High Speed, Energy Consumption and Emissions

High speed, energy consumption
and emissions
Study and Research Group for Railway Energy and emissions
Author: Alberto GARCIA
-------------------------------------------
Company: FFE
Fundacin Ferrocarriles Espaoles
-------------------------------------------
Document: 1
st
delivery
-------------------------------------------
Date: 21 December 2010
RENFE Foto: Patier
INTERNATIONAL UNION OF RAILWAYS (UIC)
16 rue Jean Rey - F-75015 PARIS
Tel: +33 (0)1 44 49 20 20
Fax: +33 (0)1 44 49 20 29
PASSENGER DEPARTMENT - HIGH SPEED ACTIVITY
Legal deposit: November 2010
ISBN 978 2 7461 1900 0
www.uic.org

High speed, energy consumption and emissions

This report is in response to the
contract signed on January 20,
2010 between the Union
International des Chemins de Fer
(High Speed Committee) and the
Fundacin de los Ferrocarriles
Espaoles (FFE), the aim of
which is to conduct a study
entitled "High Speed, Energy
Consumption and Emissions.

The international Union of Railways
The International Union of Railways (UIC known by the acronym of the French Union
Internationale des Chemins de Fer) is the global partnership for cooperation among
key international actors in the railway sector.
Founded in 1922 with the aim of moving towards standardization and improving
systems of railway construction and operation of interoperable, currently holds within it
to 171 members, including national railways, operators, infrastructure managers, public
transport companies and others.
In recent years the organization has redesigned its objectives and has put special
emphasis on issues such as liberalization and globalization of the global rail sector, or
the new challenges arising in the rail its key role in a stage of sustainable development
and combating climate change.
The main aims of the UIC High Speed Department are to coordinate high speed
activities carried out by UIC members and to contribute to logical development of high
speed systems. It performs the following activities: updating databases (lines, rolling
stock, traffic, etc.), high speed world maps, benchmarking and other working teams,
communications and contacts, website and high speed brochures as well as other
publications.

Fundacin de los Ferrocarriles Espaoles (Spanish Railway
Foundation)
The Fundacin de los Ferrocarriles Espaoles (FFE, Spanish Railway Foundation) was
created on February 20, 1985 by RENFE (National Network of Spanish Railways)
currently split into Renfe Operadora and the Administrador de Infraestructuras
Ferroviarias (ADIF, Administrator of Railways Infrastructures)- and Ferrocarriles de Va
Estrecha (FEVE, Narrow Gauge Railways). Since 2002 it has been a National Public
Sector Foundation.

The mission of the FFE is to promote knowledge and use of the railway through:
research and education, technological services, the recovery and alternative use of the
railway heritage, cultural activities, periodic publications and specialized books. It also
manages the Library, the Railway History Archive and the Railway Documentation
Centre, as well as the Madrid Delicias and Vilanova i la Geltr (Barcelona) Railway
Museums.
The Spanish Railway Foundations Board of Trustees currently comprises: Renfe
Operadora; Administrador de infraestructuras Ferroviarias (ADIF), Ferrocarriles de Va
Estrecha (FEVE), Ferrocarriles de la Generalitat de Catalunya (FGC), Ferrocarriles de
la Generalitat Valenciana (FGV), Eusko Tren; Metro de Madrid S.A., Ferrocarril
Metropolit de Barcelona S.A., Asociacin de Empresas Constructoras de mbito
Nacional (SEOPAN), MAFEX, Serveis Ferroviaris de Mallorca (SFM), Euskal Trenbide
Sarea, Ferrocarriles de la J unta de Andaluca (FJ A); Asociacin Ferroviaria de
Certificacin, CETREN and INECO.
Department of Research, Training and Cientific Cooperation
The mission of the Department of Research, Training and Scientific Cooperation
involves performing activities aimed at promoting the development of the railway,
providing the necessary tools for education, promoting the positive image of the
railway, and carrying out studies and research projects to serve the interests of the
railway sector as a whole, both at home and abroad, and especially in Ibero-American
countries.
Its main lines of action are: studies and research projects, sectorial promotion and
cooperation, training and communication activities.
The own Departments research activity is carried out by the following study and
research groups:
1. Study and research group for energy and emissions in transport
2. Study and research group for economics and transport operation
3. Study and research group for geography and rail traffic
4. Study and research group for the sociology of transport
In recent years, these groups have worked on many different projects (some completed
and others still in progress), the most important being: EnerTrans, ElecRail, Reactiva,
Observatorio del Ferrocarril Espaol (Spanish Railway Observatory), Securemetro,
Hope in Stations, Viadintel, Aeroave, Balasto artificial (Artificial Ballast), etc.

Table of Contents
Table of Contents ......................................................................................................................................................... 5
1. Introduction and background ........................................................................................................................... 7
1.1. The aims of this report ................................................................................................... 7
1.2. Background .................................................................................................................... 8
1.3. Scope of the study ......................................................................................................... 8
1.4. Assumptions and general comments ............................................................................. 9
1.4.1. Electric traction and emissions .......................................................................................................... 9
1.4.2. General comments ............................................................................................................................ 9
2. Does high speed require a lot of energy? ..................................................................................................... 10
2.1. The problem in the www .............................................................................................. 10
2.2. The square and cube rules ...................................................................................... 10
2.3. Lack of information about the energy consumption of the high speed train? ............... 11
3. Comparison of the high speed railway systemand the conventional systemin terms of energy
consumption ............................................................................................................................................................... 14
3.1. Empirical verification .................................................................................................... 15
3.1.1. Empirical relationship between energy consumption and speed .................................................... 15
3.2. Articles based on specific cases .................................................................................. 18
3.2.1. The case of different trains on different lines .................................................................................. 18
3.2.2. The case of different trains on the same line .................................................................................. 20
3.2.3. The case of the same train on different lines .................................................................................. 20
4. Technical reasons for lower consumption.................................................................................................... 21
4.1. Addends of the consumption function .......................................................................... 21
4.1.1. Energy needed to overcome mechanical resistance on straight track and on curves .................... 21
4.1.2. Energy needed to overcome intake air resistance in the train ........................................................ 23
4.1.3. Energy needed to overcome aerodynamic drag ............................................................................. 23
4.1.4. Energy dissipated by the brake and not utilized .............................................................................. 24
4.1.5. Energy consumed by the trains auxiliary services ......................................................................... 28
4.1.6. Multiplier factors in the consumption function ................................................................................. 28
4.1.7. Energy loss coefficient .................................................................................................................... 28
4.1.8. Shorter distances between the same points ................................................................................... 30
4.2. Coefficients and exponents of the consumption function ............................................. 31
4.3. The effect of speed ceteris paribus (all other things being equal) ................................ 33
4.4. Energy used in constructing the vehicle ....................................................................... 34
4.5. General view of the speeds effect and comparisson with conventional railway .......... 37
5. Analysis of the effect of the shift of traffic fromother modes to high-speed rail ..................................... 39
5.1. Modal spilt HS-plane .................................................................................................... 39
5.2. Optimization of speed from the energy consumption point of view .............................. 40
6. Best practices and recommendations in high speed to reduce consumption and emissions ................ 45
6.1. General recommendations (not only for high speed) ................................................... 45
6.2. Best practices applicable to high speed ....................................................................... 46
6.2.1. Increase of speeds on gradients (downgrades) .............................................................................. 46
6.2.2. Optimization of the trains exterior dimensions ............................................................................... 48
6.2.3. Air intake reduction ......................................................................................................................... 49
7. Bibliography ..................................................................................................................................................... 51


2010

7
1. Introduction and background
The purpose of the study is to identify and quantify energy and environmental
characteristics of high speed trains.
The study only covers three variables:
1. Final energy consumption.
2. Consumption of primary energy or fossil fuels.
3. CO
2
emissions.
Energy differences and advantages of conventional trains in relation to other
modes of transport will be observed first, therefore identifying factors inducing
train consumption and differences according to train and lines.
The characteristics of the high speed system with respect to conventional trains
will be analysed next, checking to see whether these accumulate or become
diluted with increasing speed, and whether their source is the infrastructure, the
vehicle or the operation.
Thirdly, we will analyse the effect of the shift of traffic from other modes of
transport to high speed trains for varying speeds, checking in each case the
effect of consumption on the whole corridor.
1.1. The aims of this report
This report analyses the energy consumption of high speed trains and seeks to
answer the following questions:
a) Does the high speed system offer the same general advantages as
conventional trains?
b) Does the high speed system entail an increase in energy consumption,
per passenger, in comparison with the conventional railway system?
c) Does the high speed railway system require less energy, per passenger
transported, than other modes of transport?
d) If it requires less energy, what is the order of magnitude of the
differences?
e) Does the high speed systems appearance on a route entail an increase
or a reduction in the energy consumption for transportation purposes,
taking into account the induced demand?
f) What is the effect of speed on the distribution of the energy used for the
vehicles construction during its operating life?
In short, the aim is to analyse whether the expansion of the high speed network
is going to exacerbate the problem of emissions and energy consumptions in
the transport sector; or whether, on the contrary, the high speed railway system
is more energy-efficient than other alternative modes and, therefore, investment
2010

8
in high speed lines and trains will help improve the sustainability of the transport
system.
1.2. Background
Thoughts on the widespread opinions regarding the energy consumption and
emissions of high speed trains, as well as the evidence which largely
contradicts such generally held beliefs, have been published in various previous
articles written by members of the Study and Research Group for Railway
Energy and Emissions of the Spanish Railway Foundation.
These publications also contain most of the points covered in greater depth in
this report, which relate to the reasons for the differences in consumption
between conventional trains and high speed trains, as well as the effect that
introducing the high speed train in a corridor has on energy consumption and
emissions.
Besides expanding on the aspects dealt with in previous publications, this report
also adds a number of considerations and data relating to the effect of speed on
the amount of energy needed (and on the emissions produced) to manufacture
and recycle the train (including the materials incorporated into the train).
Another important new feature of this report is the general outline of a method
used to quantitatively determine the optimum speed of the high speed train
from the energy point of view.
Both subjects constitute initial outlines which, as has already been done with
the topics dealt with previously, would be worth exploring in greater detail in
subsequent studies.
1.3. Scope of the study
The study aims to have worldwide applicability, given that there are no major
differences between countries in terms of the aspects covered in the report.
Nevertheless, the study uses data corresponding to the Spanish case,
eliminating -obviously- considerations that stem from purely local aspects such
as the difference in track gauge between the conventional network and the high
speed network. The use of Spanish data offers two advantages:
a) Firstly, all the data used come from the same system (which also
provides the authors of the report with abundant data and information),
thus avoiding the dysfunctions caused by the use of combined data with
an unknown degree of homogeneity.
b) Secondly, the Spanish case includes a unique variety of train types (of
different technological origin, architectures and sizes), types of service
(long distance, high speed regional trains, overnight trains and trains
which partly run on conventional lines), infrastructures (new high speed
lines with very different maximum speeds, diverse speed profiles,
upgrades, downgrades and unequal path coefficients.
2010
9

1.4. Assumptions and general comments
1.4.1. Electric traction and emissions
This report will focus solely on electric traction trains, given that this traction
system is the only one that permits high speed running in the sense defined
herein, and in order to make homogeneous comparisons it is also necessary to
refer to electric traction in the conventional railway system that is to be
compared with its high speed counterpart.
Otherwise (i.e. if the analysis referred to conventional trains with diesel traction),
some of the differences observed would reduce the clarity of the results.
Furthermore, all the high speed trains and over 80-90% of the passenger trains
currently used in continental Europe are electric traction vehicles.
By focusing solely on electric traction, we can analyse energy consumption
alone, given that the associated greenhouse gas emissions can immediately be
deduced from this energy consumption, since the applicable emission factor is
fixed for each year and for each country. Therefore, the train that consumes
less energy will be the one that emits less CO
2
, and precisely in the same
proportion.
Greenhouse gas emissions will only be analysed when the railway is compared
with other modes of transport, most of which use non-electric energy (petroleum
products), and there are two reasons for choosing this indicator to measure
energy efficiency: the importance of reducing emissions of these types of
gases, and the fact that it provides a clear indication of transports contribution
to the exhaustion of non-renewable energy sources (oil, gas and coal).
1.4.2. General comments
It should be pointed out that are no absolute, or universally valid, truths.
This study analyses specific cases, corresponding to the European case, in
which 30 years experience of high speed operation has provided conclusive
data. In other types of operation, in other geographic environments, or at other
moments in time, the results might be different.
It is also worth emphasising that the high speed train is not a substitute for the
conventional train, but instead they should seek complementarity. They can and
should coexist, and their respective infrastructures and services should be
planned in such a way as to ensure that they complement and support each
other.
During the course of this analysis we will frequently refer to the train (in order
to use a common and easily comprehensible terminology), but in doing so we
are actually referring to the railway as a whole, as a system.

