

-20

16

JUL-

20

16

AU

G-2

01

6

SEP

-20

16

OC

T-2

01

6

NO

V-2

01

6

DEC

-20

16

Ave

rage

dai

ly


All other causes


Weather 'W'

Staffing 'S'

Capacity 'C'

IFR flights





282 days of generated en-route ATFM delay (+39d) 40.6 million Euro est. delay costs (+5.6m)



Barcelona ACC en-route performance overview (2016)

0%1%2%3%4%5%6%7%8%9%

10%

1-5

16

-20

31

-35

46

-50

61

-65

76

-80

91

-95

10

6-1

10

12

1-1

25

13

6-1

40

15

1-1

55

16

6-1

70

18

1-1

85

19

6-2

00

21

1-2

15

22

6-2

30

24

1-2

45

ho

urs

per

yea

r (%

)

flights per hour

2008 2016

2015




23

An analysis of the days in July and August when total ATFM delay was greater than 2 minutes per 1 flight, as in PRR2015, shows the following: 2

Table 3-3: ATFM regulations applied by Barcelona ACC (July/August 2016) 3

Date

Sector Group

Planned sectors at maximum capacity (NOP)

Highest number sectors actually opened

Time of operation at highest configuration (hh:mm)

Period of regulations due to ATC capacity. (hh:mm)

Delay due ATC capacity (minutes)

Overlap between ATC capacity regulations and deployment of highest capacity on that day (hh:mm)

02/07 West 6 6 14:30 14:50 2671 12:50 87%

East 6 6 14:30 12:00 4449 9:20 78%

22/07 West 6 6 15:00 5:00 803 5:00 100%

East 6 6 11:00 10:00 3431 5:00 50%

23/07 West 6 6 15:00 11:00 2932 9:10 83%

East 6 6 15:00 9:40 3068 8:20 87%

31/07 West 6 6 15:00 3:40 1314 3:40 100%

East 6 6 15:00 9:00 3179 6:40 74%

13/08 West 6 6 15:00 5:00 1534 5:00 100%

East 6 6 15:00 8:40 3732 8:40 100%

In comparison to 2015, the ANSP provides the maximum number of sectors for much longer periods- 4 up to 15 hours. This is a significant improvement, especially in Sector Group East (which was usually 5 restricted to deployment of maximum sectors for less than 8 hours in 2015.) 6

The above table show that there is still room for further improvement in making sure that capacity is 7 deployed according to the traffic demand instead of rigidly providing capacity independently of 8 traffic demand. However, the predominant issue for Barcelona ACC appears to be the necessity of 9 providing additional capacity. 10

Failure to plan and implement adequate capacity for Barcelona ACC has been flagged by the Network 11 Manager in each Network Operations Plan since 2012. 12

Prestwick ACC experienced a traffic growth in 2016 (+6.6%) which was notably higher than forecast 13 (+2.9%). ATFM en-route delays in Prestwick ACC peaked during June and July with the primary reason 14 being the implementation of, and training associated with, a new iTec (interoperability Through 15 European Collaboration) air traffic management system. Performance improved notably in the 16 second half of 2016 and following the successful implementation of the new system no further 17 constraints are expected in 2017. 18

Maastricht UAC also achieved the highest traffic level on record in 2016. En-route weather was 19 responsible for significant portions of delay in May (57%), June (67%), July (39%) and August (25%). 20

21 Figure 3-17: Maastricht UAC en-route performance overview (2016) 22

23

0

2 000

4 000

6 000

8 000

10 000

12 000

JAN

-20

16

FEB

-20

16

MA

R-2

01

6

AP

R-2

01

6

MA

Y-2

01

6

JUN

-20

16

JUL-

20

16

AU

G-2

01

6

SEP

-20

16

OC

T-2

01

6

NO

V-2

01

6

DEC

-201

6

Ave

rage

dai

ly


All other causes


Weather 'W'

Staffing 'S'

Capacity 'C'

IFR flights



15.1 min delay per delayed flight (+0.2min)3.7% of flights ATFM delayed (+1.4% vs. 2015)





Maastricht UAC en-route performance overview (2016)

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

10%

1-5

21

-25

41

-45

61

-65

81

-85

10

1-1

05

12

1-1

25

14

1-1

45

16

1-1

65

18

1-1

85

20

1-2

05

22

1-2

25

24

1-2

45

26

1-2

65

28

1-2

85

30

1-3

05

32

1-3

25

34

1-3

45

36

1-3

65

ho

urs

per

yea

r (%

)

flights per hour

2008 2016

2015




24

Figure 3-18: Maastricht UAC traffic evolution (2010-2016)

100

110

120

130

140

150

160

170

180

JAN

MA

Y

SEP

JAN

MA

Y

SEP

JAN

MA

Y

SEP

JAN

MA

Y

SEP

JAN

MA

Y

SEP

JAN

MA

Y

SEP

JAN

MA

Y

SEP

2010 2011 2012 2013 2014 2015 2016

Flig

hts

(' 0

00

)

Monthly and average annual traffic evolution(Maastricht UAC )

Maastricht UAC allocated a high level of delays to adverse en-route weather phenomena during the 1 May to August period, much greater than in previous years. Adverse en-route weather phenomena 2 such as severe icing, severe turbulence, thunderstorms etc. usually necessitate the publication of 3 SIGMET (Significant Meteorological information) advising aircraft of the occurrence or expected 4 occurrence of specified en-route weather phenomena which may affect the safety of aircraft 5 operations. 6

Closer investigation of the delays attributed to adverse weather correlates with the publication of 7 SIGMETs for one or more of the FIRs in 8 which MUAC provide air traffic services: 9 Brussels FIR (EBBU), Amsterdam FIR 10 (EHAA) and Hannover UIR (EDYY). 11

The traffic growth in Maastricht UAC was 12 above the high traffic forecast which led 13 to higher traffic levels than previously 14 handled. 15

Figure 3-18 shows that Maastricht UAC is 16 handling higher levels of monthly traffic 17 year on year. This underlines the 18 importance of ensuring that capacity 19 plans are implemented in sufficient time 20 to handle the ever growing traffic levels. 21

22 23 Warsaw ACC: Following a 7.2% increase of traffic on 2015, Warsaw reached a traffic level never 24 achieved before. Delays more than doubled (+127%) and the number of days when en-route ATFM 25 delay was greater than 1 minute per flight increased from 4 in 2015 to 39 for 2016. A dramatic rise in 26 delays occurred in July with peak traffic (72k flights), and continued, albeit at a smaller levels, until 27 November. 72% of delays are attributed to staffing issues. 28

Further investigation of days with high delay in July 2016 reveals an inability to open the maximum 29 number of sectors (10) for lengthy periods of high demand, or even at all. 30

31

Figure 3-19: Warsaw ACC en-route performance overview (2016) 32

33

34

0

2 000

4 000

JAN

-20

16

FEB

-20

16

MA

R-2

01

6

AP

R-2

01

6

MA

Y-2

01

6

JUN

-20

16

JUL-

20

16

AU

G-2

01

6

SEP

-20

16

OC

T-2

01

6

NO

V-2

01

6

DEC

-20

16

Ave

rage

dai

ly


All other causes


Weather 'W'

Staffing 'S'

Capacity 'C'

IFR flights



15.0 min delay per delayed flight (+-0.0min)2.8% of flights ATFM delayed (+1.5% vs. 2015)





Warsaw ACC en-route performance overview (2016)

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

10%

1-5

11

-15

21

-25

31

-35

41

-45

51

-55

61

-65

71

-75

81

-85

91

-95

10

1-1

05

11

1-1

15

12

1-1

25

13

1-1

35

14

1-1

45

15

1-1

55

16

1-1

65

17

1-1

75

ho

urs

per

yea

r (%

)

flights per hour

2008 2016

2015




25

Figure 3-21: Karlsruhe UAC traffic evolution (2010-2016)

8090

100110120130140150160170180

JAN

MA

Y

SEP

JAN

MA

Y

SEP

JAN

MA

Y

SEP

JAN

MA

Y

SEP

JAN

MA

Y

SEP

JAN

MA

Y

SEP

JAN

MA

Y

SEP

2010 2011 2012 2013 2014 2015 2016

Flig

hts

(' 0

00

)

Monthly and average annual traffic evolution(Karlsruhe UAC )

Karlsruhe UAC: Similarly to Maastricht UAC, the majority of en-route ATFM delays were attributed to 1 adverse weather phenomena, particularly during June and July 2016. 2

3 Figure 3-20: Karlsruhe UAC en-route performance overview (2016) 4

As with Maastricht UAC, closer examination of the delays allocated to adverse weather correlates 5 with the publication of SIGMETs for the 6 Rhein UIR, wherein Karlsruhe UAC 7 provides air traffic services. 8

The traffic growth was slightly above the 9 high forecast and as a consequence 10 Karlsruhe UAC serviced more flights than 11 ever before. 12

Figure 3-21 shows how Karlsruhe UAC is 13 handling higher levels of monthly traffic 14 year on year. This underlines the 15 importance of planning sufficient capacity 16 to meet ever growing traffic levels. 17

18

Canarias ACC experienced 10.3% growth in traffic levels over 2015, which was above the predicted 19 high forecast (8%) and the highest annual level on record so far. 20

21 Figure 3-22: Canarias ACC en-route performance overview (2016) 22

0

2 000

4 000

6 000

8 000

JAN

-20

16

FEB

-20

16

MA

R-2

01

6

AP

R-2

01

6

MA

Y-2

01

6

JUN

-20

16

JUL-

20

16

AU

G-2

01

6

SEP

-20

16

OC

T-2

01

6

NO

V-2

01

6

DEC

-20

16

Ave

rage

dai

ly


All other causes


Weather 'W'

Staffing 'S'

Capacity 'C'

IFR flights



14.7 min delay per delayed flight (-1.1min)2.4% of flights ATFM delayed (+1.3% vs. 2015)





Karlsruhe UAC en-route performance overview (2016)

0%1%2%3%4%5%6%7%8%9%

10%

1-5

21

-25

41

-45

61

-65

81

-85

10

1-1

05

12

1-1

25

14

1-1

45

16

1-1

65

18

1-1

85

20

1-2

05

22

1-2

25

24

1-2

45

26

1-2

65

28

1-2

85

30

1-3

05

32

1-3

25

34

1-3

45

36

1-3

65

38

1-3

85

ho

urs

per

yea

r (%

)

flights per hour

2008 2016

2015



0

2 000

JAN

-20

16

FEB

-201

6

MA

R-2

01

6

AP

R-2

01

6

MA

Y-2

01

6

JUN

-20

16

JUL-

20

16

AU

G-2

01

6

SEP

-20

16

OC

T-2

01

6

NO

V-2

01

6

DEC

-20

16

Ave

rage

dai

ly


All other causes


Weather 'W'

Staffing 'S'

Capacity 'C'

IFR flights



20.6 min delay per delayed flight (-5.3min)1.8% of flights ATFM delayed (+0.8% vs. 2015)





Canarias ACC en-route performance overview (2016)

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

1-5

6-1

01

1-1

51

6-2

02

1-2

52

6-3

03

1-3

53

6-4

04

1-4

54

6-5

05

1-5

55

6-6

06

1-6

56

6-7

07

1-7

57

6-8

08

1-8

58

6-9

09

1-9

59

6-1

00

10

1-1

05

10

6-1

101

11

-11

5

ho

urs

per

yea

r (%

)

flights per hour

2008 2016

2015




26

Figure 3-23: Traffic and ATFM delay by weekday – Canarias ACC (2016)

0

20

40

60

80

100

120

MO

ND

AY

TUES

DA

Y

WED

NES

DA

Y

THU

RSD

AY

FRID

AY

SATU

RD

AY

SUN

DA

Y

mo

vem

en

ts p

er

ho

ur

Traffic and delay by weekday - Canarias ACC (2016)

Plan (2016) Required reference (2016)


Weekend flights are 6 times more likely to be en-route ATFM delayed

79% of all en-route ATFM delays occurs on weekends (69 % on Saturdays)

0

1

2

3

4

5

6

7

500 700 900 1100 1300

Avg

. en

-ro

ute

ATF

M d

elay

per

flig

ht

Daily flights

Weekdays

Weekends

Figure 3-24: ATFM performance (network indicators)

4%

6%

8%

10%

12%

14%

16%

18%

20%

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

% of take offs outside ATFM slot tolerance window

% regulated hrs. with actual demand/capacity >110%(excess demand)% of ATFM delays due to avoidable regulations (no excessdemand) Source: Network Manager

Out of 36 days where delays were greater than one minute per flight: 24 were Saturdays; 3 were 1 Tuesdays, 3 were Fridays; 2 2 Thursdays 2 Mondays, 3 and 2 Sundays. 4

Comparison of the 5 capacity performance on 6 Saturdays in November 7 and December 2016 8 alone shows that the 9 highest amount of traffic 10 occurred on 17th 11 December although there 12 were relatively fewer 13 delays than on the other 14 Saturdays in December. 15

Closer analysis shows 16 that the orientation of 17 the runways-in-use in the Canarias has a significant impact on the en route capacity performance. 18 When northerly runways are in use (17th December) the en route capacity performance is 19 significantly better than when traffic is landing / departing in a southerly direction. 20

The aerodrome charts for airports in the Canarias shows a significant mismatch in location and type 21 of runway exits for traffic landing in a northerly direction compared to traffic landing in a southerly 22 direction. Factors such as location and type of runway exits influence the landing, and departure, rate 23 which can create congestion in the TMA and further upstream into the en-route sectors. 24

Mis-identification of causal capacity constraints hinders mitigation and resolution of capacity 25 problems. If capacity constraints are due to the lack of rapid-exit-taxiway in southerly landing 26 configuration then allocating the delay as being due to en route ATC capacity will not lead the airport 27 authorities to build a new taxiway. 28

Similarly, if a TMA does not have sufficient holding patterns to accommodate traffic holding for the 29 airports it serves, allocating the delay as being due to ATC capacity in the en-route sectors will not 30 lead to the creation and use of suitable holding patterns through a TMA redesign project. 31

32

ATFM performance (network level) 33

The ATFM function in Europe is jointly executed by local ATFM units and the Network Manager 34 (central unit for ATFM). ATFM regulations are put 35 in place by the Network Manager to protect en-36 route sectors or airports from receiving more 37 traffic than ATC can safely handle upon request 38 of the local Flow Management Positions (FMP). 39

Figure 3-24 shows the evolution of the three 40 high-level indicators presently in use to monitor 41 the performance of the ATFM function at system 42 level. 43

In 2016, ATFM slot adherence continued to 44 improve and the regulated hours with excess 45 demand also decreased slightly. Following the 46 notable improvement in 2015, ATFM delays due 47 to avoidable regulations increased again to 2014 48 levels. 49

50


27

3.3.2 En-route Flight Efficiency 1

This section evaluates en-route flight efficiency in Pan-European airspace. En-route flight efficiency 2 has a horizontal (distance) and vertical (altitude) component and is the result of numerous 3 interactions between stakeholders with different objectives and constraints. More information on 4 methodologies (approach, limitations) and data for monitoring the ANS-related performance is 5 available online at www.ansperformance.eu. 6

There is a close link between operational efficiency and environmental sustainability. Improved flight 7 efficiency has not only an economic impact in terms of fuel savings but also an impact in terms of 8 reduced emissions (most notably carbon dioxide (CO2)) impacting on the environment. 9

With air traffic expected to double by 2035 and the airspace being finite, there is a need to make the 10 ATM system more efficient to keep up with demand and to reduce operational inefficiencies as much 11 as possible. However, as pointed out in previous reports, 100% flight efficiency cannot be achieved 12 for a number of reasons including, inter alia, safety, weather and capacity issues. 13

In view of the numerous factors and complexities involved, and with traffic levels growing again, 14 flight efficiency improvements will become more and more challenging and will require the joint 15 effort of all involved parties, coordinated by the Network Manager. 16

3.3.2.1 Horizontal en-route flight efficiency 17

Please note that the scale of the horizontal flight efficiency metric has been changed so that it now shows the level of efficiency instead of the level inefficiency. The underlying methodology remained unchanged.

