J. Mod. Transport.

(2016) 24(1):1–21
DOI 10.1007/s40534-015-0094-y

A multidimensional examination of performances of HSR

(High-Speed Rail) systems
Milan Janić1

Received: 13 October 2015 / Revised: 19 December 2015 / Accepted: 22 December 2015 / Published online: 5 February 2016
 The Author(s) 2016. This article is published with open access at Springerlink.com

Abstract This paper deals with a multidimensional 1 Introduction

examination of the infrastructural, technical/technological,
operational, economic, social, and environmental perfor- The high-speed rail (HSR) systems as the rather innovative
mances of high-speed rail (HSR) systems, including their systems within the railway transport mode, particularly as
overview, analysis of some real-life cases, and limited compared to its conventional (rail) passenger counterpart,
(analytical) modeling. The infrastructural performances have been developing worldwide (Europe, Far East-Asia,
reflect design and geometrical characteristics of the HSR and United States of America (USA)). Despite the common
lines and stations. The technical/technological perfor- name, different definitions of these systems have been used
mances relate to the characteristics of rolling stock, i.e., as follows:
high-speed trains, and supportive facilities and equipment,
• Japan The HSR system called ‘Shinkansen’ (i.e., ‘new
i.e., the power supply, signaling, and traffic control and
trunk line’) is defined as the main line along almost its
management system(s). The operational performances
entire length (i.e., route) where trains can run at the
include the capacity and productivity of HSR lines and
speed of at least and above 200 km/h. The ‘Shinkansen’
rolling stock, and quality of services. The economic per-
system’s network has been built with the specific
formances refer to the HSR systems’ costs, revenues, and
technical standards (i.e., dedicated tracks without the
their relationship. The social performances relate to the
level crossings and the standardized and special loading
impacts of HSR systems on the society such as congestion,
gauge). This HSR system represents a part of the
noise, and safety, and their externalities, and the effects in
overall Japanese Shinkansen transportation system [1].
terms of contribution to the local and global/country social-
• Europe The definition of HSR system includes (a) in-
economic development. Finally, the environmental per-
frastructure, (b) rolling stock, and (c) compatibility of
formances of the HSR systems reflect their energy con-
the infrastructure and rolling stock [2].
sumption and related emissions of green house gases, land
use, and corresponding externalities. – Infrastructure Infrastructure of the trans-European
HSR system is considered a part of the Trans-
Keywords HSR (high-speed rail) systems  Performances  European rail transport system/network. It is spe-
Multidimensional examination  Overview  Analysis  cially built and/or upgraded for the high-speed (HS)
Modeling travel. This may include connecting lines and
junctions of the new lines upgraded for the HS,
and the stations located on them, where the train
& Milan Janić speeds must take into account the local conditions.
M.Janic@tudelft.nl The HSR lines include those specially built for the
1
speeds equal to or greater than 250 km/h (Category
Faculty of Civil Engineering and Geosciences, Department of
I), those specially upgraded for the speeds of the
Transport & Planning, Delft University of Technology,
Stevinweg 1, 2628 CN Delft, P.O. Box 5048, 2600, GA, order of 200 km/h (Category II), and those
Delft, The Netherlands upgraded with the particular features resulting from

123
2 M. Janic

the topographical relief or the town-planning con- Sect. 8 environmental performances. The last section
straints (Category III). Therefore, the Category I summarizes some conclusions.
lines are exclusively considered as the real HSR
lines.
– Rolling stock The HS trains are designed to 2 A concept of performances of HSR systems
guarantee safe and uninterrupted travel at the speed
of at least 250 and 300 km/h under the appropriate The performances of transport systems can be defined as
circumstances on the Category I lines, about their ability to fulfill the needs and expectations of par-
200 km/h on the specially upgraded Category II ticular actors/stakeholders involved, which are usually
lines, and at the highest possible speed on the other users/passengers, rail operators, and the third parties. For
Category III lines. the HSR systems, these performances can generally be
– Compatibility of the infrastructure and rolling stock classified as infrastructural, technical/technological, oper-
The HS trains are designed to be fully compatible ational, economic, social, and environmental [6].
with the characteristics of infrastructure, and vice
• Infrastructural and technical/technological perfor-
versa, which influences the performances in terms
mances imply the system’s physical, constructive,
of safety, quality, and cost of services.
technological, and technical characteristics of infras-
• China According to Order No. 34, 2013 from China’s tructure, vehicles, i.e., HS trains, and supporting
Ministry of Railways, the HSR system refers to the facilities and equipment, i.e., the power supply, signal-
newly built passenger-dedicated lines with (actual or ing, and traffic control/management system(s);
reserved) speed equal and/or greater than 250 km/h. Its • Operational performances reflect the system’s capabil-
specific acronym is China railway high-speed (CRH). ities to serve the specified volumes of user/passenger
In addition, a number of new 200 km/h express demand under given conditions;
passenger and 200 km/h mixed (passenger and freight) • Economic performances express the costs and rev-
lines have been building as the components of the enues, the latter based on the charges (prices) to
country’s entire HSR network [3]. users/passengers, and their relationship(s); and
• USA (United States of America) The HSR system is • Social and environmental performances reflect the
defined as that providing the frequent express services scale of the system’s effects and impacts on the society
between the major population centers on the distances and environment, the later usually expressed in the
from 200 to 600 mile (mi) with a few intermediate monetary terms as the external costs, i.e., externalities,
stops, at the speeds of at least 150 mph (mi/h) on the if internalized by the related policies.
completely grade-separated, dedicated rights-of way
The above-mentioned performances of the HSR systems
lines (1 mi = 1.609 km). It is also considered as the
are frequently considered individually although being
system providing regional, relatively frequent services
inherently dependent and influential on each other as
operated at the speeds from 110 to 150 mph between
shown in Fig. 1.
the major and moderate population centers on the
As can be seen, according to the ‘top-down’ approach,
distances between 100 and 500 miles with some
the infrastructural performances directly influence the
intermediate stops, grade separated with some dedi-
technical/technological performances, thus causing their
cated and shared tracks using the positive train control
mutual influence as well as the influence between them and
technology [4, 5]. In both cases, the HSR system has
all other performances. According to the ‘bottom-up’
been expected to relieve congestion at the highways
approach, the social/policy performances can directly
and airports, in the latter case particularly by competing
influence the infrastructural and technical/technological
with the short- to medium-haul airline flights.
performances, thus creating the mutual influence of these
This paper deals with a multidimensional examination and all other performances.
of the infrastructural, technical/technological, operational,
economic, social, and environmental performances of the
above-mentioned HSR systems by providing their over- 3 Infrastructural performances of HSR systems
view, analysis, and limited (analytical) modeling. In addi-
tion to the introductory section, this paper consists of eight The infrastructure of HSR systems consists of lines with
other sections. Section 2 introduces the concept of per- the rail tracks connecting the stations/stops along them and
formances of the HSR system(s). Section 3 analyzes the the end stations/terminuses, both considered exclusively as
systems’ infrastructural, Sect. 4 technical/technological, the above-mentioned Category I of the HSR lines. The
Sect. 5 operational, Sect. 6 economic, Sect. 7 social, and lines and stations constitute the HSR network spreading

123 J. Mod. Transport. (2016) 24(1):1–21

A multidimensional examination of performances of HSR (High-Speed Rail) systems 3

Technical/
lines with long segments of tunnels and/or bridges such as
technological those in Japan.1 In addition, they have been expected to
Infrastructural
increase the capacity of HSR lines, operating speed, reduce
the maintenance costs through reducing the frequency of
maintenance operations, and consequently increase the
Operational
level of safety.

3.2 Stations
Economic
The HSR stations mainly characterized by location and
design enable facilitation of the HSR system with its
Environmental users/passengers. The main aspect of location as the nodes
Social
of corresponding HSR network is their number along
particular lines. Then, it is their micro-location in urban
Bottom-up Top-down areas/cities and often at airports, which should enable safe,
Fig. 1 A potential relationship of the performances of HSR systems efficient, and effective accessibility by individual (car) and
[6] mass urban public transit systems (bus, tram, light rail,
metro, and regional rail).2 Furthermore, it is their func-
tional design, which includes (i) the track and platform
over a given region, country, and/or a continent. Table 1
technical aspects (number, arrangement, dimension, safety,
provides an illustration of the progress so far in developing
and electrical, signaling, and communication systems); (ii)
the HSR networks at particular continents.
the user/passenger service and comfort aspects (accessi-
As can be seen, the longest HSR network currently
bility, inter-modal transfer, security, ticketing and travel
operating and being under construction is in Asia, mainly
information, station facilities, etc.); and (iii) the environ-
thanks to the fast developments in China, followed by that
mental aspects (choice of building/construction materials
in Europe. The last are those in both Americas and Africa.
and protection of the local environment from noise) [11,
12].
3.1 Lines
In particular, an additional important aspect of design of
the HSR stations is the arrangement of tracks and platforms
The lines as links connecting particular stations as the
for users/passengers. In general, two main concepts have
nodes of HSR network are mainly characterized by their
been used: the side platforms facing the track(s) by one
three-dimensional layout and geometry of tracks. The most
side and the island platforms facing the tracks by both
relevant parameters of geometry of tracks are the distance
sides, as shown in Fig. 2a, b, respectively [11].
between their centers, gauge, the maximum axle load,
The safety aspect of design of the HSR stations is
gradient, the minimum horizontal and vertical radius of
important for users–passengers and accompanies standing
curvature, the maximum cant and the maximum cant gra-
on the platforms in cases when the non-stopping trains are
dient, and the length of transition curves corresponding to
passing by at relatively high speeds. These people could be
the minimum curve radius. For example, in Europe, except
affected (sucked toward a passing train if standing too
track gauge (1,435 mm), all other parameters are depen-
close to the platform edge) by air streams generated by the
dent on the maximum design speed. In addition, the HSR
HS trains. For example, some research indicated that the
tracks can be broadly ballasted and ballast less [7, 8]. The
people standing on the platforms at the distance of 2 m
former are present at the most already built HSR lines,
from the HS train passing by at the speed of 240 km/h
while the latter have been considered particularly for the
could be under a real risk [13].

