CEN/TR 14067-7:2021

Railway applications - Aerodynamics - Part 7: Fundamentals for test procedures for train-induced ballast projection

This document discusses:
-   economic aspects of ballast projection;
-   comparison of methods in France and Spain for rolling stock;
-   infrastructure assessment methods;
-   review of available literature;
-   next steps and recommendations regarding standardization and research.

Bahnanwendungen - Aerodynamik - Grundlagen für Prüfverfahren für zuginduzierten Schotterflug

Applications ferroviaires - Aérodynamique - Principes généraux pour des procédures d'essais vis-à-vis des projections de ballast générées par la circulation des trains

Železniške naprave - Aerodinamika - 7. del: Osnove preskusnih postopkov za zaščito pred letečim drobirjem, ki ga sproža vlak

General Information

• Status: Published
• Publication Date: 27-Apr-2021
• ICS: 45.060.01 - Railway rolling stock in general
• Technical Committee: CEN/TC 256 - Railway applications
• Drafting Committee: CEN/TC 256/WG 6 - Aerodynamics
• Current Stage: 6060 - Definitive text made available (DAV) - Publishing
• Start Date: 28-Apr-2021
• Due Date: 03-Jun-2021
• Completion Date: 28-Apr-2021
• Directive: 2008/57/EC - Directive 2008/57/EC of the European Parliament and of the Council of 17 June 2008 on the interoperability of the rail system within the Community (Recast)
• Directive: 2016/797/EU - Directive (EU) 2016/797 of the European Parliament and of the Council of 11 May 2016 on the interoperability of the rail system within the European Union (Text with EEA relevance)
• Mandate: M/483 - Mandate for programming and standardisation addressed to the European standardisation bodies under directive 2008/57/EC in the field of the interoperability of the rail system within the European Union
• Ref Project: SIST-TP CEN/TR 14067-7:2021 - Railway applications - Aerodynamics - Part 7: Fundamentals for test procedures for train-induced ballast projection

Overview

CEN/TR 14067-7:2021 - "Railway applications - Aerodynamics - Part 7: Fundamentals for test procedures for train‑induced ballast projection" - is a CEN technical report that defines the fundamentals for assessing and testing the risk of ballast projection caused by trains. The report reviews the phenomenon (both ice‑fall and aerodynamically‑induced projection), summarizes historical incidents, examines economic impacts, compares national approaches (notably France and Spain), and gives recommendations for standardization and further research. It supports homologation and infrastructure assessment for high‑speed and regional operations.

Key topics and technical focus

• Phenomenon description: mechanisms of ballast projection, including ice accumulation and Couette‑type airflows under high‑speed trains.
• State of the art: incident summaries, tracks and sleeper types, and conditions linked to events (speeds, weather, ballast height).
• Economic assessment: cost of damage, cost of mitigation for rolling stock and infrastructure, and cost–benefit considerations.
• Homologation and test concepts: existing technical approaches, responsibilities between rolling stock manufacturers and infrastructure managers, and monitoring methods used during approval (e.g., air‑speed measurements, impact microphones, accelerometers, video inspection).
• National method comparisons: detailed comparison of practices in France, Spain, Italy, Belgium and others, with annexed summaries.
• Research and standardization recommendations: gaps identified, literature review and next steps to harmonize test procedures.
• Normative linkage: refers to EN 14067‑4 for related aerodynamic test requirements.

Practical applications - who uses this standard

• Train manufacturers - for vehicle design validation and homologation testing to demonstrate resistance to ballast projection.
• Infrastructure managers - for track maintenance criteria (e.g., ballast height), risk assessment and mitigation planning.
• Regulators and certification bodies - to define approval procedures and interface requirements between vehicles and infrastructure, especially for operations above typical high‑speed thresholds.
• Testing laboratories and research institutes - to develop, run and interpret aerodynamic and impact monitoring campaigns.
• Maintenance planners - to inform maintenance specifications that reduce projection risk (ballast profile, sleeper type).

Related standards and references

• EN 14067‑4:2013+A1:2018 - Railway applications - Aerodynamics (requirements and test procedures on open track).
• TSI LOC&PAS and TSI INF - referenced interface and safety requirements relating to operation at high speeds (>250 km/h internally discussed).

Keywords: ballast projection, railway aerodynamics, train‑induced ballast, test procedures, homologation, high‑speed trains, infrastructure assessment, EN 14067.

