OPERATIONAL AND COST MODELS FOR HIGH-SPEED
RAIL AND MAGLEV SYSTEMS
FADI EMIL NASSAR
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
To my parents,
Emil and Maggie
I would like to extend my deep gratitude to Dr. Fazil
Najafi, my supervisory committee chairman, for his guidance
and encouragement. Without his continuous support this study
would not have been possible. Sincere appreciation goes to Dr.
Paul Thompson, Dr. Ralph Ellis, Dr. Joseph Wattleworth, and
Dr. Sharma Chakravarthy for their suggestions and comments and
for serving on my supervisory committee.
Additionally, I like to thank Mr. Fritz Polifka, managing
director of Transrapid testing facilities in Germany; Mr. Rolf
Hellinger of the German National Railways (DB); Mr. Gabriel
Courbey of the French National Railways (SNCF); Mr. Joseph
Silien, a representative of ABB of Sweden; Mr. John Harrison
of Parsons Brinckerhoff Quade & Douglas; Mr. Nazih Haddad and
Mr. Jack Heiss of the Florida Department of Transportation;
Mr. Ron Mauri, Mr. Pete Montague and Mr. Jim Milner of the
United States Department of Transportation; and Mr. Robert
Casey of the High Speed Rail/Maglev Association. The technical
assistance provided by these individuals was very valuable.
Finally, much gratitude is owed to my parents, Emil and
Maggie, for their love and support during the entire course of
my study; and to D6rte Sage for adding colors to my life.
TABLE OF CONTENTS
LIST OF TABLES
LIST OF FIGURES ........................
1 INTRODUCTION ......................
General Background ................
Problem Statement .................
Study Objectives ..................
Research Methodology ..............
2 HIGH-SPEED RAIL AND MAGLEV SYSTEMS
The Japanese Bullet Train .........
The French TGV ....................
The German ICE ....................
The ABB X-2000 Tilt Train .........
The German Transrapid Maglev System
Japanese HSST Maglev System .......
The Japanese MLU Maglev System ....
The American Maglev Systems .......
3 COMPARISON OF HSGT SYSTEMS ........
Conventional and High-Speed Trains
Maglev and High-Speed Rail ........
EMS and EDS Maglev Technologies ...
Tilt and Nontilt Trains ...........
4 REVIEW OF HSGT LITERATURE .......................
Argonne National Laboratory Maglev Study ........
Carnegie Mellon University Study ................
USDOT Safety Reports ............................
TRB Safety Reports .............................. 77
TRB Special Report 233 ................. ......... 78
Supertrains Book ................................ 80
5 TRAIN DYNAMICS ................................... 84
Resistive Forces Acting on HSR Trains ........... 85
Resistive Forces Acting on Maglev Trains ........ 90
HSR and Maglev Tractive Power ................... 92
Acceleration on Level Terrain ................... 94
Acceleration on Grades ...... ......... ............. 101
Allowable Speed on Curves ....................... 106
Transitional Spiral Curves ...................... 111
Vertical Curves ................................. 112
6 OPERATIONAL SIMULATION MODEL .................... 116
Classification of Simulation Models ............. 116
Purpose of the Operational Simulation Model ..... 118
Inputs to the Model ............................. 119
Structure of the Model .......................... 121
Collection of Technical Data .................... 133
Model Validation ................................ 142
Study Cases ..................................... 146
7 COST STRUCTURE OF HSR AND MAGLEV PROJECTS ....... 167
Capital Cost of HSR Systems ..................... 168
Operation and Maintenance Costs of HSR Systems .. 190
Capital Cost of Maglev Systems .................. 194
Operation and Maintenance Costs of Maglev Systems. 206
Capital Costs of Implemented HSR Systems ........ 208
Capital Cost Estimates of U.S. Corridor Studies 210
Estimates of Operating and Maintenance Costs .... 217
Development of a Preliminary Cost Database ...... 217
8 CAPITAL COST ESTIMATION PROCEDURE ............... 219
Limitations of the Traditional Method ........... 219
Advantages of a Database Software ............... 220
Proposed vs. Traditional Cost Estimation Methods 224
Structure of the Cost Estimation Method ......... 227
9 CONCLUSIONS AND RECOMMENDATIONS ................. 237
Conclusions ..................................... 237
Recommendations ................................. 243
REFERENCES ........................................... 245
A LISTING OF THE OPERATIONAL SIMULATION MODEL
WRITTEN IN FORTRAN ............................ 254
B LISTING OF THE OPERATIONAL SIMULATION MODEL
INPUT MENUS WRITTEN IN BASIC .................. 279
C ROUTE ALIGNMENT DATA OF THE TAMPA-ORLANDO
CORRIDOR OBTAINED FROM FDOT AND ROUTE INPUT
FILES FOR THE SIMULATION MODEL ................ 286
D SAMPLE SPEED PROFILE OUTPUT OF THE OPERATION
SIMULATION MODEL .............................. 297
E TECHNICAL DATA RELATED TO THE TRANSRAPID MAGLEV
SYSTEM OBTAINED FROM THE MANAGING DIRECTOR OF
TRANSRAPID TESTING FACILITY IN GERMANY ........ 304
F TECHNICAL DATA RELATED TO THE TGV HIGH-SPEED
RAIL SYSTEM OBTAINED FROM THE FRENCH NATIONAL
RAILWAYS (SNCF) ............................... 316
G TECHNICAL DATA RELATED TO THE ICE HIGH-SPEED
RAIL SYSTEM OBTAINED FROM THE GERMAN NATIONAL
RAILWAYS (DB) ................................. 326
H TECHNICAL DATA RELATED TO THE X-2000 HIGH-SPEED
RAIL SYSTEM OBTAINED FROM A U.S. REPRESENTATIVE
OF THE SWEDISH ABB TRACTION COMPANY ........... 331
I MAGLEV MODULE OF THE COST MODEL WRITTEN IN
DBASE IV ...................................... 342
J PRELIMINARY UNIT COST DATABASE COMPILED FROM
VARIOUS STUDIES ............................... 363
K RELATED CORRESPONDENCE .......................... 370
BIOGRAPHICAL SKETCH .................................. 372
LIST OF TABLES
4-1 Cost of Vehicle Congestion to the Highway
User ...................................... 67
4-2 Cost of Air Congestion ....................... 67
4-3 Probability Estimates of Undesired Events
in Maglev Operation ......................... 73
4-4 Maglev Risk Assessment Estimates ............. 74
5-1 High-Speed Rail and Maglev Comparison Matrix .. 96
5-2 Relationship Between Lateral Acceleration and
Unbalanced Superelevation for HSR Systems ... 114
5-3 Relationship Between Lateral Acceleration and
Unbalanced Superelevation for Transrapid
Maglev Systems ............................. 115
6-1 Comfort Parameters for Various HSR and maglev
Technologies ............................... 144
6-2 Simulation Results of Cases A Through D ...... 165
7-1 Escalation Cost Factors to January 1993 ...... 169
7-2 Capital Cost of the FHSRC Proposal Submitted
to the State of Florida .................... 172
7-3 Electrification Cost per Mile for Two Tracks
of a High-Speed Rail System ................ 180
7-4 Power Substation Cost per Mile ............... 181
7-5 Signaling Cost per Mile per Single Track ..... 182
7-6 Unit Costs of Railroad and Roadway Bridges
for the X-2000 Tilt Train .................. 186
7-7 Cost Structure of a Maglev Guideway .......... 200
7-8 Maglev Capital Cost Items of the California-
Nevada System .............................. 204
7-9 Cost Estimates of Maglev Infrastructure Based
on Corridor Study Reports .................. 216
LIST OF FIGURES
1-1 Proposed European High-Speed Rail Network
for the Year 2015 ........................... 3
1-2 The High-Speed Rail Planned Lines in South
Korea and Taiwan ............................ 4
1-3 Projected Total Aircraft Delay for Major
Airport by 1996 ............................. 7
1-4 Top 50 Air Traffic Routes Under 600 Miles .... 8
2-1 Illustrations of HSR and Maglev Systems ...... 24
2-2 TGV-SE Distribution of 1988 Operating
Revenues .................................... 31
2-3 X-2000 Tilt Body and Self-Steering Trucks .... 38
2-4 Transrapid Support and Guidance Systems ...... 42
2-5 Japanese MLU002 Maglev Vehicle and Guideway .. 47
2-6 Magnetplane Vehicle and Guideway ............. 51
3-1 Attractive EMS and Repulsive EDS Levitation
Modes ...................................... .. 57
4-1 Conceptual Plan for Connecting Hub Airports
With Maglev Systems ........................ 64
5-1 Resistive Forces Acting on Transrapid ........ 93
5-2 Tractive Effort and Total Resistance of TGV .. 97
5-3 Train Acceleration on Level Terrain and
Grades ...................................... 99
5-4 Forces Acting on a Train Travelling on a
Curve ..................................... 102
5-5 Car Body Tilt Angle of a Tilting Train ....... 109
6-1 Sample Route Data Input File ................. 123
6-2 First and Second Input Data Screens of the
Operational Simulation Model ............... 124
6-3 Flowchart of the Operational Simulation Model 125
6-4 Simulation Process Using Deceleration Curves
and the Look-ahead Distance ................ 129
6-5 Backward Safe-Speed Curves ................... 131
6-6 Sample Output of the Operational Simulation
Model ..................................... 134
6-7 Tractive Effort Curve for the TGV-A Trainset .. 136
6-8 Computation of Acceleration Rates from Speed-
Distance and Speed-Time Profiles ........... 139
6-9 Transrapid Energy Consumption Rates vs. Speed
for Four Trainsets ......................... 141
6-10 Noise Emission Levels of Various
Transportation Systems ..................... 143
6-11 Transrapid Run Case A ........................ 148
6-12 TGV Run Case A ............................... 149
6-13 ICE Run Case A ............................... 150
6-14 X-2000 Run Case A ............................ 151
6-15 Transrapid Run Case B ........................ 152
6-16 TGV Run Case B ............................... 153
6-17 ICE Run Case B ............................... 154
6-18 X-2000 Run Case B ............................ 155
6-19 Transrapid Run Case C ...... ................... 156
6-20 TGV Run Case C ............................... 157
6-21 ICE Run Case C ............................... 158
6-22 X-2000 Run Case C ............................
6-23 Transrapid Run Case D ........................
6-24 TGV Run Case D ...............................
6-25 ICE Run Case D ...............................
6-26 X-2000 Run Case D ............................
6-27 All System Runs for Case D ...................
7-1 Cross-Section of TGV Track ...................
7-2 TGV Fasteners and Concrete Block .............
7-3 Embankments and Structures to Reduce TGV
7-4 TGV Track Block Sections and Automatic
7-5 ICE Track, Elevated Structure and Roadway
7-6 Single and Double-Track Tunnels for TGV ......
7-7 Cost Distribution of the California-Nevada
Maglev Project .............................
7-8 Cross-Section of Transrapid Guideway and
7-9 Typical Transrapid Roadway Crossing ..........
7-10 Capital Cost Components of the TGV Atlantic
7-11 HSR and Maglev Corridor Studies in North
7-12 Capital Cost Components of the Florida TGV
Proposed System ............................
7-13 HSR and Maglev Cost Components for a
Pennsylvania System ........................
8-1 E-R Diagram for a Hypothetical HSR
Construction Project .......................
8-2 Translating the E-R Diagram into Relations ... 223
8-3 Cost Summary of Main Items per Route Section .. 225
8-4 Overall Structure of the Cost Estimation
Procedure .................................. 228
8-5 Code-Generating Menu for Maglev Capital
Cost Items ................................. 229
8-6 Code System Used to Locate Items ............. 232
8-7 An Illustration of a Code File ............... 234
9-1 Energy Consumption Rates of Various Modes .... 240
9-2 Profiles of the Old and New Paris-Lyon
Rail Lines ................................. 242
Crushed rock that supports the rails and
distribute the train's weight.
Track is divided into sections for signaling
Overhead contact wire and the supporting
structure to supply electric current to
Another word for a train.
