07-06-02 UIRR d-fine ZCCT-Roadmap Study FINAL 29.11.2022

Roadmap to Zero-Carbon Combined Transport 2050 Investment needs to enable decarbonisation of inland freight transportation through zero-carbon combined transport

November 2022, on behalf of

Contents Management Summary

page 3

Introduction

page 4

1.

page 5

Demand for freight transport 2020-2050 1.1 1.2 1.3 1.4 1.5

2.

Development of freight transport Projection for modal split Projection of rail freight volume Development and projections for combined transport Implications for the combined transport sector

Public investments into the rail infrastructure and the potential for capacity improvements page 11 2.1 Capabilities of the existing railway infrastructure 2.2 Projection of capacity increases due to impacts of current legislative developments and regulatory undertakings 2.3 Gap assessment – capacity projection compared to the freight demand growth 2.4 Regulatory changes needed to enable rail freight growth 2.5 Investment assessment for ensuring a resilient rail network

3.

Investment plans for road infrastructure and road capacity utilisation page 18 3.1. European road network utilisation and congestion 3.2 Planned investments into the road infrastructure 3.3 Road infrastructure usage and costs related to combined transport

4.

Investment needs for intermodal assets

page 21

4.1 Capacity and utilisation of current intermodal infrastructure 4.2 Improvements to the terminal infrastructure 4.3 Additional terminal infrastructure and a comprehensive network of intermodal terminals 4.4 Impact on railway rolling stock 4.5 Qualitative gap analysis and investment assessment 5.

Conclusions

page 28

5.1 Projections for freight transport, upcoming capacity bottlenecks, and possible improvements 5.2 Investment assessment for a resilient infrastructure for rail and combined transport sector

Acknowledgments The study was conducted by d-fine in close cooperation with UIRR (International Union for Road -Rail Combined Transport s.c.r.l.). Financing by UIRR supported by ERMEWA, Stena Line, Paneuropa, RailFlow, Modility, 4PL International, Trafikverket, and Künz through the CT4EU Campaign is gratefully acknowledged. In addition, we thank the International Transport Forum (ITF) and Dr. Roland Klüber for discussion and contribution to the study.

2

Management Summary The European Green Deal sets the overarching aim to make Europe the first carbon neutral continent on the planet. To get there, a carbon emission reduction by 55% has been agreed on until 2030 (on the 1990 base), while complete carbon neutrality is pledged by 2050 (the reduction target for the transport sector is 90 %) [1]. The Russian invasion of Ukraine and the ensuing war has added an additional objective beyond decarbonisation: a pronounced aim to boost energy efficiency. The use of energy generated on the continent came on top, as Europe must also reduce its dependency on external sources of fossil energy. The need to boost energy efficiency while reducing emissions equally targets households, economic actors, and governments. The task in transportation, the sector whose share of Europe’s carbon footprint grew significantly in the past years, promises to be particularly challenging. In addition, it is expected that the transformation alongside growing transport demand will not interfere with the smooth functioning of society and the economy. The European Union projects a 27 % increase in freight transport by 2030 and 51 % by 2050 relative to 2020. Various actors offered solutions.

By today, two distinct alternatives seem to remain in the race: fuel cell electric vehicles propulsion based on green hydrogen and zero-carbon combined transport (ZCCT) using electric freight trains, electric transhipment and battery-powered trucks for the first and last mile connections all taking power directly from the electricity grid. The two have been compared in a recent study [2], which found that ZCCT is the substantially more energy efficient of the two. In order to reap the benefits of ZCCT, a consistent and sustained shift from road to rail is needed. The share of rail freight transport is projected to double until 2050, with combined transport filling 70 % of the freight trains. In absolute numbers, this means that the tonne-kilometres in rail freight transport are expected to triple by 2050 and combined transport will grow by 360 % over the coming 30 years. As a result of the increasing transport volumes on rail, the number of freight trains is projected to double.

3

The required capacity on the network can be met with the public investment of € 490 billion into the TEN-T rail infrastructure (of which € 330 billion are already planned and € 160 billion are needed for elimination of bottlenecks) and about € 47 billion dominantly coming from private actors into transhipment terminals and other intermodal assets 1. Additional efforts are needed to implement digitalisation solutions, new operating processes, and an upgraded regulatory framework to enable decarbonisation and energy efficiency in inland freight transportation. The regulatory framework upgrades deemed essential include measures to level the playing field and a robust supportive recast of the Combined Transport Directive.

Reduced demand of truck hours and long -distance truck drivers is relative to the projected demand based on the EU Ref. Scenario without an accelerated shift to rail. The number of drivers is estimated based on truck hours assuming 2000 hours per year and driver. Legislative train schedule for the revision of the TEN“ ” the CT D “ ” A tion is planned to start in Dec 2024, full application is assumed in the process of the rail capacity regulation. Obligation for MS authorities to accept digital freight information according to the eFTI Regulation is foreseen for August 2025, full application and wide use is estimated for mid 2030s. TEN-T standards are planned to be implemented in the core network by 2030, in the comprehensive network by 2050.

1 This investment represents a lower boundary as a dditional investments can become nec-

essary to further drive operational efficiency (e.g., through digitalisation) or avoid negative effects due to the higher infrastructure utilisation (e.g., noise protection measures) .

4

Introduction The European Commission announced ambitious climate goals, aiming to achieve carbon neutrality and a reduction of greenhouse gas emissions of 90 % for the transport sector by 2050 (compared to 1990) [1]. These goals are not comfortably achievable through an organic evolution of mode-specific technologies but require fundamental changes such as a significant shift of transport modes. This study is part of a series highlighting the technical, regulatory and investment requirements to enable sustainable and even ZCCT. It was shown that contemporary door-to-door combined transport (CT) offers a CO 2 savings potential of up to 89 % compared to the Euro 6 dieselpowered trucking alternative. This is due to the fact that CT is significantly more energy efficient (up to 64 %), while it uses electric rail freight that is powered by electricity increasingly generated by renewable means [2]. Thus, modal shift has enormous potential to reduce emissions (25 million tonnes of CO 2 equivalents per percentage point of road transport shifted to rail [3]), and the technologies to enable even completely ZCCT already exist today [4]. The EU Reference Scenario projects an increase in freight transport in Europe of 27 % by 2030 and of 51 % by 2050 compared to 2020 [5] (Table 1, Figure 1).

