HIGH SPEED DEPARTEMENT
OPERATING HIGH-SPEED LINES
CARRYING MIXED TRAFFIC:
EXPERIENCE GAINED AND CURRENT TRENDS
Prof. Dr. Ing. A. López Pita
Polytechnic University of Catalonia
UIC – High Sped Department – Mix traffic on high speed lines – October 2001
1. TYPOLOGY OF THE OPERATING OF HIGH-SPEED LINES 3
2. CURRENT EXPERIENCE IN THE OPERATING OF HIGH-SPEED
2.1 Germany ………………………………………………………… 8
2.2 Spain ……………………………………………………………. 18
2.3 France …………………………………………………………… 24
2.4 Italy ………………………………………………………………. 29
3. TRENDS IN THE CRITERIA USED FOR THE PLANNING AND
OPERATING OF NEW HIGH-SPEED LINES
3.1 Germany ………………………………………………………… 32
3.2 Spain …………………………………………………………….. 35
3.3 France …………………………………………………………… 35
3.4 Italy ………………………………………………………………. 37
4. CHIEF FACTORS TO BE CONSIDERED IN OPERATING
HIGH-SPEED LINES CARRYING MIXED TRAFFIC
4.1 Experience in operating conventional lines carrying
mixed traffic …….……………………………………………… 43
4.2 Formulation of the problem …………………………………. 48
4.2.1 Geometrical parameters ……………………………… 50
4.2.2 Effects of mixed traffic on damage to the track …. 52
22.214.171.124 The seventies ……………………………….. 52
126.96.36.199 The eighties and nineties …………………. 58
4.2.3 Impact of goods traffic on the capacity of a
high-speed line …………………………………………. 68
4.2.4 Operating criteria ……………………………………… 71
5. METHODOLOGICAL PROPOSAL FOR TAKING DECISIONS
WITH RESPECT TO THE OPERATING OF A HIGH-SPEED
LINE CARRYING MIXED TRAFFIC ………………………………… 75
6. SYNTHESIS ……………………………………………………………. 83
1. TYPOLOGY OF THE OPERATING OF HIGH-SPEED LINES
At the present time there are in Europe more than 3000 km of new lines suitable
for carrying high-speed traffic. In Japan the total length of equivalent lines is in
excess of 1800 km.
As will be seen from Table 1 each of the lines is suitable for a different type of
traffic, and the maximum speed at which trains are run varies within the range
from 230 to 300 km/h.
Examination of this table clearly shows that operating practice falls into two
categories, viz. the conventional ones of passengers and freight. The former is
applied on lines which accept only passenger traffic, whereas the latter relates
to the normal practice adopted in the operating of conventional lines up to
maximum speeds of 200 km/h, which involves running both passenger and
goods trains, usually referred to as mixed traffic.
We feel, however, that there is some confusion about the practical application of
the concept of mixed traffic run on high-speed lines. It would therefore seem
advisable to define the practice more precisely, since the terms ‘passenger’ and
‘freight’ may not be sufficient in themselves to dictate the technical and
economic viability of operating a mixed traffic line.
From a certain angle it is possible to conceive of high-speed trains suitable for
carrying goods. In this case a single type of rolling stock could be used subject
to the behaviour of the track and the capacity of the line.
At the opposite extreme is the AVE line between Madrid and Seville, used
solely by passenger trains, but by both high-speed trains and conventional
locomotive-hauled trains with more or less conventional rolling stock.
What we then have is a train operating system encompassing different
maximum speeds and different train formations. These two elements might well
have an adverse effect not only on the line capacity but also on the damage to
the track, thus leading to an increase in maintenance costs.
HIGH-SPEED LINES COMMERCIALLY OPERATED IN EUROPE AND
JAPAN, AND TYPOLOGY OF THE TRAFFIC CARRIED
COUNTRY LINE TRAFFIC FIRST YEAR IN
(km) CARRIED OPERATING OPERATION
Hanover – Würzburg Passenger 200 to 280
(327) Freight 100 to 160 1991
GERMANY Hanover – Berlin Passenger 200 to 280
(264) Freight 100 to 120 1998
Mannheim - Stuttgart Passenger 200 to 280
(100) Freight 100 to 160 1991
BELGIUM French border - Brussels Passenger 300 1997
SPAIN Madrid - Seville Passenger 200 to 300 1992
Paris - Lyons Passenger 300 1981/83
(417) Freight 160 to 200 1997
TGV Atlantique Passenger 300 1989/90
FRANCE (282) Freight 160 to 200 1997
TGV Nord Passenger 300 1993
TGV Rhône-Alpes Passenger 300 1994
TGV Interconnexion Passenger 270 1994
TGV Méditerranée Passenger 300 2001
Rome - Florence Passenger 250
ITALY (Direttissima) Freight 100 to 140 1977/84/92
Tokyo – Shin - Osaka Passenger 270 1964
Shin–Osaka–Okayama Passenger 230 1972
JAPAN Okayama-Hakata Passenger 230 1075
Tokyo-Monoka Passenger 240 1991
Omiya-Niigata Passenger 270 1982
Source: Prepared by the author using data supplied by the railway companies
This prompts us to conclude that the concept of mixed traffic – with its normal
connotation as regards conventional lines – might not be sufficiently meaningful
in the case of high-speed lines.
In the light of the above considerations we feel it would be advantageous to
redefine the classification of high-speed lines, where the crucial factor would not
be the kind of train running on them but rather the typology of the stock used.
Within this context we should like to propose the classification presented in
Table 2, which would make it possible to place the different high-speed lines
currently operating commercially into four groups: T1 to T4.
It is important to stress that this classification pays no heed to theory, but is
based on the practical consequences which each type of traffic has for the
technical and economic operating of the line. There is no doubt that from the
standpoint of wear and tear on the track it cannot be expected that specialised
trains with an axleload of 12 to 17 t running at 300 km/h would have a similar
effect on the track as conventional formations with axleloads of 20 to 22 t and
top speeds of 100 to 120 km/h.
CURRENT HIGH-SPEED LINE OPERATING SYSTEMS
TYPOLOGY TYPES OF TRAIN TECHNICAL
OF TRAFFIC FORMATION CHARACTERISTICS
In commercial service
* Trains designed * Paris – Lyons
Vmáx= 270/300 km/h
specifically for * TGV Atlantique
passengers P= 16/17 t axleload * TGV Nord and Interconnexion
* TGV Rhône-Alpes
(T1) * Tokyo – Shin – Osaka
* Shin – Osaka – Hakata
Exclusively * Tokyo – Morioka
passenger * TGV Méditerranée
* Omiya - Niigata
Planned in the short term
* Cologne – Frankfurt
* Trains designed Vmáx = 300 km/h In commercial service
specifically for Madrid - Seville
P= 16/17 t axleload
* Conventional Vmáx= 160/220 km/h
locomotive-hauled Madrid - Barcelona
passenger trains Locomotive : 20/22 t axleload
(T2) Coaches : 12 à 14 t axleload
* Trains designed Vmáx= 250/300 km/h In commercial service
specifically for * Hanover - Würzburg
P= 16/17 t axleload
passengers * Mannheim - Stuttgart
* Hanover – Berlin
* Rome - Florence
* Conventional Vmáx= 160/220 km/h
locomotive-hauled Planned in the short term
Locomotive : 20/22 t axleload
passenger trains * Karlsruhe – Basle
Coachess : 12 à 14 t axleload * Florence – Milan
* Barcelona – French border
Passengers + * Conventional Vmáx= 100/160 km/h
Locomotive : 20/22 t axleload
(T3) Wagons : 16 à 20 t axleload
* Trains designed Vmáx= 250/300 km/h
P= 16/17 t axleload
* Conventional Vmáx= 160/200 km/h Certain sections of the Paris-Lyons line
locomotive-hauled and the TGV Atlantique (since October
Locomotive : 20/22 t axleload 1997)
(T4) Wagons : 16 t axleload
Source: A. López Pita (2001)
So the following types of line are involved:
a) T1 lines, used solely by trains designed specifically for
These include the Paris-Lyons, TGV Atlantique, TGV Nord,
Interconnexion, TGV Rhône-Alpes and TGV Méditerranée lines and also
the Japanese high-speed lines. This service will also be provided on the
b) T2 lines, used only by specifically designed passenger trains
but also to some extent by conventional locomotive-hauled
These include the Madrid-Seville line used jointly by AVE trains and the
locomotive-hauled TALGO 2000 train. In the short term the same will
also apply to the new Madrid-Barcelona line.
c) T3 lines, used not only by trains designed specifically for
passengers, but also by conventional locomotive-hauled
passenger and freight trains
This group includes the new German lines Hanover-Würzburg,
Mannheim-Stuttgart and Hanover-Berlin, as well as the Italian line
between Rome and Florence. The Karlsruhe-Basle line will shortly be
added. In the longer term there will then be the Milan-Florence line and
after that the new line between Barcelona and the French border
d) T4 lines, used only by trains designed specifically for
passengers and conventional freight trains.
This is the position as it has existed in practice on certain sections of the
new Paris-Lyons and TGV Atlantique lines since October 1997.
Later in this paper we shall be examining the technical and economic
repercussions of each of these arrangements.
2. CURRENT EXPERIENCE IN THE OPERATING OF HIGH-SPEED
MIXED TRAFFIC LINES
The technical planning of the new high-speed lines in Germany began in the
early seventies. The Ministry of Transport and DB jointly analysed the criteria to
be applied in the preparation of the plans for the new railway infrastructure.
It should be recalled that the initial idea was to focus on the possibility of
running high-speed passenger trains in conjunction with operating freight trains
at 200 km/h carrying lorries. The objective was to examine ways of reducing
lorry traffic on German roads. The heavy financial cost of introducing this
operating system (Figure 1) and constructing the necessary new lines made it
impossible in practice to go any further along this route.
