Rail Accident Report

Derailment at Liverpool Street station, London

23 January 2013

Report 27/2014
December 2014
This investigation was carried out in accordance with:
l the Railway Safety Directive 2004/49/EC;
l the Railways and Transport Safety Act 2003; and
l the Railways (Accident Investigation and Reporting) Regulations 2005.

© Crown copyright 2014

You may re-use this document/publication (not including departmental or agency logos) free of charge
in any format or medium. You must re-use it accurately and not in a misleading context. The material
must be acknowledged as Crown copyright and you must give the title of the source publication.
Where we have identified any third party copyright material you will need to obtain permission from the
copyright holders concerned. This document/publication is also available at www.raib.gov.uk.

Any enquiries about this publication should be sent to:

RAIB Email: enquiries@raib.gov.uk

The Wharf Telephone: 01332 253300
Stores Road Fax: 01332 253301
Derby UK Website: www.raib.gov.uk
DE21 4BA

This report is published by the Rail Accident Investigation Branch, Department for Transport.
Derailment at Liverpool Street station, London
23 January 2013

Contents

Summary 5
Introduction 6
Preface 6
Key definitions 6
The accident 7
Summary of the accident  7
Context 8
Events during the accident  11
The investigation 14
Sources of evidence 14
Investigation timing 14
Key facts and analysis  15
Background information 15
Identification of the immediate cause  18
Identification of causal factors  24
Identification of underlying factors 40
Observations 43
Summary of conclusions  44
Immediate cause  44
Causal factors  44
Underlying factors 44
Additional observations 45
Previous RAIB recommendation relevant to this investigation 46
Actions reported as already taken or in progress relevant to this report 47
Actions reported that address factors which otherwise would have resulted in a
RAIB recommendation  47
Learning points 48
Recommendations 50

Report 27/2014 3 December 2014

Liverpool Street
Appendices 51
Appendix A - Glossary of abbreviations and acronyms 51
Appendix B - Glossary of terms 52
Appendix C - Key standards current at the time 56
Appendix D - Switch diamond crossings 57

Report 27/2014 4 December 2014

Liverpool Street
Summary

Summary
Shortly after 10:00 hrs on Wednesday 23 January 2013, train 1P18, the 10:00 hrs
Greater Anglia service from London Liverpool Street to Norwich, derailed 260 metres
from London Liverpool Street. The train comprised nine coaches pushed by a
locomotive, and had just left platform 13. A total of 17 wheelsets derailed on a tight
curve and, as the train proceeded, all the wheelsets were guided back onto the correct
rail within a distance of 40 metres.
The driver was unaware of any problem until the senior conductor told him that
passengers had reported a rough ride and the signaller advised him that the signalling
system had identified a problem at a set of points used by the train when leaving
Liverpool Street. The driver then stopped and examined his train at Shenfield, but saw
nothing unusual. No one appreciated that there had been a derailment until the train
was examined by a specialist inspector when it arrived at Norwich and, at about the
same time, a signal maintenance team found track damage close to Liverpool Street
station.
The train derailed on the curve because the track fixings had deteriorated over a
period of time. This tight curve and other non-standard trackwork at Liverpool Street
should have triggered consideration of mitigation measures to deal with the associated
enhanced derailment risk. The investigation found that no consideration had been
given to these enhanced risks because the maintenance management staff did not
have the knowledge necessary to appreciate the need for, and to undertake, this
activity. This lack of knowledge had not been appreciated by more senior staff. The
Network Rail procedures for establishing a track inspection and maintenance regime
for non-standard track did not require the regime to be independently checked.
The RAIB has identified six learning points and three recommendations. One learning
point relates to effective communication between train and incident controllers when
dealing with events which could be associated with urgent safety issues. A second
learning point restates the relevance of Network Rail’s existing requirements for
verifying maintenance management staff competencies relevant to risk assessing
track assets. Three learning points refer to the need for a complete record of assets
requiring maintenance, the importance of looking for signs of rail movement when
inspecting track and the correct use of data obtained from a commonly used track
geometry measurement device (an Amber trolley). The final learning point refers to
the need for proper archiving of inspection records.
The three recommendations are all addressed to Network Rail. The first relates to
providing assurance that suitable inspection regimes are established, recorded and
validated for non-standard track assets. The second recommendation is intended to
ensure assessment of management staff’s safety critical track related competencies
to ensure they have the necessary experience and knowledge to perform that role.
The third recommendation seeks a review and, if necessary, improvement of the
competency assessment processes applicable to managers with safety critical roles
linked to the maintenance of assets other than track.

Report 27/2014 5 December 2014

Liverpool Street
 Introduction
Introduction

Preface
1 The purpose of a Rail Accident Investigation Branch (RAIB) investigation is to
improve railway safety by preventing future railway accidents or by mitigating their
consequences. It is not the purpose of such an investigation to establish blame
or liability.
2 Accordingly, it is inappropriate that the RAIB’s reports should be used to assign
fault or blame, or determine liability, since neither the investigation nor the
reporting process has been undertaken for that purpose.
3 The RAIB’s investigation (including its scope, methods, conclusions and
recommendations) is independent of any other investigations, including those
carried out by the safety authority or railway industry.

Key definitions
4 All dimensions in this report are given in metric units, except speeds and locations
which are given in imperial units, in accordance with normal railway practice.
Where appropriate the equivalent metric value is also given.
5 The report contains abbreviations and technical terms (shown in italics the first
time they appear in the report). These are explained in appendices A and B.

Report 27/2014 6 December 2014

Liverpool Street
The accident

The accident
Summary of the accident
6 Shortly after 10:00 hrs on Wednesday 23 January 2013, train reporting number
1P18, the 10:00 hrs Greater Anglia service from London Liverpool Street to
Norwich, derailed 260 metres after departing from London Liverpool Street
station. The train comprised nine coaches pushed by a locomotive, and had
just left platform 13. A total of 17 wheelsets (19 wheels) had derailed on a
sharply curved section of track within 2035B points on the approach to 2035C
points (figures 1 and 2). 2035B points form part of a switch diamond crossing
(paragraphs 61 and 62).
7 The driver was unaware of the derailment and continued to drive the train
normally. All of the derailed wheels were guided back onto the correct rail within
a distance of about 40 metres. The driver remained unaware of any problem
until he was contacted by the on-board senior conductor. The senior conductor
asked him to stop and examine the train at Shenfield because passengers had
reported a rough ride. The driver was also given a similar request by the signaller
in response to the signalling system identifying a problem at 2035C points. The
driver stopped and examined his train at Shenfield, but he saw nothing unusual.
The train then continued to its final destination before anyone appreciated that
there had been a derailment.
8 In addition to minor damage to the train wheels, the bogie of one coach, parts of
2035C points and approximately 40 metres of track were damaged. Significant
disruption to services using Liverpool Street station continued until 05:45 hrs the
following morning.

Location of accident

© Crown Copyright. All rights reserved. Department for Transport 100039241. RAIB 2014

Figure 1: Extract from Ordnance Survey map showing location of accident

Report 27/2014 7 December 2014

Liverpool Street
The accident

2035C points

Area of
derailment

Incident train
direction of travel

Figure 2: Derailment site showing path of train 1P18 (2035B points in centre of image)

Context
Location
9 Liverpool Street station is the London terminus for trains serving East Anglia,
local services to north-east London and express services to Stansted Airport. A
complex junction just outside the station routes trains between the three pairs of
tracks approaching the station and the eighteen terminal platforms (figure 3). The
junction consists of many sets of points linked with short sections of plain track.
The maximum permitted speed for trains using this junction is 15 mph (24 km/h).
10 Signalling at Liverpool Street is by track circuit block and colour light signals
controlled from Liverpool Street Integrated Electronic Control Centre (IECC).
11 Operating incidents are managed by Network Rail and Greater Anglia staff
located in the Anglia Integrated Control Centre (AICC).
Organisations involved
12 Network Rail owns and maintains the track and signalling which is operated as
part of its Anglia Route.
13 Abellio Greater Anglia Ltd (referred to as Greater Anglia in this report) operated
the train which derailed, and employed both its driver and senior conductor. It
also operated an empty train which passed over the derailment site shortly after
the accident.
14 Greater Anglia and Network Rail’s Anglia Route freely co-operated with the
investigation. There were some delays in obtaining information from Network Rail
national centre staff due to their workload.

Report 27/2014 8 December 2014

Liverpool Street
 The accident

Figure 3: Liverpool Street station and detail of derailment area

Trains involved
15 The derailed train comprised an unpowered driving van trailer (DVT) at the front
of the train and eight Mark 3 coaches propelled from the rear by a class 90
electric locomotive (figure 4). In this report, the DVT is designated vehicle 1 and
the following coaches are designated vehicles 2 to 9 (figure 5).
16 A second train, an empty Class 321 electric multiple unit, reporting number 5V00,
passed over the derailment site a short time after train 1P18, but did not derail.
17 Examinations of the train involved in the derailment by the RAIB and Greater
Anglia found no evidence of pre-accident defects which could have contributed to
the derailment.

Report 27/2014 9 December 2014

Liverpool Street
The accident

Figure 4: A DVT leading a train formed of Mark 3 coaching stock. A class 90 locomotive is attached
at the rear of the formation (not the train involved or the location of accident) (image courtesy of Kev
Gregory/Railway Herald)

Loco V9 V8 V7 V6 V5 V4 V3 V2 DVT

Direction of travel Derailed

bogie

Figure 5: Train 1P18 (derailed wheels shown in red)

Staff involved
18 The key Network Rail maintenance staff involved with this investigation and their
position within the organisational structure are shown in figure 6.
19 The route asset manager [track] (RAM[T]) had 33 years railway experience. This
includes working in general track maintenance grades, eight years within a track
design office and 13 years in senior track maintenance positions. He became
RAM[T] in 2010.
20 The infrastructure maintenance delivery manager (IMDM) had 35 years
experience in track engineering. His initial experience comprised six years track
design and 16 years in track maintenance technical grades. He then became a
track maintenance engineer in 1994 and held senior track maintenance positions
through several business reorganisations. He was appointed Romford IMDM in
October 2012.