A tunneling technique in which the track in
built in a deep cutting, roofed with a
reinforced concrete structure, and then
covered with minimum amount of soil to restore
Braking in which the traction motors are used
as generators and convert kinetic energy into
electric energy. It does not involve friction.
Braking system based on magnetic attraction.
It does not involve friction.
Electrodynamic levitation system for maglev.
Electromagnetic levitation system for maglev.
Federal Aviation Administration.
Florida Department of Transportation.
Federal Railroad Administration.
High-speed ground transportation system.
Intercity Express, a German high-speed rail
Magnetically levitated train.
The current collecting device located on the
roof of a train.
French National Railways.
United States Department of Transportation.
The fleet of trains.
A French high-speed rail system.
The transverse part of the track
infrastructure to which the rails are
attached. Also called sleeper.
A train consisting of a fixed formation of
several passenger sections (usually
articulated) and a power car at one or both
ends. It is handled as one unit.
A German EMS maglev system.
The structure housing the axles and
suspensions and upon which rests the car body.
Sometimes referred to as bogie.
A Swedish high-speed rail system with a cab
body tilt mechanism.
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
OPERATIONAL AND COST MODELS FOR HIGH-SPEED
RAIL AND MAGLEV SYSTEMS
Fadi Emil Nassar
Chairman: Dr. Fazil Najafi
Major Department: Civil Engineering
High speed ground transportation (HSGT) has been
recognized by European and Southeast Asian countries as a
necessary component of a balanced transportation policy. The
governments of these countries took the lead in promoting and
funding HSGT systems. In the United States, it was hoped that
the private sector will undertake the development and
financing of HSGT systems. Such hopes, however, did not
materialize because none of the proposed projects was
implemented. Currently, there is a renewed public and private
effort to reexamine the merits of high-speed rail (HSR) and
magnetically levitated (maglev) systems.
Both HSR and maglev systems offer comfortable and safe
ground transportation at high speed. These systems, however,
differ significantly in technology, cost structure, and
operational and cost sensitivity to design changes.
The dissertation objectives are to describe the
characteristics of the various systems operational today,
develop an operational simulation model, study the cost
structure of HSR and maglev projects, and formulate a
simplified cost procedure for quick and efficient preliminary
feasibility analysis of potential corridors.
The operational model simulates the operation of HSGT
systems, taking into consideration systems' performance
characteristics, comfort criteria, and route alignment.
Default performance parameters for 14 systems were obtained
and included in the model. New systems can be added easily.
The simplified cost procedure relies on a database
software. The procedure's overall structure is described and
illustrated with a model written in dBase IV. The model is
interactive. Menus are provided for efficient data input. Cost
estimation is performed in two steps: first, menus are used to
generate a file containing the codes of the project's main
items for each segment of the route; second, the model
interacts with a separate unit cost database and replaces item
codes with corresponding unit costs. Finally, menus allow the
user to generate various types of cost summaries. The cost
model supports programs that can be used to automatically
capture input errors, upgrade unit costs, and perform cost
adjustments to reflect job condition, soil type, location and
High-speed ground transportation (HSGT) systems have been
recognized by European and Southeast Asian countries as an
important component of a balanced transportation policy. The
governments of these countries took the lead in investing
directly or indirectly in the development and/or
implementation of HSGT systems. Japan and France were the
first countries to start the HSGT race, motivated primarily by
energy conservation goals and to reduce their dependency on
imported oil. The technical and commercial success of the
early Japanese and French high-speed lines led to a growing
worldwide interest in this new form of ground transportation.
High-speed rail (HSR) and magnetically levitated trains
(maglev) are the two modal components of HSGT. The HSR is an
advanced conventional steel wheel and steel rail system
engineered for speeds of 240 to 350 km/h (150 to 220 mph).
Three decades of revenue service have established HSR as the
safest and most energy-efficient transportation mode.
Nevertheless, maglev technology is being advanced as an even
safer and more efficient transportation mode. Maglev trains
generate lower noise and vibration levels than HSR systems.
They are also capable of climbing steeper grades and
consequently cause little disruption to the environment.
Maglev systems are the fastest and most technologically
advanced ground transportation mode. They rely on magnetic
forces to support the train above the guideway, providing for
contactless operation that permits speeds of up to 500 km/h
(310 mph). Electromagnetic forces provide levitation,
guidance, propulsion and braking (Johnson et al., 1989).
Nations all around the world are expressing interest in
HSGT systems. The European Community has proposed an ambitious
master plan to link most major West European cities with an
HSGT service. It is estimated that by the year 2015, the
European HSGT network will consist of over 28,800 km (18,000
mi) of routes including 18,880 km (11,800 mi) of new lines or
lines upgraded for speeds of more than 250 km/h (156 mph)
(Vranich, 1991). The proposed European HSR network is
illustrated in Figure 1-1. Japan is continuously upgrading its
rail links and planning new HSR lines. It is also developing
two maglev systems using conventional and superconducting
magnetic technologies. South Korea is designing an HSR system
between Seoul and Pusan. The line is scheduled for completion
in 1998 at a cost exceeding $8.3 billion. Taiwan is planning
a 320 km (200 mi) HSR system between Taipei and Kaohsiung at
a cost estimated between $12 billion and $15 billion (Vranich,
1991). The planned HSR lines in South Korea and Taiwan are
shown in Figure 1-2. Australia is considering an HSGT link
Figure 1-2: The HSR Planned Lines in South Korea and Taiwan
Source: ICE, 1992.
between Sydney and Melbourne. China has announced a joint
venture with Japanese firms to plan a 280 km (175 mi)
high-speed train connecting Fuzhou with Xiamen by 1997
(Speedlines, April 1993). France and Great Britain will soon
be linked by the Eurotunnel, an all-weather HSR system
operating in a tunnel dug under the English Channel. The
Eurotunnel is the largest, privately funded civil engineering
project in the world. The consortium in charge of the project
includes over ten major international construction companies.
The project management is assumed by Bechtel International
In the United States, interest in HSR and maglev systems
is growing as a response to air and highway congestion. The
Department of Transportation (USDOT) predicts a continuous
increase in travel demand, further straining the already
congested highway facilities; and the Federal Aviation
Administration (FAA) estimates an even larger increase in air
travel. According to the American Association of State Highway
and Transportation Officials (AASHTO), surface travel demand
is expected to at least double by the year 2020 (Vranich,
1991). The House of Representative's bill H.R. 1087 (U.S.
House of Representatives, 1991) states that according to
USDOT, over two billion hours are lost annually because of
highway congestion, costing the economy $80 billion. A study
conducted at the Argonne National Laboratory (Johnson et al.,
1989) for the U.S. Department of Commerce states that in 1986,
the cost of all air traffic delays amounted to nearly $5
billion. The H.R. 1087 bill estimates an annual loss of $8
billion due to airport congestion by 1987. The Argonne study
also states that 22 airports are expected to exceed three
million hours of annual passenger delay by 1996. Figure 1-3
shows the projected total aircraft delay at major airports.
The ability to expand highway capacity on congested
corridors has frequently been hampered by high costs and
limited rights-of-way. Proposals to build new airports or
expand existing ones have faced strong community opposition
(TRB, 1991). In corridors where highway and airport expansions
are prohibitively expensive or infeasible, an HSGT link could
provide needed capacity to insure adequate mobility to the
traveling public. One HSR line can carry as many passengers as
six lanes of highway, depending on operation frequency.
Furthermore, HSR and maglev systems are potentially attractive
competitors to air travel for distances between 150 km and 950
km (about 100 mi and 600 mi). They offer travelers the
advantages of frequent departures, comfort, economy, safety
and reliability in all weather conditions. While airports are
built outside urban areas, train stations are usually located
near the center of the city, thus significantly reducing the
overall travel time (i.e., perceived trip time) of short
trips. The Argonne study (Johnson et al., 1989) estimates that
between 38% and 50% of airline trips consist of distances of
less than 960 km (600 mi) such as those shown in Figure 1-4.
3 Washington, D.C.
Salt Lake City
////// New York (JFK)
///////// New York (LGA)
/////////////////A San Francisco
/////////////////A St. Louis
3 20 40 60 80 100 120 140 160 180
Aircraft Delay (x1000 h)
Figure 1-3: Projected Total Aircraft Delay For Major Airports by 1996
Source: Price, 1987
Although both HSR and maglev systems offer comfortable
and safe ground transportation at high speed, they differ
significantly in performance, cost structure, maintenance
requirements, noise emission and cost sensitivity to design
changes. Maglev systems are typically elevated. HSR systems
operate mostly at grade. As a result, the incremental costs
for bridges, overpasses, and other special structures vary
considerably between these two modes.
Maglev systems are more expensive than HSR systems, but
they are capable of higher speeds and faster acceleration
rates. Traveling at high-speed requires not only fast trains,
but more importantly a track alignment with gentle horizontal
and vertical curves to provide a comfortable ride. Curves
cause lateral and vertical accelerations that reduce passenger
comfort. Furthermore, operation at high-speed requires
compatibility among track, equipment, sensors, communication
and signaling systems (TRB, 1991). Unless track quality, route
alignment, and signaling systems permit extended travel at
maximum speed, the improvement in travel time will be marginal
regardless of the technology used.
Travel time is the most important operational parameter
in a corridor study. It is also a key factor in predicting
ridership. Typically, travel time has been estimated by state
officials in two extreme ways: the first is by conducting a
detailed feasibility analysis, usually by hiring consultants;
the second is by performing some simplistic calculations based
on the travel distance and the train's top speed. Detailed
corridor studies are time consuming and costly. Furthermore,
these completed studies have little value when the considered
HSGT technologies are upgraded, or when changes in route
alignment or station locations are desired. On the other hand,
simplistic calculations based on the system's top speed and
the estimated travel distance can be misleading. The reason is
that the proposed HSGT alignments usually follow interstate
rights-of-way or existing railroad tracks in order to minimize
environmental and social impacts. However, the horizontal and
vertical curves of the interstate system and existing railroad
tracks are designed for relatively low speeds. This can
severely limit the train's allowable speed and significantly
increase travel time beyond what is acceptable to justify the
high capital cost of an HSGT system. Therefore, alignment
improvements are necessary on existing tracks or highway
rights-of-way to accommodate HSGT systems. These improvements,
however, must be carefully selected based on cost and
operational impact analysis.
Preliminary cost estimates of HSR or maglev corridors
have necessitated detailed and costly studies. This is
because the topography and conditions of each corridor are
unique. Past studies have shown a wide cost disparity between
proposed systems, especially for maglev projects (as described
in Chapter 7). There is a need to develop a simplified cost
estimation technique that is not as detailed or costly as a
full corridor studies, but that provides much better cost
estimates than a comparison with the cost per mile of other
projects. The cost estimation method should be based on an
understanding of all the main cost components of an HSR or
maglev system and should be designed to interact with a
separate cost database of main items. The cost database should
be easily maintained and upgraded. Several steps are required
to develop this simplified cost model and the unit cost
database of main items. In Chapter 8, the overall structure of
the model is outlined and a sample program is developed to
illustrate the model.
In summary, the selection of the appropriate HSGT system
and the optimal route alignment requires the evaluation of a
large number of alternate route/system configurations. This
type of analysis is most efficiently performed with the help
of computer-based techniques.
The objectives of this study are to examine the
characteristics of existing HSR and maglev systems and to
develop a simplified microcomputer-based analysis technique to
perform fast and low-cost preliminary HSGT corridor
evaluations. As previously mentioned, a preliminary study of
an HSR or maglev corridor requires numerous simulation runs to
select the appropriate HSGT technology and the most effective
route alignment that balances cost and performance. Today's
powerful microcomputers provide the ability to run
sufficiently accurate simulations of train operation for
varying route conditions. Furthermore, the advancement in
microcomputer database software makes it possible to develop
a menu-controlled cost model that allows rapid preliminary
cost estimates of entire projects as well as selective design
changes. The ability to consider numerous design alternatives
requires that both the cost and simulation models be simple to
use, menu controlled, and designed to easily incorporate new
systems and updated cost data.
