Working Paper for Optimizing Core TEN-T by Including Non-EU Countries

ACCESSIBILITY IMPROVED AT BORDER CROSSING FOR THE INTEGRATION OF SOUTH EAST EUROPE - ACROSSEE

WP3 INSTITUTIONAL PLATFORM AND ADMINISTRATIVE COOPERATION Report Action 3.1.7 Working paper for the Optimization of the proposed Comprehensive and Core TEN-T

Date: 11.03.2015 Version: [to be set as final by LP]

Involved ACROSSEE partners:

FEDERAL MINISTRY FOR TRANSPORT INNOVATION AND TECHNOLOGY , AT Author: Dr. Helmut Adelsberger (contractor of bmvit)

CENTRAL EUROPEAN INITIATIVE EXECUTIVE SECRETARIAT, IT

ARISTOTLE UNIVERSITY OF THESSALONIKI (AUTH) – RESEARCH COMMITTEE – SPECIAL ACCOUNT FOR RESEARCH FUNDS OF AUTH, GR

University "POLITEHNICA" of Bucharest, RO

Institute of Traffic and Transport Ljubljana, SI

VENICE INTERNATIONAL UNIVERSITY, IT

The information in this document is provided as is, and no guarantee or warranty is given that the information is fit for any particular purpose. The user thereof uses the information at its own risk and is solely liable for its usage.

Table of Contents:

1 EXECUTIVE SUMMARY ...... 5 2 INTRODUCTION: ...... 5 2.1 The Framework of Acrossee ...... 5 2.2 Political Framework ...... 6 2.3 Approach of this document ...... 6 3 THE PLANNING METHODOLOGY FOR THE TRANS-EUROPEAN TRANSPORT NETWORK (TEN-T) ...... 8 3.1 The Comprehensive Network ...... 8 3.2 The Core Network ...... 9 4 THE IMPLEMENTATION OF INFRASTRUCTURE PROJECTS ...... 14 5 EXTENSION OF THE TEN-T INTO THE WESTERN BALKANS ...... 15 6 POSSIBLE CONSEQUENCES FOR EU MEMBER STATES AROUND THE WESTERN BALKANS ...... 25 7 EXTENSION OF THE TEN-T TO MOLDOVA AND WESTERN UKRAINE ...... 27 8 POSSIBLE CORE NETWORK CORRIDORS ...... 32 9 TEN-T AND TRANSPORT DEMAND ...... 35 10 COHERENCE WITH NATIONAL CONSIDERATIONS ...... 37

1 EXECUTIVE SUMMARY

Action 3.1.7 in the framework of ACROSSEE aims at the identification of a possible future TEN-T Core Network for those parts of the ACROSSEE project area which are located beyond EU external borders in South-Eastern Europe (Western Balkans as well as Moldova and parts of Ukraine) and its cross-border optimization in the entire project area. Applying the planning methodology which had been developed for the TEN-T within the EU territory and which is depicted and explained in the paper, results in a proposal of a possible Core Network. In a consistent planning procedure in line with the methodology, urban core nodes and core ports are identified and connected with each other in the entire project area, using the agreed Comprehensive Networks of SEETO as well as of MD and UA as a base. In a few cases this reveals presumptive gaps in these basic networks, which should be further investigated and discussed for possible future adaptations. New node- node relations may even lead to proposals to investigate and adapt parts of the TEN-T Core Network already identified. As a consequence and in line with previous TEN-T planning within the EU, a proposal is developed also for supplementing the existing set of Core Network Corridors by additional branches and by additional Core Network Corridors. This planning procedure is supplemented by some information on possible EU co-financing, depending on the legal status of the individual countries. Within the limits of accuracy, the outcome of these procedures seem to match quite well with the results of a recent TRANSTOOLS calculation, however special attention should be paid to transport infrastructure in agglomeration areas, where large scale traffic models are usually predestined to fail. Within this framework, there is also a reasonable congruence of the results of this study with national planning and priorities, in particular in the context with the proposed Core Network Corridors.

2 INTRODUCTION:

2.1 The Framework of Acrossee The project “ACROSSEE”, meaning “Accessibility improved at border CROSsing for the integration of South Eastern Europe”, has a focus on border crossing transport along the main arteries of the region. It consists of five working packages: 1. Transnational project and financial management, 2. Communication activities, 3. Institutional platform and administrative cooperation, 4. Transport model, 5. Cross border analysis. It is co-funded by the European Union within the Transnational Cooperation Programme for South-East Europe, the project area that covers Albania, Austria, Bulgaria, Croatia, Greece, Hungary, Italy, Romania, Serbia, Slovenia and Ukraine. Under the leadership of the Central European Initiative (CEI), the project comprises partners, associated institutions and observers from the countries of the project area as well as from supranational level.

2.2 Political Framework After the steps of EU enlargement 2004 and 2007, Croatia became an EU Member State in 2013, while currently Serbia and the Former Yugoslav Republic of Macedonia (FYROM) and – beyond the project area – Turkey are candidate countries. At the longer run, one may expect that all or at least most of the countries of so-called “Western Balkans” will be part of the EU, while this is not necessarily the case for Moldova and/or Ukraine; regarding these last two countries, the EU geo-strategically policy objectives are quasi- uncertain, and decision level is at higher level, depending mostly on the EU external conditions. Regardless the current political and legal situation of the individual countries in South-Eastern Europe, in particular in the Danube Region, there is “geographical continuity”: Inland waterways, determined by the most important rivers follow the physical shape of the country. Road and rail transport, cross the borders, as well, but in many cases infrastructure deficiencies (missing links, bottlenecks, poor technical parameters, …), poor services and in particular the borders are serious obstacles to both passenger and freight transport flows. Although it may be an advantage in the sense of environmental impacts of transport that, due to these problems, a great part of transport in the relation Turkey / Middle East – Central and Western Europe has shifted to the maritime detour, there are serious economic disadvantages: Longer transport times entail economic losses in origin and destination regions, interjacent countries are bypassed. With poor accessibility on road and rail, they suffer from low attractiveness as location for economic activities. Transport infrastructure in South-Eastern Europe has many deficiencies as regards configuration, technical parameters and status of maintenance of the network, as well as quality and reliability of operation and services. In general, there is a general tendency in all parts of Europe that countries give priority rather to those border crossing links directed towards more central countries than to more peripheral ones. However, it is evident that wherever delays of border crossing are in the order of one or several hours, improvements of border infrastructure as well as administration and procedures of border control would be the most effective and efficient measures. As, in general, only little investments are needed to achieve high benefits, this should be the short-term solution. Corresponding gains in overall transport times may then be significant enough to be the key to shifting traffic flows, yielding the necessary socio-economic cost-effectiveness of infrastructure investments also in other parts of the network. At the long run, internalising the external costs of transport, combined with tight intermodal connectivity, would be a key factor for modal shift to rail, inland waterways and sea, to boost sustainability in transport.

2.3 Approach of this document This document is based on the work commissioned by the Austrian Federal Ministry of Transport, Innovation and Technology (BMVIT) within ACROSSEE work package 3. It is aiming to show the consequences of applying the rules and tools that have been developed for the TEN-T network in EU-MS to the whole region of the Western Balkan, Ukraine and Moldavia. In a second step the infrastructure needs derived from the considerations according to TEN-T methodology will be compared with the infrastructure needs defined by the capacity analyses of WP4. Finally, there will be given room for the representatives of all countries concerned to express their view on the proposal. The criteria set up in the TEN-T planning methodology shall be applied to analyse the status quo of the rail infrastructure and to identify a potential future TEN-T Core Network relevant for South-Eastern Europe. Finally, options for relevant improvements, including the possibilities of funding or financing such measures, shall be shown.

Repeating analyses of the status of the transport network and identification of projects and investment needs is not within the scope of this investigation. This has been carried out since the late 1990’s in previous studies such as TINA (only covering acceding countries at that time) [1], TIRS [2], REBIS [3] (which besides is currently being updated) or in the annual reports of SEETO. In this context, it is essential to mention the importance of the Pan-European Corridors (PETCs), which were finally, determined 1997 in the Third Pan-European Conference of Ministers of Transport, as the infrastructure backbone for the extension of the EU. Although PETCs lost importance with EU enlargement and faded out during the recent TEN-T Policy Review, they are still an appropriate means of reference in this context, in particular in the context with non-EU Member States. For the current EU Member States of the Region, a TEN-T Comprehensive and, as a subset hereof, a TEN-T Core Network to be implemented by 2030, have been identified in the Regulation (EU) No 1315/2013 [4] of the European Parliament and of the Council of 11 December 2013 on Union guidelines for the development of the trans-European transport network and repealing Decision No 661/2010/EU. A general revision of the Core Network is foreseen to start in 2023, but earlier modifications resulting from accession of new Countries have to be effected immediately. Furthermore, Council and Parliament, who have adopted the Regulations, have the right to request the Commission for changes at any time, provided of course a sufficient majority supports them. In the area of the non EU-Member States, in particular in the Western Balkans region, there is only the SEETO [5] Comprehensive Network. Those parts of the transport networks of Ukraine and Moldova which are considered as of common European interest, have been incorporated in the “Eastern Partnership Initiative” of the EU. They do not necessarily fulfil the criteria specified for the TEN-T. The task is to assess the existing transport network in the entire project area, consisting of the Western Balkans, Moldova and the western parts of Ukraine, as well as the neighboring EU Member States. For this purpose, the TEN-T planning methodology, as published by the EU Commission [6], will be applied on the “SEETO Comprehensive Network”, the trunk networks of Moldova and western parts of Ukraine and the TEN-T of the relevant EU Member States, to identify the potential TEN-T Comprehensive and Core Network for an enlarged EU. According to the TEN-T planning methodology, these are the links which should fulfill all requirements according to their high strategic functionality and importance, e.g. direct alignment as far as possible, sufficient speed and capacity, intermodal connectivity, environmentally friendly performance. This investigation and comparing existing infrastructure with the quality criteria specified in Regulation (EU) 1315/2013 will help to identify relevant infrastructure deficiencies and to develop proposals for an optimized integrated transport network for the SEE project area. Priority shall be given to sections that continue or connect TEN-T Core Network links towards and/or within adjacent EU Member States, yielding a high European benefit. Further, the focus will be on railways. Railways and inland waterways are considered the backbone of the TEN-T Core Network and of the Core Network Corridors. While there is little space to adapt the shape of the inland waterway network, such options do exist for the rail part, however limited by financial resources. Against the background of the obligation to implement the TEN-T Core Network by 2030 and the poor financial resources available, a clear distinction is necessary between projects to be realized before 2030 and visions of desirable long-term network improvements. As the basis for the further investigations, in the following, the TEN-T planning methodology, as mentioned above, will be described and explained or commented, as relevant for the project area. Then it will be applied in order to develop a proposal for an optimum comprehensive and core network in the area. Comparing with this the existing network, deficiencies will be shown and corresponding suggestions for improvements developed. However, it will not be possible to consider all detail aspects thoroughly. Therefore, in-depth investigations and assessment would be necessary in each individual case, before finally deciding on financing and implementing corresponding projects.