Report 27/2014 10 December 2014

Liverpool Street
21 The infrastructure maintenance engineer (IME) was an electrical engineer, who

The accident
joined the railway in 2000. He worked in the fixed plant asset management
organisation for eight years including a year as the area electrification and plant
engineer on Wessex route. He became Romford IME in 2008.
22 The track maintenance engineer (TME) responsible for Liverpool Street joined the
railway in 1984. He initially worked for a track renewal team before transferring to
a track maintenance team and advancing through the grades to become TME in
2006. The TME is responsible for maintenance of track from Liverpool Street to
Chelmsford.
23 The section manager [track] (SM[T]) joined the railway in 1984. He advanced
through the track maintenance grades and reached the position of SM[T] in
1999. The SM[T] is in charge of the teams that undertake track maintenance at
Liverpool Street.
24 The assistant track maintenance engineer (ATME) had over 30 years railway
experience as a structures engineer before moving into the track maintenance
discipline when he became ATME in 2012. The ATME provided the TME, the
SM[T] and other track maintenance staff with technical support.
25 The train services manager employed by Greater Anglia and the incident
controller employed by Network Rail were among the AICC staff who dealt with
the derailment.
External circumstances
26 The derailment occurred where the track is sheltered from natural sunlight and
rain by buildings built over the railway. The environment is dirty and dimly lit by
artificial lighting.

Events during the accident

27 At approximately 10:00 hrs on Wednesday 23 January 2013, train 1P18 departed
from platform 13. As the train was passing over 2035B points, and approaching
2035C points (figure 3), on a right-hand curved section of track about 260 metres
from the platform end, 19 wheels of the train derailed (figure 5). Of these, 15
wheels were re-railed within 5 metres. The remaining four wheels were on a
bogie which travelled in a derailed state for approximately 40 metres (around
6 seconds running time), before forcing its way through (running through) the
converging rails of 2035C points and re-railing (paragraph 66)).
28 The senior conductor was towards the rear of train 1P18 and felt the train move
in an unusual manner. The driver of the train was unaware of this rough ride and
the damage incurred to 2035C points and continued to drive the train as normal.
29 As the train ran through 2035C points there was a loss of detection which was
indicated on the signaller’s display at the IECC. This indication is a warning to the
signaller that the points are possibly not in a position which provides safe passage
for trains. The signalling system then prevented the signaller from routing trains
over the points. The signalling shift manager reported this to the Network Rail
incident controller at the AICC who, at 10:05 hrs, requested attendance from the
local signalling maintenance team.

Report 27/2014 11 December 2014

Liverpool Street
 30 As the senior conductor walked through the train she received reports from
The accident

passengers of a rough ride and dust falling from the ceiling. At 10:13 hrs the
senior conductor contacted the Greater Anglia train services manager at the AICC
and expressed a concern that the train could have derailed. However, at the
time of the call, the train was apparently running normally so they agreed that the
senior conductor would ask the driver to make an unscheduled stop at Shenfield
to examine the train.
31 The senior conductor contacted the driver and requested him to examine the train
because of the rough ride report. Shortly after this, the driver was also contacted
by the signaller who reported the loss of detection at 2035C points and also
asked him to examine the train.
32 At approximately 10:30 hrs the train stopped with the left-hand side of the
train alongside platform 3 at Shenfield. The driver contacted the signaller and
arranged for train movements to be stopped on the adjacent track. After the
signaller had confirmed that the train movements had been stopped, the driver
climbed down on to the track and walked along the right-hand side of the train.
The driver looked for loose or hanging objects which could have damaged
the points. Although some of the right-hand wheels had marks caused by the
derailment which would have been visible on close inspection, he was not looking
for such damage as it was not yet appreciated that a derailment had occurred. At
the rear of the train, the driver climbed on to the platform and started to walk back
towards the front of the train. The driver continued looking for defects, but the
platform restricted his view of the left-hand wheels.
33 After completing his examination, the driver reported back to the signaller that
he had found no defects and considered the train safe to continue. The signaller
gave the driver permission to proceed and the train continued to Norwich
running at normal speeds of up to 100 mph (160 km/h). The Greater Anglia train
service manager contacted one of Greater Anglia’s technical riding inspectors
and requested him to examine the train in more detail when the train arrived at
Norwich.
34 Although they were located at adjacent workstations in the AICC, the Greater
Anglia train services manager did not inform the Network Rail incident controller
about the senior conductor’s concerns that the train could have derailed, but did
mention a rough ride.
35 At around 10:17 hrs, the signal maintenance team at Liverpool Street began
investigating the continued loss of detection at 2035C points. A loss of detection
can be caused by a number of mechanical misalignment or electrical faults.
The derailment had not left any obvious marks in the immediate vicinity of the
points, so the maintenance team had no reason to suspect that a derailment had
occurred. Believing the fault to be with the point operating machine, they made
routine maintenance adjustments to the machine mechanism. These adjustments
re-established detection which allowed the signaller to route a train over the
points.

Report 27/2014 12 December 2014

Liverpool Street
36 At 10:35 hrs train 5V00, an empty class 321 electric multiple unit departed from

The accident
platform 11. This was the first train to pass over the derailment site after the
signal maintenance team had re-established detection. Train 5V00 passed over
the derailment site without derailing or suffering damage, but as it passed over
2035C points, detection was again lost. This again prevented the signaller from
routing trains over the points and the signal maintenance team resumed their
investigations.
37 The signal maintenance team re-examined the point operating machine and, after
not finding a fault, began looking for faults elsewhere. The signal maintenance
team found that a crank forming part of the mechanical points drive mechanism
had become detached from the track. This fault could not be fixed by the signal
maintenance team so, at 12:31 hrs they requested the attendance of track
maintenance staff.
38 While awaiting assistance from track maintenance staff, the signal maintenance
team walked towards the station and observed significant damage to the track
consistent with a derailment. This was reported to the AICC at about 12.47 hrs.
39 The incident controller had not appreciated that a derailment might have occurred
until damage was reported to the AICC by the signal maintenance team. At
approximately the same time, about 2 hours 45 minutes after the derailment,
the technical riding inspector at Norwich reported to the AICC that he had found
damage to wheels and bogie equipment consistent with a derailment.
40 The loss of detection at the damaged points prevented services using
platforms 11 and 12; this trapped three empty trains at the station (each platform
can accommodate more than one train). All other trains were forced to use
alternative routes around the damaged points, causing disruption to services for
the remainder of the day. Two of the six lines serving Liverpool Street remained
closed until repairs were completed at 05:45 hrs on 24 January.

Report 27/2014 13 December 2014

Liverpool Street
 The investigation
The investigation

Sources of evidence
41 The following sources of evidence were used:
l witness statements;
l the train’s on-train data recorder data;
l data from the Control Centre of the Future;
l site photographs and measurements;
l Network Rail’s records of track maintenance and inspections;
l examinations of the track on the day of the incident and when permanent
repairs were carried out in November 2013; and
la review of previous RAIB investigations that have relevance to this accident.

Investigation timing
42 The reasons for the derailment were established on the day of the accident in
sufficient detail for Network Rail to begin implementation of the precautionary
actions described in paragraphs 168 and 169. With the RAIB’s agreement,
Network Rail implemented emergency repairs shortly after the derailment to allow
rapid reopening of the railway. These repairs included replacement of rail fixings
damaged in the accident using components positioned so that they obscured,
but did not destroy, the witness marks needed to gain the full understanding of
the derailment mechanism described in this report. The RAIB decided to delay
collection of this witness evidence until November 2013, the earliest date when it
could be obtained without disrupting planned maintenance activities.

Report 27/2014 14 December 2014

Liverpool Street
Key facts and analysis

Key facts and analysis

Background information
Organisational Structure
43 Network Rail has structured its maintenance workforce into individual routes,
each with their own route managing director (RMD) and associated management
structure. The key Network Rail maintenance staff involved with this investigation
and their positions within the organisational structure are shown in figure 6.
44 Below each RMD are several functions including two relating to infrastructure
maintenance:
l asset management, to balance maintenance, renewal and enhancement
activities for the best return on investment; and
l maintenance delivery, the workforce directly responsible for ensuring safety of
the infrastructure and implementation of inspection and maintenance activities.
45 The asset management function is intended to use technical expertise,
experience and records of asset deterioration to decide whether maintenance or
renewal would best meet the safety and operational requirements of the railway.
The team is split into specialist disciplines and covers the whole Anglia route. The
track team responsible for Liverpool Street were based in offices near to Liverpool
Street station, but have since moved to offices in Stratford.
46 The RAM[T] leads the Anglia route track asset management team responsible
for the cost-effective management of track assets. The RAM[T] also provides
technical assistance and mentoring to the route TMEs. He reports through the
director of route asset management (DRAM) to the RMD.
47 The RAM[T] is also responsible for arranging the briefing of local maintenance
staff about technical standards information originating from the Network Rail
national centre in Milton Keynes.
48 The Anglia route maintenance teams are divided into geographical areas and
specialist disciplines. For the track at London Liverpool Street the maintenance is
undertaken by the maintenance delivery unit (MDU) based at Romford.
49 The Romford MDU is responsible for the infrastructure on the lines from
London Fenchurch Street to Tilbury and Southend, and from Liverpool Street to
Chelmsford on the main line to Norwich. The MDU works in accordance with
Network Rail standards, Network Rail guidance and local knowledge to develop
an inspection and maintenance regime. This is intended to ensure the safe
operation of the railway and to meet the reliability and performance targets set by
the RAM[T].
50 The MDU at Romford is led by the infrastructure maintenance delivery manager
(IMDM) who is accountable1 for safety of the line, and manages all maintenance
disciplines, within this MDU.

1
The accountable person typically ensures appropriate processes are in place. Tasks are actually undertaken by
the responsible person, sometimes with assistance.

Report 27/2014 15 December 2014

Liverpool Street
Key facts and analysis

Route Managing Director

(RMD)

Director Route Asset Route Infrastructure

Management Maintenance Director
(DRAM) (RIMD)

Route Asset Manager Infrastructure Maintenance

[Track] Delivery Manager
(RAM [T]) (IMDM)

Infrastructure Maintenance
Engineer
(IME)

Track Maintenance
Engineer
(TME)

Assistant Track Maintenance

Section Manager [Track]
Engineer
(SM [T])
(ATME)

Principal Technical Officer Senior Technical Officer

Section Supervisor Team Leader
[Track] [Track]
[TRACK] [TRACK]
(PTO [T]) (STO [T])

Technical Team Track Staff

Track Technical Teams

Figure 6: Organisation structure relevant to track maintenance

51 The Romford infrastructure maintenance engineer (IME) reports to the IMDM and
is responsible for managing the staff who ensure the railway remains safe for
operational use within the MDU area. His remit covers all engineering disciplines.
Discipline-specific maintenance engineers report to him, including two track
maintenance engineers. One of these covers the Fenchurch Street, Tilbury and
Southend lines and the other covers the Liverpool Street to Chelmsford area,
including the accident site.
52 Each TME is responsible for developing the inspection and maintenance regime
for his geographical area and implementing this regime. The TME responsible
for the Liverpool Street to Chelmsford area has support from an assistant track
maintenance engineer (ATME) and two section managers [track] (SM[T]),
each with their own support teams. The ATME provides input throughout the
geographical area covered by the TME, but the same area is split geographically
between the two SM[T]s.

