WP3: Rail Passenger Transport - TOSCA Project

Technology Opportunities and Strategies toward Climate-friendly trAnsport
FP7-TPT-2008-RTD-1
Coordination and Support Action (Supporting)
Deliverable D4 (WP3 report No 1)
Rail passenger transport
Techno-economic analysis of energy and greenhouse gas reductions
Royal Institute of Technology (KTH), Stockholm, Sweden
2011-03-29
Evert Andersson, Mats Berg, Bo-Lennart Nelldal, Oskar Fröidh
Dissemination level
Public PU X
Restricted to other programme participants (including Commission Services)
Restricted to a group specified by the consortium (including Commission
Services)
Confidential, only for members of the consortium (including Commission
Services)
PP
PE
CO

Coordinator
Dr. Andreas Schaefer
University of Cambridge
Institute for Aviation and the Environment
AIM Group Dept of Architecture
1-5 Scroope Terrace, Cambridge CB2 1PX, UK
Tel: +44-1223-760-129
Fax: +44-341-2434-133
E-mail: as601@cam.ac.uk
Internet: www.toscaproject.org
Contact:
Royal Institute of Technology (KTH)
Dept. of Aeronautical and Vehicle Engineering
Teknikringen 8
SE - 100 44 Stockholm, Sweden
Internet:
www.kth.se/ave
Prof. Evert Andersson
Tel: +46 8 790 7628
Fax: +46 8 790 7629
E-mail: everta@kth.se
Prof. Mats Berg
Tel: +46 8 790 8476
Fax: +46 8 790 7629
E-mail: mabe@kth.se
The TOSCA project (Technology Opportunities and Strategies toward Climate-friendly trAnsport)
aims to identify promising technology and fuel pathways to reduce transportation-related
greenhouse gas emissions through mid century about 2050. An important building block of this
project is the techno-economic specification of low-GHG emission transportation technologies and
other means, which are input into a scenario analysis. TOSCA considers all major modes of
passenger and freight transport, along with transportation fuels and technologies capable of
enhancing infrastructure capacity. This report on rail passenger transport is thus one out of a
number of such techno-economic studies.
Deliverable D4 – WP3
II

CONTENTS
CONTENTS
ABBREVIATIONS AND DEFINITIONS
ABSTRACT 1
1. INTRODUCTION 3
2. REFERENCE SYSTEMS CHARACTERISTICS 4
3 TECHNOLOGY DEVELOPMENT 6
3.1 General 6
3.2 Technology trajectories 7
3.3 Technology descriptions 8
3.4 Incremental improvements and higher speed 12
3.5 Technology scenarios to be further considered and investigated 13
4. CONSTRAINTS FOR REDUCING GHG EMISSIONS 16
5. RESULTS 18
5.1 Technology readiness and R&D requirements 18
5.2 Energy and GHG emissions 19
5.3 Cost for reducing GHG emissions 24
5.4 Scalability 30
5.5 Social acceptability 31
5.6 User acceptability 32
6. SUMMARY AND CONCLUSIONS 34
6.1 Summary 34
- Low-GHG electricity 34
- Eco-driving 35
- Space efficiency 35
- Energy recovery 35
- Low-drag 36
- Low-mass 36
- Incremental improvements 36
- Combination of measures 37
- Higher speeds and combinations 37
- Incentives needed 37
6.2 Conclusions 37
REFERENCES 39
III
IV
Deliverable D4 – WP3
III

ABBREVIATIONS AND DEFINITIONS
Articulated train
Carbody
Train unit where two carbody ends are supported by the same running gear
(wheels, suspension etc).
The part of a rail vehicle, above the running gear, where the payload
and/or the train crew is carried.
CO2-eq
Carbon dioxide (CO2) equivalents of greenhouse gases, so that the global
warming potential of a mix of gases can be directly compared.
DMU Diesel Multiple Unit, i.e. a train unit consisting of self-propelled rail cars
powered by diesel engines.
Dual mode
In this context, a rail vehicle being able o run both in the electric mode (fed
by an external overhead electrical wire) or by diesel engines.
Eco-driving
Driving style with intention to reduce or minimise energy use.
Energy recovery Train technology where the (electric) motors are used as generators while
braking, to send generated energy back to the electric supply system or to
storage on board the train.
Energy use
Net energy intake to the train or to the railway system, after subtraction of
energy recovered when braking.
EMU
Electrical Multiple Unit, i.e. a train unit consisting of self-propelled rail cars
powered by an electric drive system.
EU-12 The newest 12 member states of the EU, joining after year 2000.
EU-15 The first 15 member states of the EU, joining before year 2000.
EU-27 All 27 member states of the EU (2009).
GHG
Greenhouse gas
High-speed train In this context, trains with an admissible top speed of at least 250 km/h.
Incremental improvement Improvements in small steps.
KTH
Kungliga Tekniska Högskolan, the Technical University of Stockholm, also
called the Royal Institute of Technology.
Load factor
Same as ‘seat occupancy rate’, i.e. occupied seat-km divided by the total
number of run seat-km of a train. Load factor is usually determined over a
period of time.
MEUR
Million Euros
pkm
Passenger-km
Railenergy
EU-funded project on energy efficiency for railways. www.railenergy.org
R&D
Research and Development
Sales price
Price to customer, excluding taxes and other duties, with general price
levels and currency rates as in 2009.
Specific energy (or GHG) Energy use (or GHG emissions) per passenger-km.
STEC
Simulation of Train Energy Consumption (software developed by KTH).
Technology readiness In this context, time when a new product or system is developed to a state
where it can be offered for sale.
Top speed
Maximum operational speed
UIC
International Union of Railways.
WP
Work Package (of the TOSCA project).
Deliverable D4 – WP3
IV

ABSTRACT
In Stage 1 of the EU/FP7-funded project TOSCA (Technology Opportunities and Strategies
toward Climate-friendly trAnsport) the techno-economical feasibility of different technologies
and means to reduce greenhouse gas (GHG) emissions is being analysed for the different
modes of transport. This is made in the long-term perspective until 2050, with 2009 as the
reference year. This is the report on rail passenger transport, applicable to the European Union
(EU-27).
The present report has been subject to review among railway experts, representing train
suppliers, railway operators as well as academia. They have also responded to a questionnaire.
Further, a workshop was held, where the report with assumptions and results was discussed.
In the analysis presented in this report it is estimated that a number of efficient improvements
that, individually and in combination, are available in order to significantly reduce energy use
and the resulting GHG emissions on the rail passenger market until 2050. The analysis has
considered different technologies and means:
– low air drag
– low train mass
– energy recovery
– eco-driving, including traffic flow management
– space efficiency in trains (increasing payload per metre of train)
– incremental improvements of energy efficiency, in particular reduced losses.
Despite anticipated higher average train speeds in the future these combined approaches will,
according to the analysis, have the potential to reduce the average specific energy use per
passenger-km (pkm) in the order of 45–50 % in the very long term until 2050. As a consequence
also the direct and indirect GHG emissions will be reduced. The highest reductions are
possible in city and regional rail operations. Reductions are more limited in high-speed operations,
because of the advanced technologies already applied. However, high-speed rail has
today a comparatively low energy use per passenger-km, partly due to its high average load
factor. To be consistent with other work packages of TOSCA, energy use and GHG emissions
are measured per passenger-km, assuming representative load factors in different operations.
Irrespective of reductions in energy use, by far the most effective mean of reducing GHG
emissions from European rail operations is the provision of low-GHG electricity, which
depends on technologies that will be introduced outside the railway sector; the end result being
dependent of the degree of GHG reductions in the future European mix of electricity.
A 60–80 % reduction of GHG content in future average European electric power, is approximately
within the different scenarios of TOSCA WP6. The resulting CO 2 -eq emissions are then
estimated to 4–11 g CO 2 -eq per passenger-km, which is very favourable.
The results might be seen as optimistic, but they are due to the combination of a number of
technologies, which to some extent also reinforce each other. It should also be pointed out that
some options are not analysed in detail at this stage, and not included in the above estimations,
i.e. (1) increased average load factor on trains, (2) modular trains with capacity according to
actual need, (3) dual-mode operations (diesel and electric) and hybrid trains, (4) introduction of
bio fuels in diesel engines and (5) additional electrification of railways or increased investment
in rail infrastructure. A thorough analysis of these options involves many uncertain technical,
economical, operational and societal considerations. They are all out of the scope and time
frame of this study. In particular heavy investment in rail infrastructure above levels of 1995-
2010 would further reduce per-unit operation cost, specific energy use and GHG emissions.
Deliverable D4 – WP3 passenger 1

Magnetic levitation is excluded from further analysis, since it is considered not to be an
efficient mean of reducing energy use and GHG emissions. These systems are not even interoperable
with existing rail systems and have serious cost implications.
Cost and incentives
In most cases the proposed technologies have modest implications for the asset price of trains
and will in the long term pay for themselves through lower cost of maintenance and energy.
However, many train operators and leasing companies are expected to be conservative and
hesitate to introduce new technologies that will result in additional investment and technical
risks. In the rail sector there is currently a tendency to prioritize the first cost (investment cost)
and pay less attention to the long-term life cycle cost and to the income side of the account.
For the above reasons some evident economic incentives (or possibly legislation) within the
EU would be useful to help bring railways onto a still more ambitious environmental path. In
particular, low-drag and low-mass trains, as well as dual-mode trains, may need additional
incentives to be introduced.
The energy performance is often unclear and uncertain for the corporate decision maker when
purchasing a new train.. Therefore it is essential to have a standard for mandatory declaration
and certification of energy performance, as is done for some road vehicles. Such
standardization would contribute to confidence in energy-saving features.
Support for R&D is important. Substantial research is needed for low-mass trains, and partly
also for reduced air drag and eco-driving.
Acceptability
It is anticipated that society and users would not present any strong positive or negative
attitudes to the proposed measures, provided that noise and vibration emissions are kept within
accepted limits. However, public resistance might, from time to time, occur against new
railway infrastructure undertakings, in particular if new railway links are planned.
Infrastructure upgrading
EU-27 has a low market share (8 %) for passenger rail transport (including metros and
tramways) in comparison to some other highly developed countries outside EU, in particular
Japan (32 %) and and Switzerland (16 %). In order to achieve a considerably higher market
share, European passenger rail must be developed according to “best practice” in the world,
through reduced journey time as well as improved reliability and cost efficiency.
If a considerable modal shift to rail is envisaged a major restriction is the need for capacity and
overall performance in the rail transport system and, consequently, funding for improvement of
the railway infrastructure. A comparatively large share of investments in transport
infrastructure must be directed toward rails systems. New high-speed railway links are needed
in addition to upgrading of the existing infrastructure. Besides enhancing capacity, also speed,
noise abatement, reliability and general flexibility must be improved.
Upgrading the existing infrastructure is also partly necessary to fully implement proposed
technologies. On some rail networks it is desirable to enlarge the loading gauge (useful cross
section of vehicles) and on some networks the electric supply systems should be upgraded in
order to fully be able to recover electric energy. These improvements are anticipated to be
consistent with average 1995-2010 levels of rail infrastructure investment (0.3 % of GDP).
Generally the cost and GHG implications for improving and maintaining new or existing rail
infrastructure is not part of this study, as is also the case for the equivalent studies for road, air
and maritime transport. Current track-access charges are however included in cost estimations
for rail operations. Taxes and similar duties are not included.
Deliverable D4 – WP3 passenger 2