Together with an accelerating modal shift towards rail, this will pressure the existing rail and transhipment infrastructure. This study analyses the projection of freight transport demand and investigates how the current infrastructures can cope with an increased shift towards sustainable means of transport – rail freight and CT. It is evaluated, to which extent infrastructure capacity can be increased through different intensive and extensive measures and a gap analysis indicates which measures and investments are needed. The study is structured into four chapters. The first chapter models freight transport demand, rail freight, and CT volumes up to 2050 with a focus on EU decarbonisation objectives. Chapter 2 – 4 deal with capacity utilisation and planned improvements for rail and road infrastructure, as well as intermodal-specific technologies, respectively. The findings are summarised and conclusions are drawn in Chapter 5.

50%

4.000

45%

3.500

40% 35% 30% 25% 20% 15% 10% 5%

Freight volume transported in billion gross-tkm

Share of long-distance surface freight (> 300 km)

Figure 1: Projection of transported freight volumes in billion gross -tkm for surface freight, rail freight and CT. Projections based on the EU Reference Scenario for freight transport and own modelling.

3.000 2.500 2.000 1.500 1.000 500

0%

0

2020

2025

2030

2035

2040

2045

2.377

2.750

3.050 3.200

3.350

3.500 3.600

Inland Freight > 300 km 1.688

1.950

2.150

2.300

2.450

2.500 2.600

Rail share in long distance transport

Rail Freight

382

500

650

750

900

1.000

Combined Transport

Combined Transport

174

Inland Freight

370

2050

1.150 800

5

1.

Demand for freight transport 2020-2050 Freight transport demand, especially by rail, will increase continuously in the coming years. Based on the EU Reference Scenario 2020 [5], the objectives of the rail sector [3], and the European Commission’s Roadmap to a Single European Transport Area [6], rail freight transport will double by 2050 compared to 2020. One factor contributing to this is the growth of combined transport (CT), which can be assumed to increase in share from 43 % in 2018 to 70 % of total rail freight transport by 2050. In consequence, the modal shift together with the increasing freight transport demand will result in the tripling of rail freight transport volumes in terms of tonne-kilometres by 2050 (relatively to 2020), while the number of intermodal consignments will grow by 360 % resulting in a growth of transport volumes on CT road legs and more transhipments in terminals.

1.1

Development of freight transport until 2050 The volume of surface 2 freight transport in Europe amounted to 2,377 billion gross-tonne-kilometres in 2020 [1]. Several independent studies 3 modelling the European 4 freight transport market project an increase of more than 50 % for surface freight by 2050. Demand for surface freight transport is projected to increase

EU- Reference Scenario 2020 The Reference Scenario 2020 relies on both carbon price signal extension to road transport and buildings as well as strong intensification of energy and transport policies with emission caps set in line with cost-effective contributions in each sector.

Europe has set itself ambitious climate targets in the Green Deal [7]. With the European Climate Law [8] and the “Fit for 55” policy package [1], the legislative basis has been established to deliver on the climate targets – a 55 % reduction of GHG emissions by 2030 compared to 1990 and climate neutrality by 2050. The EU Reference Scenario 2020 [5] will be used to provide a model for the freight transport development until 2050 5. The modelling is calibrated based on historical data up to 2020 and projected transport volumes for the different modes among other social and economic indicators in steps of five years up to 2050. The EU Reference Scenario can be considered moderate compared to other models for freight transport demand, as it foresees a lower increase in transport volumes. However, it still projects an overall growth of 51 % in total freight transport volumes until 2050 6. The projections presented in this work are a result from the interplay of a number of external impacts that are difficult to forecast and quantify, including demographics, ageing and urbanisation, development and breakthrough of technologies, relocation of industries and new

2 Surface transport refers to the accumulated transport by road, rail, and IWW. 3 These include: the OECD ITF Outlook 2021 [9], Fraunhofer ISI: Reference and Pro Rail Sce-

narios for European Corridors to 2050 [19], Capacity4Rail consortium: Requirements toward the freight system of 2030-2050 [10]. 4 European countries includes to the members of European Economic Area (consisting of EU27 countries plus Iceland, Liechtenstein and Norway) as well as Switzerland and European microstates. The ITF Transport Outlook additionally includes Turkey, Ukraine, Moldova, Serbia, Montenegro in the model for European freight transport. 5 Meanwhile an adapted projection of freight transport that is in line with the Sustainable and Smart Mobility Strategy [12] and the “Fit for 55”-package [1] is available with the EU Green Deal Mix Scenario. However, as this differs only slightly from the EU Reference Scenario in terms of total freight volume and only models up to the year 2030, the EU Reference Scenario is chosen as basis for this work. 6 The projections are based on data collected before the Covid-19 pandemic and the war in Ukraine. Emerging uncertainties are not considered an results may be skewed upwards.

6

production concepts, and the possible regionalisation of trade and the resulting flow of goods [9]. Several developments unique to Europe are caused by the change in type of cargo transported from bulk to manufactured goods. Components and finished products will make up more than half of the freight transport in 2050, while the transport of bulk cargo decreases [9, 10]. This shift will affect the transport techniques due to the decreasing cargo density. Intermodal loading units can hold any type of cargo carried in trucks today, which will subsequently lead to an increase of demand for intermodal transport. OECD – High Ambition scenario The ITF/OECD Transport Outlook Reshape scenario considers a strong set of decarbonisation policies with pro-active measures in response to environmental challenges in the transport sector and in support of the United Nations Sustainable Development Goals (UN SDGs).

Other forecasts for freight demand in Europe include the ITF Transport Outlook of 2021, which already included pandemic effects. The study projects a comparatively slower pace of growth until 2030 followed by a strong boom leading to a 120 % increase in freight volumes by 2050 compared to 2020. Another method for projecting freight transport demand is based on its direct correlation with GDP growth, which has been observed over multiple decades [11]. From the latest GDP projection of the OECD Economic Outlook, a growth estimate for freight transport can be derived . Assuming that this development continues, the GDP projection until 2050 results in a growth in freight demand of 61 % [12]. The modelling and projection in the following chapters will , however, only be based on the EU Reference Scenario. The model was developed with a detailed and accurate representation of the conditions and circumstances in Europe 1 and is moreover in good agreement with GDP projections. Additionally, due to various sociodemographic effects that are particularly relevant for Europe (e.g., ageing of society, immigration policies), the moderate growth during the 2030 – 2050 period is considered more probable.

