Introduction
1 The sole purpose of a Rail Accident Investigation Branch (RAIB) investigation is to
prevent future accidents and incidents and improve railway safety.
2 The RAIB does not establish blame, liability or carry out prosecutions.
3 This report contains the findings of the RAIB investigation into the incident that occurred
at Lewes on 30 November 2005 when a train passed a signal at danger by approximately
150 metres. It is Part 2 of an investigation into adhesion-related incidents during autumn
2005.
4 On 25 November 2005, an incident occurred at Esher involving a train passing two
signals at danger. The circumstances of that incident bore some similarity to the incident
at Lewes. The drivers of both trains had alleged that severe adhesion problems had been
the cause of the overruns. RAIB has also undertaken an investigation into the incident at
Esher (reported in Part 1 of the autumn adhesion investigation).
5 During the early stages of the investigation into the two signal passed at danger (SPAD)
incidents, it became apparent that there had been a higher number of adhesion-related
SPADs and a much higher number of adhesion-related station overruns in autumn 2005
than had occurred in autumn 2004. A separate report, Part 3 of the autumn adhesion
investigation, has been prepared to address the causes of the high number of adhesionrelated incidents in autumn 2005. Some of the issues identified in the investigation of the
incident at Lewes have much wider relevance than for that incident alone. Where relevant,
this report into the incident at Lewes contains references to the analysis of performance
during autumn 2005 reported in Part 3.
6 Access was freely given to staff, data and records by Network Rail and Southern Railway
in connection with this investigation.
7 Appendices at the rear of this report contain:
l explanation of acronyms and abbreviations (Appendix A);
l explanation of technical terms (shown in italics the first time they appear within the
body of this report) (Appendix B);
l a list of relevant Railway Group Standards (RGS), current at the time the incident
occurred.
8 Reference is made in the report to levels of adhesion between wheel and rail. This is
normally expressed as a coefficient of friction (symbol µ). The lower the value of µ, the
lower the adhesion between wheel and rail. Typical values for µ for dry rail would be at
least 0.20. In wet weather, this can fall to 0.10. Under severe low adhesion conditions,
the value of µ can drop below 0.03. As trains rely on the coefficient of friction between
wheel and rail to stop, the level of adhesion available is critical to the rate at which the
train can decelerate. Many modern trains have four or five fixed braking rates available
to the driver, the lowest of which will normally achieve a deceleration rate of 0.3 m/s2
and
the highest a rate of at least 1.2m/s2
. A braking rate of 0.3 m/s2
can only be achieved if
the value of µ is at least 0.03. The value of µ would need to be at least 0.12 to sustain an
emergency braking rate of 1.2 m/s2
.
Summary of the report
9 The RAIB investigation into the SPAD at Lewes on 30 November 2005 which resulted in a
‘near miss’ with another train has been undertaken in parallel with an investigation into the
SPAD incident at Esher on 25 November 2005 and a general investigation into the causes
of adhesion-related station overrun and SPAD incidents during autumn 2005. This report
focuses on the results of the investigation into the Lewes incident alone.
10 Refer to Figure 2 for a diagram of the location of the incident. At approximately 19:07 hrs
on Wednesday 30 November 2005, train 2D45, the 18:54 hrs Southern Railway service
from Brighton to Hastings passed signal LW9 at danger at platform 3 in Lewes station.
Train 2D45 stopped over the crossover located to the east of Lewes station, passing signal
LW9 by a distance of approximately 150 metres and running through 75 and 77 points.
Signal LW9 was at danger to protect the movement of train 2F21, the 19:07 hrs service
from Lewes to Seaford, which was departing from Platform 5 and routed through 76 and
77 points towards the down line. The driver of train 2F21 heard train 2D45 approaching
in parallel on the down line and realising that the two trains were on a converging path,
stopped train 2F21 at the tips of 77 points, some 30 metres from the potential point of
conflict. The vigilance and prompt action of the driver of train 2F21 is commended.