2010

10
2. Does high speed require a lot of
energy?
There is undoubtedly a widespread belief among the general public that the
high speed train requires a lot of energy in order to operate, even an excessive
and disproportionate amount considering the benefits it provides.
This common misconception serves as a pretext to attack the high speed train
and propose, as an alternative to constructing new high speed lines, the
improvement of conventional lines, or simply the strengthening of other modes
of transport.
2.1. The problem in the www
An example of the state of opinion on this subject and how these ideas spread
through society can be found in some of the entries that Spanish schoolchildren
submitted to the Railway Museums high speed competition in 2006. In order to
find the information they needed, the youngsters searched the Internet and
came away with the ideas they felt were most suited to the task.
Several entries coincided in reproducing two statements, one of which we have
taken directly from an entry submitted by 4
th
year students at the Gustavo
Adolfo Becquer secondary school in Algete (Madrid): The high speed train
consumes a disproportionate and unsustainable amount of energy. The AVE
that runs between Madrid and Seville at a maximum speed of 300 km/h (at an
average of 209 km/h) has a power output of 8,000 kW: it consumes as much
electricity, measured in kilowatts/hour, as a city of 25,000 inhabitants (sic).
Along a very similar line of argument, the 1
st
year students at the Cardenal
Herrera Oria secondary school in Madrid, citing ecologist sources, claimed
that increasing a trains speed from 100 to 400 km/h means that the power has
to be multiplied by 64. Yet besides its huge energy consumption (), it is clear
that the AVE is linked to the extremely high consumption of electricity and,
therefore, to nuclear energy.
2.2. The square and cube rules
There are two ideas relating to the links between energy and the railway
operation of trains that are very widespread, even among many engineers and
railway experts.
a) The first idea, which we could call the power rule or cube rule, is that
the power of the train would increase with the cube of its speed.
b) The second idea, which we could call the energy rule or square rule,
indicates that energy consumption would increase in proportion to the
square of the speed. This rule would correspond to a consumption
induced solely by quadratic resistances, such as those of an
aerodynamic nature.
2010

11
2.3. Lack of information about the energy consumption of
the high speed train?
Roger Kemp, a professor of engineering at the University of Lancaster (Great
Britain), has repeatedly insisted on the energy-related harmfulness of the high
speed train, with affirmations such as the one contained in this provocatively
titled article Take the Car and Save the Planet (Kemp, R. 2004) and in The
European High Speed Network (Kemp, R., 1993), in which he obtains various
comparative results that do no favours at all for the high speed train.
The quantitative basis of this professors claims is the consumption attributed in
1993 to a hypothetical high speed train running at different speeds between
London and Edinburgh.
The graph he presented in this study shows an increase closely related to
speed, together with some very high energy consumption values.

Figure 1. Comparison between the journey time and the energy consumption of a train
between London and Edinburgh for various maximum speeds. Energy is expressed in
kWh/seat and the journey time in hours. Source: Kemp (1993)
With an energy use per seat of 57 kWh for the whole journey, a joint train and
transmission efficiency of 65%, and with a power station efficiency of 0.40,
Kemp calculates a fuel consumption per seat of 22 litres on the train journey
from London to Edinburgh at 350 km/h. He then compares it with a VW Passat
130 TDI carrying two occupants, to which he attributes a fuel consumption of
2.8 litres/100 km (and a wheel to tank efficiency of 80%), and reaches the
conclusion that the consumption is the same: 822 litres per passenger). He
completes the comparison with a single-class Airbus 321-100 to which he
attributes a fuel consumption of 16 litres per seat, and with the same wheel to
tank efficiency (80%), resulting in a consumption of 20 litres per seat for the
1,5
1,7
1,9
2,1
2,3
2,5
2,7
2,9
3,1
3,3
3,5
200 220 240 260 280 300 320 340
Maximum speed (km/h)
J
o
u
r
n
e
y

t
i
m
e

(
h
)
30
35
40
45
50
55
60
65
70
E
n
e
r
g
y

c
o
n
s
u
m
p
t
i
o
n

(
k
W
h
/
s
e
a
t
)
Journey time
Energy
consumption
3.5
3.3
3.1
2.9
2.7
2.5
2.3
2.1
1.9
1.7
1.5
2010

12
journey from London to Edinburgh; in other words, less than the high speed
train!
It would not be worth analysing these figures any further were it not for the fact
that they are practically the only ones to have been published in relation to the
high speed trains energy consumption. Therefore, they have been used by
many authors.
In view of this, it would seem advisable to point out that Kemp has not taken
into account the fact that the same amount of primary energy used to produce
electricity does not have the same effect on greenhouse gas emissions or on
the contribution to the exhaustion of fossil fuels as when the primary energy
used for transportation is derived from oil. In our opinion, the mere comparison
of primary energy is not relevant for any practical purpose.
Comparison with Spanish 16-years experience
Nevertheless (and disregarding numerous methodological aspects that could be
called into question), the most important difference with respect to our own
studies lies in the high speed trains consumption, as Kemps 1993 estimate
does not seem to tally with Spains high speed operation experience over the
last 16 years.
In fact, let us consider the route from Madrid to Barcelona (620 kilometres,
somewhat longer than the London-Edinburgh route used by Kemp) and take a
series 103 high speed train with 397 seats running at various maximum speeds.
We calculate the journey time with the Aplica simulator and the consumption
with the Alpi2810 (both extensively checked against actual figures). To ensure
homogeneity, the time calculation is based on a margin in the middle of the UIC
band (8%); and the consumption is calculated in the trains wheel-rims
(including the auxiliary services) and for each train seat so that the figure 1
shows is strictly comparable.
In spite of the greater distance (+3.3%), the result shows:
a) Slightly shorter journey times than Kemps (between 6 and 16 minutes),
which can be attributed to logical differences, due to the layout of the line
or the margins adopted.
b) A considerably lower energy consumption, with differences of up to 56%.
The 103 from Madrid to Barcelona, running at 350 km/h, consumes
25,08 kWh per seat in wheel-rims, less than half the figure presented by
Kemp, although it should be pointed out that we have assumed the
existence (as in reality) of a regenerative brake, whereas this advantage
would almost certainly not have been considered in Kemps study (based
on a 1981 French TGV that did not have a regenerative brake). Yet even
without considering the contribution of the regenerative brake, the AVEs
consumption between Madrid to Barcelona, at 350 km/h, would be 27,47
kWh/seat, 52% less than the figure indicated by Kemp.
These differences can be seen in the following table.
2010

13
Table 1. Comparison between journey times and energy consumptions at different
speeds for a journey from London and Edinburgh (Kemp 2004]) and from Madrid to
Barcelona (independently produced).
Max.
Speed
(km/h)
Consumption
Lon-Edim.
(kWwheel/seat)
Consumption
Mad-Bcn
(kWwheel/seat)
with reg.brake
Consumption
Mad-Bcn
(kWwheel/seat)
without
re.brake
Time (h)
London-
Edimburgh.
Time (h)
Madrid-
Barcelona
Time
difference
MadBcn-
LonEdim.
(min)
Consum.
Difference
MadBac (with
reg brake) -
LonEdim.
(kWh/p)
Consum.
Difference
MadBac
(without reg
brake) -
LonEdim.
(kWh/p)
225 30 16.57 21.77 3.25 3.10 9 -13.43 -8.23
250 36 18.40 23.00 2.95 2.85 6 -17.60 -13.00
300 48 22.00 25.50 2.65 2.50 9 -26.00 -22.50
350 57 25.08 27.47 2.55 2.30 16 -31.92 -29.53

The following graph shows how the differences occur across the whole range of
speeds.

Figure 2. Comparison of energy consumption at different speeds for a journey from
London to Edinburgh (Kemp, 2004) and from Madrid to Barcelona (independently
produced).

15
20
25
30
35
40
45
50
55
60
65
200 220 240 260 280 300 320 340
Maximum speed (km/h)
E
n
e
r
g
y

c
o
n
s
u
m
p
t
i
o
n

(
k
W
h
/
p
l
a
z
a
)Cosump.
London
Edinburgh
(Kemp
1993)
s 103
consumption
Mad-Bcn
without reg.
brake
s 103
consumption
Mad-Bcn
withreg.
brake
E
n
e
r
g
y
c
o
n
s
u
m
p
t
i
o
n
(
k
W
h
/
s
e
a
t
)
2010

14
3. Comparison of the high speed
railway system and the conventional
system in terms of energy
consumption
In view of this background, we will try to compare the high speed railways
energy consumption and greenhouse gas emissions with those of the
conventional railway. This comparison will be based on empirical verification
and results of analyses carried out with simulators, the aim being to explain the
reasons for various types of consumption differences.
The comparisons entail a difficulty deriving from the fact that the cases being
compared are differentiated by other variables (besides speed). In fact, the high
speed trains and the lines on which they run are different from the conventional
ones; and the differences, besides speed, may relate to the layout, the power
supply, the number of stops, etc.
As we have explained, the high speed system requires certain curve-free
layouts, certain especially light and aerodynamic trains, and a power supply
system that can provide high outputs. Without these attributes, high speed
would not be possible, and therefore the comparison must incorporate all these
attributes.
In short, the comparison between the consumptions of the various trains will be
made in the normal operating environment of each one: the characteristic lines,
their sizes, their occupancy rates...
Later on, and in order to understand the reasons for the differences, each one
of the causes will be analysed separately in an attempt to attribute parts of the
differences to each one of the two categories of consumption inducers (defined
by Minayo, 2008): the direct effects of the speed and the effects of the high
speed system.
2010

15
3.1. Empirical verification
An initial approach could involve observing reality from a general perspective, in
order to find out whether the common misconceptions have any basis in reality
or whether, on the contrary, reality offers different results.
3.1.1. Empirical relationship between energy consumption and speed
As regards the common misconception regarding energy consumption
(square rule: the energy consumption of the trains would increase with the
square of their speed), we have made a comparison between various types of
trains that cover a variety of real situations.
An initial comparison, of an economic nature, can be obtained from Renfes
2005 Report
1
by extracting the data corresponding to traction energy use and
the income and traffic figures of each one of Renfes Business Units. These
Units cover commuter, regional, long distance and high speed trains, with
relatively different average speeds from one Unit to another, but with a certain
degree of uniformity within each Unit.
The results appear in the following table, which confirms the existence of an
inverse relationship between speed and the percentage of total costs and total
income that the energy cost represents.
Specifically, in the case of the High Speed Business Unit, whose trains logically
have the highest commercial speed
2
, energy costs account for only 5.25% of
income
3
, as opposed to 16.32% in the freight trains or 12.08 % in the commuter
trains.