The development of a simplified computer-based analysis
technique requires the following steps:
Studying train movement dynamics and analyzing the
characteristics of the various HSR and maglev systems
Developing an operational simulation model that can
quickly and accurately simulate an HSR or maglev train
movement along any route. The simulation should take into
consideration the performance characteristics of the
system, track alignment, station locations and
constraints such as allowable lateral acceleration and
Developing the overall structure of a computerized cost
model that interacts with a separate cost database to
produce quick preliminary cost estimates of HSR and
The objective of the study is not to develop a detailed
unit cost item database of main HSGT construction items. This
is currently not possible in the United States because no HSR
or maglev system has been built in North America, and foreign
experience cannot be easily nor accurately translated. The
objective is instead to develop the overall structure of a
cost estimation procedure and illustrate its operation with a
sample program. In addition, a preliminary cost database was
developed by adapting cost estimates from previous HSR and
maglev corridor studies conducted in North America.
The cost estimation procedure has the advantage of
relying on a database software. Preliminary HSGT cost
estimates have typically used spreadsheet programs (Schmelz,
1989). Spreadsheet programs are easy to use. They have,
however, inherent limitations that can restrict a quick
evaluation and representation of the costs involved in a
project. A spreadsheet program presents a "two-dimensional"
layout of a project's cost items. A database program is
capable of developing a "three-dimensional" view of the
project's cost items by storing cost information of each route
segment in a different record of the file. A database software
offers additional advantages such as the ability to develop
menus for quick interaction and easy input of data, the
capability to incorporate programs written in a host language
to perform specific tasks, and the powerful query capability
possible by creating relational databases.
The main feature of the cost model is its ability to
automatically integrate new and updated cost databases as long
as the coding system is respected. While item costs vary and
a cost database needs to be updated regularly, HSR and maglev
projects generally have a consistent cost structure. The cost
model is based on this cost structure and is developed by
studying the main components of typical HSR and maglev
projects. Future improved unit cost databases of main items
will enable the model to produce more accurate estimates
without modifying its internal structure.
The study was completed according to the following main
Task 1: Review of HSGT systems (Chapter 2)
The literature was reviewed to study the characteristics
of the various HSR and maglev systems. The review emphasis was
on operational systems or systems certified as operationally
ready. Few academic books on the subject are available. The
information was mainly obtained directly from the French,
German and Swedish HSGT manufacturers; and from reviewing
newsletters, scientific papers, periodicals and USDOT reports.
The most advanced HSR and maglev systems were identified, and
a description of each system is presented.
Task 2: Comparison of HSGT Technologies (Chapter 3)
The information gathered in Task 1 was useful to perform
a systematic comparison of the various HSGT technologies. This
includes a review of the differences between HSR and
conventional rail, a comparison of HSR and maglev systems, an
assessment of the attractive and repulsive maglev
technologies, and an examination of the advantages and
disadvantages of tilt trains. The reasons for the wide cost
disparity between the comparable French TGV (Train a Grande
Vitesse) lines and the German ICE (Intercity Express) lines
are also described.
Task 3: Review of HSGT Literature (Chapter 4)
The literature review pertaining to HSGT technology dealt
primarily with government reports and studies published by the
Transportation Research Board (TRB). The proposals to build
and operate HSGT systems in the U.S. were more focused on
ridership estimates and financial feasibility than on
technical analysis. Therefore, they were not reviewed in this
task. Nevertheless, these proposals were used in task 6 to
analyze the cost structure of HSGT projects and develop a
preliminary cost database.
Some of the studies reviewed are the Argonne National
Laboratory's report sponsored by the U.S. Department of
Commerce; the Carnegie Mellon's study and the TRB Special
Report, both sponsored by USDOT; the various safety reports
produced by USDOT and the Federal Railroad Administration
(FRA); studies performed by the Transportation Research Board
(TRB); transcripts of congressional hearings; and the book
titled Supertrains: Solutions to America's Gridlock written by
Joseph Vranich, the current president of the High Speed
Task 4: Study of Train Dynamics (Chapter 5)
The simulation of an HSGT train operation along a route
is deterministic because it is based on the laws of dynamics.
The performance of an HSR or maglev train is governed by
mechanical limitations and comfort limits. The train operation
depends on the system's acceleration and deceleration rates,
maximum speed, acceleration on grades, and allowable speed on
vertical and horizontal curves. A train acceleration is a
function of its tractive force minus the sum of all resistive
forces. The tractive force is based on the engine's rated
power. Resistive forces are estimated by various mechanical
and aerodynamic formulas developed empirically or through
field tests. Acceleration decreases with speed due to air
resistance, which becomes the dominant resistive force at
higher velocities. Grades reduce acceleration because they add
a new resistive force equal to the component of the weight
vector parallel to the ground.
A train traveling around a curve is subject to a
centrifugal force acting outward. This force can be partially
balanced by track superelevation. Higher speeds or sharper
curves cause an excess lateral force that causes lateral
acceleration and is sometime referred to as unbalanced
superelevation. Similarly, vertical curves cause upward and
downward accelerations. The allowable lateral and vertical
accelerations are governed by passenger comfort criteria and
In addition to curves and grades, the simulation process
takes into consideration station locations and externally
imposed speed restrictions. A train must respond to changes in
route condition several miles before it reaches them. This is
achieved by developing at each route section a deceleration
curve based on comfortable deceleration rates. Deceleration
curves are compared with a route's allowable speeds to
determine the proper time for the train to begin decelerating
so it will not exceed the route's speed limits.
Task 5: Development of the Operational Model (Chapter 6)
The purpose of the operational simulation model is to
simulate the movement of an HSR or maglev system along a route
and compute travel time and operating speed for each section
of the route. The model's inputs consist of route data, system
performance data, and comfort criteria such as banking angles
and allowable lateral accelerations. The model's outputs
include a speed profile and a summary table stored in an ASCII
file that can be imported by most spreadsheet programs to
produce graphical representations of results.
The program is written in FORTRAN because the simulation
process requires extensive computations with floating numbers
and FORTRAN is suitable for this kind of application. The
FORTRAN language, however, does not allow easy interface with
the user and is incapable of generating input menus.
Therefore, menus were created using the BASIC language to
provide a graphical and easy way to input data and capture
input errors. Because the FORTRAN simulation program uses all
the available random access memory (RAM) available to a
personal computer, the interface between the BASIC and FORTRAN
programs is insured through a batch file that causes these
programs to run sequentially.
A good deal of effort was spent on collecting performance
data from the various HSGT manufacturers, located mainly in
Europe. This effort was quite successful. Technical data
describing the newest TGV and Transrapid trains (fully tested
but not in revenue service yet) were obtained, as well as
technical data related to the ABB X-2000 tilt train that
started trial operation between Boston and New York in March
1993 and was tested in Florida in June 1993.
The technical data describing the performance
characteristics of the various technologies were obtained in
several formats. Some HSGT manufacturers provided speed-time
or speed-distance profiles, others provided charts or
equations of the tractive and resistive forces, and some
provided graphs of acceleration vs. speed. Several spreadsheet
programs were developed to translate these various types of
data into the uniform input format acceptable by the model.
The operational simulation model was tested on
hypothetical corridors and the simulated acceleration and
deceleration operations were compared to manufacturers' speed
profiles. A small spreadsheet program was also developed to
verify the model's simulation results. After the model's
results were validated, it was used to simulate the operation
of several HSR and maglev systems along the Orlando-Tampa
corridor. The route alignment data were obtained from the
Florida Department of Transportation (FDOT).
Task 6: Cost Structure of HSGT Projects (Chapter 7)
The various HSGT corridor studies performed in the U.S.
were reviewed to develop a preliminary cost database of main
items and to study the cost structure of typical HSR and
maglev projects. Some of the corridor studies reviewed are
The Florida TGV proposal (1988)
The Florida HSR Corporation proposal (1990)
The California-Nevada Transrapid study (1990)
The Texas High-Speed Rail study (1989)
The TRB Special Report 233 (1991)
Furthermore, several European corridor studies were
reviewed. These studies, however, were not used to generate
cost data for the U.S. market. It is important to note that
there are great uncertainties related to the cost of HSGT
systems in North America. No HSR or maglev system has been
built, and cost estimates vary greatly among feasibility
studies. In fact, even the same study produced different
estimates months apart. For instance, the Florida High-Speed
Rail Corporation reported in its July 1990 proposal to the
State of Florida an estimated total cost of $3.6 billion
($10.3 million per km or $11.1 million per mile) for a 517-km
(323-mi) HSR system. These draft cost estimates were reduced
to $2.5 billion ($5 million per km or $8.1 million per mile)
a few months later in October 1990, for a slightly revised
system. Similarly, the cost of the Orlando Maglev
Demonstration Project was estimated in 1991 to be about $450
million. In 1992, the estimated cost was raised to $600
million. There are reports today that the newest estimates
have further risen sharply by over 20% compared to the cost
estimates made in 1992.
Both the Orlando Maglev Demonstration Project and the
Texas TGV Project were approved for construction by Florida
and Texas authorities, respectively. Detailed cost estimates
were requested from the two companies in charge of these
projects. Understandably, this effort was not successful due
to the confidentiality surrounding both projects.
Task 7: Cost Estimation Procedure (Chapter 8)
The overall structure of a computerized procedure to
produce quick and efficient preliminary capital cost estimates
of HSGT projects is outlined. The procedure relies on a
database software to compute capital costs. A sample
microcomputer model was developed to illustrate the procedure.
The capital cost section of the sample model was developed
using dBase IV, which is one of the best known and most
feature-rich database software. A spreadsheet software was
used to write programs to estimate operating costs based on
the information provided in Chapter 7. Because most operating
costs can be considered uniformly distributed along the route,
there was no need for a segmental-type analysis and therefore
a database software was not necessary.
The capital cost model is based on a clustering method
and a coding system. Items are grouped into greater project
units. For example, an HSR track consists of rail, ballast,
subballast, ties, and fasteners; maglev electrification
includes power supply, substations, control power, electric
switches, and power feeders. The coding system is a
hierarchical classification method (tree structure), similar
to the method used by libraries for coding books. Each code
has several alphanumerical components. Each component is a
subgroup of the component to its left. Menus allow successive
selections of desired branches until the desired detail level
is reached. There is no limit to the database's level of
detail because the lowest nodes can be broken down still
further, and new menus added, when more detailed information
is available. The code for future, more specific items can be
developed by adding alphanumerical digits to right side of the
"parent" code, without modifying the initial digits that are
used by the model to locate and manipulate items.
The cost model is menu controlled. Layers of menus are
used to select the various components of the system and to
store this information in an efficient way. For example, a
guideway type that runs from section 800 to section 1,300 of
the route will be entered only once. The model will
automatically enter the guideway code for all 500 sections of
the route; and the guideway cost for each section will
subsequently be adjusted by internal programs to reflect the
work condition, soil type, and location of that section.
After all the project's capital items are selected, the
model generates a large file where each record (row)
represents a route segment and each field (column) represents
a cost item of that section. This generated file includes only
the alphanumerical codes of selected items. In the final step,
the model interacts with a cost database and generates a new
file by replacing each code with its corresponding value from
the cost database. The resulting file contains all the cost
items for all route sections. Prewritten commands are used to
produce all types of useful cost summaries.
Task 8: Conclusions and Recommendations (Chapter 9)
The simulation results are presented, as well as a
general comparison of the performance and cost of the various
HSR and maglev systems. Internal programs that can enhance the
operation of the cost model are outlined, and steps needed to
develop and maintain a national unit cost database are
described. Recommendations are formulated that call for the
development of a uniform coding system for the various items
and the need for a structured way to estimate and report item
costs in feasibility studies. A more detailed description of
item costs makes it easier to adjust them to other projects.
HIGH-SPEED RAIL AND MAGLEV SYSTEMS
This chapter provides a description of the various HSR
and maglev systems. The emphasis is on operational systems and
systems certified as operationally ready. Two of the
technologies considered are illustrated in Figure 2-1.
The Japanese Bullet Train
Japan was the first country to develop a high-speed
train, known as the Bullet Train or Shinkansen, as a way to
improve mobility between its densely populated cities. Japan,
similar to other Asian countries, has a very high population
density, a growing transportation demand, and already
saturated transportation facilities. The Japanese government
subsidized the development of the Bullet Train because it
responded to specific needs such as energy efficiency,
nonpolluting electric power, and high line capacity.