Figure 3-25 shows the horizontal en-route flight efficiency for the actual trajectory and the filed flight 18 plan for the EUROCONTROL area11. 19

While remaining at very high levels (the 100% level is a theoretical value), after a continuous 20 improvement over the past years, the value of horizontal flight efficiency slightly decreased in 2016 21 compared to 2015. At Pan-European level, horizontal flight efficiency in filed flight plans decreased 22 from 95.5% in 2015 to 95.4% in 2016. At the same time, the efficiency of actual trajectories 23 decreased stronger from 97.3% to 97.1% in 2016. 24

25 Figure 3-25: Horizontal en-route flight efficiency (Pan-European level) 26

11 The Pan-European airspace analysed in this section refers to the CFMU area.

Jan

-16

Feb

-16

Mar

-16

Ap

r-1

6

May

-16

Jun

-16

Jul-

16

Au

g-1

6

Sep

-16

Oct

-16

No

v-1

6

De

c-1

6

Flight Plan Actual trajectory

30 day mov. avg. (Flight Plan) 30 day mov. avg. (Actual trajectory)

Horizontal en-route flight efficiency (Pan-European level)

Source: PRC Analysis

ATC industrial action2016

97.1% flight efficiency in actual flown trajectories (-0.2% pt. vs. 2015)

95.4% flight efficiency in flight plans (-0.1% pt. vs. 2015)

95

.4%

95

.4%

95

.5%

95

.5%

95

.4%

96

.9%

97

.0%

97

.4%

97

.3%

97

.1%

93%

94%

95%

96%

97%

98%

99%

100%

20

12

20

13

20

14

20

15

20

16

effi

cien

cy (

%)

Flight Plan

Actual trajectory



28

Figure 3-26: Flight efficiency by State (2016)

91%

92%

93%

94%

95%

96%

97%

98%

99%

100%

91% 92% 93% 94% 95% 96% 97% 98% 99% 100%

Flig

ht

pla

n e

ffic

ien

cy (

%)

Trajectory (actual) efficiency (%)

FRA Full implementation (H24)

Other

Gap between actual and plan

Free route airspace (FRA) benefits on flight efficiency (2016)

1.6 %pt. higher average flight efficiency in FRA States (2016)

1.0 %pt. smaller gap between flight plan and actual in FRAStates (operations closer to plan)

EUROCONTROL area

The analysis of daily values shows a weekly pattern with higher efficiency during the weekend and 1 lower efficiency during the week (which has been the subject of detailed analysis in PRR2015). 2

Although the effects of ATC industrial action on specific days in 2016 are clearly visible on the right 3 side of Figure 3-25, at Pan-European level, the annual value for horizontal flight efficiency improves 4 by merely 0.03% points if days with industrial action are removed from the analysis. 5

A possible indirect reason for the deterioration is linked to the rising congestion, leading to more and 6 more cases in which the trade-off between length of the trajectory and delay is solved in favour of 7 longer trajectories to avoid congested airspace. 8

With the current route network to a large extent designed on a structure based on ground-based 9 navigation aids, technological developments on board of new aircraft have outpaced the way the 10 current ANS system is operated resulting in a sub optimal utilisation of the aircraft capabilities. The 11 implementation of Free Route Airspace (FRA), which would now be possible throughout the entire 12 EUROCONTROL area, gives the aircraft 13 operators more freedom in the choice 14 of the flight plan and the possibility to 15 avoid some of the restrictions imposed 16 by a rigid route network. This leads to 17 a more flexible environment which 18 responds more dynamically to changes 19 in traffic flows. 20

Although flight efficiency will never be 21 100%, the benefits that the 22 implementation of FRA can bring in 23 terms of flight efficiency gains and 24 resulting reductions in costs, fuel burn 25 and emissions are substantial. 26

Figure 3-26 shows the level of flight 27 efficiency in in actual trajectories (X-28 Axis) and filed flight plans (Y-Axis) by 29 State in 2016. States in which FRA is 30 available 24 hours are shown in red. 31

The benefits are clearly visible. On 32 average, States where FRA has been 33 fully implemented all day show a 1.6 34 percent point higher flight efficiency compared to the other States were FRA has not been fully 35 implemented. 36

Furthermore, it can also be seen that the gap between the flight plan efficiency and the efficiency in 37 the actual flown trajectory (the vertical distance between a point and the diagonal) is narrower than 38 for the other States (1.0 percent point smaller gap). Actual operations closer to plan improves the 39 level of predictability for all players involved with a positive impact on capacity and resource 40 utilisation. 41

The notable gap between flight plans and actual flown trajectories, which has been highlighted in 42 previous years, is clearly more prominent in States where FRA has not been fully implemented all 43 day. 44

This provides evidence that, while the inefficiencies are the result of complex interactions between 45 airspace users, ANSPs and the Network Manager, FRA enables a better match between the planning 46 and operational phase. 47

48

http://ansperformance.eu/references/acronym/fra.html


29

Figure 3-28: Local and network effects on flight efficiency by State (2016)

0

2

4

6

8

10

12

14

16

18

20

Fin

lan

dM

alta

De

nm

ark

Latv

iaSw

eden

Esto

nia

Bo

snia

-Her

zego

vin

aIr

elan

dC

roat

iaH

un

gary

Ro

man

iaG

eorg

iaP

ort

uga

l (C

on

tin

enta

l)Se

rbia

an

d M

on

ten

egro

Lith

uan

iaP

ola

nd

Bu

lgar

iaSl

ove

nia

Mo

ldo

vaFY

RO

MA

lban

iaG

reec

eN

orw

ayA

ust

ria

Slo

vaki

aC

zech

Rep

ub

licA

rmen

iaU

krai

ne

Ger

man

yN

eth

erla

nd

sFr

ance

Ital

yB

elgi

um

Spai

nC

ypru

sU

K (

Co

nti

nen

tal)

Turk

ey

Swit

zerl

and

Avg

. ad

dit

ion

al k

m p

er f

ligh

t (a

ctu

al t

raje

cto

ry)

Local and network effects on flight efficiency (2016)

Local inefficiencies within airspace - Full FRA implementation (H24)

Local inefficiencies within airspace - Other (FRA partial or No implementation)

Network effects linked to interfaces with other States or TMAs

Figure 3-27 shows the horizontal en-route flight efficiency on the actual trajectories by State for 1 2016. Those States where FRA is fully implemented all day are highlighted in red12. 2

3 Figure 3-27: Horizontal en-route flight efficiency (actual trajectory) by State (2016) 4

Flight efficiency is expressed as a ratio of total distances and is therefore not influenced by traffic 5 volume or individual flight length. The absolute values of the additional distance per flight and per 6 State provide a more complete picture and explain which States influence more the overall value for 7 the EUROCONTROL area. 8

The scatter plot on the left side of Figure 3-27 provides a link between the three quantities. It shows 9 the flight efficiency of the actual trajectory (X-axis), the average additional distance per flight (Y-axis), 10 and the total additional distance of the Member State (the size of the bubble). France combines a 11 below average flight efficiency with long average flight segments (and a high traffic volume) which 12 consequently results in a substantial amount of total additional kilometres in 2016 (the bubble for 13 the EUROCONTROL area would be the sum of all the bubbles). 14

All else being equal, if the nine States below the EUROCONTROL average could have improved the 15 flight efficiency of the actual trajectories by 0.2 percent points in 2016, the saved distance would 16 have been equivalent to 8.2 million kilometres in 2016 and flight efficiency in the EUROCONTROL 17 area would have improved by 0.1 percent points. On the other hand, the same improvement of 18 0.2 percent points by the 19 nine best States would 20 improve system wide flight 21 efficiency performance by 22 merely 0.02 percent points. 23

The Horizontal Flight 24 Efficiency methodology 25 considers the entire flight 26 extension and not local 27 Great Circle Distances. It 28 allows therefore a 29 breakdown of local and 30 network effects. 31

Figure 3-28 shows the 32 results on a per flight basis. 33

12 Please note that Italy is not shown in red as FRA was only fully implemented in early December 2016. The

resulting benefits are therefore expected to be visible in the analysis of 2017.

99

.1%

98

.9%

98

.9%

98

.8%

98

.8%

98

.7%

98

.7%

98

.6%

98

.5%

98

.5%

98

.5%

98

.4%

98

.4%

98

.4%

98

.3%

98

.3%

98

.3%

98

.2%

98

.1%

98

.0%

98

.0%

97

.9%

97

.8%

97

.8%

97

.7%

97

.7%

97

.4%

97

.4%

97

.2%

97

.0%

96

.5%

96

.4%

96

.1%

96

.1%

95

.9%

95

.8%

95

.7%

95

.1%

91%

92%

93%

94%

95%

96%

97%

98%

99%

100%

Fin

lan

dM

alta

Den

mar

kLa

tvia

Swed

enEs

ton

iaB

osn

ia a

nd

Her

zego

vin

aIr

elan

dC

roat

iaH

un

gary

Ro

man

iaG

eorg

iaP

ort

uga

l (C

on

tin

enta

l)Se

rbia

an

d M

on

ten

egro

Lith

uan

iaP

ola

nd

Bu

lgar

iaSl

ove

nia

Mo

ldo

vaTh

e fo

rme

r Yu

gosl

av…

Alb

ania

Gre

ece

No

rway

Au

stri

aSl

ova

kia

Cze

ch R

epu

blic

Arm

enia

Ukr

ain

eG

erm

any

Net

her

lan

ds

Fran

ceIt

aly

Bel

giu

mSp

ain

Cyp

rus

UK

(C

on

tin

enta

l)Tu

rkey

Swit

zerl

and

Flight efficiency (actual trajectory), additional distance per entry and total additional distance per State (2016)

EUROCONTROL area (2016)

Full FRA implementation (H24)Other (FRA partial or No implementation)

States below EUROCONTROLarea efficiency level

Sweden

RomaniaPoland

Greece

Norway

AustriaSlovakia

Ukraine

Germany

Netherlands

France

Italy

Belgium

Spain

Cyprus

UK (Continental)

Turkey

Switzerland

1

3

5

7

9

11

13

15

17

19

95% 96% 97% 98% 99% 100%

Ave

rage

ad

dit

ion

al k

ilom

eter

per

flig

ht

Flight efficiency actual trajectory (%)

Bubble size refers to the total additional kilometers flown by State


30

Figure 3-29: Example distribution of maximum filed flight levels

In general, States implementing FRA show a very low local component (the coloured part of the 1 bars), while other States present potential for reduction of those local inefficiencies. 2

There is potential for additional reduction in the length of the trajectories by reducing the network 3 component (grey part of the bars). This requires the joint effort of all involved parties, best 4 coordinated by the Network Manager. 5

According to the ATM Master Plan, Free Route Airspace on a H24 basis should be implemented 6 throughout the entire EUROCONTROL area by 2021. As highlighted in PRR2015, ANSPs should work 7 actively with the Network Manager and the Deployment Manager to deliver FRA across the entire 8 EUROCONTROL area including necessary cross-border implementation as soon as possible. 9

Research is ongoing to better understand and quantify the individual contributing factors (flight 10 planning, awareness of route availability, civil-military coordination, etc.) in order to identify and 11 formulate strategies for future improvements. A crucial prerequisite for the development of a better 12 understanding is the collection of better data on the activation of special use airspace and on route 13 availability when the flight plan was submitted by airspace users (shortest available route). 14 15

3.3.2.2 Vertical en-route flight efficiency 16

In order to address a growing stakeholder interest to better evaluate the vertical component of flight 17 efficiency, this section presents a first evaluation of vertical en-route flight efficiency. Because of the 18 distinct nature of the different phases of flight, specific methodologies were developed for the 19 analysis of vertical flight efficiency during climb and descent on the one hand and for the analysis of 20 en-route vertical flight efficiency on the other hand. More information on methodologies is available 21 online at www.ansperformance.eu. 22

The focus of the following section is on the en-route phase rather than on the climb and descent 23 phases which is addressed in more detail in Chapter 4 of this report. It is also important to point out 24 that the analysis in this section does not aim at quantifying the total amount of vertical en-route 25 inefficiencies in the EUROCONTROL area nor does it identify all underlying reasons for the observed 26 inefficiencies. Instead, it enables an understanding to be gained of the potential level of vertical flight 27 inefficiencies on specific airport pairs, in order to evaluate some specific cases in more detail. 28

The main assumption for the analysis of en-route vertical flight efficiency is that level capping due to 29 ATC constraints i inefficient during the flight planning. Based on the assumption that flights on airport 30 pairs with similar Great Circle Distance (GCD) should be able to reach similar cruising altitudes, the 31 methodology compares the maximum filed flight levels of flights on a specific airport pair and flights 32 on reference airport pairs with a similar GCD and without RAD (Route Availability Document) 33 constraints. 34

Figure 3-29 illustrates the 35 distribution of observed 36 maximum filed flight levels 37 on a given airport pair 38 (blue line) and the 39 reference distribution 40 based on airport pairs with 41 a similar GCD (red line). 42

This representation allows 43 determining the share of 44 flights that are filing lower 45 than the reference flights 46 (impacted flights) and also 47 the altitude difference 48 between them. 49

Although in a number of cases the flights on the given airport pair show a higher maximum flight 50 level than the reference distribution, the focus is on vertical inefficiencies represented by the red 51 shaded area in Figure 3-29. 52



31

Figure 3-30: Results for the top 20 airport pairs in terms of total VFI

Toulouse (TLS)- Paris (ORY)

EHAM-LFPG

LFPG-EHAM

LEBL-LEMD

London (LHR)-Amsterdam (AMS)

EDDF-LFPGLFPG-EDDF

EDDF-EGLL

LFMN-LFPO

EDDT-EDDF

EDDM-EDDK

EGLL-EBBR

EHAM-EGLL

LEMD-LEBL

EGKK-EHAM

LTAC-LTBA

EDDF-EDDT

EGLL-EDDF

EDDM-EDDT

200

300

400

500

600

700

2000 4000 6000 8000 10000 12000

Nu

mb

er o

f fl

igh

ts

Vertical flight inefficiency per flight (feet)

Top 20 airport pairs in terms of total vertical flight inefficiency(30/04/2015 to 27/05/2015)

Bubble size refers to the total vertical inefficiency on the respective city pair

Figure 3-31: Distribution of maximum filed flight levels for LFBO-LFPO

Figure 3-32: Distribution of maximum filed flight levels for EGLL-EHAM

The total vertical flight inefficiency (VFI) is then based on the number of impacted flights and the 1 altitude differences. To account for statistical uncertainties, the lowest and highest 10% of the flights 2 (grey areas in Figure 3-29) are not considered in the analysis. A more detailed explanation of the 3 methodology can be found at www.ansperformance.eu. 4

The methodology was applied 5 to all airport pairs within the 6 ECAC area that have at least 7 1,000 flights per year. 8

The analysis was carried out 9 for the May 2015 AIRAC cycle 10 (30/04/2015 to 27/05/2015). 11

Figure 3-30 shows the results 12 for the top 20 airport pairs by 13 total vertical flight 14 inefficiency. 15

The number of flights is 16 shown on the Y-axis and the 17 vertical inefficiency per flight 18 is shown on the X-axis. The 19 size of the bubble refers to 20 the total vertical flight 21 inefficiency on the respective 22 airport pair during the 23 analysed period. 24

The flights from Toulouse (TLS) to Paris Orly (ORY) showed the largest total vertical flight inefficiency 25 with each flight filing 5,325 26 feet below the reference 27 flights on average. 28

The distributions of the 29 maximum filed flight levels on 30 the two airport pairs Toulouse 31 (TLS)-Paris (ORY) and London 32 Heathrow (LHR) to Amsterdam 33 (AMS) (highlighted in red in 34 Figure 3-30) are shown below 35 in more detail. 36

Flights from Toulouse to Paris 37 Orly cannot file higher than 38 FL345 according to the Route 39 Availability Document (RAD) 40 which explains that the 41 maximum filed altitude is 42 FL340 (Figure 3-31). 43

Further investigation revealed 44 that one airline filed FL280 as 45 their maximum flight level, 46 probably because of an old 47 restriction in their flight 48 planning system. Around 30% 49 of the flights are filing at FL300 50 or FL320 but the reason for 51 this is not immediately clear. 52



32

Flights from London Heathrow (LHR) to Amsterdam (AMS) are level-capped below FL235, according 1 to the RAD. In practice, this results in almost all flights filing at FL230 as can be seen in Figure 3-32. 2 This constraint is probably put to avoid that these flights would enter MUAC airspace. 3

On average, flights are filing 6,550 feet lower than the reference flights. 4

The methodology will be further developed in order to increase the stability of the reference 5 distributions. Future outputs will include time series of the inefficiencies over several AIRAC cycles 6 and a first quantification of the measured inefficiencies in terms of fuel burn and CO2 emissions. 7

3.3.3 Short term ATFCM measures (STAM) 8

Definition: An approach to smooth sector workloads by reducing traffic peaks through short-term 9 application of minor ground delays, appropriate flight level capping and exiguous rerouting to a 10 limited number of flights. 11

Short-term ATFCM measures (STAMs) can reduce the complexity of anticipated traffic peaks. FMPs 12 analyse the associated lists of flights to anticipate ATC workload and identify actions to be taken in 13 order to reduce the traffic complexity generated by those flights. 14

Aircraft operators have expressed concerns that, although they generally support the concept of 15 applying specific localised measures to avoid systematically applying more cumbersome regulations, 16 the adverse effects of STAM measures need to be monitored so that they can be considered in the 17 overall service quality of ANS operations. 18

The PRC agrees with this approach and has investigated how and where the adverse effects of STAM 19 are, or could be, recorded. 20

Minor ground delays: The application of minor ground delays on departing traffic at the behest of 21 local/national ATC can lead to an increase in the taxi-out time of the aircraft concerned as they have 22 to queue until the departure conditions are met: therefore the local performance indicator for Taxi-23 Out-additional time would increase. If the departing aircraft is delayed at the gate instead of during 24 the taxi-out, then the adverse impact would be captured in the ATC pre-departure delay ‘IATA code 25 89’, which is also monitored as a local airport performance indicator. 26

Flight level capping: Currently there is no metric for quantifying the ad-hoc flight level capping 27 arrangements between ATC units as part of STAM. This is somewhat similar to non-availability of 28 requested flight level due to safety (conflicting traffic) or weather (turbulence etc), or indeed tactical 29 change in cruising level per pilot request. Any flight level capping arrangements promulgated through 30 the Network Manager, for example the application of a RAD restriction or the application of a 31 scenario, can identify the impact of the restrictions on filed flight plans (see en-route vertical flight 32 efficiency later in this report). 33

Re-routing: If the re-routing constraints are propagated through the Network Manager resulting in 34 changes to flight plans, then this will become visible and measurable for the horizontal flight 35 efficiency metric that is based on the last filed flight plan. If the re-routing is of a tactical nature, such 36 as STAM re-routing, it will become visible through the horizontal flight efficiency metric based on 37 actual trajectory. Basing both horizontal flight efficiency metrics on achieved distance enables the 38 identification and reporting of performance at local level. 39

In summary, most of the adverse effects of STAM can be monitored through different performance 40 indicators and the PRC will work towards developing new metrics and improving existing ones so that 41 eventually, all ANS constraints can be identified and monitored. 42

3.4 Civil Military cooperation & coordination 43

3.4.1 PRC Review on behalf of Provisional Council. 44

To meet the increasing needs of both sets of stakeholders in terms of volume and time and to 45 maximise the use of finite resource airspace, close civil/military co-operation and co-ordination 46 across all ATM-related activities is crucial. 47

Following the PC recommendation, stemming from PRR 2015 [Ref. 5], to evaluate how the current 48


33

arrangements could be further improved to benefit both civil and military stakeholders; the PRC 1 carried out a review of existing civil/military co-operation and co-ordination procedures [Ref. 6] 2 within EUROCONTROL Member States. 3

The questionnaire focused in particular on the information available to the Level 2 actors in airspace 4 management: to the airspace managers involved in the pre-tactical activities and in the allocation of 5 airspace to satisfy the requirements of both civil and military airspace users. 6

It was structured around 9 specific criteria relevant to individual aspects of civil military coordination 7 and cooperation. All criteria are linked; the information obtained in each allows the different entities 8 (civil, military) to share information and take effective decisions for the benefit of all airspace users. 9

The summary of the identified main issues from the feedback received through the questionnaire is 10 shown in Figure 3-33. They suggest that there is scope for improvement in the overall processes 11 related to the management of airspace. In particular, the main issues relate to: 12

the lack of impact assessments regarding restricted or segregated airspaces and the effect 13 they have on general air traffic, in terms of available ATC capacity and route options; 14

the absence of clear national / regional strategic objectives for both OAT and GAT at ASM 15 level 1; and, 16

the haphazard flow of information throughout the ASM process (availability of the right 17 information to the relevant parties at the right time). 18

There is a need to ensure a functioning feedback loop to ensure that results and issues observed at 19 ASM level 3 are fed back to the previous two levels (strategic, pre-tactical) in order to improve 20 processes where necessary for the benefit of all airspace users. 21

22

Figure 3-33: Identified improvement areas for civil/military cooperation and coordination 23

3.4.2 Additional questions on civil military coordination and cooperation. 24

In preparation for PRR 2016, the PRC invited Member States to provide additional information on 25 cases where military booking requests were adjusted, or cancelled, because of conflicts with GAT 26 traffic demand - by providing answers to the following two questions: 27

Number of times that specific airspace booking requests for military operations and training, 28 were conflicting with GAT traffic demands, and which directly led to the mission being 29 cancelled; 30