1
The reason for using the 2005 Report is that it is the last one in which the data
corresponding to Suburban, Regional, Long Distance and AVE are presented separately, as
these four Business Units have since been grouped together in two areas of activity:
Suburban and Medium Distance and Long Distance and High Speed, the data being
published for each area of activity. Moreover, since 2005 the medium distance high speed
trains (Avant product) have been managed by the Suburban and Medium Distance area,
whereas up until then they had been managed by the AVE Business Unit. Therefore, the data
contained in the 2005 Report are the last ones with the old structure which, given the high
number of different types of trains, offers more relevant information for our purposes. It
should be pointed out that the price of energy has risen sharply since 2006, which has
increased energy costs as a percentage of income (in 2006, the energy costs for passenger
trains accounted for 7.44% of income, as opposed to 6.86% in 2005). However, this does not
affect the relationship with speed, which is the purpose of this study.
2
In 2005, the High Speed Business Unit includes both the AVE trains and the Shuttles and
Talgo 200 that partly use the high speed lines.
3
For comparative purposes, it is worth pointing out that in 2007, Iberia (the Spanish national
airline) recorded energy costs of 1.144 billion euros, which represents 21.8 % of their
operating costs (on all routes) and 20.7% of their operating income; in other words, the
aviation energy costs/income percentage is approximately 3.7 times higher than that of the
high speed train.
2010

16
Table 2. Weith of energy costs on several passenger trains

Average
speed Traffic
Traffic
income
Income
per traffic
unit
Engry
expenses
Total
expenses
%
Energy /
Incomes
% Energy
/
expenses
Energy
expenses
per traffic
unit
km/h
Gvkm
(Gtkm)
M /
year
c/v.km o
c/t.km M / year M / year
%GE/
GT % GE/ IT
c/ v.kmt
t.km
Commuter 53.29 8,417 574 6.82 69.35 556 12.08% 12.47% 0.82
Regional 71.21 2,745 222 8.10 14.88 216 6.69% 6.88% 0.54
Long
distance 89.29 6,322 426 6.73 33.10 480 7.78% 6.89% 0.52
AVE-High
speed 159.99 2,325 262 11.26 14.58 276 5.57% 5.28% 0.63
Freight 54.50 11,071 323 2.91 52.68 409 16.32% 12.89% 0.48
Total 69.07 30,880 1,807 5.85 184.59 1,937 10.22% 9.53% 0.60
As the table shows, Renfes energy costs represent only 9.53% of the total
business income, whereas in other modes of transport, namely the car, the bus
and the aeroplane, energy accounts for approximately 50%, 25-30% and 20-
30% of the total costs, respectively. In the case of the railway, this percentage
has dropped significantly over the years, given that energy costs accounted for
28% of total costs in 1958.
The analysis of energy costs in relation to income has a weakness, given that
the unit income generated by each traffic unit (passenger-km) varies
significantly according to each type of traffic. In fact, income per traffic unit rises
sharply when the average speed increases, as can be seen in the table.
Even though this inducer (income per passenger-km) offers an example of the
passengers willingness to pay for an increase in the average speed of the
trains (even if they had to pay more for an increase in energy consumption), it is
true that from the technical point of view, the amount of energy consumed per
passenger-km seems more important than energy costs as a percentage of
total income.
2010

17
In this respect, the following graph shows how the High Speed Business Unit,
the one whose trains run at the highest speed (160 km/h), is precisely the one
that records the lowest consumption per passenger-kilometre, whereas in the
rest of the business units, the consumption per passenger-km increases as the
average speed decreases.

Figure 3. Comparison of energy consumption (Wh) per passenger-kilometre in various
types of trains in ascending order of average speed. Independently produced.
This fact should not come as a surprise, because if we compare various road
transport services, the fuel consumption of a car in the urban driving cycle, with
an average speed of 40 km/h, is about 60% higher than the same cars
consumption in the intercity driving cycle with an average speed of 100 km/h.
The same occurs in the case of buses and coaches: in urban services (20 km/h)
the energy consumption is approximately 30% more than on intercity routes
with average speeds of around 70 km/h.
2010

18
3.2. Articles based on specific cases
In recent years, the author has published various articles about the energy
consumption of all types of high speed trains with the help of a self-developed
energy consumption simulator, practical verification of the resultant data having
revealed a high level of accuracy. Some of these articles, together with others
produced by other authors, allow the energy consumption of the high speed
train to be compared with that of the conventional train in specific cases.
We have chosen three cases: one involving different trains (high speed and
conventional) on different lines (high speed and conventional); another involving
different trains (conventional and high speed) on the same line; and finally,
identical trains (high speed) on two different lines (high speed and
conventional).
3.2.1. The case of different trains on different lines
The first of these articles (Garca lvarez, 2005) aimed to analyse the energy
guzzler accusations that are frequently levelled against the high speed train
and the campaigns against these trains that are supported by those who
demand, with a view to reducing transport consumption and emissions, the
expansion of improved conventional lines or speed limits for trains. Therefore,
the article compares the consumptions of the high speed train with those of the
conventional train.
Thus, by means of simulation validated with real data, the energy consumption
of a conventional train (consisting of a series 252 electric locomotive and 7
Arco cars) running at 200 km/h on an improved conventional line (Barcelona
to Alicante, on what is known as the Mediterranean Corridor) was compared
with the consumption of a high speed train (Talgo 350, Renfe series 102) on the
Madrid-Lleida section of the Madrid-Barcelona high speed line. In both cases,
the trains have the same capacity (316 seats).
The average speed reached by the high speed train is 32.6 % higher, in spite of
which it consumes less energy per kilometre, both when measuring the energy
imported to the pantograph (-7.2%) and when recording the net energy
(deducting the energy exported to the network) as it leaves the power plant
4
(-
15.7%). Given the aim of the experiment, the effect of the shorter length of the
high speed lines in comparison with the conventional lines was not taken into
account (for example, the high speed line is 16.5% shorter in the case of the
Madrid-Lleida section), which would make the difference even greater in favour
of high speed.

4
The difference between the energy produced at the power station and the energy imported
to a trains pantograph lies in the energy conversion and transmission losses that occur from
the moment it leaves the power station to the moment it enters the train, minus the energy
regenerated by braking (negative). In short:
Energy generated at the power station = Energy imported to the pantograph + losses in the
network and catenary energy regenerated in braking.
2010

19
To separate the influence on this result of the trains characteristics on the one
hand, and those of the line and the service on the other, the consumption of the
same high speed train (Talgo 350) running on the improved conventional line is
also simulated. The difference in consumption turns out to be smaller than in
the real case, but the consumption is still lower (-7.4% in power station busbars)
in the case of the high speed line. All of which leads to the conclusion that it
cannot be said the energy consumption of the high speed train (at 300 km/h or
more) is essentially different from that of the improved conventional train
(running at maximum speeds of around 200 km/h), provided that the service
characteristics are homogeneous.
The following graph shows the differences between each one of the addends in
the consumption of the two trains.

It can be seen that, in spite of a lower average speed, consumption is higher in
the improved conventional train, both in the trains wheel-rims and at the power
station outlet. Certainly, consumption due to aerodynamic drag is much higher
(+71.6%) in the high speed train, but all the other consumption addends are
lower: mechanical resistances (-30.4%), auxiliary services (-28.9%), losses in
the locomotive and in the network (-30.7%) and, most of all, energy dissipated
in the brake (-57.29%), this being the addend with the biggest reduction in
consumption.
2010

20
3.2.2. The case of different trains on the same line
Another comparison, between a new high speed train and a conventional train
on the same line (albeit improved), can be found in Andersson and Lukaszewicz
(2006), who have carried out in-depth analyses of energy consumption in trains
in Sweden.
These authors show that on the same route (Vters to Stockholm), a
conventional train took 78 minutes and consumed 0.042 kWh/seat-km in 1994,
whereas in 2004 the X2000 tilting high speed train, with one more stop, took 53
minutes (-32%) and consumed 0.030 kWh/seat-km (-28%).
3.2.3. The case of the same train on different lines
In Garca lvarez and Martn Caizares (2007) we analysed the compared
consumption of the same train on two lines (a high speed line and a
conventional line) between the same points, and in two specific cases: that of
an Alvia 120 train (Renfe series) between Lleida and Roda de Bar, and that of
a Talgo 200 between Crdoba and Antequera. In December 2006, these
(variable track gauge) trains stopped running on the conventional lines and
began to operate on the new high speed lines.
In this case, on changing from the conventional line to the high speed line and
running between the same stations, the Alvia and Talgo 200 trains achieved
energy consumption reductions in the pantograph of 16% and 8% respectively,
which is compatible with an average speed increase of 64% and 78%
respectively.
In the same article we also analysed the consumption of electricity produced by
these trains, measured at the power stations outlet busbars. In the case of
Lleida to Roda, the reduction in consumption at the power station is -40%,
greater than the reduction measured in the pantograph (-16%). Something
similar occurs in the case of the Crdoba route: consumption in the power
station bars is 27% less (in the pantograph the reduction in consumption is 8%).
The differences are due to the fact that the high speed lines are electrified at a
higher voltage (25 kV as opposed to 3 kV), which means that losses are lower.
Moreover, the regenerative brake is used more on high speed lines than on
conventional lines.

2010

21
4. Technical reasons for lower
consumption
In all the real cases analysed (different trains and lines; different trains on the
same line; and the same train on different lines), the result of the analysis
coincides with the overall result observed for all Renfes trains: the trains that
run at higher speeds consume less energy.
As this result clearly contradicts the existing common misconception, it is
worth analysing in detail the reasons for this lower consumption.
4.1. Addends of the consumption function
The trains dynamics is explained in detail in Garca lvarez (2004), and the
physical and technical reasons that explain the differences in consumption
between the high speed train and the conventional train are analysed in Garca
lvarez (2006). Using data from both articles we will explain the reasons for the
consumption by taking a look at each one of the addends that make up the
energy consumption according to the underlying model in the ALPI2810
simulator.
The consumption function is constructed by analysing the energy released by
the train, to which we add the losses that occur from the moment the energy
leaves the power station to the moment it leaves the train. In each case, we will
analyse the relationship with the mass, the distance travelled and the trains
speed, and also with the characteristics of the vehicles and the lines that are
necessary for operating at high speed.
This will allow us to produce a table, with approximate coefficients, that sums up
the different consumption in each one of the parts that explain it.
4.1.1. Energy needed to overcome mechanical resistance on straight
track and on curves
The energy needed to overcome mechanical rolling resistances depends on the
specific value of such resistances, the mass and the distance travelled.
Along the whole route (straight and curve) it can be assumed that this
resistance is proportional to the trains mass, the constant of proportionality (a)
ranging from 1.2 to 2 daN/t for conventional trains and from 0.5 to 0.9 daN/t for
high speed trains.
On curves, the energy required is proportional to the mass and to the distance
travelled, and inversely proportional to the curve radius. In Garca lvarez
(2004,[7]) it is proposed that each line be modelled so that it is defined as an
equivalent coefficient of curves (a
c
) that would be added to the specific
coefficient of the mechanical resistances. The values of coefficient a
c
would
vary between 1.424 daN/t (Betanzos-Ferrol line) and 0.08 daN/t (high speed
2010

22
line from Madrid to Barcelona). In the Enertrans project
5
, values of 1.2 daN/t
and 0,09 daN/t are proposed for intercity lines and high speed lines,
respectively.
The value of these mechanical resistances can be assumed, therefore, to be
independent of the speed
6
and can be expressed as follows:
c
c
rm
l M
R
k
L M a E + =

And if the specific coefficient of curves a
c
is defined as:
L
l
R
k
a
c
c
c
=

,
the energy needed to overcome the mechanical and curve resistances for a
route of length L can be expressed as
L M a a E
c rm
+ = ) (
Coefficient a depends on the rolling stock (not on the line), being much lower in
high speed vehicles.
Coefficient a
c
depends on the line (not on the rolling stock)
7
; specifically, it
depends on the number of curves and on their radius and length. Therefore,
this coefficient is lower on high speed lines, given that the most important layout
restriction on such lines is precisely the absence of curves, so that the trains
can run at high speed.
The vehicle mass M does not seem to be closely related (per standard unit of
supply) to the speed or the type of line or type of rolling stock. In any case, if
there were a relationship, it would be of a lower mass per standard units of
supply in high speed trains.