Table 1 Development of the HSR networks at particular continents

[14, 50]
Status Continent World 1
In the year 1972, the ballastless ‘slab track’ had been developed and
a
Europe Asia Others applied to the Sanyo Shinkansen line; in the year 2007, the ‘slab
track’ was used for 1244-km-long line, which shared about 57 % of
In operation (km) 7,351 15,241 362 22,954 the Shinkansen network [9]. In China, both ballast and ballastless slab
Under construction (km) 2,929 9,625 200 12,754 tracks have been used [10].
2
Total (km) 10,280 24,866 562 35,708 For example, the new CRH South Guangzhou station on the
Hangzhou–Shenzhen line (China) has 15 platforms with 28 tracks and
a
Latin America, USA, Africa is the largest in Asia at the moment [10].

J. Mod. Transport. (2016) 24(1):1–21 123

4 M. Janic

Platform lengths are much shorter and vary, for example, from
Stop track
280 km between Berlin and Hamburg (Germany) to
770 km between Paris and Marseille (France) [14]. How-
Barriers Passing tracks ever, the experience so far has shown that the average
travel distances on some of these long Chinese lines have
Stop track been about 560–620 km, which appears comparable to
Platform some of their (long) European counterparts [15].
(a)
Track 20

4 Technical/technological performances of HSR

Platform Tracks
systems
Track 21

Track 22 The technical/technological performances of HSR systems

relate to their rolling stock, i.e., high-speed trains (HSTs)
Platform
Tracks and supportive facilities and equipment, i.e., power supply,
Track 23 signaling, and traffic control/management system(s).
(b)
Fig. 2 Simplified schemes of arrangements of the platforms and
4.1 Rolling stock
tracks at the HSR station. a Line station with the side platforms and
two passing and two stopping tracks [11]. b Begin/end station/ The HSR rolling stock, i.e., trains, are characterized by an
terminus with the inland platforms (Tokyo Shinkansen) [24] optimized aerodynamic shape; fixed composition and bi-
directional set; self-propelling, concentrated, or distributed
power; interior signaling system(s); several braking sys-
3.3 Network tems; power electronic equipment; control circuits; com-
puter network; automatic diagnostic system; particularly
The above-mentioned lines and stations constitute the HSR high level of reliability, availability, maintainability, and
infrastructure network, which spreads over a given country safety (RAMS); maintenance by inspection in fixed time
(http://www.johomaps.com/eu/europehighspeed). Table 2 intervals and preventively; and compatibility with infras-
shows some characteristics of the main grid (eight national tructure (track and loading gauge, platforms, catenary, etc.)
backbone lines) of the HSR network in China. [1]. Table 3 provides the selected technical/technological
The specificity of this (Chinese) compared to the other specifications for different HS trains. As can be seen, the
HSR rail networks worldwide, particularly those in Europe, maximum design speed varies from 250 to 350 km/h. The
is the length of lines between the end stations/terminuses, locomotives are powered by the electric energy. These are
which varies from 1,000 to 2,400 km. In Europe, these the so-called multi-system locomotives interoperable for at

Table 2 Some characteristics of the main grid of China rail high-speed (CRH) network [10, 15] (https://en.wikipedia.org/wiki/High-speed_rail_
in_China/)
Relation Orientation Length of line (km) Design speed (km/h)

Beijing–Harbin N–S 1,800 350

Beijing–Shanghai N–S 1,318 350
Beijing–Hong Kong N–S 2,383 350
Hangzhou–Shenzhen N–S 1,499 250/350
Sub-length 7,000
Qingdao–Taiyuan E–W 940 200/250
Xuzhou–Lanzhou E–W 1,434 250/350
Chengdu–Shanghai E–W 2,066 200/250
Kunming–Shanghai E–W 2,056 350
Sub-length 6,496
Total length 13,469
a
North–South (N–S); East–West (E–W)

123 J. Mod. Transport. (2016) 24(1):1–21

A multidimensional examination of performances of HSR (High-Speed Rail) systems 5

Table 3 Technical/technological characteristics of different HS trains [51] (http://en.wikipedia.org/wiki/ICE_3; http://en.wikipedia.org/wiki/

Siemens_Velaro; http://www.trainweb.org/tgvpages/tgvindex.html; http://en.wikipedia.org/wiki/New_Pendolino)
Type of HS train Build date Max. speeda Supply voltageb Traction Length/weight Configurationc Performance
(year) (km/h) (kW, kV) (m, ton) (–, seats) metrics
(kW/seat)

TGV PSE 1978–1985 300/270 25 kV 50 Hz AC 6,450, 25 200, 385 1 ? 8 ? 1, 385 18.34

(Paris Sud-East) 1.5 kV DC 3,100, 1.5
TGV LA Poste 1981–1984 270 25 kV 50 Hz AC 6,450, 25 200, 345 – –
1.5 kV D 3,100, 1.5
TGV Atlantique 1989–1992 300 25 kV 50 Hz AC 8,800, 25 238, 484 1 ? 10 ? 1, 485 18.14
1.5 kV DC
TGV Resau 1992–1996 300 25 kV 50 Hz AC 8,800, 25 200, 386 1 ? 8 ? 1, 377 23.34
1.5 kV DC
Eurostar 1993–1995 300 25 kV 50 Hz AC 12,200, 25 394, 752 1 ? 18 ? 1, 794 15.90
3 kV DC
1.5 kV DC, 750 V DC
TGV Duplex 1995–1997 320 25 kV 50 Hz AC 8,800, 25 200, 380 1 ? 8 ? 1, 545 16.15
1.5 kV DC
Thalys 1995–1998 300 25 kV 50 Hz AC 8,800, 25 200, 385 1 ? 8 ? 1, 377 23.14
1.5 kV DC
ICE 3 1998–1999 330 15 kV 16.7 Hz AC 8,000, 1.5 201, 435 1 ? 8 ? 1, 441 18.14
1.5 kV DC
ICE 3 M 2000 330 25 kV 50 Hz AC 3 kV DC 8,000, 25 201, 435 1 ? 8 ? 1, 430 18.60
ICE Velaro CNd 2004 350 25 kV 50 Hz 9,200, 25 200, 447 1 ? 8 ? 1, 610 15.31
AVE 1991–1992 300 25 kV 50 Hz AC 8,800, 25 200, 392 1 ? 8 ? 1, 320 26.75
3 kV DC
KTX 1997–2002 300 25 kV 50 Hz AC 13,200, 25 381, 701 1 ? 18 ? 1, 935 14.12
(TGV Korea)
N700-I (Japan) 2007 330 25 kV 60 Hz 9,760, 25 204.7, 365 8, 636 15.35
ETR 600d 2008 250 25 kV 50 Hz AC 5,500, 25 187, 387 2 ? 3 ? 2, 430 12.79
3 kV DC
a
Design speed
b
AC: alternating current, DC: direct current
c
Power car(s)–trailers–power car(s)
d
Operating in China—version CRH2C

least two different electric power supply systems. The the installed (and required) traction. This indicates that the
traction power varies from 5,500 to 13,200 kW/train set. HS trains with higher seating capacity do not need to have
The length of a train set is predominantly about 200 m, and the proportionally stronger traction.
the corresponding weight is between 350 and 450 tons. As well, Fig. 4 shows the relationship between the max-
Typical configuration of an HS train set is 1 power car ? 8 imum designed and operating speed of the HS trains [16].
trailers ? 1 power car. The performance metrics vary As can be seen, the speeds ranging from 200 to
across the considered set of HS trains from 12 to 23 kW/ 320 km/h coincide with each other for many HS trains.
seat. Nevertheless, generally, with the increase of the maximum
In addition, Fig. 3 shows the relationship between the design speed, the positive difference between this and the
performance metrics and the seat capacity of the selected maximum operating speed tends to increase. This particu-
HS trains. larly happens for the speeds ranging between 270 and
As can be seen, the performance metrics expressed by 380 km/h. Consequently, at particularly high maximum
the installed traction per seat (kW/seat) decreases more design speeds (above 300 km/h), it is likely to expect the
than proportionally with the increase of the number of lower maximum operating speeds for about 10 %–20 %, as
seats, thus indicating economies of the train size in terms of shown in this case.

J. Mod. Transport. (2016) 24(1):1–21 123

6 M. Janic

28 system components can further be divided into two main

components: the HSR electrical infrastructure and the HS
Performacne metrics (kW/seat)
26
24 rolling stock traction equipment [17].
22 PM (s)=395.095 s -0.495
20 R 2=0.710 4.2.2 Signaling systems
18
16 The different HSR signaling systems have been applied in
14 different countries. For example, each European coun-
12 try has its own HSR signaling systems: in France it is
10 Transmission Vole Machine (TVM), in Germany
300 400 500 600 700 800 900
LinienZugBeeinflussung (LZB), in Spain German’s LZB
Seat capacity (train)
(for speeds up to 300 km/h) and Electrique Bureau CABine
Fig. 3 Relationship between the performance metrics and the seat (EBICAB) (for speeds up to 220 km/h), and in Italy Blocco
capacity of HS trains (Table 3) Automatico a Correnti Codificate (BACC) (for speeds up to
250 km/h). In addition, the European rail traffic manage-
ment system (ERTMS—Level 1 and/or 2) has been intro-
duced on the particular lines in different countries as an
400
alternative and/or a complement to the existing national
Maximum operating speed (km/h)

systems [17].
350 V 0=V d
R 2=1
The type of signaling system influences the length of a
block of the track, which can be occupied exclusively by a
300
single HS train. The number of such successive empty
250
blocks determining the (breaking) distance between any
V 0=-0.0036V 2d+2.709V d-205.16 pair of HS trains moving in the same direction depends of
R 2=0.789
200 their maximum operating cruising speed and the breaking/
deceleration rate(s).
200 250 300 350 400
Maximum design speed (km/h)
4.2.3 Traffic control/management system
Fig. 4 Relationship between the maximum design and maximum
operating speed of the selected HS trains (http://en.wikipedia.org/ In general, at the HSR rail lines/networks the rail traffic
wiki/List_of_high-speed_trains)
control/management systems is fully computer supported
and can include the following main components: TOC—
train operation controller; PC—power controller; STC—
4.2 Supportive facilities and equipment signal and telecommunication controller; CCC—crew and
car utilization controller; PSC—passenger service con-
The main supportive facilities and equipment of the HSR troller; and TSMC—track and structure maintenance con-
system in the given context are power supply, signaling, troller. These components are usually accommodated in the
and traffic control/management system. same room with the corresponding staff [18].