Report 27/2014 16 December 2014

Liverpool Street
53 The dates of appointment of key staff relevant to the investigation are shown

Key facts and analysis

in table 1. The relevance of these appointment dates is considered in
paragraphs 121 to 150.

Position Appointment
IMDM October 2012
IME 2008
TME 2006
ATME April 2012
SM[T] 1999
Table 1: Appointment dates of key staff at Romford MDU

Track Standards
54 Network Rail requires the inspection and maintenance of track to be in
accordance with its standard NR/L2/TRK/001, ‘Inspection and Maintenance of
Permanent Way’. This standard prescribes the inspections, intervention limits
and actions required to prevent derailments due to track defects and to optimise
track performance, cost and asset life.
55 Network Rail standards are updated as necessary in response to the introduction
of new technology, the application of new methodologies or in response to a
previous incident requiring a new working practice. Where possible Network
Rail issues the latest version ahead of the date when the company requires full
compliance. This allows those people affected to prepare before full compliance
is necessary.
56 NR/L2/TRK/001 had been updated several times and issue five was current
at the time of the derailment. Issue six had been issued but, at the time of the
derailment, compliance had not yet been mandated. The version history of
NR/L2/TRK/001 is shown in table 2.

Issue number Issue date Compliance date

Initial Issue August 2005 August 2005
Issue 2 October 2005 October 2005
Issue 3 26 August 2008 26 August 2008
Issue 4 05 December 2009 05 December 2009
Issue 5 02 June 2012 01 September 2012
Issue 6 01 December 2012 02 February 2013
Table 2: Issue dates of Network Rail standard NR/L2/TRK/001

57 Each standards update usually has an associated briefing pack which outlines
the changes made from the previous issue. For issue 5 of NR/L2/TRK/001 the
standard had been rewritten into a new format which was substantially different to
previous formats. The differences between the versions of NR/L2/TRK/001 and
how they are relevant to this investigation are covered in paragraphs 137 to 150.

Report 27/2014 17 December 2014

Liverpool Street
 58 Network Rail issues briefing updates, grouped by engineering discipline, which
Key facts and analysis

lists those standards that have been updated in the preceding period. It is the
responsibility of all Network Rail staff to ensure they are aware of any updates
to standards which apply to them by attending organised briefing sessions or by
self-learning, and to implement those changes by the compliance date. Local
managers maintain a record of staff attending briefings and follow up staff
absence from briefings as necessary.

Identification of the immediate cause2

59 The derailment occurred as train 1P18 negotiated the small radius curve at
2035B points because the outer rail fixings of the left-hand rail on the curve
were unable to resist the lateral forces acting at the wheel/rail interface.
The forces were sufficient to widen the track gauge such that the right-hand
wheels on the trailing bogie of the third vehicle dropped between the rails.
60 The following evidence supports this:
l Marks on the track.
l Marks on the wheels of train 1P18.
l Marks on the sleepers and damage to the track fastenings indicating lateral
movement of the track fixings.
l Computer simulations of similar derailments which show that significant gauge
spread forces are developed on small radius curves (paragraph 72).
61 When the RAIB examined the site it found evidence that the derailment had
occurred on a section of movable switch rail that formed part of 2035B points.
62 These points, together with 2035A points, formed the switch diamond crossing
illustrated in figure 7. A description of a switch diamond crossing is included at
appendix D.
63 The derailment evidence comprised closely spaced marks on the right-hand rail
indicating that all the derailed right-hand wheels of the train had dropped inside
this rail approximately five metres beyond the tip of the moveable rail (figures 8
and 9). The RAIB has concluded from the wheel face (the outside face) damage
on the train, and a lack of derailment marks elsewhere on the right-hand rail, that
the 17 right-hand train wheels derailed in turn as they reached this location. At
the point of initial derailment the track is on a right-hand curve and the left-hand
wheels were running along a section of rail that should be fixed.
64 The first of the wheelsets exhibiting wheel face damage were on the trailing bogie
of vehicle 3 (figure 5). Damage to the track fixings of the right-hand rail (figure 9)
and the absence of damage on the left-hand wheels of this bogie show that these
wheelsets continued with the right-hand wheels dropped inside the right- hand
running rail while the left-hand wheels remained on the rail. Damage to the
track fixings indicate that these wheels ran derailed for about 4.5 metres before
encountering a fishplate forming part of a rail joint. As they struck the fishplate,
the wheels were raised back up to rail level and guided back on to the running rail
(figures 10 and 11).

2
The condition, event or behaviour that directly resulted in the occurrence.

Report 27/2014 18 December 2014

Liverpool Street
 Key facts and analysis

Figure 7: Track layout at incident site

Report 27/2014 19 December 2014

Liverpool Street
Key facts and analysis

Point of
derailment

2035B
Points

Path of
train 1P18

2035A
Points

Figure 8: Incident site showing switch rails set for train 1P18 (alternative route shown in yellow)

Direction of travel

Tread drop-in marks

Path of derailed wheels

Figure 9: Wheel tread drop-in marks and damage caused by derailed wheels

Report 27/2014 20 December 2014

Liverpool Street
 Key facts and analysis

Figure 10: Derailment mechanism (excepting leading bogie of vehicle 4)

C:\RPB Stuff\RAIB Diagrams.dgn 20/11/2014 17:20:46

Report 27/2014 21 December 2014

Liverpool Street
Key facts and analysis

Wheels strike fishplate

and are forced back into
normal running position

Fishplate

Path of derailed
train wheels

Figure 11: Rail joint which re-railed some wheels

65 Damage to all wheels on the leading bogie of vehicle 4 indicate that this bogie
fully derailed. Rail head marks (paragraph 63) and marks on the right-hand
wheel faces show that the right-hand wheels initially followed the same path as
vehicle 3. However, instead of the right-hand wheels re-railing at the rail joint, the
left-hand wheels climbed over the left-hand rail. This is evidenced by damage to
the flanges of the left-hand wheels, marks left across the left-hand rail head and
damage to the track fixings on the outside of the left-hand rail.
66 The extent of damage to track fixings shows that all wheels on this bogie then
continued in a derailed state for approximately 40 metres until converging rails at
2035C points forced the wheels to run through the closed points and guided them
back into a normal state on the running rails (figure 12).
67 On the remaining 13 wheelsets that derailed, damage was limited to the
right-hand wheel face. This indicates that these wheelsets followed a similar
derailment path to those of the trailing bogie on vehicle 3 (paragraph 64).
68 Scuff marks and depressions indicative of rail fixing movement were evident on
the upper surface of sleepers near the point of derailment. The extent of these
marks observed by the RAIB showed that the left-hand rail fixings had moved
outwards, increasing the distance between the running rails by up to 35 mm.
This rail was intended to remain in a fixed position but it was able to move as a
consequence of deterioration of the rail fixing (paragraphs 82 to 84).

Report 27/2014 22 December 2014

Liverpool Street
 Key facts and analysis

Figure 12: Derailment mechanism for vehicle 4 leading bogie

C:\RPB Stuff\RAIB Diagrams.dgn 20/11/2014 17:21:12

Report 27/2014 23 December 2014

Liverpool Street
 69 Conventional rolling stock, like the coaches of train 1P18, can generate large
Key facts and analysis

lateral forces at the wheel/rail interface when they negotiate small radius
curves. This force will act to spread the rails apart if they are not fully restrained
and evidence of gauge spread at this location is provided by scuffing marks
(paragraph 68).
70 Although the track in the derailment area was tightly curved, with a radius of
125 metres, the inner (right-hand) rail was not equipped with a check rail. Had
such a rail been provided, it would have been placed close to the right-hand rail
(figure 13) and contacted the back of the right-hand wheels. This would have
reduced the lateral forces applied to the left-hand (outer) rail and prevent the
right-hand wheels dropping between the rails.

Check Running
rail rail

Running
rail

Figure 13: Typical check rail

71 Current Railway Group standard GC/RT5021 normally requires a check rail on

curves less than 200 m radius. However, the standard reflects Network Rail’s
view that this is not viable on movable rails such as the right-hand rail at this site.

Identification of causal factors3

Behaviour of the track
72 The RAIB has investigated a number of derailments on small radius curves where
widening has occurred due to the lateral forces exerted by the passage of trains.
The RAIB investigation of the derailment at Liverpool Central underground station
on 26 October 2005 (RAIB report number 14/2006), found that these lateral
forces (gauge spreading forces) generally increase if the track gauge widens.
The vehicle dynamic study commissioned by the RAIB as part of the Ordsall Lane
Junction investigation (RAIB report number 07/2014) also found a similar gauge
spreading force. In both these instances the forces combined with other factors
and resulted in a different type of derailment from that at Liverpool Street.

3
Any condition, event or behaviour that was necessary for the occurrence. Avoiding or eliminating any one of
these factors would have prevented it happening.

Report 27/2014 24 December 2014

Liverpool Street
73 In the vicinity of the derailment, the RAIB found that the distance between the

Key facts and analysis

rails was wider than normal, and there were signs of deterioration allowing
movement in the rail fixings intended to resist such movement. The RAIB has
concluded that the following factors are the most likely explanation of how the
rails became sufficiently separated for the right-hand wheels on vehicle 3 to drop
inside the rails:
l the wider than normal track gauge on the curve with no trains present (static
gauge) (paragraphs 75 to 80);
l the degraded condition of the rail fixings in the vicinity of the derailment,
which reduced the strength of the connection between the rails and sleepers,
making it easier for the rails to move apart as train 1P18 passed over
(paragraphs 81 to 97); and
l an increasing gauge spreading force due to the combined effects of
excessive static gauge and rail movement due to the degraded rail fixings
causing additional deflection of the left-hand rail under the passage of train
1P18 (paragraphs 98 to 101).
74 The RAIB did not observe any defect with train 1P18 which would indicate that it
was significantly more at risk of derailment by gauge spread than any other trains
that were regularly routed over the curve. It is therefore likely that similar gauge
spreading forces were being generated by trains preceding the passage of 1P18,
and that these forces were themselves the cause of the deteriorating condition of
the both the track gauge and the rail fixings.
Wider than normal static track gauge
75 The wide static track gauge, which had developed on the curve between
2035B and 2035C points, reduced the margin that was available for the rails
to safely deflect when train 1P18 passed over. This was a causal factor.
76 Most rails at Liverpool Street, except switch rails, are fixed to wooden sleepers
using a retaining system consisting of a baseplate which is screwed to the
sleeper using chairscrews and spring clips which clamp the foot of the rail to the
baseplate (figure 14). This type of fixing is intended to resist the forces imposed
on the rail by train movement.
77 To ensure the safe passage of trains the distance between the rails (gauge)
must be maintained to ensure adequate support for the train wheelsets. This
distance can be increased both by design and by the normal deterioration of track
in service. As an increased gauge can increase the risk of derailment, Network
Rail standard NR/L2/TRK/001 defines the limits at which intervention should take
place to maintain a suitable gauge and associated safety margin.
78 The original track design drawings showed an increase in the standard track
gauge of 1435 mm to 1441 mm at the derailment site. This increase is normal
practice and is intended to ease the passage of trains negotiating small radius
curves.