Table 1: Projections for surface 1 freight and rail transport volumes for 2030 and 2050. Increase relative to the base value of 2020 is given in brackets [5, 18, 21, 19]. 2020

2030

2050

Freight volume in billion tonne-kilometres

2,377

3,000 (27 %)

3,600 (51 %)

Long distance freight >300 km in billion tonne-kilometres7

1,688

2,200 (29 %)

2,600 (54 %)

Rail freight modal share8

16 %

23 %

32 %

Rail

23 %

30 %

44 %

382

650 (69 %)

1,150 (201 %)

Rail freight in million tonnes

1,250

2,000 (61 %)

3,200 (161 %)

Intermodal rail share in total rail freight10

43 %

58 %

70 %

Intermodal rail in billion tonne-kilometres

174

370 (114 %)

800 (361 %)

’ long-distance modal share9

Rail freight in billion tonne-kilometres

7 Due to the non-distinction between short and long-distance freight transport in the EU

Reference Scenario, the projections are based on the assumptions that the share of long and short-distance road transport would stay constant in the baseline scenario and that the stronger shift to rail affects on ly long-distance road freight transport. 8 The modelling of the share of rail transport is based on various recent European targets explained in Chapter 1.2. 9 Long-distance freight transport refers to transport distances above 300 km. 10 The modelling of the CT share is based on projections for transport volumes of different cargo types explained in Chapter 1.4.

7

1.2

Projection for modal split The European Commission aims to achieve a more environmentfriendly transport sector and backed this goal with a wide range of farreaching regulatory measures. One of the Commission’s objectives set out in the Sustainable and Smart Mobility Strategy, is to enable a fundamental change in modal share with setting a milestone of doubling rail freight until 2050 [13] – a goal that will already be achieved when projecting the rail freight transport volumes of the policy scenario for the European Green Deal Mix Scenario [14] to 2050 using the growth rates of the EU Reference Scenario. The rail freight sector itself has set a more far-reaching goal. The members of the Rail Freight Forward 11 initiative committed themselves to an increase of rail modal share from 18 % in 2015 to 30 % by 2030. In this context, they addressed that more multimodal solutions and the required rail infrastructure are needed to achieve their goal [3]. The aforementioned ITF Transport Outlook also exceeds the goal of the Sustainable and Smart Mobility Strategy. This model projects the rail share in freight transport to increase to 23 % by 2030 and almost 30 % by 2050. Together with the very optimistic development of the overall transport market, this represents significantly more than a doubling of rail freight traffic by 2050.

The study by the Capacity4Rail consortium 12 analysed that a rail share of 31 – 36 % would be necessary to achieve the European Commission’s climate targets by 2050 [10]. Based on these aspects, a doubling of the rail share in freight transport from 16 % in 2020 to 32 % in 2050 is assumed for the modelling in this study, which is in line with the EU's climate targets [6] and still remains on the lower end of the rail freight share needed to achieve the climate objectives. This corresponds to a yearly increase in rail freight’s market share by 0.5 percentage points and a 5 % annual growth rate for Combined Transport. The analysis of long-distance transport (>300 km) 13 provides a realistic picture of the types of freight movements that are relevant for a shift to rail. In road freight transport, transhipments on long-distances are currently responsible for 60 % of transport [15]. Already today, rail accounts for 23 % of long-distance freight transport. It can be assumed that the change in the modal split toward rail is almost exclusively driven by the shift of long-distance transport from road to rail. For 2050, the modelling of this study results in the following modal split for long-distance transport: 44 % rail, 41 % road, and 15 % transport on inland waterways (IWW).

The “Transport 2050” plan of the European Commission published in 2011 called for 30% of road freight over 300km to be shifted to other modes like rail and waterborne transport by 2030, while 50% should be achieved until 2050 [16]. The 44% rail freight market share projected in this study – together with the inland waterway performance, promises to deliver the 50% objective.

11 Rail Freight Forward is an association of European rail freight companies, representing

90 % of the European rail freight market. 12 Capacity for Rail (C4R) is a consortium that features rail freight companies, academies, and universities. 13 Transports over 300 km are classified as long-distance freight transports. Eurostat data [14] on road freight transports by distance class is used and its distribution is assumed to be constant until 2050 (60 % long distance). In addition, it is assumed that every form of rail and IWW transport can be categorised as long -distance transport.

8

1.3

Projection of rail freight volume A strong increase of rail freight volume arises from the general objective of increasing the market share of rail. A tripling of rail freight tonne-kilometres is projected until 2050 A 70 % increase of rail transport volumes in tonne-kilometres is projected until 2030, while a tripling is anticipated until 2050 (relative to 2020, see Table 1). The projections for transport volumes can also be translated into further characteristics that are relevant for estimating the capacity demand for rail infrastructure and terminals. In 2021, the average distance for rail freight transport was 307 km 14 in Europe [17]. For the last decade, an annual increase in distance of 0.5 % can be observed [18]. Using this as a starting point, the average transport distance should develop towards 355 km in 2050. On this basis, the projection yields 2 billion tonnes of freight transported by rail in 2030 – an increase of 61 % compared to 2020. Until 2050, an increase by 160 % to 3.2 billion tonnes can be expected (Table 1). Rail infrastructure for an additional 1,2 billion train-kilometres is needed until 2050 On average, today’s freight trains carry a net payload of 555 gross tonnes [19, 18] 15, which result in 2.2 million freight train journeys in 2020 (Table 2).

The average cargo volume carried reflects the current composition of rail freight of conventional bulk cargo and containerised intermodal loads and the circumstance that conventional wagons typically travel empty nearly 50% of the time [20]. CT trains on average carry a payload of 658 tonnes per journey [19], the modelled increase of CT (see Chapter 1.4 for more details) is used to obtain a projection for the average payload per train. Building on this, the number of trains and train-kilometres have been derived. For 2050, the projection yields 2.3 billion train-kilometres in freight transport corresponding to 14.7 thousand trains per day 16 – an increase of 140 % compared to 2020. Assuming an optimal utilisation of 740 m train length, the number of trains required is lowered to 13,800.