11 Nobody was injured in the incident and there was no damage to the rolling stock. Points
75 and 77 were damaged when train 2D45 ran through them. After a conversation
between the signaller and the driver of train 2D45, the train was moved clear of 77 points
and the driver relieved of duty as part of Southern Railway’s standard response to serious
incidents. The driver of train 2D45 had made no allegation against the brakes on the unit
involved and a Southern Railway fitter confirmed that the unit was in a safe condition to
move. After a brake test, the train was taken at a maximum speed of 30 mph (50 km/h) to
Glynde where it terminated and passengers were conveyed to their destinations either by
following trains or by road transport. The train was taken empty to Eastbourne sidings.
12 The driver of train 2D45 was driving in accordance with the professional driving policy in
force at the time within Southern Railway. The actions of the driver were neither causal
nor contributory to the SPAD.
13 Southern Railway’s Class 377 fleet is equipped with a wheelslide prevention (WSP)
system, which has the objective of minimising stopping distances under low adhesion
conditions.
14 The Class 377 unit was also equipped with a sanding system, designed to work with the
WSP system to improve levels of adhesion. Before autumn 2005, Southern Railway had
recognised a weakness in the sanding characteristics of the Class 377 unit (sanding was
available for a maximum of 10 seconds, irrespective of the duration of WSP activity) and
had started a programme to extend sanding times to a maximum of 60 seconds (when the
WSP system was active). The unit involved in the SPAD at Lewes had not been modified
at the time of the incident.
15 Post-incident testing of the unit involved in the SPAD at Lewes by Bombardier (the
manufacturer of the Class 377 fleet) indicated that the key systems on the train, braking,
WSP and sanding had performed in accordance with the specification for those items of
equipment. The standards that apply to the design and operation of these systems, the
Train Operating Company’s (TOC) involvement with their specification, their optimisation
for low adhesion conditions and the way in which they are tested to demonstrate that they
are fit for purpose are issues that have been considered in the Part 3 report.
Rail Accident Investigation Branch
www.raib.gov.uk
Report 25 (Part 2)/2007
January 2007
16 The treatment of the railhead by Network Rail on the line where the incident occurred
involved water jetting and applying a layer of Sandite using a Multi-Purpose Vehicle
(MPV). Railhead treatment of Falmer bank had been undertaken approximately eight
hours before the SPAD occurred. Treatment commenced at a point approximately
800 metres beyond the location where train 2D45 started experiencing adhesion problems.
The effects of Sandite are gradually eroded with the passage of trains. The time lapse and
traffic density between treatment and the incident would have limited, if not negated, the
benefit obtained over those parts of the bank that had been treated by the time train 2D45
passed.
17 Railhead swabbing of the area where the incident had occurred (which was undertaken
soon after the incident) found localised and limited evidence of contamination from
vegetation and hydrocarbons. It is likely that the effect of the contaminant was
exacerbated by light rain, which commenced as train 2D45 approached Falmer bank.
18 Data gathered from the On Train Monitoring and Recording (OTMR) equipment from
the unit involved indicated that train 2D45 experienced severe low adhesion conditions
for a distance of approximately 2,500 metres. It is likely that available levels of adhesion
were less than 0.02 (see paragraph 8), whereas normal dry rail would offer at least 0.20.
Severe low adhesion conditions are discussed in the Part 3 report.
19 The actions of the signaller involved in the Lewes incident were neither causal nor
contributory to the SPAD. However, there are lessons to be learned with relevance to
the training of signallers in handling emergency Cab Secure Radio (CSR) messages and
the circumstances under which Network Rail performs routine drugs and alcohol testing.
There are also lessons to be learned with regard to co-ordination between Network Rail
and TOCs following near-miss incidents.
20 Three recommendations specific to the Lewes incident are made to improve safety, all of
which relate to matters arising from the incident, as referred to in paragraph 19. The Part
3 report contains a number of recommendations relevant to the causal and contributory
factors associated with the Lewes incident, but with broader application.
The Investigation
Summary of the incident
21 Shortly after 19:07 hrs on Wednesday 30 November 2005, train 2D45 passed signal LW9,
located at the end of Platform 3 at Lewes station, at danger.
22 Refer to Figure 2 for a diagram of the area. After passing Signal LW9 at danger, train
2D45 ran through 75 points which were not set for the passage of the train. Train 2F21
had departed from Platform 5 at Lewes station on time at 19:07 hrs and was approaching
76 points when the driver heard train 2D45 approaching and, realising that the two trains
were on a conflicting route, stopped some 30 metres from the point of conflict. Train 2D45
ran through 77 points (which had been set for train 2F21 to depart from Platform 5 towards
Seaford), stopping with the front of the train approximately 30 metres beyond the points.