3 THE PLANNING METHODOLOGY FOR THE TRANS-EUROPEAN TRANSPORT NETWORK (TEN-T)

The trans-European transport network (TEN-T) features a dual layer network structure, comprising the comprehensive and, as a subset hereof, the Core Network. Whereas the Comprehensive Network was formed on base of Member States’ proposals, in line with some principle guidance from the Commission, the Core Network was determined by the Commission itself, selecting the strategically most important nodes and links from the Comprehensive Network, by applying a special planning methodology, which had been developed for this purpose, together with expert groups. Minor adjustments resulted from negotiations with Council and Parliament, including the final Trilogue. The following description of this methodology reflects its final status [6]. The current TEN-T as set up in Regulation 1315/2013 [4], was identified strictly in line with the criteria of this planning methodology. It may therefore be assumed that this methodology will also be applied for future adaptations or extensions of the network, in particular in the context of further steps of EU enlargement.

3.1 The Comprehensive Network The Comprehensive Network represents the basic layer of the TEN-T and includes components for all transport modes – rail, road, inland waterway, air and maritime as well as their connecting points and corresponding traffic information and management systems. The corresponding criteria are specified in Articles 9 to 37 of the TEN-T Regulation. The Comprehensive Network, essentially, results from updating and adjusting the previous TEN-T, as re- cast in Decision N° 661/2010/EU [7] of the European Parliament and the Council of 7 July 2010 on Union guidelines for the development of the trans-European transport network. In the Western Balkans, the “SEETO Comprehensive Network” may become the future TEN-T Comprehensive Network, possibly with a few modifications, following the rules specified below. For Moldova and Ukraine, it might be necessary to revise the existing networks accordingly. Updating and adjustment need to abide by a number of rules: a. Updating the current TEN-T to reflect progress in its implementation and adjust it where necessary to changes in national planning, in coherence with EC planning; b. Adding selected and well-defined missing links and nodes, especially in Member States which have acceded to the EU since 2004, where necessary to ensure a homogeneous network planning, a sound modal balance and the interconnection of national networks, and to contribute significantly to TEN-T objectives. Special attention shall be given in this context to network density, which in principle shall correspond to NUTS 2 zones, connecting their main urban nodes as directly as possible, and reflecting spatial distribution of population and of economic and industrial activities. Inland waterways have to comply with Class IV acc.to UN-ECE, as a minimum. (It is a characteristic; in particular of the railway infrastructure in South-Eastern Europe, that in many parts, the existing network does not have the density needed to comply with this criterion. A number of missing links may be assumed, some of which may affect shaping a functional Core Network, according to European and macro-regional needs.) c. Eliminating dead ends and isolated links in the current TEN-T if not justified by geographical particularities, either by removing such links or by extending them to close network meshes.

d. Ensuring that minimum standards for infrastructure and equipment be met in accordance with relevant legislation currently in place (e.g. rail interoperability, road tunnel safety, inland waterway categorisation). (There seems to be a considerable backlog regarding existing infrastructure standards in South- Eastern Europe.) e. Revising the selection of seaports which are open for commercial traffic, according to at least one of the following specific criteria, where annual traffic volumes represent the average of the latest three- years totals for which data covering all Member States are available from EUROSTAT: o Passengers: Seaports connected to the land component of the Comprehensive Network with an annual traffic volume exceeding 1‰ of the total annual EU maritime passenger traffic. o Freight: Seaports connected to the land component of the Comprehensive Network with an annual traffic volume – either for bulk or non-bulk cargo handling - that exceeds 1‰ of the corresponding total annual cargo handled in EU ports. o Seaports located on islands, on condition that they provide accessibility at NUTS 3 or archipelagos level. o Seaports located in outermost regions or peripheral areas, provided their road-distance from another TEN-T port is at least 200 km. f. Revising the selection of airports open to commercial traffic, according to at least one of the following specific criteria, where annual traffic volumes represent the average of the latest three- years totals for which data are available from EUROSTAT: o Passengers: Airports with an annual traffic volume exceeding 1 ‰ of the total annual EU air passenger traffic. o Freight: Airports with an annual traffic volume exceeding 2 ‰ of the corresponding total annual cargo handled in EU airports. o Airports located on islands. o Airports located in peripheral or landlocked areas, provided their distance from another TEN-T airport is at least 100 km or, in case they are connected to a high-speed railway line, at least 200 km. g. For inland ports open to commercial traffic, located on a TEN-T inland waterway and be interconnected with other TEN-T road or rail infrastructure, the volume threshold set out in Decision N° 661/2010/EU remains unchanged. h. Adding any multimodal platform which provides free access to any logistics operator and fulfils one of the following specific criteria: o its transhipment volume exceeds 800.000 tons per year. o it is connected to three TEN-T modal network components or it is the main platform of a NUTS 2 region, connected to two TEN-T modal network components.

3.2 The Core Network

The Core Network is a subset of the Comprehensive Network, overlaying it to represent the strategically most important nodes and links of the trans-European transport network. It is multimodal – i.e. it includes all transport modes and their connections, as well as relevant traffic management systems. The criteria the Core Network has to fulfil are set up in Article 38 to 41 of the TEN-T Regulation.

There are two classes of main nodes:  Primary main nodes (P), fulfilling the corresponding criteria (size, importance) and therefore selected before shaping the network. (They are the vertices of the Core Network, spanning its links.)  Secondary main nodes (S), resulting from shaping the network, e.g. as parts of primary main nodes or crossing or branching points of links or as interfaces of different modes. (Secondary main nodes do not entail additional Core Network links, except local last mile connections.)

The Core Network is identified in the following steps:

Step 1 Identifying of the primary main nodes (P) of the Core Network: These are the nodes of the highest strategic importance in the EU, to be identified in the first step of the planning procedure:  main nodes for passengers and freight,  main nodes for freight only,  main nodes for passengers only.

Step 2 Identifying the links between the primary main nodes: Multimodal links are selected from the Comprehensive Network to connect the primary main nodes, following the corresponding (potential) main traffic flows. The entire inland waterway network which complies with UN-ECE Class IV as a minimum is considered part of the Core Network. The "Motorways of the Sea" are the maritime dimension of the TEN-T. As far as they fulfil the function of Core Network links or of sections thereof (e.g. linking Core Network main nodes across the sea), they are considered part of the Core Network, as well. Secondary main nodes (S) result from geographical and modal intersections of Core Network links as identified according to this methodology.

The main nodes of the Core Network: Primary main nodes are marked with (P), secondary main nodes by (S).

A. Main nodes for passenger and freight traffic: A.1 (P) The capital city of each EU Member State and cities with EU capital function; A.2 (P) Every "Metropolitan European Growth Area" (MEGA in the ESPON Atlas 2006); A.3 (P) Any conurbation or city cluster which, including its environs as defined by the corresponding LUZ ("Larger Urban Zones", acc. to Urban Audit and EUROSTAT) exceeds 1 million inhabitants; A.4 (P) The main city of an island or of a group of islands which is a NUTS 1 region, with at least 1 million inhabitants; A.5 (P) The main border crossing point per mode between each EU Member State with external border and each of its neighboring non EU Member States. Border crossing points only serve as auxiliary points for network planning, but do not provide any other core node function.

(In South-Eastern Europe, apart from capitals and except Istanbul, there are extremely few large cities as characterized in A.2 and A.3, however there are many small capital cities potentially fulfilling criterion A.1.) In the framework of this dual layered approach, urban nodes play an important role within the multimodal Core Network, with regard to their infrastructure both for passengers and for freight. They are particularly relevant in the following respect:  They connect network links – both of the core and the Comprehensive Networks;  They interconnect transport modes, thus enhancing multimodality;  They connect long distance and/or international with regional and local transport (passengers and freight).  The quality of these connections contributes decisively to a well-functioning transport system, in particular to enhance public transport mobility chains and to achieve EU climate goals.

B. Main nodes for freight traffic: B.1 (S) A sea or inland port or a road-rail terminal of a main node according to one of the criteria A.1 – A.4; B.2 (P) A sea or inland port with an annual transshipment volume of at least 1 % of the total transshipment volume of all EU seaports, if interpolating linearly between bulk and non- bulk, complies with the formula: vb/tb + vn/tn ≥ 1 (where vb is the volume of bulk, tb the threshold for bulk, vn the volume of non-bulk and tn the threshold for non-bulk). Seaports which are immediate neighbors and together fulfil the volume threshold, even if individually they would not, may be considered as a cluster, if they have common hinterland connections, except for the "last mile", or if they cooperate closely, e.g. under common management, or supplement each other in function. (In the maps, such cluster ports are indicated individually.) B.3 (P) The largest seaport along each continuous coastline ("façade") of insular Member States and non-insular NUTS 1 regions with access to the sea where no ports are classified according to the criteria B.1 or B.2. This only applies to such façades or coastlines relevant at European scale (e.g. peninsulas longer and wider than 200 km), not taking into account detail coast shapes. B.4 (S) Inland ports which have relevant interface function to Core Network rail links for freight and/or to maritime transport, to be connected to the corresponding modes. B.5 (S) Road-rail terminals which are located in the area of branching or crossing points of Core Network rail links for freight or which are located in the neighborhood (e.g. in the same town) of a core sea or inland port. B.6 (S) Airports with an annual airfreight volume of min. 1 % of the corresponding EU total. B.7 (S) Seaports which are core inland ports according to B.4 and inland ports which are seaports according to B.3.

C. Main nodes for passenger traffic: C.1 (S) The main airport of each urban main node according to A.1 – A.4; C.2 (P) Airports with an annual passenger volume of min. 1% of the corresponding EU total; C.3 (P) The cities relative to Core Network seaports according to the criteria B.2 or B.3, if their population exceeds 200.000 inhabitants in the corresponding LUZ;

C.4 (S) Core Network seaports which have a relevant bridgehead function for (cross-border) passenger ferry connections within the Core Network.

The Links of the Core Network

The following criteria apply on road and rail, only. The land-based Core Network links (road, rail) will be complemented by the "Motorways of the Sea", to give due access to insular Member States and to shortcut connections to or between main nodes on islands or peninsulas. Core Network links shall be of highest importance for long-distance traffic. They thereby contribute to a more homogenous and balanced accessibility structure throughout the Union.