1 INTRODUCTION
Rail passenger transport in the EU-27 totalled in 2007 a volume of 480 billion passenger-km,
of which metros and trams made accounted for some 85 billion (EU, 2009). The market share
was almost 8 % of the total passenger transport volume (in passenger-km, intra-EU), although
with large variations between countries. In comparison some non-EU countries with well
established and managed rail networks had considerably higher market shares; as shown in
Table 1-1 below.
Table 1-1
Market shares in passenger transport by 2007 (passenger-km, pkm)
Source: EU (2009): EU energy and transport in figures
Road (car, bus) EU-27 83 %
Air EU-27 9 %
Rail EU-27 8 %
cf. Switzerland 16 %
Japan (approx) 32 % (2006)
The market share for the rail mode declined in Europe for a long time until the mid 1990´s.
After 1995 the decline has continued in the new member states (EU-12). However, in the old
member states (EU-15) rail passenger transport has recovered to some extent, as result of new
EU and national rail policies leading to investments in new trains and infrastructure, as well as
the ongoing attempts of market liberalization. The most successful countries in this respect are
France, UK and Sweden. Table 1-2 shows the development of the passenger rail market
(excluding trams and metros) in some areas and countries in the period 1995–2007, in
comparison to the average development of all transport modes.
Table 1-2
Development of passenger transport (passenger-km) in EU-27 in the
period 1995-2007, rail and average of all modes. Source: EU (2009)
Average all modes EU-27 +22 %
Rail EU-27 +13 %
EU-12 (new members, average) -33 %
EU-15 (old members, average) +25 %
France+UK+Sweden +52 %
It is concluded that these new policies have started a revitalization of railways in Europe,
although not yet in all member states.
Rail passenger transport is widely considered to have low specific energy consumption and low
direct or indirect greenhouse gas (GHG) emissions, in comparison to competing modes of
transport. Although this is not always true, it is usually the case in electric train operations, in
particular when the electricity mix is dominated by renewable primary energy sources or
nuclear power. Diesel trains are usually not outstanding in energy and emission performance,
but can in many cases also be very competitive in the energy and GHG respect.
The main reasons for the low energy consumption are the following technical characteristics
that are inherent in the basic rail technology:
- Rolling resistance is very low with smooth hard wheels running on smooth hard rails.
- Most vehicles in a long train are (to a large extent) shielded from the air drag by other
vehicles ahead or behind.
Deliverable D4 – WP3 passenger 3

- An electrical wire (so-called catenary) can be positioned over the track, thus facilitating
a simple mean of electricity supply and energy recovery. In some cases the overhead
catenary is replaced by a third rail for supply of electricity.
- Electric energy can, with modern technology, be partly recovered and fed back to other
trains, by using the electrical motors as generators when braking (retarding or
maintaining speed on descents).
There are also technical features in today’s technologies that are less favourable with regard to
energy and GHG emissions. For example, the vehicle mass and space per passenger seat is
usually high, some trains have an unfavourable exterior shape with regard to aerodynamic
drag, and there are still substantial losses in the power equipment. In some cases (e.g. in lower
traffic flows) the quite long trains may not be fully utilized. Furthermore, about 10–12 % of the
total rail passenger transport (in passenger-km) in Europe still relies on diesel powered trains
(i.e. not using external sources of electricity). In conclusion the favourable inherent technical
characteristics of the rail mode have for a long time not been fully utilized. A reason for not
prioritizing energy issues is likely that railway companies have, until quite recently, felt that
the energy cost is not an essential part of the total cost. Recently, however, there is a steadily
increasing awareness of the economical and ethical implications of energy consumption and
GHG emissions.
For the above reasons there is potential for further improvement, partly be relatively simple
means. Some of the necessary technologies and measures are known and partly already
implemented. Due to increasing awareness of environmental responsibility as well as
economical incentives they are expected to be more widely used in the future.
2 REFERENCE SYSTEMS CHARACTERISTICS
To facilitate a consistent evaluation of different technologies for the reduction of energy
consumption and its resulting GHG emissions, four reference trains are defined in Table 2-1.
These reference trains are estimated to represent around 95 % of the total rail passenger
transport volume in Europe (in passenger-km, pkm). Metro trains are included in the group
‘Local city trains’, being quite similar with regard to energy and operational characteristics.
Based on EU and national statistics, the approximate market shares for different types of trains
and train services are estimated within TOSCA WP3; see Table 2-1.
It should be noted that the defined reference trains are typical new trains that are put into
service in the reference year 2009. They will to some degree differ in performance from the
average fleet of trains in service. The techno-economical lifetime of European trains are
usually in the order of 25 years, sometimes longer. The principle of using the average new
reference vehicles delivered in 2009 is used throughout the Tosca project, also in workpackages
for road, air and maritime transport. In all cases there will be delays in their
introduction.
Energy estimations are made from several sources (Andersson et al, 2006), (Kemp et al, 2007),
(Lukaszewicz, 2001), (Lukaszewicz et al, 2009), (RSSB, 2007), (Railenergy, 2009) as well as
un-published data. Energy data have been judged and confirmed with the KTH simulation
software STEC (Simulation of Train Energy Consumption). Included in this stage of TOSCA
WP3 is the direct energy for propulsion, auxiliaries and comfort, also including losses in the
railway’s dedicated electric supply system.
Cost estimates are made by experts within WP3, with cost models used at KTH. Note that
operating cost at this stage excludes energy cost, as energy cost is a variable in Stage 2 of
TOSCA.
Deliverable D4 – WP3 passenger 4

Table 2-1
Reference characteristics (average new 2009 systems in operation)
High-speed
train unit
Electric
Intercity or
regional
Electric
Intercity
or
regional
Diesel
Local
city train
Electric
Train configuration a 2 PU + 8 Tda L + 8 T 6 DMU 12 EMUa
Train length (m) 200 230 150 214
Capacity (seats) ∙ load factor 510 ∙ 0.65 480 ∙ 0.40 330 ∙0.40 750 ∙ 0.35
Operational mass, incl. average load (tonnes) 410 480 280 430
Typical max operational speed b (km/h) 300 160 140 140
Average scheduled stopping distance (km) 120 24 20 3
Average speed, incl. stops (km/h) 200 105 85 55
Running distance per year (1000 km) 500 250 200 120
Max power at rail (kW) 9000 6000 1500 8500
European rail market share (% pass-km) f 20 43 12 25
Energy use – at train intake c (MJ/pass-km) 0.22 0.32 0.73 0.35
Energy – from public grid (MJ/pass-km) 0.24 0.36 - 0.38
GHG emissions (g CO 2 -eq per pass-km) 31 d 46 d 65 49 d
of which direct emissions 0 0 54 0
Sales price of train unit (MEUR) 28 16 12 14
Operating costs e (EUR per pass-km) 0.063 0.101 0.127 0.095
Average lifetime (years) g 25 25 25 25
a
DMU = Diesel multiple unit cars; EMU = Electric multiple unit cars; L = locomotive; PU = Power unit;
T = trailer car; Index “a” = articulated train; Index “d” = double decker.
b Maximum speed is not reached on all sections of the line and not on all lines.
c
Net electricity as intake to the train, i.e. energy recovery is subtracted. Intake to the railway facility into
converter or transformer station) is assumed to be 10 % higher on average.
For liquid fuels: energy content as delivered into the fuel tank.
d Average EU-27 electricity mix (2009): 460 g CO2 -eq per kWh or 128 g per MJ, at intake to consumer.
e Excl. energy, i.e. train capital + maintenance + crew + track & station & dispatch charges + train formation +
f
sales and adm.
In percent of the considered part of passenger rail transport. Trams and local rural trains, as well as night hotel
trains (together around 5–7 %) are excluded from consideration in this context.
Comments
There is a large variety of train types throughout Europe, and the reference selection is not
representative for each member state of the EU. Of the selected trains the ‘Intercity or regional’
loco-hauled electric train is expected to have slightly higher energy use and indirect GHG
emissions than the European average, while the selected ‘High-speed train unit’ is expected to
be slightly better in these aspects than the average.
Average energy cost in European electric rail operations is in the order of 10 % of total cost.
Deliverable D4 – WP3 passenger 5

3 TECHNOLOGY DEVELOPMENT
3.1 General
In this section technologies and operational measures are presented, which offer a technical
potential for reducing energy consumption and (indirect or direct) GHG emissions. Most of
these technologies are listed and described in the ‘Energy efficiency strategies for rolling stock
and train operation’ (UIC, 2010), published continuously by UIC (International Union of
Railways) as a result of on-going research and development. Some of these technologies and
operational measures are also preliminary results from the EU-funded project Railenergy
(Railenergy, 2009). Earlier research at KTH (Andersson, 1994), (Lukaszewicz, 2001),
(Lukaszewicz et al, 2009) is also used.
All the above-mentioned technologies are subject to critical evaluations and judgement within
TOSCA WP 3. These assessments were complemented with simulations and other estimations
where necessary.
A special case where development outside the railway sector is important, is the prospected
change of the electricity mix in Europe’s energy market. In 2009 the average CO 2 -eq emissions
within the EU-27 electricity generation was about 435 g per kWh of electricity, according to
TOSCA WP4 (Perimenis et al, 2010). At the consumer level (taking into account losses in the
electricity transmission) the emissions were 460 gCO 2 -eq per consumed kWh, or 128 g per MJ
of electricity. Very significant reductions in railway GHG emissions can be achieved with a
declining share of fossil fuels and increasing shares of bio fuels and wind power, likely also
solar, geothermal and nuclear power. Another promising technology is the introduction of
carbon capture and storage (CCS) for fossil fuel power plants. Low-carbon electricity is further
discussed as a technology denoted PJ in Section 3.3.
The different technologies and measures (PA – PK presented below) are referred to as technology
trajectories, including the projected approximate year of technology readiness, i.e. when a
technology is ready for commercial sale. In some cases the energy-saving technology is
already partly introduced. The indicated amount of improvement should be interpreted as the
potential for further improvement in comparison with the reference cases of 2009 in Table 2.1.
Before marketing the improved technologies they must be integrated into products, even if the
basic technology is available in 2009. In these cases the improved technology is anticipated to
be ready for commercial sale by about 2015, provided that corporate decisions on product
development are taken by 2011–12. Another 4–5 years are needed for delivery and introduction
in full-scale regular service. In order to be considered as “standard” for newly delivered
systems will finally take some additional 3–5 years.
Summing up this chain means that new systems ready for market introduction (sales) by 2015
can be a standard for modern systems by about 2022–2025. Even then most of the older
technology, say from the reference year 2009, will be used in daily operations, although less
intensively. However, compared with the situation by the reference year an improvement will
take place, because even older technology was used in 2009. The practical lifetime of trains –
i.e. with intensive use – is usually in the order of 25 years, although many trains will survive
for a longer time as operational reserve or used at peak periods.
In some cases commercial introduction is anticipated at a later stage, by 2020 or 2025. A
second estimate is on the very long term (2050) anticipating a continuous or stepwise development.
This is further explained in Section 3.5.
Note that the full implementation of most technologies will usually take some 20-25 years
from the first commercial delivery, due to the long life of railway equipment. In some cases
Deliverable D4 – WP3 passenger 6

improvements of the rail infrastructure are required, that sometimes take even more time. All
transport modes have long time scales for their introduction, although this time scale is usually
shorter in road transport than for the air, maritime and rail modes.
Commercial introduction on a larger scale may for some technologies require stronger
economic incentives and/or legislation. However, the liberalization and deregulation of the
European rail transport market is expected to force operators to make their operations more
efficient, which in many cases is expected to reduce also the specific energy use and GHG
emissions.
Finally, only technologies and measures that are foreseen today (2010) are included. In the
very long term some new technologies, not known today, may be invented and developed. On
the other hand, some of the technologies described below may for one reason or another not be
developed, or developed to a lower level than anticipated.
3.2 Technology trajectories
The following chart summarizes the technology trajectories for technology domain “rail
passenger transport”. It shows the various technology opportunities independently.
PA. Low-drag train (aerodynamic)
PB. Low-mass train
PC. Energy recovery
PD. Space-efficient train (long-distance)
PE. Modular short train
PF. Eco-driving - driving advice or automatic operation
PG. Dual mode and hybrid (diesel/bio fuel + electric)
PH. Bio fuels in diesel engines
PI. Electrification of non-electrified lines
PJ. Low-GHG electric power
PK. Magnetic levitation
2010 2015 2020 2025 2030 2035 2040
2050
Figure 3-1
Technology readiness for various technologies
Deliverable D4 – WP3 passenger 7