23 Nobody was injured in the incident. Points 75 and 77 were damaged when train 2D45 ran
through them.
Background
24 the general layout of the area where the incident occurred. Lewes is
located at the junction of two routes, one from London via Haywards Heath and the other
from Brighton. The two routes have their own platforms at Lewes, the junction being
located to the east of Lewes station.
25 Lewes station has 5 platforms, with Platforms 1 and 2 being used by trains from and to
London respectively and Platforms 3 and 4 being used by trains from and to Brighton
respectively. Platform 5 can be used by trains running to and from Brighton but is also
used by trains approaching from the east that terminate at Lewes
26 Trains are signalled in accordance with the Track Circuit Block regulations and controlled
by Lewes Signal Box. A mixture of two and three aspect signalling is provided in the area
and the maximum line speed through Platform 3 at Lewes station is 10 mph (16 km/h)
because of limited clearances between passing trains and a bridge that crosses over the
line.
27 The train involved in the incident was the 18:54 hrs Southern Railway service from
Brighton to Hastings. The train reporting number was 2D45. Train 2D45 comprised a
Class 377 four-car ‘Electrostar’ unit No. 377 456. Class 377 units are equipped with a
brake controller with step 1, step 2, step 3(full service) and emergency positions, a WSP
system and equipment for depositing sand to assist in braking or traction. The WSP
equipment is a software-driven system which detects when the train is experiencing
adhesion difficulties and modulates the braking effort on sliding wheels in an attempt to
match the braking rate to the available adhesion. See paragraphs 79 and 80 for further
information on the WSP system. The WSP system on the Class 377 also triggers sanding,
providing that step 2, step 3 or emergency braking has been selected by the driver. At the
time of the incident, the duration of sanding on unit 377 456 was limited to ten seconds for
any discrete period of WSP activity.
28 The weather on 30 November was overcast and drizzly. At 19:00 hrs, the temperature in
the area was around 8°C.
Evidence
29 The RAIB took evidence from the following sources:
l OTMR download from unit 377 456 for the period of the incident;
l OTMR download from unit 377 456 for the period covering dynamic testing of the WSP
and sanding equipment on 7/8 December 2005;
l Staff involved in the incident;
l Network Rail;
l Southern Railway, the operator of train 2D45;
l Bombardier Ltd, who manufactured unit 377 456;
l Knorr Bremse, who supplied the braking, WSP and sanding equipment fitted to 377 456.
30 The RAIB also reviewed the investigation reports prepared by Network Rail and Southern
Railway into the incident and discussed with the Investigating Officer from each company
the contents of their report.
Events preceding the incident
31 On the evening of 30 November 2005, the train service in the vicinity of Lewes station was
operating normally. At around 19:00 hrs, the timetabled sequence of departures eastwards
was:
l Train 1F34 17:53 hrs Victoria to Hastings (scheduled departure 19:01 hrs) from
Platform 1;
l Train 2F21 19:07 hrs Lewes to Seaford from Platform 5;
l Train 2D45 18:54 hrs Brighton to Hastings (scheduled departure 19:10 hrs) from
Platform 3.
After the departure of train 1F34, the signaller at Lewes set the route for train 2F21 to
depart from Platform 5.
32 Meanwhile, at around 19:03 hrs, train 2D45 departed from Falmer, its last stopping point
before Lewes, approximately 4.5 miles (7 km) away. The route from Falmer to Lewes
involves the descent of Falmer bank which at its steepest drops at a rate of 1 in 84. Line
speed is 70 mph (112 km/h) for the first three miles and then reduces to 55 mph
(88 km/h). After departing Falmer station on time, the driver used the speed controller to
achieve and maintain the maximum line speed of 70 mph (112 km/h). The train did not
slip and accelerated normally to line speed during the three minutes following its departure
from Falmer. In the vicinity of Ashcombe track parallel (TP) hut, the driver applied the
brakes in step 1 to reduce speed for the 55 mph (88 km/h) permanent speed restriction
located 1000 metres ahead.
