NETWORK OPERATIONS REPORT Main Report - Draft for Consultation Draft for Consultation

Draft for Consultation Edition Number: 1.0 20/04/2018

DOCUMENT CHARACTERISTICS Document Title Document Subtitle Edition Number Edition Validity Date NETWORK Main Report - Draft OPERATIONS 1.0 20/04/2018 for Consultation REPORT 2017 Note This document is a draft version of the Network Operations Report 2017. It is made available for stakeholder consultation Status STATUS AND ACCESSIBILITY Accessible via Working Draft Intranet Draft Extranet Proposed Issue Internet (www.eurocontrol.int) Released Issue Intended for TLP STATUS Detail Red Highly sensitive, non-disclosable information Amber Sensitive information with limited disclosure Green Normal business information White Public information 2018 The European Organisation for the Safety of Air Navigation (EUROCONTROL). This document is published by EUROCONTROL for information purposes. It may be copied in whole or in part, provided that EUROCONTROL is mentioned as the source and the extent justified by the non-commercial use (not for sale). The information in this document may not be modified without prior written permission from EUROCONTROL. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue ii

Table of Contents DOCUMENT CHARACTERISTICS... II TABLE OF CONTENTS... III LIST OF FIGURES IN MAIN DOCUMENT... IV NOTICE... VI 1 EXECUTIVE SUMMARY... 7 2 INTRODUCTION & SCOPE... 9 3 NETWORK OVERVIEW... 10 3.1 2017 BY MONTH... 10 3.2 TRAFFIC 2017... 12 3.3 DELAYS... 13 3.3.1 ALL AIR TRANSPORT DELAYS (AIRLINE VIEW)... 13 3.3.2 ATFM DELAYS... 15 3.3.2.1 EN-ROUTE ATFM DELAYS... 17 3.3.2.2 AIRPORT/TMA ATFM DELAYS... 18 3.4 CAPACITY... 19 3.5 FLIGHT EFFICIENCY... 20 4 TRAFFIC IN DETAIL... 22 4.1 NETWORK CONTRIBUTORS... 23 4.2 ROUTING ASPECTS... 24 4.3 EXTRA-NM AREA PARTNERS... 25 4.4 AIRPORT TRAFFIC EVOLUTION... 26 4.5 AIRLINE INDUSTRY... 27 4.6 FLIGHT REDUCTIONS... 30 5 EN-ROUTE PERFORMANCE... 31 5.1 HOT SPOTS... 32 5.2 PLANNED EVENTS AND DISRUPTIONS... 34 5.2.1 EN-ROUTE PLANNED EVENTS... 34 5.2.2 EN-ROUTE DISRUPTIONS... 35 5.3 CAPACITY EVOLUTION... 37 5.4 ACC ANALYSIS... 39 5.4.1 DEMAND AND CAPACITY MONITORING SUMMER... 42 5.4.1.1 MARSEILLE... 42 5.4.1.2 NICOSIA... 43 5.4.1.3 KARLSRUHE UAC... 44 5.4.1.4 BREST... 45 5.4.1.5 BORDEAUX... 46 5.4.1.6 MAASTRICHT UAC... 47 6 AIRPORTS... 48 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue iii

6.1 HOT SPOTS... 50 6.2 AIRPORT DISRUPTIONS... 54 6.3 NETWORK OPERATIONS SUPPORT... 56 6.3.1 GREEK ISLANDS SUMMER... 56 6.4 AIRPORT CDM IMPLEMENTATION AND ADVANCED ATC TOWER IMPLEMENTATION... 57 6.5 INFORMATION EXCHANGE BETWEEN AIRPORTS AND NM AIRPORT CORNER PROCESS... 58 7 FLIGHT EFFICIENCY... 60 7.1 AIRSPACE DESIGN... 62 7.2 AIRSPACE CHANGES VS. FLIGHT PLANNING... 63 7.3 ACTUAL TRAJECTORY... 65 7.4 CONDITIONAL ROUTES (CDR)... 66 7.5 FREE ROUTE OPERATIONS... 70 7.6 ROUTE AVAILABILITY DOCUMENT (RAD)... 72 8 NETWORK MANAGER... 74 8.1 CAPACITY (DELAY REDUCTIONS)... 75 8.2 ENVIRONMENT (FLIGHT EFFICIENCY)... 76 9 ATFM COMPLIANCE... 78 9.1 ATFM DEPARTURE SLOTS... 78 9.2 ADHERENCE TO FLIGHT PLAN SUSPENSIONS... 79 9.3 ATFM EXEMPTIONS... 80 9.4 MISSING FLIGHT PLANS... 81 9.5 MULTIPLE FLIGHTS... 82 10 REFERENCES... 83 List of Figures in Main Document Figure 1: Average daily traffic in 2017... 10 Figure 2 Monthly ATFM delay in 2017... 11 Figure 3: Average daily traffic per year... 12 Figure 4: Average departure delay per flight 2013-2017... 13 Figure 5 Breakdown average delay per flight (2017)... 13 Figure 6: Average departure delay per flight 2017... 14 Figure 7: Percentage of delayed flights: ATFM & All Causes... 14 Figure 8 : Average daily ATFM delay (2016 vs. 2017)... 15 Figure 9 Average daily traffic and ATFM delay per flight (En-route and Airport) 2008-2017. 15 Figure 10: Proportion of ATFM delays in 2017... 16 Figure 11: 2017 average daily en-route ATFM delays... 17 Figure 12: 2017 average en-route ATFM delay per flight... 17 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue iv

Figure 13: 2017 average daily airport/tma delays... 18 Figure 14: Traffic, Delay and Effective capacity... 19 Figure 15: Average route extension due to airspace design (RTE DES)... 20 Figure 16: Yearly evolution of flight-planning indicator (KEP)... 20 Figure 17: Yearly evolution of the actual trajectory indicator (KEA)... 21 Figure 18 : IFR Flights per day in NM Area... 22 Figure 19: Main changes to local traffic on the European network... 23 Figure 20: Biggest changes in traffic patterns in 2017... 24 Figure 21: Daily Flight Change and Growth in 2017 (ECAC area)... 25 Figure 22: Traffic Development per Market Segment... 27 Figure 23: Crude Oil and Fuel Prices Evolution... 28 Figure 24: Deflated ticket prices in Europe... 29 Figure 25 Monthly Rate of Operational Cancellations 2016-2017... 30 Figure 26 Network En-route delays vs Scenarios applied... 31 Figure 27: Top 20 en-route ATFM delay locations during 2017... 32 Figure 28: Top 20 en-route ATFM delay per flight locations during 2017... 33 Figure 29 Annual Traffic, delay and capacity evolution... 37 Figure 30: ECAC 'Effective Capacity' Evolution per Month (2006-2017)... 38 Figure 31 - Summer sector scheme (average) - Marseille ACC Saturdays (vs. 2016)... 42 Figure 32 Weekly En-route delay per flight - Marseille ACC Summer... 42 Figure 33 - Summer sector scheme (average) - Nicosia ACC Sundays (vs. NOP plan)... 43 Figure 34 Weekly En-route delay per flight - Nicosia ACC Summer... 43 Figure 35 - Summer sector scheme (average) - Karlsruhe UAC Saturdays (vs. NOP plan).. 44 Figure 36 Weekly En-route delay per flight Karlsruhe UAC Summer... 44 Figure 37 - Summer sector scheme (average) Brest ACC Saturdays (vs. 2016)... 45 Figure 38 Weekly En-route delay per flight Brest ACC Summer... 45 Figure 39 - Summer sector scheme (average) Bordeaux ACC Fridays (vs. 2016)... 46 Figure 40 Weekly En-route delay per flight Bordeaux ACC Summer... 46 Figure 41 - Summer sector scheme (average) Maastricht UAC Saturdays (vs. NOP plan) 47 Figure 42 Weekly En-route delay per flight Maastricht UAC Summer... 47 Figure 43: Top 20 airport delay locations during 2017... 50 Figure 44: Top 20 airport delay per flight locations during 2017... 52 Figure 45: Route efficiency KPI per AIRAC cycle... 61 Figure 46: yearly evolution of airspace design indicator... 62 Figure 47: Potential yearly savings/ losses in nautical miles (NM) due to airspace design... 62 Figure 48: Yearly evolution of flight-planning indicator (KEP)... 63 Figure 49: Yearly savings/ losses in nautical miles (NM) due to improved flight planning efficiency... 63 Figure 50: Yearly evolution of the actual trajectory indicator (KEA)... 65 Figure 51 - Yearly savings/ losses in nautical miles (NM) due to improved actual trajectory efficiency... 65 Figure 52: Evolution of CDR availability... 66 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue v

Figure 53: Rate of CDR availability (RoCA) in 2017... 66 Figure 54: RAI (%) 2017 per AIRAC cycle.... 66 Figure 55: RAU (%) 2017 per AIRAC cycle... 66 Figure 56: Five year RAI evolution... 67 Figure 57: Five year RAU evolution... 67 Figure 58: CDR availability vs. usage in 2017... 68 Figure 59: PFE: 2017 Monthly Distance savings (nautical miles per flight)... 68 Figure 60: PFE: 2017 Monthly time savings (minutes per flight)... 68 Figure 61: PFE 2017 vs. 2016 for planned traffic... 69 Figure 62: PFE 2017 vs. 2016 for actual traffic... 69 Figure 63: PFE: 2017 Fuel economy and CO2 emissions... 69 Figure 64: Map Free Route Airspace Deployment by end 2017... 71 Figure 65 NMOC Delay Savings 2017... 75 Figure 66: ATFM Departure Slot Monitoring for 2016 and 2017... 78 Figure 67: Top 20 ADEPs - Flight Plans Suspensions for 2016 and 2017... 79 Figure 68: ATFM Exemptions for State Aircraft Monitoring for 2016 and 2017... 80 Figure 69: Missing Flight Plans for 2016 and 2017... 81 Figure 70: Multiple Flight Plans for 2016 and 2017... 82 Notice Traffic and Delay Comparisons: All traffic and delay comparisons are between the reporting year (2017) and the previous year, unless otherwise stated. NM Area: All figures presented in this report are for the geographical area that is within Network Manager s responsibility unless otherwise stated. Summer season: Figures referring to the summer season in this report are for the period May to October (incl.), unless otherwise stated. Reporting Assumptions and Descriptions: For further information on the NM Area and the regulation reason groupings, go to the Reporting Assumptions and Descriptions document i available on the EUROCONTROL website. Abbreviations: Abbreviations and acronyms used in this document are available in the EUROCONTROL Air Navigation Inter-site Acronym List (AIRIAL) ii. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue vi

1 EXECUTIVE SUMMARY There were 10.6 million flights in the Network Manager (NM) area in 2017, representing an increase of 4.4% 1 compared to 2016. The months of July, August and September had the highest monthly traffic ever recorded, each totalling more than one million flights. The network had its busiest day 30 June with 35,937 flights. All the main market segments contributed to the strong traffic growth. The high growth in the network was mainly due to the traffic originating in Western Europe (especially the South-west axis flows) together with a recovery of traffic in the Russian Federation. Airline reported delay (all causes) was 12.31 minutes per flight, an increase of 9% in comparison to 2016. Reactionary delay was responsible for 44% of the airline delay, heavily influenced by the impact of first-rotation delays. Despite the high demand, the en-route ATFM delay was just slightly higher than the previous year, at 0.88 minutes per flight (0.86 min/flt in 2016). This represents a 4% increase in capacity. A great number of en-route weather events combined with capacity and staffing issues in some ACCs accounted for much of the delay in 2017. En-route weather delay increased by 30% compared to 2016 and was the double of 2015 weather delay. On the other hand, delays due to industrial action and ATM system changes/implementations decreased significantly compared to 2016. Weekends continued to be the most constrained period in the network. A great part of the en-route delay was generated on weekends, with a peak on Saturday. High traffic was forecasted in several capacity constrained areas for the summer and ANSPs introduced RAD restrictions and scenarios to better manage the demand and available capacity. There were over 11,600 scenarios applied in 2017, 61% of which were flight level restrictions. There were 50% more scenarios than in 2016 and the double of those of 2015. Nevertheless, ATFM regulation usage increased with over 1 million regulated flights, a 30% increase on the same period for 2016. Around 80% of the regulated flights had an ATFM delay of less than 15 minutes (75% in 2016). However, some airlines were able to absorb part of the ATFM delay leading to on-time arrival at stand. A number of ACCs which had been identified in the Network Operations Plan (NOP) as having capacity short-comings for the summer season implemented measures and performed better than expected. These include Athens, Makedonia, Lisbon, Skopje and Zagreb ACCs all with traffic growth above 9%. Among the ACCs with capacity issues, Nicosia and Marseille failed to deliver their NOP commitments, opening fewer sectors than the previous year. Karlsruhe did not comply with the capacity plans agreed in the NOP but the UAC duly fore-warned NM on its staffing issues for the summer. Brest and Bordeaux have performed better than 2016 with increased number of sectors opened (including first-rotation) but still have periods with insufficient capacity in the face of high demand (e.g. Saturdays). Maastricht, despite the flexible sector 1 The growth is calculated based on average daily number of flights to remove the effect of the leap year in 2016. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 7

schemes, was operating at full capacity in 2017 and at times was not able to cope with the high geographically imbalanced demand. Airport ATFM delay decreased by 3.6% in 2017. Airport capacity and weather contributed to 83.9% of the total airport delays. Amsterdam/Schiphol was the main generator of airport ATFM delays, with adverse weather events heavily impacting the operations at this airport. Despite the decrease in traffic, Istanbul/Sabiha Gökcen and Istanbul/Ataturk remained the main contributors to airport capacity delays. Industrial action contributed to 8.5% of total en-route delay in 2017. There were roughly 800,000 minutes of delay attributed to strikes, representing a 33% decrease when compared to 2016. An estimated 4,600 flights were cancelled during the strike days. The French industrial actions in March, September, October and November contributed to most of the industrial action delays in 2017. The new ATM system implementation at Bordeaux ACC (ERATO), which started in 2016, was concluded in May. The good network cooperation with the ACC ensured a smooth transition, with less than 100,000 minutes of ATFM delay generated. The new VCS implementation at Maastricht UAC generated 110,000 minutes of ATFM delay throughout the year. Overall, there were fewer delays due to system upgrades and other planned events. These accounted for 0.03 minutes per flight of ATFM delay, the result of a wellcoordinated Transition Plan for Major Projects. NM delivered absolute en-route ATFM delay savings of 14% through individual flight improvements, above its commitment in the Network Performance Plan (NPP). Summer months were very busy for the NMOC with daily capacity issues to mitigate and a high number of weather events. This was reflected in the high number of NM helpdesk requests received (33,825 in July). Flight efficiency performance improved during 2017. The actual trajectory indicator (KEA) improved and is now under the 2.78% target (NM and SES). The route extension indicator based on the last filed flight plan (KEP) was still off-target (0.18pp for NM). Nevertheless, this represents the best KEP since 2014. Conclusions Although en-route capacity increased by 4%, for summer 2018 more is needed and in particular there needs to be network solutions to address the current and future issues. Network partners agree that: ANSPs need to resolve continuing local capacity issues and be capable of best-in-class performance. This will involve: adopting modern working practice and rostering processes; improving airspace and systems implementation practice; providing more capacity at the right time, i.e. weekend. In addition, the network needs to do things differently Short/medium term: find novel ways to cope with high traffic growth through regional ACC cooperation; Medium/long term: change the European way of delivering capacity through new network concepts, including cross border arrangements. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 8

2 INTRODUCTION & SCOPE The purpose of this document is to provide an overview of the European ATM network performance in 2017 in the areas of traffic evolution, capacity offered by the Air Navigation Service Providers and Airports, delays and flight efficiency. Airspace users opinion on the network performance is also included. The report analyses the annual results in light of the main events that took place in the course of the year. The document structure is as follows: Section 1: Executive Summary. Section 2: Introduction & Scope. Section 3: Network Overview contains the annual performance of the European ATM network: traffic, capacity, delays and flight efficiency. Section 4: Traffic in Detail is a detailed analysis of traffic growth in 2017 in the NM area and adjacent regions. Section 5: En-Route Performance Analysis is an analysis of network en-route performance: events and disruptions; capacity and ACC performance. Section 6: Airports is an analysis of the performance of airport operations. Section 7: Flight Efficiency is an analysis of network flight efficiency. Section 8: Network Manager is NM s contribution to achieved performance results. Section 9: ATFM Compliance provides a view on the compliance to the ATFM Implementing Rule. Section 10: References. Annex I: Airspace Users View outlines their perspective on how the network performed in 2017. Annex II: ACC contains a traffic and capacity evolution for each ACC in 2017. Annex III: Airports contains capacity, delay, arrival/departure punctuality status and a NM performance assessment of each of the significant airports in 2017. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 9