3.3 Technology descriptions
The potential for reducing energy use or GHG emissions is presented in percent per unit load
(passenger-km), assuming the same average speed, in comparison with the reference system of
2009. Change of speed is taken into consideration in Sections 3.4 and 3.5. In most technology
descriptions below (except FH, FI. FJ) energy savings is the immediate target, but these
savings directly translate into GHG reductions proportionally.
At this stage technologies are mainly described independently of cost effectiveness. In Sections
4 and 5 economic, social and user considerations are also taken into account.
PA. Low-drag train
Technologies are basically available for reducing air drag by 15–30 %, compared with the
reference trains in Table 2-1. Some of them will need further maturity and acceptance (UIC,
2010) (Diedrichs, 2010). Low air drag is most important for high-speed long-distance trains,
but has a significant impact also on fast regional trains and local trains as the two latter trains
have today usually a modest aerodynamic performance. The low-drag train is estimated to have
a long-term energy-savings potential of around 15 % – with an intermediate step at about 10 %
– compared with the reference ‘High-speed’ as well as for the ‘Intercity and regional’ segment.
PB. Low-mass train
Lower mass will reduce energy consumption and related GHG emissions per pkm, in particular
on stopping trains (local trains and metros), although re-generation of braking energy in
electric operations to some extent will reduce the positive impact of low mass. Technologies
are today not available for a major reduction of train mass (around 20 %), but may become
available as a result of massive R&D. New materials and/or designs are needed and likely also
economic incentives. The potential of energy savings is estimated to 10 % in local or metro
trains - ‘City trains’- less in other train types, all compared with reference trains in Table 2-1.
PC. Energy recovery
Today’s modern electric passenger trains use the (electric) motors as generators when braking,
thus feeding back electric energy to other trains on the line, or (sometimes) to the public
electric grid. Alternatively, electric energy can be stored onboard in batteries or supercapacitors,
for storage of braking energy until the next acceleration; the latter is of special
interest for non-electric trains (diesel fuel and similar) with no external electric connection. All
this is called energy recovery or regeneration. Energy recovery braking will also reduce the
maintenance of the mechanical brakes.
This technology is partly introduced already today. The feed-back ability is however usually
smaller than optimum due to the limited power (in kW) of the propulsion equipment of the
train and/or by the limited number of powered axles, as well as of the limited recovery capacity
of the electric supply chain. With the rapid development of electric power equipment it is
increasingly technically and economically feasible to install more power in trains and in the
supply network, and also to improve possibilities of feeding electricity back to the general grid.
Energy recovery has been subject to extensive simulations in this study and is estimated to be
able to save energy use and related GHG emissions by up to 20 % over the long-term, in
relation to today’s ‘Local city trains’. This potential is smaller in high-speed operations with
only few stops, although also downhill gradients contribute to energy recovery. In some cases
the electric supply system must be modified or rebuilt. A continuous improvement of
regeneration and energy recovery is anticipated over the years.
Energy recovery interacts with other technologies, such as ‘Eco driving’, ‘Low drag’ and ‘Low
mass’. To fully understand and evaluate the impact of Energy recovery combination scenarios
are needed; see further Section 3.5.
Deliverable D4 – WP3 passenger 8

PD. Space-efficient train
According to (RSSB, 2007) and author’s estimations European long-distance high-speed trains
have an average of about 2.2 seats per metre of train length, while most Japanese high-speed
long-distance trains (Shinkansen) have an average of 3.3. About half of this effect is due to the
use of wide carbodies in Japan but also due to less space for catering facilities and a more
space-efficient interior layout in general. In Europe double-decker trains are used to a certain
extent to improve space utilization, while wide-bodied trains (exterior width > 3.2 m) are only
possible on a small number of rail networks, because of space restrictions by the rail
infrastructure. Double-decker trains currently accommodate 2.5–2.7 seats per metre of train in
the high-speed versions; sometimes more on regional trains. With a consequent use of spaceefficient
train interiors - including intelligent seat design - some 10–15 % of energy and GHG
emissions (per seat-km) could be saved (Kottenhoff et al, 2009) compared with reference
trains. Note however, that the reference high-speed train in Table 2-1 is a double-decker which
already has a comparatively high amount of seats per unit length of train. In local ‘City trains’,
and many regional trains, there is already a high degree of space utilization, including
sometimes also standing passengers, so in these cases no further improvement is anticipated.
A particular mean of improved space utilization is the exchange of locomotive-hauled trains
for so-called multiple unit trains (EMU=Electrical, DMU=diesel), having their propulsion
equipment installed in the passenger cars. Usually train performance can also be enhanced
(acceleration, speed, regenerative braking, etc) and the train mass per seat can be reduced.
PE. Modular short train
Today many trains have a length of 5–16 cars (or equivalent if shorter carbodies than normal)
due to the required capacity during peak hours. A large proportion of these trains is running in
the same formation all the time, without shortening the train length during off-peak hours. This
is due either to the fixed long train configuration, which does not allow any decoupling (in selfpropelled
multiple units), or to the cost and time needed for decoupling cars in a loco-hauled
train. If trains are made in shorter self-propelled modules (3–6 cars) these modules can quite
easily be de-coupled, and the train length is halved when the full size is not needed. A halfsize-train
would save some 40 % of energy compared with a full-size one. To some extent
these principles are already applied in today’s operations, but they could be extended to further
areas. With an assumed 20 % (more than today) of the trains running at half-size some 6–8 %
of energy should be saved. However, it is at this stage uncertain to what extent these principles
can be effectively used within Europe. In some EU member states safety regulations presently
require some 5–10 m of empty (i.e. no passenger) space at each end of the train, which would
neutralize the benefits of shorter trains.
PF. Eco-driving - driving advice or automatic operation
Optimization of driving style means, for example, coasting before braking and downhill
approach, use of regenerative brakes as the ordinary brake, running slowly when time allows,
etc. Such optimization is estimated to have a saving potential in the order of 10–15 % in its
first generation, compared with the average manual driving of 2009 which is the reference in
Table 2-1 (Railenergy, 2009) (Lindemann, 2010) (Luijt, 2010).
Driving can be optimized by training for skilled driving, by on-line computerized support and
advice to drivers or by automatic train operation. The latter may be used on metros and
dedicated commuter railways. To a small extent this technology is already commercially
introduced, but is estimated to be improved and fully implemented in modern trains within the
next 10 years. These improvements are relatively inexpensive to introduce and are applicable
to most types of trains.
Deliverable D4 – WP3 passenger 9

At a later stage (say by 2025 and beyond) this technology may be co-ordinated with rail traffic
control, i.e. traffic flow management, which would lead to further improvement. Such measures
would also enhance railway’s transport capacity to some extent. In the long term (2050) ecodriving
is estimated to save a further 5–10 % compared over the first generation, used as a
single technology. However, there are interactions with other technologies; see further
combination scenarios in Section 3.5.
PG. Dual mode and hybrid trains
In a dual mode train electricity is used on electrified sections while diesel or bio fuels are used
on non-electrified parts of the operation. Today most trains running on both electrified and
non-electrified sections use diesel power only. Depending on the share of electrified sections in
the actual operation, and carbon content of electricity, energy savings and emission reductions
can be in the order of 20–50 % and sometimes more, compared with pure diesel operation.
Another possibility is hybrid diesel-electric propulsion with on-board energy storage. These
technologies are available today, but are sparsely used. It is expected to be further matured and
accepted during the next 10 years. Maximum market penetration is limited to diesel-hauled rail
services, i.e. around 10–12 %. Further incentives are likely necessary for a wider introduction.
PH. Bio fuels in diesel engines
As an alternative to diesel fuel, diesel engines can be powered with liquid or gaseous bio fuels,
which can reduce life-cycle GHG emissions. WP4 of TOSCA (Perimenis et al, 2010), suggests
Hydrogenated Vegetable Oil (HVO) as the most promising alternative available today. The net
life-cycle content of CO2-eq is about 45 % lower than for ordinary diesel fuel, while
production cost for HVO is about 50 % higher. In the longer term a new generation of bio
fuels, produced from wood, is anticipated to be available.
Most important, the future availability of biomass for energy purposes depends on the amount
of land that is not needed for food production (non-food areas) and other purposes. This is a
very uncertain issue. Therefore, the biomass potential is also highly uncertain and requires
much further analysis. This is out of scope in this study on rail passenger transport.
Due to the limited market penetration in rail operations (an absolute maximum of 12 %, i.e. the
share of diesel operations) and, in particular, the high uncertainties in future large-scale
availability, this option will not be further analysed in this report. The bio fuel option will be
further evaluated by WP6 in Stage 2 of TOSCA, as a scenario variable.
PI. Electrification of non-electrified lines
Electric rail operations are usually more energy efficient than less GHG-intensive than diesel
operations, in particular if electricity is partly generated by other means than fossil fuels. This
is shown in Table 2-1, where the electric ‘Intercity or regional’ train has lower GHG emissions
per passenger-km than the diesel train, despite higher speed and higher train mass per seat.
Today, several European countries have very limited part of their rail networks electrified.
Plans for improvements have been presented, for example in the UK. In these countries
substantial reductions of GHG emissions from rail operations are expected, in particular if
‘Low-GHG electric power’ is used in the future (see further FJ below). Massive electrification
to cover, say, 95 % of all European rail passenger transport volumes (instead of the present-day
88 %) would reduce GHG emissions by 3–8 % depending on the electricity mix.
The limited general impact – because of the low additional market penetration – and the
associated cost of electrification is a matter of optimization (i.e. what lines should be converted
to electric operation). This is however outside the scope of this study. Therefore no further
analysis of electrification is made here. This does not exclude that certain lines could and
should be electrified after a thorough analysis.
Deliverable D4 – WP3 passenger 10

PJ. Low-GHG electric power
Electric power is by 2009 produced by a number of means in Europe, some of them using
fossil fuels (mainly coal and natural gas), some of them with renewable energy (hydro, with
increasing shares of wind and biomass). Nuclear power is also an essential source of nonrenewable
energy, but with zero direct GHG emissions. Fossil fuels had by 2007 about 50 %
share of the total. The electric power production is part of the European Trading Scheme
(ETS), which means that GHG emissions will have a monetary price. Tomorrow’s electric
power mix must have substantially diminishing dependence of fossil fuels if GHG emission
targets are to be met. Further, long-term technology options include carbon capture and storage
(CCS) as well as geothermal energy, and possibly also solar and sea wave power. Also, an
increasing share of combined heat and electric power (CHP) production (with fossil or bio
fuels) increases the efficiency and reduces specific emissions related to electric power.
The GHG emissions from today’s electric power mix are estimated in TOSCA WP4 as being
460 gCO 2 -eq per kWh of electricity at consumer's outlet. GHG emissions from future
European electric power is a scenario variable in stage 2 and will be determined by WP6 in cooperation
with other work packages.
Substantially reduced average GHG emissions from electric power production will be a very
effective means of reducing emissions from the European transport sector, not only for
railways but possibly also for passenger cars in the road sector. For example, a reduction of
GHG content per kWh of electricity by 80 % will reduce specific emissions of electric trains
by the same amount. Maximum market penetration in the rail sector is about 90 % (i.e. current
and future diesel operation is excluded).
PK. Magnetic levitation (Maglev)
Magnetic levitation has been technically developed for commercial introduction since about
2000, both in Germany and Japan. However, only one commercial magnetic railway for higher
speeds has been built, i.e. the Shanghai airport Maglev system. In addition, a long-distance line
of 200 km is reported to be planned outside Shanghai. Magnetic levitation trains achieve
commercial top speeds in the range of 400 to 500 km/h; the horizontal curves may be tighter
and the gradients steeper than on conventional railways for the same design speed. Despite
these advantages, the cost of constructing magnetic railways is high, partly due to the
installation of continuous high-powered linear drives and levitation equipment along the line.
Another drawback is the non-existing interoperability with conventional railway systems. In
addition, the difference in average speed between Maglev and conventional high-speed rail
has been successively reduced in comparison with the situation 40 years ago, when the
development of magnetic levitation started.
Energy consumption is estimated to be in the same order as on conventional high-speed stateof-the-art
trains at the same speed (Kemp et al, 2007). Other sources end up with almost the
same conclusion, or somewhat higher energy consumption for magnetic rail, all comparing the
same speed. However, if the full speed potential of magnetic levitation is to be achieved the
total energy demand will increase. This is in spite of the fact that the above-mentioned
technology options PA, PB, and PD (see above) are already included in existing proposals for
magnetic levitation systems.
The option of Maglev trains will not be further studied in the context of Tosca, as this option is
not suited for reduction of energy consumption and GHG emissions.
Deliverable D4 – WP3 passenger 11