Events during the incident
33 As soon as the driver placed the controller into step 1, the WSP system became active
indicating that the level of adhesion between wheel and rail was insufficient to support
the level of braking required for a step 1 brake application. At this stage, the train was
approximately 2500 metres from Lewes station and travelling at 70 mph (112 km/h).
34 Train speed fell only to 60 mph (96 km/h) over the next 700 metres at which point the
driver (having used step 2 braking in the interim) employed full service (step 3) braking
with the train still 300 metres from the start of the speed restriction and 1800 metres from
Lewes station. A period of 23 seconds had elapsed between the driver selecting step 1 and
step 3 braking. After a further 22 seconds, the driver selected emergency braking with the
train approximately 1300 metres from Lewes station, and still travelling at 58 mph
(93 km/h).
35 Had conditions been normal, the initial Step 1 brake application would have slowed the
train for the 55 mph (88 km/h) speed restriction without the need for a higher braking
step. Had the driver applied the same braking technique under normal conditions as was
employed on the evening of 30 November, train speed would have fallen below 50 mph
(80 km/h) over the 700 metres following the initial brake application.
36 As the train exited Kingston tunnel signal LW7 came into view showing a single yellow
aspect. The driver made an emergency call via the CSR to the signaller at Lewes. At this
stage, the train was approximately 1200 metres from Lewes station and still travelling at
55 mph (88 km/h). The call was made between 19:06 hrs and 19:07 hrs.
37 The driver alerted the signaller to the fact that train 2D45 was sliding and asked the
signaller to clear the next signal (signal LW9 on Platform 3 at Lewes station, which
protects the junction where the line from Brighton converges with the line from London).
The signaller replied that another train (train 2F21) had already been signalled out of
Lewes station and that it would not be possible to clear LW9 as train 2F21 was already
departing.
38 The signaller attempted to make an emergency call via the CSR to stop train 2F21 before
it reached 76/77 crossover and crossed into the path of train 2D45. The CSR emergency
stop facility has two buttons, one red and one yellow. If a signaller wishes to send an
emergency stop message to a specific train, the red button is pressed. The system then
asks for the signaller to input the train reporting number of the specified train and, once
confirmed, the message is sent. If a signaller wishes to stop all trains within the area
controlled by the signal box (or within a designated zone), the yellow button is pressed
and, once confirmed, the emergency stop message is sent to all trains.
39 During the incident at Lewes, the signaller intended to send an ‘all trains stop’ message,
but pressed the red button on the CSR equipment in error. Thinking that the ‘all trains
stop’ message had been sent, the signaller took no further action to stop train 2F21. The
emergency message had not been sent to (or received by) train 2F21 as the CSR system
was waiting for a train reporting number to be keyed in by the signaller. Train 2F21
continued out of Platform 5 towards 76 and 77 points.
40 At around this time, train 2D45 entered platform 3 at Lewes travelling at a speed of
approximately 40 mph (65 km/h). It passed signal LW9 at danger and ran through 75
points. As it ran through the points, the rate at which the train was slowing increased
significantly.
41 The driver of train 2F21, which had departed from Platform 5 under clear signals, heard
train 2D45 approaching and, realising that the two trains were on a collision path, applied
the emergency brake on train 2F21. Train 2D45 ran through 77 points, stopping with the
front of the train located 30 metres beyond the points. Meanwhile, train 2F21 had stopped
with its leading wheels in the vicinity of 76 points, approximately 30 metres away from
train 2D45.
42 From the time that the driver of train 2D45 first applied the brakes to the time that the
train stopped, approximately 120 seconds had elapsed. During this period, the train
travelled approximately 2750 metres. The length of the overrun beyond LW9 signal was
approximately 150 metres.
Events after the incident
43 After train 2D45 had stopped, the driver and signaller conversed briefly. The signaller,
who was the only person on duty in Lewes signal box, had a number of actions to
perform as a result of which the RT3189 signal passed at danger form was completed
approximately 75 minutes later. It is necessary for the RT3189 form to be completed
before the train can be moved past another stop signal, which meant that passengers were
detained on board during this time. In the interim, the driver of train 2D45 had been asked
to move the train for a short distance to clear 77 points (which had also been run through)
so that an assessment of the damage could be made.