Avg daily traffic EUROCONTROL 3 NETWORK OVERVIEW 3.1 2017 BY MONTH The year of 2017 was marked by very high traffic in the network. The growth observed at the end of 2016 (4.6% in December 2016) was maintained throughout 2017 with especially high figures (above 5%) during the summer months. Despite the high traffic levels, the en-route delay per flight had only a slight increase of 2.3% comparing to 2016. During the first half of the year, the en-route ATFM delay was roughly ¾ of that of the same period in 2016, which had been affected by a great number of industrial actions and a major ATM system implementation (ERATO in Brest). The second half of 2017 was marked by capacity, staffing and weather delays, causing 24% more ATFM delay than the same period in 2016. With the exception of the period between July and October, the network performance in terms of ATFM delay was better than the previous year. 35000 4.8% 3.0% 4.5% 3.3% 4.6% 5.1% 5.2% 5.0% 4.5% 5.0% 4.2% 1.9% 30000 25000 20000 15000 10000 5000 0 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC 2016 2017 Figure 1: Average daily traffic in 2017 From January to April, capacity and staffing delays were still relatively low. A number of events created some peaks in ATFM delay, namely: industrial actions in France, ATM system issues in Lisbon, operational configuration changes and new VCS system implementation in Maastricht and the implementation of ERATO in Bordeaux. In April, Nicosia reached a monthly en-route delay of 1.95 min/flt, showing the first signs of capacity issues that would affect this ACC throughout the year. At the end of May, a quarter of the year-to-date en-route delays was due to ATC disruptions, especially industrial actions. The start of the IATA summer schedules brought additional traffic growth and weather events that were to affect the network during the following months. En-route weather and en-route capacity were the main causes of en-route delay in the network in June (28% each). A record network traffic of 35,937 flights was achieved on 30 June. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 10

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC ATM delay per flight (min/flt) EUROCONTROL Traffic in July averaged 33,721 flights per day, the highest figure of the year. This represented an increase of 5.2%, which was higher than the high-forecast scenario for the network. En-route ATFM delay reached its yearly peak at 1.98 minutes per flight, increasing 19% comparing to 2016. Weather was the main driver for this increase, affecting Karlsruhe, Marseille, Brest and Vienna ACCs. High delays attributed to capacity and staffing issues were recorded in Karlsruhe, Marseille, Brest and Nicosia ACCs. This high traffic and high delay situation was to remain in August - though slightly better. 3.0 2.5 2.0 1.5 1.0 0.5 0.0 1.2 1.0 1.0 0.8 1.7 1.1 0.9 0.8 1.5 1.2 2.7 2.0 2.6 2.8 1.6 2.2 1.3 2.1 1.1 1.4 0.7 0.6 1.4 1.1 2016 2017 Figure 2 Monthly ATFM delay in 2017 In September, four ACCs were still generating en-route delays higher than 1 minute per flight (Marseille, Brest, Nicosia and Karlsruhe). As the end of summer approached, daily staffing issues were recurrent in Karlsruhe and Nicosia ACCs. Capacity, staffing and weather issues resulted in an increase of 85% of the en-route delay in the network comparing to 2016. French ATC industrial actions occurred in September and continued in October affecting mainly Marseille, Brest and Bordeaux. November and December were again months with ATFM delay decrease comparing to the same months in 2016. Some capacity and staffing issues occurred in Canarias and Karlsruhe. Traffic growth recorded its lowest figure in December (1.9%) ending a one year cycle with very high traffic growth. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 11

Average Daily Traffic (Flights) EUROCONTROL 3.2 TRAFFIC 2017 30,000 29,000 28,000 27,000 26,000 25,000 24,000 23,000 22,000 26,427 26,215 26,685 27,094 27,844 29,057 2012 2013 2014 2015 2016 2017 Year Figure 3: Average daily traffic per year There were over 29,000 daily flights in the NM area in 2017 (Figure 3), an increase of 4.4% 2 compared to the previous year, reaching a total 10.6 million flights. The summer months (May to October incl.) were the highest ever recorded, with July, August and September totalling more than one million flights each. All the main market segments contributed to the strong flight growth. The high growth in the network was mainly due to the traffic originating in Western Europe (especially the South-west axis flows). Spain, UK and Germany were the biggest contributors to the network growth. In addition, there was an intense growth of flights in the Russian Federation due to a fast economic recovery which started at the beginning of 2017. The recovery of flights to South-Eastern states (Turkey, Egypt, Tunisia) was partly influenced by this growth. There was a slight slow-down of traffic growth at the end of the year, due to Ryanair capacity cuts and the ceasing of operations of Monarch Airlines, Air Berlin and Niki. In addition, adverse weather conditions led to a high number of flight cancellations in early December. 2 The growth is calculated based on average daily number of flights to remove the effect of the leap year in 2016. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 12

Average departure delay per flight in minutes Average departure delay (min) per flight 2013 2014 2015 2016 2017 EUROCONTROL 3.3 DELAYS 3.3.1 ALL AIR TRANSPORT DELAYS (AIRLINE VIEW) This section presents the all air transport delay situation as seen from the airlines by using the data collected by the Central Office for Delay Analysis (CODA) from airlines. Data coverage is 68% of the commercial flights in the ECAC region for 2017. 15 10 Average Departure Delay per Flight 2013-2017 12.31 11.30 9.34 9.70 10.50 5 0 Average reactionary delay per flight (CODA) Average delay per flight Other Primary delay All Causes (CODA) Average ATFM en-route delay per flight (CODA) Average ATFM en-route delay per flight (NM reported) Figure 4: Average departure delay per flight 2013-2017 15 10 5 0 Breakdown Average Delay per Flight 2017 0.88 0.92 0 0 Average En- Route Delay/Flight Average En- Route Delay/Flight Reactionary delay Primary Delay (excl En-Route) 6.87 Average primary delay 12.31 Passenger experience Based on airline data, the average departure delay per flight from All-Causes was 12.31 minutes per flight, an increase of 9% in comparison to 2016 where the average delay was 11.30 minutes per flight. Primary delays counted for 56% (or 6.87 min/flt), with reactionary delays representing the smaller remaining share of 44% at (5.44 min/flt). Figure 5 Breakdown average delay per flight (2017) Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 13

Percentage of flights delayed on departure 2013 2014 2015 2016 2017 ATFM NM Percentage of flights delayed on departure: All Causes 201701 201702 201703 201704 201705 201706 201707 201708 201709 201710 201711 201712 Avg departure delay/flight (min) EUROCONTROL Average Departure Delay per Flight 2017 20 15 10 5 0 Further analysis of the past 12 shows that the average All-Caus route ATFM delay reported by airlin 0.92 minutes per flight. This is highe compared to the NM reported avera route ATFM delay of 0.88 minutes p in 2017. Average reactionary delay per flight (CODA) Average delay per flight Other Primary delay All Causes (CODA) Average ATFM en-route delay per flight (CODA) Average ATFM en-route delay per flight (NM reported) Figure 6: Average departure delay per flight 2017 Percentage of delayed flights: ATFM & All Causes 10% 25.0% 8% 20.0% 6% 15.0% 4% 10.0% 2% 5.0% 0% 0.0% The percentage of flights delayed from All-Causes increased with those exceeding 15 minutes increasing by 1.6 percentage points to 21.8%. Those exceeding 30 minutes also increased with 10.8% of flights being delayed in 2017. > 15min by ATFM Restrictions (NM) > 30min by ATFM Restrictions (NM) > 15min:All Causes (CODA) > 30min:All Causes (CODA) Figure 7: Percentage of delayed flights: ATFM & All Causes 3 ATFM delays reported by airlines could be different than the NM calculated ATFM delays due to difference in methods: ATFM delays of NM are the (flight) planned delays ; the airlines report the actual experienced ATFM delay on departure. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 14

Average delay per flight (mins) Average Traffic ATFM Delays (min) EUROCONTROL 3.3.2 ATFM DELAYS 50,000 45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000-42566 43533 23871 25519 (56.1%) 18695 18014 (43.9%) (41.4%) 2016 2017 EN-ROUTE AIRPORT (58.6%) Average daily ATFM delay increased by 2.3% in 2017 compared to 2016. The average daily en-route ATFM delay increased by 6.9% but the average daily airport ATFM delay decreased by 3.6%. Figure 8 : Average daily ATFM delay (2016 vs. 2017) 3.0 2.5 2.0 1.5 1.0 0.5 0.0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 EN-ROUTE 1.59 0.96 2.02 1.14 0.63 0.53 0.61 0.73 0.86 0.88 AIRPORT 0.75 0.64 0.86 0.67 0.48 0.39 0.42 0.69 0.67 0.62 TRAFFIC 27818 26103 26329 27146 26427 26215 26685 27094 27844 29057 29500 29000 28500 28000 27500 27000 26500 26000 25500 25000 24500 Figure 9 Average daily traffic and ATFM delay per flight (En-route and Airport) 2008-2017 The average ATFM delay per flight was 1.50 minutes, a 2% decrease compared to 2016 (Figure 9). En-route ATFM delay was 0.88 minutes per flight (2% increase), above the 0.5 minutes per flight SES capacity target. Despite the high traffic increase relative to 2016 (4.4%) the en-route delay remained at similar figures. When comparing to 2008, there was significantly less en-route delay in 2017 (-45.7%). Airport ATFM delay per flight was 0.62 minutes, a decrease of 7.5% compared to 2016. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 15

AIRPORT STAFFING (ATC) AIRPORT DISRUPTIONS AIRPORT EVENTS ENROUTE DISRUPTIONS ENROUTE CAPACITY AIRPORT DISRUPTIONS (ATC) ENROUTE EVENTS AIRPORT CAPACITY (ATC) ENROUTE DISRUPTIONS (ATC) ENROUTE STAFFING (ATC) ENROUTE WEATHER AIRPORT CAPACITY AIRPORT WEATHER ENROUTE CAPACITY (ATC) 164 326 373 473 593 596 779 1442 2520 4179 5890 6745 8367 11086 0 2000 4000 6000 8000 10000 12000 En-route ATC capacity (25.5%), enroute weather (13.5%), airport weather (19.2%) and airport capacity (15.5%) were the main reasons for ATFM delay in 2017. Figure 10: Proportion of ATFM delays in 2017 Karlsruhe ACC was the main generator of ATFM delay in 2017 contributing to 18.6% of the average daily en-route ATFM delay, mainly due to en-route ATC capacity, en-route ATC staffing issues and en-route weather. In addition to Karlsruhe, en-route ATC capacity affected also Marseille, Brest, Maastricht and Barcelona. As for en-route ATC staffing, Nicosia and Langen ACCs were also affected. Amsterdam/Schiphol was the main generator of airport ATFM delay (5.6% of the total), representing an increase of 66% compared to 2016. The main reasons for this delay were airport weather and airport capacity. En-route weather delays affected mostly Karlsruhe, Maastricht, Marseille, Wien and Barcelona ACCs. Amsterdam/Schiphol, London/Heathrow, London/Gatwick, Istanbul/Ataturk and Frankfurt Main were the most affected airports by weather events. ATC industrial actions in France in March and in the autumn months affected mainly the French ACCs of Marseille, Brest and Bordeaux. See Planned Events and Disruptions for more information on disruptions and other events that affected the network throughout 2017. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 16

3.3.2.1 EN-ROUTE ATFM DELAYS ENROUTE WEATHER 4347 5890 ENROUTE EVENTS 2148 779 ENROUTE DISRUPTIONS 900 473 ENROUTE CAPACITY 423 593 ENROUTE DISRUPTIONS (ATC) 2913 2520 ENROUTE STAFFING (ATC) 3464 4179 ENROUTE CAPACITY (ATC) 9675 11086 0 5000 10000 15000 Av. Daily Delay (min) ENROUTE WEATHER 0.16 0.20 ENROUTE EVENTS 0.08 0.03 ENROUTE DISRUPTIONS 0.03 0.02 ENROUTE CAPACITY 0.02 0.02 ENROUTE DISRUPTIONS (ATC) 0.10 0.09 ENROUTE STAFFING (ATC) 0.12 0.14 ENROUTE CAPACITY (ATC) 0.35 0.38 0.00 0.10 0.20 0.30 0.40 Av. Daily Delay (min) Figure 11: 2017 average daily en-route ATFM delays Figure 12: 2017 average en-route ATFM delay per flight The 6.9% increase in the average daily en-route ATFM delay is due to an increase of enroute weather delays (35.5%) and en-route capacity and staffing delays (16.9%). Delays due to en-route disruptions and events decreased by 37% compared to 2016. This represents 2,200 fewer minutes of daily delay approximately the same amount of additional delay generated by capacity and staffing issues compared to 2016. An overview and information on individual ACCs can be found in En-Route Performance and in Annex II - ACC. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 17

3.3.2.2 AIRPORT/TMA ATFM DELAYS Airport ATFM delay decreased by 3.6% in 2017 compared to 2016. Airport capacity and weather contributed 83.9% of the total airport delays in 2017. Airport weather was the main reason of delay. Even though the airport capacity delays decreased by 20.2% compared to the 2016, Istanbul/Sabiha Gökcen and Istanbul/Ataturk remained the main contributors to airport capacity delays. The average daily airport ATFM delay due to airport weather increased by 4.3% in 2017 - Amsterdam/Schiphol alone generated 1,337 minutes of weather delay per day. Adverse weather conditions particularly impacted airport operations in January, September and December. AIRPORT WEATHER 8022 8367 AIRPORT EVENTS 142 373 AIRPORT DISRUPTIONS 238 326 AIRPORT CAPACITY 8448 6745 AIRPORT DISRUPTIONS (ATC) 483 596 AIRPORT STAFFING (ATC) 343 164 AIRPORT CAPACITY (ATC) 1019 1442 0 2000 4000 6000 8000 10000 Av. Daily Delay (minutes) Figure 13: 2017 average daily airport/tma delays ATFM delays due to airport staffing (ATC) and airport capacity (-0.2%) decreased. There were increased ATFM delays due to airport weather (+4.3%), as well as airport capacity (ATC), airport disruptions (ATC), airport disruptions and airport events. During 2017 NM continuously provided support and recommendations to major airports facing local capacity challenges and/or high delay levels. NM gave special attention to some regions and airports. NM focussed especially on continuous implementation of the Greek Action Plan and implemented the airport function within the NMOC which provided tactical support on hot-spot airports (see Greek islands Summer). An overview and information on individual airports can be found in 6 Airports and in Annex III - Airports. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 18

3.4 CAPACITY Figure 14: Traffic, Delay and Effective capacity In 2017 the effective capacity indicator increased by 4% over the whole European ATM network (an increase of 3% for the summer season), when compared to the corresponding periods of 2016. The capacity at European level is quantified using the effective capacity 4 indicator of the Performance Review Commission (PRC). 4 The "effective capacity indicator corresponds to the volume of traffic that could be accommodated with an average of 1 min en-route delay/flt, taking into account all causes. It is described in PRR 5, Annex 6. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 19

3.5 FLIGHT EFFICIENCY 3,50% 3,00% 2,50% 2,00% 1,50% 1,00% 0,50% 0,00% 3,04% 2,96% RTE-DES 2,80% 2,63% 2,55% 2,47% 2,36% 2011 2012 2013 2014 2015 2016 2017 Figure 15: Average route extension due to airspace design (RTE DES) The average route extension due to airspace design decreased from 2.47% in 2016 to 2.36% in 2017, exceeding already the target set for 2019. The indicator reached a historically low level in December 2017 with 2.31% and allowed potential average savings of nearly 7,000 nautical miles per day. 5.00% 4.50% 4.00% 3.50% 3.00% 2.50% 2.00% 4.91% 4.87% 4.86% 4.70% 4.74% 4.82% 4.62% 2011 2012 2013 2014 2015 2016 2017 Figure 16: Yearly evolution of flight-planning indicator (KEP) The flight planning indicator (KEP) measures the average route extension based on the latest filed flight plan. It decreased from 4.82% in 2016 to 4.62% in 2017 (for the NM area). The targets of 4.44% for SES area, 4.17% for NM area were not met (Figure 16). The indicator decrease shows potential gains of approximately 5.1 million fewer nautical miles planned. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 20

3.40% 3.20% 3.00% 2.80% 2.60% 2.40% 2.20% 2.00% 3.31% 3.20% 3.14% 2.72% 2.77% 2.93% 2.77% 2011 2012 2013 2014 2015 2016 2017 Figure 17: Yearly evolution of the actual trajectory indicator (KEA) The actual trajectory indicator (KEA) decreased to 2.77% (Figure 17) for the NM area, thus meeting the target (2.78%). The indicator decrease shows actual gains of approximately 3.7 million fewer nautical miles flown. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 21

4 TRAFFIC IN DETAIL Figure 18 : IFR Flights per day in NM Area There were 10.6 million flights controlled in the NM area. The summer months (May to October incl.) were the highest ever recorded, with July, August and September each totalling more than one million flights. All market segments contributed to the strong flight growth, with the exception of the military flights. In 2017 as a whole, average daily flights remained 4.4% 5 above the 2016 traffic levels. The busiest ever traffic levels in Europe occurred during the summer months (4.8% growth on average) with the months June to October (incl.) growing faster than STATFOR s February forecast iii. Although the low-cost segment remained the main driver of growth throughout the year (except in December), the charter and all-cargo segments recorded the fastest growth reaching double digit increases during the summer. There was an intense growth of flights in Western Europe and in the Russian Federation (the latter owing to a faster economic recovery than initially anticipated). Also to be noted was the recovery of flights to North- African States (mostly Egypt). Traffic growth started to slow down with the start of the winter schedules, initially due to Ryanair capacity cuts and then the ceasing of operations of Monarch Airlines, Air Berlin and Niki. In addition, in early December, adverse weather conditions led to high number of flight cancellations. 5 The growth is calculated based on average daily number of flights to remove the effect of the leap year in 2016. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 22