4.2.1 Power supply system

5 Operational performances of HSR systems
The power supply system is an integrated system including
the high-voltage electric power lines, substations, contact The main operational performances of HSR systems are
line, HS trains, and the remote command and control sys- demand, capacity, and quality of services, the latest as an
tem ensuring efficient, reliable, and safe supply of electric outcome from the dynamic interaction between the former
power to the HSR lines and trains, and consequently two. These performances can be considered for an indi-
operations. The electrified networks for the HSR lines vidual line/route and/or for the entire network serving a
generally use the alternate current (AC) or direct current given region, i.e., country.
(DC). As given in Table 3, the typical voltage and fre-
quencies are 25 kV 50 Hz AC, 1.5 kV DC, and 15 kV 5.1 Demand
16.7 Hz AC. The latest has been installed in Germany and
supplied from the dedicated high-voltage network called The demand for HSR services consists of the self-generated
the ‘Railway Frequency.’ The above-mentioned general demand and the demand attracted from other transport

123 J. Mod. Transport. (2016) 24(1):1–21

A multidimensional examination of performances of HSR (High-Speed Rail) systems 7

300 52
Eruope (HSR)

Searved demand (×109 p-km/year)

250 China (CRH) 50

Demand (×10 9 p-km/year)

Europe (HSR):
D(LN )=0.0127L N +22.837
200
48 D(GDP)=0.002GDP 2-1.827GDP+460.03
R 2=0.982
R 2=0.875
46
150
44
100
China (CRH): 42
D(LN )=0.0087L N +9.19
50
R 2=0.993 40
0 38
3,000 6,000 9,000 12,000 15,000 470 480 490 500 510 520 530 540
Length of the network (km)
GDP (×10 12 yen/year)

Fig. 5 Relationship between the annual passenger demand and the

Fig. 7 Relationship between the satisfied passenger demand by
length of HSR networks in Europe and China (period 1995–2014)
Japanese Tokaido Shinkansen HSR system and the national gross
[3, 50, 52]
domestic product (GDP) (period 2001–2015) [19]

modes on the competitive routes such as individual car, and the other above-mentioned demand-stimulating fac-
conventional railways, and air passenger transport (APT). tors. Figure 7 shows the relationship between GDP and the
In general, the self-generated demand for HSR services satisfied HSR passenger demand in Japan during the
has been stimulated by expansion of the HSR network and observed period [19].
increase of the welfare in terms of the national gross As can be seen, the served passenger demand has
domestic product (GDP). Figure 5 shows the relationship increased more than proportionally with rising of GDP,
between the served passenger demand and the length of thus indicating that GDP has generally been, is, and will
HSR network in Europe and China. continue to be a strong generator of demand in the given
In both regions, the served passenger demand has grown context.
linearly with the increase of the length of HSR networks. In The attracted and satisfied HSR passenger demand from
terms of absolute values, the served passenger demand in other transport modes on the competitive routes has
China has exceeded that in Europe during the relatively resulted from their competition. Figure 8 shows the pas-
short period of time (7 years), which has indicated the very senger market share of HSR compared to that of APT
strong user/passenger preference to the new CRH speed dependence on the line travel time.
system as shown in Fig. 6. As can be seen, the relative market share of HSR (that of
In Europe, the served passenger demand has continu- APT is complement to 100 %) has decreased linearly
ously been growing during the specified period of time. In (Europe, Japan) and more than linearly (China) with the
China, since the start of implementing the CRH speed increase of the line/route travel time within the given
network, the corresponding passenger demand has been range.
growing tremendously and very quickly exceeded that in
Europe. In both cases, this has been possible primarily 5.2 Capacity
thanks to expanding the HSR network as shown in Fig. 6
The capacity of HSR systems can generally be calculated
200
for their components of infrastructure—stations, lines/
Estimate routes, and rolling stock. In general, for the infrastructure
150 Europe (HSR)
Demand (×10 9p-km/year)

China (5 routes: Cities/airports-Beijing)

100
China (CRH) MSHSR(τ ) = -5.9116τ 2+21.235τ +81.629
R2 =0.861
80
100
Market share of HSR (%)

60

50
40
Europe (9 routes); Japan (1 route)
MSHSR(τ ) = -20.213τ +122.55
20 R2 =0.817
0
1,990 1,995 2,000 2,005 2,010 2,015
0
Time (years) 1 2 3 4 5 6
Line/route travel time (h)
Fig. 6 Development of the satisfied passenger demand in the
European HSR and Chinese CRH speed network (period Fig. 8 Relationship between the market shares of HSR and APT, and
1990–2014) [50, 52] the line/route travel time [40, 53–55]

J. Mod. Transport. (2016) 24(1):1–21 123

8 M. Janic

components, the ‘ultimate’ and ‘practical’ capacity can be The maximum operating speeds of HS trains are usu-
considered. Both are dependent on the operational rules ally about 250–350 km/h. The deceleration3 rate a-
and procedures providing a safe separation of trains while varies, i.e., it generally increases with the decrease of
operating along the lines and at the stations in the same speed during the breaking phase of trip. The buffer
and/or different directions. These rules specify the mini- zone (typically of the length of Sb/j = 100 m) is the
mum time separation between occupying the same section distance added to the braking distance of trailing HS
of the line(s) and/or of the station(s), which mainly influ- train to allow a margin for its safe separation from the
ences their corresponding capacities. In addition, the leading train (i) [22, 23]. The train length is typically
transport work and productivity can be considered as the L = 200 or 400 m. The latter is the length of Eurostar
measures integrating in some way the capacity of infras- and 2-unit German-designed Velaro train operating in
tructure components and that of operations of the rolling China (Table 3).
stock. • Station along the line/route If the leading train (i) is to
stop and the trailing train (j) is to pass through a station
5.2.1 Infrastructure components along the line, the ‘reference location’ for counting
trains, i.e., calculating the capacity, can be the exit
5.2.1.1 ‘Ultimate’ capacity signal of the station. The ‘ultimate’ capacity of the
station in this case can be estimated as follows [6, 20]:
• Line The ‘ultimate’ capacity of a given HSR line/route
is defined by the maximum number of trains, which can
T
pass safely through the selected ‘reference location’ on ls=l ðTÞ ¼ ; ð1cÞ
tij=s=min
the line where it is counted under given conditions, i.e.,
usually constant demand for service. This capacity can where tij/s/min is the minimum time interval at which the
be estimated as follows [20, 21]: successive trains (i) and (j) pass in the same direction
T through the station (min).
ll ðTÞ ¼ ; ð1aÞ The minimum time (tij/s/min) in Eq. 1c can be estimated
tij=min
as follows: the leading train (i) after being dispatched
from the station should be at least at the minimum
where i, j are the leading and trailing trains in the breaking distance of the trailing train (j) at the moment
sequence of two successive trains (ij) passing through when this arrives at the exit signal of the station, which
the ‘reference location’ for their counting, respectively, in this case will allow it to proceed. In such case, the
which can be any location along the open line/route; time (ti j/min) in Eq. 1b can generally be extended by the
tij/min the minimum time interval at which the dwell time of the train (i) at the station as follows:
successive trains (i) and (j) moving in the same
direction pass through this ‘reference location’ (min); " #1=2
and T is the period of time for calculating the ultimate Sb=j þ Li Vj
tij=s=min ¼ si þ þ ; ð1dÞ
capacity of particular infrastructure component (h). 2aj ðVj Þ a
j ðVj Þ
This minimum time interval (tij/min) in Eq. 1a is mainly
influenced by the HS train’s maximum operating speed, where si is the dwell time of the leading train (i) at the
acceleration and deceleration/braking performances, station (min).
length, the way of its control, and also the spacing The other symbols are analogous to those in the pre-
and design of the stations/terminuses, gradients along the vious equations.
line/route, and type of traffic control (signaling) system. At most HSR systems, the dwell time is typically
In general, this time can be estimated as follows [22]: s = 2–3 min at the stations located along the lines/
routes and s = 5 min for those located at airports, the
Vj Sb=j þ Li
tij=min ¼ þ ; ð1bÞ latter mainly due to enabling users/passengers to handle

aj ðVj Þ Vj

where i, j are the leading and trailing HS trains,

respectively, of the pair of successive trains (in); Vj is 3
For example, it can be a- = 0.30 m/s2 for the speeds between
the maximum operating speed of the trailing train V = 350 and 300 km/h (first 1,000 m of breaking distance),
(j) (km/h); a-(Vj) is the average deceleration rate of the a- = 0.35 m/s2 for the speeds V = 300–230 km/h (second 1,000 m
trailing train (j) at the maximal braking rate (m/s); Sb/j of breaking distance), and a- = 0.6 m/s2 for the speeds
V = 230–0 km/h (the rest of 6,000–7,000 m of breaking distance).
is the ‘‘buffer’’ zone for the trailing train (j) (m); and Li Consequently, the average deceleration rate of a- = 0.5 m/s2 is
is the length of the leading train (i) (m). usually used in these calculations [22].

123 J. Mod. Transport. (2016) 24(1):1–21

A multidimensional examination of performances of HSR (High-Speed Rail) systems 9

their baggage. This time also includes the time for 26

a=0.3 m/s2-line
closing the doors, setting up the conflict-free exit path, 24 a=0.5 m/s2-line
22 a increases with speed-line
and dispatching the leading train (i). Arrival/departure station