Number of times that specific airspace booking requests for military operations and training, 31 were conflicting with GAT traffic demands but where adaptations in either the timing or the 32 location of the mission enabled the mission to be completed as required. 33

Only ten Member States replied to the follow up question but all of them stated that no adjustments 34 were made due to conflicts with GAT traffic demand. The replies support the observations from the 35 questionnaire that there is scope for improvement in terms of impact assessment and in the 36 formulation of strategic objectives for civil/military coordination and cooperation. 37

PRC Survey on civil/military coordination and cooperation

Strategic ATFM - ASM Level 1

Pre-tactical ATFM - ASM Level 2

Tactical - ASM Level 3

16% of the States have not yet identified all relevant airspace impacting on GAT

34% of the States have not yet carried out an impact assessment of specific airspace on GAT

35% of the States have no agreed national strategic objectives for civil/military use

71% do not provide feedback to level 1

35% of the States do not notify the Network Manager of all airspace management decisions impacting route availability and capacity

39% of the States do not notify the Network Manager of all airspace management updates impacting route availability and capacity

52% of the States do not carry out any post-operations monitoring for GAT

32 of the 38 EUROCONTROL Member States eligible (84%) completed the questionnaire(28 coordinated replies)

Detailed questionnaire with 9 specific criteria and 38 questions


34

3.5 Conclusions 1

Traffic in the EUROCONTROL area increased for the third consecutive year in 2016. Of the 39 ANSPs 2 included in the analysis, 25 showed an increase in traffic compared to 14 ANSPs which showed a 3 traffic decline. In absolute terms, ENAIRE (Spain), NATS (UK), and DSNA (France) experienced the 4 highest year on year growth in 2016. DHMI (Turkey), UKSATSE (Ukraine) and ROMATSA (Romania) 5 reported the highest absolute decrease in 2016. 6

It is remarkable that 35 of the 63 Area Control Centres (ACCs) reported their highest traffic levels on 7 record in 2016, surpassing the previously highest levels dating back before the start of the economic 8 crisis in 2008. ACCs with growth rates above 10% in 2016 were Palma, Lisbon, Canarias, and Dublin 9 ACC. 10

En-route ATFM delays in the EUROCONTROL area increased for the third year in a row in 2016 11 (+20.9% vs 2015). The percentage of flights affected by ATFM en-route delays increased from 3.9% to 12 4.8% but the delay per delayed flight decreased slightly from 18.8 minutes to 18.0 minutes in 2016. 13

ATC Capacity/Staffing related constraints remained by far the main driver of en-route ATFM delays 14 (55.3%), followed by weather related constraints (18.3%), ATC disruptions/ industrial actions (12.3%), 15 and Event related constraints (9.1%) which also include delays due to ATC system upgrades. 16

Three quarters of the en-route delays were generated by four air navigation service providers: DSNA 17 (41.6%), DFS (13.0%), Maastricht (11.4%), and ENAIRE (9%). 18

The most constraining ACCs in 2016 were Brest, Nicosia, Bordeaux, Brussels, Barcelona, Prestwick, 19 Maastricht UAC, Warszawa, Canarias, Karlsruhe UAC, and, Marseille. Together, they accounted for 20 70.1% of all en-route ATFM delays but only 30.1% of total flight hours controlled in the 21 EUROCONTROL area. 22

After a continuous improvement over the past years, horizontal flight efficiency slightly decreased in 23 2016 compared to 2015. At Pan-European level, horizontal flight efficiency in filed flight plans 24 decreased from 95.5% in 2015 to 95.4% in 2016. At the same time, the efficiency of actual 25 trajectories decreased stronger from 97.3% to 97.1% in 2016. 26

At Pan-European level, the effects of ATC industrial action on specific days in 2016 are clearly visible 27 but the overall impact on horizontal flight efficiency remains within 0.03% points. 28

The benefits that the implementation of FRA can bring in terms of flight efficiency and related 29 reductions in fuel burn, emissions and costs are substantial. The average of horizontal en-route flight 30 efficiency is 1.6% better for member States in which Free route airspace (FRA) is fully implemented 31 all day. Furthermore, most of the gains are already realised in the flight planning phase - the gap 32 between flight planned and actual flown trajectory efficiency is 1.0% point narrower than for the 33 other States. 34

All else being equal, if the nine States below the EUROCONTROL average could have improved the 35 flight efficiency of the actual trajectories by 0.2 percent points in 2016, the saved distance would 36 have been equivalent to 8.2 million kilometres in 2016 and flight efficiency in the EUROCONTROL 37 area would have improved by 0.1 percent points. 38

In order to address a growing stakeholder interest to better evaluate the vertical component a first 39 evaluation of vertical en-route flight efficiency has been carried out. The analysis did not aim at 40 quantifying exactly the total level of vertical en-route inefficiencies in the EUROCONTROL area but to 41 gain an understanding of the potential level of vertical flight inefficiencies on specific airport pairs in 42 order to evaluate some specific cases in more detail. The results, expressed in terms of feet (total 43 VFI) and feet per flight (VFI per flight), showed clear differences in airport pairs. The methodology will 44 be further developed in order to gain a better understanding of the measured inefficiencies, also in 45 terms of fuel burn and CO2 emissions. 46

Close civil military cooperation and coordination is a crucial enabler to improve capacity and flight 47 efficiency performance. The PRC has identified that areas for further improvement relate to the lack 48 of impact assessment in terms of capacity and route options for restricted/segregated airspace, the 49 absence of clear strategic objectives and the lack of feedback throughout the ASM process. 50

Chapter 4: Operational ANS Performance at Airports

PRR 2016 - Chapter 4: Operational ANS Performance - Airports 35

4 Operational ANS Performance at Airports 1

SYSTEM TREND (TOP 30 AIRPORTS IN TERMS OF TRAFFIC) 2016 Trend change vs. 2015

Average daily movements (arrivals + departures) 22 365 +2.7%

Arrival flow management (per arrival)

Average Airport Arrival ATFM Delay 1.36 -0.1 min

Average Additional ASMA Time (without Turkish airports) 2.15 -0.1 min

Average time flown level during descent (without Turkish airports) 3.1 +0.2 min

Departure flow management (per departure)

Avg. ATC Pre-departure Delay (based on airline delay data) 1.0 +/-0.0 min

Average additional Taxi-out Time (without the Turkish airports) 3.7 +0.2 min

Average time flown level during climb (without Turkish airports) 0.5 +/-0.0 min 2

4.1 Introduction 3

The economic downturn starting in 2008 led to a notable downward adjustment of airport capacity 4 expansion plans over the next 20 years. At the same time, air traffic demand in Europe in 2035 is 5 forecast be some 50% higher than in 2012 which makes the provision of sufficient airport capacity 6 one of the key challenges for future air transport growth [Ref. 7]. 7

There are difficulties in achieving infrastructure growth at the locations where capacity is needed and 8 even in regions where expansions are possible the difficulty will increase as the population grows. 9 Hence, in view of the expected shortfall of airport capacity, the optimised use of available capacity is 10 crucial to keep delays to a minimum. Operational ANS performance plays a key role in balancing 11 traffic with available capacity at airports within the given infrastructural and environmental 12 constraints and in the integration of airports in the European air transport network. 13

This chapter provides a review of operational ANS performance at major European airports. The 14 evaluation of future airport capacity requirements (e.g. new runway, taxiways, etc.) is beyond the 15 scope of this report. 16

As part of the regular operational ANS performance review at European airports, this chapter 17 presents an evaluation of the top 30 airports in terms of IFR movements in 2016, which have the 18 strongest impact on network-wide performance. Together those airports accounted for 45.5% of 19 total European movements in 2016. Any unusual performance observed at an airport not included in 20 the top 30 is commented on in the respective sections of the chapter. 21

Further information on the underlying methodologies and data for monitoring the ANS-related 22 performance at all reviewed airports is available online at www.ansperformance.eu. 23

For the interpretation of the analysis in this chapter, it is important to point out that the observed 24 outcome is the result of complex interactions between stakeholders (airlines, ground handlers, 25 airport operator, ATC, slot coordinator, etc.), which make a clear identification of underlying causes 26 and attribution to specific actors difficult. While at airports, ANS is often not the root cause for an 27 imbalance in capacity/demand (e.g. adverse weather, policy decisions in the airport scheduling 28 phase, traffic demand variation) the way air traffic is managed impacts on airspace users (time, fuel 29 burn, costs), the utilisation of capacity, and the environment (emissions). 30

Hence, the analyses in the respective sections of this chapter should not be interpreted in isolation, 31 but as an integral part of the overall operational performance observed at the airport concerned. 32

At congested airports, one of the primary tools for balancing operational capacity and demand is the 33 airport slot coordination process. But even after unaccommodated demand is removed by allocating 34 airport arrival and departure slots in the strategic phase, there is an important trade-off between the 35 maximised use of scarce capacity and the acceptable level of operational inefficiencies to be 36 considered. 37


PRR 2016 - Chapter 4: Operational ANS Performance - Airports

36

Depending on the commercial value of the airport slots, the number of contingency slots can be close 1 to zero during certain peak periods or in some cases throughout most of the day. 2

The closer airports operate at maximum capacity, the more severe is the impact in terms of 3 operational inefficiencies if reduced capacity is available (due to weather, etc.) or if demand is higher 4 than planned due to variability of traffic demand. 5

The following sections evaluate ANS-related inefficiencies on the departure and arrival traffic flow at 6 the top 30 airports. The performance indicators used for the analysis in this chapter are below. 7

8

Arrival flow management Departure flow management

Expected benefits

• Reduction of airborne terminal holdings

• Support to fuel efficient descent trajectory

• Maximise airport throughput

• Optimum taxi routing (distance & time)

• Minimise ANS-related departure delays

• Optimise push back time sequencing

• Optimum taxi routing (distance & time)

• Adherence to ATFM departure slots

Related indicators

• Airport ATFM arrival delay

• Additional Arrival Sequencing and Metering Area (ASMA) time

• Average level time in descent

• ATC-pre departure delay

• Additional taxi-out time

• ATFM slot adherence

• Average level time in climb

Supporting projects/ initiatives

• Continuous descent operation (CDO)

• Performance based navigation (PBN)

• Arrival manager (AMAN/XMAN)

• Airport Collaborative Decision Making (A-CDM)

• Departure manager (DMAN)

• Continuous climb operations (CCO)

Figure 4-1: ANS-related operational performance at airports (overview) 9

The indicators relate to the optimisation of the inbound and outbound traffic flow and are also part 10 of the SES performance scheme. Complementary to the four standard indicators, an analysis of 11 continuous climbs and descents is provided. A separate study looking at the en-route aspect of 12 vertical flight efficiency can be found in Chapter 3. 13

Through the Global Air Navigation Plan (GANP) [Ref.8], ICAO has established a framework for 14 harmonising airborne and ground-based capabilities. The Aviation System Block Upgrades (ASBUs) 15 comprise packages of capabilities with clearly defined measurable operational improvements, 16 necessary equipage on the ground and in the air, and associated standards and operational 17 procedures. 18

The focus of the current implementation roadmaps are the ASBU Block 0 and 1 Upgrades. With a 19 view to operational ANS performance at airports these upgrades include the following modules. 20

21

Table 4-1: ASBU Performance Improvement Areas and Block upgrades 22

ASBU Performance Improvement Area

Block 0 (2013) Block 1 (2018)

Airport Operations

optimised approach procedures, incl. vertical guidance

increased runway throughput through optimised wake turbulence separation

improve traffic flow through sequencing (AMAN/DMAN)

optimised airport accessibility

further enhanced enablers

increased runway throughput through optimised wake turbulence separation

improved airport operations through departure, surface, and arrival management

Efficient Flight Path improved flexibility and efficiency in descent profiles using CDO and CCO

improved flexibility and efficiency

improved traffic synchronisation

23

24

Capacity management (throughput)

Arrival flow management Departure flow management

Approach (ASMA)

Airport arrival ATFM delay

ATC-related departure

delay

Taxi-out additional

time

Optimisation of departure

sequencing

Balancing ATFM delays at origin airport vs. local

airborne holdings


37

4.2 Traffic evolution at the top 30 European airports 1

Figure 4-2 shows the evolution of average daily IFR movements at the top 30 airports in absolute and 2 relative terms13. On average, movements (arrival + departure) at the top 30 airports in 2016 3 increased by 2.7% compared to 2015. 4

Amsterdam airport, with a global increase in traffic of 6.2%, reported the highest traffic level on 5 record and became the busiest airport in Europe in terms of IFR movements in 2016. The observed 6 growth was mainly due to a substantial increase in low-cost traffic. 7

8 Figure 4-2: Traffic variation at the top 30 European airports (2016/2015) 9

Antalya, previously in the top 30 airports, has suffered a major drop in traffic of almost 30% of 10 movements in total and up to 40% in the summer season showing the tourism downfall mainly due 11 to the migration crisis, escalating security concerns and political problems. This resulted in Warsaw 12 entering the top 30 airports, reinforced by a significant increase in Warsaw traffic of almost 10%. 13

The two Istanbul airports, which reported a remarkable traffic growth over the past years, were 14 affected by the situation in Turkey and by the capacity constraints, showing a slowdown in this yearly 15 traffic increase. 16

Traffic at Brussels (BRU) airport decreased by -6.5% over 2015 as a result of the reduced capacity 17 following the terrorist attacks in March 2016, causing a decrease of 30% and 40% in March and April 18 respectively. 19

As in previous years, the number of passengers at the top 30 airports in 2016 (+4.6% vs. 2015) 20 increased at a higher rate than flights (+2.7%) which is consistent with the observed increase in 21 average aircraft size and passenger load factors. 22

As a result of this ongoing trend, passenger numbers at the top 30 airports were in 2016 27.5% 23 higher than in 2008 which is remarkable considering the fact that movements are merely 1.0% above 24 2008 levels. 25

13 Please note that the airport ranking is based on total IFR movements, which is different from ACI Europe

statistics, based on commercial movements only.

13

40

13

09

12

98

12

65

12

43

10

70

10

33

85

8

84

1

76

5

72

6

71

8

67

0

66

1

64

9

64

1

61

6

59

6

59

3

58

5

53

9

52

5

50

3

50

0

49

9

49

8

48

9

46

1

45

6

42

0

0

200

400

600

800

1 000

1 200

1 400

1 600

Am

ster

dam

(A

MS)

Par

is*

(CD

G)

Lon

do

n*

(LH

R)

Fran

kfu

rt*

(FR

A)

Ista

nb

ul A

tatu

rk (

IST)

Mu

nic

h*

(MU

C)

Mad

rid

* (M

AD

)

Ro

me*

(FC

O)

Bar

celo

na*

(B

CN

)

Lon

do

n*

(LG

W)

Co

pen

hag

en (

CP

H)

Zuri

ch*

(ZR

H)

Osl

o*

(OSL

)

Vie

nn

a (V

IE)

Par

is O

rly

(OR

Y)

Sto

ckh

olm

(A

RN

)

Ista

nb

ul (

SAW

)

Bru

ssel

s* (

BR

U)

Du

esse

ldo

rf*

(DU

S)

Du

blin

(D

UB

)

Pal

ma

(PM

I)

Man

ches

ter

(MA

N)

Ber

lin T

ege

l (TX

L)

Gen

eva*

(G

VA

)

Lisb

on

(LI

S)

Ath

ens

(ATH

)

Lon

do

n (

STN

)

Hel

sin

ki*

(HEL

)

Mila

n*

(MX

P)

War

saw

(W

AW

)Avg

. dai

ly m

ove

me

nts

(2

01

6)

5.9

%

0.5

%

-0.1

%

-1.4

%

0.0

% 3.6

%

2.9

%

-0.7

%

6.3

%

4.3

%

4.0

%

1.5

%

1.2

%

-0.6

%

1.2

% 3.5

% 5.6

%

-6.8

%

3.3

%

8.5

% 10

.5%

10

.8%

0.6

%

0.6

%

9.7

%

7.2

%

6.5

%

-0.6

% 3.6

%

9.4

%

-15%

-10%

-5%

0%

5%

10%

15%

20%

25%

-60

-40

-20

0

20

40

60

80

100

Ch

ange

vs.

pre

vio

us

year

(%

)

Ch

ange

vs.