4.1 NETWORK CONTRIBUTORS Figure 19 shows the main contributors to the local traffic growth for 2017 (referring to local traffic only, ie. excluding overflights). The vast majority of States contributed to the strong growth of traffic in Europe adding each more than 50 daily flights to the network in 2017 compared with 2016. Figure 19: Main changes to local traffic on the European network In 2017 the top three contributors to the network s local 6 traffic growth were Spain (+212 daily flights), the United Kingdom (+176 daily flights), and Germany (+137 daily flights) boosted by strong growth on flows within Western Europe. Since September, the German and UK contributions slowed down (internal for Germany, arrivals-departures for the UK) due to the failures of Air Berlin and Monarch Airlines, respectively. Italy added 108 daily flights, and Portugal (excluding Azores) added 98 daily flights and kept their contribution constant throughout the end of the year. Ukraine s local traffic was up 20% thanks to the state s flows to Egypt and to Turkey along with its Eastern European flow that remained strong. Flows to and from Turkey have increased their contribution for 2017 with 86 extra daily flights to the European network mainly due to the recovery on the flow from the Russian Federation. Although not shown on the graph, only two states reported fewer flights in 2017: Norway which saw 22 fewer daily flights due to a weak domestic flow and Denmark with 18 fewer daily flights due to weak international flows to North-Western Europe (Germany, UK, Sweden and Norway). 6 IFR movements: internals, international arrivals and departures, excluding overflights. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 23

4.2 ROUTING ASPECTS Figure 20: Biggest changes 7 in traffic patterns in 2017 The most noticeable change was the recovery of traffic between the Russian Federation and Turkey which increased by 243% (+101flights/day) in 2017 as a whole (vs. 2016) owing mostly to the resumption of charter flights (+743% on 2016) as the travel ban 8 was no longer in force, and flight counts returned to pre-crisis levels. Overall, there was a 23% increase on flows between Europe and the Russian Federation. This growth had an impact on overflights over Romania and Bulgaria (+150%) as well as over Ukraine (+118%). Additional traffic leading to strong growth on the West axis continued. The five busiest flows added a combined 236 flights per day and grew by 7.5% in 2017 (vs. 2016) with flows between UK and Spain (+45 flights/day), internal Spain (+29 flights/day) and between Germany and Spain (+20 flights/day) contributing most to the growth. On the downside, flights between Germany and Turkey decreased by 8.1% and saw 21 fewer flights per day in 2017 compared with 2016. In general flows between Turkey and Western Europe remained in deficit during most of the year, but started to show signs of recovery from November onwards. Traffic flows between Europe and Egypt recovered as adverse travel advice was alleviated, this led to a 41% growth although flight levels remained 34% lower than in 2010. Leisure traffic was the main contributor to the overall growth as it climbed 89%. The flow between Germany and Egypt was the most dynamic and went up 56% (+18 flights/day). 7 Width of the arrow in Figure 20 shows the relative size of the flow. Value on the arrow is the growth rate (2017 vs. 2016). 8 Russia lifted ban on charter flights to Turkey at the end of August 2016 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 24

4.3 EXTRA-NM AREA PARTNERS Figure 21: Daily Flight Change and Growth in 2017 (ECAC area) As shown in Figure 21, the United States remained the number one destination outside Europe in 2017, closely followed by the Russian Federation, in particular during the summer months. The United States recorded 490 daily departures on average (+24 flights/day), an increase of 5% compared with 2016 partly due to long-haul low-cost flights growing by 38% on the North Atlantic flows in 2017 (vs. 2016). The Russian Federation recorded 423 departures per day, an increase of 22.6% on 2016 mainly due to the return of charter flights to Turkey. The recovery of traffic started at the beginning of 2017. Flows between Europe and the United States and between Europe and the Russian Federation accounted for 32% of all departures from ECAC to countries outside Europe. The United Arab Emirates was the third extra-european partner with 164 daily departures on average, an increase of 1.2% on 2016. Other states in the Middle-East such as Oman, Iran, Qatar and Kuwait altogether recorded a 16% increase on their flows from/to Europe (they are not shown on Figure 21 as these are small flows). Although not shown on the graph, flows between Europe and Egypt recorded the best progression with an increase of 40.2% in 2017 (vs. 2016) owing to a robust recovery of leisure traffic from Germany and Ukraine mainly. The recovery of traffic started at the beginning of 2017. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 25

4.4 AIRPORT TRAFFIC EVOLUTION Departures from the airports in the network increased by 3.8% in 2017. Nº ICAO ID AIRPORT NAME TFC % Nº ICAO ID AIRPORT NAME TFC % 1 EHAM AMSTERDAM/SCHIPHOL 696 3.9% 26 LSGG GENEVA 252 0.8% 2 LFPG PARIS CH DE GAULLE 661 0.9% 27 LIMC MILANO MALPENSA 245 7.5% 3 EGLL LONDON/HEATHROW 652 0.5% 28 EFHK HELSINKI-VANTAA 242 5.2% 4 EDDF FRANKFURT MAIN 651 3.0% 29 EPWA CHOPINA W WARSZAWIE 235 11.9% 5 LTBA ISTANBUL-ATATURK 618-0.6% 30 EDDT BERLIN-TEGEL 235-6.4% 6 EDDM MUENCHEN 551 3.0% 31 EDDH HAMBURG 211 1.4% 7 LEMD ADOLFO SUAREZ MADRID-BARAJAS 531 2.7% 32 LTAI ANTALYA 211 25.6% 8 LEBL BARCELONA/EL PRAT 443 5.2% 33 LKPR PRAHA RUZYNE 197 9.4% 9 LIRF ROMA/FIUMICINO 407-5.1% 34 LFMN NICE-COTE D'AZUR 195 2.1% 10 EGKK LONDON/GATWICK 392 2.4% 35 EDDK KOELN-BONN 190 3.8% 11 LSZH ZURICH 361 0.6% 36 LLBG TEL AVIV/BEN GURION 189 17.4% 12 EKCH KOBENHAVN/KASTRUP 355-2.2% 37 LEMG MALAGA/COSTA DEL SOL 185 10.8% 13 ENGM OSLO/GARDERMOEN 344 2.7% 38 EGGW LONDON/LUTON 185 3.4% 14 ESSA STOCKHOLM-ARLANDA 341 6.6% 39 EGPH EDINBURGH 174 4,8% 15 LOWW WIEN SCHWECHAT 329-0.3% 40 EGBB BIRMINGHAM 163 8.0% 16 EBBR BRUSSELS NATIONAL 319 7.1% 41 EDDS STUTTGART 162-0.6% 17 LFPO PARIS ORLY 318-2.2% 42 LIML MILANO LINATE 159-0.6% 18 EIDW DUBLIN 305 4.5% 43 GCLP GRAN CANARIA 159 6.0% 19 EDDL DUESSELDORF 303 2.4% 44 LROP BUCURESTI/HENRI COANDA 159 8.2% 20 LTFJ ISTANBUL/SABIHA GOKCEN 293-4.9% 45 LTAC ANKARA-ESENBOGA 154 12.4% 21 LEPA PALMA DE MALLORCA 286 5.9% 46 LFLL LYON SAINT-EXUPERY 154 2.0% 22 EGCC MANCHESTER 279 6.5% 47 LHBP BUDAPEST LISZT FERENC INT. 140 6.9% 23 LPPT LISBOA 278 11.7% 48 EDDB SCHOENEFELD-BERLIN 137 6.2% 24 LGAV ATHINAI/ELEFTHERIOS VENIZELOS 260 4.4% 49 LFBO TOULOUSE BLAGNAC 136 8.8% 25 EGSS LONDON/STANSTED 258 5.3% 50 LFML MARSEILLE PROVENCE 134 1.5% Table 1: Top 50 airports per average daily departure traffic in 2017 Amsterdam/Schiphol remained the busiest airport in terms of average daily departures in 2017 with 696 average daily departures. Traffic at Amsterdam/Schiphol airport showed continuous growth since 2010. Paris Charles de Gaulle also remained the second busiest with 661 average daily departures. Six of the top ten airports (Amsterdam/Schiphol, Frankfurt/Main, Munich, Madrid/Barajas, Barcelona/El Prat and London/Gatwick) had an increase in average daily flights in 2017. Paris/Charles De Gaulle, London/Heathrow and Istanbul/Ataturk, all capacity constrained, remained at similar levels as 2016. Overall, the largest increase was at Antalya airport by 25.6% recovering from the reduction of flights in 2016, but the traffic levels were still lower than in 2015. Tel Aviv/Ben Gurion (17.4%), Ankara (12.4%), Warsaw (11.9%), Lisbon (11.7%), Malaga (10.8%) and Prague (9.4%) airports recorded significant traffic increase in 2017. A significant decrease of traffic was recorded at the airports of Berlin/Tegel (-6.4%), Rome/Fiumicino (-5.1%) and Istanbul/Sabiha Gökcen (-4.9%). These reductions were mainly caused by flight cancellations (Alitalia) and suspension of operations (Air Berlin and Borajet) of airlines operating in those airports. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 26

4.5 AIRLINE INDUSTRY The traffic share of each market segment is changing year-on-year. Traditional schedule accounted for the highest number of flights in 2017 (52.1%) while five years ago its share was at 55%. On the other hand, the low-cost segment has increased its market share from 25% in 2012 to over 30.6% in 2017. With the exception of the military segment, all the other market segments contributed to the strong flight growth in 2017 (Figure 22). Figure 22: Traffic Development per Market Segment Although the low-cost segment remained the main driver of flight growth in 2017 (+485 flights per day on average), it lost its dominant position in terms of percentage growth in favour of the commercial non-scheduled (charter) and all-cargo segments which grew by 9.1% and 8% respectively when low-cost increased by 5.7% in 2017 (vs. 2016). The low-cost segment has grown consistently between 5% and 10% until October 2017. It then decelerated during the last quarter (0.6% in November and -0.6% in December). This was impacted by the collapse of Monarch Airlines and Air Berlin, and by the cuts in capacity of Ryanair iv in the winter by flying 25 fewer aircraft from November onwards in an attempt to avoid roster-related cancellations. The charter segment had the fastest growth and recovered from April 2017 owing to the recovery of Russian traffic growth, together with the growth of Eastern European states, and the recovery of Egypt and Turkey as touristic destinations. The all-cargo segment boosted by a strong intra-europe cargo market (+9% on 2016) jumped to an overall growth rate of 8% in 2017 vs. 2016, but with widening seasonal variation: 14% growth in summer growth, while November and December fell below 4%. The business aviation segment reported its first year of traffic growth since 2011 and recorded growth with an annual growth rate of 6.2%. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 27

Traditional scheduled was the weakest segment as it grew at an annual rate of 3% in 2017 (vs. 2016) although it reached its strongest growth rate for the last 10 years and grew even faster than low-cost in November and in December 2017. Military IFR was the only segment to decline in 2017 as it went down by 3.9% compared with 2016. Low-cost carriers were indeed the ones adding the most flights to the European network (three of the top five airlines). Ryanair added 162 flights per day and remained by far the main contributor although in 2016 it recorded 200 additional flights per day (+23% on 2017). easyjet UK added 85 flights per day and Wizz Air added 65 flights per day. Two traditional scheduled carriers completed the list and they were LOT Polish Airlines (+62 flights/day) and Lufthansa (+46 flights/day). Figure 23: Crude Oil and Fuel Prices Evolution As Figure 23 shows, oil prices averaged out at 49 per barrel in 2017 and were higher than in 2016 ( 41 per barrel) but remained below the peaks of 2011-2014. Oil prices fluctuated from 51 per barrel during the first quarter of 2017 down to 47 per barrel during the second quarter and reached their lowest during the summer months (Jun-Aug) at 43 per barrel and their highest in November and December, at 54 per barrel. Some temporary shortages in refining capacity meant that the margin of fuel over oil price increased: the end result was fuel prices increasing more than oil over the year. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 28

Figure 24: Deflated ticket prices in Europe As Figure 24 shows, airline ticket prices started to increase with the start of the summer (May-August) +3.5% on average and in December (+4%). Compared with 2016, airline ticket prices in Europe were on average 3% more expensive in 2017 (-3.6% in March to +14.3% in April due to the Easter shift). Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 29

4.6 FLIGHT REDUCTIONS In 2017, the operational cancellation rate was 1.5%, a decrease of 0.1p.p. comparing to 2016. On 23 January, there were disruptions at Amsterdam Schiphol, London Heathrow and London Gatwick due of fog. Further weather related disruptions also occurred at Istanbul Ataturk on the 10 January. Figure 25 Monthly Rate of Operational Cancellations 2016-2017 Seasonal weather (mainly snow and high winds) impacted several airports throughout February, especially Amsterdam Schiphol, Istanbul Ataturk, London Gatwick and London Heathrow. On 23 February the storm Doris affected the network and airlines cancelled flights at all major London airports, as well as Amsterdam Schiphol and Frankfurt Main. French ATC industrial action from 6 March to 10 March generated cancellations in France; NM estimates there were 2250 fewer flights during the action. The Italian ATC industrial action on 20 March generated en-route ATFM delay and airport ATFM delay in Italy. NM estimates there were 420 fewer flights during the action. The industrial action by ground handling on 13 and 14 March at Berlin Tegel and Berlin Schönefeld caused the cancellation of approximately 920 and 500 flights respectively. Weather events impacted airlines during the summer season. Oslo experienced weather issues on 24 April while London Gatwick was affected on 28 May. On 22 June the London airports, Amsterdam/Schiphol and Frankfurt s airports operations were also heavily impacted by weather. Later on the summer, Amsterdam Schiphol experienced low visibility issues on the 16 and 30 August. Overall, the summer saw fewer significant disruptions. In September a peak in cancellations occurred following French ATC industrial actions in Marseille, Brest and Reims ACC s on the 11, 12, 13 September as well as 21 September. Towards the end of the year December saw an increase in the rate of cancellations following a month where winter storm affected north western Europe, notably London Heathrow, London Stansted, London Luton, Manchester, Amsterdam Schiphol and Oslo airports. There was also an Italian strike on the 15 December. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 30

En-route delay (min) Nr of Scenarios EUROCONTROL 5 EN-ROUTE PERFORMANCE High traffic was forecasted in several capacity constrained areas for the summer and ANSPs introduced RAD restrictions and scenarios to better manage the demand and available capacity. There were over 11,600 ATFM scenarios applied in 2017, 61% of which were flight level restrictions. There were 50% more scenarios than in 2016 and the double of those of 2015 (Figure 26). Reims, Madrid, Maastricht, Brest and Paris have all applied more than 1,000 scenarios each. 30,000 16,000 25,000 20,000 16,225 19,725 23,871 25,519 11,627 14,000 12,000 10,000 15,000 10,000 5,000 5,201 5,846 7,899 8,000 6,000 4,000 2,000-2014 2015 2016 2017 - Avg. En-route delay (min) Nr of Scenarios Figure 26 Network En-route delays vs Scenarios applied Despite the increase in ATFM scenarios, ATFM regulation usage increased with over 1 million regulated flights, a 30% increase on the same period for 2016. 80% of the regulated flights had an ATFM delay of less than 15 minutes (75% in 2016). Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 31

KARLSRUHE UAC MAASTRICHT UAC MARSEILLE ACC BREST U/ACC BORDEAUX ALL ACC Proportion of total en-route delay NICOSIA ACC BARCELONA ACC LANGEN ACC REIMS U/ACC WIEN ACC LONDON ACC LONDON TMA TC GENEVA ACC MADRID ALL ACC PARIS ALL ACC LISBOA ACC/UAC ATHINAI CONTROL BRUSSELS CANAC ZURICH ACC WARSZAWA ACC Average en-route ATFM delay (min) EUROCONTROL 5.1 HOT SPOTS 20% 15% 10% 5% 0% 4740 3377 3262 2568 1283 1090 1068 779 709 672 488 455 406 396 391 316 255 253 249 240 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0 ENROUTE CAPACITY (ATC) ENROUTE STAFFING (ATC) ENROUTE DISRUPTIONS (ATC) ENROUTE CAPACITY ENROUTE DISRUPTIONS ENROUTE EVENTS ENROUTE WEATHER PREV YEAR Figure 27: Top 20 en-route ATFM delay locations during 2017 Figure 27 shows the top twenty en-route ATFM delay generating locations for 2017 with respect to total ATFM delays. Figures are in minutes and they represent the average daily delays for the individual locations. The top twenty delay locations generated 90.1% of en-route ATFM delay in 2017. The top six locations (Karlsruhe, Maastricht, Marseille, Brest, Bordeaux and Nicosia) generated 64% of all en-route ATFM delay. Of the top six ACCs, Karslruhe, Maastricht, Marseille and Nicosia increased their average daily en-route delay (175.4%, 25.2%, 156.2%, and 95.7%, respectively) while Brest and Bordeaux have decreased their en-route delays (-45.9% and -26.1%, respectively). For more details on the summer performance of these ACCs please see 5.4.1 Demand and Capacity Monitoring. Other significant decreases on the top 20 ACCs in terms of en-route ATFM delays occurred in Paris (-64.6%), Brussels (-67.9%) and Warsaw (-69.9%). Wien ACC recorded the highest increase, with an ATFM delay 5 times higher than 2017, mainly due to the impact of weather events during the summer. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 32