Standards Content (Sample)

SLOVENSKI STANDARD
01-julij-2021
Železniške naprave - Aerodinamika - 7. del: Osnove preskusnih postopkov za
zaščito pred letečim drobirjem, ki ga sproža vlak
Railway applications - Aerodynamics - Part 7: Fundamentals for test procedures for train-
induced ballast projection
Bahnanwendungen - Aerodynamik - Grundlagen für Prüfverfahren für zuginduzierten
Schotterflug
Applications ferroviaires - Aérodynamique - Principes généraux pour des procédures
d'essais vis-à-vis des projection de ballast causés par la circulation des trains
Ta slovenski standard je istoveten z: CEN/TR 14067-7:2021
ICS:
45.060.01 Železniška vozila na splošno Railway rolling stock in
general
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.

CEN/TR 14067-7
TECHNICAL REPORT
RAPPORT TECHNIQUE
April 2021
TECHNISCHER BERICHT
ICS 45.060.01
English Version
Railway applications - Aerodynamics - Part 7:
Fundamentals for test procedures for train-induced ballast
projection
Applications ferroviaires - Aérodynamique - Principes Bahnanwendungen - Aerodynamik - Grundlagen für
généraux pour des procédures d'essais vis-à-vis des Prüfverfahren für zuginduzierten Schotterflug
projections de ballast générées par la circulation des
trains
This Technical Report was approved by CEN on 19 April 2021. It has been drawn up by the Technical Committee CEN/TC 256.

CEN members are the national standards bodies of Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia,
Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway,
Poland, Portugal, Republic of North Macedonia, Romania, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Turkey and
United Kingdom.
EUROPEAN COMMITTEE FOR STANDARDIZATION
COMITÉ EUROPÉEN DE NORMALISATION

EUROPÄISCHES KOMITEE FÜR NORMUNG

CEN-CENELEC Management Centre: Rue de la Science 23, B-1040 Brussels
© 2021 CEN All rights of exploitation in any form and by any means reserved Ref. No. CEN/TR 14067-7:2021 E
worldwide for CEN national Members.

Contents Page
European foreword . 3
1 Scope . 4
2 Normative references . 4
3 Terms and definitions . 4
4 Symbols and abbreviations . 4
5 General aspects of ballast projection and state of the art . 4
5.1 Introduction . 4
5.2 Summary of studies and incidents (by countries, manufacturers) . 5
5.3 Overview of ballasted track systems in Europe . 9
5.4 Ice accumulation induced ballast projection . 15
6 Economic judgement of damage . 18
6.1 Cost of reported damage . 18
6.2 Cost of homologation, measures to rolling stock and infrastructure . 22
6.3 Cost benefit analysis . 26
7 Homologation concepts . 27
7.1 General. 27
7.2 Existing technical approaches . 27
7.3 Responsibilities, interests and intended interface definitions . 28
7.4 Conceptual approaches. 29
8 Comparison of existing methods . 32
8.1 France . 32
8.2 Spain . 36
8.3 Italy . 46
8.4 Belgium . 47
8.5 Other countries . 47
8.6 Comparison of existing methods . 48
8.7 Conclusion drawn from French and Spanish assessments . 48
9 Available background . 48
10 Conclusion and next steps . 49
Annex A (informative) Summary comparison of existing methods addressing ballast
projection . 51
Annex B (informative)  Review of ballast projection papers . 56
Bibliography .103

European foreword
This document (CEN/TR 14067-7:2021) has been prepared by Technical Committee CEN/TC 256
“Railway Applications”, the secretariat of which is held by DIN.