3.4 Incremental improvements and higher speed
General trends
Despite slightly higher average speeds and a modest priority for energy-saving measures in the
rail sector until recently, specific energy consumption has declined. For example, in the period
1990–2007 the Deutsche Bahn reduced its specific primary energy use by about 20 % in
passenger transport and about 35 % in freight (UIC, 2008), while Swedish railways reduced its
specific final electric and diesel energy use by about 11 % on average (SIKA, 2008). The
German case means an average saving of about 2 % per year, including efficiency improvements
both in electric power generation and in train efficiency, as well as higher load factors.
The Swedish case relates to train and feeding systems efficiency only and is about 0.7 % per
year. At least half of these improvements are estimated to be due to larger steps in new
passenger train technology, while the rest is due to continuous incremental improvements.
In the future increased attention is expected to be paid on energy and emission issues.
Therefore, the general historical trend of incremental improvement is supposed to continue at
least at the same rate as in the Swedish case, i.e. by about 0.3 % per year in relation to the total
rail energy consumption, excluding increased efficiency due to larger steps and improvements.
Incremental improvements – in particular reduced energy losses
Incremental improvements – besides the larger measures PA–PI in the railway system as
described above – are possible. Examples include: comfort energy reduction, traction and
auxiliary systems with reduced losses as well as reduced losses in the railway’s electric supply
systems. Until 2050 it is anticipated that a saving of 12 % will be reached compared with the
2009 reference, considered here as a single measure. This is equivalent to 30 % reduction of
the losses in electric rail operations, as losses, auxiliaries and comfort constitute about 40 % of
total energy consumption as an average. It is estimated that an energy saving of around 7 %
(out of 12) would be in reach for new trains until 2025 (UIC, 2010) (Railenergy, 2009), with
another 5 % assumed until 2050. Reduction in comfort energy can be facilitated by recovery of
heat, better controlled ventilation, heat pumps etc. Also losses in electric power equipment (on
trains and in supply systems) are assumed to be reduced, for example by introduction of
permanent magnet motors and more efficient converters. Similar improvements (totally 12 %
until 2050) are expected in diesel railway operations as well, due to improved efficiency in
diesel engines and in energy use for comfort and auxiliaries.
Rapid improvements are constrained by the long life of railway equipment (vehicles and fixed
installations). In the long term (say more than 15–20 years) the potential is higher if
progressive decisions are taken within the next 5–10 years. Many of these measures will be
motivated also for economic reasons.
Higher speed
There is a strong tendency to reduce travel time for several reasons, in particular increased
attractiveness against other modes but also improved productivity for trains and train crew.
These positive factors contribute to a modal shift to rail. In many cases (although not all)
reduced travel time will imply increased top speed, which, in turn, will imply increased energy
use with all other conditions equal. However, train technology is usually adapted to speed.
For the ‘Intercity or regional’ (electric) segment top speeds are estimated to increase (on
average) by 0.9 % per year, i.e. by 15 % until 2025 and by about 40-45 % until 2050. This
follows a general trend in top speeds for the last 150 years and there is no tendency or reason
that this continuing development should decline for the intercity or regional segment. This
segment is heterogeneous with top speed ranging from 100–120 km/h up to 220–240 km/h in
2009. The typical top speed is however quite low – about 160 km/h. For 2050 typical top
Deliverable D4 – WP3 passenger 12

speeds in the range of 220–250 km/h are expected. This may an underestimation, as some
networks already today partly run this class of trains at top speeds in the range 200–230 km/h
(Germany, France, Spain, Sweden, UK, etc).
For ‘High-speed trains’ the commercial top speed in Europe (2009) is 320 km/h, while a more
common operating top speed is 300 km/h that is also defined as the reference case in Table 2-1.
In China trains for 380 km/h are already on order, although it is uncertain when and to what
extent such speeds will be utilized commercially in the near future. European high-speed trains
are assumed to have a typical top speed around 370 km/h in 2050, corresponding to an average
annual increase of 0.5 % per year. Thus, the historical trend of about 0.9 % long-term annual
speed increase is assumed to be broken. This is due to assumed difficulties to abate noise
emissions as well as increased cost for infrastructure investment and maintenance.
In local train services however, with shorter stopping distances, the top speed will increase at a
slower rate, estimated to be 0.3–0.4 % per year. In diesel operations the average speed is
assumed not to increase at all, as the more competitive routes are assumed to be electrified,
although exceptions may occur.
It should be pointed out that the average speed, including stops and other delays, will usually
increase at a lower rate than the top speed. This will be shown in Table 5-2 of Section 5.2.
Load factor - seat occupancy
Many railway operations have a comparatively low average seat occupancy rate or load factor
(30–50 %) if compared for example with airlines. High-speed trains however, usually have an
average load factor of 60–75 %, which is more comparable with domestic airlines. Too many
empty seats are, of course, not desirable neither from an economic nor from a specific energy
point of view. With deregulation, increasing competition and more business-oriented railway
companies the load factor would improve, by flexible fares and by other means. This is to a
large extent what has happened to the airlines in Europe and North America after deregulation.
Increased load factor is not a “technology”, its future potential is hard to quantify with
precision and is therefore not considered in the following analysis. This factor could be an
option of additional improvement in energy and GHG emissions per passenger-km.
3.5 Technology scenarios to be further considered and investigated
Some of the technologies described in Section 3.3 are very promising for the future, as they
offer large energy and GHG savings in at least some types of services and also experience high
possible market penetration, i.e. in several types of rail services with considerable total market
share. These promising technologies constitute five scenarios (1–5) to be further considered
and investigated in this study.
1 PA. Low-drag
2 PB. Low-mass
3 PC. Energy recovery
4 PD. Space efficiency
5 PF. Eco-driving
The following combined scenarios are also studied:
6 Pcomb electric: PA + PB + PC + PD + PF + Incremental
7 Pcomb HS electric: As (6) + Higher Speed
8 Pcomb as (7) + low GHG electric power
9 Pcomb diesel: PA + PB + PC + PD + PF + Incremental
Deliverable D4 – WP3 passenger 13

Note that technology ‘Modular short train’ (PE) or ‘Electrification of non-electrified lines’ (PI)
is not explicitly part of the scenarios, since it is not possible within the scope of this study to
estimate their future feasibility and implementation. PE and PI are therefore also nonquantified
options for additional improvement.
Note also that ‘Bio fuels’ (PH) is not either part of the scenarios at this stage, because the
potential of this option is dependent on uncertain variables being assessed in Stage 2 of
TOSCA. The same is e.g. the case for ‘Low-GHG electric power’ (PJ) but a tentative
improvement is anyhow studied in Scenario 8 above, just to understand the possible impact of
low-GHG electric power.
Scenarios 1 to 5 with single measures are comparatively easy to estimate. The combined scenarios
are, however, much more complicated since there are considerable interactions and
dependences between the different measures, if combined. In this context a thorough and
validated simulation software (STEC) is used, taking all factors into consideration in the same
run. In the simulations with STEC the full estimated potential of improvement until year 2050
is used.
As it is very hard to find qualified estimates on rail technology for 2050, the total time 2009-
2050 is divided in two periods. Due to the current lack of future estimates for the railway
sector, the following procedure is being applied:
Period 1 The techno-economic potential for in-service introduction during the next 10–15
years (i.e. until 2025) is estimated according to the earlier mentioned sources.
Period 2
Development is assumed to continue after 2025. For most technologies it is
assumed that 2/3 of the first period achievement can be achieved in the second
period, i.e. 2025–2050, provided there are no obvious physical or economic
obstacles. The exception is technology PB (Low-mass train) that requires more
substantial research before technology readiness (see Table 3-1), which will
delay most to its possible introduction to Period 2.
This means that, for most technologies, 60 % of the total achievement until 2050 is what today
is known or assumed as appropriate technologies, possible to introduce in regular service
during the Period 1, while 40 % is assumed to be achieved during Period 2. As the latter period
is longer than the first one, the annual rate of improvement is assumed to be slower in the
second period. This assumption is justified as the first steps are considered as technically and
economically quite well-founded, while the following steps would likely require more
advanced technologies, also being closer to the physical and economic limits.
If – for example - the total combined improvement is 50 % reduction of specific energy in the
total period (41 years), then the energy will be reduced by 30 % under the first period and by a
further 20 % (of the original amount) in the second of 25 years. This ends up with 2.2 %
average reduction per year until 2025 and 1.4 % reduction per year during 2025–2050. With
the same percentage annual rate the assumed 50 % reduction until 2050 is equivalent to an
annual reduction by 1.7 % per year.
Table 3-1 presents examples of changes for improvement in the first period (until 2025). These
changes have been judged as possible with reference to (mainly) UIC Energy efficiency
strategies for rolling stock and train operation as well as (Railenergy, 2009) and (Diedrichs,
2010), i.e. essentially the same as in Section 3.3. They are included in the earlier mentioned 9
scenarios and are investigated both as individual technologeis and in combination. Results are
presented in Section 5.
Deliverable D4 – WP3 passenger 14

It should be noted that it is not for sure that all these changes in technology will really be
implemented: At this stage they will only be studied in the context of their potential of
reducing energy use and GHG emissions.
Table 3-1
Examples of changes until 2025, in relation to reference trains.
High-speed Intercity Local city
or regional
Aerodynamic drag - 16 % - 21 % a
Seats per metre of train + 9 % +22 % a + 0 %
Train mass per metre of train - 0 % - 4 % b - 7 % b
Per cent of energy recovery at braking c 75 % to 84 % (2050) 44 % to 72 % (2050)
Energy loss in power systems, per car e - 18 % - 18 %
Energy use for auxiliary and comfort e - 18 % - 18 %
Eco-driving
Speed profile smoothing and coasting before
braking, within a time margin of 3 % d
Increase in top speed 15 % 26 % 12 %
__________________________________________
a Including transition from locomotive-hauled electric trains to multiple units.
b 1/3 of total reduction of tare train mass (until 2050) is estimated to be introduced as state-of-the-art on new trains
until 2025, while 2/3 is introduced in the period 2025–2050.
c In relation to the theoretically possible amount after energy losses, air resistance and rolling resistance, braking
only with electric regenerative brakes. Note that it is assumed that also the reference “state of the art” case have a
substantial average amount of energy recovery in electric rail services, which is not the case in all recent rail
operations. The average real improvement in relation to the present situation for new trains could therefore in
some cases be larger.
d
Driving style is adjusted until 3 % of time-tabled time is used. Usually the total time margin is in the order of 10
%. The use of 3 % of time-tabled time for Eco-driving is an average, sometimes it can be done, sometimes not and
sometimes more, depending on the actual time-keeping situation for the actual train.
e As an example, energy efficiency in the train is originally 82 %, i.e. losses are 18 %. With 18 % reduction of
losses (until 2025) the resulting energy efficiency increases to 85.2 %. In similar way energy efficiency of the
railway’s supply system is increased from 91 % to 92.6 % until 2025. In the very long term (until 2050) energy
losses are assumed to be reduced by 30 %.
Deliverable D4 – WP3 passenger 15

4 CONSTRAINTS FOR REDUCING GHG EMISSIONS
Most of the proposed technologies are essentially neutral in the public and end user opinions.
The end users or rail operators are not expected to show any opposition or resistance to trainbased
technologies such as lower mass, energy recovery, eco-driving, modular trains or
multiple units instead of conventional locomotives and cars, and not even hybrid or dual-mode
trains. On the contrary, most end users and operators will likely react positively to a more
pronounced aerodynamic design, newly designed multiple unit trains and advanced
technologies in general. Higher speeds and reduced travel time is a very positive factor and
will strengthen the competitiveness of rail systems. All this together would most likely strengthen
the rail business opportunities through a more positive public image.
In many cases the new technologies will also lead to per-unit cost reductions, through lower
energy bills or higher capacity for the same length of train. This is also a strongly positive
factor. Also the strengthened environmental image around trains and rail services is positive,
provided that external noise and vibrations remain within acceptable limits.
There are not even any obvious obstacles for the scalability of proposed technologies, either
regarding natural or human resources or regarding social equity. The main restriction will be
rail infrastructure capacity and financial resources for infrastructure investment. Regarding
infrastructure capacity a lot of improvements are needed to generate a major increase of
transport capacity, i.e. new high-speed lines and rail freight corridors, double tracking of
single-track lines, improved signalling systems etc. A larger loading gauge (wider and/or
higher cross section of trains) is also highly beneficial. Loading gauge is most restrictive in the
UK, but also continental Europe is restricted compared with most other parts of the world. This
is due to different obstacles as platforms, tunnels, bridges and others.
Regarding aerodynamic design we should point out a possible resistance against covers of
bogies, end couplers etc that might be troublesome and time-consuming in operation and
maintenance, if not designed and handled in the right way. On the other hand, sometimes
covers are beneficial to protect sensitive equipment from snow, dust etc. Also, very long nose
and tails of a train, as may be required for superior aerodynamics at very high speed, will
reduce space for paying passengers, in particular for short trains (6 cars and below). Thus long
nose or tail would to some extent hamper the economic and efficient use of the train.
It is not even for sure that mass reduction is always profitable from a pure economic point of
view, if more advanced materials and manufacturing processes must be used. There is also a
risk of safety implications and a more modest comfort regarding noise, vibrations and others. A
substantial amount of research and development is required. The positive effects are highest for
stopping trains with much acceleration and stops (suburban and metro trains).
There are mainly two issues identified that could be troublesome in the customer and public
opinions. The first one is the issue of improved space utilization. If the space for each
passenger is felt to be reduced to an extent that is hampering the feeling of privacy or comfort
or ability to eat, read, use lap-tops or carry luggage, it will certainly be negative. Done in the
right way it must not be negative however. There are, for example, possibilities of smart spacesaving
chairs (Kottenhoff et al, 2009) and of more rational and space-efficient food services
than today, while still maintaining good comfort. However, there is some risk that improvement
is not made in a proper way so that passenger acceptance will really be reduced, or that
decision makers feel that this is the case. However, space utilization can also be improved by
using multiple unit trains instead of a locomotive plus cars, or double deckers, which is usually
not felt as negative.
Deliverable D4 – WP3 passenger 16