Report 27/2014 25 December 2014

Liverpool Street
Key facts and analysis

Rail wheel

Spring clip
restraining rail foot
Rail
Baseplate

Sleeper
Chairscrew

Chairscrew

Figure 14: Rail fixing system

Location image
(photograph taken September 2013)

Path of
1P18

Ingrained dirt at Scuff marks

edge of fixing, evidencing rail
discounting movement in
sudden movement direction of white
during/shortly arrows then return
before derailment in the opposite
direction

Rail movement evidence

(photographed immediately
after derailment)

Figure 15: Wide gauge rail fixing with ingrained dirt around baseplate discounting possibility of recent
movement

Report 27/2014 26 December 2014

Liverpool Street
79 The static track gauge was measured throughout the derailment site immediately

Key facts and analysis

after the derailment and a value of 1474 mm was found at the drop in (derailment)
marks. If found during routine inspection, Network Rail standard NR/L2/ TRK/001
would require remedial action within 36 hours. Although gauge was not
measured immediately before the derailment, the build up of dirt immediately
behind the baseplate (figure 15) discounts the possibility that the baseplate first
moved during, or shortly before the accident. Network Rail records show that the
static gauge was measured on March 2012 and found to be less than 1455 mm
(paragraph 122) indicating that significant deterioration had occurred over the
following ten months.
80 The inspection regime intended to detect, and trigger correction of, excessively
wide static gauge at Liverpool Street is described in paragraph 102.
Degraded condition of rail fixings
81 The degraded condition of the rail fixings, on the curve between 2035B and
2035C points, made it easier for the rails to move apart when train 1P18
passed over. This was a causal factor.
82 Forces imposed on outer rails of curves, such as the left-hand rail at the location
of the derailment, can cause the rail fixings to deteriorate. The chairscrews
may break, or the sleepers may deteriorate and allow the chairscrews to move.
This failure allows the baseplate to move under the lateral forces imposed as
wheelsets pass. The curvature of the rail will then tend to pull the baseplate back
towards its correct position after wheelsets pass.
83 This repetitive movement leaves a scuff mark and/or depression on the sleeper
beyond the normal position of the outer edge of the baseplate (commonly known
as ‘shuffle’, figure 16). This effect is well understood and normally mitigated by
the regular inspection and maintenance processes required by Network Rail
standard NR/L2/RK/001.
84 The scuff marks and depressions observed after the accident on the upper
surface of sleepers near the point of derailment were associated with the greatest
displacement of the baseplate which probably occurred during the derailment.
These are likely to have obscured any scuffing and depressions associated
with previous shuffle. The observed marks and depressions show that, during
the derailment, the left-hand rail fixings moved laterally by up to 35 mm beyond
the static gauge. The measured static gauge and shuffle marks varied at each
sleeper, but when both are added together, the combined maximum measured
was 1508 mm at the point of derailment. This maximum included 4 mm of
wear on the left-hand rail, an amount within the limits permitted by Network Rail
standards.
85 As the greatest baseplate displacement had probably occurred during the
accident, it was not possible to determine the extent of any shuffle marks that
may have been visible before the accident. However, evidence of pre-derailment
shuffle is provided by witness marks left by sheared chairscrews found in five
sleepers after the accident. These chairscrew marks were beneath the baseplate
so would have not been visible before the accident.

Report 27/2014 27 December 2014

Liverpool Street
Key facts and analysis

Figure 16: Baseplate shuffle and sleeper marks

Report 27/2014 28 December 2014

Liverpool Street
86 The chairscrews had broken a short distance below the underside of the

Key facts and analysis

baseplate leaving a short length of chairscrew protruding downwards into the
sleeper. Subsequent movement of the baseplate was evidenced by grooves cut
into the upper part of the sleeper. These grooves remained after repairs were
undertaken immediately following the derailment and were seen by the RAIB
while permanent repairs were being undertaken in November 2013 (figure 17).

Groove evidencing
movement in two directions Initial chairscrew
position

Extent of shuffle marks Outline of initial baseplate

prior to derailment position prior to derailment
remedial works

Figure 17: Marks caused by pre-derailment shuffle (post-derailment marks shown on figure 18)

87 These pre-accident grooves were much wider than the chairscrews indicating that
the sides of the groove had been worn outwards by many cycles of movement.
Some of the grooves had developed in more than one direction and this is
inconsistent with them being caused only by movements during the derailment.
88 A sample of the sheared chairscrews were examined after the derailment and the
fracture faces examined. The sample included a screw with a fully rust covered
fracture surface, a screw with a clean fracture across the full width of the screw
and a partially rusted screw with a remainder of the surface still shiny.
89 The fully rust covered fracture surface indicates there were screws which had
fractured for some time before the derailment and remained in the baseplate as
described in paragraph 87. The screws with a clean surface across the full width
of the fracture face indicate the bolt had recently failed, possibly as a result of
an overload as a consequence of the derailment. The part rusted fracture face
indicates a crack which had formed in the screw allowing rust to penetrate, but
which finally failed under load during the derailment.

Report 27/2014 29 December 2014

Liverpool Street
 90 It is possible to prevent shuffle by screwing metal blocks, known as gauge
Key facts and analysis

stops, into the sleeper against the outer edge of the baseplate4. These provide
additional resistance to the outward movement of the baseplate caused by train
loading. No gauge stops were found after the accident, there was no evidence
of any being fitted in the past and, at the derailment location, the distance
between the outer edge of the baseplate and the sleeper end was too narrow to
accommodate gauge stops.
91 Where wide gauge caused by gradual deterioration is found, Network Rail
standard NR/L2/TRK/001 requires the fitting of tie bars as a short term control
measure. Tie bars clamp to the underside of both running rails and provide
additional restraint to further gauge widening (an increase in the distance between
the rails). The fitting of tie bars should be recorded and a permanent repair
should be made as soon as possible, and no longer than six months from fitting.
No tie bars had been provided in the vicinity of the derailment site, although they
had been provided elsewhere in the Liverpool Street station area.
92 Sleepers at the derailment site did not show any evidence of poor quality or
rotten wood. The condition of the sleeper wood was discounted as a factor in the
derailment.
93 As part of the repairs undertaken immediately after the derailment, Network Rail
turned the existing baseplates through 180 degrees and re-fixed them to the
existing sleepers. The asymmetrical chairscrew hole pattern in the baseplate
meant the chairscrews were then screwed into a previously unused part of the
sleeper. Turning the baseplates also meant that the baseplate edges were in a
slightly different position on the sleeper. It was therefore possible for the RAIB to
distinguish between the effects of shuffle before and during the derailment and
that which occurred later.
94 Shuffle continued to occur in the 10 months between re-fixing the baseplates
immediately after the derailment and the implementation of permanent repairs in
November 2013. The marks made by the lateral movement during post-accident
shuffle are shown in figure 18. The post-accident shuffle was associated with
sheared chairscrews which were found during the relaying works in November
2013 (figure 19).
95 The inspection regime intended to identify, and trigger correction of, shuffle is
described at paragraph 102.
96 The combined distance of 1508 mm between the rails (paragraph 84) is
less than the 1515 mm required for a standard wheelset to drop inside the
rails. Post- accident measurement confirmed that the derailed wheelsets on
vehicles 3 and 4 would all require 1515 ±1 mm to drop in between the rails,
so it is necessary for the gauge to be greater than the 1508 mm described in
paragraph 84. It is probable that the distance between the rails increased above
1508 mm due to rotation of the right-hand switch rail.

4
The use of gauge stops is not referenced within the Network Rail standards. However, the RAIB is aware that
gauge stops are fitted on Network Rail infrastructure. The Permanent Way Institute textbook ‘British Railway Track,
Volume 4, Plain Line Maintenance’, 7th Edition, describes the good practice of fitting gauge stops on curves where
high lateral forces are generated.

Report 27/2014 30 December 2014

Liverpool Street
 Key facts and analysis
Mark due to Outline of baseplate position after
post- accident shuffle derailment remedial works

Chairscrew positions after baseplate

turned through 180 degrees

Figure 18: Marks caused by post-derailment shuffle (pre-derailment marks shown on figure 17)

Full length chairscrews

Sheared chairscrews

Figure 19: Sheared & intact chairscrews

Report 27/2014 31 December 2014

Liverpool Street
Key facts and analysis

Figure 20: Derailment sequence

Report 27/2014 32 December 2014

Liverpool Street
97 The right-hand switch rail is intended to move laterally according to the direction

Key facts and analysis

in which 2035B points are set (figure 7). The rail support system allowing this
movement does not prevent the top of the right-hand rail moving outwards due
to rotation at the foot of the rail. The right-hand rail is not normally subjected to
sufficient gauge spreading force (paragraph 69) to cause movement of this type.
However, in this instance, it is likely that the displaced position of the left-hand rail
meant that the outer edges of the right-hand wheels were running on the inner
edge of the switch rail. In this position, the downward load from the wheels, the
angled shape of the wheels and the curved rail head combining with the gauge
spreading force would result in the top of the switch rail being pushed outwards
(figure 20, inset).
Increased gauge spreading force
98 The increase in the gauge spreading force, arising from the static wide
gauge and rail deflection due to the reduced strength in the rail fixing,
resulted in additional deflection of the rails when train 1P18 passed over.
This was a causal factor.
99 The vehicle dynamics study commissioned by the RAIB following the Liverpool
Central and Ordsall Lane Junction derailments identified that widened track
gauge (paragraph 79) generates a greater gauge spreading force on the rails
than track of standard gauge. This tendency for widening gauge to cause higher
gauge spreading forces meant that loads on the track fixings at Liverpool Street
were increased due to the wide static gauge (paragraph 75). They then further
increased as movement of the outer rail due to fixing deterioration (paragraph 81)
led to a further gauge increase. There was also a tendency for gauge spread
forces to increase under successive wheelsets of a train. This was because each
wheelset was encountering track which had been subject to the gauge widening
effects of previous wheelsets.
100 Although gauge spread was occurring at the derailment site, outside this
immediate area the outer rail remained secured to its fixings so the rail sprang
back towards its original position after the wheelsets passed. This could occur
both between successive wheelsets of the same train and between successive
trains. However, the cumulative effect of successive movements led to increasing
fixing deterioration, and thus both permanent deformation (the wide static gauge)
and an increasing maximum deflection under load. These effects increased until
the deflection under load was sufficient to cause the derailment.
101 The on train data recorder shows that there was an application of power from
the locomotive at the time that vehicle 3 passed through the derailment site (a
normal occurrence). This application of power from the locomotive at the rear
would have changed the distribution of forces along the train and at the wheel/rail
interface and is a possible reason why this was the first vehicle to derail.