20

15

(ab

solu

te)

change vs. 2015

Source: Network manager; PRC analysis

Average daily movements (2016)

* A-CDM implemented airport Airports with highest traffic level on record in 2016


38

Figure 4-4: Capacity utilisation at top 30 European airports

0

0,2

0,4

0,6

0,8

0,4 0,5 0,6 0,7 0,8 0,9

Pe

ak L

oad

Ind

ex (

PLI

)

Base Load Index (BLI)

London Heathrow

4.3 Capacity management (airports) 1

4.3.1 Coordination levels 2

In general, the expansion of airport capacity faces various challenges ranging from administrative 3 (e.g. regulatory requirements, planning rules) to environmental sustainability requirements (e.g. 4 noise abatement). While increasing airport capacity in Europe is a must in the long term, there is 5 already the need to make the best use of existing capacity at congested airports. 6

The objective of airport coordination is to ensure the limited airport resources are efficiently used to 7 benefit the greatest number of airport users. 8

Airports are categorised as Level 1 (Non-Coordinated Airport), Level 2 (Schedules Facilitated Airport) 9 and Level 3 (Coordinated Airport). Figure 4-3 shows the distribution of these coordination levels 10 across airports in Europe with more than 20,000 annual movements. 11

Currently almost half of these airports are fully 12 coordinated or coordinated in certain cases (only 13 summer season, only certain hours in the day, 14 etc.) and this number is expected to grow given 15 the lack of capacity to cope with increasing 16 demand. 17

This represents more than 60% of total 18 movements in Europe, and nearly 75% of 19 movements at these airports above 20,000 20 movements. Amongst the top 30 airports, only 21 Athens is not coordinated. The coordination 22 process therefore plays an important role for the 23 capping and distributing of demand in the 24 strategic phase which may have an impact on 25 performance on the day of operations. 26

4.3.2 Throughput and declared capacity 27

While the airport capacity declaration process targets the balance between the demand and the 28 declared capacity in the strategic phase, the actual achieved throughout provides an understanding 29 of the real utilisation of the capacity. 30

Figure 4-4 provides an indication of the 31 capacity utilisation at European airports. The 32 Base Load Index (BLI) refers to the share of 33 time an airport operates above a defined 34 base level (15% of the reference capacity) 35 and the Peak Load index (PLI) provides an 36 indication of the share of time the airport 37 operates above peak level (80% of reference 38 capacity). 39

More information on the methodology is 40 available on www.ansperformance.eu. 41

Considering the achieved levels of 42 throughput across the European top 30 43 airports, a diverse picture emerges. 44

While a number of airports show the classical throughput peaking behaviour with a consistent base 45 level throughout (0.65 < BLI < 0.8), London Heathrow (LHR) in the top right corner shows a clear 46 exceptional capacity utilisation, which needs to be considered when interpreting the results given in 47 this chapter. 48

Figure 4-3: European airports coordination level

(>20.000 movements/year)

Level 3 - Fully coordinated

27%

Level 3 -Special

conditions48%

Level 2 -Schedule faciliated

8%

Level 2 -Special

conditions5%

Level 1 - Not coordinated

12%

75% of traffic is handled by Coordinated/ IATA Level 3 airports (49%)

Source: PRU analysis based on IATA data

http://ansperformance.eu/references/acronym/bli.html

http://ansperformance.eu/references/acronym/pli.html



39

The next section focuses on arrival throughput at airports which is usually more challenging than 1 departure throughput. It compares the declared peak arrival capacity to actual throughput at the top 2 30 European airports. It provides an understanding of the distribution of the arrival throughput 3 including the “peak service rate” which can be achieved in ideal conditions and with a sufficient 4 supply of demand. 5

Figure 4-5 shows the declared peak arrival capacity (brown bars) in 2016 together with the observed 6 arrival throughputs (06h00 – 22h00 local time) shown as box plots which give also an indication of 7 the degree of dispersion of the arrival throughput. 8

9 Figure 4-5: Arrival throughput at the top 30 airports 10

Confirming the observation from Figure 4-4 on page 38, the analysis in Figure 4-5 shows London 11 Heathrow (LHR) with the highest median arrival throughput of all airports and with a small spread 12 close to the peak declared arrival capacity which suggests a high intensity operation all day long. 13 Moreover it is quite remarkable that this performance was achieved with two runways operated in 14 segregated mode. 15

16

Figure 4-6: Evolution of arrival throughput at the top 30 airports (2016) 17

Both Istanbul airports show also a narrow distribution of the hourly throughput with a peak service 18 rate above their declared capacities. As shown in the historic evolution of arrival throughputs 19 (median and peak service rate) in Figure 4-6, these two airports also show a great increase of their 20 throughput in the last 8 years during which Sabiha Gökçen quadrupled its hourly throughput. This 21 and other indicators shown in the report highlight the urgent need for capacity deployment at 22

68 64 45 55 35 58 48 54 38 36 48 40 40 48 34 42 20 33 30 48 33 33 35 25 30 23 31 48 40 260

10

20

30

40

50

60

70

Am

ster

dam

(A

MS)

Par

is*

(CD

G)

Lon

do

n*

(LH

R)

Fran

kfu

rt*

(FR

A)

Ista

nb

ul A

tatu

rk (

IST)

Mu

nic

h*

(MU

C)

Mad

rid

* (M

AD

)

Ro

me*

(FC

O)

Bar

celo

na*

(B

CN

)

Lon

do

n*

(LG

W)

Co

pen

hag

en (

CP

H)

Zuri

ch*

(ZR

H)

Osl

o*

(O

SL)

Vie

nn

a (V

IE)

Par

is O

rly

(OR

Y)

Sto

ckh

olm

(A

RN

)

Ista

nb

ul (

SAW

)

Du

esse

ldo

rf*

(DU

S)

Du

blin

(D

UB

)

Bru

ssel

s* (

BR

U)

Pal

ma

(PM

I)

Man

ches

ter

(MA

N)

Ath

ens

(ATH

)

Gen

eva*

(G

VA

)

Ber

lin T

egel

(TX

L)

Lisb

on

(LI

S)

Lon

do

n (

STN

)

Hel

sin

ki*

(HEL

)

Mila

n*

(MX

P)

War

saw

(W

AW

)

Arr

ival

s p

er

ho

ur

Arrival throughput at the top 30 airports in 2016

Peak declared capacity Source: PRU analysisSource: PRU analysis

* A-CDM implemented airport

Airports with more than 200k movements per year (2016)

0

10

20

30

40

50

60

70

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

Amsterdam(AMS)

Paris* (CDG) London*(LHR)

Frankfurt*(FRA)

IstanbulAtaturk (IST)

Munich*(MUC)

Madrid*(MAD)

Rome*(FCO)

Barcelona*(BCN)

London*(LGW)

Copenhagen(CPH)

Zurich*(ZRH)

Oslo* (OSL) Vienna (VIE) Paris Orly(ORY)

Arr

ival

s p

er h

ou

r

Evolution of arrival throughput at the top 30 airports in 2016

Source: PRU analysisSource: PRU analysis


Peak Service Rate (99th percentile)

Median (50th percentile)

0

10

20

30

40

50

60

70

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

20

08

20

12

20

16

Stockholm(ARN)

Istanbul(SAW)

Duesseldorf*(DUS)

Dublin (DUB) Brussels*(BRU)

Palma (PMI) Manchester(MAN)

Athens (ATH) Geneva(GVA)

Berlin Tegel(TXL)

Lisbon (LIS) London (STN) Helsinki*(HEL)

Milan* (MXP) Warsaw(WAW)

Arr

ival

s p

er h

ou

r


40

Istanbul, together with improved planning and monitoring of operations. 1

Although the analyses in Figure 4-5 and Figure 4-6 provide a first indication of the operations at the 2 airports, it is acknowledged that other factors such as runway layout, mode of operation, and 3 available configurations (many runways may not be operated independently), as well as the societal 4 factors such as noise and environmental policies, would need to be considered in a more detailed 5 analysis. 6

A number of initiatives to further increase airport throughput including, inter alia, time based 7 separation, improved wake vortex separation standards are being implemented at various capacity 8 constrained airports across Europe, and it will be interesting to monitor the benefits of those 9 initiatives in terms of performance in future reports. 10

For instance, the European Wake Vortex Re-categorisation (RECAT-EU) implemented at Paris Charles 11 de Gaulle airport in March 2016 aims at safely increasing airport capacity by redefining wake 12 turbulence categories and their associated separation minima, creating more categories than the 13 traditional ICAO ones. This would allow more accurate and efficient spacing delivery with potential 14 benefits in both runway throughput and safety. 15

4.4 ANS-related flight efficiency constraints at and around airports 16

4.4.1 Arrival flow management 17

This section analyses ANS-related inefficiencies on the arrival flow at the top 30 European airports in 18 terms of arrival ATFM delay and additional ASMA time. 19

Please note that software release 20.0 of the Network Manager on 04 April 2016 introduced a change to improve the accuracy of the ATFM delay calculation for operational purposes which resulted in a reduction of delay compared to the old methodology as of April 2016. More information on the change is available online at www.ansperformance.eu.

Figure 4-7 shows the arrival ATFM delay (top of figure) and the additional ASMA time (bottom of 20 figure) per arrival at the top 30 European airports in 2016. 21

22 Figure 4-7: ANS-related inefficiencies on the arrival flow at the top 30 airports in 2016 23

On average, 6.8% of the flights arriving at the top 30 airports were delayed due to ATFM arrival 24

0

2

4

6

8

10

12

14

Am

ster

dam

(A

MS)

Par

is*

(CD

G)

Lon

do

n*

(LH

R)

Fran

kfu

rt*

(FR

A)

Ista

nb

ul A

tatu

rk (

IST)

Mu

nic

h*

(MU

C)

Mad

rid

* (M

AD

)

Ro

me*

(FC

O)

Bar

celo

na*

(B

CN

)

Lon

do

n*

(LG

W)

Co

pen

hag

en (

CP

H)

Zuri

ch*

(ZR

H)

Osl

o*

(OSL

)

Vie

nn

a (V

IE)

Par

is O

rly

(OR

Y)

Sto

ckh

olm

(A

RN

)

Ista

nb

ul (

SAW

)

Du

esse

ldo

rf*

(DU

S)

Du

blin

(D

UB

)

Bru

ssel

s* (

BR

U)

Pal

ma

(PM

I)

Man

ches

ter

(MA

N)

Ath

ens

(ATH

)

Gen

eva*

(G

VA

)

Ber

lin T

egel

(TX

L)

Lisb

on

(LI

S)

Lon

do

n (

STN

)

Hel

sin

ki*

(HEL

)

Mila

n*

(MX

P)

War

saw

(W

AW

)

2016

2015

0

2

4

6

8

10

12

14Other (allother codes)

Weather(C,D)

ATC other(codes IRVT)

Cap./ staffing(codes CSG)

2015

Airport ATFM arrival delays

1.0

0

0.7

7

0.8

7

1.4

7

1.3

6

2.2

1

2.1

4

2.0

5

2.2

2

2.1

5

201

2

2013

2014

2015

2016

Turkish airports are not included in the top 30 ASMA additional time

ASMA additional time


no

t av

aila

ble

no

t av

aila

ble

Airports with more than 200k movements per year (2016)* A-CDM implemented airport

ANS-related inefficiencies on the arrival flow at the top 30 airports in 2016 (min per arrival)

http://ansperformance.eu/references/definition/atfm_delay.html

http://ansperformance.eu/references/definition/additional_asma_time.html



41

regulations in 2016 (+0.3%pt. vs 2015). At the same time delays were on average shorter than in 1 2015 resulting in a decrease of the average delay per delayed arrival decreased by 2.4 minutes to 2 reach 20.1 minutes in 2016. 3

As could be assumed from the results in Figure 4-7, Istanbul Sabiha Gökçen (42.5%) and Istanbul 4 Atatürk airport (20.3%) had by far the highest share of arrivals delayed due to arrival ATFM 5 regulations in 2016. 6

7 Figure 4-8: Arrival ATFM delayed arrivals at the top 30 airports (2016) 8

While the European average of the additional ASMA time ranges around 2 minutes per arrival 9 throughout the last years, significant variations can be seen at local level. 10

11 Figure 4-9: Five most contributing airports in 2016 (Arrival ATFM delay/ ASMA add. time) 12

At a global scale, the inefficiencies in the arrival flow at the top 30 airports resulted in 5.6 million 13 minutes of arrival ATFM delay (84% of the total arrival ATFM delay in Europe) and 7.8 million minutes 14 of additional ASMA (excluding Istanbul airports for which there is no ASMA data) in 2016. While the 15 arrival ATFM delay minutes affect aircraft on the ground, the extra minutes spent in the ASMA area 16 have an important environmental effect and associated fuel cost for the airspace users. 17

The 5 highest contributors to each of these indicators accounted for approximately half of total delay 18 at the top 30 airports which is to some extent linked to the high traffic volume at those airports. 19

Overall, 30% of the total minutes of arrival ATFM in Europe were generated by regulations at the two 20 Istanbul airports mainly for capacity reasons, while they only accounted for 9% of the traffic. The 21 arrival regulations at Heathrow and Gatwick were mainly due to weather. 22

The main contributor for the additional ASMA time in Europe was London Heathrow which 23 accounted for 25 % of the total minutes of the top 30 airports, while its traffic share was less than 24 3%. This is a consequence of the mode of operations at Heathrow, ensuring a constant demand to 25 maximise runway throughput. 26

12

.0%

3.1

% 7.2

%

3.9

%

20

.3%

2.6

%

2.3

%

1.0

%

9.4

%

8.4

%

0.2

%

16

.4%

4.1

%

4.6

%

10

.4%

1.2

%

42

.5%

3.5

%

0.7

% 5.1

% 7.7

%

0.4

%

1.9

% 5.1

%

1.6

% 4.6

%

3.2

%

1.4

%

0.0

%

2.2

%

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

0

10

20

30

40

50

60

70

80

90

Am

ster

dam

(A

MS)

Par

is*

(CD

G)

Lon

do

n*

(LH

R)

Fran

kfu

rt*

(FR

A)

Ista

nb

ul A

tatu

rk (

IST)

Mu

nic

h*

(M

UC

)

Mad

rid

* (M

AD

)

Ro

me*

(FC

O)

Bar

celo

na*

(B

CN

)

Lon

do

n*

(LG

W)

Co

pen

hag

en (

CP

H)

Zuri

ch*

(ZR

H)

Osl

o*

(OSL

)

Vie

nn

a (V

IE)

Par

is O

rly

(OR

Y)

Sto

ckh

olm

(A

RN

)

Ista

nb

ul (

SAW

)

Du

ess

eld

orf

* (D

US)

Du

blin

(D

UB

)

Bru

ssel

s* (

BR

U)

Pal

ma

(PM

I)

Man

ches

ter

(MA

N)

Ath

ens

(ATH

)

Gen

eva

(GV

A)

Ber

lin T

egel

(TX

L)

Lisb

on

(LI

S)

Lon

do

n (

STN

)

Hel

sin

ki*

(HEL

)

Mila

n*

(MX

P)

War

saw

(W

AW

)

shar

e o

f ai

rpo

rt A

TFM

arr

ival

del

ayed

arr

ival

s

Ave

rage

ATF

M a

rriv

al d

elay

per

del

ayed

arr

ival

avg. ATFM arrival delay per delayed flight % of delayed arrivals

Arrival ATFM delayed arrivals (top 30 European airports in 2016)

6.8% arrival ATFM delayed arrivals at the top 30 airports (+0.3 %pt. vs 2015 )

20.1 min delay per arrival ATFM delayed arrival (-2.4 min vs 2015)

25.0%

7.3%4.9% 4.9% 4.8%

0%

10%

20%

30%

London* (LHR) London (LGW) Frankfurt (FRA) Munich (MUC) Amsterdam(AMS)

Share Additional ASMA Time Top30

7.8 million minutes of add. ASMA Time

5 most contributing airports account for 47% of add. ASMA Time (top 30 airports)

0.3 million tonnes of add. Fuel Burn

0.9 million tonnes of add. CO2 Emissions

22%

13%10%

8%6%

0%

10%

20%

30%

Istanbul (SAW) Istanbul (IST) Amsterdam(AMS)

London (LHR) London (LGW)

Share Arrival ATFM Delay Top30

Cap./ staffing (codes CSG) ATC other (codes IRVT) Weather (C,D) Other (all other codes)

5.6 million minutes of Arrival ATFM Delay

5 most contributing airports account for 59% of airport ATFM arrival delay (top 30 airports )


42

Regional Greek airports 1

Although not included in the top 30, it is noteworthy to highlight the performance observed at a 2 number of small Greek airports. As already pointed out in 2015, those regional airports continue to 3 generate very high delays during summer with a notable impact on the network. Overall, nine smaller 4 Greek airports (Mikonos, Zakinthos, Skiathos, Khania, Kefallinia, Santorini, Iraklion, Makedonia, 5 Diagoras) accounted again for 5.3% of total European airport arrival ATFM delays in 2016, while 6 handling only 1% of the traffic. 7

The observed airport ATFM arrival delays are linked to capacity issues but since the airports are fully 8 coordinated during the summer months the continuous application of ATFM regulations to manage 9 demand should not occur. Even though there is a high level of seasonality at those airports, there is a 10 need to proactively address the issues in order to avoid a repetition of high delays in summer 2017. 11

4.4.2 Departure flow management 12

This section analyses ANS-related inefficiencies on the departure flow at the top 30 European 13 airports in terms of ATFM departure slot adherence, additional taxi-out time, and, ATC pre-departure 14 delays at the gate. 15

4.4.2.1 ATFM departure slot adherence 16

ATFM departure slot adherence ensures that traffic does not exceed regulated capacity and increases 17 overall traffic flow predictability. ATFM regulated flights are required to take off at a calculated time 18 (ATC has a 15 minute slot tolerance window [-5 min, +10 min] to sequence departures). 19

Figure 4-10 shows that although the share of ATFM regulated departures at the top 30 airports 20 (brown bar) increased in 2016 from 11.4% to 14.6% the share of regulated flights departing outside 21 the ATFM slot tolerance window (red line) further decreased from 8.7% to 8.1% which is positive in 22 terms of network predictability. 23

24

Figure 4-10: ATFM slot adherence at airport (2016) 25

Although with a comparatively small share of ATFM regulated departures in 2016, Istanbul Sabiha 26 Gökçen (39.2%) and Istanbul Atatürk airport (27.0%) showed again the highest share of departures 27 outside the ATFM slot tolerance window, followed by Paris Orly (ORY), which suggests scope for 28 improvement. 29

10

.6% 1

4.9

%

5.7

% 9.1

%

27

.0%

4.5

%

4.0

% 9.4

%

5.9

% 9.9

%

2.1

%

8.2

%

1.8

%

4.0

%

16

.6%

4.6

%

39

.2%

4.5

%

5.7

%

4.3

%

3.3

% 9.0

%

8.9

%

7.3

%

14

.6%

8.7

%

7.8

% 11

.7%

1.9

% 5.5

%

Am

ster

dam

(A

MS)

Par

is*

(CD

G)

Lon

do

n*

(LH

R)

Fran

kfu

rt*

(FR

A)

Ista

nb

ul A

tatu

rk (

IST)

Mu

nic

h*

(M

UC

)

Mad

rid

* (M

AD

)

Ro

me

* (F

CO

)

Bar

celo

na*

(B

CN

)

Lon

do

n*

(LG

W)

Co

pe

nh

agen

(C

PH

)

Zuri

ch*

(ZR

H)

Osl

o*

(O

SL)

Vie

nn

a (V

IE)

Par

is O

rly*

(O

RY)

Sto

ckh

olm

(A

RN

)

Ista

nb

ul (

SAW

)

Bru

sse

ls*

(BR

U)

Du

esse

ldo

rf*

(D

US)

Du

blin

(D

UB

)

Pal

ma

(PM

I)

Man

che

ster

(M

AN

)

Ber

lin T

egel

(TX

L)

Gen

eva*

(G

VA

)

Lisb

on

(LI

S)

Ath

ens

(ATH

)

Lon

do

n (

STN

)

Hel

sin

ki*

(HEL

)

Mila

n*

(M

XP

)

War

saw

(W

AW

)

% of ATFM regulated departures

% of ATFM regulated departures outside the ATFM tolerance window

Average (ATFM regulated departures outside the ATFM tolerance window)


Airports with more than 200k movements per year (2016)

Source: NM; PRU analysis

ATFM slot adherence at the top 30 airports (2016)

14.6% of the flights departing at the top 30 airports were ATFM regulated (+ 3.2% pt. vs 2015)

8.1% of the ATFM regulated flights departed outside the slot tolerance window (-0.6% pt. vs 2015)

11

.4%

14

.6%

8.7

%

8.1

%

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

20

10

20

11

20

12

20

13

20

14

20

15

20

16

Top 30 average

http://ansperformance.eu/references/definition/atfm_slot_adherence.html

http://ansperformance.eu/references/definition/additional_taxi-out_time.html

http://ansperformance.eu/references/definition/atc_pre-departure_delay.html

http://ansperformance.eu/references/definition/atc_pre-departure_delay.html


43

Figure 4-12: Five most contributing airports in 2016 (taxi-out add. time)

15% 8.0% 7.3% 7.3% 5.9%

London*^ (LHR) London* (LGW) Paris* (CDG) Rome* (FCO) Barcelona* (BCN)

Share Additional Taxi-Out Time Top 30

12.8 million minutes of add. Taxi-Out Time

5 most contributing airports account for 43% of total add. Taxi-Out Time (top 30 airports)