The subsequent studies implemented by DB and the German Ministry of
Transport were to show, during the first half of the 1970’s, that “as a first step,
maximum speeds of between 200 and 250 km/h for the passenger traffic and of 120
km/h in the case of the freight traffic would be perfectly adequate to meet
This then formed the basis for the planning of the first high-speed lines in
Germany, work on the section between Hanover and Würzburg being begun in
1973 and on the section between Mannheim and Stuttgart in 1976. A few years
later, in 1982, DB were to add that: “the polycentric structure of Germany meant that
there were no concentrations of population large enough to economically justify the
provision of a new line dedicated exclusively to carrying passenger traffic,” as can be
seen from Figure 2.
However, the population density along the Dortmund-Dusseldorf-Cologne axis,
coupled with other factors, resulted in the planning of the new Cologne-
Frankfurt line designed exclusively for passenger traffic.
Taking as a point of departure the two maximum speeds used as a reference,
viz. 250 km/h for passengers and 120 km/h for goods, the geometrical features
of the planned line (BLIND, 1980) were derived from the application of the
following mathematical expressions:
V2pass. - 1202
Rnormal = 11,8 ------------------------ ≈ 7.000 m
V2pass. - 802
Rminimum = 11,8 ------------------------ ≈ 5.100 m
It was accepted that the speed of freight trains would be 120 km/h in normal
radius curves on plain line, and 80 km/h in minimum radius curves. Side by side
with this it was accepted that the sum of the cant deficiency and of the excess
cant would be 89 mm and 130 mm respectively.
As regards the order of magnitude of the maximum falling gradients to be
adopted in the planning of the new lines, GLATEZEL (1980) observed that this
variable would have a marked effect on line capacity. In view of this DB felt that
12.5 ‰ should be the maximum permissible for mixed traffic operation. This
was dictated by the loss in speed which would result from any steeper
gradients, and also by the time it would take goods trains to start moving in the
event of their having needed to stop on such gradients. As a rough guide,
Figure 3 reproduces the graph prepared by ZEUGE (1975) of DB based on the
capabilities of the traction vehicles and rolling stock used at the time.
The Hanover-Würzburg and Mannheim-Stuttgart lines were opened for
commercial services in June 1991. What operating programme has been
employed on these two routes since then?
SPEED-DISTANCE DIAGRAM IN RESPECT OF CERTAIN
GERMAN TRAIN COMPOSITIONS
According to HEINISCH (1992) and JÄNSCH (1992) the maximum authorised
running speeds on the German high-speed lines at the time they were opened
for traffic were as follows (Table 3):
MAXIMUM AUTHORISED SPEEDS
ON THE FIRST GERMAN HIGH-SPEED LINES
TYPE OF TRAIN TRAIN MAXIMUM SPEED
Passenger ICE 250/280
Freight Intercargo (ICG) 120
Intercargo Express (ICGE) 160
Source : JANSCH (1992)
During the first year in operation (June 1991 to June 1992) the Kassel-
Göttingen section of the Hanover-Würzburg high-speed line carried traffic which
was more or less equally distributed between passenger trains and goods
trains, as can be seen from Table 4.
TRAFFIC ON THE KASSEL-GÖTTINGEN HIGH-SPEED LINE
AVERAGE NUMBER TOTAL
TYPE OF TRAIN TRAIN OF MOVEMENTS CIRCULATIONS
PER WORKING DAY
Passenger ICE 16 38
IC + FD 22
Freight ICG + ICGE 35 35
Source : JANSCH (1992)
From the standpoint of the traffic distribution over the day, HEINISCH (1992)
reported that goods trains operated for the most part at night, whereas
passenger trains ran throughout the day, so that the need to use passing tracks
was not very frequent.
It should be pointed out that at the time the first high-speed lines were opened
DB introduced the pilot service under the name of “Intercargo Express”. This
service was provided by freight trains running at top speeds of 140/160 km/h
and making use of the new section between Hanover and Würzburg. Figure 4
shows the repercussions which this had on the service between Bremen and
Hamburg to the north and between Stuttgart and Munich to the south. The
“Intercargo Express” service catered for combined traffic in 80% of the cases,
while the remaining 20% consisted of domestic goods traffic using sliding-wall
On account of the use of the new 160 km/h track on the routes mentioned
(Hamburg-Munich ≈ 800 km, and Bremen-Stuttgart ≈ 700 km) it was possible to
reduce the journey time by almost 2 hours, thus placing the railways in an
advantageous position as compared to the road by providing an adequate
timetable (depart ≈ 20 h, arrive 6 h). In fact the Hamburg-Munich run took 8
hours and the Bremen-Stuttgart run 7 hours. There are two trains a day on each
From the technical standpoint the stock used consisted of four-axled container
wagons type Sgas-y 703 and two-axled sliding-wall wagons type Hbills-y 307
(Figure 5). According to KRAMER and WACKERMANN (1992) increasing the
speed from 120 to 160 km/h required the adoption of better technology in
respect of both running gear and brakes (Table 5), as can clearly be seen from
Figure 6. Consequently, the financial investment in wagons was higher. The
above-mentioned authors indicated that the profitability of services of this type
would not be known until the pilot scheme programme had been completed.
EQUIPMENT ON WAGONS USED BY THE GERMAN RAILWAYS
FOR OPERATING AT 160 km/h
• Disc brakes
• Automatic load-proportional brake
• Electropneumatic brake control
• Wheel-slide protection device
Source : STIELER, G. (1992/93)
Ten years after the advent of high-speed services in Germany the speed of
freight trains on the new lines is 120 km/h for the most part, the commercial and
economic expectations of 160 km/h services not having been vindicated.
However, since the beginning of 2000 a “Parcels Intercity” train belonging to
Danzas Euronet and DB Cargo has been operating between Hamburg and
Munich nightly at 160 km/h, covering the distance of almost 800 km separating
the two cities in 7 hours 43 minutes.
The basic object of building the first high-speed line in Spain was to resolve the
problems associated with the lack of capacity on the existing single-track
section through Despeñaperros gorge (Figure 7) on the line between Madrid
The planning of the line was embarked upon in the mid-eighties, and the public
authorities responsible approved its construction at the end of 1986. So it is not
surprising that, being initially equipped with RENFE gauge track, the operating
system should have been made compatible with both passenger and goods
In view of the fact that the top running speed envisaged for the fastest
passenger trains was of the order of 250 km/h the geometry of the new Spanish
infrastructure was strongly influenced by the German criteria in respect of new
Consequently a maximum gradient of 12.5 ‰ was decided upon. Examination
of Figure 8 will show the necessity of a steep drop in altitude between the
Sierra Morena and Cordoba, making it essential on certain sections to apply the
maximum falling gradient continuously over significant distances: 10 to 22 km.
As regards the radius of plain line curves it was not recommended to follow the
German practice, so as not to excessively increase the breaks in construction
due to topographical difficulties on the few hundred kilometres between
Brazatortas and Cordoba. Consequently, although the normal practice was to
adopt a minimum standard radius of 4000 m and in exceptional cases one of
3250 m, the minimum radius was set at 2300 m on the section between
Ademuz and Villanueva de Cordoba (Figure 9).
In December 1988 Spain decided to introduce the standard international gauge
on all its new lines, as a result of which the route was, for obvious reasons,
used only for passenger services.
The new 471 km long high-speed line between Madrid and Seville was opened
for commercial services in April 1992, and was initially used only by the AVE
long-distance passenger trains.
However, this type of service was subsequently increased during the different
periods indicated in Table 6 by incorporating AVE shuttle trains (between
Madrid and Ciudad Real and Puertollano) and 200 km/h Talgo trains between
Madrid and Malaga, Cadiz, Huelva and Algeciras, the frequency of service of
these trains having increased little by little.
DEVELOPMENT OF COMMERCIAL SERVICES
ON THE NEW MADRID-SEVILLE LINE
PERIOD SERVICE (per day in each (per day in each
April 1992 AVE 6 20
Madrid - Seville
October 1992 AVE shuttle
Madrid-Ciudad Real-Puertollano 5 10
January 1993 Talgo 200 Madrid-Malaga 6
July 1993 Talgo 200 Madrid-Cadiz-Huelva 1 1
June 1999 Talgo 200 Madrid-Algeciras 1 1
Source: Prepared by the author using data supplied by RENFE
In addition to these services there are also a number of Talgo trains operating
on certain night routes between Andalusia and Catalonia.
It may therefore be said that at the current time there are: 30 AVE series (on the
busiest section, Madrid-Puertollano) and eight Talgo 200 services, which thus
means that 79% of the traffic consists of high-speed trains and 21% of
conventional trains of the type capable of 200 km/h.
In terms of tonnage carried by the track, and on the basis of a normal AVE train
weight of 420 t and a conventional Talgo train weight of the order of 280 t, the
AVE train tonnage represents 82% of the total and the Talgo 200 stock 18%.
The planning of the new lines in France suitable for carrying traffic at very high
speed, commenced in the mid-sixties. Since then the general practice has been
to reserve the new route layouts exclusively for specially designed high-speed
It should be mentioned that the only exception to this arrangement – and one
which we shall come back to later – is the section in the vicinity of the city of
Tours, which is on the TGV Atlantique line; this is also designed for carrying
The main reasons for adopting the above practice, which were outlined by
SNCF at the end of 1976, may be summarised as follows:
1) The reason for providing the new line was that the existing line was
saturated, and consequently it was considered particularly advisable to
envisage, for both lines, an operating system which would offer the
highest possible line capacity, thus delaying any future saturation.
2) A study undertaken in connection with the former line between Paris and
Lyons had shown that by eliminating goods trains it would be possible to
more or less triple the passenger movement capacity between the two
cities. Where two parallel lines are available it is obvious that the
optimum capacity of the system as a whole can be achieved if each of
the lines specialises in a specific type of traffic.
3) The damage sustained by the track as a result of traffic is caused
essentially by the heavy axleloads.
It may be stated that a line carrying wagons with 30 t axleloads and
locomotives with axleloads of up to 30 t in some cases, would be
preserving geometrical characteristics adequate for operating high-speed
trains at 250 km/h or more, at least not without incurring prohibitive
maintenance costs incompatible with intense traffic.