Table 2: Impact of the increase in freight volumes in rail and CT on infrastructure utilization. Increase relative to the base value of 2020 is given in brackets [1, 18, 21, 19]. 2020

2030

Rail freight traffic in number of trains per day

6,100

9,500 (+54 %)

14,700 (+140 %)

Rail freight traffic in million train-kilometres

700

1,100 (+62 %)

1,900 (+177 %)

Transported intermodal loading units (ILU) in millions

21.1

43.0 (+104 %)

84.5 (+301 %)

27

56 (+109 %)

120 (+350 %)

Freight on CT road-legs in billion tonne-kilometres

2050

14 The statistics on freight train journeys also include short distances, e.g., for shunting

operations. Additionally, international trains may be counted as separate trains with a shorter distance in case of a change of the train number when crossing the border. 15 The average gross payload per train is derived from statistical data for transported freight in tkm and reported train-km (Eurostat data for 2019) [17]. 16 No improvements in payload per train have been assumed for the projection of train numbers. Possible improvements are discussed in Chapter 2 of this study.

9

1.4

Development and projections for combined transport Currently a change in the types of cargo carried in freight trains is ongoing. Volumes of manufactured goods are growing twice as fast as conventional bulk cargo [9, 22]. The goods in the fast-growing segments are typically transported in intermodal loading units (ILU). Intermodal rail significantly contributes to rail freight transport In 2018, intermodal rail transport accounted for 43 % 17 of rail freight tonne-kilometres [23] and with the accelerated growth in demand for CT, the share of intermodal rail within total rail freight is expected to increase further [9]. By 2030, intermodal rail will be responsible for up to 58% of rail freight tonne-kilometres (see Table 1). This forecast, based on historic trends, is in good agreement with the projection of CT’s market share in the ITF Transport Outlook 18. Considering the projected demand for CT, the share of intermodal rail is expected to reach 70 % of total rail freight by 2050. This figure is used for the modelling within the scope of this study.

1.5

Implications for the combined transport sector With this increase in CT share, intermodal rail transport volumes are expected to more than double by 2030 (Table 1) and reach 800 billion tonne-kilometres in 2050, translating to an increase of 360 % compared to 2020. This growth of CT will impact multiple areas of the intermodal sector. Along with the number of trains, transhipment capacities and CT road legs around the transhipment terminals will expand dynamically. Terminal capacity for transhipping 84.5 million intermodal consignments is needed The required transhipment 19 capacity for CT is derived from the freight projections using an average weight per ILU 20 and an average cargo density. By 2030, transhipment capacity for 43 million consignments 21 will become necessary (Table 2). By 2050, the demand is expected to rise to 84.5 million consignments, an increase of 300 % compared to 2020. Transport volumes on CT road legs are projected to triple With the growth of CT, the demand for first and last-mile road legs within the catchment area of the terminals will also increase. An average road leg of 70 km 22 is assumed when modelling the anticipated growth in CT road legs. This results in a demand for the use of road infrastructure of 55 billion tonne-kilometres for 2030 and 120 billion tonne-kilometres for 2050 (Table 2). On the other hand, 420 billion

17 Fraction of CT volume of total rail freight (in tkm). 18 To determine the growth of CT share, the reported rail cargo types are classified as typi-

cally containerized or non -containerized goods. 19 The terms handling and processing are used interchangeably. 20 The average weight per TEU is derived from historical Eurostat data for rail freight transport in containers for the period from 2010 to 2020. The resulting average of 13.8 t/TEU is assumed to stay constant over time. 21 The expected numbers of processed consignments in CT are based on the assumption of two transhipments in a typical CT transport chain . 22 In unaccompanied CT, road legs are estimated to equate to 10 – 15 % [18] of the rail leg distance, which is 900 km for international CT within the EU [23]. Regarding UIRR members, 80 % of their services are unaccompanied CT. Distances in rail transport and CT rail legs are assumed to develop at the same rate as between 2010 and 2020 (annual growth of 0.5 %).

10

tonne-kilometres of long-distance trucking is projected to be eliminated by 2050 as a consequence of the increased shift to rail. When subtracting the additional 40 billion tonne-kilometres on CT road legs due to this shift, a net impact on the road infrastructure of minus 380 billion tonne-kilometres can be found (see Chapter 3.3 for details) . The growth of transport demand challenges the rail infrastructure The increasing transport demand will pressure the existing rail and terminal infrastructure. The following chapters analyses to what extend the current and planned infrastructure and regulation are sufficient to accommodate the projected increases. In addition, it is derived which regulatory changes and investments are required to prepare the transport system in Europe for the challenges of the future.

11

2.

Public investments into the rail infrastructure and the potential for capacity improvements The rail sector can meet the challenges of increasing freight volumes with a relatively limited quantity of extensive expansion of infrastructure 23. The implementation of various capacity improvements, most of which is already planned in the TEN-T programme, have the potential to unlock substantial additional infrastructure capacities. The required capacities can be created in combination with regulatory changes favourable to freight transport – underpinning that the rail freight sector is able to handle a rail freight modal share of 32 % (44% in terms of long-distance transport) in 2050.

2.1

Capabilities of the existing railway infrastructure In 2020, approximately 2.2 million freight train journeys were performed on the European rail network (see Chapter 1.3), which corresponds to 690 million train-kilometres. Freight transport makes up 20 % of the traffic on the rail network Rail freight accounts for only 1 in 5 train-kilometres on the rail network in Europe [24]. In 2019, 18.2 thousand train-kilometres were recorded per line kilometre, with only 3.7 thousand train-kilometres attributable to freight trains. The level of utilisation varies greatly between countries 24: The Netherlands (50 thousand train-km per line-km), Austria (31.4) or Denmark (30) recorded much higher utilisation rates than less densely populated Member States like Romania (6.6) or Greece (4.8) [24].

Revised TEN-T Goals Among others, the TEN-T Goals include the finalisation of all 10 core TEN-T rail freight corridors, facilitation of use of 740-metre-long trains, P400 loading gauge and 22,5t axle load, as well as full electrification of the rail network and the implementation of the European Rail Traffic Management System (ERTMS).