44 It was necessary to decide whether train 2D45 should be allowed to continue in service.
One of the questions asked during completion of the RT3189 form is whether the unit is
fit to continue. The driver of train 2D45 did consider that the unit was fit to continue in
service and the senior Southern Railway manager who attended the incident was aware by
this stage that other trains had experienced low adhesion conditions on Falmer bank. A
Southern Railway fitter confirmed that the unit was in a safe condition to continue. The
train could not be taken the short distance back to Lewes station because it had damaged
the points and was at risk of derailment if it reversed. The only other option would have
involved evacuating passengers from train 2D45 down ladders onto the track in the dark,
which would have been a hazardous operation.
45 The driver in charge of 2D45 was relieved on site as part of Southern Railway’s standard
response to a serious incident. Following a brake test, another driver took the train
forward at a maximum speed of 30 mph (50 km/h) to Glynde, where it terminated and
passengers were conveyed to their destinations either by following trains or by road
transport.
46 The driver of train 2D45 underwent drugs and alcohol screening, the results of which
were negative. The signaller was not subject to drugs and alcohol screening. This issue is
discussed further in paragraph 98.
47 Network Rail undertook railhead swabbing on Falmer bank. A total of 50 swabs were
taken between Ashcombe TP Hut and Lewes station, covering the entire section of line
over which 2D45 had experienced low adhesion conditions.
48 Unit 377 456 was taken to Eastbourne for an OTMR download and stabling overnight. A
static test of the WSP and sanding equipment was undertaken that night with a satisfactory
result and the sand boxes were checked and found to be almost full. On the following day,
the unit was returned to Lovers Walk Depot at Brighton for further testing.
49 Overnight on 7/8 December, and at the RAIB’s request, dynamic testing of the WSP and
sanding systems was undertaken on a test run between Brighton and Eastbourne and back.
The run was was subject to special monitoring by the signallers involved to ensure that it
could be undertaken safely.
Analysis – events during the incident
50 The RAIB determined that there were four principal issues to be investigated relating to
this specific incident:
a) the actions of the signaller;
b) the actions of the drivers (trains 2D45 and 2F21);
c) the condition of the railhead;
d) the performance of the rolling stock. The actions of the signaller
51 The regulation of trains at Lewes around the time of the incident was in accordance with
the working timetable.
52 Upon receipt of the emergency CSR message from the driver of train 2D45, the signaller
had very little time in which to react. In the time available, the signaller attempted to stop
train 2F21 by using the emergency call facility on the CSR. Paragraphs 38 and 39 explain
why the message was not sent.
53 Signallers are trained in the use of CSR emergency equipment. Their competence is
assessed before thay are appointed to a signal box and the assessment includes questions
relating to use of the equipment in emergency scenarios. Following appointment,
signallers are subject to continuous assessment to ensure that they remain competent
for all duties. The assessment cycle covers three years and includes a mixture of formal
assessment and gathering of evidence from situations that the signaller has encountered
while at work. Every 13 weeks, signallers on Network Rail’s Sussex route (where Lewes
signal box is located) are released for a day to attend the signallers’ training centre at
Redhill for a safety briefing, discussion of specific issues relating to their job and to sit an
assessment paper based on the topics selected for discussion.
54 Another element of the continuous assessment of signaller’s competence involves
Signalling Managers asking questions on how signallers would respond to specific
scenarios, including the use of CSR equipment in an emergency. The signaller involved
in the incident at Lewes had been subject to such an assessment on the use of CSR
emergency equipment on 20 October 2005 (6 weeks before the incident) and had correctly
answered all questions.
55 Questioning of experienced signallers on how they would respond to emergency situations
can be undertaken in different ways. More benefit can be gained if the questioning
requires the signaller to describe his or her actions in detail (e.g. requiring the signaller
to describe how an emergency CSR call would be made rather than simply accepting the
response that an emergency CSR call would be made). However, the use of any form of
questioning has its limitations and particularly where it relates to the individual’s response
in the stressful circumstances of an emergency
56 Given that use of the CSR emergency equipment is comparatively rare, it is in general
difficult for signallers to gain any practical experience. The Sussex Route’s training centre
for signallers at Redhill has a simulator, which enables the instructor to set up a range of
operating scenarios for signallers to practise their skills. Operating scenarios involving use
of the emergency CSR equipment can be simulated. Currently, the centre is not being used
to assess signallers in use of the CSR emergency equipment during the training day they
spend at Redhill every 13 weeks. The centre is rarely used by experienced signallers; its
principal role is in training new signallers. The assessment of signallers in responding to
emergency situations continues to be based on a set of theoretical questions
The actions of the driver – train 2D45
57 The actions of the driver of train 2D45, described in paragraphs 33-37, have been analysed
using data obtained from the train’s OTMR equipment.