D. Links for passengers and freight: D.1 Neighboring urban main nodes according to A.1 – A.3 shall be connected with each other on road and rail. (Two main nodes are considered as "neighboring", if the corresponding relevant (existing and/or potential) traffic flows between them follow a direct line, not passing through a third main node located somewhere in between.) More distant main nodes will thus be indirectly connected with each other, by which the network will be formed. D.2 In any case, each land border line between two neighboring EU Member states shall be crossed by at least one multimodal Core Network link. D.3 Border crossing points according to a.5 are connected with their corresponding hinterland main nodes acc. to A.1 – A.3, following the relevant traffic flows. D.4 Land connections may be supplemented by links of the "Motorways of the Sea", to connect insular Member States or urban main nodes on islands acc. to A.4 or seaports acc. to B.3 with core seaports of the mainland, or to shortcut detours around bays.

E. Links for freight: E.1 Sea ports acc. to B.2 or B.3 shall be connected to only one hinterland main node each, following the most relevant traffic flows. Connections between ports are not foreseen, but may result from the overall itinerary of a Core Network link. In countries with railways, hinterland connections of Core Network ports must include both road and rail. E.2 The local links of sea and inland ports as well as road-rail terminals according to B.1 and B.4 ("last miles") are considered part of the Core Network. E.3 The entire inland waterway part of the Comprehensive Network.

F. Links for passengers: F.1 In Member States which have railways, airports of urban main nodes acc. to C.1 have to be connected to the rail network by end of 2050, if their annual passenger volume exceeds 1% of the corresponding EU total. F.2 For passengers, seaport cities acc. to C.3 shall be connected to the same hinterland urban node to which the corresponding seaport is linked for freight (acc. to criterion E.1).

G. Omission of links: Core Network links according to D, E or F shall not be foreseen, if: G.1 the link does not exist ("missing link") and its implementation would not be justified by its functionality (e.g. within a potential long distance transport corridor), or not be economically viable, not environmentally sustainable or not feasible by 2030; G.2 the link exists, but does not comply with the requirements of its intended function within the Core Network and its upgrading would not be justified by its function, e.g. within a potential trans-European transport corridor, or would not be feasible by 2030; G.3 the link exists, but the corresponding traffic flows between the relative nodes are negligible (e.g. because of long distance and/or small size of nodes) or can be bundled on other (parallel) links which are in the Core Network due to other (more dominant) functionalities. (Based on this criterion and on G.2, some of the potential direct railway connections of neighboring capital cities, in particular in the Western Balkans, will not be feasible within the planning horizon 2030. As far as there is no such direct railway link in the Comprehensive Network, which could be upgraded to the required standards by 2030, the corresponding missing link will have to persist.) Applying these criteria for the modes individually will allow deviating from the principle of multimodality at the level of links. Some links may comprise only either road or rail.

H. Routing of the links: H.1 Only links of the Comprehensive Network may be selected for the Core Network. (This results in the problem that, as pointed out in the context with the Comprehensive Network, criterion b. above, missing links or large detours may persist in some cases, even in the Core Network.) H.2 The links should be as straight and direct as possible, to follow the relevant long-distance traffic flows, to enhance effectiveness and efficiency of transport, to support territorial cohesion and to contribute to minimum greenhouse gas and polluting emissions as well as to sustainable land use. Exceptions are permitted to fulfill criterion D.2. (Wherever a new link is foreseen to close a gap of the network, the “relevant long-distance traffic flows” shall be understood as potential, according to forecasts after its implementation.) H.3 Detours would be justified to bypass unavoidable obstacles and ecologically sensitive spaces (such as Natura 2000 sites) and to string additional smaller cities, airports, freight terminals, etc.), if not too distant from the direct line and if the disadvantages due to additional detours do not exceed the benefits of improved regional or local accessibility. (Detours may also be acceptable, if a shortcut should prove to be not feasible by 2030, due to technical, political or financial reasons.) H.4 Preferably, the links should follow infrastructure already existing, under construction or planned. Traffic flows shall be bundled wherever possible, considering topographical conditions, environmental impacts, users' needs and potential bottlenecks. H.5 Rail links may have different itineraries for passenger and freight transport, taking into account specific technical parameters (gradients, speed, …) and particular operational situations such as bypasses of nodal areas with high passenger traffic. Deviating from criterion d.1, this may even include bypassing urban main nodes, at a larger scale, to avoid capacity bottlenecks and to follow real cargo flows.

(There are quite some examples in the existing Core Network for rail, that freight and passenger lines, within the same functional corridor, are parallel at a mutual distance in the order of up to 100 km or even more.)

4 THE IMPLEMENTATION OF INFRASTRUCTURE PROJECTS

Within the EU, implementation of the TEN-T Core Network by end of 2030, at the latest, is mandatory, according to the TEN-T Regulation, Art. 38. To achieve this goal, Regulation (EU) No 1316/2013 [8] of the European Parliament and of the Council of 11 December 2013 establishing the “Connecting Europe Facility” (CEF), amending Regulation (EU) No 913/2010 [9] and repealing Regulations (EC) No 680/2007 and (EC) No 67/2010, foresees co-funding through grants and guarantees or other innovative financial instruments and identifies nine multimodal TEN-T Core Network Corridors. These Core Network Corridors are an implementation tool aiming at facilitating the coordination of investments in the infrastructure and improvement of the management of cross-border issues. They cover a great part of the Core Network, linking together, according to a functional geographical context, in particular the large, most complex border crossing projects, and include great part of the rail freight corridors as first defined in Regulation (EU) 913/2010 and partly adapted when integrated into the Core Network Corridors. Whereas Core Network Corridors comprise all available modes of transport, strategic priority will be given to maritime transport, inland waterways and railways. Concentrating CEF funding on these modes of transport, shall induce a sustainable development of transport infrastructure and operation, according to the environmental and climate goals of the White Paper on Transport [10]. Core Network Corridors have a special governance structure, each one headed by a “European Coordinator”, who chairs the corresponding “Corridor Forum”. Whereas these Core Network Corridors cover the territory of all 28 EU Member States, one of them also includes sections passing through Switzerland, another one sections in Norway. As Core Network Corridors do not form a strategic level different from the Core Network, there is, in principle, no difference as regards funding and financing of Core Network projects within or beyond Core Network Corridors. However, given the power of the European Coordinators and the influence of the Corridor Fora, it may be easier to get funds for projects along Core Network Corridors than for ordinary Core Network projects. Accordingly, the EU will support Member States in their obligation to implement the Core Network by 2030, by cross-border coordination of measures to implement projects and provide multimodality, interoperability and technical innovation along the Core Network Corridors, by possible co-funding of the Core Network projects of high European added value, including the border crossing sections in neighbouring countries. With respect to the previous TEN-T budget, the CEF budget is more than three times as high, comprising (including inflation) some 26 B€, of which about 11 B€ are earmarked for core network projects in cohesion countries. Apart from this, the Cohesion Fund foresees some 24 B€ for transport projects, including the TEN-T Comprehensive Network. In third countries, moreover, projects can be funded from the CEF if there is a clear common interest, which might be the case in particular if continuing the core network at cross-border points or for the deployment of EU traffic management systems. Further, there are funds available from the Instrument for Pre-Accession Assistance (IPA) for candidate and potential candidate countries and some smaller funds also for countries of the Eastern Partnership Initiative. Mutual interest in projects would be key for possible financial support.

IPA has been a funding mechanism since 2007 which replaced previous programmes such as ISPA, PHARE and CARDS. Funds are made available to EU candidate countries (Albania, Macedonia, Montenegro, Serbia, Turkey) and potential candidate countries (Bosnia and Herzegovina and Kosovo), however under different conditions.

5 EXTENSION OF THE TEN-T INTO THE WESTERN BALKANS

While in EU Member States, including those in South-Eastern Europe, Regulation (EU) 1315/2013 indicates a Comprehensive Network, with a Core Network as its subset, the SEETO maps only show a “Comprehensive Network”. Similarly, also the network based on the corresponding agreements of the Eastern Partnership Initiative does not distinguish between Comprehensive and Core Network elements. (The so called TEN-T Core Network Corridors, which are formed from Core Network sections linked according to main traffic flows and covering the most complex border crossing projects within the EU, have been defined in Regulation (EU) 1316/2013 in the context of CEF.) By applying the TEN-T planning methodology, with the same criteria which had been used for the current EU, on the Western Balkans, Moldova and the part of Ukraine which is covered by the project area, a comparable proposal for a TEN-T Core Network can be made for these countries. According to criterion A.5 of the TEN-T planning methodology, only one border crossing point per mode and per borderline between each EU Member State and each adjacent non EU Member States was identified as a proxy for the latter. Therefore, any EU enlargement entails applying the TEN-T methodology both on the new Member States, taking into account all neighbouring (see D.1 above!) Core Network nodes in present EU Member States. In the current context, the complete integration of the Western Balkans into the EU is assumed and the methodology to the entire SEETO Comprehensive Network applied. This procedure, which goes beyond just adding the networks of acceding countries to the existing Core Network, may reveal deficiencies of the network, e.g. missing border crossing links, with negative impacts in the entire macro-region. Network modifications even within present EU Member States are a possible consequence. With view to the morphology of the region, the expected traffic demand and the financial possibilities, it will be necessary to apply the criteria indicated in paragraph G in the planning methodology above to omit some of the links identified. This also includes direct connections between countries with a common borderline, according to the criterion D.2. According to the criteria A.1, A.3, B.1 and B.3 of the TEN-T planning methodology, the following cites (in alphabetical order), including their airports, would be Core Network main nodes in the Western Balkans. As postulated in D.1 of the planning methodology, each of these cities would have to be connected with its corresponding neighbouring main nodes, including the core seaports (where applicable), in the region and its relevant neighbourhood: Belgrade: Budapest, Bucharest, Podgorica, Pristina, Skopje, Sofia, Timisoara, Zagreb; Podgorica: Bar (core port); Belgrade, Pristina, Sarajewo, Tirana, Zagreb; Pristina: Belgrade, Podgorica, Skopje, Tirana; Sarajewo: Belgrade, Budapest, Ploce (core port), Podgorica, Zagreb Skopje: Belgrade, Pristina, Sofia, Thessaloniki, Tirana; Tirana: Athens, Durres (core port), Podgorica, Pristina, Skopje, Thessaloniki.

As the annual passenger volumes of all relevant airports are clearly below the threshold of 1 % of the corresponding EU total, according to criterion F.1, no rail connection is required. Further, according to criterion G.1 of the TEN-T planning methodology, seaports must be connected only to that one hinterland main node, which is the most relevant one for the port. Based on information from SEETO, TENtec, internet and maps, the following table 1, which makes reference to the former PETCs, describes the status of the links between these main nodes, which are potential Core Network links (provided that implementation by 2030 is possible). Regarding the Western Balkans, one may distinguish two sets of PETCs relevant for the Western Balkans: those passing from northwest to southeast (nos. IV, VII and X) and those which are crossing them from southwest to northeast (nos. V and VIII), both including their branches. (As the Danube flows through the non-EU Member States Serbia, Moldova and Ukraine, the governance structure of PETC VII is still active.)