Capacity (trains/h)
20
• End terminus/station The ‘ultimate’ capacity in this
18
case can be estimated as follows [6, 20]: 16
14
T 12
ls=arr ðTÞ ¼ ; ð1eÞ
tij=min=arr 10
8
where tij/min/arr is the minimum time interval at which 240 260 280 300 320 340 360
Maximum operating speed (km/h)
the successive trains (i) and (j) arrive at the entry signal
of the given end station/terminus (min) as the ‘reference Fig. 9 Relationships between the ultimate capacity of the HSR line/
location’ for their counting. route, begin/end station/terminus, and the maximum train operating
speed
The time (tij/min/arr) in Eq. 1e can be estimated as fol-
lows:
The other symbols are analogous to those in the
previous equations.
" #1=2
Vj Sb=j þ Li Equation 1h indicates that the minimum time between
tij=min=arr ¼  þ þsij þ sb ; ð1fÞ departures of the successive trains (i) and (j) from the
aj ðVj Þ 2aj ðVj Þ
begin station/terminus should be set up as the maximum
where sij is the time for changing the route of trains of two time periods: the time the leading train (i) needs
(i) and (j) arriving at the end station/terminus of the to reach the minimum breaking distance from the
given line/route (typically 10 s); and sb is the time of trailing train (j) and the time for setting up a safe
blocking the entrance of the end station/terminus by departure path for this trailing train (j).
other trains(s) (typically 25 s). Figure 9 shows examples of the above-mentioned ‘ul-
The other symbols are analogous to those in the pre- timate’ capacities of the HSR line/route and begin/end
vious equations. station/terminus dependent on the train’s maximum oper-
• Begin terminus/station The ‘ultimate’ capacity in this ating speed calculated by Eq. 1.
case can be estimated as follows [6, 20]: As can be seen, the line/route capacity decreases with
the increase of speed if the same average deceleration/
T acceleration rate is applied (a = 0.5 m/s2 for the speeds of
ls=dep ðTÞ ¼ ; ð1gÞ
tij=min=dep V = 250–350 km/h). However, if this rate increases with
the increase of speed (a = 0.5 m/s2 for the speed of
where tij/min/dep is the minimum time interval at which
V = 250 km/h, a = 0.3 m/s2 for the speed of V =
the successive trains (i) and (j) pass the exit signal of
270 km/h, a = 0.4 m/s2 for the speed of V = 300 km/h,
the given station/terminus as the ‘reference location’ for
and a = 0.5 m/s2 for the speeds of V = 320 and 350 km/
their counting (min).
h), the capacity generally tends to increase. In the latest
The time (tij/min/dep) in Eq. (1g) can be estimated as
case, the capacity again decreases due to applying the same
follows:
deceleration/acceleration rate to the increasing speed.
( 1=2 Similar happens with the arrival and departure capacities of
Sb=i þ Li Vi begin/end station/terminus, respectively. In all cases, the
tij=min=dep ¼ max þ ; sj=r
2aþi ðVi Þ
þ
ai ðVi Þ ð1hÞ train length is assumed to be L = 400 m and the buffer
 distance Sb = 100 m [22]. Consequently, the line/route
þsj=gl þ sj=cf þ sj=d ;
capacity can be estimated as the minimum of the above-
where sj/r is the time for setting the exit path for the mentioned four ‘ultimate’ capacities. In practice, the ‘ul-
trailing train (j) in a given departing sequence (ij) timate’ capacity of the HSR lines/routes and stations is
(usually 10 s); sj/gl is the time for setting the green light typically l = 13–15 trains/h.4 In addition, the required
for trailing train (j) in a given departing sequence (ij) number of tracks at the end/begin station/terminuses can be
(usually 25 s); sj/cf is the time of blocking exit of the determined as the product of the above-mentioned
station/terminus for departing trailing train (j) by other
incoming and outgoing trains (usually 60–75 s); and sj/d
is the dispatching time of the trailing train (j) in a given 4
The number of the Shinkansen ‘‘Nozomi’’ services has been
departing sequence (ij) (usually 30 s). scheduled to be 10 dep/h during the peak hours [19].

J. Mod. Transport. (2016) 24(1):1–21 123

10 M. Janic

‘ultimate’ capacities and the HS train’s dwell time at the and s is the seat capacity of a train operating on a given line
stations—stop time at the line and the turnaround time5 at (seats/train).
the begin/end station(s). The other symbols are analogous to those in the previous
equations.
5.2.1.2 ‘Practical’ capacity The ‘practical’ capacity of a The train’s turnaround time (s) increases with the
given HSR line/route is defined as the maximum number of increase of the operating time along the line/route (the ratio
HS trains, which can be accommodated during the speci- between the length of line/route and the operating speed),
fied period of time under conditions when each of them is the number and duration of intermediate stops, all in both
imposed an average delay [25]. However, in this case, the directions, including those at the beginning and end station/
mutual interferences between the HSR services of equal terminus, and vice versa. The train’s seat capacity is usu-
priority operating on the above-mentioned Category I lines ally constant per service frequency indicating the above-
causing their delays are prevented by the stability of mentioned homogeneous HS train fleet on a given line/
timetable. This implies that the maximum permissible route. For example, if the given line/route operates at the
delay of leading train in the sequence of two trains is service frequency of f(T) = 15 trains/h, and if the average
defined in a way not to cause an additional delay of the turnaround time per train is sl = 4 h, the required number
following train. As such, this delay indicates some kind of of trains will be mrs(T) = 15 9 4 = 60. In addition, if the
the system’s margin allowing delays of the HS trains average train’s seat capacity is s = 485 (TGV Atlantique,
anyway. The longer delays causing disruption of the see Table 3), the total number of required seats will be
timetable occur generally due to other causes. ms(T) = 29,100.

5.2.2 Rolling stock 5.3 Transport work and productivity

The capacity of HSR rolling stock reflects its size expressed The transport work and productivity of a given HSR line/
by the number of trains of a given seating capacity required route can be calculated for the supply and demand sides.
to operate under the conditions specified in the timetable. On the supply side, it counts the total offered number of
These conditions are usually characterized by the service seats during a given period of time. On the demand side, it
frequency during the given period of time (h, day) and the counts the total number of used seats under the same
train’s turnaround time along the given line/route. Conse- conditions.
quently, the required number of rolling stocks/trains to Based on Eq. 2, the transport work on a given line for
carry out at the specified service frequency on a given line, the supply (s-km) (seat-kilometers) and demand (p-km)
mrs [T; f(T)], can be estimated as follows [6, 26]: sides, respectively, can be calculated as follows [26]:
mrs ½T; f ðTÞ ¼ f ðTÞs; ð2aÞ TWSðTÞ ¼ f ðTÞ  s  d and
ð3aÞ
TWDðTÞ ¼ f ðTÞ  s  qðTÞ  d:
where f(T) is the train service frequency on a given line
during time (T) (trains/h; trains/day) and s is the average Similarly, the productivity of both supply and demand
turnaround time of a train along a given line (h). sides of a given line expressed as the volumes of seat-km/h
The service frequency f(T) in Eq. 2a can be either and pax-km/h, respectively, can be calculated as follows:
considered to be equal to the line/route ‘ultimate’ capacity
TPSðTÞ ¼ f ðTÞ  s  v and
determined by Eq. 1 or set up to satisfy the expected ð3bÞ
TPDðTÞ ¼ f ðTÞ  s  qðTÞ  V;
demand as follows [26]:
  where d is the length of a given line (km) and V is the
DðTÞ
f ðTÞ ¼ min ll ðTÞ; ls=l ðTÞ; ls=arr ðTÞ; ls=dep ðTÞ; ; operating speed of HS trains on a given line (km/h).
qðTÞs
The other symbols are analogous to those in the previous
ð2bÞ
equations.
where D(T) is the expected user/passenger demand on a As can be seen, the transport work increases with the
given HSR line during time (T)(pax); q(T) is the average increase of the length of line, service frequency, seat
load factor on a given line during time (T) (q(T) B 1.0); capacity per frequency, and load factor. The productivity
increases with the increase of the service frequency, seat
5
This time is used for disembarking the incoming passengers and capacity, load factor per frequency, and the average train
their baggage, cleaning the interior of the train, replenishing water, operating speed, and vice versa. For example, for the HS
restock, king victuals, changing the crew, and embarking the outgoing trains, each with the seat capacity of s = 485 seats and the
passengers and their baggage. It is typically about 20 min at most
HSR systems. In Japanese HSR system (Shinkansen), it is about average load factor q = 0.80, operating on the line of
12 min [24]. length of d = 500 km at the operating speed of

123 J. Mod. Transport. (2016) 24(1):1–21

A multidimensional examination of performances of HSR (High-Speed Rail) systems 11

V = 300 km/h and the service frequency in the single 7

Before HSR
direction of f(T) = 15 trains/h, the transport work on the 6 By HSR t(d)=0.021d 0.871
line’s demand and supply sides during the period of 1 h R 2 =0.978
5
will be TWS = 15 9 500 9 485 = 3,637,500 (seat-km)

Trip time (h)

and TWD = 15 9 500 9 485 9 0.80 = 2,910,000 (pax- 4
t(d)=0.009d 0.945
km), respectively. The corresponding productivity under 3
R 2=0.984
the same conditions will be TPS = 15 9 485 9 300 = 2
2,182,500 (seat-km/h) and TPS = 15 9 485 9 0.80 9
1
300 = 1,746,000, respectively (pax-km/h).
0
0 200 400 600 800
Line/route length (km)
5.4 Quality of service
Fig. 10 Relationship between the trip time by the HS and conven-
The quality of service provided by HSR systems can be tional rail, and length of line/route in Italy [56]
expressed by the attributes such as schedule delay, trip time
in combination with the reliability and punctuality of ser-
100
vices, the comfort on board HS trains, and accessibility of
the HSR stations [27]. 98

96

Punctuality (%)
5.4.1 Schedule delay
94
The schedule delay is defined as the difference between the 92
desired and the available time of boarding a chosen HSR
service. Under an assumption that the users/passengers 90

familiar with the timetable arrive uniformly during the time 88

between any two successive HS trains’ departures on the Japan railways Renfe Korean HS train TGV france
& Eurostar
same line/route/direction, this delay can be roughly esti-
mated as follows [28, 29]: Fig. 11 Punctuality of services—the ratio—of the selected HSR
T systems [30]
SDðTÞ ¼ ; ð4aÞ
4f ðTÞ
where all symbols are analogous to those in the previous increase of the line/route length, which in the given
equations. For example, for the service frequency of case amounts to 33 %–42 %.
f(T) = 1 train/h, the schedule delay will be • Punctuality of the HSR services can be expressed by
SD(T) = 15 min; for the service frequency of f (T) = 15 two attributes: (i) the ratio of the number of transport
trains/h, the schedule delay will be SD = 1 min (T = 1 h services carried out on time, i.e., according to the
or 60 min). timetable, or with the specified maximum or average
delays, and the total number of services carried out, and
5.4.2 Trip time, reliability, and punctuality (ii) the average delay per delayed service. Both
attributes are recorded during a given period of time
• Trip time by HSR systems is much shorter than that of (day, month, year) under given conditions. The expe-
their conventional counterparts at the same lines/routes. rience so far has shown that these services in general
The potential time savings on a given route can be and on the particular lines/routes have been highly
estimated as follows: punctual as shown in Fig. 11 [30].
As can be seen, the Japanese HSR system has generally
D ¼ d=ð1=VCON  1=VHSR Þ; ð4bÞ
been the most and the UK’s the least punctual. In
where d is the length of a given line/route (km); VCON addition, Fig. 12 shows an example of the punctuality
is the operating speed of the conventional passenger of the Japanese HSR system expressed by the average
train (km/h); and VHSR is the operating speed of the HS delay per service.
train (km/h). Figure 10 shows an example for this in As can be seen, the average delay per HSR service has
Italy. varied from 0.3 to 0.5 min. In addition, the average
As can be seen, the difference in trip time by the delay of the Shinkansen HSR system has been about
conventional and HSR trains increases with the 0.6 min per service over the last decade [24, 31].