58 As soon as the driver placed the brake controller into step 1, the WSP system became
active, indicating that the level of adhesion available was insufficient to support the brake
demand being made. The driver was aware that the speed of the train was not decreasing
and progressively increased the brake to step 2, step 3 and emergency.
59 This technique was in accordance with Southern Railway’s professional driving
policy. The initial brake application had been well in advance (1000 metres) of the
commencement of the speed restriction of 55 mph (88 km/h), although the route
description issued by Southern Railway for the Brighton to Lewes route advises drivers to
‘bring their trains under control’ (ie perform a running brake test) at least one mile before
the commencement of the speed restriction in autumn. Having started to experience WSP
activity, the driver increased the braking effort and allowed the WSP system to manage the
wheelslide.
60 In comparison with the actions of the driver involved in the SPAD at Esher on 25
November 2005 (see the Part 1 report) the driver of train 2D45 held each brake step for a
relatively long period (11-12 seconds as compared with 6-7 seconds on the train involved
in the Esher incident) before moving to the next step. However, the actions of the driver
at Lewes were guided initially by the need to achieve only a modest reduction in speed
from 70 mph (112 km/h) to 55 mph (88 km/h) for a permanent speed restriction whereas at
Esher, the driver was braking to stop at a red signal.
61 On realising that there was a possibility that signal LW9 could be passed at danger, the
driver made an emergency call to the signaller via the CSR. The RAIB has considered
whether the CSR emergency message could have been sent earlier, which might have
allowed the signaller to stop train 2F21 before it departed from Lewes. Figure 3, shows a
timeline of the key events leading up to this incident:
62 As signalling between Falmer and Lewes is of the three aspect variety, it was not until the
signal preceding LW9 (signal LW7) was sighted that the driver knew whether there was a
possibility of passing Signal LW9 at danger. The timeline in Figure 3 shows that within
20 seconds of selecting emergency braking, and at the same time as the train passed the
AWS magnet located 183 metres from Signal LW7, the driver initiated the emergency call.
But it was only within this period of 20 seconds that two key facts would have become
apparent to the driver:
l the train was still decelerating at a very slow rate, despite the recent selection of
emergency braking;
l signal LW9 at Lewes station was at danger.
63 The driver therefore had no reason to make an emergency call to the signaller earlier.
64 The driver did not sound the train horn as train 2D45 approached Lewes station. The
driver was unable to control the speed of the train at this point and the rule book states
that in such circumstances, the driver should sound a series of long blasts on the horn.
RSSB, Rule Book GE/RT 8000, Module TW1, Preparation and movement of trains, Clause 15.5. Issue 4, April
2005.
Caution at AWS magnet for LW7
signal. CSR emergency call initiated
1:07:05
Emergency
brake
1:0:5
1:07:15
Step
brake
1:0:2
Step 2
brake
1:0:12
1:0:00 1:07:00
Step 1
Brake
1:0:00
Figure 3: Timeline of key events before driver of 2D45 made CSR emergency call to signaller. It is likely that the driver of train 2F21 would have been alerted earlier to the approach
of train 2D45 had the driver of train 2D45 sounded the horn on passing through Lewes
station. In the event, this had no effect on the outcome of the incident as the driver of
2F21 was alerted to the approach of train 2D45 by the noise of the blowdown valves
operating on unit 377 456 as the WSP system applied and released the brakes. In other
circumstances, it would have been more critical for the driver of train 2D45 to have
complied with the rule book requirement as the train approached and ran through Lewes
station.