5
The aim of the EnerTrans research project was to obtain an accurate model that reveals the
energy consumption (and its associated emissions) of the Spanish transport system, according
to the significant variables on which it depends. Participants include the Fundacin de los
Ferrocarriles Espaoles, Universidad de Castilla-La Mancha, Universidad Pontificia de Comillas
de Madrid, ALSA, Universidad Politcnica de MadridINSIA, Fundacin General de la
Universidad Autnoma de Madrid, Fundacin Agustn de Betancourt and Fundacin
Universidad de Oviedo. The project is subsidised by the Centro de Estudios y Experimentacin
de Obras Pblicas (Ministry of Public Works, project n PT-2006-006-01IASM).
6
In France, a part of the mechanical resistances is considered to be proportional to the speed
(Bernard and Guiheu, 1976, [6]), but this is not a widespread practice, and the proposed
values of coefficient a already include the part that would depend on the speed.
7
For certain specific vehicles, such as the Talgo trains with guided axles, or for certain rolling
gear arrangements, the constant k
c
may depend on the rolling stock, but both trains,
whatever the value of k
c
may be, can run at any speed within their respective speed ranges,
and therefore the relationship between k
c
and the rolling stock will not be taken into account
in this analysis.
2010

23
4.1.2. Energy needed to overcome intake air resistance in the train
The energy needed to overcome the resistance produced by the entry of air into
the train is proportional to the amount of air that enters the train (Q, in m
3
/s),
which is a parameter determined by passenger comfort; to the air density (d); to
the trains speed; and to the distance travelled (L).
It can therefore be expressed as follows:
L V d Q k E
ea ea
=
where k
ea
is constant. Q is related to the size of the train (number of seats), but
is independent of the trains speed and the type of line it runs on.
4.1.3. Energy needed to overcome aerodynamic drag
The energy needed to overcome aerodynamic drag depends on the square of
the instantaneous speed and on the air density; it does not depend on the
trains mass.
It can be broken down into two addends, one of them due to pressure (normal
forces) and another due to friction (shear stresses).
Energy needed to overcome drag due to pressure forces
Drag due to pressure forces occurs both at the front and the rear of the train,
and depends (with a proportionality coefficient c
p
) on the trains cross-section
surface area (St), and on its form (coefficient C
x
). It also depends on the air
density and, on the part of the route corresponding to tunnels, on a tunnel factor
(T
f
). The energy consumed on the route due to this cause is:
= dl V Tf S c E
f p rap
2

In conventional trains, the value of coefficient c
p
is around 0.0022
daN/[(km/h)
2
.m
2
], whereas in high speed trains the value is closer to 0.00096
daN/[(km/h)
2
.m
2
], assuming a cross-section surface area of about 12 m
2
in both
cases.
Energy needed to overcome friction drag
Friction drag due to shear stresses occurs on the trains wet surface (skin),
whose surface area can be calculated on the basis of its height (H), length (L
t
)
and width (W) as follows:
t m
L W H S + = ) 2 (
The energy needed to overcome this drag on a route is:
= dl V T S c E
f m f raf
2

In conventional trains, the value of coefficient c
f
(which depends on surface
continuity and quality) is around 0.00003 daN/[(km/h)
2
.m
2
], whereas in high
speed trains the value is closer to 0.000021 daN/[(km/h)
2
.m
2
], assuming a wet
perimeter of 11m in both cases. It should be pointed out that for trains with a
length of more than 200 m, the quotient c
f
tends to fall.
2010

24
For all the aerodynamic drags, Lukaszewicz proposes a coefficient C value
proportional to (/2)x(8.3+0.057L
t
,) for conventional trains, and a value
proportional to (/2)x(4.7+0.050L
t
) for high speed trains, L
t
being the length of
the train in metres. For a 200 metre train, this means that the value of
coefficient C in a high speed train is 25.3% lower than the same value in a
conventional train
4.1.4. Energy dissipated by the brake and not utilized
On any given route, the train needs additional energy to increase its speed and
to climb upgrades. However, instead of being lost, this energy accumulates in
the train in the form of kinetic and potential energy respectively, and can be
used to overcome the rolling resistance, in which case it is not wasted.
As the energy needed to overcome rolling resistances has already been
considered in the previous addends (whatever their origin may be: external or
energy accumulated in the train itself), it can be understood that the energy lost
is only that which is dissipated by the trains brake. This dissipation occurs in
two cases: when the train brakes to reduce its speed (in order to stop or comply
with a local speed limit) or when it brakes to descend a steep downgrade
without exceeding the maximum speed. Therefore, the energy dissipated by the
brake is the sum of the energy dissipated by braking to reduce speed and the
energy dissipated by braking to descend steep downgrades.
Kinetic energy dissipated in speed reductions
The kinetic energy dissipated in speed reductions depends on the vehicles
mass, the rotating masses
8
, the distance between technical and commercial
stops and the distance between equivalent stops (parameter of the line that
characterises the homogeneity of the authorised speed profile -see Garca
lvarez for more information-).
( ) ( ) L V
D D D
M M E
pt pe
pc
rot frepar

+ + + =
2
1 1 1
2
1

From this energy we can deduce the amount used to overcome rolling
resistance during the deceleration process.
The amount of energy dissipated in the speed reduction depends on the mass
(and on the rotating masses), the number of stops (D
pc
) and the homogeneity of
the speed profile (D
pe
).

8
Rotating masses are those which turn with a rotation speed proportional to the translational
speed of the vehicle (axles, wheels, brake discs, etc.) When the train accelerates, these
masses increase their rotation speed, and therefore they must be accelerated angularly. This
effect is taken into account by adding the equivalent rotating masses (M
re
) to the trains
mass. These masses usually have an equivalent value of between 4 and 10% of the trains tare
mass. For more information, see Garca lvarez 2004.
2010

25
Figure 4. Speed profile on the Mediterranean Corridor. The speed reductions are
equivalent to an additional stop every 50 km. Source: Garca lvarez, 2005
Figure 5. Speed profile of the Madrid-Barcelona and Madrid-Seville high speed lines,
equivalent to a stop every 550 km and 120 km, respectively. The homogeneity of the
speed profile is one of the main reasons for the lower consumption of high speed trains.
Source: Garca lvarez
Kinetic energy dissipated when braking on gradients with a
value greater than p
e

For a train, the gradient of repose (p
e
) is defined as the gradient for which the
train, without braking or applying traction, maintains a speed equal to its
maximum speed on the line section. It is understood that the higher the trains
authorised speed on the line, the greater the gradient of repose is. If the actual
existing gradient (p
r
) is greater than the gradient of repose (p
r
>p
e
), the train
must brake so as not to exceed the maximum speed, and brake more (the
greater the difference p
r
-p
e
, the more energy it loses).
Therefore, the lower the gradient of repose, the more the train will brake on
gradients, i.e. the lower the maximum speed, the more it will brake. Thus, for
example, when the authorised speed of a series 102 train on the Madrid
Barcelona high speed line changes from 220 to 300 km/h, the energy dissipated
by the brake on gradients is reduced by 32%.

0
50
100
150
200
250
0
2
8
6
7
9
4
3
2
2
2
7
0
8
6
5
8
2
3
3
0
1
2
5
9
2
7
1
4
0
6
4
3
1
6
5
1
7
2
1
7
3
8
7
5
1
7
9
8
6
7
1
8
7
1
8
8
2
0
1
1
4
0
2
4
0
3
1
4
2
5
1
9
9
6
2
7
4
7
9
7
3
0
3
2
6
4
3
8
2
5
4
0
4
1
2
2
2
9
4
3
7
1
6
2
4
4
7
8
2
8
4
6
5
5
6
3
4
9
0
5
2
5
5
0
2
2
2
5
0
5
0
10
0
15
0
20
0
25
0
30
0
35
0
40
0
0 10
0
20
0
30
0
40
0
50
0
60
0
K
m
km/h
2010

26
The energy dissipated by the brake on gradients can be expressed as follows:
[ ] ( ) M
e
p
t
p
p
l g
frepend
E =
where p
e
is greater as the speed increases.
Energy not lost due to economical driving
If the train decelerates or descends the gradients without braking, it can use
some (or all) of the accumulated kinetic and potential energy to overcome the
rolling resistance; the energy consumed is therefore reduced, since the amount
of energy dissipated by the brake is reduced. As a result, however, the train
loses time, which means that over the whole route it uses more time than the
necessary minimum. This is what is called economical driving (usually
achieved by coasting): the longer the journey time, the lower the energy
consumption
9
.
In practice, the effect of economical driving can be modelled as a coefficient,
equal to or less than 1, which multiplies the energy lost during the braking
process.
( )
frepend frepar condecono
E E k Ef + = , where 1 0
condecon
k
If coasting occurs at high speeds, little time is lost, yet the energy saving is high.
At low speeds, if the train coasts instead of braking, a lot of time is lost and yet
little energy is saved. Therefore, economical driving can save more energy at
high speeds, and therefore the high speed train can cover a considerable part
of the route by coasting (or freewheeling) without losing much time.
On a typical high speed line, a time margin of 6% (minimum recommended by
the UIC) permits K
condecon
values of around 0.4 to 0.6, whereas on a line where
the maximum speed is 160 km/h, with the same margin, values of around 0.8 to
0.9 can be obtained for K
condecon
.
The economical driving programme on the Madrid-Seville AVE reduced energy
consumption by 8% while at the same time increasing the average speed. A
high speed train can coast along 64% of the Madrid-Seville route (losing 7
minutes without consuming a single kilowatt-hour), while the Madrid-Toledo
Avant train can do the same over 68% of the route, losing just 4 minutes of its
minimum running time.
Energy recovered by the regenerative brake
If the train has a regenerative brake, some of the energy dissipated by the
brake can be recovered: either to be used by other trains, or to be returned to
the public power network.
Only a part of the energy dissipated by the brake is converted into electricity by
the regenerative brake (should it exist), as it is always necessary to use, to a
greater or lesser extent, the friction brakes. Friction brakes are used mainly at

9
Besides coasting, economical driving also admits other variants (which cannot be analysed
in detail in this article), and numerous studies have dealt with the optimization of economical
driving according to the track profile, the rolling stock and other circumstances, for example:
Lukaszewicz (2001) and Aragn (2005).
2010

27
low speeds (for example, between 50 km/h and the stop). Therefore, the part of
the energy dissipated by the brake that is converted into electricity is greater at
high speed.
The energy dissipated by the brake that is not regenerated is:
( )
frepend frepar condecono on NOgeneraci frnoreg
E E k k E + =
The value of the NOgeneracin (k
NOgenreracin
) coefficient is 1 if the train does not
have a regenerative brake. If it does, the following can be assumed,
approximately (V
mx
. being the trains maximum speed on the line, in km/h):
2
max
2
50
V
k
on NOgeneraci
=
From this formula we can deduce that for a maximum speed of 100 km/h, 75%
of the energy dissipated by the brake is converted into potentially usable
electricity, whereas if the maximum speed is 300 km/h, this percentage rises to
97%.
The actual degree of utilization of the braking energy converted into electricity
by other trains depends on the length of the electrical sections (the longer they
are, the greater the possibility of recovery), and on the traffic density (the higher
the density, the more likely it is that when one train is generating energy with
the brake, there is another one that needs it for an acceleration process).
In practice, it can be modelled as a coefficient k
aprovereg

( )
frepend frepar condecono n NOgeneraci en aprovecreg do aprovfrena
E E k k k E + = ) 1 (
At high speed, alternating current electrification allows the energy produced by
the regenerative brake to be returned to the public network, and (although the
railway company is not paid for the energy returned to the network) from the
energy efficiency point of view, it can be regarded as utilized energy.
Therefore, on AC lines, it can be assumed that K
aprovregen
=1, whereas with direct
current, if there are no devices for storing energy or returning it to the network, it
can reach values of 0.5 to 0.9, depending on the traffic density and the length of
the electrical sections.
In short, and with regard to the utilization of potential and kinetic energy, the
above can be summarised as follows:
The high speed train has to dissipate less kinetic and potential energy,
as it makes fewer stops and can descend downgrades at higher speeds.
Of the energy it has to dissipate, it can lose a smaller part of it in the
brake (for the same journey time), as it can perform economic driving
more efficiently.
Of the energy that, in spite of this, it dissipates in the brake, it can
convert a larger part of it into electricity (as the electric brake is more
usable at high speeds).
The energy thus converted into electricity can be utilized to a greater
degree; as it can be returned to the power network, the degree of
utilization is practically 100%.
2010