4) The greatest loads exerted by rolling stock on the track are applied in
curves on account of the centrifugal force. The carrying of mixed traffic
would make it necessary to reach a compromise as regards the cant and
use a cant intermediate between the one required for fast passenger
trains and the one suitable for freight trains. This is unacceptable from
both the technical and the ride point of view.
To overcome this incompatibility it would be possible to resort to the
device of increasing the curve radius, but that would mean a marked
increase in the cost of the line.
5) Eliminating freight trains from the line simplifies the problems associated
with the operating safety, an aspect to which special attention is being
In the light of these considerations SNCF adopted a form of geometry for the
new Paris-Lyons line suitable for the traffic it would be called upon to carry, i.e.
passenger trains only, using the available engine power and providing the
performances which it would be advisable to ensure in order to achieve
commercial viability. In particular:
• minimum curve radius: 4000 m
• maximum falling gradient: 35 ‰
Applying this falling gradient criterion a saving of 30% was made in
infrastructure costs as compared to a solution involving maximum falling
gradients of 15 ‰.
The basic guiding principles applied in the construction of the new Paris-Lyons
line were again adopted when planning the second high-speed line in France,
viz. the TGV Atlantique. The maximum permissible gradient was reduced from
35 ‰ to 25 ‰ on account of the different topographical features of the line.
The concept of a dedicated passenger line was retained, with the exception of
a short 17 km section in the vicinity of Tours, so that from the functional
standpoint mixed passenger and freight traffic is acceptable.
What was the reason behind such a decision?
Published arguments have emphasised the need to reduce the number of
conventional passenger and goods trains passing through the railway junction
of St. Pierre-des-Corps as this situation was starting to give rise to saturation
With this in mind three links were envisaged between the new line and the
existing ones, as can be seen from Figure 10 which shows the general
location, and from Figure 11 which shows the details.
The particular aims of each link will be as follows:
1) Saint-Pierre-des-Corps link
To enable TGV trains coming from Paris to enter the St. Pierre and Tours
2) Montlonis link
To allow conventional trains not needing to stop at St. Pierre to pass on the
3) La Ville-aux-Dames link
To enable freight trains coming from the south of the country to use the new
line and thus obtain direct access to the St. Pierre-des-Corps marshalling yard.
It may therefore be said that this new 17 km section of line constituted an
exception to the normal French criterion of using the new routes exclusively for
passenger traffic. The particular features of this exception are as follows:
a) The extension is a very short one covering only a few kilometres, so that
a direct and continuous check can be kept on the rate of deterioration of
the geometrical quality of the track, thus making it possible at any time to
discontinue or reduce the number of conventional train movements.
b) There are figures for the type and the number of conventional trains
which will be able to run over the line at high speed.
Ten years after its introduction into service this section was still used by only a
negligible number of freight trains.
The other planned and finished high-speed lines in France are still all dedicated
exclusively for use by passenger traffic. The chief design criteria in respect of
track geometry are summarised in Table 7.
It should, however, be stressed that since the end of 1997 some sections of
both the Paris-Lyons line and the TGV Atlantique line have been used every
night by two conventional freight trains running at a maximum speed of 160/200
km/h. These are made up of specially modified type 613 wagons mounted on
Y37A bogies and hauled by class BB22200 locomotives with TVN 430
signalling. The hauled load (consisting of urgent parcels) is of the order of 300
tonnes. The axleload of the locomotives is 22.5 t and that of the wagons 16 t.
OF THE HIGH-SPEED LINES IN SERVICE
PARIS-LYONS TGV TGV NORD TGV TGV RHONE TGV
ATLANTIQUE INTERCON- ALPES MEDITER-
Maximum speed in
(km/h) 270/300 300 300 270 300 300
Minimum plain line
curve radius (m) 3200 4000 4000 3250 4000 4500
Normal plain line
curve radius (m) 4000 6000 6000 4000 6000 7000
Maximum gradient ‰ 35 25 25 25 35 35
Track gauge (m) 4.2 4.2 4.5 4.2 4.5 4.8
Source: Prepared by the author from data supplied by SNCF
The Italian railways are probably the first in Europe to have had to define
operating criteria (for the new line between Rome and Florence). It should be
pointed out that in the 60’s the idea of high-speed trains used to such a wide
extent as they are today was not fully recognised.
So it is reasonable to mention that the Direttissima was not originally thought of
as being a line specially dedicated to high-speed traffic, as was the case with
the Paris-Lyons line, but merely as a means of continuing the railway tradition of
mixed traffic. On the other hand it was designed using geometrical standards
consistent with the long life of new railway infrastructure.
However this may be, in the mid-seventies FS stated that:
“The alternative consisting in the provision of a track specially
designed for high-speed traffic, which would have offered the
advantages of a basic interval timetable and lower construction
and operating costs, would not have been able to fully meet the
need for achieving maximum capacity while at the same time still
catering for freight traffic.
The solution therefore opted for was one which would immediately
provide a four-track line.”
The well-known result was that the Direttissima was introduced into commercial
service in a number of phases:
1981 ------------ first section -------------- 150 km
1984 ------------ second section ----------- 74 km
which meant that the line was not fully operational until 1992, while the ETR 450
sets capable of operating at 250 km/h had come into service four years earlier.
When the line was fully open the traffic carried each day and in each direction
on the Direttissima consisted basically of:
6 ETR 450 sets running at 250 km/h
15 IC or EC sets running at 200 km/h
30 express night trains and fast day trains running at 160 km/h
10 container trains
At the present time the typology and the characteristics of the stock used on the
Italian high-speed line are as indicated in Table 8.
PRINCIPAL CHARACTERISTICS OF VEHICLES USING
THE ROME-FLORENCE HIGH-SPEED LINE
TYPE OF STOCK CHARACTERISTICS
Axleload of most heavily loaded vehicle: 17 t
ETR 500 Axleload of the other vehicles: 12 t
Vmax = 250 km/h
PASSENGER TRAIN SPEED AXLELOAD
LOCOMOTIVE (km/h) (t)
E 646 140 18
E 652 160 17.6
E 402 200 22
PASSENGER COACHES Axleload: 16 t
FREIGHT TRAIN SPEED AXLELOAD
LOCOMOTIVE (km/h) (t)
E 636 100 17
E 445 120 19
WAGONS Axleload : 22.5 t
Source: Prepared by the author from data supplied by DI MAMBRO (FS)
3. TRENDS IN THE CRITERIA USED FOR THE PLANNING AND
OPERATING OF NEW HIGH-SPEED LINES
The above section outlined the experience gained in operating high-speed lines
carrying mixed traffic. We shall now go on to synthesise the criteria currently
adopted in the planning, construction and operating of lines at present under
construction or at an advanced development stage.
According to the revised planning schedule of July 1999 the opening of the new
lines for commercial operation and the modernisation of certain sections of
existing lines would take place within the time indicated in Table 9.
As regards the line geometry design criteria and the operating system
envisaged in each case it may be stated that:
1) The Cologne-Frankfurt line was planned with a minimum plain line radius of
3500 m, and in exceptional cases of 3250 m (similar to the radius adopted
the Paris-Lyons line) and with maximum gradients of 40 ‰. In some tunnels
the gradients were reduced to 28 ‰ (BELTER et al (1999)).
2) The Cologne-Frankfurt line will be reserved exclusively for passenger traffic,
not only on account of its gradients but also because of the estimated traffic
demands by the end of the decade (20 to 25 million passengers) on this
3) The other lines are designed for carrying mixed traffic, and consequently a
normal maximum gradient of 12.5 ‰ has been adopted, with the absolute
limit being set at 20 ‰, as is the case on the Nuremberg-Ingoldstadt section
ESTIMATED DATES OF OPENING TO COMMERCIAL TRAFFIC
OF UPGRADED AND NEW LINES IN GERMANY
LINE YEAR TO BE OPENED MAXIMUM SPEED
(km) TO COMMERCIAL (km/h)
Cologne – Frankfurt (175) 2002
≈ km 0 to km 5 80 – 160
≈ km 5 to km 27 200
≈ km 27 to km 28 250
≈ km 28 to km 162 300
≈ km 162 to km 168 220
≈ km 168 to km 172
≈ km 172 to km 175
Karlsruhe – Basle (210) 2004 200/250
Berlin – Halle/Leipzig 2022 200
Nuremberg – Leipzig
* Nuremberg –Ebensfeld (83) S.D. 200
* Ebensfeld – Erfurt (109) 250
* Erfurt – Leipzig/Halle (123) 2008 200/250
Nuremberg – Ingolstadt-Munich 2003
* Nuremberg –IngolstadT (83) 300
* Ingolstadt – Munich (88) 160/200
(with the exception of
a very short 2 km
section where it will
be 130 km/h)
Leipzig – Dreden ( ) 2002/2003 200
Cologne – Aachen (70) 2002/2005 70% at 250
Source: Prepared by the author from data obtained from (1) DB PROJEKT (1997); (2), (3)
and (4) JANSCH (1999); (5) H.D. KONNINGS (1999); SPANG et al (2000); (7) BLIND (1992).
The second high-speed line in Spain will link Madrid with Barcelona (≈ 625 km).
Its geometry has been planned to have a 7000 m curve radius on plain line, with
a view to permitting speeds of up to 350 km/h to be reached. The maximum
gradient will be 25 ‰.
On the other hand, the new international link between Barcelona and the
French border (Perpignan ≈ 170 km) will feature a normal radius of 4000 to
6000 m and a maximum gradient of 18 ‰.
On the former line (Madrid to Barcelona) an operating system geared principally
to passenger traffic is envisaged. The train compositions will consist of: sets
designed specially for high-speed operation (> 300 km/h) and conventional
modern locomotive-hauled trains with a maximum speed of 200/220 km/h.
Depending on demand, some car-carrying or freight services could be
scheduled, but the stock used for these would need to be suitable for speeds of
between 180 and 220 km/h.