0.2 million tonnes of add. Fuel Burn

0.6 million tonnes of add. CO2 Emissions

As was the case already in 2015, in contrast to almost all other fully A-CDM implemented airports, 1 Paris Charles de Gaulle (CDG) and Paris Orly (ORY) showed again a surprisingly high share of 2 departures outside the ATFM slot tolerance window. 3

4.4.2.2 ANS-related inefficiencies on the departure flow 4

Figure 4-11 shows the local ATC departure delays (top of figure) and the taxi-out additional time at 5 the top 30 airports. Although the level of inefficiencies cannot be reduced to zero, on average, the 6 less fuel efficient taxi-out additional time is almost four times higher than the local ATC departure 7 delays at the gate which suggests scope for further improvement. 8

9

Figure 4-11: ANS-related inefficiencies on the departure flow at the top 30 airports in 2016 10

A-CDM should help transfer some of these minutes of extra Taxi-Out time to delay at the gate, 11 reducing the emissions and costs. However, this effect is not visible in Figure 4-9, where A-CDM 12 implemented airports show similar Taxi-Out performance to non CDM airports. Geneva maintains an 13 additional Taxi-Out time of 3 minutes per departure with respect to 2015, despite having 14 implemented A-CDM in March 2016. 15

In 2016, the total additional Taxi-16 Out time at the top 30 airports was 17 12.8 million minutes with an 18 associated fuel burn of 0.2 million 19 tonnes. Figure 4-10 shows the 5 20 main contributors in 2016, all of 21 them A-CDM airports, which 22 accounted for 43% of the total 23 additional Taxi-Out time. The traffic 24 share within the top 30 airports was 25 23%. London Heathrow showed a 26 similar behaviour in the departure 27 queue as in the arrival flow, with 28 long additional times. However, 29 being an A-CDM airport, the holding 30 should mostly take place at the gate. 31

0

1

2

3

4

5

6

7

8

9

Am

ste

rdam

(A

MS)

Par

is*

(CD

G)

Lon

do

n*

(LH

R)

Fran

kfu

rt*

(FR

A)

Ista

nb

ul A

tatu

rk (

IST)

Mu

nic

h*

(MU

C)

Mad

rid

* (M

AD

)

Ro

me*

(FC

O)

Bar

celo

na*

(B

CN

)

Lon

do

n*

(LG

W)

Co

pen

hag

en (

CP

H)

Zuri

ch*

(ZR

H)

Osl

o*

(O

SL)

Vie

nn

a (V

IE)

Par

is O

rly

(OR

Y)

Sto

ckh

olm

(A

RN

)

Ista

nb

ul (

SAW

)

Du

esse

ldo

rf*

(DU

S)

Du

blin

(D

UB

)

Bru

ssel

s* (

BR

U)

Pal

ma

(PM

I)

Man

ches

ter

(MA

N)

Ath

ens

(ATH

)

Ge

nev

a (G

VA

)

Ber

lin T

ege

l (TX

L)

Lisb

on

(LI

S)

Lon

do

n (

STN

)

Hel

sin

ki*

(HEL

)

Mila

n*

(MX

P)

War

saw

(W

AW

)

2016

2015

0

1

2

3

4

5

6

7

8

9

2016

2015

Local ATC pre-departure delays (at gate)

0.8 0.8 0.9 1

.0

1.0

3.8

3.7

3.5 3.7 3

.9

2012

2013

2014

201

5

2016

Additional taxi-out additional time

Source: NM; PRC

no

t av

aila

ble

no

t av

aila

ble

no

t av

aila

ble

* A-CDM implemented airport Airports with more than 200k movements per year (2016)

Turkish airports and Osloare not included in top 30

taxi-out time

ANS-related inefficiencies on the departure flow at the top 30 airports in 2016 (min per departure)


44

Local ATC pre-departure delay addresses the effect of capacity/demand imbalances surrounding the 1 departure process. The local ATC departure delay is derived from off-block delays attributed to IATA 2 delay codes reported by airlines, more specifically code 89. 3

The pre-departure delay in Figure 4-11 is calculated according to CODA data as reported by the 4 participating airlines. Nevertheless the pre-departure delay is also reported by the airports under the 5 Performance Scheme through the Airport Data Flow (currently Istanbul airports do not provide the 6 data). This data flow allows for allocation of delay according to the IATA delay codes plus special 7 codes 999 and ZZZ14, and it is required information for all departures. 8

A high share of delays attributed to unknown or unspecified reasons15 might hide a higher share of 9 pre-departure delay attributable to code 89 (local ATC) or other codes. 10

Figure 4-13 shows the breakdown of total minutes of pre-departure delay in 2016 as reported by the 11 airports. It shows a varying picture of the delay allocation at different airports across Europe. While 12 airports like Düsseldorf and Berlin Tegel attribute all their delay to “999” and “ZZZ” codes, in other 13 airports like Zurich, Stockholm, Oslo or Manchester more than 90% of the minutes of delay are 14 allocated to identified reasons. In Heathrow that share is 10%, leaving unexplained 90% of the 15 minutes of delay. 16

This inconsistency in the reporting is not linked to A-CDM implemented airports or those with an 17 APOC16, where delay clearance could potentially be performed. 18

19

Figure 4-13: ATC Pre-departure delay reporting at the top 30 airports 20

Additionally, the chart shows the minutes of delay that should have been reported according to the 21 specifications but nevertheless are not. This could be due either to flights where reported delay is 22 less than the actual or delayed flights for which no delay is reported at all. The latter is only possible 23 in the old reporting mechanism that is still used by some airports. It is the case for Amsterdam, for 24 which there is no information regarding pre-departure delay at the moment. The transition of all 25 reporting airports to the new Airport Data Flow is ongoing. 26

14 Code 999: Delay code clearing is not possible.

Code ZZZ: No delay code information is available/attainable. 15

IATA delay code 99 or specified ambiguity codes in the airport data flow specification, i.e. codes 999 or ZZZ. 16

APOC: Airport Operations Centre

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Man

ches

ter

(MA

N)

Osl

o*

(OSL

)

Sto

ckh

olm

(A

RN

)

Zuri

ch*^

(ZR

H)

Vie

nn

a (V

IE)

Ro

me*

(FC

O)

Lisb

on

(LI

S)

Mila

n*

(MX

P)

Gen

eva*

^ (G

VA

)

Lon

do

n (

STN

)

Hel

sin

ki*

(HEL

)

Co

pe

nh

agen

^ (

CP

H)

War

saw

(W

AW

)

Du

blin

(D

UB

)

Mu

nic

h*

(MU

C)

Fran

kfu

rt*

(FR

A)

Lon

do

n*

(LG

W)

Par

is O

rly

(OR

Y)

Bru

ssel

s*^

(BR

U)

Mad

rid

* (M

AD

)

Bar

celo

na*

(B

CN

)

Ath

ens^

(A

TH)

Pal

ma

(PM

I)

Par

is*

(CD

G)

Lon

do

n*^

(LH

R)

Du

esse

ldo

rf*^

(D

US)

Ber

lin T

ege

l (TX

L)

Am

ster

dam

(A

MS)

Ista

nb

ul A

tatu

rk (

IST)

Ista

nb

ul (

SAW

)

Min

ute

s o

f re

po

rted

del

ay

Pre-departure delay reporting (airport data flow) (top 30 airports)

Min. Dly 89

Min. Dly Other

Min. Dly 99

Min. Dly 999

Min. Dly ZZZ

Unreported min.of delay

5.6 million min unreported pre-departure delay

Category “Other” groups all IATA delay codes different from 89 or 99 * A-CDM implemented airport ^Airport with APOC

No

n-r

epo

rtin

g ai

rpo

rt

No

n-r

epo

rtin

g ai

rpo

rt

40% of pre-departure delay unreported or allocated to unidentified reasons

13.7 million min pre-departure delay without identified reason


4% of ATC pre-departure delay (code 89)


45

Figure 4-14: Average time flown level per flight at the top 30 airports

0

1

2

3

4

5

6

7

Am

ster

dam

(A

MS)

Par

is*

(CD

G)

Lon

do

n*

^ (L

HR

)

Fran

kfu

rt*

(FR

A)

Ista

nb

ul A

tatu

rk (

IST)

Mu

nic

h*

(M

UC

)

Mad

rid

* (M

AD

)

Ro

me*

(FC

O)

Bar

celo

na*

(B

CN

)

Lon

do

n*

(LG

W)

Co

pen

hag

en^

(C

PH

)

Zuri

ch*^

(ZR

H)

Osl

o*

(OSL

)

Vie

nn

a (V

IE)

Par

is O

rly

(OR

Y)

Sto

ckh

olm

(A

RN

)

Ista

nb

ul (

SAW

)

Du

esse

ldo

rf*^

(D

US)

Du

blin

(D

UB

)

Bru

ssel

s*^

(BR

U)

Pal

ma

(PM

I)

Man

ches

ter

(MA

N)

Ath

ens^

(A

TH)

Gen

eva*

^ (G

VA

)

Ber

lin T

egel

(TX

L)

Lisb

on

(LI

S)

Lon

do

n (

STN

)

Hel

sin

ki*

(H

EL)

Mila

n*

(MX

P)

War

saw

(W

AW

)

Ave

rage

tim

e fl

ow

n le

vel (

min

)

Descents

no

t a

vaila

ble

no

t a

vaila

ble

Climbs

0

1

2

3

4

5

6

72016 results

2015 results

Figure 4-15: Median CDO/CCO altitude at the top 30 airports

0

10

20

30

40

Am

ster

dam

(A

MS)

Par

is*

(CD

G)

Lon

do

n*

^ (L

HR

)

Fran

kfu

rt*

(FR

A)

Ista

nb

ul A

tatu

rk (

IST)

Mu

nic

h*

(M

UC

)

Mad

rid

* (M

AD

)

Ro

me*

(FC

O)

Bar

celo

na*

(B

CN

)

Lon

do

n*

(LG

W)

Co

pen

hag

en^

(C

PH

)

Zuri

ch*^

(ZR

H)

Osl

o*

(OSL

)

Vie

nn

a (V

IE)

Par

is O

rly

(OR

Y)

Sto

ckh

olm

(A

RN

)

Ista

nb

ul (

SAW

)

Du

esse

ldo

rf*^

(D

US)

Du

blin

(D

UB

)

Bru

ssel

s*^

(BR

U)

Pal

ma

(PM

I)

Man

ches

ter

(MA

N)

Ath

ens^

(A

TH)

Gen

eva*

^ (G

VA

)

Ber

lin T

egel

(TX

L)

Lisb

on

(LI

S)

Lon

do

n (

STN

)

Hel

sin

ki*

(H

EL)

Mila

n*

(MX

P)

War

saw

(W

AW

)

Med

ian

CD

O/

CC

O a

ltit

ud

e (f

eet)

Results 2016

Results (2015)Descents

no

t a

vaila

ble

no

t a

vaila

ble

Climbs

0

10

20

30

40

Tho

usa

nd

s

Based on the above, the Local ATC Pre-departure delay only accounts for 4% of the total pre-1 departure delay of the Top30 airports in Europe. 2

4.4.3 Vertical flight efficiency during climb and descent 3

This section reports on possible complementary indicators for the measurement of the vertical 4 dimension of flight efficiency during the climb and descent phase. Eliminating intermediate level-offs 5 and diversions for arrivals can save substantial amounts of fuel and reduce CO2 emissions. More 6 information on the methodology and data are available on www.ansperformance.eu. 7

Figure 4-14 shows the average time flown level per flight within a 200NM radius around the airport. 8 Generally, climb outs 9 (top bar chart) are less 10 subject to level offs than 11 descents (bottom bar 12 chart). For descents, a 13 significant amount of 14 level flight can be 15 observed. 16

It is worth noting that at 17 Stockholm (ARN), Athens 18 (ATH) and Helsinki (HEL) 19 the amount of level 20 flight in the descent 21 phase is comparatively 22 low. They all have an 23 average time flown level 24 of less than 1 minute. 25

While Figure 4-14 illustrates the time dimension of vertical flight efficiency, Figure 4-15 provides an 26 understanding about the median altitudes at which continuous descent operations (CDO) started and 27 at which continuous climb operations (CCO) ended. 28

For this metric, the 29 altitude of the lowest 30 level segment during the 31 climb/descent of each 32 flight has been used. 33

The choice for the lowest 34 level segment is due to 35 the fact that this level 36 segment has the highest 37 environmental impact, 38 mainly in terms of fuel 39 consumption. Indeed, if 40 we would have a level 41 segment with a fixed 42 duration, the lower the 43 altitude of the level 44 segment, the higher is 45 the fuel burn. 46

It can be seen that climbs (top figure) are performed more efficiently than descents (bottom bar 47 chart). 48

Most airports have their median CCO altitudes above FL300 which is close to the nominal cruising 49 altitude of jet aircraft. For arriving traffic, the median CDO altitude is notably lower for all considered 50



46

Figure 4-16: Monthly average time flown level per flight to/from EHAM

0.00.51.01.52.02.53.03.5

Jan

Feb

Mar

Ap

r

May Jun

Jul

Au

g

Sep

Oct

No

v

Dec

Ave

rage

tim

e fl

ow

n le

vel (

min

)

Descents

Climbs

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.52016 Results

2015 Results

Figure 4-17: Monthly median CDO/CCO altitude to/from EHAM

05

101520253035

Jan

Feb

Mar

Ap

r

May Jun

Jul

Au

g

Sep

Oct

No

v

Dec

Med

ian

CD

O/

CC

O a

ltit

ud

e (

feet

)

2016 Results

2015 Results

Descents

Climbs

0

5

10

15

20

25

30

35

Tho

usa

nd

s

airports, which is probably due to the application of arrival procedures and the use of holding stacks. 1

It is worth noting that the average time flown level per flight during descent for the Paris airports 2 (CDG, ORY) is in 2016 much higher than in 2015. After consulting DSNA, it became clear that the 3 update rate of the surveillance data changed significantly on 02/09/2015. Since that day, France is 4 providing data with an overall update interval of 1 minute instead of 3 minutes. Further examination 5 of the impact of the update rate on the results revealed that the higher the update rate (smaller 6 update interval), the more correct level flight is being detected since the available trajectory data 7 give a more accurate representation of the true trajectory. Because a lower update rate can “hide” 8 certain level segments, overall more level flight is being detected. 9

Despite the legal requirement to provide surveillance data based on 30 seconds reporting interval, 10 France is only making them available with a 1 minute update interval, being the only EUROCONTROL 11 State not complying. 12

Case study - Amsterdam Schiphol 13

The number of flights to and from Amsterdam (AMS) has increased significantly. Nevertheless, the 14 average time flown level is for the second year in a row comparatively low. 15

Figure 4-16 shows that the 16 monthly values for average time 17 flown level per flight remain 18 stable during 2015 and 2016, 19 although a slightly decreasing 20 trend for the descent value can 21 be observed towards the end of 22 2016. 23

Figure 4-17 presents the monthly 24 median CDO/CCO altitudes. Also 25 these values are quite constant 26 in 2015 and 2016. However, it is 27 remarkable that the median CDO 28 altitude is pretty low. 29

To get a better view on the 30 altitudes of the level flight 31 segments, the vertical 32 trajectories of flights arriving at 33 Amsterdam airport in July 2016 34 are plotted in Figure 4-18. The 35 detected level segments are 36 highlighted in red. Most level 37 flight happens at 2000 and 3000 38 feet which explains the low 39 median CDO altitude values. The 40 level segments at these altitudes 41 are part of the approach 42 procedures during the day in 43 peak hours. According to LVNL, 44 the level segments are used 45 during radar vectoring to increases capacity. Also, because the Schiphol TMA is relatively small, LVNL 46 is unable to implement the level segments at higher altitudes. 47

Figure 4-19 presents the top down view of the arrival trajectories into Amsterdam airport. This figure 48 illustrates that the level segments are happening primarily in the final stages of the approach. The 49 green lines indicate the FIR boundaries. 50


47

Figure 4-18: Vertical trajectories of Amsterdam

(EHAM/AMS) arrivals

Figure 4-19: Horizontal trajectories of Amsterdam

(EHAM/AMS) arrivals

4.5 Conclusions 1

Controlled movements at the top 30 airports in the EUROCONTROL area (in terms of traffic) 2 increased for the third consecutive year in 2016. Average daily movements increased by +2.7% 3 compared to 2015 which corresponds to 594 additional movements each day. 4

At the same time, the number of passengers continued to outpace traffic growth in 2016 (+4.6% vs. 5 2015). As a result of this ongoing trend, passenger numbers at the top 30 airports were in 2016 6 27.5% higher than in 2008 which is remarkable considering the fact that movements are merely 1.0% 7 above 2008 levels (the year with the highest level of traffic on record so far). 8

Ten of top 30 airports (Amsterdam, Istanbul Ataturk, London Gatwick, Stockholm Arlanda, Istanbul 9 Sabiha Gökçen, Dublin, Berlin Tegel, Geneva, Lisbon, and Warsaw) reported their highest traffic level 10 on record surpassing the traffic levels observed before the economic crisis starting in 2008. 11

Antalya experienced a 30% reduction in traffic in 2016 which resulted in Warsaw entering the top 30 12 airports in terms of traffic instead. Amsterdam reported a 5.9% increase in traffic in 2016 which 13 made it the airport with the most commercial movements in Europe in 2016. A number of airports 14 (Manchester (MAN), Palma (PMI), Lisbon (LIS), Warsaw (WAW), and Dublin (DUB) experienced high 15 growth rates above 8% in 2016. 16

The two Istanbul airports, which reported a remarkable traffic growth over the past years, were 17 affected by the situation in Turkey, resulting in a notable slowdown in traffic growth. Of the top 30 18 airports, six showed a traffic decrease in 2016. The highest decrease in traffic among the top 30 19 airports in 2016 was observed for Brussels (BRU) airport with -6.5% compared to 2015 as a result of 20 the reduced capacity following the terrorist attacks in March 2016. 21

The global implementation roadmaps driven by the ICAO Global Air Navigation Plan (GANP) and the 22 supporting Aviation System Block Upgrades (ASBU) modules cast their light on Europe. This 23 strengthens the level of enabler deployments such as Airport Collaborative Decision Making (A-CDM) 24 or procedural changes in form of continuous climb or descent operations, time based separations 25 and the use of improved wake vortex categorisations. 26

Despite the positive effect the aforementioned enablers are expected to have on performance, the 27 substantial traffic increase at some airports contributed to higher levels of operational inefficiency 28 and resulted in somewhat higher additional times during descent and in the taxi-out phase compared 29 to 2015. 30

Average airport arrival ATFM delay decreased slightly in 2016 at the top 30 airports but is still heavily 31 concentrated among a few airports. Five airports (Istanbul Sabiha Gökçen, Istanbul Ataturk, 32 Amsterdam, London Heathrow, and London Gatwick) accounted for 59% of the airport arrival ATFM 33 delay reported for the top 30 airports. At Istanbul Sabiha Gökçen, 42.5% of all arrivals in 2016 were 34 airport ATFM delayed compared to Istanbul Ataturk with 20.3% of the arrivals being delayed. 35