J. Mod. Transport. (2016) 24(1):1–21 123

12 M. Janic

5.4.3 Accessibility

Average delay per service-train (min)

1.0
Tokaido
Tohoku, Joetsu, Nagano
0.8
Accessibility of stations is an important attribute of the
0.6
overall quality of services provided by the HSR systems. In
most cases, the new dedicated HSR stations are usually
0.4 located and designed to fit as good as possible within the
surrounding urban and/or sub-urban layout on one hand
0.2
and enable the satisfactory quality of accessibility on the
0.0
other. In some other cases, the parts of conventional rail-
1998 2000 2002 2004 2006 2008 2010 way stations have been appropriately upgraded and adapted
Time (year)
to serve the HSR services. In both cases, the quality of
Fig. 12 Punctuality of services—the average delay—of the Japanese accessibility needs is expected to be efficient, effective, and
HRS system (period 1999–2008) [24, 57] safe. This implies a reasonable (acceptable) time and costs
from/to the doors of users/passengers by a variety of urban
and sub-urban transit modes (car, taxi, and frequent,
0.15 punctual, reliable, and safe, i.e., without incidents/acci-
Failure rate number (×10 6 km)

dents due to known reasons, bus, tram, metro, regional rail,

0.12
etc.), respectively.
0.09
5.4.4 Comfort on board the HS trains
0.06
The comfort offered to their users/passengers on board of
0.03 the HS trains usually includes the booked seats and the
very limited number of stops along the lines/routes com-
0.00
1986 1988 1990 1992 1994 1996 1998 2000 pared to those at the conventional train counterparts. As far
Time (year) as the comparison with the ATP system as the main
competitor on the short- and medium-haul liens/routes is
Fig. 13 Reliability of the HSR rolling stock (East Japan Railways—
period 1987–2000) [24] concerned, the attributes for comparison have usually been
the distance between seats and internal mobility, diversity
As can be seen, the average delay per HSR service has and type of services, noise on board, and the potential
varied from 0.3 to 0.5 min. In addition, the average delay impact on health. Table 4 summarizes these for both
of the Shinkansen HSR system has been about 0.6 min per systems/modes.
service over the last decade [24, 31]. As can be seen, the HS trains have generally possessed
• Reliability of the HSR services can be expressed as the higher comfort on board than their aircraft counterparts.
ratio between the realized and planned transport
services during a given period of time (day, month,
and year) under given conditions. This is dependent on 6 Economic performances of HSR systems
the rate of failure of rolling stock due to any system’s
internal and/or external reasons causing cancelation or The economic performances of HSR systems include their
long delays of the affected services. Figure 13 shows an costs and revenues. The costs are imposed by implemen-
example of the Japanese HSR system. tation and operation of the systems. The revenues obtained
mainly by charging users/passengers cover the costs and
As can be seen, this rather very low failure rate has provide some funds for updating the system and the profits
fluctuated during the observed period with an average of for particular stakeholders involved. In any case, both
0.084 failures/106 km.6 revenues and costs need to be balanced in order to guar-
6
antee the economic and financial stability of the system.
This has been achieved by maintaining the rolling stock at four
levels: (i) daily inspection (every 2 days), i.e., inspection of the wear
parts (pantograph strip, refreshing water/waste); (ii) regular inspec- 6.1 Costs
tion (every 30 days or 30,000 km) (test of conditions and function,
inspection of the important parts/components without decomposi- The total costs of a given HSR system generally consist of
tion); (iii) inspection of bogie (every 1.5 year or 600,000 km) (bogie
the infrastructure and operating costs. The infrastructure
parts by decomposition); and (iv) the overall inspection (every
3 years or 1,200,000 km) (inspection of the overall rolling stock by costs include: (i) planning the system and acquisition and
decomposition) [32]. preparing the land; (ii) building the lines and stations

123 J. Mod. Transport. (2016) 24(1):1–21

A multidimensional examination of performances of HSR (High-Speed Rail) systems 13

Table 4 Some attributes of the comfort on board of the HS trains and Table 5 The average infrastructure cost of the HSR lines worldwide
commercial passenger aircraft [30] [3, 35, 42]
Attribute System/mode Country Cost (million €/km)
HS train Aircraft Built (in service) lines Under construction lines

Distance between seats (cm) 87–97 78–85 Austria – 18.5–39.6

Internal mobility (general) Higher Lower Belgium 16.1 15.0
Services (food, drink, internet, etc.) (general) Similar France 4.7–18.8 10.0–23.0
Noise (dBA) 62–69 70–82 Germany 15.0–28.8 21.0–33.0
Impacts on health (general) Lower Higher Italy 25.0 14.0–65.8
Netherlands – 43.7
including tunnels and bridges, and the supportive facilities Spain 7.8–20.0 8.9–17.5
and equipment including the signaling systems, catenaries Japan 20.0–30.0 25.0–40.0
and electrification mechanisms, and communications and South Korea – 34.2
safety installations; and (iii) maintenance of the entire Chinaa 8.4–12.3b/11.3–22.0c –
infrastructure and supporting facilities and equipment [33]. a
27 HSR lines in operation
The operating costs include acquiring, operating, and b
Design speed: 250 km/h
maintaining the rolling stock, selling services, and admin- c
Design speed: 350 km/h; 1 RMB (Chinese Yuan) & 0.12 €
istration. The costs of labor, material, and energy have the
largest share in the total costs [33].
Table 5 gives an indication of the average infrastructure 6.2 Revenues
cost of the already built and planned HSR lines, which do
not include the cost of planning, and acquisition and The HSR systems obtain revenues from different sources
preparation of the land. such as the transport-based charging users/passengers,
As can be seen, the average infrastructure cost for both merchandise, and others [37]. In particular, the prices for
already built and under-construction HSR lines has sig- users/passengers are set up to cover the systems’ total
nificantly varied in both European and non-European, i.e., operating cost in cases of the lack of subsidies. The latter can
two Asian countries. In Europe, the lowest cost has been in be used as an element for enabling stronger competition with
France and Spain, and much higher in Italy, Germany, and the other transport modes such as the conventional rail and
Belgium. It can be shown that the average infrastructure particularly APT, both on the above-mentioned competitive
cost has been 18 million€/km. In addition, the average cost lines/routes. Figure 14 shows relationship between the
of building the new HSR lines in Asian countries (Japan, annual revenues and the annual satisfied demand of the HSR
South Korea, except China) has been slightly higher than systems in different countries [19].
those in particular European countries [34, 35]. As well, the As can be seen, the revenues have generally linearly
average maintenance cost per unit of length of the HSR increased with increasing of the volumes of satisfied
system infrastructure has also highly varied, mainly demand at an average of 17.44 ¢US$/p-km, which is in line
depending on the length of lines. Some estimates indicate with the above-mentioned corresponding costs.
that the average maintenance cost in European countries
has amounted from about 13–72 thousands/year [35, 36]. 6.3 Balancing revenues and costs
The average cost of operating the HSR services has also
differed throughout the European counties and rest of the The HSR systems intend to operate in the profitable way,
world as well. This cost has been mainly influenced by the i.e., to cover their costs by revenues. Figure 15 shows an
local pricing of the particular above-mentioned inputs and example of the profitability of the Japanese HSR operating
type of the HS trains. Some estimates indicate that this both HSR and conventional rail services.
average operating cost for 12 types of the HS trains oper- As can be seen, despite a relatively high variations the
ating in the corresponding European countries has been: profitability has generally increased with increasing of the
C ¼ 0:14626 €/seat-km. In this total, the cost of mainte- volume of the company’s output during the given period of
nance of the rolling stock has shared about 8.5 %. Under an time. This case could be used as an example how the HSR
assumption that the average load factor was: h = 0.8 (i.e., system can be profitable in the medium- to long-term
80 %), the total average operating costs of the HSR ser- period—by careful balancing the revenues and costs while
vices throughout Europe would be: C ¼ 0:183 €/p-km [34, at the same time increasing the scale of operations to sat-
35]. isfy the growing user/passenger demand.

J. Mod. Transport. (2016) 24(1):1–21 123

14 M. Janic

50,000 shows a scheme of changing the distance and time of

exposure to noise by an HS train of an observer.
Revenue (× 10 6 $ US/year)
40,000
R(D)=0.1744D+2209.8
2
The shadow polygon represents an HS train of length (L)
R =0.980
30,000 passing by an observer (small triangle at the bottom) at the
speed (V). He/she starts to consider noise of an approaching
20,000
train when it is at distance (b) from the point along the line,
10,000 which is at the closest right angle distance (c) from him/her.
The consideration stops after the train moves behind the
0
0 50,000 100,000 150,000 200,000 250,000 above-mentioned closest point again for the distance (b).
Demand (×10 6 p-km/year) Under such circumstances, the distance between the obser-
ver and the passing-by HS train changes over time as follows:
Fig. 14 Relationship between the annual revenues and the satisfied
passenger demand of particular HSR systems—Japan, France, q2 ðtÞ ¼ ðL=2 þ b  V  tÞ2 þ c2
Germany, UK, USA (period 2012–2015) [19] ð5aÞ
for 0 \t\ ¼ ðL þ 2  bÞ=V;

where the last term represents duration of the noise event,

4.0
i.e., the time needed for a train to pass by the observer (The
Net income/profits (¢$ US/p-km)