1 Scope
This document discusses:
— economic aspects of ballast projection;
— comparison of methods in France and Spain for rolling stock;
— infrastructure assessment methods;
— review of available literature;
— next steps and recommendations regarding standardization and research.
2 Normative references
The following documents are referred to in the text in such a way that some or all of their content
constitutes requirements of this document. For dated references, only the edition cited applies. For
undated references, the latest edition of the referenced document (including any amendments) applies.
EN 14067-4:2013+A1:2018, Railway applications - Aerodynamics - Part 4: Requirements and test
procedures for aerodynamics on open track
3 Terms and definitions
For the purposes of this document, the terms and definitions given in EN 14067-4 apply.
ISO and IEC maintain terminological databases for use in standardization at the following addresses:
• ISO Online browsing platform: available at https://www.iso.org/obp
• IEC Electropedia: available at https://www.electropedia.org/
4 Symbols and abbreviations
For the purposes of this document, the symbols and abbreviations given in EN 14067-4 apply.
5 General aspects of ballast projection and state of the art
5.1 Introduction
The phenomenon of ballast projection has been caused in the past by lumps of ice or accreted snow falling
from the train structure during extreme winter conditions, which then strike the ballast bed causing
ballast to be ejected upwards, impacting the train underside or passing trains and leading to damage.
Typically, this damage includes breakage of train underbody equipment, failures of train systems or
reduced efficiencies, breakages of station or train windows, and impact damage to train or trackside
structures. This type of ballast projection occurs to regional trains as well as to high speed trains and has
been well-known for a long period in railways world-wide.
However, in the early 2000s there were a number of significant incidents of ballast projection involving
high speed trains which were not caused by ice fall, but seemed to arise from aerodynamic causes.
Substantial damage was caused to the underside of an ICE 3 train in one particular incident in Belgium in
2003. This phenomenon seems to be solely a high speed train phenomenon. The relevant contributory
factors involve:
— the aerodynamic design of the train, particularly the train underbody;
— the air speeds generated under the train, due to the Couette-type flow created by the high speed
train passing over the static track bed;
— train-induced pressures acting on the ballast;
— track vibrations caused by the traction engines and wheel passing over the rails.
The types of damage from this sort of incident principally includes: ballast stone impacts to the
underbody structure and damage to the underbody equipment, pipes and cables of high speed trains, and
damage to wheels and rails when ballast stones are trapped between them. Although possible, there
seems to be little evidence of collateral damage to other trains or of injury to trackside workers.
It should be noted that although the aerodynamically-induced ballast projection incidents have resulted
in some spectacular damage to trains, there is evidence of minor impact damage to train underbodies at
lower train speeds that appears to be deemed tolerable by train maintainers. Furthermore, since the first
upsurge in these incidents, there has been a complete cessation with no further incidents since 2004. This
reduction in incidents coincides with measures introduced by many European Infrastructure Managers
to reduce ballast levels relative to the top of sleepers.
Nevertheless, there is a widespread concern that this is still a valid train/infrastructure issue needing
certain controls, which is supported by its inclusion in the LOC&PAS and INF TSIs as an issue, (albeit
currently as an open point). Consequently, various national rules regarding the issue have been
developed, mainly focused on rolling stock, leading to a burden on train manufacturers trying to
introduce trains into different countries, as they are required to apply different methods to confirm their
trains’ performance with regard to ballast projection.
Within the revision published in June 2019, ballast projection is addressed in TSI LOC PAS and TSI INF as
an interface issue relevant for operation with train speeds >250 km/h. The issue is connected to the
essential requirements of safety and technical compatibility.
5.2 Summary of studies and incidents (by countries, manufacturers)
5.2.1 General
Table 1, reproduced from Claus [1], summarizes some of the major incidents up to 2006 of both types of
ballast projection in this century (those due to winter weather are not exhaustive).
Table 1— Reported past ballast projection incidents
Date Train Location Speed Track Weather Remarks
Type Type Conditions
2001 ICE 3 Fulda- 230 km/h Mono-block Winter
Göttingen, sleepers, conditions,
Germany lowered snow
ballast
2003 KTX South Korea 300 km/h Mono-block No snow See [Kw03] for
sleepers details
2003 ICE 3 Lille-Calais, 320 km/h Bi-block Winter
France sleepers conditions,
snow
2003 ICE 3 Belgium 300 km/h Mono-block No snow Speeds up to
sleepers, 275 km/h did
ballast not not cause
lowered problems in
double
traction
2004 ICE 3 France 320 km/h Bi-block No snow During
sleepers homologation
test runs
2004 ICE 3 Mannheim- 250 km/h Mono-block Winter Foreign parts
Stuttgart, sleepers, conditions, in the track
Germany lowered snow have been
ballast found
2004 ETR 500 Rome- 300 km/h Mono-block No snow New track
Naples, Italy sleepers, with ballast
ballast not above the
lowered sleepers, see
Fig 1
2006 ICE-T Hamburg- 230 km/h Mono-block Winter
Berlin, sleepers, conditions,
Germany lowered snow
ballast
The following sections give additional details of incidents and the current status of ballast projection
measures in different European countries.
5.2.2 Italy
During initial runs of the ETR 500 on the then newly constructed Rome-Naples high speed line in 2004 a
ballast projection incident occurred. The sleepers were of the mono-block type, as shown in the
photograph in Figure 1 taken at the time of the incident. It can be seen that the ballast was not specially
lowered between the sleepers, and significant amounts lay on top of the sleepers themselves.
On the first high speed line, the 25 kV Turin-Milan route, problems arose for 300 km/h running of the
ETR 500. Three levels of ballast height reductions were investigated with ballast impacts being
monitored using microphones. The best ballast level was chosen for the maintenance target. Maintenance
procedures were modified to ensure that ballast stones are properly placed, but the intervals between
track maintenance have not been increased. Since the initial problems, there have been no further
problems with flying ballast for trains regularly running up to the maximum speed of 300 km/h.
Homologation for the ETR 400 (also known as ETR 1000) running above 300 km/h required special
measures and configurations of track to ensure test running up to 360 km/h without problems.
There are no national notified technical rules for ballast projection in Italy. If ballast projection occurs,
the operator of the train has to establish a settlement with the infrastructure manager.