28
4.1.5. Energy consumed by the trains auxiliary services
The trains auxiliary services are systems that consume energy for technical
purposes (compressors, motor ventilators, etc.) or for the comfort of passengers
(heating, air-conditioning, lighting, etc.)
The energy consumption of these services is proportional to the time during
which they operate; and therefore, if the average speed increases, the
consumption per kilometres decreases in the same proportion as the average
speed increases.
The energy consumed by the auxiliary services can be modelled as the result of
the average power used by the auxiliary systems multiplied by the time during
which they operate.
V
L
k P T k P E
uso aux uso aux confort aux
= =
In principle, the energy consumption of the auxiliary services is not directly
related to speed; thus, in the verification of a typical high speed case, a 50%
increase in the average speed means a 29% reduction in the energy consumed
by the auxiliary services.
It should be pointed out that the auxiliary services continue to consume energy
during stops, even though the train does not cover any distance at all, and
therefore an increase in this time (including the time at the passengers disposal
before the train sets off again) means a higher consumption per kilometre on
the part of the auxiliary services. As high speed trains make fewer and shorter
stops than conventional trains, and as the time spent at the terminus is divided
among more kilometres, the energy consumed by the auxiliary services is also
reduced for this reason.
4.1.6. Multiplier factors in the consumption function
So far we have specified the addends that make up the consumption function,
and the sum of all of them (with their sign) provides the net energy consumed in
the trains wheel-rims (or at the inlet of the auxiliary services) to move the train
over a distance of one kilometre.
To find out the total energy consumed, we would have to multiply by the energy
loss coefficients from the moment the energy leaves the power station to the
moment it reaches the train, and also by the total distance travelled.
4.1.7. Energy loss coefficient
Power stations have to produce an amount of energy for a train which is the net
energy required at the trains wheel-rim (or at the inlet of the auxiliary services)
plus the losses of any kind that occur until the electricity is converted into
energy that can eventually be used. Two main types of losses can be identified
(each one with an associated loss coefficient): in the power networks and in the
locomotive.
2010

29
Energy lost in the power networks
The energy lost during the voltage change processes and during the
transmission of the energy from the power station to the train is proportional to
the energy that reaches the locomotive, and the loss coefficient depends on the
trains operating voltage (higher voltage means fewer losses), and also on the
electrification characteristics (cross-section of the conductors, distance between
substations, etc.)
The diagram in Figure 8 shows the equivalent efficiencies in each one of the
energy transformation phases and, consequently, the incremental amount of
energy that needs to be produced in order to offset these losses.
In Spain, high speed lines are electrified at 25 kV a.c. and conventional lines at
3 kV d.c. The higher voltage means fewer electrical losses during the
transformation and transmission processes. The additional amount that has to
be produced, on top of the amount consumed by the pantograph, is: 22.6 % for
trains operating at 3,000 V d.c. and 8.8% if the trains operate at 25 kV, a.c.
(Pilo, 2006).

Figure 6. Losses during the voltage change processes and transmission of electricity to
the train. Source: Pilo (2006)
Energy lost in the locomotive
In motor vehicles, energy losses occur due to the efficiency of the various
systems (converters, reducer motors, etc.)
In principle, efficiencies in electric traction vehicles are usually independent of
the engine operating conditions, and therefore of the speed and the speed
profile. In practice, however, there are fewer energy losses in high speed
systems, given that, on the one hand, engines tend to be bigger (and the bigger
the engine, the fewer the losses) and, on the other, the modern technologies
2010

30
that are usually applied to high speed trains generally make better
performances possible.
4.1.8. Shorter distances between the same points
In all the addends of the consumption function, the distance travelled is
multiplied by the corresponding monomial. Therefore, it can be said that the
energy consumed by a train is proportional to the distance travelled
10
.
High speed layouts imply shorter distances than those of conventional lines
between the same points. If the distance is shorter between the same points,
energy consumption will be lower in same proportion, all other factors being
equal.
Table 3. Comparison between distances by high speed line, by conventional railway line,
on straight track (displacement) and by road on Spanish high speed routes. Source:
Garca lvarez and Fernndez Gonzlez (2008)
Displacement Road Conv. R HS R HS R/ Conv. R/ HS R/ HS R/
Madrid to (km) (km) (km) (km) disp. disp. Conv. R road.
Segovia 63.0 96.0 101.6 68.3 1.08 1.61 0.67 0.71
Valladolid 153.0 206.0 249.4 179.2 1.17 1.63 0.72 0.87
Barcelona 502.3 617.0 708.0 621.0 1.24 1.41 0.88 1.01
Zaragoza 260.0 309.8 338.0 306.7 1.18 1.30 0.91 0.99
Lleida 370.0 461.6 532.0 442.0 1.19 1.44 0.83 0.96
Camp. Tarragona 410.0 461.6 595.0 521.0 1.27 1.45 0.88 1.13
Huesca 320.0 380.4 421.0 392.0 1.23 1.32 0.93 1.03
Cuenca 130.0 169.8 200.0 188.1 1.45 1.54 0.94 1.11
Albacete 221.8 266.0 278.7 314.0 1.42 1.26 1.13 1.18
Valencia 301.2 354.0 489.1 390.8 1.30 1.62 0.80 1.10
Alicante 360.0 420.0 454.7 483.0 1.34 1.26 1.06 1.15
Murcia 350.0 400.0 459.9 525.0 1.50 1.31 1.14 1.31
Toledo 66.2 89.6 90.2 75.2 1.13 1.36 0.83 0.84
Sevilla 387.0 535.8 571.0 470.5 1.22 1.48 0.82 0.88
Crdoba 277.0 398.4 440.0 343.0 1.24 1.59 0.78 0.86
Ciudad Real 150.0 208.3 262.0 173.0 1.15 1.75 0.66 0.83
Mlaga 393.0 535.3 633.0 512.9 1.31 1.61 0.81 0.96
HSL 1 (North) 1.13 1.62 0.70 0.79
HSL 2 (Northeast) 1.22 1.38 0.88 1.02
HSL 3 (Levante) 1.40 1.40 1.01 1.17
HSL 4 (Andalusi) 1.21 1.56 0.78 0.87
Static HSL AVERAGE 1.26 1.47 0.87 1.00
EFFECTIVE HSL AVERAGE (Weighted with passenger-km) 1.27 1.45 0.88 1.01
The table with data taken from shows that, in Spain, the average distance of
high speed lines is 13% shorter than that of the conventional line between the
same points, if measured in static terms (as a simple average of the route
coefficients), and 12% if the effective route coefficient is measured (the
coefficients weighted by the passengers-kilometre anticipated on each route). It

10
Strictly speaking, it would be necessary to add the energy consumed by the auxiliary
services while the train is stationary but with these services operating; for example, during
commercial stops or the period during which passengers board the train before it departs
from the station of origin, or while they get off the train at the destination.
2010

31
can be seen that on some routes (such as Madrid to Segovia) the high speed
distance is as much as 23% shorter, and only on certain Levante routes
(Albacete, Murcia, Alicante) is the high speed distance greater, due to the
peculiar trunk topology of this high speed line.
4.2. Coefficients and exponents of the consumption
function
The development of the consumption function shows that each one of the
addends has a different relationship with the high speed system:
In some addends, energy consumption is directly related to speed: it
increases with speed (e.g. energy consumed due to air intake); or with its
square (energy consumed to overcome aerodynamic drags); or it
decreases linearly with speed (as in the case of the energy consumption
of auxiliary services).
In other cases, the relationship between consumption and high speed
does not derive directly from speed, but from the characteristics of some
of the high speed subsystems (of the vehicles, of the infrastructure or of
the operating system).
To analyse the reasons for the consumption differences between a high speed
train and a conventional train, it is advisable to bear in mind that the
fundamental differences between the two types of systems manifest themselves
in three areas: rolling stock; different layout and new infrastructure, as opposed
to improvement of the existing infrastructure; and a different form of operation
(fewer stops).
The high speed train is a system, and therefore it should be analysed as a
whole. It is not worth comparing only some of its variables (we would reach the
conclusion that minimum consumption occurs with very light, modern and
expensive trains on lines with favourable layouts and by operating them at very
low speeds and with no commercial stops, but then we would have to ask
ourselves why a line and train of these characteristics has been
constructed/manufactured).
In turn, the differences (in rolling stock, infrastructure and operation) between
the high speed system and the traditional railway system can be divided into
two categories:
Some differences are intrinsic to high speed: either in the vehicle (lower
mass per seat, lower specific rolling resistance), or in the infrastructure
(larger curve radii), or in the operation (fewer stops).
Other differences usually arise due to the greater modernity of high
speed lines, although conventional lines and trains should not be
excluded from them, at least in the medium term (e.g. alternating current
electrification with higher voltages, use of the regenerative brake, greater
vehicle efficiency, greater utilization of the trains, etc.)
If we consider the characteristics of the rolling stock, using the abbreviation
CT to refer to a train with the characteristics of conventional train, and HST
2010

32
for a generic train with the typical characteristics of high speed trains, we can
analyse each one of the addends and each one of the multiplier factors of the
consumption functions to work out the approximate absolute or relative values
corresponding to all four possible combinations: high speed train on
conventional line (CL), conventional train on high speed line (HSL), high speed
train on high speed line, and conventional train on conventional line.
We also analyse the relationship of each addend (according to its physical
nature) with the mass and the speed.
Table 4. Values of consumption function coefficients and speed exponents in each
monomial. Independently produced
Coefficientes: abosolute values Coef. Relative values/LC-TC Exponents
Line and vehicle characteristics CL-CT CL-HST HST-CT HSL-HST CL-CT
CL-
HST HSL-CT
HSL-
HST Leng Mass Spd
Ermr mechanical straight resistance 1.2 a 2 1.2 a 2 0.5 a 0.9 0.5 a 0.9 1 1.00 0.44 0.44 1 1 1
Ermr mechanical curve resistance 0.25 0.25 0.09 0.09 1 1.00 0.36 11.11 1 1 1
Total E. mechanical resistance 1.825 1.825 0.79 0.79 1 1.00 0.43 0.43 1 1 1
Eea resistance to aire intake 0.034 0.034 0.034 0.034 1 1.00 1.00 1.00 1 0 1
Erap resistance to aerod. Presure 0.0022 0.00096 0.0022 0.00096 1 0.44 1.00 0.44 1 0 2
Eraf aerodin. Resistance 0.00003 0.000021 0.00003 0.000021 1 0.70 1.00 0.70 1 0 2
Efp losses of brake on stops 1 0.72 0.42 0.31 1 1 2
Efp losses of brake on slopes 1 0.79 0.39 0.20 1 1 -1
Er energy regenerated by brake 1 0.71 0.40 0.29 1 1
Eaux energy consumed by aux. 1 0 -1
Current and brake charact.
C.C.-No
R.B. C.C.-R.B.
C.A-
No.R.B. C.A.-R.B.
C.C.-No
R.B.
C.C.-
R.B.
C.A.-
No.R.B.
C.A.-
R.B.
Brake recovery 0.35 1 0.35 1 1 2.86 1 2.86
Losses in the network 12317 12.317 11.016 1.02 1 1 0.89 0.89