The utilisation of the existing rail network is already relatively high in several countries, and it must be assumed that existing capacities are extensively utilised under the current path allocation and traffic management scheme. Especially countries on the busiest rail freight corridor (Rhine-Alpine RFC), record utilisation rates above average (20,000 50,000 train-km per line-km). Data on trains crossing at least one border shows, that high utilisation rates are present ( Table 3). In 2022, 677,600 border crossings of trains were recorded 25. Different indicators show that rail freight transport under the current path allocation system is close to its capacity limits In rail transport, punctuality and infrastructure reserves are closely connected. When approaching the limit of capacity, actual transportation time becomes unstable [25] and punctuality declines. The European Commission has set a punctuality target within its proposal for a revised TEN-T Guidelines Regulation [26]. According to which the average dwell time of freight trains crossing an internal EU border should not exceed 15 minutes by 2030, while at least 90 % of freight trains should arrive at their destination with a delay of less than 30 minutes [27].

23 Apart from construction to upgrade to TEN-T standards, the required capacity can be

achieved without extensive construction through reallocation of passenger train paths. 24 Differences also stem from geographical position and state of the infrastructure. 25 Trains crossing borders can be used as an estimate for the number of trains and, thus, for line utilization. The numbers might be biased by double counting when using multiple corridors as well as by non-recording of domestic transport. Number of border crossings is not only influenced by utilization but also by the number of countries and respective borders on the corridor.

12

Table 3: Number of trains crossing at least one border on a TEN-T corridor in 2021 [67]. Corridor Name

Trains crossing a border

Corridor Name

Trains crossing a border

Rhine-Alpine

97,765

Orient-East Med

123,643

North Sea-Mediterranean

25,194

North Sea-Baltic

87,484

Scandinavian-Mediterranean

46,743

Rhine-Danube

102,970

Atlantic

14,931

Amber Rail Freight

30,639

Baltic Adriatic

88,618

Alpine-Western Balkan

16,404

Mediterranean

29,848

Czech-Slovak

13,409

Total

677,600

These targets are missed by far today. The present punctuality rate (arrival delayed by less than 30 minutes) in Europe reaches 60 % for domestic freight trains and 53 % for international freight trains – which means that only every second freight train arrives on time 26. The congestion rate is another indicator showing that the capacity limits are currently being reached 27. The share of railway lines classified as congested increased from 1040 km in 2015 to 2261 km in 2018 – the congested line kilometres have thus more than doubled in only 3 years [24]. The example of Germany shows that several important routes for freight transport are overloaded or have even been declared congested (Figure 2) – especially seaport hinterland routes. These routes need capacity enhancements to cope with the current and future increase in rail freight traffic instead of preventing the increase of rail share

2.2

Projection of capacity increases due to impacts of current legislative developments and regulatory undertakings The European Commission announced the goal to complete the TEN-T core network by 2030 and the comprehensive network covering all European regions by 2040. Among construction and the achievement of standards for infrastructure, the policy objectives focus on application of new technologies and digital solutions [28]. Archiving the goals set in the TEN-T Guidelines Regulation (EU REG 1315/2013) in the core network by 2030 and the comprehensive network by 2050 is estimated to require investments at about € 1.5 trillion for all modes. Current investment plans show that 78.8 % of the TEN-T funding are used for rail projects [77]. The European Commission proposal for a revised TEN T Guidelines Regulation expands the policy objective of the TEN-T regulation. Among others, the core network standards are extended to the comprehensive network which is also expanded by additional sections with need for upgrade. Pledged investments have to be made in time The analysis accompanying the impact assessment for the revision proposal of the TEN-T Guidelines Regulation concluded, that the original investment goals and regulatory measures are not sufficient for

26 Additionally, approx. 11 % of international freight services are cance lled. 27 The threshold for declaration of routes as congested varies for different cou ntries. In

general, a route is declared congested according to Article 47.1 of Directive 2012/34/EU when it is not possible to satisfy the requests after coordination and consultation [65].

13

achieving the Sustainable and Smart Mobility Strategy objectives [29]. The assessment focused on the removal of bottlenecks and estimates additional costs of € 16.4 billion per year, which represents an increase of 33 % of annual investments [27]. The transformation of the rail infrastructure to comply with the TEN-T technical standards on the comprehensive network is a major cost factor accounting for 78 % of the additional costs. However, these investments include measures that are specific to passenger transport such as the introduction of a minimum line speed of 160 km/h for the passenger core network, which is estimated to contribute 31 % (€ 5 billion per year) of additional spending. Figure 2: Map of the capacity utilisation on the German rail network. Congested segments are marked in red. Tracks with a load factor of over 125 % are shown in purple, with 115 – 125 % in orange and with 108 – 115 % in blue. Other tracks of the network are shown in grey [66].

Implementation of the TEN-T standards has the potential to significantly increase infrastructure capacity The existing railway infrastructure still offers capacity reserves, which can be exploited through infrastructure management techniques within the framework of TEN-T by 2050 (Table 4). Improving the signalling system and the traffic management system will increase line capacity. The introduction of the European Rail Traffic Management System (ERTMS) has the potential to increase capacity by up to 25 % [30]. Combined with reduced signalling block lengths, the overall capacity impact can be increased to 37 % [31]. Improving traffic management to enable a steady speed of 100 km/h for freight trains offers the possibility of increasing line capacity by up to 66 % [32]. The introduction of P400 provides the opportunity to use high-capacity intermodal loading units, like high cube pallet-wide European containers or semi-trailers [33]. P400 will allow transportation of these units on the entire TEN-T network, enabling rail operators to fill their trains more efficiently, as they will enable higher transport volumes for low density cargo [32]. These effects will have a measurable capacity impact and will increase the ability of the network to accommodate the projected modal shift.

14

Table 4: Possible measures to increase capacity within the existing rail infrastructure. Description

Current situation

Potential for improvement

European rail traffic management system (ERTMS) [27, 30, 32, 34, 35, 3] ERTMS is the system of standards for management and interoperation of signalling for railways. It consists of two main subsystems, namely the European Train Control System (ETCS) and the Global System for Mobile Communication for Railways (GSM-R). Three different levels28 of ERTMS application exist.