NICOSIA ACC MARSEILLE ACC KARLSRUHE UAC BREST U/ACC MAASTRICHT UAC BORDEAUX ALL ACC BARCELONA ACC WIEN ACC CANARIAS ACC/FIC REIMS U/ACC PALMA ACC LANGEN ACC GENEVA ACC LISBOA ACC/UAC ATHINAI CONTROL BRUSSELS CANAC MAKEDONIA CONTROL MADRID ALL ACC AMSTERDAM ACC(245-) En-route delay per flight (min) ZAGREB ACC EUROCONTROL Nicosia and Marseille had the greatest en-route delay per flight - both with more than 1 minute per flight (Figure 28). Brest, the main generator of en-route delay in 2016, recorded 50% less delay in 2017. Paris, Brussels, Warsaw, Scottish and Olso ACCs have also decreased considerably their en-route delay per flight. 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 1.11 1.08 0.93 0.88 0.67 0.49 0.45 0.29 0.26 0.26 0.26 0.22 0.22 0.20 0.17 0.15 0.15 0.13 0.13 0.13 ENROUTE CAPACITY (ATC) ENROUTE STAFFING (ATC) ENROUTE DISRUPTIONS (ATC) ENROUTE CAPACITY ENROUTE DISRUPTIONS ENROUTE EVENTS ENROUTE WEATHER PREV YEAR Figure 28: Top 20 en-route ATFM delay per flight locations during 2017 Of the top twenty delay locations, the largest increases in en-route delay (more than 100%) were in Marseille, Karlsruhe, Wien, Geneva, Makedonia and Zagreb ACCs. En-route ATC capacity and staffing (61%) were the main reasons of en-route ATFM delay per flight for the top six ACCs followed by en-route weather (20%) and en-route disruptions and events (8%). En-route ATC capacity delays were higher than the network ATC capacity delay per flight (0.38 min/flt) in Marseille, (0.60 min/flt), Brest (0.54 min/flt) and Nicosia (0.52 min/flt). Barcelona (0.31min/flt) and Maastricht (0.28min/flt) complete the top five ACCs in terms of ATC capacity delays in 2017. En-route staffing issues affected mainly Karlsruhe (0.39 min/flt) and Nicosia (0.50 min/flt) ACCs. En-route delays due to weather affected mostly Karlsruhe (0.26min/flt), Maastricht (0.21 min/flt), Wien (0.21 min/flt) and Marseille (0.18 min/flt) ACCs. En-route ATC disruptions and En-route events affected mostly the French ACCs of Brest (0.25 min/flt), Marseille (0.24 min/flt) and Bordeaux (0.24 min/flt). Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 33

5.2 PLANNED EVENTS AND DISRUPTIONS En-route ATFM delays due to planned events (system upgrades/transition projects) and disruptions decreased by 37% in 2017 (see 3.3.2.1 En-Route ATFM Delays). In the following paragraphs there is an overview of the main events and disruptions in 2017. 5.2.1 EN-ROUTE PLANNED EVENTS The new ATM system implementation at Bordeaux ACC (ERATO) continued in 2017 and went through its final transition stage in May. The good network cooperation with Bordeaux ACC ensured a smooth transition, with less than 100,000 minutes of ATFM delay generated during the whole duration of the upgrade (30,000 minutes in 2017). The PSS system implementation at Langen ACC generated 42,800 minutes of delay. Maastricht UAC introduced operational configuration changes (3 rd layer in the East of the Brussels sector group) and implemented a new VCS system, which generated 18,000 and 110,000 minutes of ATFM delay, respectively. Table 2 shows the system upgrade/transition projects that imposed capacity reductions in several ACCs and that were included in the NOP Transition Plans v. Major Projects / Special Events 9 January - March April - June July -September October - December Albania Tirana ACC ATM system upgrade France - Bordeaux ACC Implementation of ERATO system Germany - Langen ACC Implementation of PSS system EBG10 Germany Karlsruhe ACC Implementation of icas KAR Ireland Shannon ACC Low level FRA Maastricht MUAC Third layer (Brussel sector group) Switzerland Geneva ACC Virtual center Switzerland Zurich ACC Harmonisation release Virtual center UK London TC Implementation of ExCDS North sect. UK Scottish ACC PLAS Deployment 2 Table 2 System Upgrade / Transition Projects 9 Does not include postponed projects and should not be considered as exhaustive Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 34

5.2.2 EN-ROUTE DISRUPTIONS Industrial action contributed to 8.5% of total en-route delay in 2017 (including indirect delays in the neighbouring ACCs due to on-loading traffic). There were roughly 800,000 minutes of delay attributed to strikes, representing a 33% decrease when compared to 2016. An estimated 4,600 flights were cancelled during the strike days. The French industrial actions in March, September, October and November contributed to most of the industrial action delays in 2017. Table 3 shows unplanned events or disruptions 10 certain ACCs in 2017. that imposed capacity reductions in Date Location Event Traffic Impact (Cancellations) ATFM Delay Impact 06 January Brest ACC Radar and frequency failure None 1,169 minutes 13 January Reims ACC Frequency failure None 1,489 minutes 14 January Padova ACC Radar failure None 4,948 minutes 26/27 January 14 February Malmö ACC 19-24 February Nicosia ACC Radar maintenance None 3,977 minutes Non-availability of ATM back-up system due to weather phenomena None 5,616 minutes Brest ACC Frequency failure None 1,775 minutes 23 February Karlsruhe ACC Frequency failure None 1,082 minutes 06/07/08/09 /10 March France ATC industrial action 2.250 flights French ACCs 300,198 minutes Neighboring ACCs 23,849 minutes 11 March Geneva ACC Radar failure None 5,257 minutes Lisbon ACC - 32.655 minutes 11/12/14 Lisbon ACC ATM system failure None Canarias ACC 5,722 minutes March Sevilla ACC 2,882 minutes 20 March Italy ATC industrial action 420 flights Italian ACCs 6,582 minutes 06 April Brest ACC ATM system failure None 1,307 minutes 10/11 April Brest ACC Communication system failure None 6,309 minutes 17 April Bordeaux ACC Frequency failure None 1,067 minutes 27 April Bordeaux ACC Surveillance issue None 1,315 minutes 27 April Marseille ACC Surveillance issue None 1,164 minutes 10 May Lisbon ACC Flight data processing system instability None 12 May Romania ATC industrial action 210 overflights rerouted 2,418 minutes Bucharest ACC - 2,413 minutes Beograd ACC 1,183 minutes 13 May Marseille ACC Frequency failure None 1,371 minutes 17 May Greece ATC industrial action None Greeks ACCs - 8,067 minutes 23 May Lisbon ACC Communication failure between Lisbon ACC and Santa Maria ACC None 30 May Romania ATC industrial action 400 overflights rerouted 2,398 minutes Bucharest ACC 5,236 minutes Beograd ACC 2,396 minutes 01-30 May Geneva ACC Radar instability None 7,346 minutes 01-30 May Zurich ACC Radar instability None 2,272 minutes 06 June Brussels ACC ATC equipment failure None 3,039 minutes 13 June Bordeaux ACC FDPS failure None 8,975 minutes 19 June Paris ACC Frequency failure None 1,723 minutes 10 The main source for the event description is the remark field on the NM ATFM Regulation (ANM) Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 35

Date Location Event Traffic Impact (Cancellations) ATFM Delay Impact 21 June Shanwick Oceanic Instability of telecommunication system in Gander Oceanic None 3,488 minutes 22 June Brest ACC Frequency failure None 2,900 minutes 25 June Marseille ACC Frequency failure None 4,048 minutes 09 July Marseille ACC Frequency failure None 1,077 minutes 10 July Beograd ACC Radar failure None 2,165 minutes 15 July Shanwick Oceanic Communication failure in Ottawa Communication Center None 8,435 minutes 21 July Bordeaux ACC Frequency failure None 1,101 minutes 06 August Maastricht ACC ATC equipment failure None 20,698 minutes 20 August Lisbon ACC FDPS failure None 1,985 minutes 11/12/13 September 21 September France ATC industrial action 300 flights French ACCs 106,891 minutes Neighboring ACCs 17,457 minutes France ATC industrial action None French ACCs 15,902 minutes 01 October Karlsruhe ACC FDPS failure None 15,478 minutes 02 October Lisbon ACC Surveillance system failure None 3,131 minutes 09/10/11 October France ATC industrial action 1.200 flights French ACCs 243,066 minutes Neighboring ACCs 31,841 minutes 27 October Munich ACC Communication system failure None 1,144 minutes 15 November 16 November 15/16/17 November 15 December 19-30 December 09/10/11 December Langen ACC Communication system failure None 6,464 minutes Brest ACC Frequency failure None 1,087 minutes France ATC industrial action 150 flights French ACCs 36,274 minutes Neighboring ACCs 5,731 minutes Italy ATC industrial action 285 flights 6,265 minutes Lisbon ACC Frequency failure None 7,542 minutes Brest ACC SSR code allocation issue between Madrid and Brest ACCs None Table 3: Unplanned Events/Disruptions 3,949 minutes Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 36

Movements per day En-route delay per flight (min) EUROCONTROL 5.3 CAPACITY EVOLUTION The capacity at European level is quantified using the "effective capacity" 4 indicator of the Performance Review Commission (PRC) that takes into account traffic and delay evolution. Between 1999 and 2017, traffic increased by 32%, the effective capacity of the network increased by 66% and the average en-route ATFM delay per flight decreased by 81%. In 2017 the effective capacity indicator increased by 4% over the whole European ATM network when compared to 2016. 35000 30000 25000 20000 15000 10000 5000 0 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 Traffic Av ER delay Effective Capacity EU delay target Figure 29 Annual Traffic, delay and capacity evolution The "effective capacity indicator takes into account en-route ATFM delays, for all reasons, including weather, disruptions and significant events: system failures, industrial action, implementation of new ATM systems. Figure 30 shows the monthly evolution of the effective capacity of the European ATM for the last 10 years. The highest value ever of this indicator was in September 2016 (Figure 30), while July-August 2017 saw the highest daily traffic values of the last 20 years. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 37

The year of 2017 was had the highest overall traffic and average daily traffic since the statistics have been gathered in Eurocontrol (last 20 years). 2013 was the year with lowest average delay per flight. Figure 30: ECAC 'Effective Capacity' Evolution per Month (2006-2017) Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 38

5.4 ACC ANALYSIS In the European Network Operations Plan (NOP) 2017 2019/21 vi values for each ACC: there are two delay The required en-route delay/flight performance to achieve annual network delay target in 2017 (0.5 min/flight). This is also known as the delay breakdown. The forecast delay based on 2017 NOP capacity planning, excluding disruptions such as industrial action and technical failures. An overview of the ACC performances in 2017 is in Table 4 and shows the traffic growth, capacity and delay for each ACC. Those ACCs that exceeded their reference value are highlighted in amber. The actual delay in 2017 was higher than the breakdown value reported in the NOP 2017-19/21 for only 15 out of 65 ACCs. EN-ROUTE DELAY TRAFFIC CAPACITY COUNTRY ACC ACC Code Breakdown 11 Forecast 12 Actual Forecast Summer 14 (Low/High) 13 Actual Annual 15 Actual NOP Plan Actual NETWORK NETWORK ALL_DNM 0.50 0.73 0.88 2.9% 4.9% 4.4% n/a n/a ALBANIA TIRANA ACC LAAAACC 0.09 0.03 0.00 +0.6% / +3.2% 3.1% 4.2% 5% 0% ARMENIA YEREVAN ACC UDDDACC 0.01 0.01 0.00 +6.0% / +9.5% 45.6% 41.0% sufficient 0% AUSTRIA WIEN ACC LOVVACC 0.20 0.20 0.29 1.0% / +4.5% 10.2% 9.7% 3% 1% AZERBAIJAN BAKU ACC UBBAACC 0.01 0.01 0.00 2.5% 4.6% 5.3% sufficient 0% BELGIUM BRUSSELS ACC EBBUACC 0.06 0.13 0.15 +1.7% / +4.3% 5.9% 6.4% 6% 13% BOSNIA SARAJEVO ACC LQSBACC 0.01 0.01 0.00 2.9% 13.1% 13.6% sufficient 8% BULGARIA SOFIA ACC LBSRACC 0.05 0.01 0.00 +2.1% / +4.5% 3.5% 3.3% 4% 2% CROATIA ZAGREB ACC LDZOACC 0.25 0.24 0.13-0.2% / +3.2% 9.0% 9.3% 3% 7% CYPRUS NICOSIA ACC LCCCACC 0.25 1.40 1.11 +5.7% / +10.3% 11.6% 11.9% 5% 3% CZECH REPUBLIC PRAGUE ACC LKAAACC 0.09 0.09 0.05 +2.3% / +5.6% 1.5% 2.0% 0% 3% DENMARK COPENHAGEN ACC EKDKACC 0.07 0.01 0.00 +1.5% / +3.5% 2.0% 1.7% 2% 0% ESTONIA TALLINN ACC EETTACC 0.03 0.02 0.02 +2.5% / +5.7% 10.0% 7.7% sufficient 2% EUROCONTROL MAASTRICHT UAC EDYYUAC 0.17 0.56 0.67 +2.8% / +5.1% 4.1% 4.1% 3% 2% FINLAND TAMPERE ACC EFESACC 0.09 0.01 0.00 +2.2% / +4.3% 8.1% 7.2% sufficient 0% FRANCE BORDEAUX ACC LFBBALL 0.13 0.07 0.49 +3.0% / +6.1% 5.5% 6.1% 14% 3% 11 The required en-route delay/flight performance to achieve annual network delay target in 2017 (0.5 min/flight), also known as delay breakdown - NOP 2017-2019/21 12 Forecast delay based on 2017 capacity planning including disruptions such as industrial action and technical failures at a statistical level of 0.1 min/flt - NOP 2016-19/20 (June 17 edition) 13 Low/High traffic forecast STATFOR Feb 2017 used for NOP capacity planning, variation in % compared to 2016. When not available, Base forecast is provided. 14 May to October (inc.) 15 Growth calculated based on the average daily traffic for 2017 and 2016 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 39

EN-ROUTE DELAY TRAFFIC CAPACITY COUNTRY ACC ACC Code Breakdown 11 Forecast 12 Actual Forecast Summer 14 (Low/High) 13 Actual Annual 15 Actual NOP Plan Actual FRANCE REIMS ACC LFEEACC 0.19 0.18 0.26 +1.4% / +4.2% 2.5% 3.2% 12% 8% FRANCE PARIS ACC LFFFALL 0.14 0.06 0.12 +1.0% / +3.6% 0.6% 1.4% 12% 1% FRANCE MARSEILLE ACC LFMMACC 0.15 0.07 1.08 +2.0% / +5.0% 4.7% 5.2% 12% -4% FRANCE BREST ACC LFRRACC 0.10 0.42 0.88 +3.8% / +6.6% 7.6% 8.0% 13% 10% FYROM SKOPJE ACC LWSSACC 0.20 0.16 0.03-0.5% / +3.4% 16.9% 18.3% 5% 3% GEORGIA TBILISI ACC UGGGACC 0.01 0.01 0.00 2.6% 14.9% 13.5% sufficient 0% GERMANY LANGEN ACC EDGGALL 0.23 0.27 0.22-0.2% / +2.7% 3.6% 3.3% 2% 5% GERMANY MUNCHEN ACC EDMMACC 0.20 0.04 0.04 +1.1% / +3.9% 4.0% 3.5% 1% 2% GERMANY KARLSRUHE UAC EDUUUAC 0.25 0.54 0.93 +2.0% / +4.7% 3.6% 3.9% 2% -2% GERMANY BREMEN ACC EDWWACC 0.07 0.04 0.12 +2.7% / +4.7% 2.4% 0.0% -1% 0% GREECE ATHINAI ACC LGGGACC 0.19 0.36 0.17 +1.2% / +3.9% 9.0% 9.1% 10% 8% GREECE MAKEDONIA ACC LGMDACC 0.15 0.15 0.15 +0.3% / +3.0% 7.4% 8.3% 3% 5% HUNGARY BUDAPEST ACC LHCCACC 0.05 0.03 0.01 +1.4% / +4.5% 6.5% 5.9% 2.5% 8% IRELAND DUBLIN ACC EIDWACC 0.04 0.02 0.00 +3.6% / +1.2% 4.2% 4.6% 3% 0% IRELAND SHANNON ACC EISNACC 0.05 0.01 0.00 +2.6% / +4.4% 0.5% 1.8% 2% 0% ISRAEL 16 TEL AVIV ACC LLLLACC n/a n/a 0.00 n/a 20% 19% n/a n/a ITALY BRINDISI ACC LIBBACC 0.01 0.01 0.00 +0.4% / +2.8% 10.3% 9.5% 8% 10% ITALY MILAN ACC LIMMACC 0.08 0.02 0.00 +0.9% / +4.3% 2.8% 2.2% 3% 0% ITALY PADOVA ACC LIPPACC 0.09 0.02 0.01 +1.3% / +4.3% 2.1% 2.5% 2% 3% ITALY ROME ACC LIRRACC 0.05 0.01 0.00-0.3% / +2.6% 5.3% 4.3% 2% 0% LATVIA RIGA ACC EVRRACC 0.04 0.01 0.00 +2.6% / +6.2% 10.3% 9.1% sufficient 0% LITHUANIA VILNIUS ACC EYVCACC 0.03 0.01 0.00 +2.2% / +6.0% 6.4% 5.6% sufficient 0% MALTA MALTA ACC LMMMACC 0.02 0.01 0.00 +1.6% / +3.7% 4.5% 5.2% sufficient 0% MOLDOVA CHISINAU ACC LUUUACC 0.01 0.01 0.00 +6.4% / +10.3% 21.9% 20.3% sufficient 0% MOROCCO 12 CASABLANCA ACC GMMMACC n/a n/a 0.01 n/a 6.9% 7.6% n/a n/a NETHERLANDS AMSTERDAM ACC EHAAACC 0.14 0.08 0.13 +2.9% / +3.5% 2.3% 3.1% 0.5% 1% NORWAY BODO ACC ENBDACC 0.10 0.01 0.01-2.9% / +0.7% 0.2% 0.3% sufficient 0% NORWAY OSLO ACC ENOSACC 0.14 0.04 0.03-1.3% / +1.7% 1.4% 0.9% sufficient 0% NORWAY STAVANGER ACC ENSVACC 0.12 0.04 0.00-0.1% / +1.2% -1.0% -2.3% sufficient 0% POLAND WARSAW ACC EPWWACC 0.23 0.31 0.12 +2.4% / +6.5% 4.9% 5.0% 7% 17% PORTUGAL LISBON ACC LPPCACC 0.11 0.56 0.20 +6.2% / +9.1% 8.9% 9.6% 2% 13% ROMANIA BUCHAREST ACC LRBBACC 0.01 0.01 0.01 +4.8% / +7.5% 11.0% 8.8% 0% 0% SERBIA. BELGRADE ACC LYBAACC 0.10 0.01 0.04 +0.3% / +3.3% 4.3% 5.4% 1% 0% 16 Only actual data (not planned) is reported for Israel and Morocco. Both states are now part of IFPS but they are still on a transition phase towards full ATFCM integration. Therefore, their operational plans were not part of the NOP 2017 2019/21 - their inclusion is envisaged for 2019. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 40