Western Neighbouring PETC Mode Status of Infrastructure Balkans Core Core Node (quality indications refer to sections beyond EU Member States only) Node

Belgrade Bucharest Road No direct link (via Turnu Severin – Craiova) in SEETO Network, detour via Timisoara: Belgrade – Pancevo: express way (good), Pancevo – Timisoara: 2 lanes (good), Timisoara – Sibiu – Bucharest: 2 lanes, motorway partly existing, partly under construction

Rail No direct link (via Turnu Severin – Craiova) existing at all, detour via Timisoara:

Belgrade – Pancevo: 2 tracks, electrified (medium), Pancevo – Timisoara: 1 track, not electrified (medium), Timisoara – Turnu Severin – Strehaia: 1 track, electrified, Strehaia – Craiova – Bucharest: 2 tracks, electrified

VII IWW Danube river + missing canal Danube – Bucharest

Belgrade Budapest X Road 100 % motorway (good)

X Rail Belgrade – Indija: 2 tracks, electrified (medium) Indija – Novi Sad – Kelebia - Budapest: 1 track, electrified (medium)

IWW Danube river

Belgrade Podgorica Road 2 lanes (medium – good)

Rail 1 track, electrified, low speed (medium – good)

Belgrade Pristina Road Belgrade – Nis: motorway (good), Nis – Pristina: 2 lanes

Rail detour via Pozega – Kraljewo: 1 track, electrified (medium), shortcut Lapovo – Kraljewo existing, not in SEETO Network: 1 track, not electrified (poor?) Kraljewo – Pristina 1 track, not electrified (medium)

Belgrade Sarajewo Road direct road via Cacak – Gorazde: 2 lanes (poor, medium, good), detour via Velika Kopanica (HR; PETC X, Vc): Belgrade – Velika Kopanica: motorway (good), Velika Kopanica – Zenica: 2 lanes (medium), Zenica – Sarajewo: motorway (good – very good)

Western Neighbouring PETC Mode Status of Infrastructure Balkans Core Core Node (quality indications refer to sections beyond EU Member States only) Node

Rail no direct connection; detour via Strizivojna Vrpolje (HR; PETC X, Vc): Belgrade – Strizivojna Vrpolje: 2 tracks, electrified (medium), Strizivojna Vrpolje – Doboj: 1 track electrified (medium), Doboj – Zenica: 2 tracks, electrified (good), Zenica – Sarajewo: 1 track, electrified (medium)

Belgrade Skopje X Road Belgrade – Nis – Leskovac: motorway (good), Leskovac – Vranje: 2 lanes (good, motorway under construction), Vranje – Skopje motorway (good)

X Rail Belgrade – Vel. Plana: 1 + 1 tracks, electrified (medium), Vel. Plana – Aleksinac: 2 tracks, electrified (medium),

Aleksinac – Nis – Skopje: 1 track electrified (medium)

Belgrade Sofia X, Xc Road Belgrade – Nis: motorway (good), Nis – Sofia 2 lanes (medium – good, motorway under construction)

X, Xc Rail Belgrade – Vel. Plana: 1 + 1 tracks, electrified (medium), Vel. Plana – Aleksinac: 2 tracks, electrified (medium), Aleksinac – Nis: 1 track electrified (medium) Nis – Dimitrovgrad (border SRB-BG): 1 track, not electrified (medium), low capacity due to long 1-track sections, border SRB-BG – Sofia: 1 track, electrified

Belgrade Timisoara Road Belgrade – Pancevo: express way (good),

Pancevo – Timisoara: 2 lanes (good)

Rail Belgrade – Pancevo: 2 tracks, electrified (medium), Pancevo – Timisoara: 1 track, not electrified (medium)

Belgrade Zagreb X Road 100 % motorway (good)

X Rail Belgrade – Indija – Novska: 2 tracks, electrified (medium) Novska – Zagreb: 1 + 1) tracks, electrified

IWW Sava river (up to Sisak)

Podgorica Bar (port) Road 2 lanes (medium, very good)

Rail 1 track, electrified (medium)

Podgorica Pristina Road 2 lanes (very poor, medium)

Rail no direct connection (detour via Pozega – Kraljewo)

Podgorica Sarajewo Road 2 lanes (very poor – good)

Rail no direct connection (detour via Belgrade)

Maritime Short Sea Shipping Bar - Ploce (possible MoS)

Podgorica Tirana Road 2 lanes (very poor – good)

Rail 1 track, not electrified (poor – medium)

Maritime Short Sea Shipping Durres – Bar (possible MoS)

Podgorica Zagreb Road Podgorica - Dubrovnik – Ploce : 2 lanes (medium – very good), Ploce – Zagreb: motorway

Western Neighbouring PETC Mode Status of Infrastructure Balkans Core Core Node (quality indications refer to sections beyond EU Member States only) Node

Rail no direct connection (only detour via Belgrade)

Maritime Short Sea Shipping Bar – Rijeka (possible MoS)

Pristina Skopje Road 2 lanes (medium – good)

Rail 1 track, not electrified (medium)

Pristina Tirana Road mostly motorway or expressway (medium – very good)

Rail no direct connection (detour via Kraljewo – Pozega – Podgorica)

Sarajewo Budapest Vc Road Sarajewo – Zenica: motorway (good – very good), Zenica – Velika Kopanica: 2 lanes (medium),

Velika Kopanica – Osijek – Budapest: partly motorway

Vc Rail Sarajewo – Zenica: 1 track, electrified (medium), Zenica – Doboj: 2 tracks, electrified (good), Doboj – Strizivojna Vrpolje: 1 track, not electrified (medium),

Strizivojna Vrpolje – Osijek: 1 track, not electrified, Osijek – Pecs: 1 track, not electrified, Pecs – Dombovar: 1 track, electrified, Dombovar – Budapest: 1 – 2 tracks, electrified

Sarajewo Ploce (port, HR) Vc Road 2 lanes (medium)

Vc Rail 1 track, electrified (medium – good)

Sarajewo Zagreb Vc, X Road Sarajewo – Zenica: motorway (good), Zenice – Banja Luka: 2 lanes (medium), Banja Luka – Zagreb: motorway (except b/c section BIH-HR) (medium)

Vc, X Rail no direct connection, detour via Strizivojna Vrpolje: Sarajewo – Zenica: 1 track, electrified (medium), Zenica – Doboj: 2 tracks, electrified (good), Doboj – Strizivojna Vrpolje: 1 track electrified (medium), Strizivojna Vrpolje – Zagreb: 2 (Novska – Zagreb: 1 + 1) tracks, electrified

Maritime Short Sea Shipping Ploce – Rijeka (possible MoS)

Skopje Sofia VIII Road Skopje – G.Konjare: motorway (good), G.Konjare – Radomir: 2 lanes (medium), Radomir – Pernik – Sofia: motorway

VIII Rail Skopje – Kumanovo: 1 track, electrified (medium), Kumanovo - border MK-BG – Dolno Selo: missing (planned), Dolno Selo – Pernik: 1 track, not electrified, Pernik – Sofia: 1 track, electrified

Skopje Thessaloniki X Road 100 % motorway or expressway (medium – very good)

X Rail 1 track, electrified (medium)

Skopje Tirana VIII Road Skopje – Tetovo – Gostivar: motorway (good), Gostivar – Struga – Elbasan: 2 lanes medium), Elbasan – Tirana: 2 lanes (medium, motorway under construction)

Western Neighbouring PETC Mode Status of Infrastructure Balkans Core Core Node (quality indications refer to sections beyond EU Member States only) Node

VIII Rail Skopje – Tetovo – Kicevo: 1 track, not electrified (medium) Kicevo – Struga – Khane-e-Thines: b/c section missing Elbasan – Durres – Tirana: 1 track, not electrified (detour via Durres, very poor – medium)

Tirana Athens Road Tirana – Durres – Fier: mostly motorway expressway (good), Regulation 1315/2014 shows planned road Elbesan – Ninesh.

Ninesh – Ioannina: 2 lanes (very poor – medium), Ioannina – Patras: mostly 2 lanes, short motorway sections (Antirio bridge)

Rail Tirana – Durres – Elbesan – Pogradec: 1 track, not electrified (very poor – medium), Pogradec – Florina: b/c section missing, Florina – Platy: 1 track, not electrified, Platy – Athens: partly 2 tracks, electrified, partly 1 track, not electrified

Maritime Short Sea Shipping Durres – Patras (possible MoS)

Western Neighbouring PETC Mode Status of Infrastructure Balkans Core Core Node (quality indications refer to sections beyond EU Member States only) Node

Tirana Thessaloniki Road Tirana – Elbasan: 2 lanes (medium, motorway under construction) Elbasan – Kastoria: 2 lanes (poor – good), Kastoria – Thessaloniki: motorway (very good)

Rail Tirana – Durres – Elbesan – Pogradec: 1 track, not electrified (very poor – medium), Pogradec – Florina: b/c section missing, Florina – Platy: 1 track, not electrified, Platy – Thessaloniki: 1 track, electrified, Greek section (existing and planned) in TEN-T Core Network

Tirana Durres (port) VIII Road 100 % motorway (good)

VIII Rail 1 track, not electrified (medium)

Table 1: The status of the links between main selected nodes (Western Balkans)

Whereas all of these individual potential Core Network links comprise at least 2-lanes roads, this table shows several gaps in the railway network, a few inland waterway connections and some maritime links, along which “Motorways of the Sea” may be established. Though motorways exist only in some connections, the quality status of road is generally better than that of rail. Generally, transport infrastructure is better in the northwest than in the southeast of the Western Balkans and better along PETC X than in the mountainous areas of Bosnia-Hercegovina, Montenegro, Albania, Kosovo and FYROM. Rail links are missing mainly between the “small” capital cities in the mountainous parts of the Western Balkans, mostly offside former PETCs. Considering the rather bilateral importance of these connections and the relatively little potential traffic demand, while there is a pronounced scarcity of funds, is not likely in most of these cases, that these gaps could be closed by 2030. Realistically, they should not be included in the Core Network at this time. It seems to be more appropriate to serve these connections on road, which should be upgraded to a higher level of traffic safety.