Furthermore, the construction of new HSR lines were seen as an
effective means of developing secondary urban areas for
residential and industrial relocation, thereby relieving the
demographic and industrial growth in major cities (Roth, 1990;
French TGV HSR
German Transrapid Maglev
Figure 2-1: Illustrations of HSR and Maglev Systems
Source: Petersen, 1985.
The first Bullet Train service was initiated in 1964. It
provided a link between Tokyo and Osaka. The top speed of the
first generation of Bullet Trains did not exceed 218 km/h (136
mph). The average speed was much lower because of route
alignment and in order to reduce energy consumption and noise
emission. In later years, lighter and more powerful Bullet
Trains in addition to improved alignment and extensive use of
noise barriers permitted operations at significantly higher
speeds. The travel time on the Tokyo-Osaka line was gradually
reduced from the initial four hours to about two and a half
hours, achieved with the new Super Hikari Bullet Train.
Bullet Trains have a modular structure, which means that
cars can be added or taken off a train. All cars are
self-propelled, and therefore they can be added to a train
without affecting the train's acceleration rates. Bullet
Trains run on conventional standard gauge track of 1.435 m
(56.5 in). Rails are continuously welded to eliminate the
"ticking" sound. The propulsion is based on electric power
supplied through an overhead catenary power distribution
system. The most recent Super Hikari, Series 300 Shinkansen,
is capable of a top speed of 270 km/h (168 mph).
All Bullet Train lines are grade-separated, built on
exclusive right-of-way. Japan's difficult terrain and limited
right-of-way have required the construction of numerous
tunnels and bridges. For instance, the 320-mile line between
Tokyo and Osaka has about 66 tunnels and over 3,100 bridges,
accounting for almost one-third of the distance; and the rail
link to the north island of Hokkaido required the construction
of a $5.6 billion 32-mile Seikan Tunnel and other special
Right after the Tokyo-Osaka Shinkansen line was opened,
it experienced a phenomenal increase in ridership. It is still
operating at a profit. Additional Bullet Train lines serving
less populated cities were gradually added to the rail
network. The operations of some of these newer lines are still
subsidized by the government. Japan has today over 1,300 miles
of high-speed tracks, and more lines are planned. The Bullet
Train lines carry over 125 million passengers annually. They
have a perfect safety record and their on-time arrival rate is
about 99% (Vranich, 1991).
The French TGV
The Paris-Lyon route is France's most important and
heavily travelled corridor. In the early 1960s, the
double-track rail line that served the two cities was near
saturation, but the demand was rising steadily. To provide
additional capacity, the French government decided to
construct an entirely new line, exclusively dedicated to
carrying passengers at high speed. The new line starts on the
outskirts of Paris and Lyon and utilizes existing lines and
stations within the cities. The Paris-Lyon line started
operation in 1981, and the full 439 km (273 mi) segment was
completed in September 1983. The French HSR technology, known
as the TGV (Train a Grande Vitesse), significantly outperforms
the Japanese Bullet Train and remains to date the fastest rail
system in commercial operation. The TGV reached a speed record
of 515 km/h (320 mph) on May 18, 1990.
Construction work on another line to serve the French
Atlantic seaboard from Brest to Hendaye started in 1985. The
TGV Atlantic line became fully operational in 1991. While the
Paris-Lyon line has no tunnels, the TGV Atlantic line has 7.6
km (4.75 mi) of tunnels and 3.5 km of viaducts (Ross, 1990;
Chambron, 1986). Additional lines crossing the French borders
to Belgium, Germany, and Great Britain through the Eurotunnel
are near completion. The TGV North will reduce the 2 hour 25
minute trip between Paris and Brussels to 1 hour 30 minutes,
and the 4 hour 50 minute trip between Paris and Cologne to 2
hours 50 minutes. The travel time from Paris to London will be
reduced by more than half, to about three hours. The rolling
stock for these new lines is based on the TGV-A. The expected
top operational speed will be increased to 320 km/h (200 mph).
The Eurotunnel trainsets, called the Transmanche Super Train
(TMST), will consist of 18 passenger cars constructed with
composite materials. Each TGV-TMST trainset is designed to
carry about 800 passengers.
The first TGV rolling stock developed for the Paris-Lyon
line is known as the TGV-SE (southeast). The TGV-SE operates
at a maximum commercial speed of 270 km/h (168 mph). Each
TGV-SE trainset is composed of 10 units and carries 386
passengers. A new generation of rolling stock known as the TGV
Atlantic or TGV-A was developed for the Atlantic line. The
TGV-A incorporates several technical advances, the most
important being the reliance of synchronous, self-commutating
traction motors. The three-phase synchronous motors are
lighter, more powerful, and more energy efficient than the DC
asynchronous motors used in the TGV-SE. They are also easier
to maintain because they do not contain a commutator, which
requires reprofiling roughly every 300,000 km (187,500 mi) (La
Vie du Rail, 1990). In a synchronous motor, the role of the
rotor and stator are reversed, and the rotor (field coil)
rotates at the same speed as the revolving field generated by
the stator (stationary part or armature). The synchronous
motor can also be used as a fail-safe electric brake by
reversing its operation to function as a generator.
Other improvements in the TGV-A rolling stock include a
more aerodynamic design, two additional passenger sections per
trainset providing a total seating capacity of 485, a novel
air-sprung suspension that provides an inter-car damper
mechanism, a microprocessor-controlled wheel slip prevention
mechanism, a microprocessor-based diagnostic system, an
improved pantograph design to collect current from the
catenary above, improved disk brakes, and fail-safe electric
brakes (eddy-current) that operate independently of the power
supply (Lacote, 1986). Only eight synchronous motors (maximum
power output is 1,100 KW [1475 HP] per motor and 8,800 KW
[11800 HP per trainset) are used in the TGV-A. The older,
smaller-size TGV-SE trainset is powered by 12 asynchronous
motors (maximum power output is 580 KW [778 HP] per motor and
6960 KW [9340 HP] per trainset). The TGV rolling stock has a
nonmodular trainset design. Each trainset is composed of a
fixed set of articulated passenger cars, propelled by one
power car at each end (push-pull operation). A fixed formation
trainset has the advantages of improved aerodynamics, lower
noise and vibration levels, and enhanced riding comfort
(Lacote, 1986). The TGV-SE trainset is composed of 10 units:
an articulated section of eight passengers cars (trailers) and
two power cars. The trainset configuration is described as
1P-8T-lP where "P" stands for power car and "T" for passenger
trailer. The TGV-A has a 10-section, articulated trailer and
two power cars. The trainset configuration is therefore
1P-10T-lP. The TGV-North and TGV-Texas have both a 1P-8T-1P
configuration, and the TGV-TMST designed to cross the
Eurotunnel has a 1P-18T-lP configuration. The TGV-Texas
trainset is based on the TGV-A technology, is capable of
operating at 320 km/h (200 mph), and has 317 passenger seats
The Paris-Lyon TGV line has attracted more ridership than
predicted. It is unanimously considered to be a huge technical
and commercial success as illustrated by the profit margin
shown in Figure 2-2. By 1990, the TGV-SE line had carried over
100 million passengers without any fatal accident. The line
continues to operate at a substantial profit. Revenues from
the TGV-SE line paid back the infrastructure cost in ten years
instead of the originally projected fifteen. All the new TGV
lines are expected to operate at profit, although their rate
of return is expected to be smaller than that of the TGV-SE
line. The TGV rolling stock has earned a reputation of
reliability, dependability, energy-efficiency, and safety. The
French National Railways' master plan calls for the building
of a 2,000-mile network of 14 new TGV lines by 2015 (La Vie du
The German ICE
Ten years after the TGV started operation on the Paris-
Lyon line, the Germans introduced the ICE train (Intercity
Express) in 1991. The first ICE operation was on newly
constructed Stuttgart-Mannheim and Hanover-Wurzburg lines.
These two lines and the Cologne-Frankfurt line (scheduled for
completion in 1996) will provide 800 km (500 mi) of new ICE
track. In addition, the German Federal Railway (DB) is in the
process of upgrading 3,200 km (2000 mi) of existing tracks for
ICE standards. The ICE trains operate at a top speed of 250
km/h (156 mph) on new lines and 200 km/h (125 mph) on upgraded
The ICE power car uses advanced three-phase asynchronous
motors designed without wearing parts. The ICE trains are
certified for maximum commercial speeds ranging from 250 km/h
to 300 km/h. The traction and braking systems of the ICE are
largely computer controlled. The train is capable of
regenerative braking that recovers energy during the braking
phase and returns it to the overhead power supply network.
European countries use different power supply environments for
their rail system. The ICE-M is designed to automatically
identify the voltage and frequency being provided by the
overhead wires and transform it to the required voltage.
The ICE passenger cars are manufactured with an aluminum
body shell that provides significant weight saving and reduced
dynamic loading. Aluminum is also employed in the major
components of the power car such as the transformers and the
gears housing. The ICE train has the lowest drag coefficient
of all HSR systems due to a streamlined design, smooth outside
surfaces, and flush-fit windows and doors. It is also quieter
than the TGV or Bullet Train because it is equipped with wheel
noise absorbers and selected components are covered with
noise-absorbing materials. The ICE cars are pressurized and
the inside pressure is regulated by the air conditioning
system (ICE, 1992).
Unlike the fixed formation of a TGV trainset, the ICE has
a modular design which provides the flexibility to adjust the
number of cars in a train to suit the demand. An ICE train has
a power unit on each end and can comprise any number of
passenger cars between 6 and 14. The ability to safely run
trains of various lengths requires sophisticated communication
and signaling systems. This is insured by heavy reliance on
computer chips and optoelectronic (fiber optics) data links.
The ICE is also equipped with a comprehensive computerized
diagnostic system that checks all the equipment throughout the
train. Each microprocessor subsystem possesses a
self-diagnostic capability, and communicates with a central
diagnostic unit that receives, analyses and stores all
relevant reports. Diagnostic reports are displayed on a screen
located in the power unit. Should a malfunction occur, the
diagnostic system identifies the cause and displays it on a
special screen in view of the driver, and simultaneously
recommend an action to be taken (ICE, 1991).
The ICE uses three braking systems designed to improve
passenger comfort and reduce maintenance cost. First, a
regenerative brake is applied, then an eddy-current brake
based on electromagnetic attraction that acts independently of
wheel/rail adhesion is used, and finally, a disk brake is
applied if necessary (Rosen, 1989).
The cost per mile of a new ICE line is two to three times
that of TGV lines. There are many reasons for this cost
disparity, the most important being the following:
The ICE uses slab tracks that allow mixed traffic with
priority freight trains. The TGV tracks are ballasted and
used for exclusive passenger use. The TGV weight is
limited to 17 tonnes per axle (18.7 tons per axle).
The signals and communications on the ICE lines must
accommodate mixed traffic and variable size ICE trains,
while only fixed TGV trainsets having exactly the same
number of cars per trainset operate on TGV lines.
The topography of the ICE lines is more difficult than
that of the TGV lines. Furthermore, because the ICE line
must accommodate freight trains with less powerful
tractive power, grades are limited to 1.25% as opposed to
3.5% for the TGV. As a consequence, the ICE lines
required more extensive earthwork, and numerous tunnels
and engineering structures.
German legal system requires public disclosures and
public hearings before plans for new rail lines can be
approved. This generated 360 lawsuits and 10,700
objections and resulted in years of delay for the ICE
lines. The French legal environment makes it easier to
acquire right-of-way for a new line. If a rail project is
considered by the French government to be "in the
national interest," objections to new lines are legally
unable to delay construction work.
The first ICE line started operation in 1991. It is too
early to develop a detailed assessment of the ICE system. All
indications, however, confirm that the ICE is a technical and
financial success. Ridership the first 100 days were 25%
higher than expected, and public support remains strong. The
reunification of Germany modified early rail expansion plans.
Current plans call for the completion of a new ICE line
between Hanover and Berlin by late 1997. Other new and
upgraded lines for the ICE system are also planned. German
rail experts anticipate a 30% ridership increase in the next
few years (ICE, 1992).