The second issue that may be troublesome is the provision of low-GHG electricity. One way is
to use nuclear power, for which a strong opposition is active on some occasions. Something
similar could possibly also be the case for carbon capture and storage (CCS), if the public
opinion feels that this storage is not sustainable and absolutely safe. However, on the positive
side of these opinions are the prospects of “greener” generation of electricity and heat.
Anyhow, if EU targets on GHG reduction are to be met, a dramatic change must take place in
electric power generation.
In addition, there is an identified risk that many train operators and leasing companies will be
conservative and hesitate to introduce new technologies that will result in additional
investment and technical risks. There is sometimes a feeling that railways and rail operations
are already superior in energy performance compared with other transport modes, so that no
substantial efforts are needed.
Another limiting factor is the increased tendency to prioritize the first cost (investment cost of
new trains) and pay less attention to the long-term life cycle cost and to the income side of the
account. In theory, at 6 % interest rate and 25 years amortization, the annual cost (annuity) for
energy-saving investment is 7.8 % of the additional investment, equivalent to a required payoff
time of maximum 13 years. In reality however the pay-off time should rather be 3–8 years,
because vehicle investors expect to have a substantial benefit before they are willing to take the
risk of investing in new technology at a higher first cost. This tendency is negative for any
useful energy-saving measure that increases train’s sales price and investment cost.
For the above reasons some evident economic incentives - or legislation - within the EU,
would be useful to help bring railways onto a still more ambitious environmental path. It
should be pointed out that economic incentives not necessarily call for subsidies; also taxes can
be used for environmentally less friendly products. If EU and national governments want to
reduce GHG emissions they have to support the change. Support for R&D is also required.
Another support for introducing energy-saving measures, with extra initial cost, is to support
and introduce a standard mandatory declaration and certification of energy consumption
performance. Such standardization will contribute to the consideration and confidence in
energy-saving features in the procurement process of new trains and in the adoption of new
operational practices.
Deliverable D4 – WP3 passenger 17

5 RESULTS
5.1 Technology readiness and R&D requirements
In this section the time of technology readiness, i.e. the technically possible first market
introduction (sales) is estimated, if corporate decisions are taken in the next 2 years. Also the
needed amount and depth of research and development (R&D) is estimated.
In many cases the basic technology for the first step already exists in the reference year 2009.
However, before marketing the improved technologies they must be integrated into products.
In some cases, also the rail infrastructure must be further developed for European conditions
and requirements, although the basic technology already exists. These considerations postpone
market introduction.
In some cases in Table 5-1 below, both company-level and EU-wide research programs are
needed, although not indicated in the table. This means that different parts of technology
development require different levels of R&D.
Table 5-1 Technology readiness and R&D requirements
Technology readiness
Most
likely
LB
UB
R&D requirements (rel. to marketreadiness)
Insignificant
Significant
(companylevel)
PA Low drag 2018 2015 2022 X
Substantial
(EU-wide
program)
PB Low mass 2020 2017 2025 X
PC Energy recovery 2015 2012 2020 X
PD Space-efficiency 2015 2012 2020 X
PE Modular short 2015 2012 2020 X
PF Eco-driving 2015 2011 2018 X
PG Dual mode 2009 2009 2009 X
PH Bio fuels 2009 2009 2009 X
PI Electrification 2009 2009 2009 X
PJ Low-GHG electric a 2025 2020 2030 X
PK Magnetic levitation b 2025 2020 never X
a Estimated year of large-scale market introduction, taking technological maturity and necessary permissions into
consideration. In particular CCS (Carbon Capture and Storage), but also other technologies, are anticipated.
b Estimated year of maturity and adaption for European conditions, however, excluding economical and financial
considerations.
Deliverable D4 – WP3 passenger 18

5.2 Energy and GHG characteristics
With the approach and assumptions outlined in Sections 3.1–3.5, supported by KTH simulation
with STEC, the potential of individual technologies, as well as their combined impact, are
evaluated. Results are presented in detail in Tables 5-2 to 5-4 and more briefly in Figures 5-1
to 5-3. Detailed results are presented in Table 5-2. In scenarios 1 to 5 individual technologies
are studied separately, while scenarios 6 to 9 represent their combined effect.
Table 5-2 Energy use per passenger-km, as estimated for the different types of rail
passenger services by 2050, measured as electricity at public grid or at the train’s
fuel tank. NB: 1 MJ = 0.278 kWh.
Bold figures are the main results to be compared.
High-speed
train unit
Electric
(MJ/pkm)
Intercity or
regional
Electric
(MJ/pkm)
Intercity
or regional
Diesel
(MJ/pkm)
Local
city train
Electric
(MJ/pkm)
Technologies and scenarios
Reference 2009 (Table 2-1) at train 0.22 0.33 0.73 0.35
at public grid 0.243 0.359 -- 0.382
1. PA Low-drag at public grid / tank 0.203 0.314 0.65 0.370
2. PB Low-mass a “ --- 0.337 0.68 0.346
3. PC Energy recovery “ 0.236 0.317 0.64 0.305
4. PD Space efficiency c “ 0.213 0.275 0.62 ---
5. PF Eco-driving “ 0.220 0.312 0.63 0.305
6. Pcomb electric (incl. incremental) 0.126 0.162 --- 0.149
7. Pcomb HS el: As (6) + higher speed 0.165 0.193 --- 0.175
8. Pcomb HS (7) + low GHG el power b 0.165 0.193 --- 0.175
9. Pcomb diesel (incl. incremental) --- --- 0.36 ---
Max operational speed 2009→2050 (km/h) 300 → 370 160 → 230 140 140 → 165
Average speed 2009→2050 (km/h) 200 → 240 105 → 123 85 55 → 60
Average load factor (%) 65 40 40 35
European rail market share (% of pass-km) d 20 43 12 25
a ‘High-speed train’ mass per metre is assumed constant, however, the number of seats per metre is assumed to
increase by 14 % to 2.9 seats per metre of train over the time until 2050. ‘Intercity or regional’ electric trains are
assumed to change from locomotive-hauled conventional trains to lighter multiple units (with a further mass
reduction per metre of 10 %), thus reducing the overall mass per seat by 25 %. The reference diesel train is
already a multiple unit; here a mass reduction of 15 % is assumed. For the ‘Local city train’ mass per metre, and
per seat, is assumed to be reduced by 20 % from 2009 until 2050.
b The GHG content of average European electricity is tentatively assumed to be reduced by 80 % until 2050.
c
‘Intercity or regional’ and ‘High-speed’ are assumed to have 2.9 seats per metre of train by 2050. ‘Local city
trains’ are assumed to have the same as in the reference case: 3.5 seats per metre train.
d At this stage assuming the same market share as in the reference year 2009; see further discussion on page 20.
Deliverable D4 – WP3 passenger 19

Figure 5-1 summarizes estimated trends of energy use over time as per cent of the reference
year, for different types of rail services, with all technologies aggregated in combination. For
electric trains higher speeds are included; for diesel trains speeds are assumed to be constant
over time.
2050 2025 2009 reference
High speed
67.9%
80.7%
100.0%
Intercity &
Regional electric
53.8%
72.3%
100.0%
Intercity &
Regional diesel
49.3%
69.6%
100.0%
Local city
45.8%
67.5%
100.0%
Figure 5-1
Trends in specific energy use (per passenger-km) over time for different types
of rail passenger services (new trains).
All technologies and influence of higher speeds are aggregated.
All types of rail passenger transport – from ‘High-speed trains’ to ‘Local city trains’ - are
competing with road transport, i.e. cars and buses. Air transport (in the range 300 - 1000 km) is
mainly competing with ‘High-speed trains’ and to some extent with fast intercity trains (i.e.
part of the ‘Intercity or regional’ segment). To compare with road and air transport respectively
energy and GHG characteristics for the rail mode is therefore divided in two tables, where
- Table 5-3a summarizes energy use and the resulting GHG emissions aggregated and
weighted over intercity, regional and city rail services, for comparison mainly with
road transport. For diesel trains only the results of combinations are presented;
- Table 5-3b summarizes energy use and the resulting GHG emissions for high-speed
rail, for comparison also with air transport.
Estimated lower (LB) and upper (UB) bounds are also indicated.
Inter-rail mode shift (i.e. between different types of rail services) is not considered in the
aggregated estimations in Figure 5-1, i.e. the percentage share in passenger-km is as for the
reference trains according to Table 2-1. In reality a shift from ‘Intercity or regional’ trains to
‘High-speed trains’ is anticipated. However, as the estimated future energy use and GHG
content (per passenger-km) only have a modest difference for these two types of trains, an
internal mode-shift will only have a modest influence (likely less than 5 %) on the total
average, although toward reduced emissions. Also, an internal mode shift from diesel to
electric trains is anticipated, although uncertain to what degree. However, as diesel trains have
just a minor market share today (about 12 %), such a limited mode shift will also only have a
minor, however positive, influence on the average GHG emissions from rail operations. The
assumed two mode shifts will both contribute to reduced GHG emissions and can be considered
as an extra margin in the estimated improvements.
Deliverable D4 – WP3 passenger 20

In most cases it is assumed that the GHG content of the electricity mix is as in 2009 (128
gCO 2 -eq per MJ el), except for scenario 8, where it is assumed to be reduced by 80 % by 2050.
Scenario 8 is just an example, as the GHG content of electricity is a scenario variable of Stage
2 of Tosca. GHG emissions from diesel trains are direct emissions (75 gCO 2 -eq per MJ fuel),
assumed to use fossil fuels only.
Table 5-3a
Energy use and GHG emissions (per passenger-km) by technology, as an
average over intercity, regional and city passenger trains, estimated for new
trains by 2050 compared with reference trains, at public grid or fuel tank.
Most
likely
Energy Use
(MJ / pass-km)
LB
UB
GHG Emissions
(g CO 2 -eq / pass-km)
Most
likely
Reference electric trains (2009) 0.368 47
1. PA Low-drag 0.334 0.32 0.35 43 41 45
2. PB Low-mass 0.343 0.33 0.36 44 42 46
3. PC Energy recovery 0.313 0.29 0.33 40 37 41
4. PD Space efficiency 0.314 0.30 0.34 40 38 42
5. PF Eco-driving 0.310 0.29 0.33 40 37 41
6. Pcomb electric (incl. incremental) 0.157 0.13 0.19 20 17 24
7. Pcomb HS: As (6) + higher speed 0.187 0.16 0.23 24 20 29
8. As 7. + low-GHG electric mix 0.187 0.16 0.23 5 4 6 a
LB
UB
Reference diesel train (2009) 0.75 54
9. Pcomb diesel 0.37 0.34 0.43 27 25 32
a
With the GHG content of average electricity reduced by 60 % (instead of 80 % as assumed tentatively) the
resulting GHG emissions are estimated to 8–11 gCO 2 -eq per net-tonne-km. The different scenarios presented in
TOSCA WP6 are approximately within the range of 60–80 % reduction.
Note that diesel trains by 2050 in Table 5-3a seem to be almost as good as the average electric
train, assuming that GHG content of electricity is the same as in 2009. This will however not
be the case if the average carbon content of European electricity is reduced. Another important
aspect is that the average speed of electric trains by 2050 is 60 % higher than for diesel trains,
which makes such a comparison inappropriate.
Deliverable D4 – WP3 passenger 21