Figure 1 — Part of the Rome-Naples line at the time of the ballast projection incident in 2004, [1]
5.2.3 Spain
Railway Gazette (2005) stated that Spanish Development Minister Magdalena Álvarez, when presenting
a report in March 2005 to the Spanish Parliament on difficulties encountered with the Madrid - Lleida
section of the Madrid - Barcelona high speed line, that train operations at speeds greater than 300 km/h
caused ‘ballast particles to be sucked up and thrown around’.
The Universidad Politecnica Madrid and Adif have studied ballast projection since 2008. Preliminary
studies were undertaken in 2008-2010, followed by the AeroTRAIN Project (2010-2012), and then the
Aurigidas Project (2012-2014). In all of the projects, detailed measurements of surface loads and
induction phenomena were undertaken with ballast risk projection analysis. These were followed by
Norm definition studies in 2015-2017. A national guideline on the issue is being prepared by Adif.
5.2.4 France
In France, the major incidents of ballast projection occurred in 2003 and 2004 during homologation tests
of the ICE 3 running at 320 km/h (single unit). Damage to the train and the track was very significant and
of a magnitude never seen to TGVs or other high speed trains running in France, even at higher speeds.
Following this, different types of studies were carried out by German and French personnel to understand
the origin of these incidents and to try to avoid or at least limit them.
As a result of this work, the underbody of ICE 3 running in France was modified and tests were carried
out (air speed measurements in the track bed during train passing, microphones, video control,
underbody inspection) in 2005, to check the behaviour of the modified train regarding ballast projection.
These tests confirmed the decrease of the aerodynamic load on the track and the frequency and intensity
of ballast projection.
A review of ballast incidents was carried out after one year of operation. Several incidents were identified,
but have always been observed during cold weather (temperature close to zero and snow) in Germany.
Due to these incidents, for all new high speed trains running in France or for test running at speeds above
320 km/h, some specific monitoring has been required since 2004 on the trains (accelerometers or
microphones for stone impact counting and video control), and air speed measurements in track have to
be made for all new high speed trains.
A French regulation, SAM X 012 [4], was published by EPSF in 2015 to describe the methodology to assess
a new high speed train regarding ballast projection. This methodology is based on air speed
measurements in the track and consists of comparing the new train with an existing train or with an
absolute criterion (cf § 7.2).
In France, ballast projection also occurs in winter due to snow/ice conditions. To limit these incidents,
speed reductions are applied on French high speed lines.
5.2.5 Germany
The major incident of ballast projection of concern to Germany actually occurred not in Germany, but in
Belgium, during homologation tests on the ICE 3 on a new high speed line. The ICE 3 had run without
incident on high speed lines in Germany. The Belgian high speed line was fitted with mono-block sleepers,
but no ballast lowering had been instigated. Successful trials had taken place with single traction units of
200 m length running at up to 300 km/h. However, when the double units were running at 270 km/h,
ballast impacts on the train underside were audible. During one run severe damage occurred to the train
underside, which led to Siemens having to investigate improved underbody designs to limit the problem.
Passenger reports on 15 Jan 2016 stated that an ICE travelling from Stuttgart to Frankfurt was hit hard
by ballast stones. The incident was reportedly caused by ice lumps dropping from the train. Damage was
caused to the train under-floor surfaces and side windows were cracked. The train continued at reduced
speed. (SOURCE: https://rail-sim.de/forum/index.php/Thread/20576-Bahn-reduziert-
H%C3%B6chstgeschwindigkeit-im-Fernverkehr/?pageNo=3:).
5.2.6 Great Britain
The High Speed 1 line (HS1) in Great Britain, runs from London to the Channel tunnel, and Eurostar trains
using the line reach a top speed of 300 km/h. A form of rail damage, known as ‘ballast pitting’, has been
reported, but no ballast projection problems have. This type of damage appears to be associated with
small particles of ballast becoming trapped between the railhead and the wheels of rail vehicles. It is
thought that the speed and the energy of trains cause an explosive crushing of the ballast particle, which
damages both the railhead and the wheel. Inspection and maintenance records of train wheels on HS1
show that wheel pitting occurs predominantly on the front and rear bogies of a train with a steady
reduction towards the centre of the vehicle. Quinn et al [2010], suggest that the vibrations and pressure
pulse associated with the approaching train play a part in making ballast particles airborne as the trains
travel in the reverse direction on the return journey.
High Speed 2 is a new line soon to begin construction in 2019, and will have a design speed of 360 km/h
and an operating speed of 330 km/h, obviously, there are not yet ballast projection problems on this line.
5.3 Overview of ballasted track systems in Europe
5.3.1 General
There are several aspects of the design of ballasted track directly under passing trains, which may be
relevant to the issue of aerodynamically induced ballast projection. These are:
— sleeper type and sleeper spacing;
— rail fastening systems, and whether these provide restraint to ballast movement;
— ballast size;
— ballast maintenance procedures.
How these aspects vary in different European countries could have an impact on risk of ballast projection
on different railway infrastructure designs.
5.3.2 Sleepers
There are a number of different types of sleepers, which are used in varying proportions by European
railways. They are used to support the rails in position on the ballast bed and are generally of similar
sizes, and differ mostly in their construction materials. These are summarized in Table 2.
Table 2 — Common sleeper types used on European ballasted track.
Sleeper material Size Notes
Timber Typically, 150 mm Different types of wood can be used.
high, 250 mm wide.
Ballast longitudinal movements restricted by the
sleeper, unless ballast above sleeper level.
Steel Similar size to wooden
sleeper.
Concrete mono-block Similar size to wooden Ballast longitudinal movements restricted by the
sleeper. sleeper, unless ballast above sleeper level.
Concrete twin- or bi- Consists of two blocks connected by a coupling rod
block or pipe to maintain separation of blocks.
Ballast free to move longitudinally between blocks.
Of particular interest is the bi-block sleeper, where there is no constraint to the movement of the ballast
in the direction of the train from the sleeper between the two end blocks.
Examples of the different sleepers are shown in Figure 2.
The sleeper spacing, which for instance varies between 600 mm and 800 mm in Great Britain, depends
on the axle loading of trains using the line, track curvature and whether there are local track formation
difficulties. This is to ensure the vertical and lateral stability of the track. Due to sleeper functionality,
there is unlikely to be any systematic difference between European countries in sleeper spacing, although
within a country there could be differences between lines depending on the factors above. For high speed
lines, it is expected that differences in sleeper spacing between countries will be relatively small, (as axle
loads will be similar, track formation will be of high quality and track curvatures will be larger than on
conventional lines).
a) Wooden sleepers b) Steel sleepers