The problem at the two Turkish airports is clearly capacity related and linked to the substantial 36


48

growth observed over the past years. The peak arrival throughput at Sabiha Gökçen airport 1 quadrupled over the past eight years making it the second busiest single runway airport in the 2 EUROCONTROL area after London Gatwick. The situation in Istanbul is expected to improve with the 3 opening of the first phase of the new Istanbul Airport which is scheduled for 2017/2018. 4

Airport arrival ATFM performance at Amsterdam and the two London airports (LHR, LGW) was to a 5 large extent affected by weather which required reducing the available capacity. 6

London Heathrow, Istanbul Ataturk, and Istanbul Sabiha Gökçen show all up with a continuously high 7 arrival throughput close to the peak declared arrival capacity. Constant operations close to maximum 8 capacity generate high delays and possibly cancellations when there is a mismatch between 9 scheduled demand and the capacity that can be made available. 10

At London Heathrow, wind is by far the most dominant factor affecting the airports arrival capacity. 11 The introduction of time based separations in spring 2015 is expected to reduce ATFM delays due to 12 high headwinds at the airport. In 2016, the average weather related ATFM delay per arrival 13 decreased slightly and it will be interesting to see the impact on time based separation on 14 performance once longer time series of data are available. 15

The high intensity operation with a clear focus on the maximisation of runway throughput at London 16 Heathrow comes at a price and the analyses of the additional time in approach and in the taxi out 17 phase show comparatively high levels of inefficiency. In fact, London Heathrow alone accounted for 18 one quarter of the total additional ASMA time in 2016. The average holding per arrival at London 19 Heathrow improved slightly in 2016 but is still above 8 minutes per arrival which is in line with a 20 deliberate decision taken during the airport scheduling process after consultation with airlines. The 21 cross border arrival management (XMAN) project was set up already in March 2014 to reduce 22 airborne holdings on specific traffic flows but it would be worth to investigate further possibilities to 23 reduce holding times at airports through a better support of the network while ensuring a continuous 24 arrival flow into the airport. 25

Due to the lack of data the additional holding time is presently not available for the two Istanbul 26 airports and it would be interesting to get this complementary information in addition to the airport 27 arrival ATFM delay to be able to provide a more balanced picture. 28

The problem with the small Greek airports generating very high delays highlighted in last year’s 29 report still persisted in 2016. Overall, nine regional Greek airports accounted for 5.3% of all airport 30 ATFM delays in the EUROCONTROL area. ATFM regulations should not be applied continuously to 31 regulate demand at airports and the issue, most likely linked to scheduling and variability needs to be 32 addressed proactively in order to avoid a repetition of high delays also in summer 2017. The PRC will 33 be monitoring the situation which now persists for several years. 34

Despite yet a higher number of ATFM regulated flight in 2016, overall ATFM slot adherence at the 35 top 30 airports improved again which is positive in terms of network predictability. 36

Whereas A-CDM implementation is considered to be an enabler to improve situation awareness and 37 performance, it is important to ensure that the available information is used to improve local 38 processes. A-CDM can also help improving the data quality which is presently an issue for the 39 measurement of ATC pre-departure delays. An evaluation of the 2016 data showed that 40% of the 40 delay reported by the top 30 airports was not attributed to a valid delay cause which clearly requires 41 improvement in the future. 42

Building on the initial analysis included in last year’s report, the vertical flight efficiency in climbs and 43 descents at the top 30 airports was also addressed. It is worth pointing out that the measure is 44 complementary to the ASMA additional time and cannot be added to get a combined measure. 45

On average, inefficiencies (expressed in average time flown level per flight) were more than 6 times 46 higher in descent than in climb with notable differences by airport. On average, level flight time 47 during descent increased slightly in 2016 to reach 3.1 minutes per arrival (+0.2 min vs 2015). 48

Chapter 5: ANS Cost-efficiency

PRR 2016 - Chapter 5: ANS Cost-efficiency 49

5 ANS Cost-efficiency (2015) 1

SYSTEM TREND 2015 Trend change vs. 2014

En-route ANS cost-efficiency performance (39 States)

Total en-route ANS costs (M€2009) 6 539 +1.5%

En-route service units (M) 133 +3.9%

En-route ANS costs per service unit (€2009) 49.2 -2.4%

Terminal ANS cost-efficiency performance (30 States)

Total terminal ANS costs (M€2009) 1 084 Time series analysis not available 17

Terminal service units (M) 6.3

Terminal ANS costs per terminal service unit (€2009) 171.6

Air Navigation Service Provider gate-to-gate economic performance (38 ANSPs)

Gate-to-gate ATM/CNS provision costs (M€2015) 8 223 +1.7%

Composite flight-hours (M) 19.0 +1.8%

Gate-to-gate ATM/CNS provision costs per composite flight-hour (€2015) 432 -0.1%

Gate-to-gate economic costs (M€2015) 9 593 +6.0%

Gate-to-gate unit costs of ATFM delays(€2015) 73 +38.7%

Gate-to-gate economic costs per composite flight-hour (€2015) 505 +4.2%

5.1 Introduction 2

This chapter analyses ANS cost-efficiency performance in 2015 (i.e. the latest year for which actual 3 financial data are available) and provides a performance outlook, where possible. 4

It provides a Pan-European view, covering 39 States18 operating 38 en-route charging zones19 that are 5 part of the multilateral agreement for Route Charges. This includes the 30 States which are subject to 6 the requirements of the Single European Sky (SES) Performance Scheme (“SES States”) and also 9 7 EUROCONTROL Member States which are not bound by SES regulations (see section 5.2 below). 8

The cost-efficiency performance of SES States in 2015 has already been scrutinised in accordance 9 with the SES Regulations and the results have been reflected in the Performance Review Body (PRB) 10 2015 monitoring report. The annual Performance Review Report published by the PRC does not seek 11 to duplicate this analysis nor assess performance against SES targets. Instead, it takes into account 12 the SES data and aggregates it with the information provided by the non-SES States to present a Pan-13 European view. The chapter also provides an outlook for the 2016-2019 period. 14

Section 5.2 presents a detailed analysis of en-route cost-efficiency performance at Pan-European 15 system level. Section 5.3 gives an evaluation of terminal ANS costs within the SES area. In order to 16 ensure consistency and comparability with indicators defined in the SES performance scheme and 17 the information provided in RP2 Performance Plans, the cost-efficiency indicators presented in 18 Sections 5.2 and 5.3 are expressed in terms of costs per service unit and in Euro 2009. 19

17 2015 coincides with the beginning of second Reference Period under the SES Performance Scheme which entails

a number of changes for Terminal cost-efficiency. For this reason, it was not possible to analyse changes in terminal cost-efficiency performance between 2014 and 2015 (see page 55 for further details).

18 This is different from the 41 EUROCONTROL Member States in 2015 since: (1) Ukraine is a EUROCONTROL

Member State which is not yet integrated into the multilateral agreement for Route Charges, and (2) Monaco en-route costs are included in the French cost-base.

19 Note that in the Route Charges system, two en-route charging zones include more than one State (Belgium-

Luxembourg and Serbia-Montenegro). Similarly, there are two charging zones for Spain (Spain Continental and Spain Canarias).

http://ansperformance.eu/references/definition/composite_flight_hour.html



PRR 2016 - Chapter 5: ANS Cost-efficiency

50

Finally, Section 5.4 provides a factual benchmarking analysis of ANSPs’ 2015 gate-to-gate economic 1 performance focusing on ATM/CNS costs which are under ANSPs direct responsibility, and including 2 the estimated costs of total ATFM delays (en-route and airport) attributable to the respective service 3 providers. 4

5.2 En-route ANS cost-efficiency performance 5

The analysis of en-route ANS cost-efficiency in this section refers to the en-route charging zones which were part of EUROCONTROL's Route Charges System in 2015 (with the exception of Portugal Santa Maria) but includes Estonia which joined EUROCONTROL on 1st January 2015 and which is part of the SES Performance Scheme.

As shown in Figure 5-1, the “SES States” refer to the 28 Member States of the European Union (EU), plus Switzerland and Norway. These States operate under the “determined costs” method which includes specific risk-sharing arrangements, defined in the Charging Regulation [Ref. 8] aiming at incentivising economic performance.

Figure 5-1: SES and non-SES States

The “non-SES States“ refer to nine States which are not bound by SES regulations but which were 6 part the EUROCONTROL Multilateral Route Charges System in 2015 (i.e. Albania, Armenia, Bosnia-7 Herzegovina, FYROM, Georgia, Moldova, Serbia, Montenegro and Turkey). For these nine States, the 8 “full cost-recovery method” applied in 2015. 9

5.2.1 Changes in reporting of en-route costs and geographical coverage between 2014 and 2015 10

It is noteworthy that the geographical coverage of RP2 now includes Croatia which was not subject to 11 cost-efficiency targets under the SES Performance Scheme during RP1. Apart from a different 12 geographical scope, it should be noted that the cost-efficiency targets for RP2 are based on the 13 Determined Unit Cost (DUC) instead of the Determined Unit Rate (DUR) concept as it was the case 14 during RP1. The main difference between DUR and DUC is that the latter does not include costs 15 associated to exempted VFR flights, while these costs were included in the DUR computation during 16 RP1. Therefore, in order to ensure consistency in time-series analysis, historic en-route ANS costs 17 (2009-2014) have been adjusted to: (a) include the costs associated with Croatia en-route Charging 18 Zone, and (b) exclude the costs associated to VFR exempted flights. These adjustments are presented 19 in the top Table of Figure 5-2 below. 20

21

TR

RS

GE

BA

AM

MK

MD

AL

ME

FR

FI

ES

SE

IT

DE

PL

NO

RO

GB

BG

IE

GR

PT

AT HU

CZ

LT

LV

SK

EE

CH

BE

HR

NL

SI

DK

CY

LU

MT

RP2 SES States

Non SES States

SES and non SES States

Total en-route ANS costs (M€2009)2009

Actuals

2010

Actuals

2011

Actuals

2012

Actuals

2013

Actuals

2014

Actuals

2015

Actuals

2015 vs

2014

2009-15

CAGR

SES States (EU-27+2) en-route costs based

on RP1 defini tion6 248 6 069 5 972 6 048 5 947 5 936 6 008 1.2% -0.7%

+ Croatia en-route costs incl. BiH 65 66 76 72 77 82 80 -2.4% 3.6%

- Costs for VFR exempted flights 11- 26- 34- 9- 9- 9- 11- 19.2% -0.2%


on RP2 defini tion6 302 6 110 6 014 6 110 6 014 6 009 6 077 1.1% -0.6%


51

1

Figure 5-2: Reconciliation between RP1 and RP2 en-route ANS costs for SES States (€2009) 2

It should be noted that Croatia en-route cost-base includes information relating to the provision of 3 ATC services in Bosnia-Herzegovina, while these costs are also included in Bosnia-Herzegovina en-4 route cost-base. 5

Similarly, Hungary en-route costs comprise information relating to the provision of ATC services in 6 Kosovo’s upper airspace (KFOR sector) since April 2014, while these costs are also disclosed in Serbia 7 and Montenegro en-route cost-base. 8

Therefore, in order to present a consistent analysis at Pan-European system level and to avoid any 9 double counting of en-route costs, it was decided to remove these costs from Croatia and Hungary 10 en-route cost-bases. For this reason, the en-route costs presented in this report for the SES States 11 slightly differ from the information published in the PRB monitoring reports. These adjustments are 12 detailed in the bottom Table of Figure 5-2 above. 13

The Tables above shows that the adjustments carried out on the historic data does not affect the 14 trends in en-route costs whether these are computed on the 2009-2015 period (-0.6% p.a.) or 15 between 2014 and 2015 (+1.1%). 16

5.2.2 Trends in actual en-route cost-efficiency performance at Pan-European system level 17

The analysis presented in this section focuses on the 38 Charging Zones that consistently provided 18 en-route costs data over the 2009-2015 period. Georgia which started to provide actual en-route 19 costs data for the year 2014 is not included in this trend analysis. 20

Figure 5-3 shows that in 2015, at Pan-European level, en-route total service units (TSUs) increased 21 faster (+3.9%) than actual en-route ANS costs (+1.5%). As a result, actual en-route unit costs 22 decreased by -2.4% compared to 2014. 23

24 Figure 5-3: Real en-route unit costs per SU for EUROCONTROL Area (€2009) 25

Total en-route ANS costs (M€2009)2009

Actuals

2010

Actuals

2011

Actuals

2012

Actuals

2013

Actuals

2014

Actuals

2015

Actuals

2015 vs

2014

2009-15

CAGR


on RP2 defini tion6 302 6 110 6 014 6 110 6 014 6 009 6 077 1.1% -0.6%

- Adjustment relating to BiH costs included in

Croatia en-route cost-base6.5- 6.9- 8.0- 7.4- 8.0- 8.1- 7.1- -12.8% 1.3%

- Adjustment relating to KFOR staff costs

included in Hungary en-route cost-basen/appl n/appl n/appl n/appl n/appl 1.7- 2.3- 36.9% n/appl

SES States (EU-28+2) en-route costs used in

PRR 20166 296 6 103 6 006 6 103 6 006 6 000 6 068 1.1% -0.6%

60.0 56.5 53.6 54.9 53.1

50.4 49.2

97 97 98 97 97

99

103

108 107

109

116

120

20

30

40

50

60

70

2009Actuals

2010Actuals

2011Actuals

2012Actuals

2013Actuals

2014Actuals

2015Actuals

En-r

ou

te r

eal c

ost

p

er S

U (

€2

00

9)Real en-route unit

costs per SU for the EUROCONTROL Area

(€2009)

En-route SU index (2009)

En-route ANScost index (2009)


2009

Actuals

2010

Actuals

2011

Actuals

2012

Actuals

2013

Actuals

2014

Actuals

2015

Actuals

2015 vs

2014

2009-15

AAGR

Total en-route ANS costs (M€2009) 6 637 6 451 6 419 6 501 6 416 6 445 6 539 1.5% -0.2%

SES States (EU-28+2) 6 296 6 103 6 006 6 103 6 006 6 000 6 068 1.1% -0.6%

Other 8 States in the Route Charges System 342 348 413 398 409 445 471 5.8% 5.5%

Total en-route service units (M SU) 111 114 120 118 121 128 133 3.9% 3.1%

SES States (EU-28+2) 99 102 107 105 107 112 115 3.0% 2.5%

Other 8 States in the Route Charges System 11 12 13 13 14 16 18 10.1% 8.1%

En-route real unit cost per SU (€2009) 60.0 56.5 53.6 54.9 53.1 50.4 49.2 -2.4% -3.3%

SES States (EU-28+2) 63.4 59.9 56.3 58.0 56.2 53.8 52.8 -1.8% -3.0%

Other 8 States in the Route Charges System 30.4 28.3 31.4 30.2 29.3 27.3 26.2 -3.8% -2.4%


52

Over the 2009-2015 period, en-route unit costs reduced by -3.3% p.a. since traffic volumes rose by 1 +3.1% p.a. while en-route costs remained fairly constant (-0.2% p.a.). Figure 5-3 shows that these 2 average changes mask different trends for the SES and non-SES States. 3

Indeed, the en-route unit costs decrease for SES States (-3.0% p.a.) was achieved by slightly reducing 4 costs (-0.6% p.a.) while traffic rose by +2.5% per year on average over the 2009-2015 period. Twelve 5 en-route charging zones operated by SES States could achieve a reduction in en-route costs between 6 2009 and 2015. This is particularly the case for Spain Continental (-5.9% p.a.), Spain Canarias (-4.5% 7 p.a.), Greece (-4.2% p.a.), Portugal (-3.1% p.a.), Switzerland (-2.8% p.a.), Belgium-Luxembourg (-2.7% 8 p.a.) and Denmark (-2.1% p.a.). 9

The en-route unit costs reduction achieved between 2009 and 2015 by non-SES States (-2.4% p.a.) 10 reflects the fact that traffic, measured in service units, increased faster (+8.1% p.a.) than en-route 11 costs (+5.5% p.a.). This was particularly the case for Turkey which benefited from a +9.8% annual 12 traffic increase over the 2009-2015 period. 13

These performance improvements should be seen in the light of (a) the cost-containment measures 14 initiated in 2009-2010 in response of the traffic downturn arising from the economic recession, and 15 (b) for SES States, the implementation of the Performance Scheme and the incentive mechanism 16 embedded in the charging scheme which contributed to maintain a downward pressure on costs 17 during RP1. 18

5.2.3 Breakdown of en-route costs by nature (2015 vs. 2014) 19

As shown in Figure 5-4, en-route costs can be broken down into the following main components:

Staff costs is the largest category and represent some 58% of total en-route costs;

The second largest category, other operating costs accounts for 24% of the total;

Capital-related costs which represent 18% of total en-route costs can be further broken down into depreciation costs (12%) and cost of capital (6%);

Finally, exceptional costs account for less than 1% of total costs.