▪ ▪ ▪ ▪

52 % of entire TEN-T net▪ work29 needs to be upgraded 46 % of all new trains are already equipped 0.17 million /km (for L-228 ) ▪ Cost for the TEN-T network in total: € 30 billion

Line capacity increase of 5 % 30-15 % with ERTMS L-2 on already optimised block sections 25 % increase in line capacity with L-231

52 % of the TEN-T network needs to be upgraded, 75 % of all upgrades in Spain, France and Italy Cost for the TEN-T network in total: € 5.1 billion

▪

Increased ability to shift long-distance road transport to rail, due to the better usability of high-capacity loading units

34 % of the total TEN-T network needs to be upgraded Cost for the TEN-T network in total: € 8.9 billion

▪

Maximum line capacity increases to 20 trains per hour 66 % increase of maximum trains per line33

89 % of TEN-T core network electrified 0.7 – 1.2 million/km Cost for the TEN-T network in total: € 92.4 billion

▪

In the report “ ” the international energy agency assessed that electrified lines carry twice as many tonne-kilometres compared to non-electric ones

▪

Full exploitation of 740 mlong trains without improved load factor35: 13 % increase Maximum capacity utilization of a 740 m-long train at 2,000 t36 : 60 % increase.

P400 gauge [23, 33, 36] The P400 reference is the used standard of ▪ measurement for semi-trailers loaded on a pocket wagon. It allows for a maximum height of 4 metres, at which a semi-trailer can be transported by rail32. Currently, semi-trailers ▪ account for 25 % of CT. 100 km/h average speed [27, 32] Currently, the average speed of freight trains is about 55 km/h. With a block length of 3,000 metres, the maximum throughput through a line section is 12 trains per hour.

▪ ▪

▪

Electrification [24, 37, 38] Although largely electrified, the rail network is interspersed with sections lacking electrification. Eliminating these would reduce the need for change of locomotives and increase capacity.34

▪ ▪ ▪

Optimal usage of maximum train weight and length [27, 20, 2] Currently, trains transport on average about 555 tonnes of payload. In a fully developed TEN-T network, the potential of long trains with a length of 740 m and a train weight of 2,000 t (22.5 t axle load), can be better be exploited.

▪ ▪

34 % of the total TEN-T network needs to be upgraded Cost for the TEN-T network in total (740-m-long trains): € 0.8 billion; (22.5 t axle ): € 150 billion

▪

28 L-1 is designed as an add-on for conventional lines equipped with line -side signals and train detectors. In contrast, L -2 does

not require line-side signals. The movement authority communication occurs directly to the on -board unit using a GSM-R. Continuous communication in L-2 allows the train to reach its optimum or maximum speed while maintaining a safe braking distance. The conceptual stage L-3 is based on moving block technology instead of classic operation in fixed intervals to continuously allow the train to monitor its own position and to achieve continuous line -clear authorisation. 29 Entire network refers to the combined TEN -T core and comprehensive network. 30 For double track lines. 31 Combined with shorter block length the impact of ERTMS L -2 increases to 37 %, implementing ERTMS L-3 with floating blocks offers capacity increase of 42 % – 50 % [31, 30]. 32 This limit is set from the base of the wagon to the top of the semi -trailer. 33 Higher speed can also be the result of block signalling due to ERTMS, which reduces the individual effects. 34 Interoperability issues can occur between different voltages used on national networks. Adjustment of traction of international trains to voltages among 750V, 1.5 kV, 3 kV, 15 kV, and 25 kV is necessary. 35 Modelled 740 m-long train under the assumption that empty trip factors and the share of semi -trailers and containers [18] stays constant for the period up to 2050.

Capacity increases stem from the higher payload of this average 740 m-long train (679 t) compared to the average projected payload (603 t).

15

Exploiting a maximum allowed train length of 740m and the maximum gross weight of trains at 2,000 t holds further potential. Compared to the current average payload per train, optimal utilisation would increase the payload by up to 60 % 36. However, beforehand infrastructural preconditions (740m train length, 22,5t axle load) for the use of optimal trains have to be established within the framework of TEN-T. More efficient timetabling is needed to enable better capacity utilisation The current timetabling system lacks efficient path allocation 37, since almost 80 % of freight train path requests have to be repeatedly modified [39]. A major redesign of the current timetabling regime is under way with the Timetable Redesign Project (TTR). The rail sector expects the TTR to increase flexibility and deliver higher quality train paths. Additionally, measures are planned to be implemented to increase the efficiency and reliability of timetables [40]. The investment including also IT-infrastructure accumulates to € 1 billion while the potential benefit is estimated to € 1.5 to € 2 billion per year [41, 39]. RNE reports, that through the measures of TTR, capacity gains of 15 % are possible [41].

2.3

Gap assessment – capacity projection compared to the freight demand growth In combination, the capacity of the existing railway infrastructure can potentially be improved to a significant extent through measures allowing for a higher number of trains running per hour , as well as enabling higher payloads and volume utilisation of trains (Table 5). It is highly advisable to initiate these measures in order to bridge the gap between infrastructural capabilities and projected freight volumes. Most capacity improvements are interdependent, which could reduce the total combined effect of the named measures. To compensate for reduced effects, additional solutions that include extensive work on the rail network and regulatory changes are discussed in the following. Socio-economic cost-benefit analysis could serve as a good basis for prioritising investments in the railway infrastructure. The proposal for the revision of the TEN-T Guidelines Regulation contains this tool.

2.4

Regulatory changes needed to enable rail freight growth The proposal for the revision of the TEN-T Guidelines Regulation is a first step in making the rail sector ready for the objectives set in the European Union. Additional measures together with regulatory changes are needed to provide a resilient rail network in 2050. The regulatory framework upgrades deemed essential include measures to level the playing field and a robust supportive recast of the Combined Transport Directive.

36 The optimally utilised CT-train (740 m, 2,000 t, no empty wagons) considers the average

share of semi-trailers and containers [18]. The resulting increase in payload is relative to the projected payload for 2050 considering the development of CT share 37 An important issue is that railway undertakings are currently required to book train paths 8 to 20 months in advance, which leads to leads to overbooking to obtain sufficient flexibility. In this way, bookings often have to be adjusted and the remaining ca pacity is needed for last-minute changes [39].