c) Mono-block sleepers d) Bi-block sleepers
Figure 2 — Examples of railway sleeper types
5.3.3 Rail fastenings
Rail fastening are used to secure the rails to the sleepers and there are a number of different designs
currently in use. Baseplates may be also used under rails to hold the rails in place and may be used in
conjunction with clip devices.
The DE (Deenik, Eisses) clip, shown in Figure 2 a), is widely used and can be fitted on concrete or wooden
sleepers. A number of manufacturers have also produced other fastening devices, such as Pandrol,
Vossloh, McKay, and Nabla. These other devices are shown in Figure 3.

a) Rail baseplate b) Pandrol clip c) Vossloh fastening

d) McKay fastening e) Nabla fastening
Figure 3 — Rail fastening systems
It can be observed that the rail fastening systems, although having differing designs, are unlikely to
influence ballast movement or provide any significant impediment to aerodynamic ballast projection.
5.3.4 Ballast Size
EN 13450:2013 Aggregates for railway ballast specifies a standard for railway ballast sizes. Although
currently withdrawn, the standard is still referenced in company standards e.g. by Network Rail in Great
Britain. Essentially, ballast is graded according to the mass of ballast which passes through sieves with
holes of varying sizes. Ballast particles range between a nominal 31,5 mm up to a maximum of either 50
mm or 63 mm. Within each of these ranges there are five sub-ranges referred to as grading categories, A
to F overall. Figure 4 shows grade category A in the range up to 50 mm, and grade F in the range up to 63
mm, and represents the full range of ballast sizes and distributions. For each grade category there is also
a further sub-range at each sieve size, indicated by a solid line at the maximum value and a chained line
at the minimum. It can be seen that most ballast particles are between 20 mm and 63 mm, whichever
grade of category they are in.
Key
X sieve size, mm
Y percentage passings by mass
cat A
cat B
Figure 4 — Distribution of ballast sizes, Categories A and F
Ballast is maintained using a number of machines, such as stone blowers. These use smaller ballast
particles, capable of passing through a 20 mm sieve, which are blown under the sleepers once the
machine has lifted them, to maintain the track level. At worst, 80 % of this type of ballast will be large
enough to pass through a 14 mm sieve. It is possible that these sized particles can find their way to the
surface of the ballast bed.
There is no evidence known to the authors of systematic ballast size and shape distributions being
different in different European countries.
5.3.5 Ballast maintenance regimes
5.3.5.1 General
An important consideration for aerodynamic ballast projection is where ballast is permitted to lie
between the running rails, which will depend on the maintenance regime for high speed lines in each
country.
5.3.5.2 Great Britain
5.3.5.2.1 Network Rail
A variety of railway sleepers are used on the standard railway lines in Great Britain. These include
wooden sleepers, concrete mono-block sleepers and steel sleepers. Figure 5 shows a still image from a
Network Rail ballast maintenance video; ballast stones can be seen on the sleepers, despite the
commentary stating “No ballast must be left on the railhead, piled against the rail webs, over the
fastenings or loose on the sleepers”. It is not clear if the instruction was eventually complied with in the
video.
Figure 5 — Still from Network Rail video on ballast maintenance
5.3.5.2.2 High Speed 1
High Speed 1 is constructed using twin-block concrete sleepers on plain line and mono-block pre-stressed
concrete sleepers in the- vicinity of turnouts and switches. It uses a mixture of ballasted and slab track.
5.3.5.2.3 High Speed 2
High Speed 2 will be constructed with a mixture of slab track and ballasted track. No information has
been obtained regarding the sleeper types to be used.
5.3.5.3 Germany
German infrastructure managers established a procedure to counteract damage by ice dropping
instigating ballast projection. In late autumn the height of ballast is checked and lowered by a sweeping
machine to 4 cm to 6 cm below the base of the rail. This maintenance work is performed according to
internal rule 820.2010 7 (6) for every railway line above 140 km/h except for curves with small radius.
As sleepers are then elevated compared to the ballast bed level, objects dropping from fast running trains
will most likely hit the sleeper only, due to their flat-angle trajectory. The impact on sleepers is acceptable
and avoids the swirl of further ballast stones from the ballast.
5.3.5.4 CER position paper
In 2015, the Community of European Railway and Infrastructure Companies (CER) and European Rail
Infrastructure Managers (EIM) produced a joint position paper on aerodynamic ballast projection,
CER/EIM (2015). This specifically addressed the open points relating to ballast projection in the
LOC&PAS and INF TSIs and set out some principles that were felt should be respected when regulating
for the issue. The paper acknowledges the full-scale track test procedure for the assessment of rolling
stock set out in Annex A of EN 14067-4:2013+A1:2018 Railway applications - Aerodynamics - Part 4:
Requirements and test procedures for aerodynamics on open track, but notes the absence of limit values or
acceptance criteria. It notes and does not oppose the philosophy espoused in EN 14067-4 of the