EN-ROUTE DELAY TRAFFIC CAPACITY COUNTRY ACC ACC Code Breakdown 11 Forecast 12 Actual Forecast Summer 14 (Low/High) 13 Actual Annual 15 Actual NOP Plan Actual MONTENEGRO SLOVAKIA BRATISLAVA ACC LZBBACC 0.10 0.04 0.03 +1.8% / +4.6% 3.9% 3.6% 2% 2% SLOVENIA LJUBLJANA ACC LJLAACC 0.22 0.03 0.00-0.2% / +4.4% 12.6% 12.9% 3% 9% SPAIN CANARIAS ACC GCCCACC 0.27 0.28 0.26 +4.3% / +6.6% 6.1% 7.5% 1% 4% SPAIN BARCELONA ACC LECBACC 0.23 0.31 0.45 +3.6% / +7.0% 5.1% 5.8% 10% 5% SPAIN MADRID ACC LECMALL 0.15 0.11 0.13 +4.0% / +6.8% 7.7% 7.7% 1% 7% SPAIN PALMA ACC LECPACC 0.19 0.20 0.26 +1.4% / +4.9% 4.5% 5.6% 3% 3% SPAIN SEVILLA ACC LECSACC 0.12 0.13 0.07 +3.9% / +6.5% 8.4% 8.8% 3% 5% SWEDEN MALMO ACC ESMMACC 0.07 0.10 0.01 +1.6% / +3.5% 6.4% 5.5% 1% 5% SWEDEN STOCKHOLM ACC ESOSACC 0.07 0.03 0.03 +0.5% / +2.5% 7.5% 6.5% 1% 0% SWITZERLAND GENEVA ACC LSAGACC 0.19 0.19 0.22 +1.5% / +4.7% 6.4% 6.0% 1% 2% SWITZERLAND ZURICH ACC LSAZACC 0.18 0.21 0.12 +1.4% / +4.2% 5.0% 4.5% 1% 4% TURKEY ANKARA ACC LTAAACC 0.15 0.01 0.00 +1.4% / +4.2% 7.4% 5.5% 10% 6% UKRAINE KYIV ACC UKBVACC 0.02 0.00 0.00 13.8% / +19.4% 25.6% 23.5% sufficient 0% UKRAINE DNIPROPETROVSK ACC UKDVACC 0.01 0.01 0.00 +11.1% / +14.7% 9.5% 8.1% sufficient 0% UKRAINE L VIV ACC UKLVACC 0.01 0.01 0.00 +4.3% / +9.9% 14.7% 16.0% sufficient 0% UKRAINE ODESA ACC UKOVACC 0.01 0.01 0.00 +13.7% / +19.1% 28.0% 25.8% sufficient 0% UNITED KINGDOM UNITED KINGDOM UNITED KINGDOM PRESTWICK ACC EGPXALL 0.13 0.14 0.07 +2.0% / +3.9% 7.3% 7.0% 11% 17% LONDON ACC EGTTACC 0.18 0.18 0.09 +2.6% / +4.8% 4.3% 4.2% 1% 6% LONDON TC EGTTTC 0.11 0.04 0.12 +1.4% / +3.4% 3.0% 3.1% 2% 6% Table 4: Overview of the ACC performances in 2017 Compared to the delay forecast, the performance of the following ACCs was better than foreseen 17 in the NOP 2017-2019/21: Lisbon, Nicosia, Warsaw, Athens, Skopje, Zagreb, London, Zurich, Malmo, Prestwick, Sevilla, Langen. The performance of the following ACCs was worse than foreseen 18 in the NOP 2017-2019/21: Marseille, Brest, Bordeaux, Karlsruhe, Barcelona, Maastricht, Vienna, Bremen, Reims, London TC, Paris, Palma, Amsterdam. Nicosia and Marseille ACCs had delays of more than 1 minute per flight, and three ACCs (Karlsruhe, Brest, Maastricht) recorded delays between 0.5 and 1 minute per flight. 17 This means the actual delay was lower than forecast delay by at least 0,05 18 This means the actual delay was higher than forecast delay by at least 0,05 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 41

H0000 H0200 H0400 H0600 H0800 H1000 H1200 H1400 H1600 H1800 H2000 H2200 Number of Sectors Minutes of Delay Average en-route delay per flight EUROCONTROL 5.4.1 DEMAND AND CAPACITY MONITORING SUMMER ATC capacity and staffing issues in six ACCs were the main cause of the en-route ATFM delays in the network (in addition to weather), particularly over the summer, and over the weekends Nevertheless, a number of ACCs which had been identified in the Network Operations Plan (NOP) as having capacity short-comings for the summer season have confirmed the proposed NOP measures and performed better than expected. These include Athens, Lisbon, Skopje and Zagreb ACCs all with traffic growth above 9%. The next paragraphs describe a more detailed view on the performance 19 of the most affected ACCs by capacity and staffing issues. These ACCs have provided feedback on the analysis to the Network Director of Operations Forum (NDOP) 20 and their views are published in Annex II ACC. 5.4.1.1 MARSEILLE Marseille ACC had 1.61 minutes of en-route delay per flight in the summer, more than 3 times the figure of 2016; and well above the planned NOP delay forecast (0.11 min/flt). Traffic increased by 4.7% with Sunday the busiest day of the week. Capacity delays happened mainly over the weekend, concentrated in the mornings and evenings, with regulations applied mostly in Marseille-East sectors. 25 20 15 10 5 0 1000 800 600 400 200 0 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.70 2.11 0.72 1.36 0.90 3.09 2.37 2017 2016 C+S Delay Figure 31 - Summer sector scheme (average) - Marseille ACC Saturdays (vs. 2016) 0.00 MON TUE WED THU FRI SAT SUN Figure 32 Weekly En-route delay per flight - Marseille ACC Summer 19 Traffic, delay and capacity figures refer to the months of May to October (incl.). Sector schemes are compared against the NOP ACC plans, when available. Average hourly delay figures include capacity and staffing delays only. 20 NDOP 18/19/12 20.03.18 Item 6.2 Action Paper: Analysis of the Network Delay Causes 2017 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 42

H0000 H0200 H0400 H0600 H0800 H1000 H1200 H1400 H1600 H1800 H2000 H2200 Number of Sectors Minutes of delay Average en-route delay per flight EUROCONTROL NM observed sector counts below 2016 levels during most of the days, but especially on Saturdays in the period 1500-1800. The traffic growth for Saturdays (2%) was below the growth for the other days - particularly concentrated at 0500 and in the period 1500-1600. The high capacity/staffing delay in Marseille occurred in periods when there were fewer sectors than last year. NM observed that delays occur each year in the summer, suggesting a regular capacity provision issue during the peak summer months. 5.4.1.2 NICOSIA Nicosia had 1.46 minutes of en-route delay per flight in the summer, the double of 2016. (NOP summer forecast was 1.9 min/flt). Traffic increased by 12% over the summer. Sunday and Thursday were the busiest days and also those with highest delays. 5 4 3 2 1 0 500 400 300 200 100 0 3.50 3.00 2.50 2.00 1.50 1.00 0.50 1.47 0.39 1.21 2.42 1.23 0.32 2.81 2017 Planned C+S Delay 0.00 MON TUE WED THU FRI SAT SUN Figure 33 - Summer sector scheme (average) - Nicosia ACC Sundays (vs. NOP plan) Figure 34 Weekly En-route delay per flight - Nicosia ACC Summer Overall, there were fewer sectors opened than what was agreed in the NOP. This was particularly noted on the 10h-18h period on Sundays, when most of the delays occurred. According to the NOP, 5 sectors were expected to be delivered during this period, but this only happened occasionally. Demand was very high on Sundays (11%) with traffic peaks that did not exist the previous year appearing at 1500 and 1900. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 43

H0000 H0200 H0400 H0600 H0800 H1000 H1200 H1400 H1600 H1800 H2000 H2200 Number of Sectors Minutes of Delay Average en-route delay per flight EUROCONTROL 5.4.1.3 KARLSRUHE UAC Karlsruhe UAC had 1.48 minuntes of en-route delay per flight in the summer, an increase of 155% comparing to 2016; and above the planned NOP summer delay forecast (0.77 min/flt). Summer traffic increased by 3.6% with the busiest day on Friday. Capacity and staffing delay was high and spread throughout the week but Saturday afternoon (12h-17h) was the most constrained period. The traffic growth on this day (5%) was spread across the period 0600-2000. 35 30 25 20 15 10 5 0 600 500 400 300 200 100 0 2.50 2.00 1.50 1.00 0.50 1.18 1.31 1.17 1.53 1.69 2.08 1.45 2017 NOP (June) C+S Delay 0.00 MON TUE WED THU FRI SAT SUN Figure 35 - Summer sector scheme (average) - Karlsruhe UAC Saturdays (vs. NOP plan) Figure 36 Weekly En-route delay per flight Karlsruhe UAC Summer Sector counts were generally in line with the June NOP capacity revision (which was 4 to 5 sectors below the levels in the first edition of the NOP). Nevertheless, the number of sectors open in busy periods deteriorated over the summer, especially on week days. It was also noted that the regulated rates decreased. Karlsruhe reported continuous staffing issues since the beginning of the year. This was particularly noted on the low share of regulations applied on elementary sectors (vs. collapsed sectors): 30% compared to almost 60% in 2016. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 44

H0000 H0200 H0400 H0600 H0800 H1000 H1200 H1400 H1600 H1800 H2000 H2200 Number of Sectors Minutes of delay Average en-route delay per flight EUROCONTROL 5.4.1.4 BREST Brest had 34% less delay in the summer of 2017 than the same period in 2016. However, the 1.05 minutes per flight of en-route delay for the summer was still above the planned NOP delay for the ACC (0.55 min/flt for the summer). Summer traffic increased by 7.6% with Saturdays being the busiest day of the week (9%). The change in demand was particularly high in the first-rotation peak in the weekend. Nevertheless, for those flights going through Brest during weekends, there was a significant drop of reactionary delays compared to 2016 (-20% on Saturdays and -12% on Sundays). 20 15 10 5 0 2017 2016 Delay 1000 Figure 37 - Summer sector scheme (average) Brest ACC Saturdays (vs. 2016) 800 600 400 200 0 2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0.74 1.65 0.54 0.91 0.88 1.47 Figure 38 Weekly En-route delay per flight Brest ACC Summer 1.20 MON TUE WED THU FRI SAT SUN NM observes that regulated rates for sectors increased by approximately 10% to 20% compared to those applied in 2016. The number of opened sectors increased, including in the first-rotations. NM believes Brest addressed many capacity issues during summer 2017. The share of regulations applied on elementary sectors (vs. collapsed sectors) increased from 59% to 72%, indicating a more efficient management of the airspace. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 45

H0000 H0200 H0400 H0600 H0800 H1000 H1200 H1400 H1600 H1800 H2000 H2200 Number of Sectors Minutes of delay Average en-route delay per flight EUROCONTROL 5.4.1.5 BORDEAUX Summer en-route delay in Bordeaux decreased by 20% compared to 2016. Despite the improvement, the en-route delay (0.63 min/flt) was still high when monitored against the NOP (0.11 min/flt for the summer). Summer traffic increased by 5.5%, with Saturday being the busiest day. Friday was also the day with most capacity delay, which remained high over the whole weekend. The morning period (0800-1200) and the evening (1600-2000) were the most delayed period on Fridays and Sundays while Saturdays had mostly morning issues. 20 16 12 8 4 0 300 250 200 150 100 50 0 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 1.38 0.80 0.51 0.63 0.54 0.30 0.27 MON TUE WED THU FRI SAT SUN 2017 2016 C+S Delay Figure 39 - Summer sector scheme (average) Bordeaux ACC Fridays (vs. 2016) Figure 40 Weekly En-route delay per flight Bordeaux ACC Summer Despite an increase in sector numbers in the morning and first-rotation compared to 2016, the ACC faced difficulties in coping with high demand in those periods. As for the evening issues on Fridays (also Sundays), there was a significant drop of sector numbers compared to 2016 (two sectors fewer on the 1600-1700 period), indicating a shift of sectors to earlier hours. Traffic increase was spread throughout the day on Fridays with a new peak at 1600 which, together with the drop-off in sectors, also contributed to the generation of delay. Mondays had similar traffic levels to those of Saturdays but significantly less delay. The number of sectors delivered was higher for that day, suggesting that the ACC was able to provide more sectors during the week than during weekends. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 46

H0000 H0200 H0400 H0600 H0800 H1000 H1200 H1400 H1600 H1800 H2000 H2200 Number of Sectors Minutes of delay Average en-route delay per flight EUROCONTROL 5.4.1.6 MAASTRICHT UAC Maastricht UAC had 1.04 minutes per flight of en-route delay in the summer, an increase of 21% compared to 2016 and very similar to the NOP forecast (0.94 min/flt for the summer). Summer traffic increased by 4.1%. The busiest day was Friday, while Saturday was the day with the lowest traffic (6%). The capacity/staffing delays were spread evenly throughout most of the week with Saturday recording the highest capacity delay. 20 16 12 8 4 0 500 400 300 200 100 2017 Planned Min. Planned Max. Delay Figure 41 - Summer sector scheme (average) Maastricht UAC Saturdays (vs. NOP plan) 0 2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0.71 0.89 1.19 1.55 1.06 0.84 1.07 MON TUE WED THU FRI SAT SUN Figure 42 Weekly En-route delay per flight Maastricht UAC Summer Sector configurations suggest that there was a slight reduction in sector counts in the weekend compared to week-days in line with the lower traffic levels. Although providing high capacity, MUAC s main issue is the geographical imbalance in demand causing at times extremely high increases in demand in some sectors. Traffic growth was concentrated in certain periods of the day and new demand peaks were observed (e.g. 1000 and 1500 on Saturdays). Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 47

6 AIRPORTS Departures from the airports in the network increased by 3.8% in 2017 (see Airport Traffic Evolution), while airport ATFM delays decreased by 3.6%. Airport capacity and weather contributed to 83.9% of the total airport delays. The integration of airports into the network progressed significantly in 2017. There was progress towards the wider A-CDM implementation in Europe: 4 additional airports fully implemented, giving a total of 26 A-CDM airports, covering 34% of the departures in the NM area (see Airport CDM and ATC Tower Implementation). In 2017, 3 airports connected to NM as Advanced ATC Tower airports, making 19 airports in total, covering 9% of departures in the NM area (see Advanced ATC Tower Implementation). NM now receives Departure Planning Information (DPI) messages for more than 42% of departures in the NM area. Greek airports had an average daily traffic growth of 5.3% in 2017 compared to 2016 for the period between April and October. The summer 2017 was once again challenging due to the long-standing problems at the Greek island airports, i.e. airport layout, passenger terminal capacity and limited numbers of aircraft parking positions. These problems are unlikely to be resolved in 2018 but ongoing developments following the privatisation of the airports may provide sufficient airport infrastructure and technological modernisation in the coming years. To overcome these difficulties, NM implemented the airport function within the NMOC, which provided tactical support on hot-spot airports (See Greek islands Summer). There was very good collaboration from airports on the provision of strategic information to NM via the Airport Corner. The focus this year was to enhance the quality of strategic airports information, instead of expanding the number of contributing airports (see Airport Strategic Information Provision). The Enhanced Information Exchange process in which airports share data with NMOC continued throughout the year. In this process, airports report foreseen capacity impacts caused by weather related or other events during the ATFM pre-tactical phase of operations. As of 2017, airports as well as airlines can subscribe to receive information on events impacting operations (see Pre-Tactical and Tactical Airports Information Exchange). The Airport Unit has started a runway throughput study for Athens Airport and Brussels Airport in 2017. The study for Athens Airport involves the HCAA, Athens International Airport and AEGEAN. The study for Brussels Airport mainly involves Belgocontrol. Both studies will be finalised in 2018. In general, the partnership with airports has further improved. Airports started expressing interest in the implementation of SESAR concepts (e.g. AOP-NOP/APOC, TBS, RECAT-EU) establishing the foundation to achieve future SESAR targets. RECAT-EU Minima Separation are applied at Leipzig-Halle Airport / Munich APP since July 2017 for specific category of aircraft (A300/B757/B767 types). London Heathrow and Vienna airports are planning RECAT-EU deployment for the coming year. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 48