As regards the international line between Barcelona and Perpignan the
operating system currently envisaged will involve the use of:
- high-speed trains (300 km/h)
- conventional stock with seated or sleeper accommodation (220 km/h)
- freight movements (principally containers and cars) (100/120 km/h).
It is envisaged that the two lines will open in phases for commercial services
during the period 2002/2004/2005.
Of the many projects for the construction of new lines suitable for carrying high-
speed traffic under consideration in France, some of the most advanced relate
to the following routes:
• TGV Est (Paris – Strasbourg)
• TGV Rhin - Rhône (Lyons – Mulhouse – TGV Sud-est)
• TGV Languedoc – Roussillon (Nîmes – Montpellier – Perpignan)
Only for the first of these is there an estimated date for the commencement of
commercial operation, the aim currently being for about 2006.
The geometrical design of the TGV Est line will be somewhat similar to that of
the existing high-speed lines in France. Before making it possible to operate
trains at 320 km/h as an initial step (350 km/h might be reached later) a route
layout with the following characteristics was planned:
Minimum: 6250 m
* Curve radius on plain line track:
Exceptional: 5556 m
* Maximum gradient: 35 ‰
The line will be reserved for passenger traffic, with an estimated demand of the
order of 11 million passengers by 2006.
As regards the TGV Rhin - Rhône, it should be pointed out that the plan is to
have three sections with different characteristics:
To link Genlis and Mulhouse (≈ 190 km)
This would initally link Genlis with Turay and later with the TGV Sud-est.
This would run the 150 km from Auxonne to Lyons and would be
intended for carrying mixed passenger and freight traffic (Figure 13).
The decision whether to run mixed traffic on the southern section would be
dictated by the fact that there are 280 trains a day currently operating on the
north-south major route and that over the coming years it is estimated that it will
be necessary to provide 500 to 600 paths a day. Hence the need for a new line
to carry mixed traffic.
As for the TGV Languedoc-Roussillon line, the new track linking Nimes and
Montpellier with Perpignan, it should be mentioned that around the periphery of
the two towns it is planned that the line will carry mixed traffic (Figure 14).
The planning criteria in respect of the design and operating of the new Italian
high-speed lines were formulated by a multi-disciplinary working party set up by
FS Headquarters in 1986. Figure 15 summarises the principal indicators
adopted for top operating speeds of 250 and 300 km/h. It will be noted that for
the higher speed a minimum plain line curve radius of 5417 m (subsequently
rounded up to 5450 m) was decided upon and a maximum falling gradient on
open track of 18 ‰, which was reduced to 15 ‰ for the two sections in tunnels.
The adoption of the above criteria for each of the planned lines resulted in the
following maximum falling gradients:
Turin - Milan: 15 ‰
Milan - Bologna: 12 ‰
Bologna – Florence 15 ‰
Rome – Naples : 15 to 18 ‰
It will be seen that the route through Modena (on the Milan-Bologna section)
prompted the adoption of a plain line curve radius of 3440 m, giving a maximum
speed of 240 km/h.
From the commercial operating standpoint the criterion adopted by the Working
Party in 1986 was to use the line for mixed traffic, and was worded as follows:
“The new high-speed lines will not be dedicated exclusively to ETR 500
sets, as is the case with the French TGV lines, but nor is it planned that
they should be totally reserved for mixed traffic.
The scheduled operating system is based on a selective traffic criterion
with trains of the same characteristics as regards speed and service:
- during the day: passenger trains
- during the night: trains made up of sleepers and couchette coaches
for long-distance journeys and certain freight trains.
For night routes a new specialised type of sleepers and couchette
coaches will be used, hauled by E402 locomotives (speed 200 km/h).”
Figures 13 and 14
Twelve years after the completion of the work of the High-Speed Working Party
the ideas formulated about the operating system of the new high-speed lines in
Italy are still valid.
This would appear to be confirmed if credence is placed on the ideas published
by MARZULLO (Italfer) and R. MANCINI (TAV) in December 1998 on this
subject, but with special reference to the new Rome-Naples line.
They stated the following in this connection:
“Priority on the high-speed line will be given to medium-distance and
long-distance daytime passenger services. Any remaining capacity will
be used during the night by another type of traffic: long-distance and
medium-distance passenger and freight services.”
As regards the daytime passenger traffic the above authors summarised (see
Table 10) the estimated services which would be provided by the ETR 500 and
ETR 480 sets.
ESTIMATED DAYTIME PASSENGER TRAFFIC ON THE NEW
ROME-NAPLES LINE USING ETR 500 AND ETR 480 TRAINS
NUMBER OF NUMBER OF
SERVICES PER SERVICES PER
SERVICE DAY WITH ETR 500 SERVICE DAY WITH ETR 480
Turin – Naples 8 Rome – Caserta 18
Turin – Bari 2 Rome – Bari 8
Milan - Naples 30 Rome - Lecce 4
Milan - Salerno 8 Rome - Taranto 4
Milan – R. Calabria 4
Trieste - Naples 4
Rome – Naples 14
Rome – Salerno 8
Rome – R. Calabria 4
TOTAL 82 34
Source: MARZULLO and MANCINI (1998)
As far as night-time passenger traffic is concerned the number of movements
expected on the new Rome-Naples line was as indicated in Table 11.
ESTIMATED NIGHT-TIME SERVICES
ON THE ROME-NAPLES HIGH-SPEED LINE
DOMESTIC SERVICES INTERNATIONAL SERVICES
(trains per day) (trains per day)
Tornio – Milan - Silice 8 Zurich - Milan - Naples 2
Venice – Bologna - Silice 2 D/Ch - Milan - Naples 2
TOTAL 10 4
Source: MARZULLO and MANCINI (1998)
Finally, in connection with freight traffic, it was pointed out that the new line
would enable the railways to fill an important slot in the market with a high-
quality service using complete train-loads of combined transport containers and
The freight trains would operate between 22 h and 6 h at a maximum speed of
120/140 km/h. Hauled with E402B locomotives and with a maximum axleload of
20 t the maximum possible hauled road in single traction mode would be 800 t.
The forecasts underlined the need to operate something like 37 trains a day
during the above-mentioned night-time period. Figure 16 presents the diagram
of programmed trains on the new Turin-Naples high-speed line, from which can
be seen the timetable slot reserved for freight trains.
4. CHIEF FACTORS TO BE CONSIDERED IN OPERATING HIGH-SPEED
LINES CARRYING MIXED TRAFFIC
4.1 Experience in operating conventional lines carrying mixed
The normal operating system with conventional railway lines has always been,
and still is, one which involves the configuration of passenger and freight train
This arrangement has been maintained in spite of the gradual increase in the
operating speed of both passenger and goods trains on lines which for the most
part were built in the 19th century.
In so far as the passenger sector is concerned, mention should be made, as a
rough guide, of the rise in the maximum authorised speeds on certain sections
of the French network, viz.:
1958 --------------- 150 km/h
1965 --------------- 160 km/h
1967 --------------- 200 km/h
In the case of 200 km/h traffic it should be pointed out that the length of the
track sections involved has increased over the years, as follows:
70 km --------------- 1967
430 km --------------- 1978
630 km --------------- 1984
At the present time the total length is in excess of 700 km. The graph
reproduced in Figure 17 depicts the geographical location of the sections of
track on the French network suitable for carrying traffic at 200 km/h.
As for the German network, on which it has also been possible for years to
operate trains at 200 km/h on conventional lines, the situation at the end of the
eighties, i.e. shortly before the introduction of the first high-speed lines, was as
indicated in Figure 18. The total length involved was some 629 kilometres
In the case of goods traffic the European maximum speed reference, for some
trains in commercial service, is as follows:
1962 --------------- 120 km/h
1975 --------------- 140 km/h
1987 --------------- 160 km/h
With the introduction of high-speed railway services the need to take full
advantage of the possibilities offered by these new lines became increasingly
evident. This, taken in conjunction with the design features of the high-speed
trains, meant that on certain sections of conventional lines the permissible
speeds of the trains mentioned exceeded the speeds authorised for
conventional rolling stock, even reaching 220 km/h.
This was in fact the case on sections between Le Mans and Nantes (Figure 19)
and between Tours and Bordeaux (Figure 20). This last figure illustrates the
differences as compared to the fastest conventional trains.
Consequently it may be stated that for less than a decade high-speed trains
running at a maximum speed of 200/220 km/h and freight trains running at
speeds between 100 and 160 km/h have co-existed on some conventional
lines, even though the trains travelling at 140 and 160 km/h represented only a
small proportion of the total traffic. By way of information, at the time when
authorisation was given for operating TGV trains at 200 km/h between Tours
and Bordeaux, there were more than 30 TGV services and more than 50 goods
trains (running for the most part at a top speed 100/120 km/h) operating on the
same line. This joint operating is all the more remarkable in view of the fact that
the distance between track centres on the route did not exceed 3.8 m.
Figures 19 and 20
In conclusion, Figure 21 provides a diagram of the maximum authorised
speeds for the TGV Atlantique trains on all the routes used by these services
and on which they pass freight trains.
So the experience gained in operating conventional lines carrying mixed traffic
may be summarised as follows:
* Maximum speed of passenger trains: 200/220 km/h
* Speed of freight trains: 100/160 km/h
* Distance between track centres: > 3.7 m
The above rough figures can be used as a reference when analysing the
operating of high-speed lines carrying mixed traffic.
4.2 Formulation of the problem
Two situations should be considered when analysing the problem of operating
mixed traffic on high-speed lines:
• The first would correspond to the possible introduction of freight trains onto
high-speed lines already being used by passenger trains. This would be the
same as the position when the French decided to authorise, with effect from
the end of 1997, the operating of a small number of freight services on
certain sections of the new Paris-Lyons and TGV Atlantique lines.
• The second situation would involve the criteria to be adopted in connection
with the planning and operating of a new high-speed line on which the
running of freight trains was also envisaged.
This document refers to this second hypothesis. In this context there would
seem to be four areas of particular interest:
1) The geometrical characteristics of the layout of the new line
2) The effects of mixed traffic in respect of the impairment of the
geometrical quality of the track
3) The impact of freight traffic on line capacity
4) The operating criteria (trains passing each other on the open track and in
tunnels, timetable slots for each type of service, etc.).