Table 5-3b
Energy use and GHG emissions (per passenger-km, at public grid), for highspeed
electric trains, as estimated for new trains by 2050 compared with
reference train, at public grid or fuel tank.
Most
likely
Energy Use
(MJ / pass-km)
LB
UB
GHG Emissions
(g CO 2 -eq / pass-km)
Most
likely
Reference high-speed train (2009) 0.243 31
1. PA Low-drag 0.203 0.18 0.23 26 44 47
6. Pcomb electric (incl incremental) 0.126 0.11 0.15 16 23 32
7. Pcomb (as 6) + higher speed 0.165 0.15 0.20 21 19 25
8. Pcomb (as 7) + low-GHG el mix 0.165 0.15 0.20 4 4 5
Results are presented graphically in Fig 5-2 (energy use) and Fig 5-3 (greenhouse gas emissions),
in both cases aggregated over all types of electric rail passenger services. In Fig 5-2
Scenario 8 with low-GHG electric power is omitted because energy use is equal to Scenario 7.
In Fig 5-3 the combination Scenario 6 (with constant speed) is omitted for simplicity reasons.
2050 2025 2009 reference
LB
UB
Low drag
100.0%
93.8%
89.7%
Low mass
100.0%
97.5%
93.8%
Energy recovery
100.0%
92.2%
87.0%
Space efficiency
100.0%
91.5%
85.8%
Eco drive
100.0%
91.1%
85.2%
All comb
44.3%
66.6%
100.0%
All comb + Higher
speed
53.7%
72.2%
100.0%
Figure 5-2
Estimated trends of energy use (per passenger-km) by technology over time,
including all combinations and higher speeds, aggregated and weighted over
all types of electric rail passenger services.
Deliverable D4 – WP3 passenger 22

2050 2025 2009 reference
Low drag
100.0%
93.8%
89.7%
Low mass
100.0%
97.5%
93.8%
Energy recovery
100.0%
92.2%
87.0%
Space efficiency
100.0%
91.5%
85.8%
Eco drive
100.0%
91.1%
85.2%
All comb + Higher
speed
53.7%
72.2%
100.0%
All comb + Higher
speed + Low GHG el
14.4%
10.7%
100.0%
Figure 5-3
Estimated trends of GHG emissions (per passenger-km) by technology over
time, including all combinations and higher speeds, aggregated and weighted
over all types of electric rail passenger services.
NB: The first six rows assumes a GHG content of EU-27 electricity mix in 2009
(128 g CO2-eq per MJ), while the lowest row assumes GHG to be reduced by 80 %.
Summary
- By far the most efficient single means of reducing greenhouse gas emissions (GHG)
originates outside the railway sector, namely the provision of Low-GHG electric power. In
the example above this reduction is assumed to be 80 % by 2050, although this is a variable
in Stage 2 of Tosca.
- The most efficient means of reducing energy use and GHG emissions within the railway
sector are Eco-driving, Space efficiency, Energy recovery as well as incremental
development in reduced losses. For ‘High-speed’ and ‘Intercity or regional’ trains
aerodynamic design to reduce air drag is also important and for ‘Local city trains’ mass
reduction.
- The combination of different means has the ability of producing a very significant longterm
reduction in energy use and GHG emissions.
- Higher speeds are in the long term expected for electric ‘Intercity or regional’ trains as well
as for ‘High-speed trains’. Some increase in speed is expected also for ‘Local city’ trains.
This will increase energy use, all other factors being equal. However, due to improvements
in technology still very significant reductions of energy and GHG emissions are expected.
Deliverable D4 – WP3 passenger 23

5.3 Cost for reducing GHG emissions
The costs for reducing GHG emissions are dependent on both the additional cost for vehicle
investment and on the change in operating cost (at this stage excluding energy). Estimations
are made with cost models developed at and used by KTH. However, it should be noted that it
is extremely difficult to make thorough and reliable predictions for the year 2050. In particular,
estimates on future sales prices of trains are uncertain. In this study future sales prices are
assumed to be the same in real terms as for the reference year 2009, provided that trains are
equivalent in performance.
Operating costs, other than depreciation and interest cost for trains, are also assumed to be the
same as in 2009 in real terms. Only those changes in operating cost (maintenance, track
charges, etc) that are motivated by changes in technology or average speed are considered in
this study. No considerations are taken to possible effects of a deregulated European rail
market, which would make rail freight services more cost-efficient.
Energy cost is excluded from operating costs, as energy cost is a variable in Stage 2 of
TOSCA. However, to provide a holistic picture, the estimated energy savings are indicated in
the right column of Table 5-4 below.
Assumptions
- Sales price for trains is proportional to suppliers cost for development and manufacture with
the same cost structure as in 2009.
- Depreciation in 25 years, 6 % interest → equal payment (annuity) of 7.82 % each year.
This is practically almost equivalent with 8 % capital cost per year, which is the resulting
capital cost by using the following simple estimate used throughout in TOSCA
C = I · (r + 1/n),
where C = annual capital cost; I = investment; r = average interest rate = 0.04;
n = life time = 25 years.
The above used interest rates are chosen from a long-term socio-economic perspective and
are not including profit margins.
- Tonnage-dependent track charges: 0.003 EUR/grosstonne-km, which is an approximate
average in Europe, according to (Nash, 2005)
- Value of saved train crew time on trains, average per time-tabled hour:
70 EUR (conductors and similar) – 100 EUR (drivers).
- No other improvements than the earlier mentioned technologies for improving energy and
GHG performance are studied, irrespective of their potential for per-unit cost
reductions. Examples of improvement not considered in this context are: higher load
factor, reduced maintenance or improved train crew or rolling stock utilization.
Comparison with road and air transport
To compare with road and air transport, cost characteristics for the rail mode are divided in two
tables:
- Table 5-4a presents cost characteristics aggregated over intercity, regional and city rail
services, for comparison mainly with road transport;
- Table 5-4b presents cost characteristics for high-speed rail service, for an approximate
comparison also with air transport.
Deliverable D4 – WP3 passenger 24

Table 5-4a Cost characteristics, aggregated and weighted over intercity, regional and city
trains (electric and diesel), as estimated for new trains by 2050, compared with
reference trains.
Note: In this table energy cost is excluded from operating cost, but is indicated as the expected
amount of energy consumption per passenger-km (pkm)
Sales price
(MEUR / train
unit)
Varia
Estimated
-tion
Operating cost d
(EUR / pass-km)
Estimated
Variation
Energy e
(MJ
/pkm)
Most
likely
Reference electric trains (2009) a 15.1 ±2 0.098 ±0.011 0.368
1. PA Low drag b 15.4 ±2 0.099 b ±0.012 0.334
2. PB Low mass 16.5 ±2.5 0.099 ±0.014 0.343
3. PC Energy recovery 15.1 ±2 0.096 ±0.012 0.313
4. PD Space efficiency 12.7 c ±2 0.089 ±0.012 0.314
5. PF Eco-driving 15.2 ±2 0.097 ±0.011 0.310
6. Pcomb electric + incremental 14.6 c ±2.5 0.088 ±0.015 0.157
7. Pcomb HS (6) + higher speed 16.5 c ±2.8 0.087 ±0.016 0.187
Reference diesel train (2009) a 12 ±1.5 0.127 ±0.014 0.75
9. Pcomb diesel + incremental 12 ±1.5 0.120 ±0.014 0.37
a
At the reference year electric trains have a total estimated market share of 88 %, and diesel-hauled trains 12 %.
b For high-speed trains it is assumed that a ‘Low-drag’ train will have 3 % less number seats due to longer front
and tail. Other types of trains will not be influenced in this respect.
c Sales price recalculated for the equivalent number of seats as for the reference of 2009. See note c) of Table 5-2.
d Operating costs include capital, maintenance, crew, track & station & dispatch, train formation, sales and
administration. Capital cost excludes profit margins.
e Energy intake to railway’s electric supply system or to fuel tank.
At the reference year 2009, the average cost of electric energy in EU-27 (taken from power grid) is
0.091 EUR /kWh or 0.025 EUR /MJ. For diesel fuel the average cost is 0.015 EUR/MJ.
In both cases taxes are excluded.
Note that the difference in sales price per train unit, as a result of changes in technology, in
most cases is smaller than the normal variation of sales price between different suppliers and
depending on the actual market situation. The estimates should therefore be interpreted as
average prices. Average prices are assumed to follow cost variations, on the assumption that
supplier‘s profit margins will not be changed.
Despite the relatively small variations in sales prices of trains with different energy-saving
technologies, it is anticipated that even these differences must be justified at train acquisition.
Even moderate price differences will be valued in relation to the documented or believed
benefits they offer.
Deliverable D4 – WP3 passenger 25

Cost characteristics over time, aggregated over all types of electrical trains, are presented
graphically in Figure 5-4.
2050 2025 2009 reference
Low drag
Low mass
Energy recovery
Space efficiency
Eco drive
All comb +
incremental
All comb + Higher
speed
100%
101%
101%
100%
100%
101%
100%
99%
98%
100%
95%
91%
100%
100%
100%
100%
94%
90%
100%
93%
89%
Figure 5-4
Estimated trends in operating costs (excl. energy cost) by technology over time,
including combinations and higher speeds, aggregated and weighted over all
types of electric rail passenger services.
Deliverable D4 – WP3 passenger 26

Table 5-4b
Cost characteristics for high-speed train services (electric),
as estimated by 2050, compared with reference train (2009).
Note: In this table energy cost is excluded from operating cost, but is indicated as the expected
amount of energy consumption per passenger-km (pkm).
Sales price
(MEUR / train unit)
Estimated
Operating cost d
(EUR / pass-km)
Variation
Estimated
Variation
Energy e
(MJ /pkm)
Most
likely
Reference high-speed train a 28 ±3 0.063 ±0.010 0.243
1. PA Low-drag b 28.5 ±3.5 0.064 b ±0.010 0.203
2. PB Low-mass --- --- --- --- ---
3. PC Energy recovery 28 ±3 0.063 ±0.010 0.236
4. PD Space efficiency 24.5 c ±3 0.058 ±0.010 0.213
5. PF Eco-driving 28.1 ±3 0.062 ±0.010 0.220
6. Pcomb electric + incremental 25 ±3.5 0.058 ±0.011 0.126
7. Pcomb HS: (6) + higher speed 28 ±4 0.057 ±0.011 0.165
a
At the reference year high-speed trains is estimated to have a market share of 20 % on the rail passenger
market; are however anticipated to increase until 2050.
b For high-speed trains it is assumed that a ‘Low-drag’ train will have 3 % less number seats due to longer front
and tail and also 2 % higher price than the reference train.
c Sales price recalculated for the equivalent number of seats. See note c) of Table 5-2.
d Operating costs include capital, maintenance, crew, track & station & dispatch, train formation, sales and
administration. Capital cost excludes profit margins.
e Energy intake to railway’s electric supply system, from public grid.
At the reference year 2009, the average cost of electric energy in EU-27 (taken from power grid) is
0.091 EUR /kWh or 0.025 EUR /MJ, taxes excluded.
Sales price evolution over time
Rail passenger vehicles are usually developed “on behalf” a specific market or customer, for
series of 100–1000 four-axle vehicles. Development cost is distributed over the actual number
of vehicles. For smaller series the sales price therefore will be higher. Typically, 40 vehicles
instead of 400 typically increase the per-unit price by 15–20 %, but this is of course dependent
on the level of development cost. For early technologies there is also a small “learning” effect,
so that production cost declines over time with each doubling of cumulative output. This effect
is estimated to be 10–15 % over a time span of 10–15 years, but very often new technologies
are more efficient and the older (and more inexpensive) technologies would become obsolete
anyhow. These conditions for the rail vehicle market are anticipated to essentially continue in
the future.
Deliverable D4 – WP3 passenger 27

Break-even price of energy for profitable technology introduction
It is important to understand which energy-saving technologies would be most beneficial in
terms of operational economics, and most likely to be introduced from a train operator’s point
of view. Therefore the necessary break-even price of electricity, where increased operational
costs are balanced by energy cost savings, is calculated for the different types of electric train
operations. Estimated operational costs of new technologies are compared with costs for
reference technologies and with the amount of energy savings.
Results are shown in Table 5-4c. The table shows break-even prices of electricity for profitable
introduction of important technologies, for average electric train operations and for specific
categories of operations where these are believed to produce results diverging from average.
Table 5-4c Break-even price of electricity for long-term profitable introduction of
technologies as estimated by year 2050.
Note: At the reference year 2009, the average cost of electric energy in EU-27 (taken from
power grid) is 0.091 EUR /kWh, excluding taxes.
Profit margin for capital cost is not included.
Average
High-speed
train unit
Electric
(EUR/kWh)
Local
city train
Electric
(EUR/kWh)
Technologies
Electric
(EUR/kWh)
PA Low-drag 0.06 0.08
PB Low-mass 0.14 0.16
PC Energy recovery – 0.15 a – 0.13 a
PD Space efficiency – 0.60 a – 0.64 a NA
PF Eco-driving – 0.07 a – 0.08 a – 0.09 a
a Negative break-even prices of electricity should be interpreted as a beneficial technology with respect to
operational cost, also if energy cost is excluded. It can also be interpreted as the payment to the operator per used
kWh that is necessary to balance the cost of not using the proposed technology.
Deliverable D4 – WP3 passenger 28