independent assessment of rolling stock (and by inference infrastructure), subject to the condition that
such an assessment permits suitable controls to be put into place on other track infrastructure for safe
operations.
The paper notes that the issue of damage resulting from ballast projection is a probabilistic issue, both in
terms of ballast flight instigation and in serious damage arising from subsequent ballast impacts. The
paper takes the view that a certain level/number of ballast impacts is tolerable and the aim should be to
limiting the rate of impacts to being ‘rare events’.
The paper then considers the infrastructure on the basis of, essentially, a reference rolling stock concept,
rather as is considered for the cross wind issue, and note that there are limited possibilities to implement
significant design changes in infrastructure, which could take decades to put in place. It then asserts that
the most effective method to tackle the issue is to lower ballast levels, relatively easily implemented at
the construction stage, but more difficult as a maintenance task on tracks with mono-block sleepers.
Following this argument, it states that ballast lowering should not be mandatory, but only recommended.
For track with bi-block sleepers, ballast gluing is proposed if normal maintenance is not sufficient.
Track fencing is suggested for sensitive areas to protect against ‘projectile ballast’. The report
acknowledges that incidents of ballast projections involve inadequate infrastructure maintenance and
indicate ballast management best practice advice. The following is an important position statement:
“It is worth mentioning that issues related to the phenomenon of ballast pick-up due to aerodynamic
effects are rarely reported and thus it is the view of EIM and CER that the validity of the currently
implemented maintenance measures should be regarded as proven by experience.“
Proposals are made for both rolling stock and infrastructure, which are summarized below.
Rolling stock — A reference limit for aerodynamic effects regarding the phenomenon of ballast pick-up
should be established for rolling stock. New vehicles that do not fulfil this criterion should not be
authorized.
Infrastructure — No requirement is needed for infrastructure for speeds up to 250 km/h. Measures
currently applied on infrastructure in order to control the phenomenon of ballast pick-up due to
aerodynamic effects for speeds in the excess of 250 km/h should be seen as TSI compliant. Measures,
such as ballast lowering and maintenance procedures for clearing sleeper tops and fastenings of ballast,
for effective design, construction and maintenance of ballast cross section are to be harmonized in the
TSI.
5.4 Ice accumulation induced ballast projection
Issues also arise from ballast being projected up from the ballast bed and causing damage to the
underside and underbody equipment of trains during winter conditions, but in these cases, aerodynamics
plays no direct part. During relatively severe winter conditions in central Europe and under more usual
winter conditions in Nordic countries, excessive falling snow or snow already on the tracks accumulates
on the bogies, coupling gears between rail vehicles and on under-surfaces of vehicles, and under the
action of wind and temperature turns to ice. Periodically, either due to the weight of ice accumulation or
because of local temperature changes, large blocks of ice fall from the train, hit the ballast bed at high
speed causing a spray of ballast to hit the vehicle, resulting in similar damage to that caused by
aerodynamically-induced ballast projection. The primary cause of the ballast projection is the fall and
impact of ice blocks into the ballast, although Kloow [2011] suggests that train underbody pressures,
lifting of snow and ice located on top of the ballast and train-induced vibrations in the track bed, reducing
friction between ballast particles, may also be influences.
The problem has long been encountered by railways, as described by Saito [2002] for instance:
“One unexpected problem started with a heavy snowfall at Sekigahara (between Nagoya and Kyoto) in
December 1965. Shinkansen trains running through snowfall at very high speeds blow up the snow on
the track. The blown snow sticks to the underfloor equipment and freezes rapidly into ice. When the train
enters a warmer region, the frozen ice thaws and drops at high speed from the train causing the ballast
to fly up and seriously damage the underfloor equipment. There was no immediate solution other than
reducing speeds to 70 km/h in snow. Later, water sprinklers were installed along the track in order to
melt the snow, but speeds in snow are still restricted even today (although not as low as 70 km/h in most
cases), because excess water sprinkling can damage soil embankments. Drawing lessons from this, the
Joetsu Shinkansen, which runs through a very snowy region north of Tokyo, was built entirely on concrete
viaducts and large amounts of warm water are sprinkled during snowfall.“
Table 1 gives brief details of some significant winter ballast projection incidents in Germany and France
in the period 2001-2006. Such events cause widespread disruption to services in countries for whom
such winter conditions are exceptional rather than normal. The winter of 2009/2010 was particularly
severe across Europe. Wiebe & Friedrich [2010] reported results for a UIC survey about the impacts on
