Conference PaperPDF Available

LEVEL CROSSING SAFETY IMPACT ASSESSMENTS (LCSIA) FOR VEHICLE AND PEDESTRIAN CROSSINGS

Authors:

Abstract

While the number of deaths and injuries at level crossings in New Zealand is relatively low compared with the national road toll and injury burden, the high severity of crashes involving trains makes it a key 'safe system' focus. It is also alarming that the proportion of crashes involving pedestrians and cyclists at level crossings has been increasing over recent years. This in part is due to the construction of a number of cycleways and shared paths that travel alongside railway lines.. In the past KiwiRail has relied primarily on the ALCAM crash estimation model to assess the increased risk of crashes at crossings due to a change in use. While ALCAM is one of the better developed level crossing models internationally, it does have it's limitations when used in isolation. The ALCAM documentation specifies that other information such as incident data and the opinions of locomotive engineers should be considered in assessing risk. In practice thee factors have rarely recieved equal importance with the ALCAM risk ratings. ALCAM also does not pickup in sufficient detail the safety impacts created by the surrounding transport network. To better inform decisionmaking KiwiRail has developed a wider assessment process that includes these factors, called a Level Crossing Safety Impact Assessment (LCSIA). This new process has been found to better reflect the crash risks at level crossings under different changes in use. This paper outlines the LCSIA process and the learnings that have occurred since it was first introduced in 2016.
Turner, Cook, Bosher 1
LEVEL CROSSING SAFETY IMPACT ASSESSMENTS (LCSIA) FOR VEHICLE AND
PEDESTRIAN CROSSINGS
Shane Turner
Abley Ltd
137 Victoria Street, Christchurch, New Zealand
Phone +64 27 4955048: shane.turner@abley.com
Eddie Cook
KiwiRail
Wellington, New Zealand
Phone: +64 4 4984495; E-mail: eddie.cook@kiwirail.com
Shaun Bosher
Stantec NZ
6 Hazeldean Road, Hazeldean Business Park, Christchurch
Phone: +64 21 330910; E-mail: shaun.bosher@stantec.com
Word count: 6,235
Submission Date: 1st August 2019
ABSTRACT
While the number of deaths and injuries at level crossings in New Zealand is relatively low
compared with the national road toll and injury burden, the high severity of crashes involving
trains makes it a key ‘safe system’ focus. It is also alarming that the proportion of crashes
involving pedestrians and cyclists at level crossings has been increasing over recent years. This
in part is due to the construction of a number of cycleways and shared paths that travel alongside
railway lines.. In the past KiwiRail has relied primarily on the ALCAM crash estimation model
to assess the increased risk of crashes at crossings due to a change in use. While ALCAM is one
of the better developed level crossing models internationally, it does have it's limitations when
used in isolation. The ALCAM documentation specifies that other information such as incident
data and the opinions of locomotive engineers should be considered in assessing risk. In practice
thee factors have rarely recieved equal importance with the ALCAM risk ratings. ALCAM also
does not pick-up in sufficient detail the safety impacts created by the surrounding transport
network. To better inform decisionmaking KiwiRail has developed a wider assessment process
that includes these factors, called a Level Crossing Safety Impact Assessment (LCSIA). This
new process has been found to better reflect the crash risks at level crossings under different
changes in use. This paper outlines the LCSIA process and the learnings that have occurred
since it was first introduced in 2016.
Keywords: Level Crossings, Railway Crossings, Pedestrian Level Crossings, Pedestrian Railway
Crossings, ALCAM
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Turner, Cook, Bosher 2
INTRODUCTION
There are over 2000 public level crossings and a large number of private level crossings in New
Zealand. While there are relatively few motor vehicle and pedestrian/cyclists crashes at level
crossings (compared with the rest of the road network), the consequence of a crash at a level
crossing is often death or a serious injury, and regularly draws media attention, especially if it
involves a vulnerable road user such as school pupil or an elderly person.
In New Zealand level crossings are managed using the same “safe system” approach that is
applied to other transport infrastructure. Namely, it is important to remember that humans make
mistakes (but shouldn’t be disproportionately punished for them) and are vulnerable to injury,
requiring a focus on harm minimisation. Applying this thinking to level crossings involves
considering the behavioural aspect of human interactions with level crossings and applying
appropriate infrastructure (e.g. engineering, vehicle technology) or non-infrastructure (e.g.
education, enforcement) treatments to each site. A shared responsibility is required to address
safety, including: rail operators, road controlling authorities (Councils and NZ Transport Agency)
and system users.
Since 2007, when the ALCAM (Australian Level Crossing Assessment Model) was formally
adopted in New Zealand, KiwiRail have been using it to prioritise the upgrading of its existing
crossings and to assess the increased risk of a change in use (e.g. significant increase in number
of vehicles, pedestrians and cyclists). To mitigate the increased risk KiwiRail has required road
controlling authorities (Councils and NZ Transport Agency) and other stakeholders (e.g.
developers) to upgrade the impacted level crossings. However, the reliance on ALCAM has not
always lead to the correct interventions. The coroners court have been critical of a few decisions
made by KiwiRail around level crossing safety, that may contributed to level crossing deaths. In
some cases the ALCAM risk rating has not adequately represented the level of risk at a crossing,