2010

33
4.3. The effect of speed ceteris paribus (all other things
being equal)
From all the above it can be deduced that if all the other variables remain
identical, both those of the infrastructure (speed profile, supply voltages, plan
and elevation of the line), and those of the train (mass, size, form,
efficiency), energy consumption increases as the maximum speed and
average speed increase.
In order to quantify this increase and assess its economic effects, we have
carried out the simulation exercise consisting in rolling a series 103 high
speed train (397 seats, with 70% occupancy) between Madrid and Barcelona at
various maximum speeds (300, 330 and 350 km/h), obtaining in each case (with
the Aplica simulator) the minimum time and the commercial time (adding to the
minimum time a margin of 13 minutes, which is in the middle of the band of the
UICs recommendations on margins). The average speed has been calculated
and, with the ALPI2810 simulator, the energy consumption measured at the
substation inlet (which is where the economic value of the consumed energy is
measured).
For each one of the stages, the journey reduction times are compared with the
increase in consumption and with the value of the incremental energy needed
for each passenger. This value, in turn, is compared with the average ticket
price.
The results appear in the table.
Table 5. Effect of increase in speed (ceteris paribus) on journey time, energy
consumption and economic cost on the Madrid-Barcelona route
MADRID-BARCELONA Max. Speed (km/h) 300 330 350
Minimum min 137 129 124
Commercial running time (Margen 13 min) min 150 142 137
Average speed km/h 248.4 262.4 272.0
Substation entry consumption kWk 12,436 13,205 13,795
Time saved on previous min -8 -5
Additional energy on previous stage kWh 770 589
Additional energy per passenger Mad-Bcn kWh 2.58 1.98
Additional energy cost per passenger /passenger 0.19 0.15
Ad. Energy cost as percentage of average ticket 0.16% 0.12%
Underlying time /h 1.45 1.78
AVE Madrid-Barcelona, direct, serie 103, 397 seats, 70% occupancy
Average ticket 120. Marginal cost 7.5 c/kWh
From the table we can deduce that:
On increasing the maximum speed from 300 to 330 km/h (+10%), the average
speed increases by 5.6% and consumption by 6.2%. On increasing the
maximum speed from 330 to 350 km/h (+6.1%), the average speed increases
by 3.6% and consumption by 4.4%. Therefore, we see that consumption (all
other variables being equal) does not increase with the square of the average
speed; instead, it increases almost linearly (exactly in proportion to Vmd
1.09
and
Vmed
1.01
).
2010

34
In the first stage each passenger would have to pay 29 eurocents to gain 8
minutes of journey time, and in the second stage each passenger would have to
pay 15 cents more to gain an extra 5 minutes. This means time values of 1.45
and 1.78 c/h, considerably lower than those shown in all the analyses of
demand, in which case it can be assumed that all the passengers would be
willing to pay these amounts, which would represent ticket price increases of
0.16% and 0.12%.
4.4. Energy used in constructing the vehicle
The energy used (and the emissions produced) during the vehicle construction
or manufacturing process (including all non-recurrent processes such as
extraction and transformation of materials, intermediate transport operations,
vehicle assembly and scrapping at the end of the vehicles operating life,
subtracting the energy value of the materials recovered in the break-up) must
be divided by the distance covered during the trains operating life, and also by
the seats offered, to obtain the energy (and emissions) repercussion of vehicle
construction.
Manufacturing energy
Thus, the energy component of vehicle construction or manufacturing, per seat-
kilometre, will be:
km seat kWh
R s
density energy man mass train
kilometre seat per energy ing manufactur . /
.
.
=
where:
train mass in kg; manufacturing energy density in kWh/kg; s is the trains
number of seats; and R is the distance covered by the train during its
operating life (in km)
The manufacturing energy density (for each ton of the trains mass) train
depends on the type of materials used, but not on the trains speed.
It can be assumed that a vehicle is used for about 7 hours on an average day,
regardless of its running speed. Therefore, the distance covered by a train
during its operating life is approximately proportional to the average speed.
As the distance covered by a train during its operating life increases with speed,
the repercussion of energy consumption (per seat-kilometre) decreases with the
average speed of the service increases.
Manufacturing emissions
Similarly, the emissions per seat-kilometre will be:
km seat kgCO
R s
factor emission manu CO mass train
emissions ringCO Manufactuu . /
.
2
2
2
=

Where:
2010

35
Train mass in kg; CO
2
manufacturing emission factor in kg CO
2
per kg; s in
the trains number of seats; and R is the distance covered by the train during its
operating life (km)
If we focus on the case of a high-speed train (30 tons for example), the energy
consumed in the manufacturing process is 31,902 kilowatts per ton and
emissions generated are 6,325 per ton kilograms CO
2
.
Manufacturing energy and emissions on several transport
modes
The energy needed and the emissions produced on manufacturing the vehicle
differ from one mode of transport to another.
Table 6. Consumption by seat kilometre depending on the transportation mode

Consumption
per ton
Emissions
per ton
Kilmetres
useful life
Consumption in
manufacturing per
seatkm
(kWh/t) (kgCO
2
/t) (km)
(kWh/seat(ton*)100km)
(l/seat (ton*)100km)
Cars (diesel)
Cars 25,497 4,818 200,000 0.42
Road 25,497 4,974 200,000 0.43
Cars (fuel)
Cars 25,494 4,818 200,000 0.44
Roads 25,497 4,818 200,000 0.30
Motorbikes
Scooter 26,977 5,183 150,000 0.11
Bus
Bus 22,973 4,952 1,000,000 0.061
Train
High Speed 31,902 6,325 15,000,000 0.23
Freight 20,883 4,642 3,250,000 2.910
-4
(*)
Boat
Container 18,531 4,567 3,000,000 2.510
-4
(*)
Aircraf
46,265 89,373 55,000,000 1.94210
-3

(*) kWh per capacity unit ton in 100 km
Comparison between manufacturing and operating energy and
emissions
In the case of a Spanish high-speed train class 100 (Alstom) with an average
tare of 392.6 tons, 332 seats and 14,052,500 miles covered in its operating life,
the percentage of energy consumption in manufacture with regard to the energy
consumed in the operating phase is approximately 4,87 %.
2010

36
Table 7. Energy consumption in manufacture in the case of high-speed train in the 100
series
Energy
Obtaining materials (MJ/veh) 26,615,706
Assembly (MJ/veh) 9,869,291
Scrapping (MJ/veh) 266,157
Total (MJ/veh) 36,751,154
Consumption per weigth (kWh/kg) 31.90
Consumption in manufacturing total (kWh) 12,524,742
Consumption in manufacturing total per seat (kWh)/Seat 37,725
Consumption in manufacturing per seat kilometre (kWh)/Seatkm 0.00268
Consumption on the operation per km kWh/km 14.92
Consumption on the operation kWh 209,663,300
% the Consumption in the obtaining materials % 3.53
% the Consumption in the assambly % 1.31
% the Consumption in the scrapping
% 0.04
% the Consumption in the manufacturing total
% 4.87
Knowing that this train has a consumption of 14.92 kilowatt hours per kilometre
in the operating process, energy consumption in the manufacturing (and
recycling) process is equivalent to what the train would consume in its operating
phase over 684,226 kilometres.
Since the high-speed train analysed covers 412,675 kilometres per year on
average, energy in manufacturing is equivalent to what it consumes in its
movement in 1.65 years.
Relationship with average speed
Consumption and emissions in the manufacturing process depend on the
materials used in vehicles and their weight, but not the speed at which it then
runs. Since consumption generated and emissions must be attributed per seat
kilometre, these emissions and consumption levels are divided by the average
annual distance travelled by trains.
Since the average annual distance covered by trains increases as their average
speed of movement increases, energy consumption produced in manufacturing
decreases.
So, for example, supposing 25 years-life for the train, in the case of a high-
speed train with a maximum speed of 250 km/h, energy consumption
distribution in manufacturing of 29.116 kWh/km and emissions of 5.773
kgCO2/km, if this maximum speed increases to 300 km/h, the consumption
level in manufacturing (and scrapping) changes to 24.738 kWh/km and
emissions to 4.905 kgCO2/km.
2010

37
Table 8. Incidence of train speed on the distributions of the energy used in construction
V max.
Distance
covered
operating life
Consumption in
manufacturing
Emissions in
manufacturing
km/h
km kWh/km kgCO2/km
200
7,381,768 1.383 0.274
210
7,667,514 1.331 0.264
220
7,952,883 1.284 0.255
230
8,229,468 1.240 0.246
240
8,497,579 1.201 0.238
250
8,765,565 1.165 0.231
260
9,025,405 1.131 0.224
270
9,285,188 1.099 0.218
280
9,629,084 1.060 0.210
290
9,972,980 1.024 0.203
300
10,316,875 0.990 0.196
310
10,660,771 0.958 0.190
320
11,004,667 0.928 0.184
330
11,348,563 0.900 0.178
350
12,036,355 0.848 0.168

Figure 7. Imputation of energy in terms of life
4.5. General view of the speeds effect and comparisson
with conventional railway
In the figure are shown the influence of the speed on the several monomions of
the energy consumption function.
2010

38
As can be seen, there are terminus not related with the speed, other increase
lineally with speed, others have a cuadratic relationship and, finally other
decrease with speed.

Figure 8. Train energy consumption and relationship with speed

(K1+Cc) x(M/s) xV
0
MECHANICAL
RESISTANCE
K2 xV
1
AIRINTAKE
RESIST.
K3x(C/s)xV
2
AERODYNAMIC
RESISTANCE
K4x(M/s)x(1/SD)xV
2
]x(1-RgC)
DECELERATE
BRAKINGLOSSES
MANUFAC.
ENERGY
+ + + +
ANCILLARY
SERVICES
K6xV
-1
DOWNHILLING
BRAKINGLOSSES
[K5x(M/s)xV
-1
]x(1-Rg)
+
TRAIN&INFR.
LOSSES
Txi
]X
[[[
]+
K7x V
-1
]x
LOAD
FACTOR
(p.km/
s.km)
TRAJEC.
COEFF.
(act.km/
ort.km)
Train energy consumption (per seat and orthodromic kilometer) and relationship with speed
+
H S O P E R A T I O N S Y S T E M
H S R O L L I N G S T O C K
H S I N F R A S T R U C T U R E
2010

39
5. Analysis of the effect of the shift of
traffic from other modes to high-
speed rail
When air services exist on a route, the largest part of the emission reductions
comes from the travellers the train attracts away from the plane. The trains
share of the market in relation to the plane over distances of 400 to 900 km
depends on the train travel time.
5.1. Modal spilt HS-plane
In the following graph we can see a characteristic curve that shows that when
the train has a journey time of less than 2 hours, it always obtains market share
over 85%; and if the journey time is more than 3 hours, the market shares falls
below 50%.
The graph shows the trains share of the train+plane market on various
European routes (prepared with data taken from Barrn, 2007, and completed
with Spanish data).