In general, the situation is much more favourable along sections of the former PETCs. In most cases, they follow the main traffic flows and comprise railway lines, even though in many cases not in good shape. Within PETC VIII, the cross-border sections AL-MK and MK-BG are planned or under construction. As shown in the table above, all PETC passing through the Western Balkans have been entirely endorsed by applying the TEN-T planning methodology, except corridor branch Xd from Veles to Florina. In the Western Balkans, a region which is surrounded by EU Member States, they even carry relevant intra- EU traffic flows, e.g. from Greece to Germany or from Romania to Italy. Priorities should therefore be focused on these sections, which quite evidently are of common interest and high European added value. The Danube river, PETC VII and now a part of the multimodal “Rhine-Danube” Core Network Corridor, has always been the natural main artery of South-Eastern Europe. Nevertheless, there are still quite a few bottlenecks of the fairway, all of them beyond the Western Balkans region, in Germany, Austria, Hungary and along the stretch between Romania and Bulgaria. The Sava-Morava-Vardar line forms a natural corridor as well, which has been used since antique times then called “Via Militaris” or “Via Diagonalis”. Along this very backbone of the Western Balkans, represented by PETC X, cities like Ljubljana, Zagreb, Belgrade, Nis, Sofia, Edirne, Istanbul, Skopje and Thessaloniki have developed and the first railway lines of the region were constructed. However, PETC X has not become a Core Network Corridor in the recent TEN-T legislation, mainly because the Alpine crossing railway section (“Tauernbahn”) with gradients up to 30 ‰ does not allow operating 740 m long loaded freight trains as specified in the TEN-T Regulation1. PETC IV, in particular the Romanian-Bulgarian border section from Timisoara to Sofia including the new Calafat - Vidin Danube bridge, was created in the 1990ies as a bypass to former Yugoslavia, to avoid passing through that region during war and political instability. It is a corridor, assembled from sections of quite different spatial functionalities like Berlin – Prague – Vienna – Bratislava – Budapest, Vienna – Budapest – Arad – Bucharest – Constanta, Sofia – Istanbul and Sofia – Thessaloniki. Now it is part of the “Orient- EastMed” Core Network Corridor. Against this background, this corridor lacks end-to-end traffic flows, which is not least up to a detour of more than 100 km, due to its curved itinerary between Turnu Severin, Craiova, Vidin and Sofia. As far as effective, these three PETCs, rather than being in competition, supplement each other in their function connecting the countries in the south-eastern quarter of Europe, including the Western Balkans, with the economically strong north-western parts of the continent and thus improving, in a sustainable way, accessibility, cohesion, economic growth, welfare and political stability in the region. While PETC V Venice – Trieste/Koper – Ljubljana –Budapest – border HU-UA and Vb Rijeka – Zagreb – Budapest are fully located within the EU and covered by the “Mediterranean” Core Network Corridor, PETC Vc passes through the non-EU Member State Bosnia-Hercegovina, where it does not belong to TEN-T. It connects the port of Ploce, which though Croatian mainly serves Bosnia and Herzegovina, with Sarajevo, the east of Croatia (Slavonia) and Budapest. PETC VIII connects the port of Durres via Tirana, Skopje and Sofia with the port of Burgos at the Black Sea. While the existing road is end-to-end, the cross-border railway sections AL-FYROM and FYROM-BG are missing. As regards the road part of PETC VIII, the existing parallel “Via Egnatia” motorway Igoumenitsa – Thessaloniki – Alexandroupolis – Kipi in Northern Greece should be taken into account, as well.

1 From perception of Slovenia not including »Tauernbahn“ into TEN-T Core network, as it has not met the technical criteria until year 2030, should not be the reason to prejudge also an exclusion to connect Western Balkans with Central Europe on PETC X. Connection from Villach to Salzburg is an important connection for further development of PETC X trough Alpine space especially for ever growing combined transport. As transport flows indicate high demand for connecting SEE area with CE, both corridors of PETC X and PETC Xa should take part of viable and considerable connector of both freight and passenger rail transport. 20

The following table 2 indicates which road and rail links would be proposed for a potential future TEN-T Core Network (red: Core Network road links, green: Core Network rail links), in line with the TEN-T planning methodology and considering the status of these links as expounded in table 1:

Main Node 1 Main Node 2 Mode of Transport Road: via Timisoara, possible shortcut via Turnu Severin or Vidin Belgrade Bucharest (to be investigated) Rail: via Timisoara Road (PETC Xb) Belgrade Budapest Rail (PETC Xb) Road Belgrade Podgorica Rail Road Belgrade Pristina Rail: via Pozega – Kraljewo (electrified) or shortcut Lapovo – Kraljevo (if added to the Comprehensive Network) Road via CACAK – Gorazde Belgrade Sarajewo Rail: via Velika Kopanica (PETC X, HR, line Beograd – Zagreb)

Road (PETC X) Belgrade Skopje Rail (PETC X)

Road (PETC X, Xc) Belgrade Sofia Rail (PETC X, Xc)

Road Belgrade Timisoara Rail

Road (PETC X) Belgrade Zagreb Rail (PETC X)

Road Podgorica Bar (port) Rail (only freight)

Road Podgorica Pristina Rail (excessive detour, implementation of shortcut not feasible)

Road Podgorica Sarajewo Rail (excessive detour, implementation of shortcut not feasible)

Road Podgorica Tirana Rail

Road Podgorica Zagreb Rail (excessive detour, implementation of shortcut not feasible)

Road Pristina Skopje Rail

Road Pristina Tirana Rail (excessive detour, implementation of shortcut not feasible)

Main Node 1 Main Node 2 Mode of Transport

Road (PETC Vc) rajewo Budapest Rail (PETC Vc)

Road (PETC Vc) Sarajewo Ploce (port, HR) Rail (PETC Vc)

Road: via Banja Luka and PETC X Sarajewo Zagreb Rail: via Velika Kopanica (PETC X, line Beograd – Zagreb)

Road (PETC VIII) Skopje Sofia Rail (PETC VIII): missing link border FYROM-BG

Road (PETC X) Skopje Thessaloniki Rail (PETC X)

Road (PETC VIII): via Elbasan (planned motorway section) Skopje Tirana Rail (PETC VIII): missing link border FYROM-AL

Road: via Elbasan (planned motorway section) Tirana Athens Rail: via Durres

Road Tirana Thessaloniki Rail: via Durres

Road Tirana Durres (port) Rail (freight and passengers, part of Tirana – Athens/Thessaloniki) Table 2: Selection of road and rail links for the potential Core Network, acc. to TEN-T planning methodology (WB)

All inland waterways of the TEN-T Comprehensive Network (≥ Class IV acc. to UN-ECE) belong to the Core Network automatically. The Tisa river up to Szeged has been included in the present Core Network already. Based on the TEN-T planning methodology, following inland ports and rail-road terminals (RRT) result as secondary core nodes: Belgrade inland port and RRT (criterion B.1), Novi Sad inland port (criterion B.4), other RRTs if any (criterion B.5). The following maps (fig. 1 - 4) visualise the proposed future Core Network for inland waterways, railways (freight and passengers) and roads.

Inland waterways Railways (freight)

Railways (passengers) Roads Fig. 1 – 4: Potential future TEN-T Core Network in the Western Balkan: Inland waterways and ports: blue; Railways: green; Road: red; additional Core network links within present EU: dotted.

As shown in these maps (Fig. 1 – 4), including the Western Balkans into the TEN-T planning would affect also the Core Network in surrounding EU Member States, e.g. the road and rail links in the Hungarian and Croatian sections of PETC Vc (dotted lines Budapest – Osijek – border HR-BIH). In the Western Balkans, the road part of the Core Network generally results denser than the rail part. The conditions in that area allow following also criterion D.2, according to which each land borderline between two neighbouring EU Member states shall be crossed by at least one multimodal Core Network link. This is not the case for rail, where mountains and low potential traffic demand are severe obstacles to establishing

a “complete” Core Network. In detail, this leads to the following observations regarding possible missing links: There are existing roads between Belgrade and Turnu Severin or Vidin, which have not been adopted for the SEETO Comprehensive Network, therefore the missing road link (as a part of a direct connection Belgrade – Bucharest) shows only on the maps. However, without being part of the Comprehensive Network, it cannot be selected for the Core Network. On rail however, such a direct link is missing in the real world. As it is not likely that such rail link would be implemented by 2030, it should not be added to the Core Network. The same applies also for the missing railway link between Lapovo and Kraljewo in the connection Belgrade – Pristina. As a precondition for closing these gaps in a future Core Network, the missing links would have to be adopted in the SEETO Comprehensive Network, which would have to be revised accordingly, after thorough strategic considerations, taking into account the corresponding costs and benefits, together with the affected countries. Apart from border crossing sections or points, bottlenecks can be found at many places, mainly resulting from the superposition of long-distance with regional or local traffic or due to insufficient capcity in sectiond between the nodes: On road, the most prominent example is the Belgrade area and the missing motorway sections along PETC V. The problem will be solved once Belgrade motorway bypass will be fully in operation, of which about 70 % will be finished by end of 2014. Congestion in Croatia south of Ploce, due to tourism, has mainly seasonal character. In this part, the problem of the Neum corridor will have to be solved. Te easiest way would be passing through the territory of Bosnia-Hercegovina, which currently is an external border, but will not be once the Western Balkans will have joined the EU. Other solutions like a bridge to Peljesac peninsula or tunneling the BIH territory would require much higher investments and entail higher maintenance costs. Existing roads and railway lines of the potential future Core Network should be upgraded and new links constructed on the base of the deficiencies of the existing infrastructure as expounded in table 1 above and of the necessary further detail studies, to be completed in 2030, at the latest. Larger infrastructure projects should be prioritised according to their location along the Pan-European Corridors, in particular PETC X and its branches, and corresponding cost-benefit analyses, based on traffic demand. As a priority, existing railways should be rehabilitated, upgraded and electrified where necessary. Further, a better modal integration should be achieved by upgrading sea and inland ports and road-rail terminals (RRTs). However, local improvements of border crossing infrastructure as well as operation and administration of border crossing, which would be extremely effective and need much lesser investments, should therefore have absolute priority.