Report 27/2014 33 December 2014

Liverpool Street
 Inspection regime
Key facts and analysis

102 The inspection regime was not sufficient to detect track fixing deterioration; this
was because:
l manual inspections did not report, and possibly did not identify, wider than
normal static track gauge or indications of a loss of strength in the fixing
between the rail and sleepers (paragraph 103);
l automated track monitoring and associated data analysis did not identify the
combined effect of widened static track gauge and loss of strength in the fixing
between the rail and sleepers (paragraph 115); and
l no consideration had been given to providing an enhanced inspection regime
for the non-standard track layout, and consequently no special mitigation
measures had been implemented, (paragraph 128).
These are now considered in turn:
Manual inspections
103 Manual inspections did not report, and possibly did not identify, wider than
normal static track gauge or indications of a loss of strength in the fixing
between the rail and sleepers (paragraphs 75 and 81). This was a causal
factor.
104 For standard track (ie track not requiring a special inspection regime due to
an enhanced derailment risk), Network Rail standard NR/L2/TRK/001 requires
inspections to be carried out at regular intervals determined by the weight,
frequency and maximum permitted speed of trains using the track. The
requirements give a hierarchy of inspections which vary in interval and content,
each carried out by different grades of track maintenance staff.
105 The inspection requirements for standard track were being applied at the
incident site before the derailment and meant that a basic visual inspection was
undertaken at weekly intervals by a patroller. This inspection was intended to
identify and report defects which could affect the safety of trains in the following
four weeks. The basic visual inspection does not require measurement of static
gauge. A visual inspection would not have the accuracy needed to detect the
changes of static gauge (up to 20 mm, paragraph 122) which had been occurring
over a ten month period at the accident site.
106 Network Rail standard NR/L2/TRK/001 required patrollers to report shuffle if
the extent of marks had ‘visibly increased’. However, they were not required to
record the amount of shuffle so this requirement relied on patrollers remembering
the amount of shuffle visible during their previous inspection.
107 It is certain that shuffle was occurring before the derailment (paragraphs 85
and 87) and that shuffle marks were visible before the derailment. The rate at
which shuffle developed before the accident cannot be established so there is no
way of knowing whether it ever increased at a rate sufficient to cause the ‘visibly
increased’ shuffle which patrollers are expected to report (paragraph 105). If this
did occur, it is not possible to determine when it occurred. There is no mention
of shuffle on available inspection records for the 10 weeks before the accident
(Network Rail was unable to provide complete records for this period).

Report 27/2014 34 December 2014

Liverpool Street
108 Network Rail standard NR/L2/TRK/001 also requires that static gauge should be

Key facts and analysis

measured at sample locations during quarterly supervisors inspections by the
section manager [track] (SM[T]). The standard requires the SM[T] to select the
location of their sample check based on local knowledge of the track conditions.
The SM[T] stated that he would normally measure gauge every quarter mile
(approximately 400 metres), at either side of points or anywhere he considered
there to be a problem. The record of the most recent inspection before the
derailment, undertaken in November 2012, contains no record of excessive static
gauge at the derailment site. As standard NR/L2/TRK/001 does not require the
location of acceptable gauge measurement to be recorded, it is uncertain whether
a measurement was taken at the derailment site.
109 The standard also requires that both shuffle and shearing of chairscrews should
be recorded during the quarterly supervisor’s inspections. Both had been noted
in other inspections undertaken by the SM[T] in other areas near Liverpool Street
station, but none were reported at the derailment site on the November 2012
inspection, the last undertaken before the derailment. The extent of shuffle,
if any, present in November 2012 cannot be established due to uncertainties
described in paragraph 107.
110 A detailed inspection of the points, including assessing track gauge, should have
been undertaken by the SM[T] every 52 weeks in accordance with standard
NR/L2/TRK/001. This standard included a specific requirement to inspect for
shuffle and indications of gauge widening.
111 Network Rail uses Ellipse software to manage track inspection and maintenance
tasks. This software records required tasks, produces task lists and records when
the tasks have been completed.
112 A Network Rail audit dated November 2012 found no records of detailed
inspections for 2035A, 2035B or 2035C points in the 52 weeks prior to the
audit. The audit also noted that a clerical error meant that inspections of these
points were not included in Ellipse and so had not been included in the task lists
supplied to the SM[T]. There is no evidence of these inspections having been
undertaken in the period between the audit and the derailment. Network Rail was
unable to identify an assurance process (other than sample based audits such as
that carried out in November 2012) which checks that Ellipse contains the correct
inspection and maintenance tasks for all assets intended to be included within the
database.
113 Standard NR/L2/TRK/001 required the track maintenance engineer (TME) to
carry out a visual inspection of all track, on foot, every two years. It also required
the TME to undertake an annual cab ride over all passenger lines although this
was not required to include all routes through junctions, all loops or all platforms.
The last TME inspection which included the derailment site took place on foot in
May 2012.

Report 27/2014 35 December 2014

Liverpool Street
 114 Network Rail standards did not give a precise specification for the activities to
Key facts and analysis

be undertaken by the TME during his visual inspections. In May 2012, three
night shifts were allocated for the TME to inspect the line from Liverpool Street
to Bethnal Green, a distance of about 1.1 miles (1.8 km) and including all the
tracks shown on figure 2. This was not intended to allow the TME to undertake
a detailed inspection of all track in this area, but to allow sample checking and
examination of areas of specific concern. The TME had not identified any issues
at the site of the derailment before the accident.
Assessment of dynamic gauge widening
115 Automated track monitoring and associated data analysis did not identify
the combined effect of widened static track gauge and loss of strength in
the fixing between the rail and sleepers (paragraphs 75 and 81). This was a
causal factor.
116 Network Rail standard NR/L2/TRK/001 requires that the track inspection process
includes an assessment of dynamic gauge, the distance between the rails when
carrying train loads. The intervention limit for triggering maintenance activities
given in this standard relates to dynamic (not static) gauge. At the incident site,
the standard required track geometry, which includes measurement of dynamic
gauge, to be assessed at a ‘nominal planning interval’ of 24 weeks and a
maximum interval of 52 weeks.
117 In most locations, track recording trains measure dynamic gauge directly and
thus record the combined effect of wide static gauge and any rail movement
under train loads due to a loss of strength in the fixings. Wide static gauge and
dynamic movement were present at the derailment site (paragraph 75 and 81),
but they could not be measured by a track recording train because the train
does not record results when travelling at or below the maximum permitted
speed of 15 mph (24 km/h) at Liverpool Street. Network Rail’s response to a
previous recommendation which relates to the assessment of track that cannot be
measured by the track recording train is given at paragraph 164.
118 Where dynamic gauge cannot be measured using the track recording train,
standard NR/L2/TRK/001 requires manual measurements to be undertaken.
Although this is not fully detailed in the standard, this requires measurement of
the static gauge and an assessment of the rail movement under traffic. These
measurements are then combined to give the dynamic gauge necessary for the
comparison to the maintenance intervention limit (paragraph 116). Post-accident
assessment by the RAIB showed that these limits had been exceeded before the
accident, but it has not been possible to determine when this condition had been
reached (paragraph 107).
119 At Liverpool Street the static gauge was recorded using a manual recording
device known as an Amber trolley (figure 21). The trolley is lightweight to allow
handling by a single person and only applies a small load to the rails. This load is
not sufficient to replicate the dynamic movement caused by train loads and so the
trolley cannot be used alone to measure dynamic gauge.
120 The trolley measures and records track gauge, and other track geometry
characteristics, while it is pushed along the track. It alerts the operator when pre-
set track geometry intervention limits, including those for dynamic gauge, have
been exceeded. Although the intervention limit relates to dynamic gauge, the
trolley is only measuring static gauge.

Report 27/2014 36 December 2014

Liverpool Street
 Key facts and analysis
Figure 21: Amber trolley in use (not at accident site)

121 The use of the Amber trolley at Liverpool Street was first implemented in 2008
by the previous assistant track maintenance engineer, he supervised the
members of the TME’s technical team undertaking the task. The lengths of
track to be measured in each shift were provided on task lists generated by
Ellipse (paragraph 111). These lists omitted some areas of track at Liverpool
Street, but did include the derailment site. The last Amber trolley measurements
at the derailment site were made in March 2012, about ten months before the
derailment. This was within the 52 weeks maximum interval required by Network
Rail standards (paragraph 116).
122 Staff used the measurement alerts given by the Amber trolley to identify areas
of track where measurements indicated a need for action according to the
intervention criteria given in standard NR/L2/TRK/001. They recorded areas
requiring action on paper forms. The derailment site is not included on these
records, suggesting that, when measured in March 2012, the static gauge was
less than the 1455 mm at which standard NR/L2/TRK/001 required remedial work
to be planned5.
123 Although the Amber trolley records both gauge and the distance travelled as
it is pushed along the track, it was not possible to determine the actual gauge
measured at the incident site in March 2012. This is because the operators at
Liverpool Street did not always record the start position and/or pass continuously
from start to finish over each section of track to be assessed. Without an
accurate start point and distance travelled record, it is not possible to determine
the measured gauge at a particular location.
5
The planning of any work at this location should have taken account of the increased gauge (1441 mm compared
to the standard value of 1435 mm) intended by the track designer to assist train cornering on the tight curve.