The ABB X-2000 Tilt Train
The X-2000 is a tilt train developed by Asea Brown Boveri
(ABB) and first used in commercial operation by the Swedish
State Railways (SJ). The purpose of developing the X-2000 was
to provide a train that can achieve higher speeds on existing
tracks, thus reducing the travel time without the need to
build a new line or extensively upgrade existing ones. The
X-2000 started operation in 1990, on a line 456 km (284 mi)
long between Stockholm and Gothenburg. The travel time was
reduced from four and a half hours to under three hours. SJ
plans to utilize the X-2000 in other Swedish corridors.
In November 1991, Amtrak signed a deal with ABB and SJ to
bring one spare X-2000 trainset to the Northeast Corridor. In
February 1993, a X-2000 trainset started a daily roundtrip
between New York and Washington; and two months later it began
operating between New Haven and Washington (Pope, 1993).
The X-2000 offers an innovative body tilting mechanism
and suspension system that enable the train to travel around
curves up to 25% faster than a nontilt train. This is possible
because the tilt mechanism partially balances the lateral
acceleration caused by traveling around a curve as explained
later in Chapter 5. Lateral acceleration reduces comfort and
must not exceed the allowable limits set by FRA.
Tilt control can be passive or active. A passive tilt
mechanism is caused directly by the unbalanced lateral forces
acting at the car body center of gravity. The Spanish Talgo
Pendular train is an example of this technology. The X-2000
employs a more complex active tilt mechanism. The advantages
of an active tilt is that the tilt operation can be controlled
and monitored. The tilt of passenger cars is mechanically
actuated and controlled by a microprocessor that activates
sequentially the tilt mechanism of each car to insure a
comfortable transition. A computer located in each car
monitors the tilt operation and performs redundant checks. The
active-tilt technology principles are also used by the
Canadian LRC and the Italian ETR train systems.
The active tilting mechanism on the X-2000 is adjusted to
cancel about 70% of the unbalanced lateral acceleration. It is
disabled at low speeds. The lateral acceleration is measured
by two sensors mounted in the center and lead truck and signal
the microprocessor to activate the tilt in passenger cars. The
power car is not equipped with a tilt mechanism to ensure a
good alignment between the pantograph and the catenary line.
In addition to active tilt, the X-2000 uses a flexible
suspension system that allows axles to move independently of
the rigid truck frame and keeps wheels aligned with the track
as illustrated in Figure 2-3. The flexible suspension reduces
the wheel rail forces and lateral-to-vertical force ratios
while negotiating curves (USDOT, 1991a).
The X-2000 trainset currently being tested in the U.S.
consists of one power car and five or six passenger cars. A
configuration with two power cars can have up to 12
intermediate passenger cars. Each power car carries four AC
asynchronous traction motors, capable of delivering 3200 KW
(4300 HP). The motors are light and designed for low
maintenance. The X-2000 top commercial speed is 240 km/h (150
The coach body is made of a steel frame. Modular
construction is utilized and components are replaced via an
opening in the roof. The floor is made out of wood and is
isolated from the steel frame to reduce noise. Side windows
are multilayer safety glass, designed to minimize the risk of
Coach Body Tilt System of the X-2000 Train
Self-Steering Trucks of the X-2000 Train
Conventional Steering Trucks
Figure 2-3: X-2000 Tilt Body and Self-Steering Trucks
Source: USDOT, 1991a.
injury during derailment. The X-2000 has three braking
systems: a disk brake, a regenerative dynamic brake, and an
The main advantage of the X-2000 is that it can operate
on existing tracks at higher speeds. This can be an
attractive option when right-of-way for new lines or improved
alignment is difficult to obtain or excessively expensive. The
X-2000, however, requires an electrified line to deliver the
required power. Its top speed on tangent sections is over 20%
lower than that of TGV or ICE.
The operation of the X-2000 in Sweden has been
successful. The trial runs on the Northeast corridor have just
began. They have, nonetheless, generated positive reactions
from the public and the media. Amtrak is planning to replace
its Metroliner fleet with 26 high-performance trains. A
decision has not yet been made on the technology, although the
X-2000 seems to be a good candidate. ABB has offered to build
the trains at its plant located in the state of New York.
Transportation officials in Michigan, Illinois, Southern
California and the Pacific Northwest have also expressed
interest in the X-2000 (Pope, 1993).
The German Transrapid Maglev System
Transrapid maglev is a new generation of ground
transportation systems. Although no maglev system has been
implemented in commercial operation anywhere in the world,
Transrapid has extensively been tested on a 31.5 km (20 mi)
single-track guideway built in the German Emsland region near
the Dutch borders.
In 1992, the German government certified the
Transrapid-07 system as operationally ready, and initially
considered implementing the Transrapid technology on a 50 km
(31 mi) line that links the airports of Bonn and Duesseldorf.
The German reunification, however, made the ICE Berlin-Hamburg
line a much higher priority, especially that Berlin replaced
Bonn as the capital of Germany. Other Transrapid feasibility
studies involved the Hamburg-Hannover and the Essen-Bonn
corridors, but no decisions were made in either case.
In the United States, the Transrapid system was proposed
to operate on a line between Las Vegas, Nevada and Anaheim,
California. In 1990, Bechtel submitted a proposal for a
franchise to build and operate the system. The financing of
the project was to be assumed by C. Itoh & Co., a major
Japanese company. The slowdown in California's economy,
however, caused the project to be canceled.
Privately financed Transrapid demonstration projects are
planned in both Orlando, Florida and Pittsburgh, Pennsylvania.
The engineering study phase is at an advanced stage in
Orlando. However, the ability to finance these projects with
private funds remains uncertain due to vast fluctuations in
capital cost estimates.
The development of the Transrapid technology was
sponsored by the German Ministry of Research and Technology.
Leading German industrial companies, the German National
Railways, and Lufthansa airlines participated in this effort.
Transrapid is a maglev train designed for speeds of 400
km/h to 500 km/h (250 mph to 310 mph) and capable of climbing
up to 10% grades. The system requires a completely dedicated
guideway that can be either elevated or near grade. Transrapid
vehicles levitate by attractive electromagnetic suspension
(EMS) and are propelled by linear induction. Transrapid
operation involves no contact between the guideway and the
vehicle; which causes less wear, less maintenance, and less
noise than HSR systems (U.S. House of Representatives, 1990).
Transrapid vehicles levitate by using the forces of
attraction between individually controlled electromagnets
arranged under the floor of the vehicle and ferromagnetic
rails (stator packages) installed under the guideway as
illustrated in Figure 2-4. A redundant electronic control
system insures a constant levitation gap of 8 mm (0.315 in).
Every section of the vehicle is equipped with 15 support
magnets and 13 guidance magnets (Meins and Ruoss, 1979).
- Coach body
M Secondary spring
( Guiding magnet
@ = Support magnet
0 Hinge point
Support magnet Stator pack with motor windings
Linear generator windings Guidance magnet
Eddy-current brake magnet (through brake) Support skids
T INKREFA sensor (vehicle location) Levitation bogies
Cabin suspension @ Pneumatic spring
Figure 2-4: Transrapid Support and Guidance Systems
Source: Meins and Ruoss, 1989.
Independent suspension is provided for each magnet. A
synchronous long-stator linear motor is used to propel and
brake the train. Thrust is controlled by changing the
intensity and frequency of the three phase current. Changing
the polarity of the electromagnetic field causes the motor to
function as a generator which brakes the vehicle without
physical contact (Ciessow et al., 1989).
All critical components are error tolerant and designed
with a high functional redundancy. Redundant systems are also
used for levitation and guidance (USDOT, 1991b). Operational
safety is enhanced by energizing only the section of the
guideway on which the train is traveling. Derailment is not
possible because the suspension system wraps around the
guideway. Transrapid signal and control systems are fully
automated. In an emergency situation, brakes can be actuated
from the central control unit.
The noise emission of Transrapid is significantly lower
than that of HSR systems. Transrapid consumes 30% less energy
than a high-speed train traveling at the same speed. Energy
consumption, however, rises rapidly at higher speeds because
of air resistance that increases with the square of the speed.
Transrapid vehicles are constructed with sandwich-type
light alloy plates. The vehicle's shell structure is made of
high-strength aluminum framing that provides high structural
stiffness. A glass-fiber plastic shell is used in the top part
of the vehicle noise and the cylindrical part of the coach
body. The side windows consist of two panes individually glued
to the coach body to improve sound and heat insulation. A
Transrapid trainset can consist of 2 to 10 sections. A 2-
section trainset has a capacity of 156 passengers, a
10-section trainset can carry a maximum of 1,060 passengers
(Gaede and Kunz, 1989).
Transrapid guideway can be made of either steel or
reinforced concrete. The guideway consists of sectional beams
of 25 m to 50 m (82 ft to 164 ft) long. Typical elevated
guideways are about 4.5-m (15 ft) high. The supporting
concrete columns tested on the Emsland track have either A or
The ability of Transrapid to climb up to 10% grades can
reduce the earthwork cost and the need for tunnels, bridges
and other special structures. On the other hand, the strict
construction tolerance adds significantly to the guideway
cost. For instance, guideway foundations in Florida can be
very expensive in a wetland or unstable-soil environment.
Impacts of temperature variations have also to be accounted
for within the tolerance limits of the guideway design.
The Transrapid system has only been operational on a test
track located in Germany and in international exhibitions.
There are no information available to evaluate the system in
revenue service operations.
Japanese HSST Maglev System
The HSST (High-Speed Surface Transportation) maglev
system is being developed by Japan airlines. HSST vehicles
levitate using attractive electromagnetic suspension (EMS)
similar to Transrapid. However, the propulsion system is
provided in the vehicle by linear induction motors instead of
the guideway, which makes the HSST less suitable for very high
The development of the HSST technology began in 1975. The
first prototype (HSST-1) was tested in 1977. The second
prototype (HSST-2) was introduced in 1978. It has eight seats
and was designed for a top speed of 100 km/h (62 mph). The
third prototype (HSST-3) has 48 seats. It was demonstrated at
the World Exhibition in Vancouver. The Vancouver track was
only 450 meter (1477 ft) long which allowed a maximum speed of
40 km/h (25 mph). The latest prototype (HSST-4) is designed
for a top speed of 200 km/h (125 mph) and accommodates 70
The HSST-3 achieved a reliability factor of 99.96% during
operation at an exposition in Japan (Vranich, 1991). The
HSST-4 was demonstrated in 1988 at an exposition near Tokyo.
A proposal to build a HSST system in Las Vegas has been
approved. If built, the HSST will use small and lightweight
vehicles. It will function as a downtown people mover and
operate at a moderate speed.
The HSST is being viewed as a fast short-distance
transportation system suitable for urban environments. It is
primarily marketed as a noiseless, low vibration people mover
system. It can operate on small turning radius and the
system's stations can be integrated with buildings and parking
The Japanese MLU Maglev System
The MLU maglev system is being developed by the Japanese
National Railways (JNR). The research begun in 1962. Ten years
later a prototype was demonstrated on a 480-m (1575-ft) long
test track. A research center and an test track were
subsequently built on the island of Kyushu near Miyazaki. The
test track was converted in 1979 to a U-shape guideway that
contains coils for propulsion and guidance in the sidewalls
and coils for suspension on the horizontal surface.
The MLU vehicles use electrodynamic suspension (EDS) for
levitation. Propulsion is provided by synchronous linear
motors. The vehicle is levitated by repulsive forces between
the magnetic field in the guideway and the superconducting
coils of the same polarity in the vehicle's underside as shown
in Figure 2-5. Superconducting dynamic levitation is
effective only at speeds above 100 km/h (62 mph). At lower
speeds, the vehicle is supported and guided by pneumatic
wheels that retract when the train's speed reaches 160 km/h
(100 mph). The vehicle levitates several inches above the
Ground coil for propulsion
Xam for apprat us
Grourd coil for suspension
Ground Coil for Propulsion and Guidance
`' Superconducting Coil
Ground Coil for Suspension
Figure 2-5: Japanese MLU002 Maglev Vehicle and Guideway
Source: Kyotani and Tanaka, 1986.