6 POSSIBLE CONSEQUENCES FOR EU MEMBER STATES AROUND THE WESTERN BALKANS

Beyond deficiencies in Western Balkans, a large-scale view shows shortcomings of network configuration also within EU Member States, along the access to this macro-region. They affect accessibility and unhampered access to the most important market places, mainly in Western Europe, from great parts of South-Eastern Europe. According to the findings of Expert Working Group 1 of the TEN-T Policy Review [11], “improving accessibility for peripheral regions does not necessarily mean the construction of transport infrastructure within those areas, but rather towards them, thus not excluding measures in central regions of the Union”. This means that also improvements along the access routes to a region might contribute to its economic and social cohesion. If on inland waterway or rail, this would also enhance sustainability. Against this background and given the particular strategic effects of the following presumed network deficiencies in key access areas to South-Eastern Europe and the Western Balkans, it might make sense to consider certain adaptations of the existing TEN-T Core Network, which Art. 54 of TEN-T Regulation foresees as from 2023. The countries involved, as well as the affected regions and infrastructure managers, are suggested to cooperate in multi-lateral working groups, including the European Commission, on a level playing field, to study and investigate jointly and thoroughly each of the following proposals. Studies already existing should be compared and evaluated, in particular if they are contradictory. Finally, all measures and their priorities should be agreed on the base of realistic, commonly accepted data, facts and studies, e.g. the latest traffic forecasts, socio-economic costs and benefits, and trade-offs among the countries and with European interests. EU co-financing options should consider the European added value of improvements, taking into account spatial integration and accessibility, regional economy, social development as well as sustainability. Starting in the northwest, there is Tauern railway Salzburg – Villach within PETC X. It used to be the most important route from Germany to former Yugoslavia, but has lost importance due the European re- orientation after the fall of the Iron Curtain and the collapse of Yugoslavia. With gradients up to 30 ‰, it is also too steep for fully loaded 740 m long freight trains (which Regulation no. 1315/2013 stimulates for freight railway lines of the Core Network). As only in recent years, the existing line was upgraded for higher speeds (for passenger trains) and higher capacity, there are no plans to replace it by a 30 – 40 km long base-tunnel (which would cost several billion Euros), even less as the existing parallel Pyhrn railway complies already currently to a large extent with Core Network freight railway standards. An economic solution could be splitting the core function, with Tauern railway for high-level passenger traffic and Pyhrn railway line for freight, according to criterion H.5 of the TEN-T planning methodology, which would not prevent light freight operation on Tauern in the existing quality and secondary passenger traffic on Pyhrn. (While the so-called “SETA” line Vienna/ Bratislava – Zagreb would be appropriate for connecting these agglomerations, Pyhrn railway axis would follow the northwest-southeast orientation of PETC X.). For the north-western access to South-Eastern Europe, shortcutting the existing 60 km PETC Xa main connection via Zidani Most of the Maribor – Zagreb link trough “Krapina railway” connection, which would be in line with the principles of TEN-T planning, may be considered a long-term investment option, not before 2030. There are several studies on this issue, based on differing assumptions, with controversial results as regards costs and benefits. So far, also discussions between countries involved have not yielded a common view. In particular, Slovenia as one of the two countries directly concerned by its potential implementation does not support it. Currently, the focus of Slovenian railway infrastructure development is prioritised on main bottlenecks on the single track railway lines (construction of II. track): Divača-Koper, Ljubljana-Jesenice, Maribor-Spielfeld and Ljubljana railway junction. Against this background, continuing thorough investigations and fair discussions might be advisable.

Next Steps for Corridor X/Xa This report is focussing on the possible development of TEN-T network, considering also long term options. We have to be aware that any high performing TEN-T freight core network corridor, enabling to cross the Alps efficiently with heavy weight, long freight trains could be realised far beyond 2030. However it is important to come to short term improvements along the corridor. One essential possibility would be the establishment of a new Rail Freight Corridor (RFC) according to Regulation (EC) No 913/2010 concerning a European rail network for competitive freight. The legal possibility to create such a new rail freight corridor is given: Regulation (EU) No 1315/2013 of the European Parliament and of the Council of 11 December 2013 on Union guidelines for the development of the trans-European transport network defines some interfaces between the TEN-T network and the RFCs. It does not suspend Regulation (EC) No 913/2010. Neither does Regulation (EU) No 1316/2013, creating the Connecting Europe Facility, which modifies the alignment of existing rail freight corridors. According to Article 5 of Regulation (EU) No 1316/2013, selection of further RFCs, Member States Concerned can apply to the Commission to establish a new corridor; the Commission has to examine the proposal. As far as appropriate for heavy freight, a new corridor has not to be restricted to the TEN-T core network. So from that point of view there is no hindrance to create a RFC X/Xa. However, inclusion on Non EU-Member States like Serbia into a new RFC would bring about some challenges. There is the need for taking certain EU-Regulations into national law for Non EU Member States to participate at a RFC. In case of Corridor X/Xa, taking into account the existing infrastructure and the transport flows, it is necessary to provide access to both branches Tauern and Pyhrn. In principle, the new RFC has to take into account the current limitations of infrastructure of the two branches. The approach of the rail freight corridor would offer the possibility to develop a common transport market study and, based on that, an infrastructure concept commonly agreed, taking into account the potential of infrastructure development on both branches (Tauern and Pyhrn), the market needs and as well the infrastructure development and available capacity in the neighbouring countries (e.g. Germany) This might provide an essential input to the discussion for the next revision of the TEN-T network, especially to investigate how technical parameters according to regulation (EU) 1315/2013 for rail freight on the core network can be met. The RFC should be used to deliver concepts of making the best use of existing infrastructure, to identify small short term investments e.g. to improve interoperability, reduce border stops etc. in order to boost the competitiveness of railways in the corridor.

Another suggestion for further investigations and discussion might be a possibly better integration of Koper into the network than currently foreseen in the TEN-T, which might strengthen its position as a seaport and enable a more efficient use of the planned Divaca – Koper railway link. Reference is made to the existing “Adria A” study [12]. Finally, the above mentioned gap between Belgrade and Bucharest exists for road exists only on paper, in the SEETO Comprehensive Network in Serbia, because two lanes national roads exist both in Romania and in Serbia. On rail, the shortest existing connections takes a detour in the order of 100 km through Timisoara. A shorter connection could be considered at the long run, certainly not for the 2030 horizon. As explained earlier, PETC IV was created in 1994 (Crete) and 1997 (Helsinki), as a basis for the 2004/2007 EU enlargement, bypassing the Western Balkans, where there was war and crisis at that time. Actually, this corridor, which is now the “Orient-EastMed Core Network Corridor”, was assembled from various sections with quite different spatial and economic functionalities and has never become really effective as a whole. In particular, the large-scale detour of some 120 km within the section Turnu Severin – Craiova – Calafat –

Vidin – Montana – Sofia seems to weaken the effectiveness of the corridor considerably. Shortcutting this routing, either by more direct access routes on both sides, mainly in Romania) or by constructing a new railway bridge close to Lom, could make the whole corridor effective. As most of these gaps exist already in the TEN-T and/or SEETO Comprehensive Networks, the first step to close them should be revising these networks accordingly, on the base of corresponding analyses. In many cases, this could be done in studies initiated within corresponding TEN-T Core Network Corridors and funded by CEF. Last, but not least, it would be up to EU Member States Germany, Austria, Hungary, Romania and Bulgaria, to remove the still remaining bottlenecks along the Danube river, in order to establish a reliable and sustainable high performance inland waterway as the very backbone of the entire Danube macro-region, which includes the Western Balkans as well as Moldova and parts of Ukraine.

7 EXTENSION OF THE TEN-T TO MOLDOVA AND WESTERN UKRAINE

Compared with the Western Balkans, the region comprising Moldova and the western part of Ukraine is not surrounded by EU Member States, but only adjacent in the east to Poland, Slovakia, Hungary and Romania, and there is, beyond association with the EU, no concrete perspective of EU accession. Further, the work done by the Eastern Partnership Initiative is much less progressed and going in depth that what SEETO has delivered already. Further, there is no borderline between the western parts and the rest of Ukraine, so it is not possible to assume border-crossing points according to criterion A.5 of the TEN-T planning methodology. As a pragmatic alternative, reference nodes outside the area which would fulfil one of the criteria A.1, A.2 or A.3 are selected to span a potential future TEN-T Core Network. Taking into account these facts, the following steps follow the same principles as for the Western Balkans, with PETC V and PETC IX as backbones. In Moldova and the western parts of Ukraine to be considered within this study, there are only two cities which may be main nodes of a potential future Core Network: Chisinau and Odessa. Chisinau is the capital city of Moldova and would therefore be a main node according to criterion A.1 of the TEN-T planning methodology. Odessa has 990.000 inhabitants, but exceeds 1 million, in its “larger urban zone” (LUZ), so that it would be a Core Network node according to criterion A.3. Lviv oblast does not belong to the area covered by the South-East Europe Transnational Cooperation Programme of the EU, but as traffic does not end at borders, even less at the arbitrarily selected edges of a project area, it is indispensable to include this province into the investigation, due to its eminent political, cultural, economic and transport related importance. Lviv with about 750.000 inhabitants might probably (depending on possible clustering with surrounding settlement areas – the “larger urban zone”) not fulfil criterion A.3. Compared with Polish cities like Gdansk, Wroclaw or Krakow, it may, however, be assumed that it would fulfil the criteria to become a “MEGA” (Metropolitan European Growth Area), so that it could be a Core Network main node according to criterion A.2. Kijiv, the capital of Ukraine, will be considered only together with Chisinau and Odessa, but not with Lviv, as this link is located outside the area of the South-East Europe Transnational Cooperation Programme. Applying the TEN-T planning methodology with the above mentioned adaptations in line with the particular situation in Moldova and the western parts of Ukraine, each of these additional main nodes (in alphabetical order) would have to be connected to the following neighbouring main nodes:

Chisinau: Bucharest, Budapest, (Kyjiv), (Lviv), Odessa, Timisoara; (Lviv): Bratislava, Bucharest, Budapest, Chisinau, Timisoara; Odessa: Bucharest, Chisinau, (Kyjiv). The following table 3 is based on information from TENtec, internet and maps, which are only partly complete (in particular as regards number of tracks) and reliable.

Moldova and Neighbouring PETC Mode Status of Infrastructure west. Ukraine Core Node (MD, UA: broad gauge 1520 mm; no quality parameters available) Core Node

Chisinau Bucharest Road Chisinau – Cimislia: 4 lanes Cimislia – Giurgiulesti – Galati – Buzau: 2 lanes, Bucau – Bucharest: 4 lanes

IX Rail Chisinau – Iasi: 2 tracks, not electrified, Iasi – Bacau – Ploesti – Bucharest: 2 tracks, electrified

Chisinau Budapest Road Chisinau – Iasi: 2 lanes, Iasi – Targu Frumos: 4 lanes, Targu Frumos – Piatra Neamt – Cluj – Oradea – Debrecen: 2 lanes Debrecen – Budapest: motorway

Rail Chisinau – Iasi: 2 tracks, not electrified, Iasi – Pascani – Suceava: 2 tracks, electrified, Suceava – Cluj: 1 track, electrified, Cluj – Oradea – Püspökladany: 1 track, not electrified, Püspökladany – Budapest: 2 tracks, electrified

Chisinau Kijiv IX Road Chisinau – Cajkivka: mostly 2 lanes, Cajkivka – Kyjiv: motorway

IX Rail Chisinau – Tiraspol – Rozdilna: 2 tracks, not electrified, Rozdilna – Zmerynka – Kyjiv: 2 tracks, electrified

Chisinau Lviv Road Chisinau – Balti – Cernivci – Ivano-Frankivsk – Lviv: mostly 2 lanes

Rail Chisinau – Iasi: 2 tracks, not electrified, Iasi – Pascani – Suceava: 2 tracks, electrified, Suceava – Cernivci – Ivano-Frankivsk – Lviv: mostly 2 tracks, not electrified