16

Table 5: Measures for capacity improvement in rail transport (in number of trains) and assessment o f cost. The improvements are expressed in percentages compared to non -TEN-T compliant infrastructure. EU Reference Scenario and own calculation

Number of trains per year (in thousand) 2020

6,100 trains per day

2030

9,500 trains per day (54 % increase relative to 2020)

2050

14,700 trains per day [13,800 if 740m overall possible] (140 % increase relative to 2020) Increase in capacity

Calculated cost (in €)

Status of implementation

P400 gauge

---

5

48 %

740m avg. trains

13 %

1

43 %

Optimal capacity usage (length and weight)

41 %

150

81 %

ERTMS (ETCS level 2)

25 %

30

48 %

Increased avg. speed (100 km/h)

66 %

49

66 %

Electrification

---

92

80 %

Timetable redesign (TTR)

15 %

1

---

<150 %

328

---

Capacity improvement Improvements for train capacity

Improvements for track capacity

Gap assessment for 2050 Cumulated capacity improvements 38

Upgrading single track lines can reduce bottlenecks, especially when routes are blocked Expanding bypass tracks and establishing double tracks – as advised by TEN-T [28] – can help to eliminate bottlenecks. As of today, 128 thousand kilometres (60 %) 39 of the TEN-T network were still single tracks [18]. Upgrading these to double tracks increases the maximum number of trains from 3 to 20 trains per hour with similar timetable speed or to 10 trains per hour on lines with heterogeneous traffic 40 [32]. Upgrading single tracks and expanding sidings can thus offer additional doubledigit percentage improvements in capacity. This option becomes particularly important when lines become blocked, as it allows overtaking instead of a total stop of operations. Total costs for upgrading single tracks to double tracks are estimated to € 6.2 million per kilometre [35]. Consequentially, upgrading the TEN-T network to comply with this standard would require a total investment of € 795 billion. However, since freight transport is only responsible for only 20 % of infrastructure utilisation, only 20 % of the total costs

38 The cumulative capacity gain represents an upper limit under the assumption that no

correlation exists between the individual measures and that capacity increases at train level lead to a reduction in the number of trains, and to the same extent reduce the utilisation of lines. 39 Germany for 2021 due to missing data for 2020 . 40 Line capacity is lower on mixed tracks used by slow trains (freight, regional) and express trains due to missing options for overtaking [31]. This bottleneck can be addressed by extension of additional tracks for overtaking.

17

should be assigned to the freight sector 41. Resulting in an investment need related to freight transport of € 160 billion until 2050. Redesignating some passenger trains paths to freight can help to close the capacity gap The capacity gains of the various measures discussed above cannot be expected to add up. Several actions will only yield an improvement but not the optimal state (e.g., average speed of 100 km/h for freight trains, exploitation of maximum length and weight). In case only a 60 % overall capacity gain could be achieved by 2050, a gap to the needed 80 % would remain.

Further selected instruments allowing for higher capacity utilisation

▪

Joining two 740-meter-long trains, with two locomotives, to pass through congested lines as a single 1480 m train require only 20 % more network capacity while carrying twice the payload on a single train path [64].

▪

Alignment of voltage and frequency for traction along with harmonisation of multi-system locomotives can help to reduce delays when crossing borders by not requiring a change of locomotives. Costs per kilometre for electrification of existing lines are estimated at € 0.6 million, changing existing electrification could thus account for similar costs [35].

One option to close this gap, could be to employ a socio-economic costbenefit analysis to reallocate certain train paths from low utilisation passenger trains to freight transport. Considering the 80:20 ratio for the use of the rail network, each percentage point of passenger train kilometres reallocated for freight transport would result in a capacity increase for freight by 4 percentage points 42. Dedicated freight lines and extended bypass options allow to double capacity in highly congested areas Speed differences of passenger and freight trains on mixed lines are one of the largest factors reducing track capacity for rail freight on the infrastructure today [22]. Freight trains are often required to stop to let passenger trains overtake. Dedicated freight tracks along major corridors and bypass routes around major urban cities could help solving this problem, as this allows up to doubling the number of circulating trains [10]. Construction costs for new lines are assessed at approx. € 8.2 million per kilometre [35]. Nevertheless, when considering extensive infrastructure expansion, e.g., for dedicated freight lines, the long construction time for new rail tracks [42] should be taken into account.

41 Note that tracks for single usage by freight as well as passenger transport exists. How-

ever, a huge majority of lines are for mixed use [42]. 42 If this measure was taken, the costs of line upgrades attributed to freight transport would also need to be increased by € 7.95 billion.

18

2.5

Investment assessment for ensuring a resilient rail network The rail infrastructure can meet the challenge of tripled freight volumes through the measures set out in the Commission’s proposal for the revision of the TEN-T Guidelines Regulation [26] and additional upgrading measures discussed in this chapter. Additional investments of € 490 billion are needed to achieve full compliance with the proposal for a revised TEN-T Guidelines Regulation and provide the infrastructure to enable a modal shift to 32% rail (44% in terms of long-distance transport above 300 km). This implies that additional investment of € 17.4 billion per year into the rail network is required, with is about 40% over today’s levels. The largest expenses relate to upgrading the existing infrastructure to allow more efficient trains (€ 151 billion), electrification of the current infrastructure (€ 92 billion), and the development of a more resilient rail network by upgrading single tracks to double tracks (€ 160 billion). The next chapter will therefore investigate the current investment plans into road infrastructure and analyse potential savings that result from the shift towards rail.

19

3.

Investment plans for road infrastructure and road capacity utilisation The proposal for the revision of the TEN T Guidelines Regulation is based on projections for freight transport that assume a rail share in freight transport of 21 % for 2050. It has already been shown that the rail infrastructure would be capable of handling a rail share of 32 %, if the foreseen TEN-T measures are delivered. Through an increase in CT, a substantial portion of the cargo that is presently carried in tr ucks can be shifted to rail, resulting in a significant reduction of the projected transport volume on road by 420 billion tonne-kilometres compared to the TEN-T base scenario in 2050. At the same time, transport volumes related to CT road-legs will increase to 120 billion tonne-kilometres.