4.2.1 Geometrical parameters
As regards the first area, the geometrical parameters of the line, we feel that the
knowledge currently available is sufficient to enable us to specify the most
adequate orders of magnitude for these parameters. It is within this context that
we shall consider:
a) the principal plain line parameters (radius of curves, cant, cant deficiency
and excess cant)
b) the steepness and length of gradients.
On the basis of our knowledge of the speeds envisaged for the passenger
services (fixed as a function of the journey time considered necessary on each
particular run to enable a high-quality service to be provided) and for the goods
trains (fixed in order to achieve a given journey time), it is feasible, in the light of
the commonly accepted values for the cant deficiency and excess cant, for the
order of magnitude of the radius of curvature and of the cant to be evaluated, as
is well known. As a point of reference a minimum radius of at least 5400 m
might be set, and a maximum cant not exceeding 100/110 mm.
As far as the maximum gradient is concerned, it is self-evident that the flatter it
is the more suitable will the line be for freight traffic, and particularly for
increasing the hauled load. As will be remembered, Germany opted for 12.5 ‰
while Italy decided on 18 ‰ on open track and 15 ‰ in tunnels.
But there is no doubt that the precise choice of geometrical parameters for each
line is greatly influenced by the economic impact which this choice might have
on investments - as has been clearly shown by a number of studies - as well as
on operating costs (e.g. possibly entailing double heading for goods trains).
It will be recalled that, as LINKERHÄGNER (1984) pointed out, the analysis of
the Rethen-Kassel section on the new Hanover-Würzburg line showed that the
adoption of 18 ‰ gradients meant a saving of only 2% in infrastructure
investments as compared to the alternative option involving a maximum
gradient of 12.5 ‰ – a saving which in no way made up for the adverse
technical and economic effects of operating with an 18 ‰ gradient.
On the Paris-Lyons line, on the other hand, the adoption of 35 ‰ gradients
represented a saving of 30% in infrastructure costs as compared to the solution
involving the adoption of maximum gradients of 15 ‰.
To sum up, it is recommended that before embarking on a choice of geometrical
parameters – and particularly of the permissible gradient where the operating of
freight trains is envisaged – an analysis should be undertaken with a view to
quantifying the actual repercussions of the geometrical characteristics, at least
on the parameters indicated in the diagram in Figure 22, viz.: infrastructure
investments, operating costs and hauled load in the case of freight trains.
SOME OF THE VARIABLES INVOLVED IN THE ANALYSIS OF THE
MAXIMUM GRADIENT OF A HIGH-SPEED LINE
CARRYING MIXED TRAFFIC
Hauled freight load
a1 a2 a3 ai Min. falling gradient ‰
Source: Prepared by the author Figure 22
On the other hand, if a certain maximum rising gradient is adopted it is essential
to have an overall picture of the line in order to fix the acceptable length of the
corresponding continuous falling gradient. Side by side with this, the acceptable
linking together of gradients of different lengths and steepness coupled with the
problems which might arise from standpoint of the braking of freight trains, may
involve wrong-direction running (Figure 23).
4.2.2 Effects of mixed traffic on damage to the track
188.8.131.52 The seventies
We shall make reference below to one of the aspects which, generally
speaking, has always been a feature of any difficulty likely to be encountered
when operating a high-speed line carrying mixed traffic, viz. the adverse effect
which freight trains might have on the geometry of the track.
As pointed out in Chapter 2.3 one of the arguments put forward in France in
support of the decision to reserve the new high-speed lines exclusively for
passenger traffic, was precisely the damage which freight wagons and
locomotives would do to the track. Consequently, as we have already
mentioned “it would not be possible to preserve the geometrical characteristics
adequate for operating high-speed trains, at least not without incurring
prohibitive maintenance costs incompatible with intense traffic.”
Within the context of this document we feel it would be advisable to analyse the
forces exerted on the track by the different types of vehicles which might need
to make use of a high-speed line carrying mixed traffic. These stresses,
considered in the vertical direction, will provide a reliable pointer as to their
effect on the deterioration of the track geometry.
A MOTIVE POWER
Two periods need to be distinguished. The first would cover motive power stock
available up to the mid-seventies, which is the time when decisions were taken
concerning the design and operating characteristics of the first high-speed line
between Paris and Lyons. The second would cover the last two decades, a
period during which new types of locomotives began to appear.
The first category would include the French BB9200 and CC6500 series of
locomotives, the German 103 series and the Italian E444 series. The second
category would comprise, among others, the German E120 series which was
introduced in 1978, the French BB2600 (Sybic) series which came into service
in 1988, the Italian E402A dating from 1986/89 and the Spanish series S252
which was introduced into service in 1992. Table 12 shows the principal
characteristics of some of the locomotives in this group.
MAXIMUM SPEED AND AXLELOAD OF SOME OF THE MOST MODERN
EUROPEAN RAILWAY LOCOMOTIVES
MAXIMUM AXLELOAD PERIOD OF
LOCOMOTIVE SPEED BASIC AREA OF USE
Rh 1016 230 21.5 Passenger-Freight 2000/2002
E 101 220 21.7 Passenger-Freight 1996/1997
BB 26000 200 22.5 Passenger-Freight 1988/1997
E 127 230 21.5 Passenger
E 145 140 21.5 Freight 1997/2000
E 152 140 21.5 Freight 1997/2001
BB 36000 220 22.5 International freight traffic 1997/1999
E 402 B 220 21.7 Passenger on the Direttissima 1997/1999
E 412 200 22 International passenger traffic 1995/1997
Re 4/4465 230 21 ---- 1994/97
S 252 220 22 Passenger 1992/95
Source: Prepared by the author from data obtained from WART (1997), MARTÍN (1993),
MARINI (1997), HERISSÉ (1996) and DE CHAREIL (1997)
The normal mean axleload of passenger vehicles is of the order of 12 to 14 t.
For the purpose of this study we shall be considering axleloads of 16, 18, 20
and 22.5 t. The relationship of the axleload to the running speed is shown in
Table 13 as recommended by the UIC.
TYPICAL FREIGHT TRAIN COMPOSITIONS
120 140 160
AXLELOAD (t) 22.5 20 18
HAULED LOAD (t) 1600 1600 1200
LENGTH OF TRAIN (m) 700 700 700
From the standpoint of the deterioration of the track, the axleload is only a
relative indicator, which is why the load actually transmitted to the track is of a
dynamic nature and is a function (apart from the nominal static weight of each
• the design characteristics of the condition of each vehicle,
• the running speed,
• the design features of the track and of its geometrical quality.
It is a well-known fact that there are mathematical formulas for determining the
dynamic stresses produced by each type of rolling stock, but there is no doubt
that it is more reliable to make use of practical quantification in actual operating
In this area the different European railway administrations have applied a
number of measures which are highly useful in analysing the contribution of
each vehicle to the deterioration of the track.
If we consider the traction vehicles of the sixties and seventies and look at the
programmes of tests undertaken by the French railways during that period we
obtain the results shown in Table 14.
VERTICAL LOADS EXERTED BY LOCOMOTIVES SERIES BB 9200
RUNNING AT 140 AND 200 km/h
VERTICAL DYNAMIC SPEED (km/h)
WHEEL LOAD (t) 140 200
Mean 11.3 14
Maximum 15.8 16.3
Source : Prepared by the author from data provided by SNCF
The details provided in Table 15 correspond to the vertical dynamic stresses
applied by BB 9200 series locomotives with a nominal wheel load of 10.3 t in
commercial service (speed 140 km/h), or in trials at up to 200 km/h. What is
more, the results published by BIRMANN (1968) in respect of the German
locomotive E 10 2999 show that for speeds up to 200 km/h the measured
vertical dynamic loads are also within the range of 14 to 16.6 t per wheel.
As regards passenger coaches, since their nominal axleload is fairly low, it may
be expected, in view of their design characteristics, that the damage they cause
to the track will be very limited.
In fact, if we consider the dynamic effect of this type of vehicle operating at the
usual maximum speeds for conventional tracks of 160 and 200 km/h, the results
will be approximately as shown in Table 15 for a 13 t axleload.
VERTICAL DYNAMIC LOADS EXERTED BY PASSENGER COACHES
Speed Maximum vertical dynamic wheel load
Source: Prepared by the author from data supplied by SNCF and taken from
In the case of older freight wagons with an axleload of 20 t, the vertical loads
applied to the track are between 15 and 16 t per wheel (PRUD’HOMME, 1970)
for speeds of between 70 and 100 km/h. These represent levels of
aggressiveness of the same order as those due to traction vehicles.
It was thus found that within the speed ranges considered for each type of
rolling stock, the stresses exerted by traction vehicles were similar to those
exerted by freight wagons. The effect of passenger coaches was relatively
Finally, the measurements taken by SNCF with high speed trains, both on
conventional lines and on the new Paris-Lyons line, confirmed that the
maximum vertical load exerted by the wheel of a TGV train running at 200 to
300 km/h was within the range indicated in Table 16.
INFLUENCE OF TRACK QUALITY
IN THE VERTICAL EFFECTS OF HIGH-SPEED TRAINS
TYPE OF RUNNING LOAD ON THE
STOCK SPEED TYPE OF TRACK TRACK
TGV 200 Major conventional line 12.15
TGV 300 Major conventional line 13.95
TGV 300 Paris-Lyons high-speed line 12.35
Source: Prepared by the author from data obtained from ALIAS (1990)
An examination of this table shows that on a high-speed line the operating of a
TGV train at 300 km/h resulted in the same level of stress as that exerted by the
same rolling stock running at 200 km/h on a conventional line.
By way of summing up the graph in Figure 24 shows the levels of vertical
loading on the track as exerted by the rolling stock which was available in the
seventies, and also by the high-speed trains, each of the trains running at the
speed normal for it in commercial service.