Figure 5-4: Breakdown on en-route ANS costs by

nature

Figure 5-5 shows that in 2015 the increase in en-route ANS costs (+1.5% or +94 M€2009) is mainly due 20 to higher staff costs (+3.1% or +115 M€2009), cost of capital (+2.4% or +9 M€2009) and exceptional 21 costs (+12 M€2009), while other operating costs (-1.9% or -30 M€2009) and depreciation costs (-1.5% or 22 -12 M€2009) were lower than in 2014. 23

24 Figure 5-5: Breakdown of changes in en-route costs (2014-2015, (€2009)) 25

Staff costs58%

Other operating

costs24%

Deprecia-tion12%

Cost of capital

6%

Exceptionalitems 0.2%

Costs bynature (2015)


3.1

%

-1.9

%

-1.5

%

2.4

%

1.5

%

3.1

%

-3.3

%

-2.1

%

3.4

% 1.1

%

4.1

% 10

.6%

6.8

%

-5.8

%

5.8

%

-20%

-15%

-10%

-05%

00%

05%

-60

-40

-20

-

20

40

60

80

100

120

140

Staff costs Other operatingcosts

Depreciation Cost of capital Exceptionalitems

Total costs

€2

00

9 (

mill

ion

)

Total

SES States

non-SES States



53

Figure 5-5 also indicates that the changes in the various en-route cost categories at Pan-European 1 system level masks diverging trends between SES and non-SES States. This is particularly the case for 2 other operating costs (-3.3% and +10.6%, respectively), depreciation costs (-2.1% and +6.8%, 3 respectively) and the cost of capital (+3.4% and -5.8%, respectively). 4

5.2.4 Actual en-route unit costs at charging zone level 5

Figure 5-6 below shows the level of en-route unit costs for each individual charging zone in 2015. En-6 route unit costs ranged from 71.9 €2009 for Italy to 18.4 €2009 for Malta, a factor of more than three 7 between these two charging zones. It should be noted that Figure 5-6 comprises en-route costs and 8 traffic data relating to Georgia which has been integrated into the Multilateral Agreement for Route 9 Charges on the 1st of January 2014. 10

11 Figure 5-6: 2015 Real en-route ANS costs per TSU by charging zone (€2009) 12

Figure 5-6 also presents the changes in en-route unit costs, TSUs and costs compared to 2014. In 13 2015, unit costs increased for 11 en-route CZs out of the 38 included in the analysis. For five charging 14 zones, en-route unit costs rose by more than +5% in 2015. This includes Moldova (+40.3%), Sweden 15 (+38.1%), Estonia (+8.3%), Armenia (+7.4%) and Finland (+6.7%). 16

While Moldova and Armenia managed to significantly reduce costs between 2014 and 2015 (-21.0% 17 and -5.3%, respectively), this was not sufficient to compensate the steep drop in TSUs (-43.7% 18 and -11.8%, respectively) and to avoid increases in en-route unit costs. For these two CZs, the large 19 decreases in traffic mainly reflect a change in traffic flows following the establishment of 20 restricted/prohibited areas in Ukraine following the MH17 accident in 2014 and the military conflicts 21 in the Eastern region of Ukraine. The changes in traffic flows also affected other CZs in the Eastern 22 European region. This was particularly the case of Bulgaria for which traffic rose by +17% in 2015 23 following a +33% increase in 2014. 24

In 2015, Finland en-route cost-base rose (+2.3%) in a context of traffic decrease (-4.1%) resulting in 25 an increase in en-route unit costs (+6.7%). The increase in Estonia unit costs (+8.3%) mainly reflects 26 the fact that en-route costs rose faster (+11.8%) than TSUs (+3.3%). 27

En-route unit costs substantially rose for Sweden (+38.1%). This is due to an increase in en-route 28 costs (+41.1%) which mainly reflects the reporting of significantly higher pension costs in 2015 29 following the use of a lower discount rate to compute the value of future pension obligations. 30 Pension issues are complex and require the utmost attention given the long term consequences of 31 pensions-related decisions. Clearly, a specific study would be required in order to better understand 32 the magnitude of ANSPs pension costs and their impact on present and future cost-bases, as well as, 33

71

.9

71

.3

71

.2

64

.8

63

.1

62

.9

61

.6

60

.7

60

.4

59

.0

57

.2

56

.3

56

.2

55

.1

52

.8

48

.9

46

.6

43

.8

42

.9

41

.6

41

.0

38

.9

36

.9

34

.9

33

.0

32

.4

30

.6

29

.7

29

.4

28

.8

28

.2

25

.9

24

.9

24

.8

23

.3

21

.4

19

.3

18

.4

49

.2

Ital

y

Swit

zerl

and

Ger

man

y

Spai

n C

anar

ias

Swed

en

Un

ite

d K

ingd

om

Slo

ven

ia

Spai

n C

on

tin

enta

l

Fran

ce

Be

lgiu

m-L

uxe

mb

ou

rg

Au

stri

a

Mo

ldo

va

De

nm

ark

Net

her

lan

ds

Fin

lan

d

Slo

vaki

a

FYR

OM

No

rway

Lith

uan

ia

Alb

ania

Cro

atia

(e

xcl B

iH)

Cze

ch R

epu

blic

Bo

snia

-H

erz.

Serb

ia a

nd

Mo

nte

neg

ro (

excl

BiH

)

Po

lan

d

Po

rtu

gal C

on

tin

enta

l

Cyp

rus

Arm

enia

Hu

nga

ry (

exc

l KFO

R)

Ro

man

ia

Gre

ece

Bu

lgar

ia

Ire

lan

d

Latv

ia

Turk

ey

Esto

nia

Geo

rgia

Mal

ta

Pan

Eu

rop

ean

-1%

0.1

%

-3%

2%

38

%

-4%

3%

-5%

2%

-1%

-4%

40

%-1

%

-6%

7%

-5%

-6%

-4%

3%

-0.3

%

-3%

-5%

-11

% -8% -5

% -1%

1%

7%

-5%

-11

% -4%

-4%

-2%

-4%

-3%

8%

-3%

-1%

-2%

-50%

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%Actual unit cost (% change vs. 2014) Service units (% change vs. 2014) Actual en-route costs (%change vs. 2014)

% c

han

ge v

s. 2

01

4A

ctu

al 2

01

5 u

nit

co

st (

EUR

20

09

)



54

on corresponding unit costs. 1

On the other hand, Figure 5-6 indicates that for nine CZs, en-route unit costs decreased by more than 2 5% in 2015: Bosnia-Herzegovina (-11.2%), Romania (-10.5%), Serbia and Montenegro (-7.6%), FYROM 3 (-6.3%), the Netherlands (-5.7%), Slovakia (-5.5%), Czech Republic (-5.4%), Hungary (-5.4%) and 4 Poland (-5.2%). For most of these CZs, the improvement in en-route cost-efficiency observed in 2015 5 is mainly due to a substantial traffic growth combined with lower or fairly constant en-route costs. 6 The two notable exceptions were: (a) Serbia and Montenegro for which en-route costs rose by +4.2% 7 in a context of substantial traffic increase (+12.7%), and (b) Poland which could significantly reduce 8 its en-route cost-base (-6.4%) while TSUs fell by -1.3%. 9

5.2.5 Pan-European en-route cost-efficiency outlook for 2016-2019 10

The objective of this section is to provide information on planned en-route unit costs at Pan-11 European system level for the period 2016-2019. It is based on data reported by EUROCONTROL 12 Member States in the en-route reporting tables submitted in November 2016 in the context of the 13 Enlarged Committee for Route Charges20. Overall, at Pan-European level between 2015 and 2019, en-14 route unit costs are expected to reduce by -1.6% per year on average. This reflects the fact that over 15 this period traffic volumes are planned to increase faster (+2.2% p.a.) than en-route costs (+0.5% 16 p.a.). 17

18

Figure 5-7: Pan-European en-route cost-efficiency outlook 2016-2019 (in €2009) 19

Figure 5-7 shows that in 2019, en-route unit costs are expected to amount to 46.1 €2009. This 20 is -23.2% lower than in 2009 (60.0 €2009). This remarkable cost-efficiency performance improvement 21 is expected to be achieved by maintaining the cost-base close to 2009 levels in a context of steady 22 traffic increase (+2.7% p.a. over the 2009-2019 period). 23

Detailed analysis indicates that over the 2015-2019 period, en-route unit costs are expected to 24 reduce for 25 en-route CZs out of the 38 included in the analysis. In particular, en-route unit costs are 25 expected to decrease by more than -4% p.a. for six CZs: Sweden (-7.2% p.a.), Moldova (-7.0% p.a.), 26 Serbia and Montenegro (-4.9% p.a.), Germany (-4.6% p.a.), Finland (-4.6% p.a.) and Italy (-4.0% p.a.). 27 For most of these CZs, the planned improvement in en-route cost-efficiency performance observed 28 over the 2015-2019 period is expected to arise from lower costs combined with a modest traffic 29 growth. On the other hand, the expected reduction in Italy en-route unit costs (-4.0% p.a.) mainly 30 reflects the substantial traffic growth planned between 2015 and 2019 (+4.9% p.a.). 31

On the other hand, en-route unit costs are expected to significantly rise for Turkey (+4.8% p.a.) since 32 en-route costs are expected to increase faster (+11.9% p.a.) than traffic volumes (+6.8% p.a.). 33

20 It should be noted that, to date, three SES States (Bulgaria, Poland and Malta) have submitted requests to the

European Commission to revise their adopted RP2 en-route cost-efficiency targets. For these three States, the information used in Figure 5-7 reflects the data provided in the November 2016 submission to the Enlarged Committee for Route Charges including the proposed revisions for the years 2017-2019.

60.0

56.5

53.6 54.9

53.1

50.4 49.2 50.1

49.1 47.7 46.1

97 97 98

97 97 99

101 101 101 101 103

108 107

109

116

120 121 123

127

131

20

30

40

50

60

70

2009Actuals

2010Actuals

2011Actuals

2012Actuals

2013Actuals

2014Actuals

2015Actuals

2016Plan/

Forecast

2017Plan/

Forecast

2018Plan/

Forecast

2019Plan/

Forecast

En-r

ou

te r

eal

co

st p

er

SU (

€2

00

9)

En-route SU index (2009)

En-route ANScost index(2009)



55

5.3 Terminal ANS cost-efficiency performance 1

The analysis of terminal ANS cost-efficiency in this section refers to the SES States (see Figure 5-8) which are required to provide terminal ANS costs and unit rates information in accordance with EU legislation [Ref. 9]. As for en-route, the SES States refers to the 28 Member States of the European Union (EU), plus Switzerland and Norway. These States report on 36 Terminal Charging Zones (TCZs).

2015 coincided with the beginning of second Reference Period under the SES Performance Scheme which entails a number of changes for Terminal cost-efficiency.

Figure 5-8: Geographical scope of terminal ANS cost-

efficiency analysis

Indeed, for the first year, the “determined costs” method is applied for terminal ANS. This method 2 includes specific risk-sharing arrangements which are aiming at incentivising economic performance. 3

In addition, in 2015 several States re-defined the number and composition of TCZs where they are 4 responsible to provide terminal ANS. These changes are summarised in Figure 5-9 below which 5 indicates that in 2015 the number of States reporting terminal ANS data increased to 30, reflecting 6 the inclusion of Croatia terminal ANS data from RP2 onwards. In the meantime, the number of TCZs 7 rose from 33 in 2014 to 36 in 2015, and the number of airports covered decreased from 230 to 173. 8

9

Figure 5-9: Changes in the reporting of terminal ANS data for SES States between 2010 and 2015 10

Examples of changes in the number or in the composition of TCZs include Italy, which went from 11 three TCZs encompassing 47 airports in RP1 (2012-2014) to two TCZs comprising 5 airports. Another 12 example is Belgium, which now reports five TCZs, while only one TCZ with the main airport (Brussels 13 Zaventem) was reported during RP1. 14

5.3.1 Terminal ANS 2015 cost-efficiency performance at terminal charging zone level 15

The terminal cost-efficiency KPI is computed as the ratio of terminal ANS costs with terminal navigation service units (TNSUs).

TNSUs are computed as a function of the maximum take-off weight ((MTOW/50)^α). Since 2015, in accordance with the Charging Scheme Regulation [Ref. 10], all States use a common formula (MTOW/50)^0.7 to compute TNSUs.

This allows for a better comparison of the level of unit terminal costs per TNSU which is achieved by the different charging zones.

Figure 5-10 shows the level of terminal ANS unit costs in 2015 for each of the 34 TCZs included in the 16 analysis. It should be noted that the two TCZs reported by UK have been excluded from this analysis 17

FR

FI

ES

SE

IT

DE

PL

NO

RO

GB

BG

IE

GR

PT

AT HU

CZ

LT

LV

SK

EE

CH

BE

HR

NL

SI

DK

CY

LU

MT

RP2 SES States

SES States

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

SES States reporting 26 28 29 29 29 30 30 30 30 30

Charging zones 28 30 31 31 33 36 36 38 38 38

Airports covered 224 226 229 229 230 173 175 175 175 175

Number of

RP1 RP2

Terminal Navigation Charges (TNC) vs. Airport Charges

Given the risk for potential misunderstanding, it is useful to differentiate between Terminal ANS charges (also called “TNC” for terminal navigation charges) and “Airport charges”, which typically include landing, passenger, cargo, parking and hangar, and noise charges, and are covered by Directive 2009/12/EC.

http://ansperformance.eu/references/acronym/tcz.html

http://ansperformance.eu/references/acronym/tcz.html

http://ansperformance.eu/references/acronym/tnsu.html

http://ansperformance.eu/references/acronym/tnsu.html

Forms/AllItems.aspx?InitialTabId=Ribbon%25http://ansperformance.eu/references/acronym/tnc.html

Forms/AllItems.aspx?InitialTabId=Ribbon%25http://ansperformance.eu/references/acronym/tnc.html


56

since (a) information relating to UK TCZ B (nine airports where terminal ANS are provided on a 1 contractual basis) is not publicly available and (b) UK TCZ C (London Approach) is not directly 2 comparable with other TCZs since the service provided is of a hybrid nature (making the transition 3 between en-route and terminal services for the five London Airports which are also part of TCZ B). 4

In addition, for three charging zones (i.e. Cyprus, Belgium and Spain) the unit costs presented in the 5 figure below do not consider other revenues which are used to subsidise all or part of terminal ANS 6 costs. 7

8

Figure 5-10: Comparison of 2015 terminal ANS unit costs by TCZ 9

Figure 5-10 indicates that in 2015, the average terminal ANS costs per TNSU amounted to 171.6 €2009 10 at system level. Figure 5-10 also shows that the unit terminal costs ranged from 955 €2009 for Belgium 11 Antwerpen TCZ to 96 €2009 for Estonia TCZ, nearly a factor of 10. 12

Caution is needed when interpreting these results since several factors on top of performance-13 related issues can affect the level of terminal unit costs in a specific TCZ. These factors include the 14 number and size of aerodromes included in the charging zone, the use of different cost-allocation 15 between en-route and terminal ANS, differences in traffic levels across TCZs and the scope of ANS 16 provided. 17

For instance, Figure 5-10 shows that the two Belgian TCZs with the highest unit terminal costs in 2015 18 only include one airport each. Similarly, while the French TCZ reflects the information relating to 60 19 airports (including regional airports), only the five main airports are included in the Italian TCZs. 20

Figure 5-11 below provides the distribution of the 34 TCZs included in this analysis based on the 21 terminal ANS costs and also shows the share of the total TNSUs served at system level (see blue 22 dashes). The three largest TCZs (France, Germany and Spain) account for nearly 50% of the European 23 system total terminal ANS costs and traffic, while at the same time, the 15 smallest TCZs represent 24 only around 8% of total terminal ANS costs (7% in terms of TNSUs). 25

Figure 5-11 also indicates that the two TCZs with the highest unit terminal costs in 2015 (Antwerpen 26 (955 €2009) and Belgium Oostende-Brugge (504 €2009)) together represent 0.6% of terminal ANS costs 27 and account for 0.1% of the TNSUs handled at system level. 28

29

0

26

7

25

5

24

4

24

3

23

8

22

0

21

7

21

7

20

9

20

7

19

3

19

0

18

3

17

9

17

8

16

9

16

8

16

3

15

7

15

5

15

0

14

6

14

2

14

2

14

1

13

6

12

6

12

6

11

9

11

6

96

1 1 3

1 1 4

2 1 2 2 1

60

1 4

1 4 3

6 4

2 1

14 16

1 3 4 5

1 1

9

1 1 1 2 0

10

20

30

40

50

60

70

-

100

200

300

400

500

Be

lgiu

m A

ntw

erp

en

Be

lgiu

m O

ost

end

e-B

rugg

e

Slo

ven

ia

Slo

vaki

a

Luxe

mb

ou

rg

Cze

ch R

epu

blic

Swit

zerl

and

Hu

nga

ry

Ro

man

ia

Cro

atia

Be

lgiu

m L

iege

Fran

ce

Bu

lgar

ia

No

rway

Be

lgiu

m B

russ

els

Lith

uan

ia

Latv

ia

Au

stri

a

Ital

y -

Zon

e 2

Cyp

rus

Gre

ece

Po

lan

d

Ger

man

y

Ital

y -

Zon

e 1

Irel

and

Net

her

lan

ds

Spai

n

Den

mar

k

Swed

en

Po

rtu

gal

Fin

lan

d

Mal

ta

Be

lgiu

m C

har

lero

i

Esto

nia

Nu

mb

er o

f ai

rpo

rts

in T

CZ

Rea

l ter

min

al A

NS

cost

per

TN

SU (

€2

00

9)

European average (2015): €171.6

Source: PRU analysisSource: PRU analysis

95

5

50

4


57

1

Figure 5-11: Distribution of terminal ANS costs and TNSUs by TCZ in 2015 2

5.3.2 Terminal ANS cost-efficiency performance: outlook for 2016-2019 3

The objective of this section is to provide information on planned terminal unit costs at system level 4 for the period 2016-2019. It is based on data reported in the terminal reporting tables submitted to 5 the EC in November 201621. 6

Figure 5-12 shows that total terminal ANS costs are expected to slightly decrease (-0.7% p.a.) between 2015 and 2019 while TNSUs are foreseen to increase at an average rate of +1.4% per annum.

As a result, terminal ANS unit costs are expected to decrease from 171.6 €2009 in 2015 to 157.5 €2009 in 2019 (or -2.1% p.a.).

Figure 5-12: Real terminal ANS costs per TNSU, total costs (€2009) and

TNSUs

7

21 It should be noted that, to date, Malta has requested the European Commission to revise the adopted RP2

terminal cost-efficiency targets. For Malta, the information used in Figure 5-12 reflects the data provided in the November 2016 submission of Terminal Reporting Tables including the proposed RP2 cost-efficiency target revisions for the years 2017-2019.

0%

5%

10%

15%

20%

25%

-

50

100

150

200

250

Fran

ce

Ger

man

y

Spai

n

Swit

zerl

and

Net

her

lan

ds

Ital

y -

Zon

e 2

No

rway

Ital

y -

Zon

e 1

Au

stri

a

Be

lgiu

m B

russ

els

Po

lan

d

Po

rtu

gal

Den

mar

k

Irel

and

Swed

en

Cze

ch R

epu

blic

Gre

ece

Hu

nga

ry

Fin

lan

d

Ro

man

ia

Luxe

mb

ou

rg

Cyp

rus

Be

lgiu

m L

iege

Latv

ia

Bu

lgar

ia

Lith

uan

ia

Be

lgiu

m A

ntw

erp

en

Cro

atia

Slo

ven

ia

Be

lgiu

m C

har

lero

i

Mal

ta

Slo

vaki

a

Be

lgiu

m O

ost

end

e-B

rugg

e

Esto

nia

Rea

l ter

min

al A

NS

cost

(M

€2

00

9)


48.4 % of total terminal ANS costs 48.7% of total TNSUs






Shar

e o

f TN

SUs

at E

uro

pea

n s

yste

m le

vel,

%

171.6 174.4165.8 161.5 157.5

101.8

98.397.7 97.2100.2

101.8

103.8

105.9

-

40

80

120

160

200

2015Actuals

2016Determ.

2017Determ.

2018Determ.

2019Determ.

TNSU index (2015)

Terminal ANScost index

(2015)


Term

inal

rea

l co

st p

er T

NSU

(€

20

09

)


58

5.4 ANSPs gate-to-gate economic performance 1

Note that the ACE data included in this section were still provisional at the time when this draft report was made available for consultation. The data will be updated in May 2017 prior to the publication of the final PRR 2016 to ensure consistency with the data that will be presented in the final ACE 2015 Benchmarking Report.