Report 27/2014 37 December 2014

Liverpool Street
 124 The last Amber trolley runs undertaken at the derailment site before the incident
Key facts and analysis

occurred prior to the appointment of the current ATME. The current ATME had
not appreciated the lack of reliable information about the location of gauge
measurements recorded electronically by the Amber trolley until they were
highlighted during the Network Rail investigation into the derailment.
125 The RAIB did not find any evidence that dynamic movements under train loading
were being measured (or estimated) and added to the Liverpool Street Amber
trolley static gauge data to give dynamic gauge.
126 Although the use of the Amber trolley was introduced after the current TME had
been appointed, the TME was also not familiar with the limitations of the data it
produced. The TME did not review the process implemented by the previous
ATME and incorrectly believed it to be a complete alternative to the track
recording train. During the course of the investigation, the RAIB identified another
TME, at a different location, who had also not appreciated the limitations of only
using the Amber trolley data to assess dynamic track gauge.
127 The SM[T] understood the need for dynamic measurement of the track, but
wrongly believed that the complex layout and restrictive access arrangements
prevented any method of dynamic assessment. Safe access to the accident
location was restricted by the intensive train service. However, it is possible to
estimate dynamic movement, without the need to watch the passage of a train, by
measuring the shuffle marks present on the sleepers (paragraph 83, figures 18
and 19) during routine or enhanced inspections. The RAIB found no evidence
that the practicalities of obtaining improved access arrangements, or applying
alternative solutions for assessing dynamic gauge, had been considered at
Liverpool Street.
The need for an enhanced inspection regime
128 No consideration had been given to providing an enhanced inspection
regime for the non-standard track layout, and consequently no special
mitigation measures had been implemented. This is considered to be a
causal factor.
129 The track layout at Liverpool Street was designed and installed in the late 1980s.
It featured some very tight curves and a high proportion of non-standard points
to fit the layout within a tight railway corridor. Current Railway Group standard
GC/RT5021, ‘Track system requirements’, generally requires a minimum design
radius of 200 metres for track used by passenger trains. However, it permits a
150 metres radius in exceptional circumstances when a larger radius cannot be
provided. If a track radius of less than 200 metres is required, a check rail should
be provided where practicable to reduce the lateral forces applied to the outer rail.
130 The derailment occurred within the movable area of non-standard points which
incorporated a 125 metre right-hand radius curve and was not provided with a
check rail. This tight curve without a check rail exposed the track to increased
lateral forces which accelerated the normal rate of wear to the left-hand rail and
associated track fixings.

Report 27/2014 38 December 2014

Liverpool Street
131 Irregularities and very small cracks in the surface of the rail head, known as head

Key facts and analysis

checking, were found during an examination by the RAIB in September 2013
(figure 22). Head checking was not seen immediately after the accident, but the
surface had been smoothed by recent rail head grinding, a process intended to
grind off the metal containing small cracks before these grow large enough to
cause a broken rail.
132 Head checking is caused by the transfer of forces along the track at the wheel/
rail interface and is commonly found in terminal stations as a result of traction and
braking forces acting longitudinally along the rail. Head checking can also occur
on the rail head of tight radius curves at other locations, due to wheels slipping
on the rail head. This slipping happens because wheels on the inner and outer
rails rotate together, but the distance travelled along the inner rail is less than that
along the outer rail. The RAIB made enquiries with other Network Rail engineers
who stated that they would not enhance the inspection regime due to head
checking alone, but would consider doing so if there were also other indicators of
high lateral forces such as shuffle or broken fastenings.

Evidence of head
checking on rail

Figure 22: Head checking found on the running rail ten months after the derailment

133 Although Network Rail mandates the interval, type and content of track based
inspections for standard track found on the mainline network, it does not provide
explicit instructions for non-standard track such as the complex junction at
Liverpool Street station. However, standard NR/L2/TRK/001 does require the
TME, assisted by the SM[T], to identify safety risks arising from non-standard
track assets and to apply appropriate mitigation measures where higher risk is
anticipated6. Mitigation can include changes to the frequencies and/or content of
the inspection and/or maintenance regimes.

6
This is a summary of detailed requirements given in sections 4, 4.1, 6.1 and 6.3 of the introductory text to
standard NR/L2/TRK/001 issue 5.

Report 27/2014 39 December 2014

Liverpool Street
 134 At the time of the derailment, no additional risk from the track layout at Liverpool
Key facts and analysis

Street had been identified and therefore no special risk mitigation measures had
been implemented. The absence of any special mitigation measures to address
the additional risk factors at this location is considered to be a causal factor.
135 Network Rail has not identified any process which would require such an
evaluation, and any associated mitigation identified as necessary, to be
documented. Such a process would enable the review and independent checking
of the assessment.

Identification of underlying factors7

Failure to ensure asset risk management
136 Neither the TME, nor the SM[T] had identified the need for a non-standard
inspection and maintenance regime at Liverpool Street.
137 Network Rail’s standards require that a track inspection and maintenance regime
should be implemented that will ensure the operational safety of the line. Issue 3
of NR/L2/TRK/001 (dated 26 August 2008) stated:
‘The track inspection regime is based on risk, both safety and commercial,
deterioration rates, anticipated failure modes and the identification of work
needed so that it may be carried out in a planned way’.
138 Issue 4 of NR/L2/TRK/001 (dated 05 December 2009) expanded on this
requirement stating:
‘It is important that the most appropriate frequency and methods of inspection
are used that will provide the SM[T] and TME with the assurance that
the infrastructure remains fit to use at its designed line speed and collect
information that will monitor condition and drive timely maintenance’.
139 The need to assess derailment risk arising from the track asset was implicit in the
need to develop an inspection regime which was based on safety risk. However
this was made an explicit requirement for the SM[T] and the TME to identify and
manage risk arising from track assets in NR/L2/TRK/001 issue 5 (dated 02 June
2012). This stated:
‘The TME and SM[T] must be able to identify risk from the track assets, assess
those risks and take action to control them. These are continuous processes
that TME and SM[T]s must follow, using results of inspections and the full range
of track asset information that is available to them’.
140 The track layout at Liverpool Street utilises many non-standard design features
that were necessary to allow the track to fit within the tight railway corridor. These
include tight radius curves and intended gauge variations, sometimes close to or
exceeding those permitted by current design standards. Some of these features,
such as the small radius curve at the derailment site, were likely to cause an
increased risk of derailment. At the time of the accident no special inspection or
maintenance plan had been implemented to control the increased derailment risk.

7
Any factors associated with the overall management systems, organisational arrangements or the regulatory
structure.

Report 27/2014 40 December 2014

Liverpool Street
141 The SM[T] and TME both believed that there were no unusual derailment risks

Key facts and analysis

associated with the track at Liverpool Street. The TME was also not aware
that the track design at the derailment site included features likely to cause an
increased risk of derailment. They therefore implemented the inspection regime
given in standard NR/L2/TRK/001 for track without such features.
142 The SM[T] and TME each had nearly 30 years track maintenance experience, but
had not received formal training in identifying unusual track layouts likely to import
an increased risk of derailment. Such training would highlight the circumstances
where track geometry, non-standard track design or the local operating conditions
might lead to an increased derailment risk.
143 Without the ability to determine risk, the SM[T] and TME were reliant on the
transfer of knowledge, or intervention by others. The TME stated that there was
no overlap with, or handover from, the previous TME. During the course of the
investigation, the RAIB has noted that the extent of handover arrangements
for TMEs at other locations has varied and does not always include direct
communication between the previous and new postholders. Network Rail
standard NR/L2/TRK/001 issue 5 requires the IME to ensure this handover
occurs, but this was issued after the appointment of the current TME.
Failure to manage technical knowledge shortcomings
144 The IME did not appreciate, and so did not manage, shortcomings in the
technical knowledge of the TME and his team.
145 Issue 5 of Network Rail standard NR/L2/TRK/001 became effective on 01
September 2012 and introduced two explicit IME responsibilities which, if
effectively implemented, would have led to the recognition, and correction, of
shortcomings in the dynamic gauge assessment and track risk identification
processes. The standard stated that IMEs must:
‘confirm that SM[T] and TME can demonstrate knowledge and understanding of
identifying risk to the track assets, risk assessment and taking action to control
risk’
‘arrange training and coaching for SM[T]s and TMEs and address shortfalls in
knowledge and understanding of risk, applying controls where necessary to
mitigate shortfalls’
146 Neither the SM[T] nor the TME had the knowledge or training to enable them to
identify potential derailment risks arising from the track asset as required by the
Network Rail standard (paragraphs 140 to 142). The TME did not appreciate that
dynamic gauge was not being assessed (paragraph 126) as required for all track
by Network Rail standard NR/L2/TRK/001.
147 Formal Network Rail competence assessments stop at the track supervisor
level, the grade below SM[T]. SM[T]s are assessed against the track supervisor
requirements, but with no additional assessment to cover their more senior
responsibilities. Consequently there was no comprehensive formal competence
assessment or management system in place for SM[T] or TME grades, so it was
necessary for the IME to assess competence using his own knowledge or arrange
an assessment by others.

Report 27/2014 41 December 2014

Liverpool Street
 148 The IME came from an electrical engineering background and had limited
Key facts and analysis

experience of track design and maintenance. However, IMEs are not required to
have comprehensive knowledge of the technical disciplines represented by all the
maintenance engineers reporting to them. They can draw on the experience of
other staff to provide such knowledge when necessary.
149 The TME was in post when the IME was appointed and the IME understood that
he was an experienced and competent member of the track maintenance staff.
This opinion was shared by many others within the MDU who had worked with the
TME for many years.
150 Network Rail standard NR/L2/TRK/001 issue 5 became effective after the IME
was appointed and gave no explicit instructions about how it should be applied to
existing post holders. Neither did the standard give examples of how the TME’s
knowledge of risk related to track assets (eg derailment risk) could be assessed if
the IME does not have personal experience in track maintenance.
151 The IME could not provide examples of checks that he had carried out, or others
had carried out on his behalf, which would provide confirmation that the SM[T]
and TME could identify derailment risk. The IME believed he could request the
RAM[T] to assess the TME for competence in identifying derailment risk, but the
RAM[T] was not asked to do so.
152 The TME, IME and RAM[T] all acknowledged that the RAM[T] role includes
the mentoring of TMEs. However, without a specific request from the IME this
arrangement would have relied on the TME asking for assistance. At Liverpool
Street there was a good working relationship between the TME and the RAM[T].
However, the TME did not believe there was a need to seek additional assistance
and did not do so.
153 In summary, the RAIB has concluded that the absence of a process to ensure the
competence of the SM[T] and the TME to identify locations with high derailment
risk, and the lack of an assurance process to validate their assessment of this
risk, led to the adoption of an inappropriate inspection regime at Liverpool
Street. As a result the inspection regime that was implemented was not sufficient
to detect deterioration of the track fixings at the derailment site before the
track failed. This absence of a competence assurance process for the track
maintenance discipline is addressed in Recommendation 2. It is possible that
a comparable competence assurance process is also required for maintenance
staff with responsibilities similar to SM[T]s and TMEs but in other disciplines.
This is particularly relevant because an IME is unlikely to have comprehensive
knowledge about all the disciplines reporting to them,. This issue is addressed in
Recommendation 3.