Chisinau Odessa Road Chisinau – Tiraspol - Odessa: mostly 2 lanes

Rail Chisinau – Tiraspol – Rozdilna: 2 tracks, not electrified, Rozdilna – Odessa: 2 tracks, electrified

Lviv Bratislava Va Road Lviv – Ushgorod: 2 lanes, Ushgorod – Kosice – Zilina – Bratislava: partly motorway (partly under construction)

Va Rail Lviv – Csop – Kosice – Zilina – Bratislava: 2 tracks, electrified

Lviv Bucharest Road Lviv – Ivano-Frankivsk – Cernivci – Suceava – Roman: 2 lanes, Roman – Buzau – Bucharest: 4 lanes

Rail Lviv – Ivano-Frankivsk – Cernivci – Suceava: 2 tracks, not electrified, Suceava – Buzau – Ploesti – Bucharest: 2 tracks, electrified

Lviv Budapest V Road Lviv – Mukacheve – Berehove: 2 lanes, Berehove – Nyiregyhaza – Budapest: mostly motorway

Moldova and Neighbouring PETC Mode Status of Infrastructure west. Ukraine Core Node (MD, UA: broad gauge 1520 mm; no quality parameters available) Core Node

V Rail Lviv – Csop – Nyiregyhaza – Miskolc/Debrecen – Budapest: mostly 2 tracks, electrified

Lviv Timisoara V Road Lviv – Mukacheve – Satu Mare – Oradea – Arad – Timisoara: 2 lanes

V Rail Lviv – Csop: 2 tracks, electrified, Csop – Satu Mare – Oradea – Arad: 1 track, not electrified, Arad – Timisoara: 1 track, electrified

Odessa Bucharest Road Odessa – Izmajil – Reni – Galati – Buzau: 2 lanes, Buzau – Bucharest: 4 lanes

Rail No direct line existing, detour via Chisinau - iasi

Odessa Kyjiv Road Odessa – Kyjiv: motorway

Rail Odessa – Kyjiv: 2 tracks, electrified Table 3: The status of the links between main selected nodes (MD, UA) Also in this region, regarding network configuration and density, the road network is more complete than the railway network. However great part even of the main road links consist of ordinary 2-lanes roads, while the share of motorways is considerably less than in the Western Balkans. On the other hand, the rail network is, in average, in better quality than in the Western Balkans, in particular in Ukraine. However, extensive detours on rail can be observed in the relations Chisinau – Bucharest via Iasi and Odessa – Bucharest via Chisinau – Iasi. Realistically, this will not change before 2030. While such direct links will not be proposed for a potential future Core Network, it would make sense, to foresee, as a long-term vision, a new railway link Chisinau – Giurgiulesti – Galati – Bucharest, with a branch Odessa – Izmajil – Reni – Giurgiulesti. Equally, from an all-European point of view, at the long run, a road and (European standard gauge) rail corridor from the Baltic Sea via Warsaw – Lviv – Cernivci – Bucharest to the Black Sea, Istanbul and the Aegean Sea, bypassing the Carpathian Mountains in the east, should be envisaged. As a priority, however, existing railways should be rehabilitated, upgraded and electrified where necessary. Further, a better modal integration should be achieved by upgrading sea and inland ports as well as road- rail terminals (RRTs), with particular focus on sections of Pan-European Corridors V and IX.

Main Node 1 Main Node 2 Mode

Road (deviating from PETC IX): via Galati Chisinau Bucharest Rail (PETC IX): via Iasi

Road: via Iasi – Cluj-Napoca - Oradea Chisinau Budapest Rail: via Iasi – Suceava – Cluj-Napoca - Oradea

Road (PETC IX): via Cajkivka Chisinau Kijiv Rail: (PETC IX): via Tiraspol

Road: via Balti - Cernivci Chisinau Lviv Rail: detour via RO (Iasi – Suceava) and Cernivci

Road: direct Chisinau Odessa Rail: via Tiraspol

Road: via Ushgorod Lviv Bratislava Rail: via Csop

Road: via Cernivci - Suceava Lviv Bucharest Rail: via Cernivci - Suceava

Road: via Berehove Lviv Budapest Rail: via Csop

Road. direct (via Izmajil – Reni – Galati) Odessa Bucharest Rail: detour via Chisinau and Iasi

Road: via Satu Mare - Oradea Lviv Timisoara Rail: via Satu Mare - Oradea

Road: direct Odessa Kyjiv Rail: direct

Table 2: Selection of road and rail links for the potential Core Network, according to TEN-T planning methodology (MD, UA)

Based on the TEN-T planning methodology, the following inland ports and rail-road terminals (RRT) are to be considered as secondary core nodes: Odessa seaport according to criterion B.1 as well as Giurgiulesti (MD) and Izmajil (UA) inland ports inland ports according to criterion B.4. RRTs would be selected from

those nominated for the Comprehensive Network, following criterion B.5 of the TEN-T planning methodology. The following maps (fig. 5 - 8) visualise the proposed future Core Network for inland waterways, railways (freight and passengers) and roads. As shown in these maps by dotted lines, extending TEN-T planning to Moldova and the western parts of Ukraine would affect also the Core Network in EU Member States, namely in Romania and Hungary.

Fig. 5 – 8: Potential future TEN-T Core Network in Moldova and western parts of Ukraine: Inland waterways and ports: blue; Railways: green; Road: red; additional Core network links within present EU: dotted.

8 POSSIBLE CORE NETWORK CORRIDORS

As described above in Chapter 4, TEN-T Regulation (EU) No 1315/2013 provides the possibility of determining so-called Core Network Corridors within the EU territory, which have to comprise all available modes of transport, consist exclusively of nodes and links of the Core Network and cover at least three Member States. The rail part of each Core Network Corridor should correspond with a rail freight corridor according to Regulation (EU) No 913/2010, which in most cases is largely the case. TEN-T Core Network Corridors are a means for infrastructure implementation; with particular focus on complex cross-border projects. Therefore, it is the CEF Regulation (EU) No 1316/2013 which defines nine such corridors. This covers corresponding governance structures with a Corridor Forum for each such Core Network Corridor and a European Coordinator on top, as well as a particular priority in funding. The following fig.9 shows the currently existing TEN-T Core Network Corridors. Given the natural alignment of the Danube inland waterway through non-EU territory, this figure shows its section passing through Serbia as a dotted line.

Fig. 9: TEN-T Core Network Corridors

Against this background, there are legal obstacles within the TEN-T, to introduce such Core Network Corridors, both in the Western Balkans and in Moldova or Ukraine, as long as these regions are not part of the EU. Functionally, it might make sense to define fictive Core Network Corridors also for the Western Balkans, maybe even for Moldova and Ukraine and to establish analogous governance structures, but on a different legal base than for the TEN-T Core Network Corridors, in order to develop potential Core Network nodes and links beyond the EU territory in a more structured way. Apart from closing the inland waterway gaps of the existing “Rhine-Danube Corridor” through Serbia, Moldova and Ukraine, it seems quite evident that, following the accession process, the former Pan- European Corridors (PETC), which also would actually fulfil the criteria specified in the TEN-T Regulation, should become – step by step – new Core Network Corridors. or extensions beyond external borders of already existing Core Network Corridors or their branches In the Western Balkans Region, this should be the following PETC and/or branches:  X Salzburg – Villach – Ljubljana – Zagreb – Belgrade – Nis – Skopje – Thessaloniki and its branches Xa Graz – Maribor – Zidani Most – Zagreb, Xb Budapest – Belgrade and Xc Nis – Sofia, which after the nine existing TEN-T Core network Corridors could become the tenth. There are questions and options on which the countries concerned could not develop a common view up to now, such as the option of upgrading Pyhrn Axis for heavy and long freight trains and questions on possible shortcuts.  Additional branches Belgrade – Pristina – Skopje and Belgrade – Podgorica Bar - Durres/Tirana should be added to include Kosovo as well as Montenegro and Albania into PETC X.  Currently, for the rail part, as described on page 26 (text box), there are activities to establish a Rail Freight Corridor according to Regulation (EU) 913/2010 along PETC X/Xa branching at Zidani Most to Ljubljana and Villach (Salzburg) on one side and Maribor and Graz (Linz/Wels) on the other side, taking into account the specificities of each branch. The instruments of the rail freight corridor can contribute to find a common approach for the infrastructure development.  Vc Budapest – Osijek – Sarajewo – Ploce (as a branch of the existing “Mediterranean Corridor”),  VIII (maybe including also existing Greek “Egnatia Odos” as additional, already existing road part), with link Sofia – Ruse – Bucharest. PETC X, including also the additional branches Belgrade – Pristina – Skopje and Belgrade – Podgorica – Bar – Durres/Tirana, as well as PETC Vc, will have to become a TEN-T Core Network Corridor at the moment when the corresponding countries accede to the EU, in order to connect all new EU MS in the region to the TEN-T Core Network Corridor system. This having been done, PETC VIII sections would have to be added to the TEN-T Core Network, according to the TEN-T planning methodology, but a new TEN-T Core Network Corridor along PETC VIII may be established at a later time, according to needs and funding possibilities. In Ukraine and Moldova, the following PETC could be considered as new or extended TEN-T Core Network Corridors:  III (Berlin/Dresden – Wroclaw – Katowice – Lviv),  V Czop – Lviv – Kyjiv (extension of current “Mediterranean Corridor”),  IX Kyjiv – Odessa/Chisinau – Bucharest – Dimitrovgrad – Alexandroupolis (possibly with an improved, i.e. more direct alignment between Chisinau and Bucharest). As an option corridor could be understood as an extension of PETC VIII from Bucharest towards Iasi – Chisinau – Odessa/Kyjiv. At the long run, an additional multimodal corridor connecting the Baltic Sea and the Black Sea beyond the Carpathians, passing through Warsaw, Lviv and Bucharest might make sense from an all-European view.

The following map fig. 10 shows the PETC system in South-Eastern Europe as it was defined in 1997, while the map fig. 11 reflects the above proposed extension of the existing TEN-T Core Network Corridors in the Western Balkans (based on PETC X, V and VIII).

Fig. 10: Pan-European Corridors in South-Eastern Europe (Source: EC DG TREN, JAAKKO PÖYRY INFRA, VTT)

Fig. 11: Proposed TEN-T Core Network Corridor system for rail in Western Balkans, based on PETC V, VIII and X:

Figure 11 presents the proposed corridor sections of the railway Core Network system for rail in Western Balkans based on V, VIII and X Pan-European Transport Corridors. In order to meet the specifications for Core TEN-T network railway lines according to Regulation (EU) 1315/2013, additional detailed investigation based on technical feasibility and socio-economic cost-benefit analysis should take place. Analysis and RFC studies should take in to consideration also infrastructural limitations of sections and propose exact development of TEN-T corridors in Western Balkans. Any possible division of specific functions on freight and passenger Core Transport network can only be made on basis of precise analysis and discussions among countries involved.