A first AOP-NOP Integration project was started with Paris Charles de Gaulle, Frankfurt and London Heathrow airports to implement extended data exchange via NM B2B services. The NM is leading a second AOP-NOP project under the CEF 2016 call including the following airports: Madrid Barajas, Barcelona, Palma de Mallorca, Amsterdam Schiphol, Brussels and Stockholm/Arlanda. In addition, a third AOP-NOP project proposal for the CEF 2017 has been launched with 7 more airports as partners. The First Rotation Optimisation Trial (FROT) at Zurich contributed to a positive network impact with fewer ATFM arrival delays (aerodrome capacity) and improved airport arrival slot compliance. Swiss Air Lines efforts to keep flight plans up-to-date improved the arrival demand picture at Zurich airport during the trial. Finally, NM continued the close and effective collaboration with airports through a number of bilateral meetings, visits and exchanges. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 49

AMSTERDAM/SCHIPHOL ISTANBUL/SABIHA GOKCEN ISTANBUL-ATATURK LONDON/HEATHROW LONDON/GATWICK BARCELONA/EL PRAT FRANKFURT MAIN ZURICH LISBOA PARIS ORLY PALMA DE MALLORCA WIEN SCHWECHAT ADOLFO SUAREZ MADRID- BARAJAS IRAKLION/NIKOS KAZANTZAKIS BRUSSELS NATIONAL LONDON/STANSTED OSLO/GARDERMOEN TEL AVIV/BEN GURION PARIS CH DE GAULLE PARIS LE BOURGET Avg Airport delay EUROCONTROL 6.1 HOT SPOTS Amsterdam/Schiphol recorded a traffic increase of 3.9% and ATFM delay increased by 66%. Airport ATFM delay per flight also increased by 60%. Airport weather was the main contributor (1,337 min/day, +72%) followed by airport capacity (883 min/day, +34.2%). Weather particularly impacted airport operations at Amsterdam/Schipol in September and December and to a lesser extent in February and January. Traffic decreased by 4.9% in Istanbul/Sabiha Gökcen airport and ATFM delays decreased by 52.3%. Airport capacity was the main contributor for the decrease in delay (1,436 min/day, - 51%) followed by weather (175 min/day, -59%). Istanbul/Ataturk airport traffic remained almost at the same level as in 2016, while ATFM delays decreased by 26. Airport capacity was the main contributor (1,006 min/day, -34%) followed by weather (454 min/day, + 8%). Despite the decrease in ATFM delay, both airports remained the highest contributors to airport capacity related delays in the network. 3500 3000 2500 2000 1500 1000 500 0 2422 1614 1463 1251 1246 771 669 596 461 445 375 357 327 272 260 243 236 233 228 224 AD CAPACITY (ATC) AD STAFFING (ATC) AD DISRUPTION (ATC) AD CAPACITY AD DISRUPTIONS AD EVENTS AD WEATHER PREV YEAR Figure 43: Top 20 airport delay locations during 2017 21 London/Heathrow traffic remained almost at the same level as in 2016. The average daily airport ATFM delay increased by 3% in 2017, of which 95% was due to weather. Adverse weather particularly affecting airport operations in December and October. London Gatwick traffic increased by 2.4% compared to 2016. ATFM delay increased by 33% and delay per flight increased by 30.2% in 2017. Adverse weather and airport capacity were the main delay causes in 2017. Airport weather delay decreased from a daily average of 660 minutes in 2016 to 596 minutes in 2017, particularly impacting airport operations in September and January. Airport capacity delay increased from a daily average of 176 21 Only airports with more than 11,000 movements/year are included Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 50

minutes in 2016 to 591 minutes in 2017, particularly impacting airport operations in July, August and September. Airport capacity regulations were applied mainly due to ground congestion. Barcelona/El Prat traffic increased by 5.2% and delays increased from a daily average of 683 minutes in 2016 to 771 minutes in 2017. Airport capacity due to environmental constraints and adverse weather conditions generated most of these delays. Airport weather delay increased from a daily average of 222 minutes in 2016 to 359 minutes in 2017, particularly impacting airport operations in September (33% of total ATFM weather delay). Frankfurt/Main airport traffic increased by 3% and ATFM delay increased by 21.6%. Adverse weather conditions caused most of the delays impacting airport operations particularly between May and October. Zurich airport traffic remained almost at the same level as in 2016, while ATFM delay and delay per flight both decreased by 26% compared to 2016. Airport capacity, weather and limited availability of the optimum runway configuration due to environmental constraints were the main delay causes. Airport capacity delay decreased from a daily average of 432 minutes in 2016 to 252 minutes in 2017. Lisbon traffic increased by 11.7% while ATFM delay nearly doubled compared to 2016. Airport capacity and airport weather were the main delay causes in 2017. Airport capacity delay increased from a daily average of 70 minutes in 2016 to 234 minutes in 2017. Lisbon airport is a new addition to the top 20 airports delay per flight locations in 2017. Paris/Orly traffic decreased by 2.2% and delay decreased by 34.5% from a daily average of 679 minutes in 2016 to 445 minutes in 2016. The airport undertook major runway reconstructions from July to September generating airport capacity delays. In addition to this, a major tower cab renovation started in November. Taxiway and tower cab works combined generated 70.2% of airport capacity delay. The airport operations were also impacted by several industrial actions in 2017. Palma de Mallorca traffic increased by 5.9% and ATFM delay increased by 15.4%. Airport weather and airport capacity issues were the main delay causes in 2017 at this airport. Airport weather delay increased from a daily average of 106 minutes in 2016 to 206 minutes in 2017, particularly impacting airport operations in September. Airport capacity delays decreased by 17% compared to 2016. Vienna traffic remained at the same level of 2016 while ATFM delay delay increased by 13%. Adverse weather conditions were the main delay cause, impacting airport operations in July and October. Madrid/Barajas traffic increased 2.7% and ATFM delay increased 24.3%. Adverse weather conditions and airport capacity issues were the main delay causes. Airport capacity delay increased from a daily average of 80 minutes in 2016 to 128 minutes in 2017. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 51

MIKONOS ZAKINTHOS/DIONISIOS SOLOMOS SANTORINI ISTANBUL/SABIHA GOKCEN CHANIA/IOANNIS DASKALOGIANNIS IRAKLION/NIKOS KAZANTZAKIS Avg Airport ATFM delay per movement AMSTERDAM/SCHIPHOL LONDON/GATWICK PARIS LE BOURGET ISTANBUL-ATATURK RODOS/DIAGORAS PAFOS/INTL LONDON/HEATHROW CANNES MANDELIEU BARCELONA/EL PRAT LISBOA ZURICH LONDON/CITY PARIS ORLY PALMA DE MALLORCA EUROCONTROL Brussels/Zaventem airport traffic increased by 7.1% and it is back to 2015 levels after recovering from the drop in 2016 caused by the terrorist attack of which happened in March of that year. ATFM delay decreased by 6.1% compared to 2016. Adverse weather conditions were the main delay cause in 2017. London/Stansted traffic increased 5.3% and ATFM delay increased by 3%. Adverse weather conditions were the main delay cause, impacting airport operations in December. Oslo/Gardermoen airport traffic increased by 2.7% compared to 2016, while ATFM delay decreased by 11%. Adverse weather conditions were the main cause of delay (86.7% of total ATFM delay). Tel Aviv/Ben Gurion airport traffic increased by 17.4% compared to 2016, while delay decreased by 36.2%. Aerodrome capacity (ATC) and airport capacity in combination with runway maintenance were the main delay causes. Paris Charles de Gaulle traffic remained at the same level of 2016 while delay decreased by 34%. Adverse weather conditions were the main delay cause. Airport weather delay increased from a daily average of 198 minutes in 2016 to 215 minutes in 2017, impacting airport operations in January and August. Paris Le Bourget is a new addition to the top 20 airport delay locations (both in total delay and delay per flight). Paris Le Bourget airport traffic increased by 2.8% and ATFM delay nearly tripled compared to 2016. ILS and DME maintenance and technical issues were the main delay cause in 2017 (83% of total airport ATFM delay). Geneva and Munich delay decreased compared to 2016 and airports are no longer in the top 20 airport delay locations. 8.0 6.0 4.0 2.0 4.29 2.89 2.87 2.75 2.03 1.84 1.74 1.59 1.52 1.18 1.17 1.03 0.96 0.88 0.87 0.83 0.83 0.79 0.70 0.66 0.0 AIRPORT CAPACITY (ATC) AIRPORT STAFFING (ATC) AIRPORT DISRUPTIONS (ATC) AIRPORT CAPACITY AIRPORT DISRUPTIONS AIRPORT EVENTS AIRPORT WEATHER PREV YEAR Figure 44: Top 20 airport delay per flight locations during 2017 1 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 52

Greek airports had a traffic growth of 4.4% during 2017 (5.3% during the summer - April to October), while ATFM delay remained almost at the same level. Mikonos, Zakynthos and Chania D. Daskalogiannis delays per flight decreased, while Santorini, Iraklion N. Kazantzakis and Rodos delays per flight slightly increased compared to 2016. Mikonos and had the same traffic levels of 2016. Airport capacity was the main delay cause at this airports. See 6.3.1 for more details on the Greek islands summer performance. Cannes/Mandelieu airport delay per flight decreased by 14.6% compared to 2016 mainly due to capacity (ATC) issues. Cannes festival accounted for 11.4% of total airport ATFM delay. London/City traffic decreased by 5.2% and delay per flight decreased from 1.01 minutes in 2016 to 0.79 minutes in 2017 mainly due to adverse weather conditions which impacted airport operations in January, February and March. Airport weather delay per flight decreased from 0.71 minutes in 2016 to 0.59 minutes in 2017. Paphos traffic remained almost at the same level while airport delay per flight increased from 0.61 minutes in 2016 to 1.03 minutes in 2017. Airport capacity issues were the main delay cause. Biggin Hill, Tel Aviv/Ben Gurion, Beauvais/Tille and Thessaloniki airports are no longer in the top 20 delay per flight locations. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 53

6.2 AIRPORT DISRUPTIONS A number of unplanned events or disruptions 22 (Table 5) imposed capacity reductions at certain airports. Technical issues generated delay at Paris/Le Bourget (ILS and DME maintenance and technical issues), Amsterdam/Schiphol (radar problems and ATC system failure), Iraklion/Nikos Kazantzakis (Non-availability of VOR), Catania (radar maintenance) and Pisa (radar problems) airports. Events that also had an impact at en-route level are listed in 5.2.2 En-route Disruptions. Date Location Event 07-Jan London/Heathrow Excessive air holding due to aircraft blocked on runway Traffic Impact ATFM Delay Impact (minutes) None 2,452 01-Feb Amsterdam/Schiphol Radar problems 86 flights 11,406 01-Feb 08-Feb 16-Feb and 23-Feb 23-Feb 28-Feb 01-Mar, 02-Mar and 03-Mar Berlin/Tegel Ground service industrial action None 5,687 Amsterdam/Schiphol Manchester Aircraft incident in combination with strong winds Vehicle fire in public tunnel beneath runways producing smoke and security considerations 8 flights 31,321 13 flights 2,056 Paris/Le Bourget ILS calibration None 2,217 20-Mar Italy ATC Industrial action 420 flights 4,481 13-Mar and 14-Mar Berlin/Tegel and Berlin/Schönefeld Ground handling staff industrial action 1420 flights 2,568 16-Mar to 19-Mar Catania Mount Etna volcanic eruption None 1,525 05-Apr 12-Apr and 19-Apr Catania/Fontanarossa Radar maintenance None 3,517 11-Apr Paris/Charles de Gaulle Radar issues None 1,498 12-Apr Palma de Mallorca and Ibiza Radar issues None LEPA - 6,177 LEIB - 1,253 26-Apr Nantes Bomb alert None 1,001 17-May Greece ATC Industrial action Unknown 3,975 18-May Amsterdam/Schiphol Fuel supply issue None 8,223 24-May 25-May 27- May and 28-May Lisbon ATC software upgrade None 5,386 26-May London/City Technical issues with baggage system None 5,790 31-May Catania/Fontanarossa Radar maintenance None 1,516 07-Jun 14-Jun 21- Jun and 28-Jun Catania/Fontanarossa, Catania/Sigonella and Comiso Radar maintenance None 5,535 22 The main source for the event description is the remark field on the NM ATFM Regulation (ANM). Only events with an impact of more than 1.000minutes of ATFM delay were considered in Table 5. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 54

Date Location Event 15-Jun Brussels Electrical power supply issues affecting approach lighting, navigation aids and terminal building supply Traffic Impact ATFM Delay Impact (minutes) None 2,106 27-Jun Tenerife/South Disabled aircraft on runway None 1,432 28-Jun Amsterdam/Schiphol Unavailability of ILS on RWY 22 None 1,851 28-Jun Paris/Le Bourget ILS calibration None 2,658 From 01-Jul to 12- Jul Bordeaux/Merignac ATC Industrial action None 4,827 05-Jul Manchester Terminal 3 evacuation None 1,339 06-Jul Dusseldorf Disable aircraft on runway None 2,734 25-Jul, 01-Aug, 02- Aug and 04-Aug Rhodes Radar issues None 3,024 04-Aug Bordeaux/Merignac Partial power failure None 1,032 11-Aug and 12Aug Pisa/San Giusto Radar maintenance combined with strong winds None 5,169 12-Aug Rome/Fiumicino Fire in airport vicinity None 1,316 17-Aug London/Heathrow Unavailability on ILS None 3,346 Sep Paris/Le Bourget ILS and DME maintenance and technical issues None 44,145 11-Sep to 13-Sep Paris/Orly French ATC industrial action 500 flights 3, 945 12-Sep and 18-Sep Barcelona ILS calibration None 1,096 26-Sep to 31-Oct Iraklion/Nikos Kazantzakis Non-availability of VOR None 11,261 19-Oct Istanbul/Ataturk ILS check in conjunction with morning capacity None 2,014 23-Oct to 25-Oct Barcelona ILS calibration None 1,165 30-Oct Lisbon ILS calibration None 1,115 Nov - Dec Tunis New radar equipment installation None 15,314 01-Nov Helsinki Unforeseen runway repair works None 1,459 16-Nov Amsterdam/Schiphol Ground radar issues None 1,555 17-Nov Ankara Technical issues with local flight planning system and communications equipment None 1,715 21-Nov Amsterdam/Schiphol ATC system failure Unknown 8,581 16-Nov Paris/Orly French ATC industrial strike Unknown 4,664 15-Dec Milano/Malpensa and Bergamo Italian ATC Industrial Action Table 5 Airport Disruptions 2017 Unknown LIME 2,302 LIMC 1,911 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 55

6.3 NETWORK OPERATIONS SUPPORT 6.3.1 GREEK ISLANDS SUMMER Summer traffic to the Greek islands summer destinations increased by 5.3% in 2017. The majority of the smaller airports were once again operating at the limit of their declared capacity during periods of peak demand. Arrival delays over the period increased by 0.9%, with an increase from 435,617 minutes to 439,520 minutes in 2017. The overall performance has improved since 2012 when the joint NM / Hellenic Civil Aviation Authority (HCAA) Action Plan was put in place. Summer 2017 was extremely challenging. The problems occurred related to airport layout, passenger terminal capacity and the limited number of aircraft parking positions. However, the insufficient ATCOs availability and the lack of radar facilities that, in consequence, required the application of procedural approach at certain destinations, were the main contributing factors for delays, in addition to the lack of airport infrastructure. These long standing problems are unlikely to be resolved in 2018 but ongoing developments following the privatisation of the airports may provide sufficient airport infrastructure and technological modernisation in the coming years. In 2017 the HCAA published a NOTAM requiring GA/BA traffic to request airport slots. This had a positive effect of better control. At the same time the HCAA requested NM/HSCA to monitor GA/BA flights to ensure they had a valid airport slot and/or were operating within the limits stated on the NOTAM. Any flight outside of these rules had a FLS issued. 25 flights had an FLS sent from a total of 5000+ GA/BA flights, which shows the procedure had an effect compared to previous years. The preparation for the summer season was done in February 2017 with close collaboration between NM, airlines, HCAA, Hellenic Slot Coordination Authority (HSCA) and Hellenic Air Navigation Service Provider (HANSP) to highlight the busiest days and busiest airports. NM re-iterated the importance of fair play by adhering to the allocated airport slot. NM provided an Airport Function within the NMOC to provide tactical support on hot-spot airports in Greece. This proved very beneficial for all operational stakeholders with tangible results in delay reduction at those airports. Close on the day cooperation with HANSP, HSCA and aircraft operators (AOs) allowed for better utilisation of capacity. NMOC requested on a daily basis 20+ flights to try and file their flight-plans with a faster speed to arrive at a given time at destination, this also allowed for better utilisation of capacity. Close cooperation with GA/BA handling companies in Greece allowed extra capacity to be filled at those airports where most GA/BA flights are only drop off/pick up and usually on the ground for approximately 15min. Therefore, NM is planning to continue with such a function in the NMOC from 2018 onwards. In the future the airport function will also support new concepts like APOC, disruption and contingency situations and crisis management. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 56