rolling stock and infrastructure in that winter. Of eleven countries providing information about the
challenges for rolling stock, six reported problems with trains which picked up ballast that then damaged
trains, and eight reported problems with ice blocks falling from trains.
It is perhaps obvious to state that problems due to falling ice blocks will increase with train speed, as the
blocks will have greater kinetic energy to impart into the ballast bed and thus into individual ballast
particles.
In the winter of 2010-11 Deutsche Bahn experienced severe disruption to train operations nationwide
caused by snow, with long distance services being significantly affected. A preventive train speed
reduction to 200 km/h resulted in delays and together with other issues, about 4 % of scheduled railway
operations (based on running distance in kilometres) could not be realized in December. Although train
speeds were reduced, damage still occurred to train underbody components (e.g. axles, pressure pipes,
windows) due to ballast projection caused by ice falling into the ballast bed. As a result, the increased
amount of maintenance work due was delayed by the additional time needed for de-icing trains in the
maintenance plant prior to service.
The melting ice also affected electrical equipment due to short circuiting. Overall, up to 10 % of long
distance trains were out of service on some days, leading to an increase in passenger demand. As other
modes of transport were also affected by the winter conditions, passenger demand increased further due
to mode transfer on important main lines by one third.
Several countermeasures were taken including technical improvements, namely impact protection on
axles, short circuit protection measures; improved heating of the couplings of the ICE2; installation of
three new de-icing stations; rental of train vehicles and several improved operational and maintenance
procedures.
Also, the German infrastructure managers established procedures to counteract damage by ice dropping
instigating ballast projection. Generally, the height of ballast is checked and lowered to about 5 cm below
base of the rail in late autumn for every railway line above 140 km/h, see 5.3.5.3. As damage increases
significantly with train speed, the so-called “Schneemeldeverfahren” (snow reporting procedure) is
established for high speed lines with ballast track. It foresees visual observation of trains passing by
traffic controllers about every 50 km, once a closed layer of snow is present. If snow is observed to be
blown up by trains passing, the trains speed is reduced (200 km/h) on that line. The same speed
reduction applies, if impacts in the underbelly region are audible to personnel on board of high speed
trains. In the case of ballast flight, the speed can be lowered further to 160 km/h locally to check damage
to the infrastructure.
Kloow (2011) and RSSB (2016) describe in some detail the range of winter problems that affect rolling
stock and infrastructure, including ballast projection due to ice fall. Kloow (2011) also details the
measures undertaken by railways to ameliorate ballast projection problems, giving both benefits and
drawbacks, which are summarized below, with particularly interesting points being included:
— Lower the ballast height:
“In Finland, Norway and Sweden the ballast level has been lowered approximately 3-5 cm below top of
the sleeper. The lowering is not done from aerodynamic reasons but primarily for letting the falling ice
and snow crush against the sleepers’ concrete edges. Lowering the macadam level has in fact proven to
be very effective for speeds up to 200 km/h and has led to the situation that some people do not consider
ballast pick-up a problem any longer.”
— Reduce train speed:
This is the procedure in Germany and France where adverse conditions only occur a few days per
year. However, this is not an acceptable solution in Nordic countries; as such conditions are more
prevalent there during average winters.
— Decrease the distance between sleepers:
This reduces the amount of ballast that can be impacted by ice falls.
— Make the underside of the train as flat as possible:
“The underpressure below the train can be reduced if the train’s underframe is designed as flat as
possible. However, bogies and ploughs will always create turbulence under the train.”
— Apply train designs that accumulate as little ice and snow as possible;
— Cover the ballast with a screen or other mechanism;
Some additional information on this type of solution is given in RSSB (2016);
— Use ballast-free tracks (i.e. slab-track) for high speed trains;
— Apply heat to surfaces on the train that accumulate snow;
— De-ice trains from packed snow and ice.
Management consultants Leigh Fisher (2012) made a presentation benchmarking NS and ProRail winter
measures against international practice. Comparisons between four European countries were made of
rolling stock measures applied to counter snow accumulation, see Table 3. Notably, speed reductions,
ballast height lowering and train design feature among the measures. The use of de-icing of trains prior
to entry into service is an increasingly used method, as shown in Table 4. Further details of de-icing
techniques are given in RSSB (2016).
Table 3 — Rolling stock engineering measures used to mitigate snow accumulation in winter in
Germany, Sweden, Great Britain and Switzerland, Leigh Fisher (2012)