The MLU001 prototype has a modular structure. A three-car
MLU001 trainset has a mass of 33 tonnes (36.3 tons) and is 29
m (94 ft) long. Each car has eight superconducting coils used
for suspension, guidance and propulsion. The MLU001 with a
three-car configuration reached a speed of 350 km/h (218 mph)
on the Miyazaki test track.
The MLU002 prototype has six superconducting coils per
side. It has a mass of 19 tons, accommodates 44 passengers and
is designed for a top speed of 420 km/h (260 mph) (Johnson et
al., 1989). The MLU002 test runs caused component failures,
and strong electromagnetic fields were measured. This
prototype was damaged by fire in October 1991. The JNR,
however, has decided to build a new prototype, the MLU002N,
and a new 43-km (26.7-mi) long test track (Vranich, 1992; New
Technology Week, 1993). A commercial version of the MLU002 may
have as many as 14 cars per trainset. Nonetheless, it will
take years before all concerns are answered and the commercial
version is approved for revenue service.
The American Maglev Systems
Basic theoretical work on maglev concepts were conducted
at the Brookhaven National Laboratory in the early 1960s. Two
researchers of this laboratory, James Powell and Gordon Danby,
developed and patented the "null-flux" system that is capable
of high lift-to-drag ratio and stiff restoring forces. This
null-flux system consists of placing the vehicle magnets
between two guideway coils. No induced current is generated if
vehicle magnets are centered between the coils. Deviations
from the center position generate opposing currents that act
as a restoring force (Powell and Danby, 1969; Danby et al.,
1974). The maglev systems being developed in Japan have
adapted some of the features described by Powell and Danby. A
group formed by these researchers have recently proposed a
32-km (20-mi) long demonstration maglev system between Kennedy
Space Center and Port Canaveral (Speedlines, April 1993).
In late 1960s, the use of superconducting magnetic
suspension for a rocket operating in an evacuated tunnel was
examined by the Stanford Research Institute and Rockwell.
Subsequently, several technologies for maglev vehicle
levitation were examined by the Stanford Research Institute.
The EDS suspension concept was determined as the most
favorable (Barbee, 1968).
Ford Motor Company became also interested in maglev
technology (Coffey et al., 1969). A study conducted by Ford
concluded that attractive suspension provides a higher
lift-to-drag ratio than repulsive suspension. The attractive
system has a low magnetic drag, but also a low tolerance for
track misalignment (Johnson et al., 1989). A conceptual design
of an EDS maglev system that can carry 80 passengers was
developed, but no further detailed analysis was pursued.
Important theoretical research was conducted at the
Massachusetts Institute of Technology (MIT) during the 1970s.
Several patents resulted from this effort. A conceptual maglev
system called the Magnetplane was developed by two MIT
researchers: Henry Kolm and Richard Thornton. The innovative
feature of the Magnetplane is its self-banking ability due to
a 0.3-m (1-ft) air gap and a semicircular guideway design as
shown in Figure 2-6. Magnetplane was demonstrated with a
1/25th-scale model operating on a guideway 120 m (400 ft)
long. In 1975, the federal government terminated funding for
all HSGT research. This action stopped most of the maglev
research being conducted at MIT, Ford, Boeing and other
companies and laboratories. In 1991, as part of the
government's maglev initiative, maglev research was resumed at
MIT and other laboratories. The research focus is related to
system concept definition for vehicles, guideways, suspension
-.c ~ -' .c
RETRACTABLE LANDING WHEELS
Figure 2-6: Magnetplane Vehicle and Guideway
Source: Johnson et al., 1989.
- ----- I ---
COMPARISON OF HSGT SYSTEMS
The comparisons presented in this chapter are between
conventional and high-speed trains, High-speed trains and
maglev systems, EMS and EDS maglev levitation technologies,
and tilt and nontilt trains.
Conventional and High-Speed Trains
High-speed rail involves more than fast trains.
Additional technological advances are necessary. The main
characteristics that distinguish HSR from conventional rail
are described in the following paragraphs.
Diesel engines are heavy and incapable of high
acceleration rates (TRB, 1991). Diesel-electric and
turbine-electric motors that generate electricity on board
also add to train's mass and maintenance cost (they can,
however, be used as a temporary solution until line
electrification is completed.) The most appropriate traction
should draw current from a catenary system instead of
generating electricity on board. Synchronous and asynchronous
three-phase electric motors have the highest power to weight
ratio and are therefore the preferred engines for HSR systems.
The train's low weight is also important for braking
performance and for reducing dynamic loading on the rails.
Integrated Braking systems
High-speed trains are typically equipped with three
braking systems. The first is a regenerative braking feeding
power back to the electric line overhead, the second is an
eddy-current brake based on attractive electromagnetic forces,
the third is a set of disk brakes. The regenerative and
eddy-current brakes are used in most cases because they do not
involve physical contact and therefore reduce wear and
maintenance. Disk brakes are used mainly in emergency cases.
At high speeds air resistance becomes the dominant
resistive force. A streamlined shape and articulated passenger
cars are used to reduce drag and provide a stiff and stable
design. High-speed trains are also equipped with primary and
secondary suspensions to improve the ride quality. The
articulated design of passenger cars further reduces the
vibration of individual train sections. Noise absorbent
material is used in the car body and component design to
reduce interior noise. Passenger cars are usually pressurized
if the route includes tunnels. Amenities such as comfortable
seats, restaurant and conference coaches, large-view windows,
convenient luggage facilities, high-quality service,
convenient and automated reservation system, on-board
telephones, and computer links are provided to attract
business travelers away from airlines.
The impressive safety record of HSR systems in service
operation is due to a number of advanced safety features. Some
of these features are a fail-safe and automatic train control
(ATC) system that continuously monitors train operation and
automatically stops the train if the operator commits an error
or if an obstacle is detected on the track; advanced
signalization controlled by redundant processors and
detectors; automatic block signaling designed to maintain a
safe distance between trains to prevent collisions; and
electronic components with self-diagnose capability that are
programmed to automatically notify the operator and the
control center (via display and warning sound and light) of
defective or malfunctioning parts.
Although HSR systems use standard track gauge and are
capable of operating on existing tracks, they can only reach
their maximum speed on new tracks designed for high-speed
operation. As previously stated, the track should be equipped
with overhead catenary and advanced signalization. Track
alignment should be as straight as possible and completely
grade separated. Steeper grades, however, can be allowed
because the powerful electric motors and the stored kinetic
energy (function of speed) permit climbing up to five percent
grades. Conventional diesel-powered trains are generally
limited to one percent grades.
Maglev and High-Speed Rail
The primary advantage of maglev technology is that the
propulsion, suspension, and guidance of the vehicle are
achieved by magnetic forces without physical contact. Traction
derives from electromagnetic forces that are generated without
moving mechanical components. The contactless operation causes
less wear and noise, and lower frictional forces; making the
system suitable for very high speeds (Magnet Schnellbahn,
1991). Other important advantages of the maglev technology are
the lightweight aluminum or composite car body which reduces
weight and improves acceleration and climbing ability (train's
weight is also reduced because the stator wires are located on
the guideway instead of the vehicle); higher banking angles
allowing higher speeds on curves; advanced electronics used in
the guidance, communication, monitoring, braking, and
automated control modes; lower energy consumption than HSR at
comparable speeds; and better distribution of dynamic loading
over the guideway (vehicle weight is distributed over wide
magnets instead of the small contract area between HSR wheels
The primary disadvantage of maglev technology is that it
requires a costly and exclusive guideway, and therefore cannot
use existing tracks and existing stations in urban areas.
Providing a new access to urban areas can be very expensive.
While HSR systems can operate on conventional tracks at lower
speeds, maglev passengers are required to transfer to complete
their trip on existing rail lines. This is the main reason why
maglev systems are being proposed primarily to link nearby
airports (Johnson et al., 1989). Another disadvantage of the
maglev technology is that maglev systems have only been
operational on test tracks and in international expositions.
Furthermore, the EDS maglev system is still in the
developmental phase. Because no maglev system has been
implemented in revenue service, the infrastructure cost and
operational expenses cannot be estimated with any reasonable
degree of accuracy. Maglev trial runs on test tracks are
infrequent and at speeds lower than the maximum speed. The
reliability of maglev vehicles and infrastructures in a
full-schedule revenue service remains to be proven (USDOT,
EMS and EDS Maglev Technologies
Maglev vehicles are levitated either by attractive
electromagnetic suspension (EMS) or repulsive electrodynamic
suspension (EMS) as shown in Figure 3-1. The German Transrapid
and the Japanese HSST use the EMS technology. The Japanese
MLU and most of the proposed U.S. maglev systems use the EDS
technology. Transrapid and MLU propulsion is provide by linear
synchronous motors that include field coils in vehicles'
underside and active stator windings on the guideway.
Levitation of EMS vehicles derives from the attractive
forces between conventional electromagnets placed on the lower
portion of the vehicle and a ferromagnetic guideway.
Levitation of an EDS vehicle derives from the repulsive
forces between superconducting magnets placed on the vehicle
and induced electric currents known as eddy currents. Systems
using EMS can levitate at rest. Systems relying on EDS require
a certain speed before induced currents levitate the train.
According to a study sponsored by USDOT (Uher, 1990), the EMS
technology is several years ahead of the superconducting
technology pursued by Japan and the United States.
The EMS systems operate with an air gap of less than a
centimeter (0.4 in) because energy consumption increases with
the square of the gap. The small air gap, however, limits
maximum operational speeds to about 500 km/h (310 mph). The
EDS systems tolerate an air gap ten times larger than EMS
systems, allowing higher speeds and more flexibility in
Another advantage of the EDS technology is its inherent
stability. If the vehicle equilibrium position is disturbed by
high wind or passenger movement, the intensity of the
repulsive magnetic field varies and creates counteracting
forces that tend to restore the vehicle to its initial
position. On the other hand, the EMS technology is inherently
unstable. Disturbances in the equilibrium position tend to
generate forces that accentuate the unbalancing forces.
Therefore, the air gap of EMS systems must be monitored
continuously and continuous adjustments of the magnetic forces
are required (Johnson et al., 1989). The power required to
maintain the air gap is relatively small. Transrapid officials
claim that it is not greater than the power consumed by the
train's air-conditioning system (Heinrich and Kretzchmar,
The lightweight superconducting magnets of the EDS system
are at present cooled by either liquid nitrogen or liquid
helium. The discovery of materials that are superconducting at
room temperature can further reduce the weight of EDS vehicles
by eliminating the need to carry on-board refrigeration.
Further, these superconducting materials will reduce energy
transmission losses for both the EDS and EMS systems.
Superconducting magnets produce an intense magnetic field that
consumes less energy than EMS systems but requires shielding
of passenger compartments. The present U-shape EDS guideway
being tested in Japan is more complex and more expensive than
Transrapid's EMS guideway. The U.S. research focuses on
developing a simpler and less expensive guideway design by
using material such as glass or carbon fiber reinforced
plastic (Phelan and Triantafillou, 1992). This is
theoretically feasible because EDS vehicles are lighter than
Further improvements in superconducting technology and
guideway design are necessary to improve the financial
feasibility of EDS systems. The EMS systems such as Transrapid
or HSST have been approved for revenue operation. These EMS
systems, however, may not be suitable or cost effective in
regions prone to earthquakes or having low cohesive soils
because of small tolerances in the guideway design that must
be insured over the entire life cycle of the project.
Tilt and Nontilt Trains
In corridors where it is not feasible to build new tracks
and straightening the existing alignment is too expensive,
operating a high-speed train may not be cost-effective nor
comfortable due to constant acceleration/deceleration
operations. If track's horizontal curvatures are the major
constraint to higher-speed operation, a tilt train technology
may be appropriate (TRB, 1991; USDOT, 1991a).
The main advantage of tilt trains is that they can travel
up to 25% faster around curves than conventional trains and
still provide the same passenger comfort level (as described
in Chapter 2). Achieving the same level of speed improvement
for nontilt trains may require curve excursions from existing
right-of-way. This often means building expensive civil
engineering structures and acquiring costly land.