Report 27/2014 42 December 2014

Liverpool Street
Observations8

Key facts and analysis

Inspection records
154 During the course of the RAIB investigation, Network Rail was unable to provide
records of some maintenance inspections. In particular, the RAIB found evidence
of work being carried out in response to findings from inspections for which the
corresponding paper records were not available. Witness evidence states that
the filing system for paper records was disorganised.
155 Network Rail standard NR/L2/TRK/001 requires maintenance reports and records
to be kept for a minimum of three years. These records not only confirm the
completeness of inspections and provide evidence for competency assessment,
they also provide historical asset data.
156 Historical asset data can be used as the input to trend analysis which, over time,
can provide information about the potential failure modes and condition of the
fixed infrastructure. In areas of non-standard infrastructure, such as the track at
Liverpool Street, trend data can be used as part of a control measure to ensure
the adequacy of the asset maintenance strategy.
Communication during the incident
157 It is possible that better communications between the service manager and
incident controller, both located at the AICC, could have led to an earlier
recognition that a derailment might have occurred, and thus earlier recognition
that the train needed a rigorous technical inspection before being allowed
to remain in service (paragraphs 29 to 39). Although these communication
shortcomings did not affect the consequences of the Liverpool Street derailment,
the consequences could have been serious in other circumstances.

8
An element discovered as part of the investigation that did not have a direct or indirect effect on the outcome of
the accident but does deserve scrutiny.

Report 27/2014 43 December 2014

Liverpool Street
 Summary of conclusions
Summary of conclusions

Immediate cause
158 The derailment occurred as train 1P18 negotiated the small radius curve at 2035B
points because the outer rail fixings of the left-hand rail on the curve were unable
to resist the lateral forces acting at the wheel/rail interface. The forces were
sufficient to widen the track gauge such that the right-hand wheels on the trailing
bogie of the third vehicle dropped between the rails (paragraph 59).

Causal factors
159 The causal factors were:
a. the wide static track gauge, which had developed on the curve between
2035B and 2035C points, reduced the margin that was available for the rails to
safely deflect when train 1P18 passed over (paragraph 75);
b. the degraded condition of the rail fixings, on the curve between 2035B and
2035C points, made it easier for the rails to move apart when train 1P18
passed over (paragraph 81);
c. the increase in the gauge spreading force, arising from the static wide gauge
and rail deflection due to the reduced rail fixing strength, resulted in additional
deflection of the rails when train 1P18 passed over (paragraph 98);
d. manual inspections did not report, and possibly did not identify, wider than
normal static track gauge or indications of a loss of strength in the fixing
between the rail and sleepers (paragraph 103, Learning points 2 and 3,
Recommendation 1);
e. automated track monitoring and associated data analysis did not identify the
combined effect of widened static track gauge and loss of strength in the fixing
between the rail and sleepers (paragraph 115, Learning points 3 and 4,
Recommendation 1); and
f. no consideration had been given to providing an enhanced inspection regime
for the non-standard track layout, and consequently, special mitigation
measures had not been implemented (paragraph 128, Recommendation 1).

Underlying factors
160 The RAIB has identified two underlying factors which led to this situation. These
are:
l Neither the TME, nor the SM[T], had identified the need for a non-standard
inspection and maintenance regime at Liverpool Street (paragraph 136,
Recommendation 2).
l The IME did not appreciate, and so did not manage, shortcomings in the
technical knowledge of the TME and the SM[T] (paragraphs 144 and 153,
Learning point 5 and Recommendations 2 and 3).

Report 27/2014 44 December 2014

Liverpool Street
Additional observations

Summary of conclusions
161 The RAIB found that records of maintenance inspections were not available due
to shortcomings in the filing system. The Network Rail standards required records
to be kept for a minimum of three years. These records confirm completeness
of inspections, assist competency assessment and can assist development of
maintenance strategies (paragraphs 154 to 156, Learning point 6).
162 Shortcomings in communications between staff in the AICC could have had
serious consequences in other circumstances (paragraph 157, Learning
point 1).

Report 27/2014 45 December 2014

Liverpool Street
 Previous RAIB recommendation relevant to this
Previous RAIB recommendations relevant to this investigation

investigation
163 The RAIB considers that earlier completion of the following recommendation
could have prevented this accident by identifying the wide static gauge and loss
of strength in the track fixing.
Derailment at Windsor and Eton Riverside, 11 October 2009
164 The RAIB investigation into a passenger train derailment at Windsor and Eton
Riverside station (RAIB report 11/2010, published on 05 August 2010) found that
dynamic track faults had not been identified by the track recording train or by an
alternative manual method. The derailment occurred at a location which was not
assessed by a track recording train because of the low speed.
165 Following this derailment, the RAIB made this recommendation:
Recommendation 2
Network Rail should develop a proposal for the periodic measurement of
dynamic gauge at potentially vulnerable locations not covered by a track
recording vehicle, and implement the identified measures, as appropriate.
166 The Office of Rail Regulation reported to the RAIB in October 2013 that this
recommendation had been ‘implemented by alternative means’ and provided the
following supporting information:
Network Rail has considered how potentially vulnerable parts of the network that
are not covered by Track Recording Vehicles (TRV) can be subject to dynamic
gauge measurement. Network Rail has delivered additional training to track
maintenance engineers and is evaluating the feasibility of direct measurement
of dynamic gauge at slower speed at potentially vulnerable locations not
covered by a track recording vehicle.
167 This investigation demonstrates that the interim solution was not effective at
Liverpool Street. Therefore, the RAIB is concerned that Network Rail had still to
develop an effective solution more than three years after the recommendation
was published (this concern was recorded in the RAIB 2013 annual report).
Network Rail has stated that a track recording vehicle capable of operating
at slow speeds was introduced on an experimental basis in 2014 and is
programmed for use in 2015.

Report 27/2014 46 December 2014

Liverpool Street
Actions reported as already taken or in progress relevant to

Actions reported as already taken or in progress relevant to this report

this report
Actions reported that address factors which otherwise would have
resulted in a RAIB recommendation
168 Network Rail undertook emergency repairs to the track in the immediate vicinity of
the track failure following the derailment. Since then, the sleepers and fixings in
the derailment area have been replaced.
169 Network Rail has reported that dynamic track gauge is now being recorded
throughout the Liverpool Street area and that an experienced track engineer is
assessing track condition, and making recommendations for enhancements to the
inspection and maintenance regime, at Liverpool Street.
170 Network Rail and Greater Anglia have stated that AICC staff have been briefed
on the need for effective liaison when dealing with events which may involve both
trains and track. Network Rail has nominated a route control manager to monitor
the quality of safety-critical communications within the AICC.

Report 27/2014 47 December 2014

Liverpool Street
 Learning points
Learning points

171 The RAIB has identified the following key learning points9:

1 The importance of prompt and effective communication between

train and track controllers when dealing with events which could be
associated with urgent safety issues (paragraph 162).

2 The need to check that Ellipse contains the correct inspection and
maintenance tasks for all assets intended to be included within the
database (paragraph 159d).

3 The need for effective management of track gauge, including in areas of

intensive train services, where this is not monitored by track recording
trains (paragraph 159d and 159e). Available techniques include:
l looking for visual indicators, (eg shuffle marks, running band position10
and head checking) which show that rails could be moving under
trains;
l identification
of broken or loose fixings (ie those which allow rails to
move when subject to train loads) by simple manual testing;
l observing rail behaviour while trains are passing;
l combining static gauge measurements with dynamic rail movements
assessed from visual indicators (possibly implemented only if visual
indicators suggest a problem);
l direct measurement of rail movement including the consideration
of technical solutions such as dynamic track gauges and electronic
monitoring where personnel access is limited (possibly implemented
only if visual indicators suggest a problem); and
l identification
of repeat defects indicative of track being inadequate for
the loads being imposed on it.

4 Staff using Amber trolley data should be aware that, although pre-
programmed to generate alerts related to dynamic gauge intervention
limits, the trolley is only recording static gauge and so could mislead
the operator. Users must assess dynamic movement by alternative
means and take this into account when assessing whether maintenance
intervention is necessary (paragraph 159d and 159e).

9
‘Learning points’ are intended to disseminate safety learning that is not covered by a recommendation. They
are included in a report when the RAIB wishes to reinforce the importance of compliance with existing safety
arrangements (where the RAIB has not identified management issues that justify a recommendation) and the
consequences of failing to do so. They also record good practice and actions already taken by industry bodies that
may have a wider application.
10
The strip of polished metal on the rail head indicating the path followed by train wheels. If displaced from the
usual position, the rail is not correctly aligned with the wheels, possibly because it is moving beneath trains.

Report 27/2014 48 December 2014

Liverpool Street
 5 The importance of IMEs managing the competence of SM[T]s and

Learning points
TMEs in accordance with the requirements given in NR/L2/TRK/001
(paragraph 160). This states that IMEs must:
l confirm that SM[T]s and TMEs can demonstrate knowledge and
understanding of identifying risk to the track assets, risk assessment
and taking action to control risk;
l arrange training and coaching for SM[T]s and TMEs to address any
shortfalls in their knowledge and understanding of risk;
l applying controls to mitigate risk where this is needed until any
shortfalls in SM[T]s and TMEs knowledge and understanding of risk
have been addressed; and
l arrange transfer of knowledge of high risk locations from previous
postholders to SM[T]s and TMEs when they take responsibility for a
new area or route.

6 The need to archive inspection and maintenance records as a minimum

in accordance with NR/L2/TRK/001, or as required to meet the
needs of the inspection and maintenance regime implemented by the
maintenance management team (paragraph 161).

Report 27/2014 49 December 2014

Liverpool Street
 Recommendations
Recommendations

172 The following recommendations are made11:

1 This recommendation is intended to reduce the risk of derailment arising

from the performance of non-standard track assets by establishing an
appropriate and independently checked inspection regime.
Network Rail should improve its management systems so that both
the identification of all non-standard track assets, and the associated
inspection regimes intended to manage any enhanced risk of derailment,
are recorded and independently checked. The scope of these inspection
regimes should include mechanisms for identifying indications of
possible gauge widening and, where necessary, assessing dynamic
track gauge (paragraphs 159d to 159f).

2 This recommendation is intended to introduce an assessment of staff in

track related safety critical roles where the role is reliant on judgements
made by that member of staff, to ensure they have the necessary
experience and knowledge to perform that role.
Network Rail should introduce a timebound programme for assessing
(and reassessing at intervals) the competence of its managers with
safety critical roles linked to track maintenance (eg section managers
[track] and track maintenance engineers), and addressing any shortfalls
arising (paragraph 160).

3 This recommendation is intended to establish whether it is appropriate to

extend the aims of recommendation 2 beyond the track discipline.
Network Rail should introduce a timebound programme for the review
of the processes used for assessing (and reassessing at intervals)
the competence of managers with safety critical roles linked to the
maintenance of assets other than track, and addressing any shortfalls
arising (paragraph 160).