6.4 AIRPORT CDM IMPLEMENTATION AND ADVANCED ATC TOWER IMPLEMENTATION During the course of 2017, 4 airports fully implemented A-CDM: Palma de Mallorca on 4 May, Stockholm Arlanda on 8 June, Hamburg on 16 August and Lyon on 14 November. This brings the total number of fully implemented airports to 26, covering 34% of departures in the NM area. More and more airports are implementing A-CDM bringing benefits not only to themselves but also to neighbouring ACCs, thanks to increased predictability. An independent study, conducted on behalf of EUROCONTROL, assessing both the local and network impact of A- CDM implementation was completed in 2016. The study can be downloaded from the EUROCONTROL A-CDM website vii. Airports that have no plans to implement the A-CDM process but still wish to integrate with the ATM network may do so as an Advanced ATC Tower airport. A number of airports are also considering this option as a first step towards full A-CDM implementation. Such airports provide a reduced set of DPI messages with a reduced set of advantages (compared to A- CDM airports). An Advanced ATC Tower airport provides Target Take-Off-Time (TTOT) estimations as well as Variable Taxi-Times (VTTs) and SIDs in use to the NMOC. These are provided from the moment that the aircraft leaves the blocks. In 2017, 3 airports connected to the Network as Advanced ATC Tower airports. These airports are Tenerife South, Malaga and Las Palmas. This brings the total number of Advanced ATC Tower airports to 19, representing 9% of departures in the NM area. The 26 A-CDM airports together with the 19 Advanced ATC Tower airports means that NM now receives Departure Planning Information (DPI) messages for almost 43% of departures in the NM area. Information on individual airports which implemented A-CDM and Advanced ATC Tower in 2017 can be found in Annex III. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 57

6.5 INFORMATION EXCHANGE BETWEEN AIRPORTS AND NM AIRPORT CORNER PROCESS As defined under the Network Manager Functions Implementing Regulation (677/2011), NM has a task to help airports take advantage of the 'network approach' to solve operational issues and enhance performance. In order to enable airports provision of information in an efficient and harmonized manner, NM has implemented the Airport Corner 10 years ago. It is an airport focused data repository supported by a secure web application. The underlying process for data provision facilitates collaboration between the local ANSP and the Airport Operator resulting in a coordinated airport view. Currently 102 European airports are participating in this process and another 22 will be invited to join the process by end 2019. Since 2015 the Enhanced Information Exchange (EIE) reporting process has been established via the pre-tactical/tactical section of the Airport Corner which allows airports to report on events in a consistent manner in the ATFM pre-tactical phase. In 2017, 35 airports had reported 134 events via the Airport Corner. As of 2017, airports as well as airlines can subscribe to receive information on events impacting operations. The process of providing detailed post-ops feedback for airports on the events, which had significant Network impact, was enhanced in 2017 by sharing it with 16 airlines. The process of reporting standard and temporary diversion capabilities via the Airport Corner was established with 48 airports in 2017. Table 6 presents a list of airport planned events that were reported via the Airport Corner and had an ATFM impact. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 58

ICAO Code GCLP Airport Name Event Name 23 Start Date End Date Gran Canaria DACT Military exercise affecting Gran Canaria airport ATFM Delay in 2017 (minutes) 16-01-2017 26-01-2017 1,920 LEBL Barcelona Runway 07L/25R pavement regeneration 20-11-2016 05-02-2017 6,551 LTBA Istanbul/Ataturk Runway 05/23 and taxiways maintenance 08-02-2017 09-02-2017 1,528 LGAV Athens Runway 03R/21L and taxiway D heavy maintenance works 15-02-2017 18-02-2017 1,018 LIRF Rome/Fiumicino Runway 16R/34L maintenance works 20-01-2017 03-03-2017 28,884 EDDL Dusseldorf WWII bomb disposal near airfield 09-03-2017 09-03-2017 2,409 LPPT Lisbon Pope visit 12-05-2017 13-05-2017 1,898 EBBR Brussels NATO Summit and Fly-by 16-05-2017 24-05-2017 1,410 EDDF Frankfurt Necessary refurbishment works at RWY 18 25-05-2017 26-05-2017 2,470 EHAM Amsterdam/Schi phol Heavy maintenance of runway 06/24, some delays in conjunction with weather 18-03-2017 31-05-2017 26,497 LPPR Porto Military Ceremony 10-06-2017 23-06-2017 1,369 LFPB Paris/Le Bourget Paris Air Show 13-06-2017 25-06-2017 6,094 EDDH Hamburg Yearly closure of RWY 15/33 15-06-2017 28-06-2017 4,632 EDDL Dusseldorf WWII bomb disposal near airfield 11-07-2017 11-07-2017 1,258 LFPO Paris/Orly Bastille day Fly-By 14-07-2017 14-07-2017 5,817 EDDT Berlin/Tegel World War II bomb disposal 29-08-2017 30-08-2017 3,071 LFPO Paris/Orly Second phase of complete renovation of runway 06/24 25-07-2017 31-08-2017 14,769 EDDH Hamburg Yearly closure of RWY 15/34 31-08-2017 13-09-2017 1,305 EGKK London/Gatwick Runway maintenance 13-09-2017 04-10-2017 3,056 GCLP Gran Canaria Regeneration of pavement of runway 03R/21L and taxiways S5 and S6 08-05-2017 15-10-2017 2,389 LROP Bucharest Runway 08R/26L resurface 17-07-2017 22-12-2017 9,251 LGTS Thessaloniki Rehabilitation work of runway 10/28 01-01-2017 31-12-2017 10,329 GCRR Lanzarote Apron works 21-11-2017 31-12-2017 2,413 EGCC EDDK LFPO Manchester Cologne Paris/Orly Occasional nightly closures of RWY 05L/23R due to essential maintenance Reconstruction of runway 06/24 and new turning pad at taxiway B Taxiway maintenance in conjunction with tower renovation Table 6 Airport reported planned events 2017 02-01-2017 05-01-2018 4,052 17-10-2016 03-18 3,054 12-10-2017 31-07-2018 59,65 23 Only events with a minimum of 1.000 minutes of ATFM delay generated are listed Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 59

7 FLIGHT EFFICIENCY This chapter provides a summary of the progress made on the implementation of the actions agreed in the joint IATA/CANSO/EUROCONTROL Flight Efficiency Plan viii, drawn up in 2008, and responds to the requirements of the SES performance scheme. The NM flight efficiency targets and objectives for 2017 included in the Network Performance Plan (NPP) 2015-2019 ix and in the Network Operations Plan (NOP) 2017-2019/21 are listed below: Route extension airspace design (DES) Target: achieve an improvement of the DES indicator by 0.57 percentage points between 2012 and 2019 Route extension last filed flight plan (KEP) Target: achieve a KEP target of 4.44% for the SES area and 4.17% for the NM area Route extension actual trajectory (KEA) Target: achieve a KEA target of 2.78% for both SES and NM areas Increase the CDR1/2 usage (CDR-RAI and CDR-RAU) NM Objective: increase the CDR availability (CDR-RAI) and CDR usage (CDR-RAU) by 5% between 2015 and 2019 Flight efficiency indicators are monitored for pure airspace design and for flight planning. The downward trend evolution of those indicators since the beginning of 2011 is shown on Figure 45. The evolution recorded on the route extension based on the last filed flight plan during 2017 was positively impacted by the implementation of Free Route Airspace. Nevertheless, negative impacts were still recorded as a result of the industrial actions, social issues that led to reduced capacities and re-routings to avoid capacity constrained and avoided/closed areas due to crisis situation. This evolution continues to demonstrate the necessity to provide sufficient capacity constantly to further improve the flight planning indicator and to reduce the gap with the airspace design indicator. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 60

Figure 45: Route efficiency KPI per AIRAC cycle A number of events in 2017 affected the network and had direct consequences on the flight efficiency evolution: Overall crisis situation in Ukraine that lead a significant number of flights to avoid the entire Ukrainian airspace moving to neighbouring countries (Turkey, Bulgaria, Romania, Poland, Slovakia, etc.); as a result of the Ukrainian crisis adjacent ACCs/ UACs were onloaded by Far Eastern traffic avoiding the Ukraine airspace leading to increased route extensions. Closure of Libyan airspace for over flights due to the security situation required procedures with impact on flight efficiency for traffic between Europe and Africa re-routed via Egypt and Tunisia. Avoidance of Syrian and Iraqi airspace due to the security situation with impact on flight efficiency for traffic between Europe and Middle East and Asia re-routed via Iran and Turkey with additional impacts on the flows from the Ukrainian crisis situation. Several French ATC industrial actions in March, September, October and November required regulations in French ACCs and protective measures in neighbouring ACCs, with impact on flight planning route extension Widespread capacity and staffing issues across the network required a high number of regulations and/or level-cap scenarios at: Karlsruhe UAC, Maastricht UAC, Brest ACC, Marseille ACC, Reims ACC, Nicosia ACC, Barcelona ACC, Bordeaux ACC. Temporary capacity reductions due to implementation/upgrade of ATM system or major airspace changes at: Maastricht UAC, Karlsruhe UAC, Bordeaux ACC required regulations, with impact on flight planning route extension. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 61

Savings Mileage flown EUROCONTROL 7.1 AIRSPACE DESIGN As part of the Flight Efficiency Plan, intensive work was undertaken by States and ANSPs in close cooperation with NM to develop and implement enhanced airspace design solutions, with over 200 airspace improvement packages being developed and implemented in the 12 months prior to summer 2017. As a result, the route extension due to airspace design (RTE- DES) continued its downward trend throughout the year, reaching its lowest level ever in December 2017 at 2.31%. Figure 46: yearly evolution of airspace design indicator The average route extension due to airspace design, RTE-DES (Figure 46) decreased from 2.47% in 2016 to 2.36% in 2017, enabling an average potential daily saving of nearly 7000 nautical miles. Potential Yearly NM Savings and Mileage Flown Airspace Design Savings (NM) 8000000 7000000 6000000 4800000 5000000 4000000 4000000 3000000 3000000 2000000 1000000 0 6900000 4,7E+09 4,6E+09 4,5E+09 4856000 4,4E+09 4,3E+09 3380000 3883000 3317371 4,2E+09 2555000 4,1E+09 1700000 4E+09 3,9E+09 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Year 3,8E+09 3,7E+09 Figure 47: Potential yearly savings/ losses in nautical miles (NM) due to airspace design Over the reporting year, this represents a potential saving of 2.55 million nautical miles (Figure 47), approximately 16 kilotons of fuel, reduced emissions of 52 kilotons of CO2, or 13 million Euros when compared to 2016. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 62

Savings (Nautical Miles) Mileage Flown (Nautical Miles) EUROCONTROL 7.2 AIRSPACE CHANGES VS. FLIGHT PLANNING The flight planning indicator (KEP) measures the length of the flight planned trajectory compared to the great circle (route extension). It reflects inefficiencies in the use of the airspace (due to RAD restrictions, CDR availability, inefficient flight-planning etc.), but also user preferences for cheaper rather than shorter routes. 5.00% 4.91% 4.87% 4.86% 4.70% 4.74% 4.82% 4.62% 4.50% 4.00% 3.50% 3.00% 2.50% 2.00% 2011 2012 2013 2014 2015 2016 2017 The average route extension based on the latest filed flight plan (KEP) decreased from 4.82% in 2016, to 4.62% in 2017 (Figure 48) for NM area. This is above the KEP 2017 targets of 4.44% for SES area and 4.17 % for NM area. Figure 48: Yearly evolution of flight-planning indicator (KEP) Figure 49 shows the corresponding yearly savings / losses and the relationship with the mileage flown over the Second Reference Period (RP2) of the SES Performance scheme. 8,000,000 6,000,000 4,000,000 2,000,000 0-2,000,000-4,000,000-6,000,000-8,000,000-2,396,182-4,418,600 5,141,498 2015 2016 2017 Savings Mileage Flown 5.3E+09 5.1E+09 4.9E+09 4.7E+09 4.5E+09 4.3E+09 4.1E+09 Figure 49: Yearly savings/ losses in nautical miles (NM) due to improved flight planning efficiency The average flight-planned distance decreased when compared to 2016, resulting in 5.1 million nautical miles less flown over the whole year. This means an average daily decrease of more than 14000 nautical miles. Over the year, this represents savings of approximately 32 kilotons of fuel, reduced emissions of 104 kilotons of CO2, or 26 million Euros when compared to 2016. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 63

The trend also reflects the combined effect of: industrial actions, special events (e.g. Ukraine crisis situation, Libyan airspace closure, etc.) and technical problems on the network and adverse weather. Despite all those factors, the trend is positive and the airspace improvements made helped to contain all the adverse effects. This situation proves the good work done in improving flight-planning options for the operators but emphasises again that more efforts must be made to improve the efficiency of the airspace utilisation and to constantly provide sufficient capacity thus ensuring that the indicator based on the latest filed flight plan follows a similar trend to the airspace design indicator. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 64

Savings (Nautical Miles) Mileage Flown (Nautical Miles) EUROCONTROL 7.3 ACTUAL TRAJECTORY 3.40% 3.20% 3.00% 2.80% 2.60% 2.40% 2.20% 2.00% 3.31% 3.20% 3.14% 2.72% 2.77% 2.93% 2.77% 2011 2012 2013 2014 2015 2016 2017 Figure 50: Yearly evolution of the actual trajectory indicator (KEA) The actual trajectory indicator (KEA) decreased to 2.77%, better than the 2017 target of 2.78% for both SES and NM areas. The continuous expansion of the Free-Route Airspace (FRA) (including continuous expansion of cross-border FRA) is a major factor in the positive evolution of the environment indicators (KEP and KEA). Figure 51 shows the corresponding yearly savings / losses and the relationship with the mileage flown from the start of the Second Reference Period (RP2) of the SES Performance scheme. The average actual distance decreased when compared to 2016, resulting in 3.7 million nautical miles less flown over the whole year. This means an average daily decrease of more than 10000 nautical miles. Over the year, this represents savings of approximately 22 kilotons of fuel, reduced emissions of 72 kilotons of CO2, or 18 million Euros when compared to 2016. 9,000,000 7,000,000 5,000,000 3,000,000 1,000,000-1,000,000-3,000,000-5,000,000-7,000,000-9,000,000 3,691,680 2015 2016 2017-6,378,669-8,630,438 Savings Mileage Flown 4.6E+09 4.4E+09 4.2E+09 4.0E+09 3.8E+09 3.6E+09 3.4E+09 3.2E+09 3.0E+09 2.8E+09 Figure 51 - Yearly savings/ losses in nautical miles (NM) due to improved actual trajectory efficiency Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 65

1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 percentage (%) percentage (%) 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 percentage (%) EUROCONTROL 7.4 CONDITIONAL ROUTES (CDR) CDR availability is an important element when considering the ASM in the Network Operations context. The charts below shows little changes in absolute figures for the evolution of CDR development as elements of the network in 2017 compared to 2016. This is due mainly to changes in CDR categories with many CDR1/2 to permit night routes opened and to the continuous network improvement process (covered by ERNIP). 4000 CDR2 CDR1/2 CDR1 CDR number of segments 2017 100 RoCA: AIRAC values 2017 3000 75 2000 50 CDR1 CDR2 CDR_all 1000 25 0 0 2013 2014 2015 2016 2017 Figure 52: Evolution of CDR availability AIRAC cycle Figure 53: Rate of CDR availability (RoCA) in 2017 RoCA for CDR1 and CDR1/2 categories has a stable high value (98%) over the entire year while RoCA for CDR2 is oscillating between 77% and 86% with an average of 79.6%. 100 RAI (%) 2017 per AIRAC cycle 100 RAU (%) 2017 per AIRAC cycle 90 90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Figure 54: RAI (%) 2017 per AIRAC cycle. The Rate of Aircraft Interested (RAI) that planned the available CDR is relatively constant at a value of approx. 74% for the entire year 2017. Figure 55: RAU (%) 2017 per AIRAC cycle The Rate of Aircraft actually Using (RAU) CDR is lower (26%). This is the result of both the ATC intervention for various reasons (expedite traffic, weather, etc.) as well as due to the expansion of FRA implementation in ECAC, making many CDRs no longer a better solution for flying Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 66