On the other hand, tilt trains use self-steering trucks
that align the wheels to the rails on curves thereby improving
the distribution of dynamic loading on the rails.
Self-steering trucks, however, limit the maximum speed
achievable on tangent tracks to a maximum of 250 km/h (155
mph). The TGV and ICE employ a truck design with a stiff
primary suspension to avoid unwanted oscillations at very high
speeds (USDOT, 1991a). Another concern is that active tilt
mechanisms add weight to each passenger car and increase the
trainset's mechanical complexity (Hopkins, 1990). This is the
main reason why some European manufacturers have abandoned
tilt train research after having experimented with few
prototypes. Generally, increased maintenance and reliability
problems should be assessed individually for each tilt train
Self-steering trucks are not applicable to maglev
systems. However, tilt maglev technology may cause aerodynamic
and vehicle stiffness problems, and also add weight and
mechanical complexity. Maglev systems can more easily and very
inexpensively increase the speed on curves by increasing the
banking angle of the supporting system. While the draft U.S.
regulations allow only a maximum track superelevation (i.e.,
banking angle) of 6 degrees for HSR systems, a track
superelevation of up to 30 degrees is being considered for the
U.S. maglev (USDOT, 1992). One degree of vehicle tilt has
similar effect on passenger comfort as one degree of track
REVIEW OF HSGT LITERATURE
This chapter consists primarily of a review of
government-sponsored studies. The reports reviewed include a
study conducted by the Argonne National Laboratory and
sponsored by the U.S. Department of Energy (1990), a study
conducted at Carnegie Mellon Research Center (1991), several
safety studies performed or commissioned by the USDOT and the
Transportation Research Board, one special report published by
the Transportation Research Board (1991), and a book written
by the current president of the High Speed Rail/Maglev
Association (1991). The literature related to train dynamics
and simulation methods is examined in the following two
Argonne National Laboratory Maclev Study
The Argonne National Laboratory study (Johnson et al.,
1989) was the first major technical evaluation of maglev
systems sponsored by a government agency since the federal
government terminated funding for all HSGT research in 1975.
The report's title is Maglev Vehicles and Superconductor
Technology: Integration of High-Speed Ground Transportation
into the Air Travel System. The study objectives were to
identify the status of maglev research in the United States
and abroad, identify potential markets, evaluate the impacts
of high-temperature superconductors (HTSC) (i.e., room
temperature superconductor), and assess the energy and
economic benefits of maglev systems.
The study concludes that there is a potential domestic
market for the maglev technology. The market, however, is not
cost-effective from one downtown area to another because
acquiring new right-of-way to provide access to urban areas
can be very costly. The report indicates that the major maglev
corridors should link airports and provide a substitute for
flights of under 960 km (600 mi). This would significantly
reduce air traffic congestion as well as noise and air
pollution. A conceptual plan for connecting hub airports with
maglev systems was developed and is shown in Figure 4-1.
It was estimated that maglev implementation can result in
a substantial energy saving because flights of up to 960 km
(600 mi) account for 45% of all energy consumed by commercial
aircraft. A model was develop to estimate energy consumption
by airlines and maglev systems. It was found that maglev
vehicles consume as little as one-third of the energy used by
aircraft over short distances. If 50% of short-haul flights
could be replaced by maglev trips, this could result in a net
saving of about 12% of the total fuel used by aircraft.
The report states that the discovery of room temperature
superconductors can enhance the technical aspect of EDS maglev
systems, reduce maglev vehicle weight by about 10%, and lower
/ g 5
energy and maintenance costs by an equal percentage.
In addition, it was found that the capital costs of
maglev systems are difficult to estimate. The report contains
a reference to cost estimates prepared for the Las Vegas to
southern California route. However, there are today
indications that these estimates are too optimistic. The
Argonne report also indicates that dedicated HSR systems are
expected to cost almost as much as maglev systems, but will
have higher operating costs. Neither of these claims can be
established, and recent cost estimates suggest otherwise
The Argonne National Laboratory is currently conducting
several maglev studies funded primarily by the National Maglev
Initiative Research Program. These studies deal with research
areas such as high-temperature superconductors maglev system,
vehicle/guideway interaction, flexible guideway, preliminary
design for maglev facility, maglev system design
considerations, and maglev potential role in short-haul
Carnegie Mellon University Study
The study conducted by the Carnegie Mellon University
(CMU) (Uher, 1990) was funded by a grant from the USDOT. The
CMU established an HSGT center in January, 1987. The center
initially evaluated a maglev technology developed by Boeing
under contract to the USDOT. Boeing discontinued the maglev
research after the USDOT stopped funding the project.
The CMU review indicated that the Boeing technology was
not suitable for a maglev regional system. Instead, the German
Transrapid technology was viewed as the most advanced maglev
system and the most appropriate for Pittsburgh.
The CMU research team predicted a significant HSGT
market, estimated at over $200 billion. This forecast was
based on two elements: highway congestion and air traffic
congestion. Highway congestion was estimated to increase
vehicle delay from 3 billion hours in 1985 to 12 billion hours
in 2005, and the attendant costs could increase from $12
billion to $46 billion as shown in Table 4-1. The cost of
delay due to air traffic congestion was estimated to increase
from $5 billion in 1986 to $13 billion in 2005 as shown in
Table 4-2. It is important to note that these estimates were
made at a time of rising demand. Recent increases in travel
demand are below these projected estimates.
The main feature of the CMU report is the proposal of a
conceptual 1800 km (1130 mi) regional maglev system for
Pittsburgh. The estimated capital cost was $29.2 billion in
1990 dollars ($26 million per mile). The annual operating
costs were estimated at $295 million (1990 dollars). It was
projected that the regional system could potentially generate
an economic activity of $96 billion during construction and
$885 million during operation.
The first phase of the proposed project would consist of
a demonstration segment linking Pittsburgh airport to the
Table 4-1: Cost of Vehicle Congestion to the Highway User
VEHICLE 1985 2005 Increase
CONGESTION (billion) (billion) (%)
Urban Freeway Congestion 1.6 8.1 406%
Signalized Arterial 1.1 3.8 245%
Total Congestion (hours) 2.7 11.9 341%
Person-Hours (a) 3.2 14.3 347%
Gallons Fuel (b) 0.8 3.6 275%
Congestion Cost (c) $12.2 $46.5 281%
(a) 1.2 persons per vehicle
(b) 0.3 gallons of fuel per vehicle-hour
(c) $3.0/person-hour + $1.0/gal
Source: Uher, 1990.
Table 4-2: Cost of Air Congestion
Source: Uher, 1990; USDOT, 1987.
DEMAND 1986 2005
Enplanements 414 millions 887 millions
Passenger Miles 359 billions 935 billions
Air Operations 29 millions 46 millions
Airports with Chronic
Delay (delay greater 11 airports 41 airports
than 8 min/operation) _
Cost of Delay $5.0 billion $13.3 billion
downtown area at a cost between $300 million and $650 million.
It was expected that the demonstration project will recover
its operating cost but not its capital cost. Plans were
proposed to develop economic activity nodes at maglev
stations. In addition, a manufacturing license would be
negotiated with Transrapid, and components of the maglev
system would be supplied by Pittsburgh-based companies. This
is consistent with one of the main objectives of the proposed
regional maglev system which consists of establishing a maglev
industrial base in Pittsburgh.
USDOT Safety Reports
The growing interest in HSR and maglev systems generated
a need to reexamine existing U.S. safety requirements as well
as the suitability of European and Japanese technologies to
the U.S. market. These advanced HSGT systems employ differing
equipment and operating procedures than conventional rails.
This required an examination of their safety in U.S. operation
and a rethinking of the federal approach to safety design and
regulations in order to avoid impending new technology as a
result of past regulations designed for conventional rails.
A series of safety reports were produced by the Federal
Railroad Administration (FRA), a division of the USDOT, which
is in charge of assuring the safety of rail systems in the
United States as stipulated in the Federal Railroad Safety Act
of 1988. Most of the safety studies were performed by the
USDOT's John A. Volpe National Transportation Systems Center.
Four HSGT safety reports were published in 1991 under the
series Moving America: New Directions, New Opportunities. The
first report, titled Safety relevant Observations on the
X-2000 Tilting Train (USDOT, 1991a), examined safety issues
related to the Swedish ABB X-2000 tilt train. The second
report, titled Safety of High-Speed Magnetic Levitation
Transportation Systems: Preliminary Safety Review of the
Transrapid System (USDOT, 1991b), addressed the safety of the
German Transrapid maglev system. The third report, titled
Safety Relevant Observations on the TGV High-Speed Train
(USDOT, 1991c), reviewed the safety of the French TGV
high-speed train system. The fourth report, titled Safety
Relevant Observations on the ICE High-Speed Train (USDOT,
1991d), reviewed the safety of the German ICE high-speed train
These FRA safety reports are of a preliminary nature.
They were initiated in response to the various corridor
feasibility studies and franchise applications to build and
operate an HSGT system in the United States. The X-2000 train
was proposed for application in the Northeast Corridor between
New York and Washington, D.C.; and in Florida to provide a
rail service between Miami, Orlando and Tampa. The Transrapid
maglev system was proposed for implementation in Orlando,
Pittsburgh, and between Anaheim and Las Vegas. The TGV system
was proposed for implementation in Florida, Texas, California
and other states. The ICE system was also proposed for
implementation in Texas, however the ICE proposal was not
selected by Texas authorities. The FRA is currently conducting
a more detailed review of these technologies, examining all
the requirements needed for their approval to operate in
revenue service in the United States.
Safety Issues Related to the X-2000 Tilt Train
The safety report related to the X-2000 tilt train
(USDOT, 1991a) presents an overall description of the system;
a preliminary evaluation of the tilt technology; and a review
of the safety features related to the tilting mechanism,
steerable trucks, braking system, and automatic train control.
Compliance with existing FRA regulations were also examined.
This includes compliance with the standards for noise
emission, track safety, marking devices, glazing (windshields
and window glass), locomotive safety, control system, and fire
This preliminary evaluation indicates that the train
operator's console exceeds U.S. industry practices in terms of
ergonomics and cabin noise standards. The noise emitted by the
X-2000 measured at 25 meters (82 ft) at 200 km/h (125 mph) was
no greater than a conventional train at 130 km/h (80 mph). The
brake system was found to operate smoothly and effectively.
The stopping distances for emergency applications were well
within U.S. accepted standards. Some modification of the train
interior were needed to meet flammability standards and
disabled access. These modifications, however, can be
performed without major difficulty. There remained certain
issues for which not enough data was available to complete the
evaluation. The X-2000 manufacturer, however, indicated their
intention to meet FRA safety requirements for any permanent
Safety Issues Related to the Transrapid Maglev Train
The safety reports related to the Transrapid maglev
system (USDOT, 1991b; FRA, 1992) present a description of the
Transrapid-07 vehicle and infrastructure, a comparison of FRA
and foreign safety standards, and an assessment of potential
maglev safety issues. The maglev technology poses two safety
concerns: first, the technology differs from the steel rail
system and requires new regulations; second, maglev systems
have only been operational on test tracks, and therefore
inadequate information is available to assess their
performance in revenue service throughout the system's life
The FRA developed a maglev safety analysis method known
as the System Safety Concept. It consists of applying
managerial and technical skills to systematically identify
potential hazards early in the system design, and to recommend
design modifications necessary to ensure safety before the
system is actually built. The hazard resolution process was
achieved by identifying hazards, assessing their severity and
probability, and developing corrective measures.
Potential safety issues were identified from operating
experiences, expert opinions, and a generic hazard checklist.
Expert opinions were used to assess the severity and
probability of each identified hazard. A risk assessment
matrix was developed to assist in the decision making process
to determine whether individual hazards can be accepted, or
they should be eliminated or controlled. Table 4-3 presents
the estimated probability of identified hazards, and Table 4-4
shows the risk matrix developed for the Transrapid system.