It will be observed that the vertical loads transmitted by traction vehicles to the
track might, at speeds of 200 km/h, reach levels 44% higher than the loads
exerted by TGV trains at 300 km/h. In the case of freight wagons the maximum
differences were of the order of 28%.
On the basis of these results, and in view of the topographical difficulties
encountered in the Paris south east corridor with the new high-speed line at Col
du Bois Clair, the expected heavy traffic demand (10/12 million passengers a
year), and since this was the first railway line intended to be operated at
260/270 km/h, there was every reason to approve the decision of the French
railways that the new route should be dedicated to carrying passenger trains.
184.108.40.206 The eighties and nineties
A feature of the last two decades of the 20th century was the introduction into
commercial service of trains specially designed for operating at maximum
speeds of between 250 and 300 km/h. Table 17 summarises the speeds and
axleloads of each train.
SPEEDS AND MAXIMUM AXLELOADS OF HIGH-SPEED TRAINS
TRAIN DESIGNATION MAXIMUM REFERENCE MAXIMUM AXLELOAD
ETR 450 250 13.5
ETR 460 250 12
ETR 500 300 17
ICE I 280 19.5
ICE 3 300
AVE 300 17.2
TGV 300 17
Source: Prepared by the author from FS, DBAG, RENFE and SNCF data
In respect of the dynamic load transmitted by the wheel, for each of these
trains, it may be stated that with a speed of 250 km/h and an axleload of 13 t,
the maximum value is less than 10 t. In the case of trains running at 300 km/h
with an axleload of 17 t the maximum wheel load is of the order of 12.5 t.
During this particular period a large number of locomotives were built which
gradually replaced those designed in the fifties and sixties. Table 18 shows
some of the classes of locomotives used in the major European countries,
designed for carrying passengers at speeds of up to 220/230 km/h, or for
PRINCIPAL MODERN LOCOMOTIVES IN EUROPE
COUNTRIES IN MAXIMUM
WHICH USED SPEED
Austria Rh 1016 230 6400
Germany E 101 220 6400
France BB 26.000 200 5600
Germany E 145 140 4200
Germany E 152 140 6400
France BB 36.000 220 5600
Italy E 402 B 220 5600
Italy E 412 200 6000
Switzerland Re 4/444.465 230 7000
Belgium T13 200 5000
Spain S 252 220 5.600
Source: Prepared by the author
At the same time there have been a number of changes in the fleet of goods
wagons, with some railways equipping themselves with vehicles capable of
140/160 km/h, as has already been pointed out.
And what about the dynamic behaviour of this new type of stock from the
standpoint of its effects on the track?
As regards the traction vehicles the results of the tests carried out with E 120
locomotives in Germany at speeds of between 120 and 200 km/h indicated
maximum vertical wheel loads of almost 14 t (NAUE, 1988).
In so far as freight wagons are concerned, Figure 25 clearly shows the
influence exerted by the geometrical quality of the track on the performance of
the stock. It can be seen that for a high-quality line, which is the category into
which a high-speed line would fall, the vertical load exerted by modern freight
wagons capable of speeds up to 140 km/h does not exceed 13 t.
As for the repercussions of the track quality, it is worth recalling that the work of
the ORE D 161 Committee resulted in the formulation of the following equation:
Qd = QE + [12 + 0,60 QE + 0,51 (V-50) (℘-0,5) ]
which gave – for the rolling stock considered in this study, i.e. wagons with 20
and 25 t axleloads – the relationship between the dynamic wheel load (Qd ) in
KN, the static load (QE) and:
V = running speed (km/h)
℘ = standard deviation of the longitudinal profile of the track (mm).
Three types of track were considered in the study:
Category 1: Very high quality ℘ = 0 to 1
Category 2: Good quality ℘ = 1 to 2
Category 3: Average quality ℘ > 2
With a wagon speed of about 110 km/h the vertical wheel load is of the order of
magnitude indicated in Table 19.
EFFECT OF THE GEOMETRICAL QUALITY OF THE TRACK ON
THE VERTICAL LOADS EXERTED BY FREIGHT WAGONS
Geometrical quality of Vertical wheel load Relative index
the track (KN)
(℘ > 2 mm) 155 122
(1 < ℘ < 2 mm) 142 112
(℘ < 1 mm) 127 100
Source: Prepared by the author from data obtained from ESVELD et al. (1988)
The changes which have taken place in the design features of railway rolling
stock, especially in respect of locomotives and wagons, has significantly
reduced the damage caused to the track. Table 21 is particularly indicative in
this regard, while Figure 26 depicts, in the form of a flow diagram, the
information contained in the table.
COMPARATIVE DEVELOPMENT OF MAXIMUM VERTICAL LOADS
EXERTED BY RAILWAY ROLLING SOTKCK
Maximum vertical wheel load
Type of stock (t)
V = 200 km/h 17.5 13 t 13.5
V = 70 km/h 14.5 t 15.7 --
V= 100 km/h -- 12.6
V= 120 km/h -- 13.3
V= 200 km/h 6.9 (TEE) to 8.3
V= 200 km/h 12.15
V= 300 km/h 13.95
V= 300 km/h -- 12.3 to 12.5
Source: Prepared by the author
There is no doubt that the differences which existed a few decades ago (in the
seventies) between high-speed trains and locomotives and wagons – whose
effects on the track might have been of the order of 44% to 28% respectively –
have now been reduced to levels of between 10 and 15% between the two
In the light of the above results it might be wondered to what extent the different
levels of vertical loads play a part in the deterioration of the track. A knowledge
of this parameter would make it possible to attempt to quantify the economic
implications, the increased maintenance and the operating costs of certain high-
speed lines used for mixed traffic.
The criterion normally adopted to quantify the effect of the vertical loads on the
track and on the deterioration of its geometrical quality is the well-known power
where QI and Qj are the two wheel loads considered and α is a coefficient which
assumes a value of 4 in the context of the deterioration of coatings whence this
analogy is taken.
However, the studies carried out, more especially in the European railway area,
during the seventies and eighties of the last century indicate appreciably lower
α values, as can be seen from Table 21, viz. between 1 and 1.2. This order of
magnitude tallies closely with Japanese experience.
VALUE OF THE EXPONENT (α) IN THE POWER RULE
Author Year Value of α
Mc CULLOGH 1972 1
ANDREYEV 1974 1 to 1.2
PENNYCOOK 1976 1
HENN 1978 1.2
JANIN 1982 1
SATO 1984 1
SHENTON 1985 1
Source: Prepared by the author
As regards the exponent of the power rule, when it is applied to traffic density
most authors agree that it should be assigned a value of 0.2 to 0.3.
To sum up, we have available to us information which will assist in quantifying
the impact of operating a high-speed line with mixed traffic on the cost of
maintaining the track geometry.
In any case, as regards the order of magnitude of these costs it is interesting to
note the graph in Figure 27, prepared for the document “Maintenance des
lignes à grande vitesse” (“Maintenance of high-speed lines”) by the UIC Way
and Works Committee in April 1996.
According to this graph, for about the first 4 or 5 years of operating a high-
speed line carrying mixed traffic this parameter would not be significantly
different from what it would be for a line used exclusively for passenger traffic.
During the period from the 5th to the 15th year in commercial service, the
increase in deterioration would not be greater than 20/25%.
Unfortunately, neither the graph mentioned nor the document containing it
makes any precise reference to the practical significance of “high-speed lines
with mixed traffic”, so that the relative deterioration referred to above cannot be
attributed to a particular type of traffic.
However this may be, the order of magnitude previously indicated (20% to 25%)
corresponds closely to the differences outlined above for the aggressive nature
of freight stock (including traction vehicles) as compared to high-speed trains,
with each modal composition operating at its usual speed, i.e. 10 to 15%, and
the application of the power rule with an exponent of 1.2.
Finally, it should be pointed out that at the time of the planning of the new line
between Barcelona and the French border (Perpignan), designed for carrying
high-speed passenger trains and goods trains, it was estimated that the
additional track maintenance costs which would be incurred on account of the
mixed traffic arrangement would be of the order of 25%.
One particular aspect of the graph in Figure 27 which should be emphasised is
the lower level of deterioration, and consequently the reduced need for
maintenance of high-speed passenger lines as compared to conventional lines
carrying mixed traffic.
In support of the above statement it might be pointed out that in 1991 SERVI
(SNCF) stated that the maintenance cost per km of high-speed line in France
represented 55% of the cost of maintenance of a conventional line carrying the
same amount of traffic (≈ 35 000 to 45 000 t/day).
Over a similar period of time THOMAS (SNCF) indicated that for the economic
conditions prevailing in 1989 the cost of maintaining the track of the Paris-Lyons
line was in the region of 94 000 francs per kilometre of line, i.e. 47 000 francs
per kilometre of track. This figure was to be later confirmed by SCHAER
(SNCF) in 1993.
Recently Professor BAUMGARTNER (2001) pointed out that the average cost
of maintaining a kilometre of track on a line designed for a maximum speed of
300 km/h was between €10 000 and €30 000 a year. The figures for the French
line lie within this range if allowance is made for the changes in economic
In conclusion it may be stated that there is no doubt that operating a high-speed
line for freight traffic will imply an increase in track maintenance costs as
compared to operating a high-speed line used exclusively for passenger traffic.
The additional cost may be of the order of 20 to 25%, but in view of the fact that
the cost of maintaining a purely high-speed line will be about 40 to 45% lower
than the cost of maintaining a conventional line carrying the same total amount
of traffic, it may be concluded that the maintenance cost of a high-speed line
carrying mixed traffic will be lower than the corresponding cost for a
conventional line carrying the same tonnage of traffic.
4.2.3 Impact of goods traffic on the capacity of a high-speed line
As has been intimated in section 2.3, one of the reasons prompting SNCF to
dedicate the new high-speed line between Paris and Lyons exclusively to
passenger traffic was precisely the fact that a study undertaken on the existing
line had clearly shown that if freight trains were eliminated it would be possible
to more less triple the capacity of the line.