The actions of the driver – train 2F21
65 The actions of the driver of train 2F21 almost certainly prevented a collision taking place.
Having left Lewes station under a green aspect, the driver had every expectation of a
normal journey towards Southerham Junction and the Seaford branch. When the noise of
the blowdown valves operating on train 2D45 became apparent, the driver of train 2F21
realised immediately that the two trains were on a collision course and applied the brakes.
The presence of mind and rapid reactions of the driver of train 2F21 were critical in
averting a collision.
The condition of the railhead
66 Network Rail identifies low adhesion sites in the sectional appendix that is prepared for
each route. In order to qualify for inclusion in the sectional appendix, a site must have
experienced more than two SPADs or more than four overruns in the last three years as a
result of leaf-fall or must have been specifically identified as a low adhesion site through
other means such as a review of OTMR data or feedback from maintenance staff or TOCs.
67 The down line between Falmer and Lewes is not included in the sectional appendix as
a low adhesion site. However, Southern Railway issues route descriptions to drivers to
provide them with key information about the routes over which they operate, including the
general layout of the line, names of specific landmarks (e.g. footpath crossings) and the
location and nature of hazards such as sections of the route where drivers might experience
low adhesion during autumn. Falmer bank is identified as a low adhesion site by Southern
Railway in the relevant route description. The driver of train 2D45 was aware that
Falmer bank was a location that could be affected by low adhesion and while not exactly
following the guidance provided by Southern Railway (see paragraph 59), still made a
brake application well before the point at which it would have been necessary had the
objective only been to slow for the speed restriction ahead.
68 Network Rail includes the down line between Falmer and Lewes in its railhead treatment
programme. On 30 November 2005, the MPV had run according to schedule, passing over
the down line between Brighton and Lewes at around 11:00 hrs, water jetting and laying
Sandite from Signal LW7R to Lewes station. There was no rail head treatment of the
first 800 metres over which train 2D45 experienced adhesion difficulties. However, given
that the effects of Sandite are gradually eroded with the passage of trains, the time lapse
between treatment and the incident and the frequency of services operated over the route
would have reduced, if not eliminated the benefit from rail head treatment.
69 During the autumn period, Network Rail operates a predictive system for railhead
conditions throughout the country, based on information supplied by the specialist
environmental and weather organisation, ADAS UK Ltd. The prediction takes account
of the likely influence of weather conditions (rain, frost, high winds) on leaf fall and the
extent of leaf fall is used as the means for predicting railhead conditions. The network
is divided into 16 areas and the prediction is area-specific. Network Rail takes further
precautions such as providing additional railhead treatment in known problem areas when
severe conditions are forecast. However, for 30 November 2005, no specific adhesion
problems were forecast for the Lewes area.
70 Railhead swabbing was undertaken almost immediately after the incident, covering the
area from Ashcombe TP hut (where the driver first made a brake application) to signal
LW9 at Lewes station. The results of the analysis were as follows:
l Ashcombe TP hut to signal LW7R, a distance of approximately 700 metres. There was
evidence of degraded leaf litter/vegetation and also sand. There was also some evidence
of hydrocarbons (diesel fuel oil).
l A27 road bridge to signal LW7, a distance of 1000 metres. Virtually no
contamination was found, but there was evidence of sand or Sandite.
l Signal LW7 to signal LW9, a distance of approximately 500 metres. Limited evidence
of leaf litter/vegetation with sand/Sandite found on the same swabs. Two swabs showed
trace elements of hydrocarbons (diesel fuel oil).
71 It has not been possible to establish why hydrocarbons were present in the swabs. The
analysis indicated that the source was diesel oil rather than mineral oil. It is possible that
diesel could have been dropped from a locomotive, the MPV or other on-track machine.
It is unlikely to have come from a diesel multiple unit because no diesel services were
scheduled to operate on the route at that time.
72 A wheel flange lubricator is provided in the vicinity of Kingston tunnel (approximately
1200 metres from Lewes station) for the purposes of minimising abrasion of train
wheels on the curves approaching Lewes station. After the incident it was found that
the lubricator was dispensing too much grease and Network Rail took it out of use. The
evidence from the OTMR download did not indicate any change in braking performance in
this area which, in any case, would only have affected the latter stages of the slide.
73 Overall, it is clear from the swab analysis that some railhead contamination was present.
74 As the train experienced problems as soon as the driver selected step 1 braking, this
indicates that the level of adhesion available at the wheel/rail interface was below 0.03.