3.5 length of HS trains is given in Table 3). If the level of noise

P(D)=0.253D-10.929
R 2=0.556
received from the train passing by an observer with the
3.0
speed (V) at the shortest distance (c) is Leq(c, V), the level
2.5 of noise at any time (t) can be estimated as follows:
2.0 Le ½qðtÞ; V ¼ Leq ðc; VÞ  8:6562 ln½qðtÞ=c: ð5bÞ
1.5 The second term in Eq. 5b represents the noise
attenuation with distance over the area free of barriers.
1.0
49 50 51 52 53 54 55 56 The total noise exposure of the observer from f(T)
Demand (×10 9 p-km/year) successive trains passing by during the period (T) can be
estimated as follows:
Fig. 15 Relationship between the annual demand and the net income/
profits—Central Japan Company (period 2004–2013) [37] f ðTÞ
X Le ½k;qðtÞ;V
Leq ½f ðTÞ ¼ 10 log 10 10 : ð5cÞ
k¼1
7 Social performances of HSR systems
As a standard approach, the noise from HS trains is
The social performances of HSR systems include the measured at the right angle distance of c = 25 m from the
impacts and effects. The impacts embrace noise, conges- track(s). Figure 17 shows the results of some such
tion, and safety, i.e., traffic incidents and accidents. The measurements across Europe depending on maximum
effects generally refer to the system’s overall welfare operating speed of the HS trains.
expressed by savings of the user/passenger time, relieving As can be seen, the noise has generally linearly
congestion from roads, and contribution to the regional increased with increasing of the train’s operating speed: at
GDP through direct and indirect employment. the lower rate for the speeds up to about 300 km/h, and at
the higher rate for the speeds above V = 300 km/h. The
7.1 Impacts variation of noise level at the given speed has been about
3–4 dBA. This noise has included the train’s rolling
The HSR system generally impacts the society/people by (wheel), pantograph/overhead, and aerodynamic noise.
noise, congestion, and safety, i.e., traffic incidents and Some additional measurements have shown that the rolling
accidents. and pantograph/overhead noise has predominated and
increased with increasing of the HS train’s speed approx-
7.1.1 Noise imately at the rate of 30lgV up to the speed(s) of about
300 km/h (some data have shown that this is 370 km/h).
The HS trains generate noise while operating at the high The aerodynamic noise depending on the HS train’s
speed(s), which comprises rolling, aerodynamic, equip- (aerodynamic) design has also increased, equalized with
ment, and propulsion sound. This noise mainly depends on the rolling noise at the above-mentioned (transition)
its level generated by the source, i.e., moving HS train(s), speed(s), started predominating and further increasing at an
and its distance from an exposed observer(s). Figure 16 approximate rate of 80lgV [38]. In addition, in cases when

123 J. Mod. Transport. (2016) 24(1):1–21

A multidimensional examination of performances of HSR (High-Speed Rail) systems 15

t=0 t1=(L+2)/V
L/2 L/2

Vt

L L

(0) (t1)

Observer-noise receiver

Fig. 16 Scheme for determining the noise exposure of an observer by passing by HS train [58]

98 routes and the entire HSR network accordingly, the HSR

96
systems are free of congestion and consequent delays due
Le=0.0625V+71.75 to the direct mutual influence of trains on each other while
94
Noise level (dBA)

R 2=0.680
‘competing’ to use the same segment of given lines/routes
92 Le=0.1169V+54.494 at the same time. However, the substantive delays due to
R 2=0.841
90 some other reasons can propagate (if impossible to absorb
88 and neutralize them) through the affected HS trains itin-
86 eraries as well as along the dense lines/routes also affecting
the other otherwise non-affected services. Under such
84
240 260 280 300 320 340 360 conditions, the severely affected services are usually can-
Train speed (km/h)
celed in order to prevent further increase and propagation
Fig. 17 Relationship between the noise and the maximum operating of their delays. On the one hand, this contributes to
speed of the passing-by HS train(s) measured at the right angle maintaining the punctuality but on the other, it compro-
distance of 25 m (Belgium, France, Germany, Spain, Italy) [59] mises the reliability of the overall services (as mentioned
above). Nevertheless, the already mentioned figures indi-
the frequent HSR services are carried out along the par- cate that both reliability and punctuality of the HSR system
ticular lines/routes, their noise becomes persistent over services worldwide have been very and in some cases
time and can be estimated from Eq. 5c. As well, the time of extremely high (The latter is the example of Japanese HSR
exposure of an observer to noise by a passing by HS train system).
can be estimated from Eq. 5a. If b = 0 m, L = 200 m, and
v = 250 km/h, this exposure time to the maximum noise 7.1.3 Safety, i.e., traffic incidents/accidents
will be about t1 = 3 s; if V = 350 km/h, this time will be
about t1 = 2 s. Experience so far has indicated that the HSR and APT
Last but not least, while considering the actual exposure system have been the safest transport systems/modes in
of the population located close to the HSR lines to noise by which traffic incidents/accidents have rarely occurred,
the passing-by HS trains, it is necessary to take into account usually due to the previously unknown reasons. This means
the noise-mitigating barriers protecting the particular land that the number of traffic incidents/accidents and related
use activities, i.e., a quiet land with intended outdoor use, a person injuries, deaths, and the scale and cost of damaged
land with the residence buildings objects, and a land with the properties both of the systems and the third parties per, for
daytime activities (businesses, schools, libraries, etc.), all by example, 109 s-km and/or p-km carried out over a given
absorbing the maximum noise levels for about period of time, have been extremely low. In particular, high
20 dB(A) (single barrier) and 25 dB(A) (double barrier). safety of the HSR systems has been provided also a prior
by designing completely the grade-separated lines and the
7.1.2 Congestion other supportive built-in safety features at both infrastruc-
ture and rolling stock. This implies that the safety has been
Thanks to applying the above-mentioned separation rules achieved on the account of increased investments and
in addition to designing timetable(s) on particular lines/ maintenance cost. As well, the HSR operators and

J. Mod. Transport. (2016) 24(1):1–21 123

16 M. Janic

Table 6 Characteristics of the HSR fatal accidents [60–62]

Country/system/number of trains Date Cause Passengers on board Fatalities Injuries

Germany/ICE/1 03/06/1998 Wheel disintegration 287 101 88

China/2 23/07/2011 Railway signal failure 1,630 40 [210
Spain/Alvia/1 24/07/2013 Derailment due to excessive speed 222 [79 139

infrastructure managers have continuously practiced a risk ambiguous and often politically challenging task. Never-
management and training approach aiming at maintaining a theless, some estimates of these externalities for the HSR
high level of safety and particularly with increasing of the systems and other transport modes in Europe have been
maximum speeds. Nevertheless, the HSR systems in dif- carried out. They have indicated that the total social
ferent countries have not been completely free from traffic externalities of HSR systems have amounted 22.9 €/103
incidents/accidents. For example, some relevant statistics p-km. In this total, the noise and traffic incidents/accidents
for the TGV system in France indicate that there have not externalities have shared about 22 % and 2 %, respec-
been accidents with the fatalities (deaths) and severe tively. Since the HSR systems are free of congestion, the
injuries of the users/passengers, staff, and/or third parties corresponding externality has not been considered. On the
since starting the HSR services started in the year 1981 other hand, for comparison, the total externalities of APT
despite the trains have been carrying out annually about have estimated to be 52.5€/103 p-km, of which the noise
10 9 106 p-km. In addition, some incidents happened on and traffic incidents/accidents externalities shared about
the HSR lines/routes such as broken windows, opening of 4 % and 3 %, respectively [39, 40].
the passenger doors during operating at the cruising speed,
couple of fires on board, collision with animals and con- 7.2 Effects
crete block on the tracks, and the terrorist attempts to bomb
the tracks. The incidents and accidents of TGV trains The effects of HSR systems have consisted of contribution
operated on the conventional tracks have been more fre- to the direct and indirect employment and consequently the
quent with fatalities, injuries, and damages of properties economic-social development and welfare, both at a glo-
but all at the relatively low scale. In these cases, the HS bal-country and the local–regional scale.
trains have been exposed to the external risk similarly to
their conventional counterparts (http://www.railfaneurope. 7.2.1 Direct employment
net/tgv/wrecks.html). Similarly, since started in 1960s, the
Japan’s Tokaido Shinkansen HS services7 have also been The direct employment relates to manufacturing, building,
free of accidents causing the user/passenger and staff and maintaining the infrastructure and manufacturing,
fatalities and injuries due to the derailments and collisions operating, and maintaining the rolling stock and supporting
of trains. This has been achieved despite the services have facilities and equipment, i.e., the main system’s compo-
been exposed to the permanent threat of the relatively nents, of the HSR systems. For example, the number of
frequent (and sometimes strong) earthquakes. employees operating the HSR services in particular coun-
Nevertheless, the fatal accidents with deaths and injuries tries is strongly dependent on the length of HSR networks
of the users/passengers and staff happened at the HSR as shown on Fig. 18.
systems in Germany, Spain, and China (one in each A can be seen, in the considered countries, the number
country). Table 6 gives the main characteristics of these of employees increases linearly with increasing of the
three accidents. length of HSR network with an average of 7.3 employees/
km.
7.1.4 Cost of the social impacts—externalities
7.2.2 Indirect employment
Quantifying the social impacts of HSR systems in the
monetary terms as externalities has usually represented an The indirect employment relates to the non-rail staff sup-
plying the HSR system(s) with different kinds of daily
7
The Tokaido Shinkansen line/route of the length of 552.6 km consuming material and energy on the one hand and that
connects Tokyo and Shin Osaka station is free of the level crossings. generated just thanks to existing of the system on the other.
The trains operate at the maximum speed of 270 km/h covering the These latter are the non-rail related economic activities
line/route in 2 h and 25 min. The route/line capacity is: ll = 13
trains/h/direction. The number of passengers carried is about 386 around and at the HSR stations such as: business services
thousand/day and 141 million/year (2011) [31]. (banking, insurance, and advertising), information and

123 J. Mod. Transport. (2016) 24(1):1–21

A multidimensional examination of performances of HSR (High-Speed Rail) systems 17

250 bypassed where this growth has been for about 1 %. This
growth has taken place primarily in the cities with the
Numver of employees (×10 3)
200 Ne=7.287L N-1250.9
2
R =0.994
information industry and higher education [44].
150