What is more, there is no doubt that the investments required for constructing a
new railway infrastructure are very high, and that consequently it is advisable to
design it using geometrical parameters which will make it possible to increase
the commercially desirable and technically achievable speeds.
In this sense the provision of high quality passenger services attracts great
demand, thus making it difficult to insert less profitable services (as freight trains
It was within this context that, in the mid-nineties, the Tokyo-Osaka high-speed
line was carrying annual traffic in excess of 130 million passengers. In Europe,
albeit to a lesser extent, but still with relatively high levels, the first three high-
speed lines built in France (TGV Sud est, TGV Atlantique and TGV Nord) each
carried more than 20 million passengers a year.
In terms of number of runs the Japanese line in question exceeded the average
of 260 trains a day at the time. As for the French high-speed network, Figure 28
shows the annual changes in the number of commercial runs on the Paris-
Lyons line which have taken place since the whole of the line was put into
commercial service. It will be observed that the number of daily movements
(220 on working days, with 250 on Fridays and up to 300 on the days at the
start and end of holiday periods) makes it difficult, if not impossible, to operate
any other goods trains than the two SERNAM night services referred to above.
It is no less certain, however, that all the new high-speed lines already
constructed carry the same volume of traffic: in Germany about 11 million
passengers, and in Spain, along the Madrid-Seville corridor, something of the
order of 5 million passengers. There is no doubt that these volumes are such
that we might, in terms of capacity, envisage the advisability and feasibility of
running freight trains on these lines, if this were commercially to be
recommended. It would seem from the data published in respect of the Spanish
corridor that the number of passenger trains per day in each direction would be
of the order of 40 movements. The different track gauge means that there would
be no advantage in operating freight trains in so far as there would no capacity
The opposite situation occurs on the Barcelona-Perpignan corridor where only
the adoption of the new international gauge infrastructure would make it
possible to offer a quality service for freight traffic. The estimated passenger
flow on this line during the initial stage would probably not exceed that figures
indicated for the Madrid-Seville corridor.
Confronted with a different set of problems on each corridor it is impossible to
avoid the need to undertake a specific study for each case with a view to taking
a decision as to whether or not freight traffic should be authorised. A possible
methodology could then formulated to enable a concrete decision to be taken.
4.2.4 Operating criteria
Let us assume that the capacity problems have been overcome: there would
still remain a certain number of highly important questions relating to operating.
Consider the following:
1) Passing speed of passenger trains and goods trains in the open air and in
2) Effect of freight trains on the maintenance frequency of installations and on
3) Signalling system
As far as the first question is concerned it should be borne in mind that current
experience in France is limited to the speed range defined by the 220 km/h
maximum speed of the TGV on certain sections of conventional line and the
140/160 km/h speed attainable by some goods trains.
On a high-speed line, however, the fastest passenger services will operate at
speeds of up to 250 and 300 km/h. The increase in the present distance
between track centres on the high-speed lines is considerable as compared to
the spacing on conventional lines (3.7 to 3.9 m), as can be seen from Table 22,
but a certain degree of caution should be exercisd as regards the validity and
rationality of extrapolating available knowledge to other cases.
DISTANCE BETWEEN TRACK CENTRES ON CERTAIN
HIGH-SPEED LINES IN EUROPE
Distance between track
Line centres Year of inauguration
TGV Sud est 4.2 1981/83
TGV Atlantique 4.2 1989/90
TGV Nord 4.5 1993
TGV Rhône-Alpes 4.5 1992/94
TGV Méditerranée 4.8 2001
Hanover - Würzburg 4.7 1991
Mannheim - Stuttgart 4.7 1991
Cologne – Frankfurt 4.7 2002
Rome - Florence 4.0 1992
Florence – Milan 5.0 2004/2006
Madrid - Seville 4.3 1992
Madrid - Barcelona 4.7 2002/2004
Barcelona - Perpignan 4.8 2005
Source: Prepared by the other using data supplied by the railways
In Table 22 it will be observed that the distance between track centres on the
new lines is between 4.5 and 5 metres.
We feel that there are three principal factors requiring specific analysis:
1) The problems emanating from the increased pressure on passenger sets
due to the higher passing speeds
2) The effects of pressure variations on freight trains when passenger trains
3) Possible incidents due, for different reasons, to variations in the load of
freight trains or to exceeding the gauge.
As regards the first point, we consider that we already have adequate tools
available for accurately modelling the problem and arriving at results and
As for the effect on freight trains of the pressure variations occurring when
passing a passenger train, particularly when the freight train is hauling lightly
loaded wagons, we should look to the experience of DBAG in this area.
Finally, in the case of incidents possibly resulting from the fouling of the gauge
by goods trains (due to shifting of the load, possible derailments, etc.) it needs
to be accepted that this is a particularly difficult problem to tackle on account of
the large number of cases which it would be necessary to consider.
In any case, work is currently being carried out in France with a view to
analysing the feasibility of authorising running speeds for high-speed trains and
freight trains higher than those stipulated in current regulations.
There is no doubt that existing traffic control systems and those planned for
introduction in the short term will make it possible to regulate operating in such
a way as to ensure that the passing speed limits are not exceeded:
a) By reducing the speed of the passenger trains when approaching an
encounter with a goods train
b) By arranging for the freight trains to make use of existing parking tracks
all along the line.
In the former case it would be necessary to analyse, for each line and route, the
effect on the journey time of the passenger trains, and the impact of this on the
ability to attract customers. In the latter case an examination should be made of
the repercussions on reliability and on the transport time of freight services.
For safety reasons it seems preferable that the two types of train should not
pass each other in the tunnels on the line.
As regards the effect of freight trains on the maintenance frequencies, it should
be pointed out that this is a problem which can easily be analysed since both
the technical and economic consequences can be quantified.
The environmental aspects are a constant source of concern, and must be
taken into account when planning the layout of the line. We are not of course
referring to the well-known aspects of building a new line, but to considering the
operating system planned for the line from the outset.
In certain cases, in the areas affected by the new railway line, it may be thought
that there will be only traffic during the daytime without any night runs. However,
this might be, the technical, and in some cases also the commercial, constraints
may make it essential to operate long-distance passenger trains, or goods
trains, at night. In this area the work will be geared towards reducing the effects
of the goods trains. On the whole the problem is similar to the one encountered
in airports, where take-offs and landings have been discontinued during certain
night periods (24 h to 6 h).
Finally, as regards signalling, and for reasons connected with the need for
flexibility in operating lines carrying different types of traffic, it will be readily
understood that there is some attraction – in spite of the additional costs
involved – of equipping freight locomotives with the systems which are currently
being put in place for inter-operational purposes (ERTMS).
5. METHODOLOGICAL PROPOSAL FOR TAKING DECISIONS WITH
RESPECT TO THE OPERATING OF A HIGH-SPEED LINE CARRYING
The construction of new railway lines does not normally involve the removal of
the existing lines along the routes where the new infrastructure has been laid.
On a given route, therefore, there will be the formerly existing lines together with
new ones offering better geometrical quality.
Operating a route on which new tracks have been included will require – if
maximum use is to be made of the existing installations – the utilisation of a
conceptual model of the traffic distribution between the two lines, so as to
optimise the commercial response to the demand for passenger and freight
transport on this route.
The allocation of traffic to one particular route or another should be the result of
a comparison between the requirements and the possibilities offered or not
offered by the existing installations to meet these requirements, including any
possible improvements which might be made. This will thus ensure the full
utilisation of the equipment and of the available railway installations.
As a preliminary to this it is pointed out that the reality of railway operating
shows that the problems of regional passenger services on some routes may be
ideally resolved by the new routes. In any case the provisions of such services
is not directly connected with the construction of a new line, on account of the
peculiar dimension of the traffic concerned.
What is more, suburban services fall into the framework of a different policy
from that embodied by long-distance services, and thus require a different
These two considerations prompt the thought that the commercial objectives to
be achieved by the railway services provided on a route, when the construction
of a new line is being considered, are focused on the precise definition of the
services to be provided in the way of medium-distance and long distance
passenger traffic, and of freight traffic.
It should be stressed, however, that in the planning of the new lines being
constructed in Europe the commercial objectives relating only to passenger
traffic are detailed, whereas those which might concern freight traffic are
contained within a general category of “improvement in transport time”, but
without any precise quantification being made, at least explicitly.
Regardless of their own technical criteria which recommend a rational
distribution of traffic between the newly constructed lines and the existing lines,
there is no doubt that the economic aspects may and must also play an
important role in such an allocation of traffic.
It is known, in fact – and has been pointed out previously – that the adoption of
certain geometrical values in the design of new route layouts may result in a
significant reduction in investment costs; but just as well known are the
envisaged restrictions which that might involve for the operating system of the
The possible reduction in track construction costs as a result of adopting certain
design criteria, as we have just mentioned, is not a general rule which is always
verified at significant economic levels, regardless of factors such as: the
topography of the terrain through which the line runs, or the relative values
considered for the curve radius or the gradients on the route, but constitutes a
field of action to be taken into consideration on account of its possible practical
It is becoming more and more necessary to pay preferential and increasing
attention to minimising the construction and operating costs of the new railway
lines being provided, on account of:
a) the extent of the economic resources required for providing a new railway
link and developing it;
b) the existing budget limitations which make it necessary to seek sources of
funding other than state contributions where the profitability of such projects
will be a decisive factor in making use of such sources.
Increasing this level may contribute towards jointly optimising the design of the
line and the planned operating of it. This is an operating system which is in turn
conditioned - as we have already pointed out - by the commercial objectives to
be achieved and by the services which, in relation to such objectives, are
provided by the existing installations.
Below we outline the principal aspects of a new general methodology for
analysing the allocation of traffic on the new high-speed lines.
For this purpose we shall consider three phases:
This phase commences with an explanation of the reasons behind the analysis
of the advantage of laying a new railway track. There are normally two
considerations prompting the analysis:
1) Problems associated with a lack of capacity of the line in commercial
2) The need to attain higher quality levels in the services provided for
These two considerations often arise simultaneously (as in the case of the
Paris-Lyons line where the section between St. Florentin and Dijon was already
carrying nearly 250 movements a day in the seventies).