The braking rate in step 1 is 0.3 m/s2 and requires an adhesion level of at least 0.03 to be
available for braking to be effective (see paragraph 8). The braking rate of train 2D45, shows that it encountered severe low adhesion conditions over a
distance of approximately 2500 metres.
75 The level of adhesion available for train 2D45 changed just before its descent of Falmer
bank. The train that ran over the same route before 2D45, the 18:34 hrs Southern
Railway service from Brighton to Seaford, also experienced WSP activity on Falmer
Bank. However, this train, which comprised a three car Class 377 unit, did not experience
WSP activity until the brake controller was placed in step 2 (braking rate 0.6 m/s2
). The
point at which the two drivers braked was almost identical. It can therefore be inferred
that adhesion conditions had deteriorated from a value between 0.03 and 0.06 when the
previous train passed to below 0.03 when train 2D45 passed.
Rail Accident Investigation Branch
www.raib.gov.uk
17 Report 25 (Part 2)/2007
January 2007
The most likely explanation for this change is that the drizzle that had commenced after
the passage of the previous train had exacerbated the impact of the rail head contaminants
present.
76 The only other explanation for the difference in performance between train 2D45 and
the preceding train is that the wheels of train 2D45 were contaminated. It has not
been possible to establish whether this was the case as the wheels were not swabbed
immediately after the incident. When the unit was inspected at Eastbourne sidings the
following day, discolouration was found on the wheels, which might have been evidence
of contamination. However, the train had been moved from Lewes to Eastbourne before
the wheels were inspected and because it would not have been possible to establish when
the possible contamination had occurred, swabbing was still not undertaken.
77 The RAIB considers that wheel contamination is unlikely to have been a cause of the
incident because:
l the train had not experienced any problems with wheel slip or slide immediately before
the incident;
l the preceding and following trains also suffered adhesion difficulties on Falmer bank,
suggesting that the source of the contamination was the rails rather than train wheels;
l if the source of contamination had been the wheels alone, the action of the WSP
equipment in releasing and applying brakes and creating friction between wheel and rail
would probably have had the effect of cleaning the wheels over a distance shorter than
the length of slide experienced by train 2D45 on 30 November.
The performance of the rolling stock
78 The Class 377 unit is equipped with dynamic and friction brakes, a WSP system and
sanding. The two types of braking are blended. The dynamic brake, acting through the
traction motors, applies initially. The friction brake is blended in as the train slows down.
The dynamic brake is inhibited and the train continues on friction braking alone if WSP
activity commences.
79 Once active, the WSP system seeks to monitor the continued presence of wheelslide by
releasing the brakes on a ‘test’ or ‘reference’ wheelset. Using measurements of speed and
acceleration derived from this wheel, the WSP system is able to estimate the true speed
of the train, which is then compared with the speed of the other wheels (which still have
brakes applied). In this way, the degree of wheelslide can be measured.
80 When wheelslide activity is detected, the WSP system endeavours to maximise and
influence the adhesion available to the train by applying and releasing the train brakes.
Wheel rotational speed is not permitted to drop below real speed by more than 20% before
the brakes are released. Wheel rotational speed is allowed to climb back towards ‘real’
speed and the brakes are applied again. The process of applying and releasing the brakes
is achieved through blowdown valves, which manage the rapid changes in air pressure
needed for the process. It has the effect of limiting wheelside while permitting a degree
of conditioning of the railhead, thereby providing a slightly cleaner rail surface for later
vehicles on the same train.
Rail Accident Investigation Branch
www.raib.gov.uk
18 Report 25 (Part 2)/2007
January 2007
81 Figure 4 shows estimated train speeds during the incident derived from the following
sources:
l Calculated speeds based on the times at which the train passed the AWS magnets, which
are located on the approach to the signal (Curve 1, green). This has been used as a
measure of train speed that is independent of the figures estimated by the WSP system
and reflected in the OTMR data.
l ‘Actual’ train speed derived from the OTMR (Curve 2, purple). In practice, this value is
an estimate of ground speed based on information about the performance of the
reference wheelset (see paragraph 79) obtained from the brake control unit (BCU).