Figure 9. Relationship between the trains share of the train+plane market and the
journey time on the main world and Spanish routes with distances of between 400 and
600 km. Source: Barrn (2007) and independently produced
The points that correspond to each one of the cases have been joined by a third
degree polynomial fit line, whose equation is:
5 . 40 89 182 . 41 686 . 4
2 3
+ + = T T T share s Train
y =4,686x
3
- 41,182x
2
+89,21x +40,559
20
30
40
50
60
70
80
90
100
1 1,5 2 2,5 3 3,5 4 4,5 5
Train travel time (hours)
T
r
a
i
n

s
h
a
r
e
:

T
r
a
i
n
/
(
T
r
a
i
n
+
p
l
a
n
e
)

(
%
)
1
3 3.5 4 4.5 5 2
2.5
1.5
2010

40
If we assume that the time (T, in hours) is the distance by railway (L in km)
divided by the speed (V, in km/h), and P being the annual number of
passengers (train and plane combined), the number of train passengers is:
) 5 . 40 89 182 . 41 686 . 4 (
100
2
2
3
3
+ + =
L
V
L
V
L
V P
passengers Train
5.2. Optimization of speed from the energy consumption
point of view
From the foregoing it can be deduced that at low speeds, increases in the
trains speed produce (due to transference of traffic) emission reductions on a
corridor providing there is competition with the plane.
This is because a speed increase at low speeds produces a more than
proportional (or at least proportional) increase in the trains market share in
relation to the plane, and an increase in the trains speed produces a moderate
increase in its absolute energy consumption. However, high speeds produce
market share increases which are less than proportional to the trains increase
in speed (especially in the case of journey times of under 2.5 or 2 hours); and
furthermore, at high speeds there are significant increases in emissions when
the trains speed increases.
It can therefore be deduced that an optimum speed will exist in each corridor
from the emission point of view, this being the speed above which there will be
no additional reductions in energy consumption and emissions when the trains
speed increases.
Obviously, this optimum speed only affects energy consumption and
emissions, and there are other optimum speeds from the point of view of
operating costs, attracting passengers, total or external costs, etc. The decision
regarding the choice of the systems optimum speed will depend in each case
on a multi-criteria analysis in which the optimization of speed from the energy
point of view is just one of the criteria to be taken into account.
The optimum speed from the emission point of view is different for each
corridor, since this optimum speed depends on the length of the route by train,
the emissions per plane passenger on the specific route in question, the
function that relates the emissions per passenger to the trains speed, the
emission factor of the electricity in the country and year in question, and the
train-plane modal split function.
2010

41
As regards the variation in energy consumption (and therefore in emissions) on
varying the high speed trains maximum and average speed, the following order
of magnitude can be estimated -all other factors being equal- for a typical
Spanish case:
=
8805 . 0
412
371 . 0
V E
M
P

Where:
PM
E is energy imported en pantgraph (kWh/km.train)
M is the train mass (in t)
and V is average speed (in km/ h)
Case example
The ideas outlined above have been applied, by way of example, to the Madrid-
Barcelona route to find out, by applying the above methodology, the maximum
(and average) speed from which an increase in train speed results in increased
emissions in the corridor.
Emissions per passenger on the plane are estimated at 70 kg CO
2
per
passenger (Garca lvarez, 2007) and are independent of train speed.
Table 9. The energy consumption of a train has been calculated for different speeds
according to the average speed (imported and net in the pantograph) and its C02
emissions per passengers.
Maximum speed (km/h) 200 220 240 260 280 300 320 340 360 380 400 420
Average speed
without stops
(km/h) 160.8 176.1 191.3 206.4 221.3 236.1 250.8 265.4 279.8 294.1 308.3 322.4
Energy intaken at
pantograph level
(kWh/km) 10.67 11.49 12.39 13.37 14.44 15.57 16.78 18.04 19.36 20.75 22.22 23.76
Net energy (kWh/km) 9.52 10.39 11.34 12.35 13.45 14.63 15.88 17.20 18.59 20.05 21.58 23.17
Energy recovered (kWh/km) 1.15 1.09 1.05 1.02 0.99 0.95 0.90 0.84 0.77 0.70 0.64 0.58
CO
2
emissions (kgCO
2
/passenger) 8.95 9.77 10.66 11.61 12.65 13.75 14.93 16.17 17.48 18.85 20.29 21.79
From these data we obtain equations linking energy consumption to average
train speed.
2010

42

Figure 10. Energy intaken at pantograph and net energy per kilometre
The points obtained have been joined by a second degree polynomial, the
equation for imported energy being:
( )
train intaken
Mass V V E + =
2 5 2 7
10 163 . 2 10 787 . 2 10 867 . 5 [kWh/km]
And for the net energy:
( )
train net
Mass V V E + =
2 5 2 7
10 749 . 1 10 611 . 2 10 867 . 5 [kWh/km]
For a train with 318 seats, assuming a load factor of 0.65, an emission factor of
337 g C0
2
/kWh (Spain, 2007) and a distance of 620 kilometres by rail,
emissions per train passenger change with the speed as follows:

Figure 11. CO
2
emissions (kgCO
2
/passenger)
6,00
8,00
10,00
12,00
14,00
16,00
18,00
20,00
22,00
24,00
26,00
120 140 160 180 200 220 240 260 280 300 320 340
E
n
e
e
r
g
y
c
o
n
s
u
m
p
t
i
o
n
p
e
r
k
m
(
k
W
h
/
k
m
)
Averagespeed withoutstops (km/h)
Energyimportedinthe
pantograph
Netenergy
16.00
18.00
14.00
12.00
10.00
8.00
6.00
20.00
22.00
24.00
26.80
280 260 240 220 200 180 300 160 320 340 140 120
y=0,0002x
2
0,0084x+5,6055
3,00
5,00
7,00
9,00
11,00
13,00
15,00
17,00
19,00
21,00
23,00
120 140 160 180 200 220 240 260 280 300 320 340
C
O
2
e
m
i
s
s
i
o
n
s
(
k
g
C
O
2
/
p
a
s
s
e
n
g
e
r
)
Averagespeed withoutstops (km/h)
CO2emissions
23.00
Y =0.0002x
2
-0.0084x +5.6055
19.00
17.00
15.00
13.00
11.00
9.00
7.00
5.00
3.00
21.00
120 140
280
300 320 340
180 200 220 240 260 160
2010

43
The points obtained have been joined by a second order polynomial, whose
equation is:
6055 . 5 0084 . 0 0002 . 0
2
+ = V V Emissions [g CO
2
/passenger]
From the data for emissions per passenger on the plane (E
p,
in tons) and the
train (E
t
, in tons), annual passengers in the corridor (P=5,000,000) and average
train speed, which determines market share (TS, en %), it is possible to
estimate the combined emissions of the train and the plane on the route
studied:

By taking the common factor:
( ) ( )
t p t p
E TS E TS
P
Emissions + =
+
100
100

By substituting the known values:
( ) ( )
t t p
E TS TS Emissions + =
+
70 100
100
000 , 000 , 5

By substituting TS and Et by equations that link them to the average speed of
the train and knowing that the train distance between Madrid and Barcelona is
620 kilometres:
( )

=
+
6005 . 5 0084 . 0 0002 . 0
5 . 40
620
89
620
182 . 41
620
686 . 4
70 5 . 40
620
89
620
182 . 41
620
686 . 4 100
000 . 50
2
2 3
2 3
V V
V V V
V V V
Emissions
t p
By deriving this equation we obtain the peak at an average speed 284.983
km/h. And for 284.983 km/h average speed, the maximum speed is 367.412
km/h, and this one is, in this particular case, the optimum speed form the
energy consumption ad emissions point of view.

=
+ t p t p
E
TS
E
TS
P Emissions
100 100
100
2010

44

Figure 12. Annual emissions of train+plane depending on train average speed
1
4
5
.
4
1
5
3
.
1
3
0
1
.
2
3
0
8
.
3
3
1
5
.
4
3
2
2
.
4
3
2
9
.
4
2
5
8
.
1
2
6
5
.
4
2
7
2
.
6
2
7
9
.
8
2
8
7
.
0
2
9
4
.
1
2
1
3
.
9
2
2
1
.
3
2
2
8
.
7
2
3
6
.
1
2
4
3
.
5
2
5
0
.
8
1
7
6
.
1
1
9
1
.
3
1
9
8
.
9
2
0
6
.
4
1
8
3
.
7
1
6
0
.
8
1
6
8
.
5
250,000
175,000
150,000
125,000
100,000
75,000
50,000
25,000
225,000
200,00
2010

45
6. Best practices and recommendations
in high speed to reduce consumption
and emissions
6.1. General recommendations (not only for high speed)
The previous sections of this report have explained the inducers of the
consumption and emissions of trains in general and of high speed trains in
particular.
In view of this, it is easy to deduce which line design, train design and operation
measures could be adopted to reduce energy consumption and emissions.
These measures can be summarised as follows:
A. In relation to line design:
- Wide radius curves to reduce rolling resistance.
- Tunnels with a generous clearance gauge to reduce additional friction
drag.
- Homogeneous speed profile to reduce the energy lost in braking to
decelerate.
- Reduced gradients to avoid braking on downgrades.
- Reduced distance to reduce rolling and aerodinamic resistance between
the same points of passenger origin and destination.
- Use of high electrification voltages and reduction of distance between
substations to reduce ohmic losses.
- Installation of reversible substations to return regenerated energy to the
power network.
B. In relation to rolling stock:
- Reduction of the mass and rotating masses per seat to reduce rolling
resistance and the energy lost in braking.
- Optimization of the aerodynamic coefficient to reduce the aerodynamic
drag.
- Reduction of the amount of air that enters the train to minimize intake air
resistance
- Increase in the size of the train to obtain reductions in mass and in
coefficient C per seat unit.
- Utilization of high-performance traction chains to reduce losses in
vehicles
- Low heat transmission coefficients in the bodies to reduce losses in air-
conditioning.
C. In relation to operation:
2010

46
- Reduction in the number of train stops to reduce the energy lost in
braking.
- Increase in the utilization coefficient to reduce consumption per
passenger.
- Implementation of a time margin that allows economical driving.
- Implementation of an economical driving system (different according to
whether or not the train has a regenerative brake) which, if possible, can
be reprogrammed in real time.
- Reduction of turnover times.
6.2. Best practices applicable to high speed
The design and operation measures outlined above are applicable to all kinds of
trains. However, some of them are especially applicable to high speed systems,
three of which can be highlighted due to their particularly powerful effects at
high speeds:
1. Increase of feasible downgrade speeds.
2. Optimization of the trains exterior dimensions (length, width and height)
to reduce coefficient C per unit of capacity.
3. Reduction of air intake according to actual train occupancy.
6.2.1. Increase of speeds on gradients (downgrades)
Trains lose a significant amount of potential energy when they have to brake on
downgrades to avoid exceeding the maximum speed.
As has been shown previously, this quantity of energy grows as the difference
between the actual gradient and what we have called the gradient of repose
increases. This is due to the fact that the potential energy dissipated on the
gradient of repose is used to overcome the rolling resistance, and only when the
actual gradient exceeds the gradient of repose must the brake be used.
As the gradient of repose is that on which the rolling resistance (which
increases with speed) equals the gravitational resistance (which increases with
the gradient), it can be concluded that the greater the gradient of repose, the
higher the trains speed is. And as the loss of energy is greater the lower the
gradient of repose, it can be concluded that the lower the downhill speed of
the train, the greater the loss of energy.
This affirmation can be understood intuitively (even by using the experience of
driving a car): the slower one has to go downhill, the more one has to brake
and, therefore, the more energy is lost.
Consequently, a strategy to reduce energy consumption is to increase the
downhill speed of the train.
This can be achieved with an adequate infrastructure design (which includes
large curve radii in downgrade areas) and, once the infrastructure has been
constructed, by allowing the train to increase its downhill speed until it reaches
the maximum speed permitted by the infrastructure.
2010

47
As the layout of the infrastructure on high speed lines is usually straight or with
very wide curves, the advantages of increasing the trains maximum speed can
be found on these lines, whereas on conventional lines the downgrade sections
often coincide with curve areas, and therefore the trains maximum speed is
usually greater than the speed permitted by the infrastructure in these areas.
Consequently, nothing is gained by increasing the trains maximum speed.
The reduction of energy consumption by increasing the trains speed was
verified on the TGV South-East line in France when, in the 1980s, the speed
was increased from 260 to 270 km/h to better utilize the potential energy on the
numerous 35 mm/m gradients. This reduction was also verified in Spain by
increasing the speed on the Madrid-Barcelona line (with long 25 mm/m
gradients, where the maximum speed went up from 280 to 300 km/h). In
addition to this direct effect of reducing the energy dissipated in braking, there is
another indirect effect derived from running, at least on gradient sections, at a
higher maximum speed. In fact, this provides, for the same journey time, a
certain time margin which can be used to perform economical driving on
horizontal sections, where the driver can opt to coast rather than reduce the
trains speed (if the utilization of the regenerative brake is not high), or to reduce
the speed in relation to the maximum (if there is high utilization of the
regenerative brake).
If it were decided that the time gained by running faster on gradients results in a
reduction of the commercial journey time, this could indirectly help to reduce
emissions in the corridor due to transference of traffic from the plane, thanks to
the train being more attractive.
The following table shows, by way of example, the effect of increasing the
maximum speed of a train (series 102 AVE) on the Madrid-Barcelona line in
both running directions.
2010