100 RAI (%) 2013-2017 100 RAU (%) 2013-2017 90 RAI 90 RAU 80 80 70 70 60 60 50 50 40 30 20 10 0 2013 2014 2015 2016 2017 Figure 56: Five year RAI evolution 40 30 20 10 0 2013 2014 2015 2016 2017 Figure 57: Five year RAU evolution The charts of RAI and RAU evolution over the past 5 years in Figure 56 and Figure 57 indicate the tendency to use less and less the CDR, since there are today better options in FRA or the DCTs. Therefore the NM objectives of increasing the availability (RAI) and use (RAU) of CDRs by 5% between 2015-2019 became obsolete. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 67

CDR availability vs. usage in 2017 1701 6000 1713 1702 4000 1712 2000 1703 CDR available CDR planned CDR used Figure 58 shows the number of CDR available for flight planning (blue line), the number that were actually flight planned (green line) and the number that were actually flown (red line). In Fig. 19 the green and red lines overlap. Approx. 15% of available CDR were used in 2017. 1711 1710 1709 1708 0 1707 1706 1704 1705 Figure 58: CDR availability vs. usage in 2017 The numbers indicating the CDR used and planned versus the CDR available show in 2017 an almost constant difference with a huge gap between availability and utilisation, a similar trend to 2016. The explanation is that in 2017 the FRA airspace in ECAC was extended significantly adding Austria, Slovenia, FYROM, Serbia, NEFAB to the already existing FRA regions. As a result the route network and implicitly the CDRs in these areas have no more relevance. 1711 PFE: 2017 NM savings per flight (AIRAC) 1712 1713 1701 40 30 20 10 0 1702 1703 1704 PFE nm/flt 1711 PFE: 2017 Minutes savings per flight (AIRAC) 1712 1713 1701 5 4 3 2 1 0 1702 1703 PFE minutes/flt 1704 1710 1705 1710 1705 1709 1706 1709 1706 1708 1707 1708 1707 Figure 59: PFE: 2017 Monthly Distance savings (nautical miles per flight) Figure 60: PFE: 2017 Monthly time savings (minutes per flight) The savings per flight in distance and in time due to CDR are strongly dependent on the network opportunities offered by the CDR but in reality the actual traffic is not always able to follow the planned trajectory that would maximize the efficiency due to various causes outside the flight planning process. With the current advances in airspace configurations Free Route Airspace and Direct routes implemented in more ECAC regions the CDRs lost their weight in improving routeing solutions. Potential Flight Economy (PFE) can be realised when using the available CDRs for planning. This is influenced mainly by the CDR availability rate (RoCA) and the awareness/ability/ willingness of the Aircraft Operators to consider the available CDRs in their FPL solutions. The indicator shows how far the real planned trajectories are from the optimum ones. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 68

1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 tons fuel /CO2 EUROCONTROL Concerning the actual traffic, the PFE is calculated with the actual flown CDRs from those available. The values may differ from the planned ones for a number of reasons (ATC intervention for direct/rerouting, delayed departure miss the CDR uptake and forcing to alter the initial FPL, weather, etc.). When making the comparison and the values are smaller, it can also signify that less potential economy is obtained when the initial trajectories are closer to optimal. The diagrams below depict the aggregated values calculated for all CDR types (CDR1, CDR1/2, CDR2) averaged by month: 50 40 PFE (NM) 2017 vs. 2016 for planned traffic PFE 2016 PFE 2017 50 40 PFE (NM) 2017 vs. 2016 for actual traffic PFE 2016 PFE 2017 30 30 20 20 10 10 0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 1 2 3 4 5 6 7 8 9 10 11 12 13 AIRAC of the year Figure 61: PFE 2017 vs. 2016 for planned traffic Figure 62: PFE 2017 vs. 2016 for actual traffic Comparing the Potential Flight Economy (PFE) year on year 2017 with 2016 one can see that the evolution in 2017 has very little variation (22-24 nm/flight) over the year due to low values of CDR used vs. CDRs available for which potential economy is calculated. The actual gain in 2017 is following in general the planned trend with similar evolution as the planned traffic. PFE: Fuel economy and CO2 emissions AIRAC 2017 100000 10000 1000 100 10 1 CO2 PFE Fuel PFE CO2 actual Fuel actual Figure 63: PFE: 2017 Fuel economy and CO2 emissions The environmental indicators of PFE translated in fuel savings and reduced CO2 emissions illustrated in the picture on the left have been calculated using the ICAO methodology for fuel burned and CO2 emissions. The curves show the effect of less CDR usage both for planning and actual flying for the causes mentioned above Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 69

7.5 FREE ROUTE OPERATIONS By the end of 2017, 51 ACCs have either fully or partially implemented Free Route Airspace operations. Full Free Route Airspace implementation Full Night Free Route Airspace implementation DCT based implementation Lisbon ACC Oslo, Stavanger, Bodo, Tampere, Tallinn, Riga,Copenhagen, Malmo and Stockholm ACCs as part of NEFRA Shannon ACC Vilnius ACC Budapest ACC Vienna and Ljubljana ACCs as part of SAXFRA Rome, Padua, Brindisi and Milan ACCs as part of FRA IT Malta ACC Belgrade ACC Zagreb ACC Skopje ACC Tirana ACC Tibilisi ACC Chisinau ACC Sofia and Bucharest ACCs as part of Danube FAB Maastricht UAC Yerevan ACC Kyiv, Lviv, Odesa, Dnipropetrovsk ACCs Reims, Brest, Bordeaux, Paris, Marseille ACCs Madrid ACC London and Prestwick ACCs Karlsruhe UAC Geneva and Zurich ACCs Athens and Macedonia ACCs Bratislava ACC Prague ACC Warsaw ACC Nicosia ACC Table 7 Free Route Airspace Operations Implementations At present, FRA has been successfully implemented across much of northern Europe, southeast, central southeast Europe, with Portugal the first country to introduce full FRA in 2009. The North Europe Free Route Airspace (NEFRA) programme completed its crossborder network that allows airspace users to file preferred trajectories irrespective of borders over Denmark, Estonia, Finland, Latvia, Norway, and Sweden in May 2017. Austria and Slovenia launched their cross-border initiative, called SAXFRA, in 2016, and plan to merge with the South East Europe (SEAFRA) covering Croatia, Bosnia-Herzegovina, Serbia, and Montenegro in the near future. Bulgaria, Romania, and Hungary have implemented crossborder free route airspace and during the next two years many more neighbouring states are due to join including Czech Republic, Slovakia, Albania, and The former Yugoslav Republic of Macedonia. Greece, Cyprus, and Turkey also plan to follow early next decade. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 70

The following map shows the European Free Route Airspace deployment status as of end 2017. Figure 64: Map Free Route Airspace Deployment by end 2017 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 71

7.6 ROUTE AVAILABILITY DOCUMENT (RAD) The Route Availability Document (RAD) is a tool that addresses how the European network airspace may be used. According to the Commission Regulation (EU) No 255/2010 x the scope of the RAD is to be a common reference document containing the policies, procedures and description for route and traffic orientation. The Network Manager Implementing Rule (Commission Regulation (EU) No 677/2011) xi makes a clear reference that the European Route Network Improvement Plan shall include route network and free route airspace utilisation rules and availability. The airspace design and airspace utilisation aspects were brought closer by the established multi-disciplinary Network Manager RAD Team guided by the Operational Stakeholders RAD Management Group. The actions performed by the NM RAD Team have facilitated a pragmatic refinement of the RAD during 2017, with full cooperation of Operational Stakeholders, aiming to overcome weaknesses in airspace design and ATM system functionality and to ensure application of the remaining restrictions only where and when required. The major RAD evolutions and developments in 2017 focusing particularly at Network level and covering the entire NM area of responsibility were as follows: Removal of Indention used as RAD Utilization definition; Adaptation of the time expression and harmonisation in entire RAD; Adaptations of Pan-Europe Annex and simplification of existence of two or more restrictions for same RAD reference ; Adaptation and simplification of Appendix 3 flow condition/s from Column FL Capping moved into Column City Pair ; Adaptation and change of restrictions identification rules in RAD; Appendix 6 - Alignment with CACD database; Appendix 7 - FUA Restriction alignment with CACD database; Appendix 7 - Management of Complex FUA restrictions; Appendix 7 - FUA restrictions utilization time buffer; Improvements in Last minutes RAD changes and Daily use of Increment File; Improvements in RAD Annex for Special Events; Improvements in data structure and format, and change management based on RAD - AURA@n-CONECT grammar; NM Release development related to Airspace Utilisation Rules and Availability (AURA) interactive process via the NOP and use of the NOP Portal as a collaborative platform to build the RAD - AURA[iii]@n-CONECT; Improvement of Dependent Applicability Function (RAD) in NMOC Systems; Further developments of the NM DCT / CDR mapping tool; Publication of B2B PTRs xls file; Monitoring and continuation of publication of harmonised text in regard to promulgation of RAD via the State AIPs; Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 72

Adaptations of RAD Terms - Definitions Document; Contributions and approval of FRA NMOC Application Guidelines - RAD related Chapter; Run of a Network impact assessment of the RAD restrictions implemented in the States and contributions to production of RAD Network Impact assessment study Document. The other RAD evolutions and developments in 2017 included the following aspects (not exhaustive): Further development of the RAD DCT Chart; Continuation of improvements in data structure and format, and change management based on RAD - AURA@n-CONECT grammar; Further adaptations in all RAD Harmonization Rules based on RAD - AURA@n- CONECT developments; Continuation of harmonisation of terminology and definitions; Continuation of improvements in RAD availability (publication) to users; Continuation of rationalisation of restrictions expression; Continuation of the pdf RAD publication. Further RAD and supporting improvement measures have been proposed for implementation in 2018 such as: Gradual improvement in RAD Utilization definition, adaptation of the expressions in the RAD and harmonisation in entire RAD; Further improvement and fine-tuning of a Network impact assessment of the RAD restrictions implemented in the States. Further NM Release developments related to Airspace Utilisation Rules and Availability (AURA) interactive process via the NOP and use of the NOP Portal as a collaborative platform to build the RAD - AURA@n-CONECT; Implementation in real operations of RAD via AURA@n-CONECT; Continuation of publication of EU / EURO restrictions; Continuation of publication of FRA DCT restrictions; Continuation of publication of B2B PTRs. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 73

8 NETWORK MANAGER In addition to the network targets defined for 2017, the NM Performance Plan defines a set of internal NM performance objectives/targets, to measure NM s contribution to the ATM network performance. In the Capacity performance area NM has the target to reduce the en-route ATFM delays by 10%. NM Operations Centre (NMOC) looks for opportunities to reduce the delays by means of proposing alternative routes (RRPs) to the airlines, manually optimising the calculated time over (CTO) or take-off times (CTOT) (these are the direct delay reduction actions). The manual CTOT changes are performed in conjunction with the FMPs/AOs and are therefore regarded as confirmed delay reductions. Re-route proposals can only deliver delay benefit if the AO accepts the proposal - this is monitored in post-ops. These techniques reduce delays at individual flight level and deliver further delay reductions at network level through the CASA optimisation algorithm (indirect snowball effect). While it is currently possible to measure the direct delay reductions initiated by NMOC, it is not possible to quantify the indirect delay reduction effect of the direct actions. The amount of delay reduced by NMOC pre-tactical planning process and the applied scenarios cannot be quantified either. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 74

8.1 CAPACITY (DELAY REDUCTIONS) NM s efforts to reduce delays increased in 2017 in proportion to the overall en-route delay increase. Delay savings were calculated conservatively, only taking into account accepted Re-routings Proposals (RRPs) and NMOC direct action (i.e. Force CTO/CTOT and Override Slot). In 2017, NMOC saved 2.1 million minutes of ATFM delay, 71% of all savings were on Enroute and 29% on Airport delays. 1.00 0.80 0.60 0.40 0.20 0.00 En-route delay savings (min/flight) Objective 10%, Achieved 14% 0.14 0.88 NM Delays savings 2017 En-route delay 2017 En-route savings exceeded 1.5 million minutes from direct actions in NMOC (1.3 million min), and RRPs proposed and followed by airlines (200,000 min), equivalent to 0.14 minutes per flight without this, the delay in 2017 would have been 1.03 minutes per flight. This equates to 14% of the annual network delay, meeting the 10% target defined in the NPP. Figure 65 NMOC Delay Savings 2017 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 75

8.2 ENVIRONMENT (FLIGHT EFFICIENCY) As part of Flight Efficiency Plan, intensive work has been undertaken by States and ANSPs in close cooperation with NM to develop and implement enhanced airspace design solutions, with over 200 airspace improvement packages being co-ordinated at network level and implemented during the 12 months of 2017. These improvement measures reduced significantly the actual losses recorded as a result of number of events (see 3.5) which had direct consequences on the flight efficiency evolution. The list below provides an overview of the major enhancements implemented in 2017. Albania - FRALB - Free Route Airspace Albania H24 above FL195 Armenia - Night Free Route Airspace above FL285 Bulgaria/ Hungary/ Romania - SEEN FRA, Phase 1a: Cross-border NIGHT Free Route Airspace within DANUBE FAB area and HUFRA area - SEEN FRA, Phase 1b. Cyprus - Nicosia Direct Route Airspace - Phase 2A. Denmark / Estonia / Finland / Latvia / Norway / Sweden - NEFAB and DK/SE FAB Free Route Airspace - NEFRA Phase 2. France - Grand Ouest - New TSA 6 Brest ACC. - PONENT sectors re-organisation Barcelona ACC. Georgia - Free Route Airspace Georgia - FRAG Phase 1. Germany - CDO Munich - EDDM airport - Step 1. - IMPROVE2+ Langen FIR - Phase 1. - Free Route Airspace preparation measures. Greece - PBN SID/STAR and approach procedures implementation Athinai FIR. - Santorini/ LGSR Instrument Procedures. Ireland - Borealis FRA Shannon CTA - Step 1. Maastricht UAC - Brussels East High sector 3rd layer. - Night Free Route Airspace Moldova - H24 Free Route Airspace above FL095 Norway - Bodø Oceanic FIR Free Route Airspace. - Airspace improvement western Norway FIR. Serbia - Revision of Military areas Beograd FIR. Switzerland - Additional sector layer Zurich / Geneva ACCs. (not in ERNIP P2) - Comprehensive Night DCT Network across Geneva ACC and Zurich ACC AoR. Turkey - Route network changes Istanbul FIR Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 76

Turkey / Iran - Airspace re-organisation at the Ankara FIR - Teheran FIR interface. - Airspace re-organisation at the Ankara FIR - Baghdad FIR interface. UK - Airspace Re-structuring United Kingdom. Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 77

9 ATFM COMPLIANCE 9.1 ATFM DEPARTURE SLOTS The overall percentage of traffic departing within their Slot Tolerance Window (STW) was 90.7% in 2017, meeting the target of 80%. However, many airports did not meet the target (see ATFM Compliance - ATFM Departure Slot Monitoring Report xii ). It is an improvement over 2016 where the compliance percentage was 89.4%. NM is working with the ANSPs for improving the level of adherence. Figure 66: ATFM Departure Slot Monitoring for 2016 and 2017 24 24 Geographical Zone : NM or Adjacent Member States Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 78

9.2 ADHERENCE TO FLIGHT PLAN SUSPENSIONS The percentage of flights suspended by FAM (Flight Activation Monitoring) but which were activated by airborne data received whilst the flight was temporarily suspended remained at 0.25% of all departures. Figure 67 shows the top airports where such situations occurred, as well as the percentage of these flights within the total number of flights at that airport. The introduction of Airport CDM has proven to be the most effective measure in bringing down the number of such flights. Figure 67: Top 20 ADEPs - Flight Plans Suspensions for 2016 and 2017 25 25 Geographical Zone : Eurocontrol or EUR28 Member States Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 79

9.3 ATFM EXEMPTIONS The overall European ATFM exempted flights decreased in 2017 to 0.61% of all departures (0.64% in 2016), which is still above the target of 0.6%. There were 21 EUROCONTROL Member States in 2017 that granted exemptions in excess of 0.60% of the State s annual departures (EU Member States will be formally notified). NM will discuss any network considerations with the State and service provider concerned. Figure 68: ATFM Exemptions for State Aircraft Monitoring for 2016 and 2017 26 26 Geographical Zone : Eurocontrol or EUR28 Member States Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 80

9.4 MISSING FLIGHT PLANS Figure 69 presents the evolution of the number of Missing Flight Plans (APL Flights), identifying those flights that entered the European airspace without a flight plan (i.e. no initial flight plan was filed successfully in IFPS) and an ATS Unit filed the Flight Plan. The percentage of such flight plans increased to 0.07% (0.06% in 2016). Figure 69: Missing Flight Plans for 2016 and 2017 27 27 Geographical Zone : ADEP or ADES NM Member States Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 81

9.5 MULTIPLE FLIGHTS NM is using the data from Flight Activation Monitoring to identify possible multiple flight plans by measuring the number of flight plans received for which no subsequent activation or airborne information is received. Figure 70 presents the evolution of numbers and proportion of these flights within the total traffic. The percentage of these flights decreased to 0.23% (0.25% in 2016). NM reviews the causes and the network impact of such cases and contacts the airlines or flight plan originators when necessary. Figure 70: Multiple Flight Plans for 2016 and 2017 Edition Validity Date: 20/04/2018 Edition: 1.0 Status: Proposed Issue 82