100 8 Environmental performances of HSR systems

50 The environmental performances of the HSR systems

generally include the energy consumption and related
0
0 5 10 15 20 25 30 35 emissions of GHG, the area of land used for settling down
Length of the network (×10 3 km) the system’s infrastructure, and the related costs consid-
ered, if internalized, as externalities. For the given HSR
Fig. 18 Relationship between the number of employees and the
length of HSR network—Japan (Central, East, West), SNCF (France), system, these performances can be considered at different
DB AG (Germany) (period 2014) [19] time and spatial scale. In the former case, this could be the
instant, short, medium, and/or life cycle assessment (LCA).
In the latter case, in combination with the former, these
retail services, research and development, higher educa- performances can be considered for the particular HSR
tion, tourism, and political institutions [30]. At the larger lines and/or the entire network [46].
scale, these businesses have created urban (both business
and housing) agglomerations around the HSR stations, 8.1 Energy consumption and emissions of green
which themselves have induced additional demand for the house gases (GHG)
HSR services. Such development has been taking place
mainly at the HSR stations already located in the larger Energy consumption and related emissions of GHG are
urban agglomerations connected by the HSR lines/routes, considered exclusively from operations of the HSR sys-
but also within them. For example, inclusion of the city of tems, which excludes those from building the infrastructure
Lille (France) in the HSR line/route Paris-Brussels has (lines) and manufacturing the supporting facilities and
brought an enormous economic development of the city equipment and rolling stock (trains) [47].
itself and its region in terms of increasing of business and In general, the HS trains consume electric energy pri-
touristic activities and related employment. In the UK, the marily for accelerating up to the operating/cruising speed
substantial economic activities have been created in the and then for overcoming rolling/mechanical and aerody-
cities 2 h from London area just thanks to the HSR [41]. namic resistance to motion at that speed. This also
includes the energy for overcoming resistance of grades
7.2.3 Contribution to the local and global economy and curvatures of tracks along the given line/route. As
and welfare well, the energy is consumed for powering the equipment
on board the trains. In particular, during the acceleration
In general, the above-mentioned employment has con- phase of a trip the electric energy is converted into kinetic
tributed to the economic-social development and welfare, energy at an amount proportional to the product of the
both at a global-country and local–regional scale. For train’s mass and the square of its speed(s). A part of this
example, at the global-country scale, the direct effects have energy recovers by regenerative breaking during deceler-
been contribution of the investments in HSR systems to the ation phase before the train stops. During cruising phase of
national GDP, which in Europe has estimated to be about a trip, the energy is mainly consumed to overcome the
0.25 % of the national GDPs. At the regional scale, this rolling/mechanical and the aerodynamic resistance, which
contribution has been about 3 % of the regional GDP [42, for a given type of HS train can be expressed as follows
43]. This contribution has been much higher in the cities [48]:
with the primarily service-oriented than in those with the
R ¼ RM þ RA ¼ ða þ bVÞW þ cV 2 ; ð6aÞ
primarily manufacturing-oriented economy [44]. In addi-
tion, the German regions with the cities of Montabaur and where RM, RA are the rolling/mechanical and aerodynamic
Limburg, with populations of 12,500 and 34,000 respec- resistance, respectively (N); W is the weight of a train
tively, have recorded growth of GDP of about 2.7 % just (tons); V is the operating/cruising speed of a train (km/h);
due to increase in their market accessibility to the larger and a, b, c are the experimentally estimated coefficients.
cities Frankfurt and Cologne thanks to the HSR services Equation 6a essentially reflects the Davis’s equation
[45]. In Japan, the HSR has generated growth of population with the corresponding coefficients. It indicates that the
in the cities of about 1.6 % compared to those being aerodynamic resistance generally increases with the square

J. Mod. Transport. (2016) 24(1):1–21 123

18 M. Janic

of operating/cruising speed. The rolling mechanical resis- For some comparison, the emission rate of an average
tance increases linearly with the increase of this speed and passenger car is around EMR = 140 gCO2/km. This is
weight of the HS train. Some experiments carried out for likely to decrease to about EMR = 130 gCO2/km by the
Shinkansen Series 100 HS trains estimated the total resis- year 2020. However, the new cars to be launched in the
tance depending on the cruising/operating speed as follows: meantime are expected to have the emission rate of about
R(V) = 8.202 ? 0.10656 V ? 0.00116232 V2 (R(V) in kN EMR = 120 gCO2/km, which is just according to the EU
and v in m/s) [40, 48]. The above-mentioned relationship proposals. In addition, this could be reduced to about
emphasizes the importance of reducing both the weight of EMR = 80 gCO2/km mainly thanks to more massive
train and its aerodynamic resistance in order to achieve introduction of hybrid cars by the year 2030, and to about
savings in the energy consumption during the longest phase EMR = 57 gCO2/km during the period between the years
of trip—cruising at high speed. 2040 and 2055 when the electric or fuel-cell cars are
Estimates of the energy consumption by different types supposed to only really contribute to the more significant
of HS trains including acceleration/deceleration/cruising reduction of the above-mentioned emission rates. Similarly
phase of a trip have differed and changed over time, just to the HS trains, this will be carried out in parallel to the
thanks to the above-mentioned permanent improvements of changing the structure of the primary sources for producing
their both characteristics (aerodynamic, weight) and oper- electric energy. In addition, the fuel efficiency and related
ations. Table 7 provides some recent estimates of this emissions of CO2 and other GHG by APT competing with
energy efficiency for different types of the HS trains. the HSR on the short- to medium-haul lines/routes will also
As can be seen, the Japanese Shinkansen is the most and be improved in the forthcoming decades. For example, the
the Eurostar the least energy efficient trains. One of the emission rate of CO2 is expected to decrease from today’s
reasons is the relatively large difference in the seat capacity average of EMR = 97–62 gCO2/s-km by the year 2025 to
between them. As an indication, at present, the average EMR = 47 and 41 gCO2/s-km by the years 2040 and 2055,
energy efficiency of an HS train is assumed to be about respectively (the emission conversion factor is 1 g of Jet A
EC = 0.033 kWh/s-km. Considering this and taking into fuel = 3.18 gCO2/s-km; the aircraft types considered are
account the emission rates of the primary sources for pro- similar to today’s A319 and B737-800 models). The
ducing electricity in Japan, the average rate of emissions of mentioned improvements are expected to be achieved by
GHG by Shinkansen trains is EMR = 42 gCO2/s-km [19]. improving the aircraft airframe and engine efficiency.
Under the analogous conditions, in Europe, this average rate Beyond the year 2050, further improvements may be
is EMR = 21 gCO2/s-km with an ambition to be reduced to expected means by introducing the alternative fuels such
EMR = 5.9 gCO2/s-km by the year 2025, 1.5 gCO2/s-km by as, for example, liquid hydrogen [6, 49]. Nevertheless, the
the year 2040, and 0.9 gCO2/s-km by the year 2055. This above-mentioned figures indicate that the HSR systems
reduction is expected to be achieved through further will remain superior in terms of energy efficiency and
improvement of the energy efficiency of HS trains and their related emissions of GHG (CO2) as compared to its com-
operations on one side and by changing type and composition petitors—passenger cars and the short- to medium-haul
of the primary sources for producing electric energy on the commercial aircraft.
other. In the latter case, the aim is to produce as much as
possible electric energy from the renewable decarbonized 8.2 Land use
primary sources [30, 47].
The HSR infrastructure directly occupies much smaller
area of land than its road–highway counterpart. For
example, if the width of an HSR line is (w) and the length
Table 7 Energy efficiency of different types of HS trains [49, 51] (d), the total occupied land can be estimated as follows:
Train type Operating Seat capacity EC (energy A ¼ wd: ð6bÞ
speed (km/h) (seats) efficiency)
(kWh/s-km) For example, if w = 25 and d = 1 km line, the total area
Shinkansen Series 700 300 1,323 0.029 of directly taken land will be A = 2.5 ha (ha—hectare) (the
AVG 300 650 0.033 average gross area of taken land is 3.2 ha). For a highway
TGV Reseau 300 377 0.031
with three lanes in both directions whose width is w = 75 m
TGV Duplex 300 545 0.032
and length d = 1 km, the directly taken land is A = 7.5 ha
Pendolino Class 300 300 439 0.033
(the average gross area of the taken land is about 9.3 ha, i.e.,
three times greater than that of the HSR line). In addition,
Eurostar Class 323 300 750 0.041
utilization of the taken land by both modes is quite different.
Velaro D 320 601 0.030
The capacity of HSR line/route in both directions is two

123 J. Mod. Transport. (2016) 24(1):1–21

A multidimensional examination of performances of HSR (High-Speed Rail) systems 19

times of 12–14 trains/h, i.e., 24–28 trains/h. If each train The particular performances have been elaborated in
carries about 600 passengers, the intensity of land use will be both descriptive and analytical ways dependent on the most
24–28 9 600/2.5 = 5,760–6,720 pax/h/ha. In case of the influential factors. In the latter case, some analytical
above-mentioned highway with the capacity of 4,500 veh/h models of particular performances have been presented. In
and the occupancy rate of 1.7 pax/car, the intensity of land addition, where considered appropriate, a comparison of
use will be 1,020 pax/h/ha, which is for about 6–7 times the performances of HSR systems with those of the com-
lower than that of HSR [40]. peting systems operated by other transport modes has been
carried out.
8.3 Externalities Finally, the HSR systems have been shown to be the
mass high-speed inter-urban transport systems serving the
The energy consumption and related emissions of GHG user/passenger demand generally efficiently, effectively,
and land use by the HSR systems have also been consid- and safely through competition and/or cooperation with its
ered as externalities. Similarly to the case of social exter- conventional rail counterpart, car, and APT, where and if
nalities, the HSR systems have been shown to be rather appropriate.
superior compared to the other competing transport modes
such as road passenger cars and APT. Some estimates have Open Access This article is distributed under the terms of the
Creative Commons Attribution 4.0 International License (http://
indicated that the air pollution associated with the climate creativecommons.org/licenses/by/4.0/), which permits unrestricted
change shares about 26 % and the land use about 30 % in use, distribution, and reproduction in any medium, provided you give
the total HSR system externalities of 0.00229€/103 p-km. appropriate credit to the original author(s) and the source, provide a
After including the above-mentioned share of the social link to the Creative Commons license, and indicate if changes were
made.
externalities, the rest to 100 % is the share of up- and
downstream, and urban externalities. The corresponding
figures for APT are 86 % for the emissions of GHG and
2 % for the land use. After including the share of social References
externalities, the rest to 100 % is the share of urban, and
up- and downstream externalities in the total of about 1. UIC (2010) Necessities for future high speed rolling stock.
Report January 2010, International Union of Railways, Paris
0.00525€/103 p-km [39, 40]. 2. EC (1996) Interoperability of the trans-European high speed rail
system, Directive 96/48/EC. European Commission, Brussels
3. Ollivier G, Sondhi J, Zhou N (2014) High-speed railways in
9 Conclusions China: a look at construction costs. China Transport Topics No.
9, World Bank Office, Beijing, pp 1–8
4. USDOT (2009) Vision for high-speed rail in America. U.S.
This paper has dealt with the multidimensional examina- Department of Transportation, Federal Railroad Administration,
tion of infrastructural, technical/technological, operational, Washington DC
economic, social, and environmental performances of the 5. Wendell C, Vranich J (2008) The California high speed rail
proposal: a due diligence report. Reason Foundation (with
HSR systems. The infrastructural performances have been Howard Jarvis Taxpayers Association and Citizens Against
related to the geometrical characteristics and design of the Government Waste), Los Angeles, California
HSR lines and stations. The operational performances have 6. Janić M (2014) Advanced transport systems: analysis, modelling,
included demand, capacity, and their dynamic relationship and evaluation of performances. Springer-Verlag, New York
7. UIC (2002) Feasibility of ‘‘Ballastless’’ track. UIC Report,
reflected through the quality of transport services provided Infrastructure Commission—Civil Engineering Support Group,
to the users–passengers. The economic performances International Union of Railways, Paris
included the cost and revenues of setting up and operating 8. UIC (2004) UIC Code 406. International Union of Railways,
the HSR system(s) and the revenues gained from charging Paris
9. Takai H (2013) 40 Years experiences of the slab track on Japa-
users–passengers. The social/policy performances have nese high speed lines. Presentation, Railway Technical Research
included the impacts and effects of the HSR systems on the Institute, Tokyo
society. The former have included noise, congestion, and 10. Takagi K (2011) Expansion of high speed rail services: devel-
traffic incidents/accidents, i.e., safety, and their externali- opment of high-speed railways in China. Jpn Railw Transp Rev
57:36–41
ties. The latter have included global and local direct and 11. Anderson T, Lindvert D (2013) Station design on high speed
indirect contributions (benefits) to the economy in the railway in Scandinavia: a study of how track and platform
widest sense. The environmental performances have technical design aspects are affected by high speed railway
embraced the energy consumption and related emissions of concepts planned for the Oslo Göteborg line. MSc Thesis,
Chalmers University of Technology, Göteborg
GHG (Green House Gases), and land use with both direct 12. Kido ML (2005) Aesthetic aspects of railway stations in Japan
and indirect impacts on the environment, and associated and Europe, as a part of context sensitive design for railways.
externalities. J East Asia Soc Transp Stud 6:4381–4396