What is more, on this route, and in spite of the journey time of only 3 h 45 m for
the 515 km separating the two cities, i.e. a commercial speed of 128 km/h, the
railway share of the total passenger market had fallen from 65% in 1963 to 48%
Faced with the objective need to analyse the feasibility of a new line the first
criteria were formulated for defining the geometrical layout plan of the new line.
Taking the passenger segment as the reference, the commercial objectives
need to be defined; in this first phase of the work these objectives relate
principally to: the journey time, the frequency of service and the typology of the
middle-distance and long-distance services.
Consideration of the journey time gives the commercial speed, and this then
gives the maximum speed to be adopted, which will thus provide an indication
of the minimum plain line curve radii and of the maximum gradients.
Furthermore, on the basis of some of the parameters indicated above and
which will be representative of the quality of the railway service, it will be
possible to forecast the traffic demand which, taken in conjunction with an
operating programme, will enable a first diagram of services to be defined and
consequently also the available residual capacity, both on an hourly and on a
By way of summary, the implementation of this first phase will provide reference
data in connection with:
a) the investment required for the construction of the new line,
b) the expected revenue and estimated operating costs,
a) the possible availability of a certain residual capacity on the new line.
If, as regards this last indicator, the capacity offered by the middle-distance and
long-distance passenger services (bearing in mind the timetable slots required
for carrying out maintenance work on the installations) were top be very limited,
it would not be possible to provide for the eventuality of operating freight trains.
If, on the other hand, there is sufficient capacity on the line then it will be
possible to proceed with the second phase as outlined below.
The potential use of the new line for freight traffic requires an initial formulation
of the commercial objectives to be achieved in this segment of the market,
essentially in respect of:
- the journey times on the line
- the timetables
- the number of services.
Once these parameters have been set, in the light of the services offered by
the other modes of transport, particularly road transport, it is then necessary to
examine the possibilities of achieving such objectives on the existing line, on
the basis of, among other things:
1) the liberalisation of certain timetable paths in so far as the essential part of
the middle-distance and long-distance passenger traffic would transfer to
the new line;
2) the incorporation of wagons and/or locomotives on the best services;
3) a reduction in the time allowed for the terminal operations.
As regards this last aspect, Figure 29 shows, in the form of a graph, that in the
transport chain, the speed during the journey is a factor which, although
significant, may become secondary as a result of the importance which may be
attached to the logistical aspects of the collection and distribution of goods.
As for the use of new railway freight stock, we have already pointed out in this
report the possibility of having wagons suitable for running at a maximum speed
of 140/160 km/h, which, taken in conjunction with the use of certain locomotives
in Table 18 would make it possible to provide better commercial services.
If the adoption of a few measures of the type indicated or others of a different
kind were to make it possible to achieve the commercial objectives of the
railway offer in the light of the market conditions occurring in each of the traffic
corridors, the analysis would be complete.
If not, it would be necessary to estimate what freight might be won by the new
line, the expected revenue and the associated costs. If the economic balance
sheet were favourable, it would then be possible to go on to the third phase.
Consideration of the potential freight traffic on the new line will determine the
a) The extra investment required as a result of the possible need to modify
certain geometrical parameters defined in phase 1, as well as the passing
or overtaking section on the assumption that the new line is operated with
specially designed passenger trains and, where applicable, with high quality
b) The modification, and possible increase in the cost of maintenance of the
installations as a result of the use of freight wagons.
c) The restrictions which might be placed on the operating of freight trains on
the new line as a result of the introduction of new passenger services, e.g.
Joint consideration of the resulting economic variables in each phase in terms
of the necessary investments, the economic balance sheet presented to the
operator of each service, and finally the commercial impact of each service,
would make it possible to support taking decisions in the interests of running
freight trains on high-speed lines. The flow chart presented in Figure 30
provides a synthesis of the briefly outlined methodology.
The ultimate object of this report was:
1) to analyse currently available experience in the area of operating high-
speed lines carrying mixed traffic,
2) to outline the current trends as regards the operating system to be adopted
for high-speed lines at the design or construction stage.
3) to analyse the principal aspects to be considered before any decision is
taken to operate a high-speed line with mixed traffic,
4) to propose an analysis methodology with a view to determining for a given
route the advantage and feasibility of adopting an operating system devised
for high-speed passenger services and for conventional freight trains.
The report began by demonstrating the objective need to accurately define the
concept of mixed traffic on high-speed lines. This is due to the fact that the
currently used terms of ‘passengers’ and ‘goods’ or ‘freight’ may not be
sufficient in themselves to determine the technical and economic viability of
operating a line with mixed traffic.
Two examples may be used to illustrate this:
a) The first would correspond to a line carrying high-speed passenger trains
with the same trains being adapted for the transport of freight. In this case,
on account of the behaviour of the track and the capacity of the line, only a
single type of rolling stock would use it.
b) The second would be the complete opposite (as in the case of the new
Madrid-Seville line), i.e. it would be a line carrying only passengers, but in
both high-speed sets and in conventional locomotive-hauled trains. Since
the operating system would then be designed for trains with different
maximum speeds and design characteristics, problems might well arise in
respect of line capacity and/or damage to the track.
Within this context what is proposed is a classification of lines (Table 2) based
on the idea that the definition should, in the case of mixed traffic, be governed
essentially be the typology of the rolling stock running on the track, and by the
aggressiveness of its effect on the track, something which will impact on the
cost of maintaining track geometry.
In the light of the above considerations the document analyses (in chapter II)
current experience in the design and operation of the following high-speed lines
or sections of lines which are open to traffic:
• Madrid-Seville (AVE sets + conventional trains consisting of passenger
coaches and a locomotive)
• Tours link line (section of the TGV Atlantique line which is designed to be
suitable for both passenger and freight traffic)
• Rome-Florence Direttissima
The following conclusions were drawn from the analysis:
1) The German high-speed lines carry freight trains most of which operate at
speeds of 120 km/h, although some – though very few – run at 160 km/h.
2) The Madrid-Seville line was planned for both passenger and freight trains
but the eventual adoption of the international standard track gauge made
the operation of freight trains inadvisable. However, specially designed
high-speed passenger trains as well as conventional locomotive-hauled
passenger trains are both operated on this line.
3) The approximately 17 km section constituting the Tours link cutting across
the TGV Atlantique line, was designed to carry freight trains, thus avoiding
certain problems connected with the lack of capacity of the conventional
line. It uses the TVM signalling system as well as trackside signalling. There
have been very few freight trains using the section in the ten years since it
was opened for commercial service.
4) The Direttissima between Rome and Florence is used to carry high-speed
stock at 250 km/h, conventional passenger trains at 160/200 km/h and
freight trains at 100/120 km/h.
Chapter III sets out the trends in the design and operating criteria for new high-
speed lines in Germany, Spain, France and Italy. The position may be
summarised as follows:
a) The new railway tracks under construction or near completion, such as the
Cologne-Frankfurt, Madrid-Barcelona and TGV Est lines, will have an
operating system dedicated exclusively to passenger traffic. On the Spanish
line, however, the high-speed sets will co-exist with conventional
locomotive-hauled passenger trains.
b) The other planned tracks or ones which are currently being developed
(Karlsruhe-Basle – Ebensfeld-Erfurt-Leipzig – Nuremberg-Ingoldstadt –
Cologne-Aachen, in Germany) will be operated with mixed traffic. A similar
operating system is envisaged for the following new lines: Barcelona-
Perpignan – Nimes loop line and Montpellier – south section of the TGV
Rhine-Rhone, and Florence-Bologna-Milan-Turin in Italy).
Chapter IV is devoted to an analysis of the principal aspects to be taken into
account before any decision is taken to operate a high-speed line with mixed
traffic. These aspects are essentially four in number:
• Choice of geometrical characteristics of the layout of the new line
• Effects of mixed traffic on the deterioration of the geometry of the track
• Repercussions of freight traffic on line capacity
• Determination of certain operating criteria (crossing speed of passenger
trains and freight trains, effect of freight trains on maintenance intervals,
environmental impact of freight trains, etc.).
As regards each of the above aspects, the principal considerations would be as
• Before selecting the geometrical parameters and in particular the maximum
gradient and length of gradient, where operating freight trains is envisaged,
proceed with an analysis of the actual impact of these parameters on the
following variables at least: investment into infrastructure, operating costs
and hauled load of freight trains.
• The operating of goods trains on a high-speed line might increase the track
maintenance costs by up to 20 or 25% as compared with a situation where
the line is used exclusively for passenger traffic. Since these costs represent
about 40% of the total cost of fixed installations (tunnels, bridges, formation,
signalling, communications, overhead equipment, power supply) the extra
cost attributable to operating mixed traffic would be of the order of 10%.
• Operating freight trains on a high-speed line can be envisaged only where
the passenger demand leaves significant empty slots in line capacity.
Passenger traffic on the European high-speed lines fluctuates between 5
and 25 million passengers a year.
• For the time being, experience in the area of the passenger trains and goods
trains passing each other is limited to the maximum speed of 220 and 160
km/h respectively. In the case of a high-speed line the traffic control systems
make it possible to regulate running and so ensure that the current passing
speeds are not exceeded, by:
- reducing the speed of passenger trains as they approach an encounter
with a freight train;
- using the existing parking tracks positioned along the line for the goods
In both cases it would be necessary to analyse, for each specific line, the
possible effect of such measures on the journey time of the passenger trains,
and their effect on attracting traffic. And also their repercussions on the
reliability and the transport times of freight services.
The report concludes with a methodological proposal in regard to taking
decisions as to whether to operate a particular high-speed line with mixed
traffic. The proposed approach embodies the idea that operating a new high-
speed line with conventional goods trains would be feasible if two basic
conditions were to be satisfied: some remaining line capacity once the
passenger service diagram has been filled, and the impossibility of achieving for
goods trains a similar level of quality on the existing lines, which would make it
possible to use the new railway infrastructure. This is the case with the new line
between Barcelona and Perpignan.