Summary and comments
- Some of the investigated technologies do not influence the total operating cost significantly
if energy cost is excluded. There is however a tendency of decreasing operating cost when
Energy recovery or Eco-driving is applied, because of less frequent use of the mechanical
brakes, which leads to reduced maintenance of these brakes. However, on some rail
networks increased energy recovery requires the electrical supply system to be rebuilt or
upgraded in order to be receptive for an increased amount of feed-back electric energy. In
the long term however this is anyhow an anticipated development.
- The most cost-saving measure is to increase space utilization (essentially this is the number
of seats per metre length of train), by changing from loco-hauled trains to multiple units
and/or by increase the space efficiency of interiors.
- Reduced mass per metre of train tends to increase cost if energy cost is excluded. This is
due to higher production cost of the train, and thus investment and capital cost, assumed to
be 20 EUR per kilogram of reduced mass (which is about half of the average price per
kilogram of the whole train). However, this estimate is not confirmed, due to a lack of
reliable sources for long-term estimations. Track charges and train maintenance costs will
be lower for low-mass trains, which partly compensates for higher capital cost. With about
50 % increased prices of electricity (relative to 2009) the ‘low-mass’ train, with assumed
costs of mass reduction, is about to break even in a long-term economic perspective. For
’Local city trains’ with frequent stops, low mass will save more energy and GHG per
average pkm, but the relatively low utilization (km per year) of these trains will limit the
benefits to moderate levels. Due to the current uncertainties on cost and the underdeveloped
technology of low mass trains, a comprehensive R&D program is recommended.
- Low drag is long-term beneficial for ‘High-speed trains’ and also for the ‘Intercity or
regional’ segment, if current energy prices are applied, and still more at higher energy
prices.
- The used interest rate for increased vehicle investment is chosen from a long-term
perspective and is not including profit margins in profit-making operating companies. This
fact indicates that additional incentives – except energy cost savings - may be required for
some technologies.
- Higher speed is usually beneficial for the operator from a cost point of view, due to
increased mileages both for trains and for the train crew. This is in spite of slightly
increasing cost of trains as well as for maintenance and energy. In addition, considerations
must be taken to the increased willingness to pay for shorter travel time; see Section 5.6, in
particular Table 5-8.
Deliverable D4 – WP3 passenger 29

5.4 Scalability
In this section technology-induced limits are considered with regard to the scalability, i.e.
potential market share in EU-27. No limits with regard to the supply of technology at a
sufficiently fast rate are considered.
The percentage market shares are approximate and should be seen as a rough estimate. For
most technologies there is a continuum with regard to the degree they are applied. Further,
what is – for example – considered as “low air drag” is dependent on what type of train this
measure is applied to; for a ‘High-speed train’ the targets are very high, while for an ‘Intercity
or regional train’ the targets are lower, and for a ‘Local city train’ still more modest. Today’s
high-speed trains have lower air drag than the ‘Local city trains’ are expected to have by 2050,
at the same speed. In the example of “low-drag” technology the market share is estimated to be
approximately 75 % as this is the sum of market shares for the rail operations where higher
targets can be motivated (i.e. slow trains are excluded). This example shows that no exact
definition of scalability is possible.
However, it is anticipated that the different technologies are implemented to approximately the
same amount (± 1/5) as shown in Table 3-1 in Section 3.5.
Table 5-5 Scalability Characteristics, i.e. maximum possible market share by 2050.
The table refers to the fraction of the market (in passenger-km) in which it
could be motivated.
Percent EU-27 Market
Most Likely LB UB
PA Low drag 75 50 90
PB Low mass 30 15 50
PC Energy recovery 90 70 100
PD Space efficient 75 40 80
PE Modular short train 20 10 40
PF Eco driving 95 80 100
PG Dual mode 5 2 10
PH Bio fuels 5 1 12
PI Electrification 6 4 10
PJ Low-GHG electric power 80 50 90
Higher speed 85 85 85
Deliverable D4 – WP3 passenger 30

5.5 Social acceptability
Table 5-6
Social acceptability
Rates from ‒ ‒ “unacceptable adverse effect” to + + “significant benefits”;
0 being “comparable to reference system”.
Values within (parentheses) indicate that counter-measures are expected to be taken in order to
produce a neutral outcome.
Social equity
implications
Generation
within EU
of jobs
Passenger safety
Noise
Privacy
Ethical issues
PA Low drag 0 + 0 0 (+) 0 +
PB Low mass 0 + 0 0 (+) 0 +
PC Energy recovery 0 0 0 0 0 ++
PD Space efficient 0 + 0 0 - ++
PE Modular short train 0 0 0 + 0 +
PF Eco-driving 0 0 0 0 0 ++
PG Dual mode 0 + 0 0 (+) 0 +
PH Bio fuels -to + 0 to + 0 0 0 - to +
PI Electrification 0 + 0 0 (+) 0 +
PJ Low-GHG electric power - to 0 0 0 0 0 0 to +
Higher speed 0 0 0 (-) 0 (-) 0 0 (-)
Comments
- Most technologies and measures are neutral or not far from neutral with respect to social
acceptability, at least after that counter-measures have been taken.
- All technologies that contribute to the competitiveness of the European rail industry (in this
case: suppliers of rail vehicles and infrastructure in relation to non-EU suppliers) are
assumed to be positive for job creation or at least job survival.
- Technologies that contribute to reduction of greenhouse gases (GHG) have an inherent
positive ethical value. Regarding bio fuels however this is uncertain, due to possible
conflicts with land use and food production; see WP4 report.
- In a narrow sense higher speed contributes to higher GHG emissions because of higher
energy use, for the same type of train. However, the historical trend is that trains are
adapted to the intended use, so higher speeds will force a faster and more ambitious
development of energy-saving features, in particular low air drag. Moreover, if higher train
speeds contribute to a higher competitiveness of rail operations, and these ‘higher speed’
operations have comparative benefits from a GHG point of view, the ‘Higher speed’ factor
would be even more positive.
Deliverable D4 – WP3 passenger 31

- Higher speed could in a narrow sense also be considered as more unsafe than lower speed,
as the consequences of an accident would most likely be more serious. However, highspeed
operations (say from 200 km/h and above) have since their introduction in the 1960´s
shown a superior world-wide statistical record on safety (Grimvall et al, 2010). This is due
to a number of safety measures that are taken when higher speeds are introduced. These are
included in the cost estimations of Section 5.4. Therefore higher speed is expected to be at
least as safe as “slower speed”.
- Noise could in a narrow sense be reduced or increased as a result of the introduction of new
technologies or speeds. However, external noise emissions are regulated by law. Therefore
appropriate measures must be taken to limit noise to the prescribed level, thus maintaining
noise levels at an essentially constant level, independent of technology and speed. The
resulting effect will therefore be more or less countermeasures, which are included in the
cost estimations of Section 5.3.
- Privacy could be negatively affected by a too tight layout of seats in Space-efficient trains.
Passengers’ privacy must, as well as comfort and working-ability issues, be taken into
consideration when space-efficient train interiors are created.
5.6 User acceptability
Acceptability of the different technologies to the end users (passengers) is rated and presented
in Table 5-7. User cost is assumed to reflect costs in Table 5-4, i.e. the rail operators will
transfer changes in the operating cost to the passenger, at least in the long term in a competitive
market situation.
Table 5-7
User acceptability at given user cost.
Ratings are from 0 (no adoption) to 5 (full adoption).
User acceptability
at given
user cost
PA Low drag 5
PB Low mass 4
PC Energy recovery 5
PD Space efficient 3
PE Modular short train 5
PF Eco driving 5
PG Dual mode 5
PH Bio fuels 3
PI Electrification 5
PJ Low-GHG electric power
Higher speed 5
a Technologies and acceptability of low-GHG electric power is outside the
railway sector and is therefore outside the scope of this particular study.
a
Deliverable D4 – WP3 passenger 32

Comment
- Most technologies and measures are expected to be fully accepted, although there are some
uncertainties on PB Low mass, due to the slightly higher cost that would occur. Cost is
however not confirmed at this stage. Some bio fuels (in particular vegetable oil) would
likely be subject to resistance because of their higher cost and possibly also conflict with
food production. Low-GHG electric power may partly also be subject to resistance if more
nuclear or wind power plants are to be built.
End user’s willingness to pay is presented in Table 5-8.
Assumptions
- Value of saved passenger time: 11 EUR per hour, constant until 2050.
- Value of leg space at the seat: 6 % of fare per 10 % of increased experienced space
(Kottenhoff, 2009).
Table 5-8
End user willingness to pay, as estimated by 2050, aggregated over all types of
passenger rail services. LB = lower bound; UB = upper bound.
End user is the average passenger.
+ means: how much additional fare would end user be willing to pay
– means: how much should end user be paid to fully adopt the technology
Most
likely
Reference electric trains ±0
Willingness to pay
(EUR / pass-km)
1. PA Low-drag a ±0 0 +0.003
2. PB Low-mass ±0 0 0
3. PC Energy recovery ±0 0 0
4. PD Space efficiency b -0.0025 0 -0.004
5. PF Eco-driving ±0 0 0
6. Pcomb electric (incl. incremental) -0.0025 0 -0.004
7. Pcomb (6) + higher speed +0.011 +0.010 +0.012
8. As 7. + low-GHG electric mix +0.011 +0.010 +0.012
LB
UB
Reference diesel trains ±0
9. Pcomb diesel -0.0025 0 –0.004
a
A streamlined appearance would have a positive impact on passenger’s perception of the train,
which likely induces a slightly higher willingness to pay. However, variations in appearance perception
are not included in other Work Packages of TOSCA, so this consideration is excluded in the subsequent analysis.
b For the ‘Intercity or regional’ segment it is assumed that loco-hauled trains are replaced by multiple-unit trains.
In addition, for all trains except ‘Local city trains’ it is assumed that 2/3 of improved space efficiency in interiors
is achieved by developed smart design and will therefore not influence passenger’s perception and willingness to
pay negatively. However, it is assumed that the perception of space (leg-room) is reduced by 5 % until 2050,
which will induce a negative willingness to pay (Kottenhoff, 2009).
Deliverable D4 – WP3 passenger 33

6 SUMMARY AND CONCLUSIONS
It is naturally a very difficult task to estimate and foresee viable and plausible technologies that
could likely be implemented in the long term from the reference year 2009 until 2050. In the
rail sector there is a lack of qualified estimations on technology in the long term. The approach
used in this study is to divide the total 41 year period in two periods, with the first period until
2025, and the second from 2025–2050, with the following assumptions:
Period 1
Period 2
The techno-economic potential for in-service introduction during the next
10–15 years (i.e. until 2025) is estimated.
Development is assumed to continue after 2025. It is assumed that 2/3 of the
first period achievement can be achieved in the second period, i.e. 2025–2050.
The exception is technology PB (Low-mass train) that requires more substantial
research before technology readiness (see Table 3-1), which will delay most of
its possible introduction to Period 2.
This means that almost 60 % of the total achievement is what today is known or assumed as
appropriate technologies, possible to introduce in regular service under the first period, while
the rest is assumed to be achieved under the second period.
This stepwise methodology is used in order to minimize uncertainties as far as possible, by
basing assumptions on what is recently known or believed to be appropriate technologies in the
intermediate term. The assumption that development will continue during Period 2 is plausible,
although the exact rate per year or decade is uncertain. Developments will in reality likely be
different for various technologies.
6.1 Summary
According to the analysis in Sections 2 to 5, the technologies and measures described below
are the most promising in order to reduce energy use and greenhouse gas (GHG) emissions
with economically viable means. They are presented in the order of their estimated potential
for GHG reductions in new trains by 2050. The first seven technologies and means presented
below are applied as a single mean with all other conditions being equal to the reference trains.
Note that energy cost is excluded from the estimated economical savings, because energy cost
is a variable in Stage 2 of TOSCA. The resulting total cost savings will therefore be higher
than indicated below.
PJ Low-GHG electricity
With a prospected GHG reduction of 50–90 % until 2050, and a possible market penetration on
all electrified rail services, this is without doubt the most promising single technology. This
technology – or combination of different technologies – is developed and provided outside the
railway sector but has nevertheless a large positive impact. Low-GHG electricity must
therefore not be neglected in the evaluation of the future GHG performance of rail operations.
Estimated overall GHG reduction from electric rail operations: 50–80 %
impact on operating cost, excluding energy: 0 %
end user average willingness to pay: 0 %
Deliverable D4 – WP3 passenger 34