48
Table 10. Effect of increasing the maxium Speedy of a train (series 102 AVE) on the
Madrid-Barcelona line in both running directions
Speed
Gradient
of repose
(mm/km)
Madrid-Barcelona Barcelona-Madrid
Energy
losts Both
directions
(kWh)
Excess
height
(mm/km)
Energy
lost
(kWh)
Excess
height
(mm/km)
Energy
lost
(kWh)
200 -6.85 3,479 1,995.3 2,537 1,454.9 3,450.2
210 -7.41 3,299 1,892.2 2,383 1,366.7 3,258.9
220 -7.99 3,118 1,788.3 2,227 1,277.4 3,065.7
230 -8.59 2,940 1,686.3 2,074 1,189.7 2,876.0
240 -9.22 2,760 1,583.2 1,922 1,102.5 2,685.7
250 -9.88 2,579 1,479.5 1,776 1,018.9 2,498.4
260 -10.55 2,403 1,378.6 1,638 939.5 2,318.0
270 -11.25 2,225 1,276.2 1,505 863.4 2,139.6
280 -11.98 2,042 1,171.5 1,371 786.6 1,958.1
290 -12.73 1,861 1,067.7 1,237 709.6 1,777.3
300 -13.50 1,680 963.7 1,107 635.1 1,598.8
310 -14.30 1,500 860.5 981 562.8 1,423.3
320 -15.12 1,326 760.7 855 490.3 1,251.1
330 -15.96 1,164 667.8 740 424.5 1,092.3
340 -16.83 1,007 577.3 630 361.4 938.7
350 -17.73 859 492.9 535 306.7 799.6
At 200 km/h (the speed to which a 6.85 mm/m gradient of repose corresponds),
the energy lost in the brake on gradients is 3,450 kWh (around 17% of the total
energy imported by the train on the route, including the energy needed for the
auxiliary services). If the same train runs at a maximum speed of 350 km/h, the
energy lost in the brake on downgrades is reduced to 800 kW/h (around 2,5 %).
6.2.2. Optimization of the trains exterior dimensions
High speed trains can adopt different architectures: concentrated traction or
distributed traction; articulated or non-articulated; single-deck or double-deck;
wide body or normal body. Also different sizes: From 250 to 1,600 seats.
As can be seen in Garca lvarez (2010), reducing the consumption of energy
per unit of capacity (i.e., per seat) can be achieved by reducing the mass per
seat, with a reduction of coefficient C per seat, and with an increase in the
trains capacity (i.e., the number of seats), which in turn affects the economies
of size consisting of a reduction of the mass per seat and of coefficient C per
seat. The same article demonstrates that the elasticity of consumption in
relation to the mass per seat is high in commuter services (0.436) and low in
high speed services (0.222). However, the elasticity of energy consumption in
relation to coefficient C per seat is high is high speed services (0.544) and low
in commuter services (0.025). These elasticity values can be seen in the table
below.
2010

49
Table 11. Elasticity of energy consumption with varing mass and coefficient C according
to type of service ( ) Coef or / /

Mass and rotating
masses Coefficient C
High speed 0.222 0.544
Conventional long
distance 0.414 0.190
Medium distance 0.390 0.072
Commuter 0.436 0.025
Metro 0.668 0.016
The same article also explains how the configurations of articulated and
distributed traction trains (even always being beneficial) have more effect on the
reduction of mass per seat and, therefore, on commuter services with frequent
stops. On the other hand, wider double-deck trains have a greater reduction in
coefficient C per seat and, therefore, display their consumption reduction effects
with greater efficiency in the field of high speed.
A train of the same capacity has a considerably lower coefficient C if it has a
wide body or two decks, and even more so if it has both. This is because
coefficient C is proportional to the trains cross-section surface area (its height
multiplied by its width) and, to a greater extent, to the trains wet surface area
(length multiplied by height times two plus width). Wide-body, double-deck
configurations increase the trains width and height, but, capacity being equal,
allow more than proportional reductions of the length, the combined effect being
to significantly reduce coefficient C per seat. It can be estimated that the energy
consumption of a wide-body, double-deck train with 350 seats is 38% less than
that of an equivalent normal-body, single-deck train, since although the former
is half a metre wider and higher, its length is reduced from 200 to 90 metres.
In addition to this, on lines with short platforms or heavy traffic density, wide-
body and/or double-deck configurations allow, for the same platform length, the
trains capacity to be increased, thereby producing two effects in terms of the
reduction of coefficient C per seat: one deriving from the smaller dimensions for
the same capacity, and the other from the greater capacity.
6.2.3. Air intake reduction
Moving trains devour a considerable quantity of air, around 10 cubic metres (12
kilograms) per person and per hour. This has two disadvantages from the
energy point of view:
The ambient temperature and humidity of the air that enters the train has
to be changed in order to meet the trains air-conditioning requirements
(temperature, humidity). The more extreme the outside temperature and,
logically, the greater the quantity of air, the more the amount of energy
required will be. This effect is obviously independent of speed.
Another effect deriving from air intake is a consequence of the fact that
the air that enters was initially still and must be accelerated almost
instantaneously up to the trains speed, which produces a retarding force
in the vehicle and, therefore, an increase in energy consumption. This
additional energy depends linearly on the trains speed (which is why it
2010

50
important in high speed systems) and on the air density (therefore it is
higher with cold temperatures), and it can represent up to 4% of the total
energy consumption of a high speed train.
Therefore, reducing the quantity of air that enters the train has an immediate
effect on the reduction of energy consumption, and this is especially
appreciable at high speeds.
Do the seats or the passengers need the fresh air?
International standards specify a quantity of refreshed air per person and hour
(between 5 and 15 cubic metres), but as the air refreshing system does not take
train occupancy into account, in order to ensure compliance with the standard,
in practice the air is refreshed by introducing the specified volume of air for each
seat instead of for each passenger. If the train is equipped with systems that
regulate air intake according to the actual occupancy of each car, a reduction in
consumption proportional to the percentage of empty seats could be achieved.
For example, for the typical utilization of 0.65, consumption could be reduced by
35%.
Intelligent air intake management systems could be used on high speed trains
not only to reduce energy consumption, but also to improve their performances.
In fact, on a long upgrade the amount of refreshed air could be reduced,
allowing the train to have additional power for traction, whereas on the
subsequent downgrade the quantity of intake air could be increased, thus
compensating for the previous lack of air and permitting an additional brake.
Carbon dioxide and air quality detection systems inside the cars (together with
the appropriate software) could help to improve air intake management and,
consequently, to optimize energy consumption.
Around 5.5% energy reduction
In the specific case of a direct high-speed train from Madrid to Barcelona (Talgo
class 102), the imported energy consumption in the pantograph is 15.53
kWh/km, with the net amount being 14.59 kWh/km. Of this energy, 1.88
kWh/km corresponds to additional drag due to air entering the train (12.87% of
net consumption), and it can also be estimated that 0.4 kWh/km (2.74% of net
consumption) corresponds to conditioning this mass of air.
So, a 35% reduction in the amount of air entering the train means a reduction of
0.657+0.140 =0.8 kWh/km, i.e. a 5.46% reduction of net energy consumption.
This reduction must be achieved with independent systems in each car that,
according to the programming established, regulate air intake depending on the
number of passengers, not the seats on the train.

2010

51
7. Bibliography
ALCOL, A. (2004): Resistencia al avance y aerodinmica de trenes, UPCO, Apuntes
Master de Sistemas Ferroviarios ICAI.
ALONSO, J. M. (2004): Conceptos aerodinmicos sobre el movimiento de los trenes.
Letter to the author dated 7 June 2004.
ANDERSON, E. and LUKASZEWICZ, P. (2006): Energy consumption and related air
pollution for Scandinavian electric passenger trains, Report KTH/AVE 2006:46,
Stockholm (Sweden).
ARAGN GURRA. E. (2005): Determinacin de las condiciones ptimas de
conduccin como forma de reducir el consumo energtico de los trenes, final-year
Industrial Engineering Project directed by Alberto Garcia Alvarez. ICAI, June 2005.
BARRN DE ANGOITI, I. (2007): Introduccin a la alta velocidad ferroviaria.
Apuntes del Master de Sistemas Ferroviarios ICAI, 6th edition.
BERNHARD, M. and GUIEU, M. (1976): Mesures rcentes de la rsistance a
lavancement de matriels roulants , in Revue Gnrale des Chemins de Fer, April
1976 issue, Pars.
GARCA LVAREZ, A. (2004): Dinmica de los trenes en alta velocidad , published
by Fundacin de los Ferrocarriles Espaoles.
GARCA LVAREZ, A. (2005): El tren de alta velocidad no es un depredador de
energa, in Dyna, June 2005, LXXX-5, pg. 33 to 38; edition updated in May 2007.
GARCIA ALVAREZ, A. (2006): Incidencia del tren de alta velocidad en el consumo
energtico y emisiones del sector transporte; speech given in Madrid on 17
November 2006 during the 6th Science Week (VI Semana de la Ciencia)
(downloadable at www.ffe.es>estudios y programas>publicaciones electrnicas).
GARCA LVAREZ, A. (2007). Consumo de energa y emisiones del tren de alta
velocidad en comparacin con otros modos de transporte. Anales de mecnica y
electricidad.
GARCA LVAREZ, A. (2007a): Normalizacin de los consumos energticos de los
trenes de viajeros, paper presented at the 3rd Railway Innovation Conference (III
Congreso de Innovacin Ferroviaria) (Tenerife, May 2007)
GARCIA ALVAREZ, A. (2007b): Consumo de energa y emisiones del tren de alta
velocidad en comparacin con otros modos, in Anales de Mecnica y Electricidad
(ICAI Engineers Association Magazine), Vol. LXXXIV, Fas. V, Sep.-Oct. 2007); and
expanded, with the same title, in Via Libre (issue 515, January 2008).
GARCIA LVAREZ, A. and FERNNDEZ GONZLEZ, E. (2008): Recorridos y cociente
entre trayectoria y desplazamiento en el transporte por ferrocarril; Notas tcnicas
Enertrans/9; January 2008, published by Fundacin de los Ferrocarriles Espaoles.
KEMP, R.J. (1993): The European High Speed Network, in Passenger Transport
after 2000 AD; published by Feilden, Wickens and Yates.
KEMP, R. (2004): Take the car and save the planet. Thought trains were always
greener than cars? Think again, IEE Power Engineer, October-November 2004.
LUKASZEWICZ, P. (2001): Energy consumption and running time for trains (Doctoral
Thesis). Railway Technology, Department of Vehicle Engineering, Royal Institute of
Technology, Stockholm.
INTERNATIONAL UNION OF RAILWAYS (UIC)
16 rue Jean Rey - F-75015 PARIS
Tel: +33 (0)1 44 49 20 20
Fax: +33 (0)1 44 49 20 29
PASSENGER DEPARTMENT - HIGH SPEED ACTIVITY
Legal deposit: December 2010
ISBN: 978-2-7461-1965-9
www.uic.org
Warning
No part of this publication may be copied, reproduced or
distributed by any means whatsoever, including electronic, except
for private and individual use, without the express permission of
the International Union of Railways (UIC). The same applies for
translation, adaptation or transformation, arrangement or
reproduction by any method or procedure whatsoever. The sole
exceptions - noting the author's name and the source -are
"analyses and brief quotations justified by the critical,
argumentative, educational, scientific or informative nature of the
publication into which they are incorporated" (Articles L 122-4 and
L122-5 of the French Intellectual Property Code).
International Union of Railways (UIC) - Paris, 2010

High Speed, Energy Consumption and Emissions

Uploaded by

Copyright:

Available Formats

High Speed, Energy Consumption and Emissions

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

High Speed, Energy Consumption and Emissions

Uploaded by

Copyright:

Available Formats

High speed, energy consumption

High speed, energy consumption and emissions

You might also like