9 TEN-T AND TRANSPORT DEMAND This chapter is aiming to synthesise the results of WP4 including the estimation of scenarios with former considerations of this report. As the method of defining the TEN-T network does not take into account capacities nor capacity shortfalls in a first step, the results shall be compared with forecasted traffic flows. WP4 will show capacity shortfalls in the network, where new infrastructure of upgrading of existing infrastructure is needed. On the other hand, the considerations based on the TEN-T methodology show as well the need for new connections and infrastructure elements. It shall be analysed in how far same infrastructure elements are concerned or if the needs due to capacity shortfalls are different to the one driven by the needs of connectivity according to TEN-T methodology. In WP4, a thorough analyses of current and future traffic flows in South-Eastern Europe was carried out for three time horizons: currently, 2020 and 2030, based on various supply scenarios. Different situations at border crossing points (BCP) and in transport infrastructure sections were considered.

From these investigations, the following scenarios resulted:

Scenario 0: current situation Scenario 1: 2020; improved BCP, infrastructure currently under realisation Scenario 2: 2020; improved BCP, infrastructure currently under realisation + planned infrastructure Scenario 3: 2020; improved BCP, infrastructure currently under realisation + selected planned infrastructure Scenario 4: 2020; current BCP, infrastructure currently under realisation + selected planned infrastructure Scenario 5: 2030; improved BCP, infrastructure currently under realisation Scenario 6: 2030; improved BCP, infrastructure currently under realisation + planned infrastructure With the focus of a potential Core Network in those countries which currently are not (yet) EU Member States, only time horizon 2030 will be considered within this investigation. This is a simplification which may neglect possible bottlenecks at earlier time horizons, but with view to the limited accuracy of the whole exercise, it seems to be justified. The investigation will be carried out separately for the Western Balkans and for Ukraine and Moldova, furthermore for road and rail, by overlaying the corresponding maps of capacity saturation and the proposed potential Core Network. There are five aspects which shall be taken into account when drawing a conclusion and which might make findings less absolute: 1. The model calculation carried out with TRANSTOOLS may imply considerable systematic inaccuracies. Despite a thorough re-calibration of the model, this experience was made already in the model calculations for the TEN-T Policy Review. 2. The network on which TRANSTOOLS was applied does cover neither the entire SEETO Comprehensive Network nor the networks of Moldova and Ukraine. 3. The available results are mapped for long infrastructure sections, in which traffic flows may vary considerably, due to local effects, in particular in the area of agglomerations. 4. The scaling of the results is done in very great steps, which do not allow distinguishing differences in traffic flows with the necessary accuracy. 5. It seems that transit traffic flows have not been sufficiently taken into consideration. This applies, in particular, to PETC X, for which many experts expect a strong growth of traffic volumes, due to developments in Turkey and beyond. (Even a standard gauge rail connection China – Turkey – Western Europe may arise over years and decades.)

Western Balkans / Road: According to Scenario 5 of WP4, the following potential (proposed) Core Network road sections may be critical (80% – 100% of capacity) or congested (> 100% of capacity): Albania: PETC VIII Tirana – Elbasan – FYROM border; Bosnia and Herzegovina: PETC Vc sections close to Doboj and Zenica; FYROM: PETC VIII Skopje – BG border, Struga/Ohrid – Kicevo – Gostivar; Kosovo: E80 Pristina – SRB border, E65 Pristina – FYROM border, Pristina – Prizren; Montenegro: E80 (coastal road) close to Podgorica, parts of E80 (to Kosovo); Serbia: PETC X Belgrade – Marcovac, Nis – Vranje; 36

Beyond these Core Network sections, WP4 results show potential bottlenecks in other road sections, some of which do not belong to SEETO Comprehensive Network. According to Scenario 6 of WP4, the following potential (proposed) Core Network road sections may be critical (80% – 100% of capacity) or congested (> 100% of capacity): Albania: PETC VIII Tirana - Elbasan; Bosnia and Herzegovina: – FYROM: PETC VIII Skopje – BG border, Kicevo – Gostivar; Kosovo: E80 Pristina – SRB border, E65 Pristina – FYROM border, Pristina – Prizren; Montenegro: – Serbia: – Beyond these Core Network sections, WP4 results show potential bottleneck in other road sections, some of which do not belong to SEETO Comprehensive Network.

Western Balkans / Rail: According to Scenarios 5 and 6 of WP4, no critical or congested sections in the railway network are to be expected. For Scenario 5, the highest rate of saturation (60% - 80%) is shown in the PETC IV/VIII section Sofia – Pernik – Radomir in Bulgaria, for Scenario 6 it is the PETC X section Skopje – Veles in FYROM.

Ukraine and Moldova / Road and Rail: WP4 shows results only for Moldova and the Odessa Oblast in Ukraine. For both Scenarios 5 and 6 of WP4, no critical or congested sections in the road and railway networks are to be expected. The maximum saturation rate shows for Scenario 5 on road, namely between 60% and 80 %, in the sections of PETC IX northeast of Dubasari – UA border (=> Kiew) and E58 east of Tiraspol – UA border (=> Odessa).

Conclusion: Even much easier and faster border crossing procedures, which will boost international traffic flows, will very likely not result in major large-scale capacity problems in the Western Balkans, in particular if all planned projects are implemented. This may however be different, if long-distance transit traffic increases more than proportionally, in particular along PETC X, due to global economic and logistic developments not forecasted in WP4. It is not a specific problem of TRANSTOOLS, but of any traffic model conceived for investigating large area, that it is not appropriate to predict local incidents. This means that nevertheless congestion on both road and rail may be expected in urban agglomerations, such as Belgrade, Skopje, Sarajewo, Tirana, etc.

10 COHERENCE WITH NATIONAL CONSIDERATIONS

The considerations based on the TEN-T methodology are strongly based on a “TOP-down approach. It seems to be necessary to give the countries concerned the possibility of expressing their priorities and views on the proposals of this report. National transport infrastructure planning is largely in line with the SEETO Comprehensive Network and, in particular with the Pan-European Corridors. That means implicitly, that there is a high level of accordance with potential TEN-T planning. However, investments are needed also in other parts of the network, as WP4 calculations indicate. The real problem is not deviations from TEN-T priorities, but the overall slow progress of infrastructure implementation. While on road, motorway sections are being constructed, anyway mainly along the PETC’s, on rail there is a wide-spread lack of maintenance, causing deterioration of track quality and railway operation. This is underlined and illustrated by the following country-by country overview, which is an abstract from WP3 – Task 3.1.5, Activity 4.4.2:

Albania: Some progress was made in the area of potential TEN-T, where the development of road infrastructure continues. Focus was PETC VIII towards FYROM and the North-South-axis of Albania. Transport infrastructure maintenance and expropriation issues remain a challenge. Overall, preparations in this area are not very advanced. Bosnia and Herzegovina: Transport infrastructure in Bosnia and Herzegovina is in great need of upgrades, in line with the indicative extension of the TEN-T to the Western Balkans region. Bosnia and Herzegovina need to focus on the development and prioritisation of projects on the routes defined in the comprehensive network of the South-east Europe Transport Observatory Network (SEETO). Active cooperation in SEETO should be reinforced.

FYROM: Attention should be paid to the development and priorisation of projects on the routes defined in the indicative extension of the TEN-T to the Western Balkans region. Upgrading road and rail links included in the South-East Europe Transport Observatory (SEETO) comprehensive network remains a priority. The government adopted the 2013-17 road investment programme. The construction work, co-financed by IPA funds, on the motorway section located along Corridor X of the South- East Europe Transport Observatory network is progressing. Construction of the rail Corridor VIII towards Bulgaria began in March 2014, financed by a loan from the European Bank for Reconstruction and Development and by the Western Balkans Investment Framework.

Kosovo: Segments 1 to 9 of the Morine-Merdare highway (Route 7) were inaugurated in November. Financing for the last segment up to Merdare has yet to be secured. A € 600 million contract to construct the Pristina-Skopje highway (Route 6) was signed in July. The new terminal at Pristina

airport was inaugurated in October 2013.

Montenegro: As regards the, further steps were taken to start construction on the priority section from Smokovac to Mateševo. Some progress has been made in the area of the potential TEN-T, in particular to continue construction of the Bar-Boljare motorway project, which connects the port of Bar with the Serbian border in the north-east of Montenegro. Significant work still remains to be done on improving road and rail links.

Serbia: Little progress has been made regarding the implementation of the action plan for the construction of Road Belgrade-Bar. There has been limited progress regarding the construction of road corridor X and works on four sections of the E-80 were delayed due to a change of contractor. On rail corridor X, project documentation is being prepared for the Novi Sad- Subotica-Hungarian border and for the section of the Nis bypass. Different loan agreements for rail projects have been concluded with third countries. Several infrastructure projects to improve navigation conditions on the inland waterways network are in progress. Overall, preparations in the field of trans-European networks remain moderately advanced.

List of references: [1] TINA Secretariat (EC): Transport Infrastructure Needs Assessment (TINA) Final Report, Vienna 1999 [2] Louis Berger S.A. (AFD, ECMT): Transport Infrastructure Regional Study (TIRS) Final Report, Issy-les- Moulineaux 2002 [3] COWI (EU): Regional Balkans Infrastructure Study (REBIS) Final Report, Copenhagen 2003 [4] Regulation (EU) No 1315/2013 of the European Parliament and the European Council of 11 December 2013 on Union guidelines for the development of the trans-European transport network and repealing Decision No 661/2010/EU [5] South-East European Transport Observatory (SEETO): Multiannual plans 2006 - 2014 [6] European Commission: Commission Staff Working Document SWD(2013) 542 final “The planning methodology for the trans-European transport network (TEN-T)”, Brussels 2014 [7] Decision No 661/2010/EU of the European Parliament and of the Council of 7 July 2010 on Union guidelines for the development of the trans-European transport network [8] Regulation (EU) No 1316/2013 of the European Parliament and the European Council of 11 December 2013 establishing the Connecting Europe Facility, amending Regulation (EU) No 913/2010 and repealing Regulations (EC) No 680/2007 and (EC) No 67/2010 [9] Regulation (EU) No 913/2010 of the European Parliament and the European Council of 22 September 2010 concerning a European rail network for competitive freight [10] White Paper on Transport “Roadmap to a single European transport area – towards a competitive and resource-efficient transport system”, ISBN 978-92-79-18270-9, European Union 2011 [11] TEN-T Policy Review, Report of the Expert Groups, Expert Group 1, Methodology for TEN-T Planning, European Commission, DG MOVE, June 2010 [12] Adria A Project: Accessibilità e sviluppo per il rilancio dell’area dell’Adriatico interno; http://www.adria-a.eu/

The geographical base of all maps if not indicated differently is the European Commission, DG MOVE, whereas the content added reflects the opinion of the author.