3.1

European road network utilisation and congestion Based on the 2019 TEN-T Performance Report [43], the road network consists of motorways (85 %) with more than two lanes (77 %) that connect urban areas (91 %). Annual average daily traffic flow (AADT 43) on most of the network (52 %) is less than 20,000 vehicles per day, while 6 % of the network have an AADT of more than 80,000 vehicles per day. A major part consists of passenger car traffic. About 30 % of road utilisation in the core network stems from heavy goods traffic 44. This share has remained constant over the last 10 years. Figure 3: Map of the AADT on the European road network. Roads utilised by more than 100 thousand vehicles per days are marked in purple, 80-100 thousand in red, 40-80 thousand in orange, 20-40 thousand in yellow, and under 20 thousand in green. Roads without data are presented in grey [43]. © O pe n St re e tMa p c on t ri b ut o rs , Cre d it : E C -GI SC O, © E u ro Ge o g ra p hi cs f o r t he a d m ini s t ra ti ve bo u nd a r ie s | © E ur o pe a n Co m m is s i on - DG MO VE – 2 018

For two-lane motorway sections outside of urban areas, a vehicle throughput of 18,000 vehicles per day was identified as a threshold for free traffic flow [44]; motorways with 2x3 lanes help to increase the average capacity to 31,000-52,000 vehicles per day [45]. This high AADT limits imply that nearly 50 % of traffic could already be handled by two-lane motorways, while 2x3 lane motorways are able to compensate for higher traffic flows in areas with higher utilisation. Consequently, free traffic flow should be possible in terms of capacity utilisation in most cases. Also, the average daily traffic flow shown in Figure 3 exceeds average utilisation only exceptionally. However, most congestion is related to traffic patterns and traffic-jam influencing events, such as

43 AADT is calculated as total volume of vehicle traffic for one year divided by 365 days. 44 Goods vehicles weighing in excess of 3.5 tonnes.

20

traffic incidents, construction work, weather [46], peak hours in urbanised areas [47, 48], or seasonal variability [48] that raise the vehicle flow or limit the capacity in the short term 45.

3.2

Planned investments into the road infrastructure In contrast to the implementation status of the TEN-T railway lines (Table 5), the completion rate of the TEN-T road network ranges from 70 % to almost 100 %. This high level was achieved thanks to € 78 billion of funding from the European Union 46. One third was allocated to projects outside of the TEN-T core network 47 [49], € 17.8 billion was attributed to construction of new roads (2,000 km) of which 50 % are outside the core network. Presently, the reconstruction of 770 km of roads and capacity improvements for the TEN-T road sections is planned [43]. Focus on the maintenance of existing road infrastructure is required With TEN-T, the length and capacity of roads in every EU Member State increased. At the same time national maintenance budgets have steadily declined [49]. In their 2020 report, the European Court of Auditors states that the Member States do not devote adequate resources to the required maintenance and upkeep, putting the condition of the core road network at risk in the medium- to long-term [49]. The TEN-T revision concluded that improvements are necessary, mainly to reduce the number of accidents due to the insufficient safety and operational reliability of the TEN-T road infrastructure.

Safety-related upgrades are needed on a length of 4,477 km (7.2 % of the comprehensive network) at an annual cost of € 693 million between 2025 and 2040 48 (€ 10.4 billion in total) to reduce road traffic fatalities. The assessment concluded that the investments would help to decrease the fatality rate by 1.2 % compared to a scenario assuming no further upgrades beyond those planned within the original TEN-T policy [27, 29]

3.3

Road infrastructure usage and costs related to combined transport Road freight transport associated with CT road legs will reach 120 billion tonne-kilometres in 2050. Of these, 40 billion tonne-kilometres are directly related to the modal shift towards rail exceeding the growth projected by the EU Reference Scenario. At the same time, due to this shift long-distance road freight transport will be reduced by 420 billion tonne-kilometres per year in and after 2050. This transport volume corresponds to 30 % of total long-distance road transport projected in 2050. Translated into utilization of trucks, this implies that at least 450 million truck hours less per year are required compared to the EU Reference Scenario 49. On the other hand, the increase in road haulage due to first and last leg in CT can be assumed to account for about 100 million truck hours 50. Because of the lower speed and additional

45 Short term congestion is estimated to account for a loss of 1 % of GDP [49]. 46 Besides grants, the EU supports development of the road network through financial

instruments such as loans and guarantees to attract private investments. 47 Comprehensive TEN-T network as well as non-TEN-T roads. 48 Funding for investments into road infrastructure shall be secured through EU funds (55 %), the private sector (5 %), toll revenues (15 %) and further national funds (25 %). 49 Calculation of truck hours assuming average parameters for long-haul trucking: weight per journey of 14.33 t [2], speed of 65 km/h [51]. 50 Calculation of truck hours assuming average parameters for CT: weight per journey of 15.85 t [2], speed of 25 km/h [51].

21

factors like, e.g., dwell time in terminals and a higher frequency of loading and unloading operations, the net reduction of road transport volumes does not lead to a proportional reduction of truck hours. Nevertheless, the shift to rail coupled with the parallel growth of CT can achieve a net reduction in truck demand of 350 million truck hours. Considering infrastructure utilization costs, the net reduction of transport volumes on road would also translate into less demand for investment into construction, upgrading, and maintenance of road infrastructure [50]. In addition, the regional nature and relatively low speed of trucks operating on the first and last leg in CT, result in a reduced wear and tear on the road infrastructure 51 [51]. Regarding vehicles, the net reduction of truck hours for the modelled higher rail share in long distance freight transport also implies a reduced demand for investment into trucks compared to the modelling of the EU Reference Scenario. While vehicles currently used in long -haul trucking are equally suitable for CT road legs, the introduction of more sustainable subsidies like BEV trucks in fleet renewal and tr ansformation becomes also easier.

The reason for this is, that CT road legs offer good possibilities for integration of charging cycles because of their characteristics (short distances, more frequent stops for loading and unloading). This is not only worthwhile in terms of sustainability, but also in terms of investments. Current projections for the total costs of ownership (TCO) of EURO 6 diesel trucks compared to BEV trucks indicate, that the latter will overtake and be the cheaper option before 2030 in all European countries [4]. The possibility to employ trucks of a regional configuration offers another potential to decrease investment needs due to lower purchase prices by about 10 %. In contrast, the demand for container chassis and loading units will increase. In 2017, only 9.5 % of the registered 240,000 semi-trailers were container chassis [52]. When considering that these travel on average 100,000 km per year 52, it becomes obvious, that not all but about half of them are required to handle the current transport volumes on road legs in rail-road CT [52]. With an increase of CT transport volumes by the factor of up to 3.5, the fleet of container chassis used in CT needs to increase from 12,000 to 60,000 – resulting in the need for private investments of € 480 million 53.

Effectively, there are investment needs in road infrastructure as well, however, these are outweighed in the net perspective as a consequence of the higher share of rail freight compared to the EU reference scenario. Maintenance expenses foreseen for road infrastructure can, thus, be redirected towards other modes. Well maintained road infrastructure is also needed for CT to function. However, the focus is shifted from the main supra-regional or even international transport axes to the regional road network within the proximity of transhipment terminals [43]. The capacity and investment analysis of these is the topic of the next chapter.