11
Those identified in the recommendations, have a general and ongoing obligation to comply with health and
safety legislation and need to take these recommendations into account in ensuring the safety of their employees
and others.
Additionally, for the purposes of regulation 12(1) of the Railways (Accident Investigation and Reporting) Regulations
2005, these recommendations are addressed to the Office of Rail Regulation to enable it to carry out its duties
under regulation 12(2) to:
(a) ensure that recommendations are duly considered and where appropriate acted upon; and
(b) report back to the RAIB details of any implementation measures, or the reasons why no implementation
measures are being taken.
Copies of both the regulations and the accompanying guidance notes (paragraphs 200 to 203) can be found on the
RAIB’s website www.raib.gov.uk.

Report 27/2014 50 December 2014

Liverpool Street
Appendices

Appendices
Appendix A - Glossary of abbreviations and acronyms
AICC Anglia Integrated Control Centre
ATME Assistant track maintenance engineer
DRAM Director of route asset management
IECC Integrated Electronic Control Centre
IMDM Infrastructure maintenance delivery manager
IME Infrastructure maintenance engineer
MDU Maintenance Delivery Unit
RAIB Rail Accident Investigation Branch
RAM[T] Route asset manager [track]
RMD Route managing director
SM[T] Section manager [track]
TME Track maintenance engineer

Report 27/2014 51 December 2014

Liverpool Street
 Appendix B - Glossary of terms
Appendices

All definitions marked with an asterisk, thus (*), have been taken from, or are based on extracts from, Ellis’s British
Railway Engineering Encyclopaedia © Iain Ellis. www.iainellis.com.

Anglia Route A sub division of the Network Rail infrastructure and the
associated resources for operation and maintenance. The
geographical area includes London to Shoeburyness,
Southend, Norwich, Cambridge, King’s Lynn and Great
Yarmouth.
Back (of a wheel) The reverse face of a rail wheel not normally visible from the
side of the train and not in contact with the running rail.
Basic visual A visual inspection of the track, carried out on foot, which aims
inspection to identify any immediate or short term actions that are required.
Often referred to as a track patrol.
Baseplate A metal plate which supports the rail on the sleeper.
Bogie (on incident An assembly of two wheelsets in a frame which is pivoted at the
vehicles) end of a long vehicle to enable the vehicle to go round curves.
Chairscrew A steel screw which attaches the baseplate to the wooden
sleeper.
Check rail A rail or other special section provided alongside a running
rail to give guidance to flanged wheels by restricting lateral
movement of the wheels.*
Colour light signal A railway signal which uses coloured lights to indicate whether
the driver has to stop, needs to be prepared to stop or can
proceed without restriction. The lights may show:
l Green - proceed, the next signal may be displaying green or
yellow;
l Yellow - caution, be prepared to stop at the next signal as it
may be displaying a stop signal when you reach it; and
l Red – stop.

Control centre of A system used by control centre staff and others which provides
the future (CCF) a visual schematic display of train position, both real-time and
historic, and presents information on train running.
Detection A failsafe arrangement that proves that a set of points are
correctly set in position. Correct detection must be obtained
before trains can pass over the points.
Driving van trailer An un-powered rail vehicle with a driving cab at one end which,
when attached at the opposite end of a train to the locomotive,
allows a single locomotive to haul the train in both directions.
Dynamic gauge The distance measured between the running rails while the
track is under load from a train. (See also static gauge.)

Report 27/2014 52 December 2014

Liverpool Street
Electric multiple An electric train consisting of one or more coaches, including at

Appendices
unit least one powered vehicle, with driving cabs at each end, which
can be coupled to other units and operated as a single train.
Ellipse A computer based asset management system used by Network
Rail to record and prioritise what maintenance is work required
to be done and when it needs to be done by.
Fishplate Specially cast or forged steel plates used in pairs to join two
rails at a ‘fishplated’ rail joint.*
Foot (of rail) The lower part of a rail section.*
Gauge (of track) The distance measured between the inside faces of the running
rails.
Gauge widening (of An increase in track gauge as a result of intended design or
track) unintended rail movement.
Head checking A more general term for a rolling contact fatigue (RCF) defect
found in the running band of the rail head.*
Incident controller A Network Rail employee who manages Network Rail’s
response to incidents, liaising with railway industry and other
parties as appropriate.
Integrated Control The co-location of Network Rail and train operating control
Centre centres.
Integrated A type of signal control system that controls the points and
Electronic Control signals for a whole route or a large geographical area by
Centre (IECC) electronic means. The signallers’ interface is normally a
monitor, keyboard and pointing device.
Intervention limit The threshold value of a defect at which remedial action is
required.
Joint (rail) A mechanical joint between two rails held in line and supported
by the fixing of two plates either side of the adjoining rail ends.
Leading bogie (or The bogie (or wheel) at the front of the vehicle (or bogie) in the
wheel) direction of travel. (See also trailing bogie (or wheel).)
Loss of detection Relating to points, the situation that exists when points are
not proved to be in the position commanded by the signalling
system.
On Train Data A data recorder fitted to a train that records information on
Recorder (OTDR) the status of train equipment, including speed and brake
applications.
Patroller A competent person whose duties are to carry out a basic visual
track inspection and minor maintenance.
Plain track A section of railway track which does not include any points.

Report 27/2014 53 December 2014

Liverpool Street
 Points A pair of moveable switches that direct trains between two
Appendices

tracks; two pairs of points are required to form a switch diamond

(figure 7).
Rail (head) The use of manual tools or a specialist train to grind the rail
grinding head to remove surface imperfections and reshape the rail to its
required profile.
Running band That part of the running rail upon which the majority of vehicle
wheels make contact. It appears as a shiny strip on the rail
head. The location and consistency of location of this strip can
tell an observer much about track geometry and track gauge.*
Running rail A rail that supports and guides the flanged steel rail wheels of a
rail vehicle.*
Run through To force a train wheel through the converging rails at a set of
(points) points when the points are not set for a movement from that
direction.
Shearing (shear Deformation (or fracture) in which parallel planes in a body
failure) remain parallel.
Shuffle (baseplate) The tendency of inadequately maintained baseplates on wood
sleepers to move laterally under traffic, so wearing the wood
away from under them.*
Signaller’s display The visual interface between the signaller and the signalling
system depicting a schematic railway layout, train positions and
status of the signalling in real time.
Signalling shift The senior post in the signalling centre operations room
manager responsible for managing the operation of the signalling centre
staff for that working shift.
Sleeper A beam made of wood, reinforced concrete or steel placed at
regular intervals at right angles to and under the rails. Their
purpose is to support the rails and to ensure that the correct
gauge is maintained between the rails.*
Spring clips The collective term for any rail fastening system that relies on
the spring action to perform its function.*
Standard track (as Track not requiring a special inspection regime due to an
used in this report) enhanced derailment risk.
Static gauge The distance measured between the running rails while the
track is not loaded by the presence of a train. (See also
dynamic gauge).
Switch (rail) The movable rail which forms part of a set of points.
Switch diamond Two tracks crossing at such an acute angle or curvature that
crossing moving rails are required to maintain clearance for the rail wheel
flange (appendix D).

Report 27/2014 54 December 2014

Liverpool Street
Technical riding A member of the train operator’s maintenance staff that usually

Appendices
inspector travels on board train services in order to provide a quick
response to minor train maintenance requirements.
Tie bar Adjustable metal bar temporarily fixed between running rails
and used to maintain gauge.
Tip (of a switch rail) The top corner at the moving end of a switch rail that is reached
first by a train travelling over the points in a diverging direction.
Track circuit block A signalling system which operates by automatically detecting
the absence of a train by electrical circuits through the track.
Track fastening A sub-component of the track fixing system used to secure
baseplates to sleepers or rails to baseplates.
Track fixings The mechanical system of components which position and
secure a rail to the sleeper.
Track recording A specially equipped train that automatically measures and
train stores track geometry information for the lines that it runs over.
Trailing bogie (or The bogie (or wheel) at the rear of the vehicle (or bogie) in the
wheel) direction of travel. (See also leading bogie (or wheel).)
Train services A member of the control centre staff employed by the train
manager operator to monitor train services and respond to incidents
relating to train services.
Wheelset Two rail wheels mounted on their joining axle.

Report 27/2014 55 December 2014

Liverpool Street
 Appendix C - Key standards current at the time
Appendices

Railway Group standard Track system requirements

GC/RT5021
Network Rail standard Inspection and maintenance of
NR/L2/TRK/001 version 5 permanent way
Network Rail standard Track Maintenance Handbook
NR/L3/TRK/002 version 7
Network Rail specification Competence and Training in Track
NR/SP/CTM/011 version 1 Engineering

Report 27/2014 56 December 2014

Liverpool Street
Appendix D - Switch diamond crossings

Appendices
D1 Diamond crossings are provided where one railway track must cross another
at the same level allowing trains to cross the opposing line, but not connect to
it. The angle between the railway tracks creates a diamond shape in the centre.
These diamond crossings are usually combined with points to form part of a more
complex junction arrangement as found on the approach to London Liverpool
Street (figures D1 and D2).

Figure D1

Figure D2

D2 To allow the rail wheel flanges to cross the opposing track, the rails are bent to
create flangeway gaps. Where the angle between the tracks is quite large the
flangeway gap is short enough between the point rails to provide continuous
guidance to the passing wheel. This design of diamond crossing, known as a
fixed diamond, does not require any moving parts.

Report 27/2014 57 December 2014

Liverpool Street
 D3 At certain locations where the running rails must cross each other at a much
Appendices

smaller angle, or if one or both tracks are curved, the distance between the point
rails becomes much longer. If the gap becomes too long there is a risk of the
wheel flange striking, or passing the wrong side of, the point rail (figure D3).

Figure D3

D4 At these locations switch diamonds are provided to avoid the need for gaps
between the rails. This type of crossing features two pairs of moveable rails to
provide a continuous path for train wheels as they cross between running rails,
thereby removing the need for the wheels to negotiate a gap (figure D4).

Intended route

Wheel guidance provided by

movable rails eliminating gaps

Rails moved to provide guidance

when alternative route is required
Figure D4

D5 These moveable rails can then by moved aside to allow the free passage of the
wheel flanges on a train crossing on the other track. The position of each pair of
moveable rails is therefore dependent on which of the two tracks is to be used by
a train and is set by the signaller. The correct positioning, locking and detection
of the moveable rails is ensured by the railway’s signalling system.

Report 27/2014 58 December 2014

Liverpool Street
This report is published by the Rail Accident Investigation Branch,
Department for Transport.
© Crown copyright 2014
Any enquiries about this publication should be sent to:
RAIB Telephone: 01332 253300
The Wharf Fax: 01332 253301
Stores Road Email: enquiries@raib.gov.uk
Derby UK Website: www.raib.gov.uk
DE21 4BA