The ATM Cost-Effectiveness (ACE) benchmarking analysis is a Pan-European review and comparison 2 of ATM cost-effectiveness for 38 Air Navigation Service Providers (ANSPs). This includes 30 ANSPs 3 which were at 1st January 2015 part of the SES, and hence subject to relevant SES regulations and 4 obligations. Detailed analysis is given in the ACE 2015 Benchmarking Report [Ref. 11]. 5

The ACE 2015 data analysis presents information on performance indicators relating to the 6 benchmarking of cost-effectiveness and productivity performance for the year 2015, and shows how 7 these indicators changed over time (2010-2015). It examines both individual ANSPs and the Pan-8 European ATM/CNS system as a whole. It is important to note that the year under review (2015) is 9 the latest year for which actual financial data are currently available. 10

The analysis of ANSPs’ economic performance in this section focuses on ATM/CNS provision costs i.e. 11 those which are under the direct responsibility of the ANSP, plus the cost of delay attributable to 12 ANSPs. 13

The analysis developed in the ACE Reports allows identifying best practices in terms of ANSPs 14 economic performance and to infer a potential scope for future performance improvements. This is a 15 useful complement to the analysis of the en-route and terminal KPIs which are provided in the 16 previous sections of this chapter. 17

Figure 5-13 shows a detailed breakdown of gate-to-gate ATM/CNS provision costs. Since there are 18 differences in cost-allocation between en-route and terminal ANS among ANSPs, it is important to 19 keep a “gate-to-gate” perspective when benchmarking ANSPs cost-effectiveness performance. 20

21 Figure 5-13: Breakdown of gate-to-gate ATM/CNS provision costs 2015 (€2015) [TBU] 22

Figure 5-13 indicates that in 2015, at Pan-European system level, gate-to-gate ATM/CNS provision 23 costs amount to some €8.2 Billion. Operating costs (including staff costs, non-staff operating costs 24 and exceptional cost items) account for some 81% of total ATM/CNS provision costs, and capital-25 related costs (cost of capital and depreciation) amount to some 19%. 26

The analysis presented in this section is factual. It is important to note that local performance is 27

Total ATM/CNS provision costs: € 8 223 M

En-route % Terminal % Gate-to-gate %

Staff costs 4 062 63.4% 1 226 67.5% 5 288 64.3%

ATCOs in OPS employment costs 2 043 - 550 - 2 594 -

Other staff employment costs 2 019 - 675 - 2 695 -

Non-staff operating costs 1 062 16.6% 309 17.0% 1 371 16.7%

Depreciation costs 799 12.5% 170 9.4% 969 11.8%

Cost of capital 468 7.3% 97 5.4% 565 6.9%

Exceptional Items 16 0.2% 14 0.8% 29 0.4%

Total 6 408 100.0% 1 816 100.0% 8 223 100.0%

ATM/CNS provision costs (€ M)

49%

51%

Staff costs64.3%

Non-staff operating

costs16.7%

Depreciation costs11.8%

Cost of capital6.9%

Exceptional Items0.4% ATCOs in OPS

employment costs

Other staff employment costs


59

affected by several factors which are different across European States, and some of these are 1 typically outside (exogenous) an ANSP’s direct control while others are endogenous. Indeed, ANSPs 2 provide ANS in contexts that differ significantly from country to country in terms of environmental 3 characteristics (e.g. the size and complexity of the airspace), institutional characteristics (e.g. relevant 4 State laws), and of course in terms of operations and processes. 5

A genuine measurement of cost inefficiencies would require full account to be taken of the 6 exogenous factors which affect ANSPs economic performance. This is not straightforward since these 7 factors are not all fully identified and measurable. Exogenous factors related to operational 8 conditions are, for the time being, those which have received greatest attention and focus. Several 9 of these factors, such as traffic complexity and seasonal variability, are now measured robustly by 10 metrics developed by the PRU. 11

The quality of service provided by ANSPs has an impact on the efficiency of aircraft operations, which 12 carry with them additional costs that need to be taken into consideration for a full economic 13 assessment of ANSP performance. The quality of service associated with ATM/CNS provision by 14 ANSPs is, for the time being, assessed only in terms of ATFM delays, which can be measured 15 consistently across ANSPs, can be attributed to ANSPs, and can be expressed in monetary terms. The 16 indicator of “economic” cost-effectiveness is therefore the ATM/CNS provision costs plus the costs of 17 ATFM delay, all expressed per composite flight-hour. In 2015, the total economic costs (sum of 18 ATM/CNS provision costs with the costs of ATFM delays) amount to some 9 593 M€ which is +6.0% 19 higher than in 2014. Further details on the methodology used to compute economic costs are 20 available in the ACE 2015 Benchmarking Report. 21

5.4.1 Trends in economic cost-effectiveness (2010-2015) 22

Figure 5-14 below displays the trend at Pan-European level of the gate-to-gate economic costs per 23 composite flight-hour (“unit economic costs” thereafter) between 2010 and 2015 for a consistent 24 sample of 37 ANSPs for which data for a time-series analysis was available22. 25

26 Figure 5-14: Changes in economic cost-effectiveness, 2010-2015 (€2015) [TBU] 27

It is noteworthy that the year 2010, which is the starting point of this trend analysis, shows a 28 relatively high level of unit economic costs for the ATM system. This mainly reflects the fact that the 29 unit costs of ATFM delays were exceptionally high that year following a sharp increase in delays for a 30 limited number of ANSPs. 31

Over the 2010-2014 period, economic costs per composite flight-hour decreased by -5.4% p.a. in real 32 terms, mainly due to the substantial decreases in unit ATFM delay costs (-23.4% p.a.). Over this 33 period, ATM/CNS provision costs remained close to their 2010 level (-0.1% p.a.) while the number of 34 composite flight-hours slightly increased (+1.0% p.a.). 35

In 2015, composite flight-hours and ATM/CNS provision costs rose in similar proportions (+1.8% and 36

22 Sakaeronavigatsia which provided data for the first time as part of the ACE 2015 cycle is not included in this

analysis.

-11.3%-5.6% -3.7% -0.7% +4.2%

0

100

200

300

400

500

600

700

2010 2011 2012 2013 2014 2015

€p

er c

om

po

site

flig

ht-

ho

ur

(20

15

pri

ces)

ATM/CNS provision costs per composite flight-hour Unit costs of en-route ATFM delays Unit costs of airport ATFM delays

+1.5%

-0.1%-1.9%

+0.3% +1.7%+3.9%

-1.9% -0.1%

+2.3% +1.8%

-37.6% -39.3%

-18.2%

+11.4%

+38.7%

-40%

-20%

0%

20%

40%

2010-11 2011-12 2012-13 2013-14 2014-15

ATM/CNS provision costs Composite flight-hours Unit costs of ATFM delays



60

+1.7%, respectively), resulting in fairly constant unit ATM/CNS provision costs (-0.1%). However, in 1 the meantime, the unit costs of ATFM delays substantially rose by +38.7% leading to a +4.2% increase 2 in unit economic costs compared to 2014. This is the first increase in unit economic costs since 2010. 3

The right-hand side chart in Figure 5-14 shows that the trend of decreasing unit costs of ATFM delays 4 stopped in 2013, and that a new cycle characterised by higher delays started (+11.4% in 2014 and 5 +38.7% in 2015 in terms of unit ATFM delays costs). This trend continued in 2016 since en-route 6 ATFM delays were +20.9% higher than in 2015. This implies that in 2016, the unit costs of delays will 7 be significantly higher than in 2015 and will negatively affect ANSPs economic cost-effectiveness. 8

Figure 5-15 below shows the comparison of ANSPs gate-to-gate unit economic costs in 2015. The 9 economic cost-effectiveness indicator at Pan-European level in 2015 amounts to €505 per composite 10 flight-hour, and, on average, ATFM delays represent 15% of the total economic costs. Figure 5-15 11 indicates that in 2015 unit economic costs ranged from €865 for Skyguide to €199 for MATS; a factor 12 of more than four. Figure 5-15 also indicates that DFS had the highest unit economic costs amongst 13 the five largest ANSPs. 14

15

Figure 5-15: Economic gate-to-gate cost-effectiveness indicator, 2015 [TBU] 16

It is important to note that, for ANSPs operating outside of the Euro zone (such as Skyguide), 17 substantial changes of the national currency against the Euro may significantly affect the level of 18 2015 unit economic costs when expressed in Euro. More information on exchange rates variations 19 and their impact on unit costs can be found in the ACE 2015 Benchmarking Report. 20

Figure 5-16 below shows the contribution of each of the 37 ANSPs to the change in ATFM delays 21 observed in 2015 at Pan-European system level (i.e. increase from 9,881 to 13,946 thousands of 22 minutes). 23

Figure 5-16 indicates that the increase in ATFM delays observed at system level in 2015 mainly 24 reflects very large increases for a few ANSPs. Indeed, more than 90% of the total increase is 25 generated by only five ANSPs (DSNA, DHMI, HCAA, MUAC and LVNL). The main factors explaining the 26 increase in ATFM delays for the top five contributors are: 27

airport capacity issues at the two Istanbul airports for DHMI; 28

the training and implementation of the ERATO stripless environment in December 2015 at 29 Brest ACC, as well as industrial action in April 2015 for DSNA; 30

ACC staffing and capacity issues during the summer period for HCAA; 31

865 864

771

711

620 607589

557 551 549 547 537 518509 490

461438 427 427 422 421 410 406

382 376 372 369 353349 338 332 326 323 319

307

242215

199

0

100

200

300

400

500

600

700

800

900

Skyg

uid

e

LVN

L

Bel

goco

ntr

ol

DC

AC

Cyp

rus

DFS LP

S

DSN

A

Au

stro

Co

ntr

ol

NA

TS (

Co

nti

nen

tal)

Slo

ven

ia C

on

tro

l

AR

MA

TS

ENA

V

ENA

IRE

Cro

atia

Co

ntr

ol

UkS

ATS

E

Alb

con

tro

l

DH

MI

HC

AA

Saka

ero

nav

igat

sia

Fin

avia

RO

MA

TSA

M-N

AV

AN

S C

R

Oro

Nav

igac

ija

NA

VIA

IR

Hu

nga

roC

on

tro

l

Avi

no

r (C

on

tin

enta

l)

PA

NSA

NA

V P

ort

uga

l (C

on

tin

enta

l)

LFV

Mo

ldA

TSA

BU

LATS

A

MU

AC

SMA

TSA

IAA

LGS

EAN

S

MA

TS

€p

er c

om

po

site

flig

ht-

ho

ur

Financial gate-to-gate cost-effectiveness Unit cost of en-route ATFM delays Unit cost of airport ATFM delays

European system average for economic cost-effectiveness: €505

European system average for financial cost-effectiveness: €432620 589 551 537 518

0

200

400

600

800

DFS

DSN

A

NA

TS(C

on

tin

enta

l)

ENA

V

ENA

IRE


61

capacity issues mainly due to shifting traffic flows for MUAC; and, 1

weather issues at Amsterdam/Schiphol airport, as well as trials and the implementation of a 2 new Voice Communication System for LVNL. 3

The right-hand side of Figure 5-16 shows that, as a result, for most of these ANSPs the share of ATFM 4 delays in economic costs in 2015 is significantly higher than the European average (15%). 5

6 Figure 5-16: ANSPs contribution to ATFM delays increase at Pan-European system level in 2015 [TBU] 7

More details on the changes in ATFM delays in 2015 can be found in Chapter 4 (Operational en-route 8 ANS performance) of PRR 2015 [Ref. 5]. 9

Figure 5-17 below shows how the unit ATM/CNS provision costs (see blue part of the bar in Figure 10 5-17 above) can be broken down into three main key economic drivers: (1) ATCO-hour productivity, 11 (2) employment costs per ATCO-hour and (3) support costs per composite flight-hour. Figure 5-17 12 also shows how these various components contributed to the overall change in cost-effectiveness 13 between 2014 and 2015. 14

Figure 5-17 shows that in 2015, ATCO employment costs per ATCO-hour (+2.9%) rose faster than 15 productivity (+1.5%), and as a result ATCO employment costs per composite flight-hour increased by 16 +1.4%. In the meantime, unit support costs fell by -0.7% since support costs (+1.0%) increased in a 17 lower proportion than the number of composite flight-hours (+1.8%). As a result, unit ATM/CNS 18 provision costs remained fairly constant (-0.1%) in 2015 at Pan-European system level. 19

20 Figure 5-17: Breakdown of changes in cost-effectiveness, 2014-2015 (€2015) [TBU] 21

More details on the changes in unit ATM/CNS provision costs at ANSP and Pan-European system 22 levels are available in the ACE 2015 Benchmarking Report. 23

7.8%7.7%

5.8%1.3%

5.6%0.6%1.4%

0.0%0.0%0.4%0.0%0.4%0.0%0.0%0.0%0.1%

1.6%0.5%

7.5%1.5%1.9%2.0%

1.2%23.0%

16.7%9.8%

14.1%24.2%

3.8%9.5%

65.8%11.1%

31.1%30.2%

41.2%18.0%

37.1%

0% 10% 20% 30% 40% 50% 60% 70%

Share of ATFM delays in economic costs (2015)

65.8%

-436-319

-47-42-26-13

-4-2-1-100

000014912121819284776

120123138

192205

254294304

563669

1 866

-1 000 -500 0 500 1 000 1 500 2 000

PANSADFS

Avinor (Continental)LFVLPS

ANS CREANS

UkSATSESlovenia Control

NAVIAIRAlbcontrol

MATSARMATS

MoldATSAOro Navigacija

LGSM-NAV

BULATSAFinavia

IAASMATSA

HungaroControlROMATSA

NAV Portugal (Continental)Skyguide

Austro ControlBelgocontrol

Croatia ControlENAV

NATS (Continental)DCAC Cyprus

ENAIRELVNL

MUACHCAADSNADHMI

Contribution to changes in ATFM delays (2014-2015, thousands of min.)

1 866

+1.5%

+2.9%

+1.4%

-0.1%

-0.7%

+1.0%

+1.8%

"Traffic effect"

ATCO-hour productivity

Unit ATM/CNS provision costs

2014-2015

"Support costs effect"

Employment costs per

ATCO-hour

ATCO employment costs per composite

flight-hour

Support costs per composite flight-

hour

Weight 69%

Weight 31%


62

5.5 Conclusions 1

PRR 2016 analyses performance in 2016 for all key performance areas, except for cost-efficiency, 2 which analyses performance in 2015 as this is the latest year for which actual financial data are 3 available. On the other hand, PRR 2016 also presents an outlook for 2016-2019 in terms of cost-4 efficiency trends. 5

The en-route cost-efficiency performance Pan-European system (39 States) improved in 2015 since 6 real en-route unit costs decreased from 50.4 €2009 to 49.2 €2009 per service unit (TSU) which 7 corresponds to a -2.4% reduction compared to 2014. This reduction is mainly due to the fact that 8 traffic (+3.9%) rose faster than en-route ANS costs (+1.5%). 9

Over the 2009-2015 period, en-route unit costs reduced by -3.3% p.a. since traffic volumes rose by 10 +3.1% p.a. while en-route costs remained fairly constant (-0.2% p.a.). This performance improvement 11 should be seen in the light of (a) the cost-containment measures initiated in 2009-2010 in response 12 of the traffic downturn arising from the economic recession, and (b) for SES States, the 13 implementation of the Performance Scheme and the incentive mechanism embedded in the charging 14 scheme which contributed to maintain a downward pressure on costs during RP1. 15

The outlook for 2016-2019 suggests that en-route unit costs are expected to decrease from 49.2 €2009 16 in 2015 to 46.1 €2009 in 2019, representing a decrease of -1.6% p.a. on average until 2019. Overall, at 17 Pan-European level between 2009 and 2019, the trend in total en-route costs is planned to remain 18 flat, while traffic is planned to increase by some +31%, implying substantial cost-efficiency 19 improvements over this 10-years cycle. 20

European terminal ANS unit costs amount to 171.6 €2009 in 2015, which is the first year of application 21 of the “determined costs” method for terminal ANS. In 2015, 30 States operated 36 Terminal 22 Charging Zones (TCZs) which included a total of 173 airports. 23

Detailed analysis shows that there are wide differences in the level of unit costs at TCZ level ranging 24 from 955 €2009 for Belgium Antwerp TCZ to 96 €2009 for Estonia TCZ. Caution is needed when 25 interpreting these results since several factors on top of performance-related issues can affect the 26 level of terminal unit costs in a specific TCZ. These factors include the number and size of aerodromes 27 included in the charging zone, the use of different cost-allocation between en-route and terminal 28 ANS, differences in traffic levels across TCZs and the scope of ANS provided. 29

The outlook for 2016-2019 suggests that total terminal ANS costs are planned to slightly decrease 30 (i.e. on average by -0.7% p.a.), while TNSUs are foreseen to increase at an average rate of +1.4% per 31 year. As a result, terminal ANS unit costs are expected to reduce by -2.1% p.a. between 2015 and 32 2019. 33

Detailed benchmarking analysis focusing on ANSPs cost-efficiency at Pan-European system shows 34 that in 2015 the gate-to-gate unit economic costs increased by +4.2%, breaking a trend of 4 years of 35 consecutive decreases. This increase is mainly due to higher ATFM delays unit costs in 2015 (+38.7%) 36 while unit ATM/CNS provision costs remained fairly constant compared to 2014. 37

38

PRR 2016 - Chapter 5: ANS Cost-efficiency 63

1 Article 1 of the PRC’s Terms of Reference, adopted in 2003.

2 EUROCONTROL STATFOR 7-year forecast, February 2016, https://www.eurocontrol.int/articles/forecasts.

3 EUROCONTROL STATFOR 7-year forecast, February 2017, https://www.eurocontrol.int/articles/forecasts.

4 Safety Review Commission, "Annual Safety Report for 2013", (February, 2014).

5 Performance Review Commission, Performance Review Report 2015 (PRR 2015). An assessment of Air Traffic Management in Europe during the calendar year 2015 (June 2016).

6 “Review of Civil Military Coordination and Cooperation Arrangements”, Report commissioned by the PRC, December 2016.

7 EUROCONTROL, “Challenges of growth 2013” report.

8 GANP Resources: http://www.icao.int/airnavigation/Pages/GANP-Resources.aspx

9 Commission Regulation (EC) No 1794/2006 of 6 December 2006 laying down a common charging scheme for air navigation services amended by Commission Regulation (EC)

10 Commission Implementing Regulation (EU) No 391/2013 of 3 May 2013 laying down a common charging scheme for air navigation services.

11 ATM Cost-effectiveness (ACE) 2015 Benchmarking Report. Report commissioned by the Performance Review Commission [to be published in May 2017]

https://www.eurocontrol.int/publications/src-annual-safety-report-2013

performance review report prr 2016

Documents

Transcript of performance review report prr 2016