The FRA initiated a review process of maglev safety
standards and guidelines developed in Germany and other
countries to assist in adapting U.S. railroad regulations to
the maglev system. The structural design standards used for
aircraft bodies were also reviewed for possible relevancy to
maglev vehicle design. It was determined that modification of
existing FRA regulations may be required to accommodate maglev
systems in areas related to emergency braking, window glazing,
signal and train control, power supply, personnel
qualification and operating rules.
More detailed information was needed by FRA to evaluate
fully the ability of the Transrapid system to perform safely
in U.S. applications. A one-year test program of the Florida
Maglev Demonstration Project was viewed as vital for obtaining
adequate data on safety issues such as the dynamic interaction
between vehicles and guideway, operation in very high winds,
soil erosion, guideway oxidation, and extreme thermal
Probability Estimates of Undesired Events in
OPERATIONAL PHASES INVOLVING PASSENGERS
EVENT Passenger Leaving/ Accessible Inaccessible
DESCRIPTION Station Arriving Areas Areas
Transfer Station of Guideway of Guideway
Vehicle D D D D
Fire in Other
Critical Element C C C C
with Object C C C C
Vehicle Collision D D D D
Guideway E E E E
Sudden Stop N/A D D D
Slow/Stop at N/A D N/A N/A
Guideway N/A D C C
Rescue D D D C
Illness/Injury C C C C
Source: USDOT, 1991b
Maglev Risk Assessment Estimates
OPERATIONAL PHASES INVOLVING PASSENGERS
EVENT Passenger Leaving/ Accessible Inaccessible
DESCRIPTION Station Arriving Areas Areas
Transfer Station ofGuideway of Guideway
Fire/Explosion in IID ID ID ID
Fire in Other
Critical Element IIIC IIIC IIC IC
with Object IIC IIC IC IC
Vehicle Collision IID IID ID ID
Guideway IHE IIE IE IE
Sudden Stop N/A IIIC IIC IC
Slow/Stop at N/A IID N/A N/A
Stranded on N/A IID IIC IC
Rescue IIID IID lID ID
Passenger IIIC IIC IIC IC
NOTE: IA, IB, IC, IA, IIB, IIlA = Unacceptable;
ID, IIC, IID, IIIB, IIIC = Unacceptable (management decision required);
IE, IIE, IIID, IIIE, IVA, IVB = Acceptable with review by management;
IVC, IVD, IVE = Acceptable without review
Source: USDOT, 1991b
conditions. This information was considered critical towards
approving Transrapid for revenue operation in the United
Safety Issues Related to the TGV Train
On May 28, 1991, the Texas TGV Consortium, headed by the
Morrison Knudsen Corporation, was awarded a 50-year franchise
to design, build, and operate a high-speed rail system in
Texas between the cities of Houston, Austin, Dallas and San
Antonio (referred to as the Texas Triangle). The French TGV
train is the technology selected by the franchisee for
operation in Texas
The FRA undertook this safety study to examine which of
existing regulations may have to be adapted to cover the
unique characteristics of the TGV, and also to identify the
aspects of TGV's design, construction and operation that need
to be changed to meet specific U.S. customer applications.
The report indicates that although the review is
preliminary, the TGV system seems to comply (or comply with
minor modifications) with FRA standards for track safety,
control of locomotive, braking system, air brake equipment,
electrical systems (except the emergency pantograph pole), cab
equipment, signal system, train control system, automatic
block signal system, interlocking, traffic control systems and
automatic train stop system. The Texas TGV trainset will
provide wheelchair accommodation and a handicapped washroom in
compliance with the Americans with Disabilities Act of 1990.
Further measurements were needed to confirm compliance with
noise emission requirements, fire protection guidelines, and
the safety windshield and glazing standards. The Windshield
and glazing safety standards at high speed are not addressed
in FRA's existing regulations. This is important because the
kinetic energy of a collision with an object is greatly
influenced by speed.
The report concludes that the proposed Texas TGV system
presents unique features compared to the customary intercity
passenger train designs found in the United States. The FRA's
existing regulations were not designed for railroad operations
in excess of 176 km/h (110 mph); and do not permit rail
passenger operations at higher speeds (Amtrak operates at
speeds up to 200 km/h (125 mph) on the Northeast Corridor
under a waiver). The Texas TGV rolling stock is capable of 320
km/h (200 mph). Furthermore, the TGV operates as a fixed
trainset with an articulated passenger section consisting of
several cars. Therefore, the on-board microprocessor network,
the speed control system, the signal system, and the braking
system are all designed based on the operation of equivalent
trainsets; and should be taking into consideration in
formulating new safety regulations. The FRA needs also to
develop standard for fault-tolerant systems, and for the
reliability of computer hardware.
TRB Safety Reports
The Transportation Research Board (TRB) published two
circulars dealing with the safety of HSR and maglev systems.
The most recent Circular #387 (TRB, 1992) consists only of
research problem statements. The problem statements were
generated at a special workshop conducted at the request of
FRA. The workshop objective was to assist FRA develop
appropriate safety regulations for HSR and maglev systems.
Approximately 70 research problem statements were identified.
These problem statements are being addressed within the
National Maglev Initiative Research Program. The safety issues
were subdivided into the following four topics:
Maglev guideway and vehicles.
HSR guideway and vehicles.
System design criteria and operation of maglev and HSR.
Issues related to signal, control and power supply.
The TRB Circular #351 (TRB, 1989), titled Safety Factors
Related to Hicrh-Speed Rail Passenger Systems, addresses only
the safety issues of HSR systems. The purpose of this effort
was to review regulations and practices in several European
nations, and to obtain information on the European experience
in operating HSR systems in revenue service.
The report describes the basic differences in railroad
safety regulations between the United States and Europe. The
European approach to railroad safety is based on "designing
safety into the system" by assessing safety requirements for
each corridor, whereas the U.S. approach consists of insuring
safety through regulations. In most European countries, the
national railroad company itself has the primary
responsibility for safety, while the government has
responsibility to oversee and to approve or disapprove the
railroad's actions. Although European governments have
regulatory power, they generally do not exercise it through
detailed regulations as is the case in the United States.
Another important distinction is that local governments in
most European countries have no jurisdiction over the
operation of the national railroad network except through
specific agreements for local service.
The report also presented a description of European
safety considerations as related to HSR systems in areas of
rolling stock, infrastructure, signalization, electrification,
and communication systems. There was, however, no comparisons
made with corresponding FRA regulations.
TRB Special Report 233
The TRB Special Report 233 (TRB, 1991), titled In Pursuit
of Speed: New Options for Intercity Passenger Transport,
provides an impartial evaluation of HSGT systems and their
potential feasibility in the U.S. market. This 179-page
special report was prepared at the request of the USDOT.
Several consultants assisted the TRB committee members in the
preparation of this document. Among the consulting companies
that participated in this effort are the Canadian Institute
for Guided Ground Transport (CIGGT), Parsons Brinckerhoff
Quade and Douglas (PBQD), SYDEC, and Analytic Services. The
report includes a review of costs, ridership, financial
feasibility, and public support of HSGT systems. Most of the
information was based on previous feasibility studies and the
various proposals for HSGT systems. The cost section of the
report was prepared with the assistance of John Harrison of
PBQD. This section and other HSGT studies performed by PBQD
were extensively used in this dissertation's HSGT cost
analysis because they constitute the only completed
theoretical (i.e., not related to any specific corridor)
maglev cost studies commissioned and funded by the USDOT.
The report concludes that the high cost of HSGT systems
and the uncertainty in forecasted demand are the main reasons
that neither public initiative nor private enterprise has
produced an operating system in the United States. The report
also states that it is unlikely that HSGT systems will pay for
themselves under reasonable ridership estimates and farebox
revenues. Therefore, public support will likely be required in
most cases. According to this report, the justification for
public support can be based on a host of secondary benefits
(often referred to as externalities) to nonusers or the
general public. These benefits include congestion relief,
environmental advantages, economic impacts, and greater
control of land use development patterns. The report
indicates, however, that further studies are needed to
evaluate these secondary benefits and devise methods to
estimate them on a site specific basis.
The book Supertrains, Solutions to America's
Transportation Gridlock (Vranich, 1991) is written by Joseph
Vranich, the current president of the High Speed Rail/Maglev
Association (HSR/MA). The book was not intended to be a
scientific study. Mr. Vranich is a strong advocate of HSGT
systems, and his involvement with HSR/MA makes him quite
knowledgeable in this area. He was called on numerous
occasions to testify in congress hearings on HSGT-related
topics. His most recent testimony was on August 6, 1992,
before the U.S. Senate's Committee on Commerce, Science, and
The Supertrain book presents a summary of the advantages
of HSR and maglev technologies and a description of the
systems operational in Europe and Japan. Although the book
contains no new technical information, it provides a unique
insight into the working of Congress and the political context
of transportation funding decisions. Government actions can
have a major impact on the implementation of HSGT systems in
the United States. As mentioned is the previous section, very
few HSGT corridor studies indicated that their proposed HSR or
maglev system will be expected to recover its capital cost.
Therefore, some sort of government support will be required in
Although legislative actions related to tax-exempt bonds
and highway right-of-way use are important for HSGT systems,
the main legislative maglev issues are related to the level of
funding and whether the appropriated funds should be used to
develop a domestic maglev technology or to implement
"americanized" foreign maglev systems. The author believes
there are justifications for the federal government to
subsidize the construction of HSGT systems. The HSGT systems
are safer, more energy efficient and less polluting than other
modes. The author enumerates all the direct and indirect
subsidies (he referred to them as hidden subsidies) to the
highway system and airline companies. Furthermore, he claims
that airplane manufacturing industries have greatly benefited
from military research programs by adapting the findings to
The legislators who favor investing in the development of
a domestic maglev technology claim that the Japanese MLU
maglev is too expensive and not yet operationally ready, and
the German Transrapid maglev uses conventional electromagnets
and must levitate with small air gaps which limits speed and
increases guideway cost. They propose developing from scratch
a new generation of superconducting maglev technology
specifically suited for the U.S. market.
The legislators who favor americanizing a foreign
technology are more interested in a quick implementation of
the already proposed maglev systems, such as the systems
proposed for Orlando and Pittsburgh. Some of them claim that
investing in developing a U.S. maglev system is mainly
designed to assist aerospace and military industries overcome
military cutbacks. They are skeptical that U.S. industries can
in a short time and cost-effectively develop a substantially
better maglev technology than what is already available. They
don't foresee significant technological breakthroughs and they
believe it will be too expensive to match the billions of
dollars invested by the Germans and Japanese and the 15-year
experience gained by their researchers. They also claim that
americanizing (obtaining a manufacturing license) a foreign
maglev technology, such as the Transrapid, can lead to a
domestic manufacturing maglev industry with almost equal
benefits to the U.S. economy as developing a new system (this
reasoning is also held by the researchers of the Argonne
National Laboratory who proposed the Pittsburgh maglev
regional network using Transrapid technology). The author
remarks that the world manufacturing process is becoming more
integrated. Even Boeing airplanes contain foreign-manufactured
parts. He therefore advises against delaying proposed systems
until a U.S. technology is ready for implementation.
Finally, the author states that maglev vehicles account
for a small percentage of the total cost of a maglev system.
The largest cost components are for guideway, earthwork, and
civil engineering structures. The spin-off from these
investments alone can generate important ramifications for the
local as well as national economy, regardless which maglev
technology is used.
This chapter provides a description of resistive forces
acting on a train, the engine power needed to accelerate a
train on a level terrain and on grades, and the allowable
speed on curves. Most units in this chapter are expressed in
SI units only.
For a train to move forward and accelerate, its motors
must produce a tractive force superior to the sum of all
resistive forces. The acceleration rate is proportional to the
resultant force and limited by the adhesion coefficient
between rail and wheel.
The total resistance is the sum of all resistive forces
acting on a train at a given time or place. Traditionally, it
was measured in pounds per ton of train. The FRA has recently
adopted the metric system, and today total resistance is
measured in SI units (Systeme International) and expressed in
KN (Kilo Newton). The SI unit "tonne" equals 1000 kg and about
1.103 U.S. short tons.
The resistive forces acting on an HSR train are different
from the forces acting on a maglev train because maglev
systems operate without physical contact. At high speed,
however, air resistance becomes the dominant resistive force