PF Eco-driving
With an estimated average GHG reduction of 15 % by 2050, as a single individual measure,
and a market penetration in almost every type of rail operations this is the second most
important technology, and the most important technology generated in the rail sector itself. It is
to some extent used by train drivers already today but can be significantly improved by using
computer-assisted advice (or in some cases automatic train operation). It is a low-cost
technology and has no adverse effects on society, operators or end users. Operating cost will be
reduced according to energy savings, although not included at this stage of TOSCA. Ecodriving
will likely also lead to reduced maintenance on train braking equipment.
This technology will be most efficient in frequently stopping ‘Local city train’ operations,
where it is estimated to reduce energy and GHG emissions by about 20 %.
Estimated overall GHG reduction from rail operations: 15 %
impact on operating cost, excluding energy: -1 %
end user average willingness to pay: 0 %
PD Space efficiency
Improved space efficiency, i.e. an increased number of seats per unit length or unit mass of the
train, can be achieved in two ways; (1) by replacing the conventional train consist of
locomotive + cars with multiple unit trains, having the propulsion in the same cars as
passengers, or (2) by improving space utilisation within the car’s interior through smart and
space-saving solutions, while still being convenient for passengers.
This technology is applicable to ‘High-speed’ as well as ‘Intercity or regional’ operations,
while no further improvement is anticipated on ‘Local city’ trains.
Estimated overall GHG reduction from rail operations: 14 %
impact on operating cost, excluding energy: - 9 %
end user average willingness to pay: - 3 %
PC Energy recovery
Energy recovery is used already today in many electric train operations, although not in all.
Due to limitations in the share of powered axles, in the propulsion and electric braking power
as well as in the receptivity of the electric supply system, the energy recovery is less than
optimum. In diesel-hauled trains there is no energy recovery at all, but this could change when
energy-storing technologies are further developed. Besides energy saving, energy recovery will
have a side-effect through reduced wear on the mechanical brake equipment, which will reduce
maintenance cost. It is important that at least about 50 % of the axles of the train are powered
and that the installed power for propulsion and electric braking is high, in order to guarantee a
high enough braking effort at electric recovery braking. The assumed power in this study is at
least 20 kW per tonne of train mass for electric trains.
This technology is applicable to all types of train services, but is most efficient on stopping
train services such as ‘Local city trains’. As a single measure it can save another 20 % of used
energy, compared with today’s reference city trains. ‘Intercity or regional’ operations have an
estimated potential for 10–12 % energy savings as an average. The full application of this
technology needs upgrading of the electric supply system on some rail networks.
Estimated overall GHG reduction from rail operations: 13 %
impact on operating cost, excluding energy: - 2 %
end user average willingness to pay: 0 %
Deliverable D4 – WP3 passenger 35

PA Low-drag
In absolute terms low air drag is most important for high-speed trains, with an estimated longterm
reduction of energy use by 17 % when applied as a single measure. The potential is large
also for ordinary trains, such as ‘Intercity or regional’. For low-speed stopping trains low drag
is less important due their low average speed. Operating cost will in most cases be reduced
according to energy savings. Further, a streamlined train design is believed to have a positive
impact on passenger’s perception of the train as a modern mean of transportation, although no
quantitative estimation is proposed in this study.
Estimated overall GHG reduction from rail operations: 10 %.
impact on operating cost, excluding energy: + 1 %
end user average willingness to pay: 0 %
PB Low-mass
Train mass per seat can be reduced in two basic ways; (1) by conceptual improvements, such
as using multiple units (with propulsion power in the same cars as passengers) instead of
locomotive-hauled trains, or by using double deckers or wide-body trains, or (2) by reducing
mass of the basic train technology by using new materials and weight-optimized structures.
Mass reduction is an effective mean of reducing energy use and GHG emissions for frequently
stopping train such as ‘Local city trains’, where a mass reduction of 20 %, as a single mean, is
estimated to reduce energy use by about 10 %. For other types of train services the energy
savings are much more moderate.
Ambitious mass reductions that requires radical changes in the basic train technology, could
lead to increasing cost for rail vehicles that would be prohibitive for their introduction. In this
study assumptions for a modest cost increase are made, however the real future cost may either
be higher or lower.
Estimated overall GHG reduction from rail operations: 6 %.
impact on operating cost, excluding energy: + 1 %
end user average willingness to pay: 0 %
Incremental improvements
Incremental continuous improvements in energy use are expected to continue, with an
estimated reduction of total energy consumption by 12 % in the long term until 2050. In
particular, the energy losses in train propulsion and electric supply systems are expected to be
reduced by 25–30 %. (i.e. from 26% to about 18 % of total input energy). This will be
beneficial not only for energy intake, but also for energy recovery. The same relative
improvements (25–30 %) are expected in auxiliary and comfort energy use, facilitated by
recovery of heat, better controlled ventilation, heat pumps etc. There are also other options for
continuous incremental improvements although not analysed in detail.
Estimated overall GHG reduction from rail operations: 12 %.
impact on operation cost, excluding energy: + 0 %
end user average willingness to pay: 0 %
Deliverable D4 – WP3 passenger 36

Combination of measures
The combination of all measures produces very significant reductions in energy use. Even if
we exclude ‘Low-GHG electric power’, the estimations and simulations end up with a
reduction of energy use and GHG by about 56 % on average, with the largest reduction for
stopping ‘Local city trains’ at 60 %. This is a result of both less intake of energy and increased
energy recovery, so that the resulting net energy is heavily reduced. Note that train speeds are
assumed to be constant at this first stage of estimation.
Estimated overall energy use and GHG reduction,
excluding effect of higher speed and ‘Low-GHG electric power’: 56 %
impact on operation cost, excluding energy: –10 %
end user average willingness to pay: –3 %
Higher speeds and combinations
With all other conditions being equal higher speeds will increase energy consumption and the
resulting GHG emissions. These negative effects are however expected to be essentially
compensated by adapting technologies to increased speed. Shorter travel time will reduce
operating cost due to higher productivity of trains and train crew. Most important, shorter
travel time will increase willingness to pay. In all, trains will be more competitive. Also, from
a socio-economic perspective shortened travel time will in many cases have benefits.
The major drawback is the need for a high-performing rail infrastructure, which in certain
cases would trigger a negative public opinion and usually require public funding. It should
however be noted that higher speed and improved infrastructure does not only apply to veryhigh-speed
rail (top speed 250 km/h and above), but also upgrading of conventional railways
for top speeds lower than 250 km/h.
Estimated overall GHG reduction, excluding effect of ‘Low-GHG electricity’ 46 %.
impact on operation cost, excluding energy: –11 %
end user average willingness to pay: +12 %
Incentives needed
The used interest rate for increased vehicle investment is chosen from a long-term perspective
and is not including profit margins in profit-making operating companies. This fact indicates
that further incentives – subsidies or penalties – in addition to energy cost savings, may be
required for some technologies, in particular for ‘low-drag’ and ‘low-mass’.
6.2 Conclusions
In the present analysis it is estimated that a number of efficient technologies, individually and
in combination, are available in order to significantly reduce energy use and the resulting GHG
emissions on the rail passenger market until 2050.
The analysis has considered different technologies and means
– reduced air drag
– reduced train mass
– energy recovery
– eco-driving, including traffic flow management
– space efficiency in trains
– incremental improvements in energy efficiency, in particular reduced losses.
Deliverable D4 – WP3 passenger 37

Despite higher train speeds in the future, these technologies will, according to the estimations,
reduce energy use and GHG emissions by 45–50 % in the long term until 2050.
By far the most efficient mean of reducing GHG emissions from rail operations is however
low-GHG electricity, that is e.g. a number of technologies to be introduced outside the railway
sector; the end result being dependent on the degree of GHG reductions in the future European
mix of electricity. This is expected to further reduce per-unit GHG emissions by far more than
the 45–50 % mentioned above.
With a 60-80 % reduction of GHG content of future average European electricity, which is
approximately within the different scenarios of TOSCA WP6, the resulting GHG emissions are
estimated to 4–11 g CO2-eq per passenger-km, which is very favourable in comparison to
other passenger transport modes.
The results might be seen as optimistic, but they are due to the combination of a number of
efficient tecnologies, which to some extent also reinforce each other. It should also be pointed
out that some options are not analysed in detail at this stage, and are not included in the above
estimations, i.e.
- increased average seat occupancy rate (load factor) on trains
- modular trains with capacity according to actual need.
- dual-mode operations (diesel and electric) or hybrid vehicles
- use of bio fuels
- additional electrification of railways or increased investment in rail infrastructure
These additional options would further strengthen the potential for reductions in energy use
and GHG emissions in future rail operations. These means can be considered as a “reserve” in
the case that some of the studied technologies and means would not be fully implemented, for
one reason or the other. Three of these additional options are applicable to diesel passenger
operations, which have a modest market share (about 12 % in EU-27). Therefore most of these
additional options (with exception of improved load factor) will have a limited impact on total
GHG emissions from railways (roughly estimated to 10-15 % of resulting GHG after other
reductions), however having the potential to substantially reduce GHG emissions from diesel
operations specifically. Including also improved load factor (applicable to most rail operations)
an additional GHG reduction 10–20 % may be possible. These figures are however highly
uncertain..
The option of magnetic levitation is excluded from further analysis, since it is considered not to
be an efficient mean of reducing energy use and GHG emissions.
Substantial research is needed for low-mass trains, to assure that the mass reductions are made
accurately to an acceptable cost. Partly substantial research is necessary also for reduced air
drag and eco-driving.
It is anticipated that society and users would not present any strong positive or negative
attitudes to most of the proposed measures. Public resistance would, however, from time to
time occur against new railway undertakings, in particular if new railway links are being built.
The main restriction is the need for rail transport capacity and improved overall performance,
and thus funding for improvement of the rail infrastructure. If a substantial mode shift to rail is
envisaged, a comparatively large share of the total transport investment resources should be
directed to the rail system. New high-speed railway links are needed in addition to
improvements of the existing rail systems.
Deliverable D4 – WP3 passenger 38

REFERENCES
Andersson E, Lukaszewicz P (2006): Energy consumption and related air pollution for
Scandinavian electric passenger trains. Stockholm, KTH/AVE 2006:46.
Diedrichs B: Personal communication with aerodynamic specialist at Bombardier
Transportation, 2010.
EU (2009): EU energy and transport in figures.
Grimvall G, et al (Eds.): Risk in Technological Systems. ISBN 978-1-84882-641-0, London
2010.
Kemp R, Smith R (2007): Technical issues raised by the proposal to introduce a 500 km/h
magnetically-levitated transport system in the UK. Lancaster University & Imperial College,
London.
Kottenhoff K, Andersson E: Attractive and Efficient Train Interiors. KTH Railway Group,
Publ 0903, Stockholm, 2009.
Lindemann M: Evaluation report of the saving potential by using the Drivers Assistant System
including test campaigns. Deliverable D4.3.3 of Railenergy, Bombardier Transportation,
Germany, 2010.
Luijt R S: ECO-driving – the implementation challenge. International conference “Railways
and Environment”, Delft, The Netherlands, 16-17 Dec 2010.
Lukaszewicz P: Energy Consumption and Running Time for Trains. KTH Railway
Technology, ISRN KTH/FKT/D-01/25-SE, Stockholm 2001.
Lukaszewicz P, Andersson E (2009): Green Train energy consumption. KTH Railway Group,
Publ 0901, Stockholm 2009.
Nash C: Rail infrastructure charges in Europe. Institute of transport studies, University of
Leeds.
Perimenis A, Majer S, Holland M, Mṻller-Langer F: Life cycle assessment of transportation
fuels. Deliverable D5 (WP4 report) of TOSCA, DBFZ, Leipzig, 2010.
Railenergy (EU FP6 project): Energy Efficiency days. Tours, 23–26 Sep 2009.
http://www.energy-efficiency-days.org
RSSB (Railway Safety & Standards Board, London) (2007): Traction energy metrics.
SIKA (Swedish Institute for Transport Analysis) (2009): Bantrafik 2008 (eng: Rail transport
2008), Report 2009:22.
UIC (International Union of Railways, Paris 2010): Energy efficiency strategies for rolling
stock and train operation. www.railway-energy.org./tfee.
UIC (International Union of Railways, Paris) (2008): Rail Transport and Environment – Facts
and figures.
Deliverable D4 – WP3 passenger 39