J. Mod. Transport. (2016) 24(1):1–21 123

20 M. Janic

13. USDOT (1999) Assessment of potential aerodynamic effects on 36. Henn L, Sloan K, Douglas N (2013) European case study on the
personnel and equipment in proximity to high-speed train oper- financing of high speed rail. In: Proceedings of Australasian
ations. U.S. Department of Transportation, Federal Railroad transport research forum, 2–4 October 2013, Brisbane
Administration, DOT/FRA/ORD-99/11, DOT-VNTSC-FRA-98- 37. JR Central (2014) Visitors guide 2014. Central Japan Railway
3, Washington DC Company, Nagoya
14. UIC (2014) High Speed Lines in the World. Updated 1st 38. Thompson JD, Eduardo LE, Liu X, Zhu J, Hu Z (2015) Recent
September 2014, UIC High Speed Department, International developments in the prediction and control of aerodynamic noise
Union of Railways, Paris from high-speed trains. Int J Rail Transp 3(3):119–150
15. Fu J, Nie L, Meng L, Sperry RB, He Z (2015) A hierarchical line 39. INFRAS/WWW (2000) External costs of transport: accident,
planning approach for a large-scale high speed rail network: the environmental and congestion costs in Western Europe. INFRAS
China case. Transp Res Part A 75:61–83 Consulting Group for Policy Analysis and Implementation, Zür-
16. Crozet I (2013) High speed rail performance in France: from ich, WWW University of Karlsruhe, Karlsruhe
appraisal methodologies to ex-post evaluation. Discussion Paper 40. UIC (2010) High speed rail: fast track to sustainable mobility.
No. 2013-26, The Roundtable on Economics of Investments in International Union of Railways, Paris
High Speed Rails, International Transport Forum, 18–19 41. Baron, T., (2009), High Speed Rail Contribution to Sustainable
December 2013, New Delhi Mobility, HAL Id, http://dumas.ccsd.cnrs.fr/dumas-00793156
17. ABB (2014) Powering the world’s high speed rail networks. ABB 42. de Rus G, Nombela G (2007) Is investing in high-speed rail
ISI Rail, Geneva socially profitable? J Transp Econ Policy 41(Part 1):3–23
18. JR Central (2011) Data Book 2011, Central Japan Railway 43. Preston J (2013) The economic of investment in high speed rail:
Company, Nagoya, Japan summary and conclusions. Discussion Paper No. 2013-30,
19. JR Central (2015) Annual Report 2015. Central Japan Railway International Transport Forum (ITF/OECD), Paris, p 38
Company, Nagoya 44. Albalate D, Bel G (2010) High speed rail: lesson for policy
20. Janić M (1984) Single track line capacity model. Transp Plan makers from experiences abroad. Research Institute of Applied
Technol 9:135–151 Economics, University of Barcelona, Barcelona
21. Parkinson T, Fisher I (1996) Rail Transit Capacity Report 13. 45. Boqué RJ (2012) High-speed rail: economic evaluation, decision-
Transit Cooperative Research Program, Transportation Research making and financing, master-Arbeit, Technische Universität
Board, National Research Council, National Academy Press, Dresden, Fakultät Verkehrswissenschaften ‘‘Friedrich List’’.
Washington DC Institut für Gestaltung von Bahnanlagen, Dresden
22. Connor P (2011) Rules for high speed line capacity or how to get 46. Yue Y, Wang T, Liang S, Yang J, Hou P, Qu S, Zhou J, Jia X,
realistic capacity figure for a high speed rail line. Info Paper No. Wang H, Xu M (2015) Life cycle assessment of high speed rail in
3, Rail Technical Webpages, PRC Rail Consulting Ltd., Sutton China. Transp Res D 41:367–376
Bonington, Loughborough 47. UIC (2010) High speed, energy consumption and emissions. UIC
23. Hunyadi B (2011) Capacity Evaluation for ERTMS (European Publications, International Union of Railways, Paris
Rail Traffic Management System) Level 2 Operation on HS2. 48. Rochard PB, Schmidt F (2000) A review of methods to measure
Bombardier Transportation Rail Control Solutions, Bombardier and calculate train resistances. Proc Inst Mech Eng Part F
Inc., Montréal, Canada 214:143–185
24. Nishiyama T (2010) High-speed rail operations in Japan, interna- 49. ATOC (2009) Energy consumption and CO2 impacts of high
tional practicum on implementing high speed rail in the United speed rail. Association of Train Operating Companies Ltd.,
States. APTA, American Public Transport Association, New York London
25. Janić M (1988) A practical capacity model of a single track line. 50. CSP (2014) China statistical yearbook 2014. China Statistical
Transp Plan Technol 12:301–318 Press, National Bureau of Statistics of China, Beijing
26. Vuchic RV (2007) Urban transit systems and technology. Wiley, 51. Siemens (2014) Valero CN high speed trains for China railways.
New York Mobility Division, Siemens AG, Berlin
27. Carol DC (2011) High-speed rail is not about trains, NC 52. EC (2014) EU transport in figures: statistical pocketbook 2014.
1-NETWORK—High Speed Rail 73: 6–7 European Commission, Publications Office of the European
28. Janić M (2000) Air transport system analysis and modelling. CRC Union, Luxembourg
Press (Taylor & Francis Group), Abington 53. de Rus G (2008) The economic effects of high-speed rail
29. Janić M (2003) High-speed rail and air passenger transport: a investments. Discussion Paper, 2008-16, OECD-ITF Transport
comparison of the operational environmental performance. Proc Research Centre
Inst Mech Eng Part F 217(4):259–269 54. IR (2012) The benefits of high-speed rail in comparative per-
30. UIC (2011) High speed rail and sustainability. UIC Publications, spective, Invensys rail. www.invensysrail.com/whitepapers/hsh-
International Union of Railways, Paris research-report.pdf
31. JR Central (2012) Data book 2011. Central Japan Railway 55. Wu J (2013) The financial and economic assessment of China’s
Company, Nagoya high speed rail investments: a preliminary analysis. Discussion
32. Yanase N (2010) High speed rolling stock in Japan: international Paper No. 2013/28, Prepared for the Roundtable on The Eco-
practicum on implementing high speed rail in the United States. nomics of Investment in High Speed Rail, 18–19 December 2013,
APTA, American Public Transportation Association, Washington New Delhi, International Transport Forum, Paris
DC 56. Cascetta E, Coppola P (2011) High speed rail demand: empirical
33. UIC (2005) High speed rail’s leading asset for customers and and modelling evidence from Italy. European Transport Confer-
society. UIC Publications, International Union of Railways, Paris ence, Glasgow, p 18
34. de Rus G (ed) (2009) Economic analysis of high speed rail in 57. Tomii N (2010) How the punctuality of the Shinkansen has been
Europe, Informes 2009. Economı́a y Sociedad, Fundación achieved. WIT Trans Built Environ 114:111–120
BBVA, Bilbao 58. Janić M, Vleugel J (2012) Estimating potential reductions in
35. Pourreza S (2011) Economic analysis of high speed rail. NTNU externalities from rail-road substitution in trans-European freight
(Norwegian University of Science and Technology), Trondheim transport corridors. Transp Res D 17:154–160

123 J. Mod. Transport. (2016) 24(1):1–21

A multidimensional examination of performances of HSR (High-Speed Rail) systems 21

59. Gautier P-E, Letourneaux FP (2010) High speed trains external railjournal.com/index.php/high-speed/wenzhou-crash-report-
noise: a review of measurements and source models for the TGV blames-design-flaws-and-poor-management.html/
case up to 360 km/h. http://pdf-ebooks.org/ebooks/high-speed- 62. Qiao H (2012) Wenzhou crash report blames design flaws and
trains-pdf.html poor management. IRJ—Int J Railw, Falmouth, Cornwall. http://
60. NDTnet (2000) ICE Train Accident in Eschede—Recent News www.railjournal.com/index.php/high-speed/wenzhou-crash-report-
Summary. e-J Nondestruct Test Ultrason 5(2). http://www.ndt. blames-design-flaws-and-poor-management.html/
net/news/2000/eschedec.htm/
61. Puente F (2014) Driver error ‘‘Only Cause’’ of Santiago accident,
says report, IRJ—Int J Railw, Falmouth, Cornwall. http://www.

J. Mod. Transport. (2016) 24(1):1–21 123