Table 4 — The use of train de-icing in winter in Germany, Sweden, Great Britain and
Switzerland, Leigh Fisher (2012)

With regard to the issue of train design, Kloow (2011) summarizes some measures from published
literature for good train designs to reduce snow and ice accumulation under trains as summarized below:
— cover moving surfaces with e.g. plexiglass on foam rubber to reduce ice formation;
— use rounded and low friction surfaces;
— use deformable low friction covers to reduce snow packing;
— avoid flat surfaces that move towards each other; introduce sharp edges on one of the flat surfaces;
— use an open bogie design;
— install air spoilers to direct air and snow flows around the bogies and car body components;
— avoid aerodynamic stagnation points in the bogie region;
— consider localized heating on critical parts of the vehicle underbody.
CEN/TR 16251:2016 “Railway applications – Environmental conditions – Design guidance for rolling stock”
provides design guidance for avoiding problems with snow and ice accumulation. It recommends an
assessment of the aerodynamic environment on a vehicle’s undersurface at the design stage. In addition,
it suggests that:
— consideration is given to design to ensure that suspension travel is not blocked by ice and snow
accumulation, perhaps using bellows to cover springs;
— air pipes should not be located next to moving parts and should be protected from damage;
— snow and ice build-up in critical areas should be prevented;
— visual inspections of snow build-up should be carried out in critical winter conditions and de-icing
considered.
6 Economic judgement of damage
6.1 Cost of reported damage
A survey was started to collect data about the cost of reported damage to rolling stock and to
infrastructure. The limited number of responses to the survey was for several reasons, e.g.:
1) Projects with incidents had unexpected cost and hence dispute about the responsibility;
2) Transparent cost reports could reveal cost of components and working processes to competitors;
3) Damage not leading to failures in operation may have been detected a long time after the incidents,
i.e. during maintenance. Hence, the origin of the damage would not be clearly stated.
As the operational speed of vehicles running on ballasted tracks increases, the number of incidents and
reported damage regarding ballast projection increases. However, it is not always possible to define the
reason for these incidents. Ballast movement can be induced by the aerodynamic load of the rolling stock,
but it is more likely to be created by ice drop-off hitting the track bed. Therefore, during winter conditions
or on vehicles running in cold climates, it is difficult to attribute the root cause of the reported incidents.
Nevertheless, one way or the other, consequences observed on the rolling stock due to ballast projection
induced damage can be summarized as:
— Damage to the air reservoir of the secondary suspension. The impact of ballast particles causes
cracks and material loss to the air reservoir of the secondary suspension. This implies that there are
regions where the thickness of the reservoir is reduced below an allowable level accordin
...