which may be more evident in the number of near misses (incidents) and/or the concerns raised
by locomotive engineers of risky behavior at some crossings.
In addition, a number of high-profile level crossing fatalities has increased the focus on level
crossing safety over the last five to ten years. In October 2016, the New Zealand Transport
Accident Investigation Commission (TAIC) added “safety for pedestrians and vehicles using
level crossings” to its watch-list of pressing concerns (https://taic.org.nz/watchlist). TAIC noted
that the process for assessing risk at pedestrian (and bicycle) crossings was not keeping pace
with infrastructure changes and increasing patronage on metropolitan passenger trains. The
increasing use of railway corridors for shared use and cycle paths has also led to an increase in
pedestrians and cyclists using these corridors and the level crossings along them.
While ALCAM does have its limitations, it is still an important part of the risk assessment
process. Many countries, and jurisdictions (States) within countries, use a hazard ranking tool
like ALCAM to assess the level of risk at their level crossings and to manage risk across their
level crossing network, at least for vehicle crossings. Sperry et al (1) contains a review of the
hazard ranking tools used in 39 states within the USA. The two most common hazard ranking
tools in the USA are the hazard index technique and the crash prediction formula, with the U.S
DOT Accident Prediction Model being the most common (used in 19 States) ranking tool. The
common factors considered in these tools are 1) highway traffic volume, 2) train volume, 3)
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
Turner, Cook, Bosher 3
warning device type, 4) crash history, 5) train speed, 6) number tracks, 7) highway lanes and 8)
road surface. Interestingly sight distance is only assessed in 9 of the 39 States, while pedestrian
volume is only considered in one State. The ALCAM vehicle model compares favorably with
the USA models, and is a combination of a crash prediction model (it uses the Peabody-Dimmick
Formula) and a hazard index. It contains seven of the eight factors above and the sight distance.
ALCAM also contains a separate pedestrian crossing model (for pedestrian only crossings and
those next to road crossings), which is a hazard index, and includes pedestrian volumes. The
ALCAM model does not currently use crash data. The new process however does include
crashes, near crash misses and other incident (e.g. driving through barriers) data from the IRIS
incident database.
To help overcome the limitations in ALCAM, KiwiRail has developed a new assessment process
called the Level Crossing Safety Impact Assessment (LCSIA). A key component of this new
process is a new risk scoring system called the Level Crossing Safety Score (LCSS). In addition
to the traditional ALCAM (level crossing risk model) score, the LCSS also looks at three
additional data sources associated with crash risk: historical crash and incident data, safety
observations made by locomotive engineers and road controlling authority (RCA) engineers, and
a more detailed site safety assessment that considers the crossing and surrounding transport
network and mix of users, including cyclists. In addition to a risk assessment, the LCSIA also
identified any safety issues that need to be considered at an existing site or as part of a crossing
upgrade. Theses safety observations can be utilized in the formal safety audit process.
KiwiRail requires that an LCSIA is completed for all level crossings that are modified/ upgraded
or impacted by increasing usage. Examples of increased usage include, increase in train
frequencies, increase in pedestrians and cycle usage due to construction of a railway corridor
cycleway or increase in vehicles due to new land-use development. A LCSIA needs to be
prepared and approved prior to the construction of any planned change to level crossings or any
operational changes (e.g. increase in train frequency).
This paper outlines the LCSIA process and also the lessons that have been learnt from
undertaking LCSIA at over 150 pedestrian and vehicle level crossings around New Zealand. The
authors of this paper were involved in preparing the LCSIA guidelines (2). This paper does
include extracts from this guideline.
LCSIA CRITERIA AND ITS LINK TO NZ LEGISLATION ON RAILWAY CROSSING
KiwiRail has developed two criteria for use in assessing level crossing safety. The first criterion
(Criterion 1) is that the Level Crossing Safety Score (LCSS) of new and modified level crossings
should be ‘Low’ or ‘Medium-Low’ after opening and ten years in the future. Only at these low
risk levels are the health and safety requirements as specified in the Railways Act (3) (“the Act”)
deemed to be meet. The main purpose of “the Act” is to promote the safety of rail operations and
to clarify the law relating to management of the railway corridor. Following recent updates, it
now also incorporates aspects of the Health and Safety at Work Act (4).
When considering the safety of rail operations in the Act (3), a key concept is that of
“reasonably practicable”, which is defined as:
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
Turner, Cook, Bosher 4
In this Act, unless the context otherwise requires, reasonably practicable, in relation to a duty to
ensure health and safety or to protect property, means that which is, or was, at a particular time,
reasonably able to be done in relation to ensuring health and safety or the protection of property,
taking into account and weighing up all relevant matters, including—
(a) the likelihood of the hazard or the risk concerned occurring; and
(b) the degree of harm or damage that might result from the hazard or risk; and
(c) what the person concerned knows, or ought reasonably to know, about—
(i) the hazard or risk; and
(ii) ways of eliminating or minimising the risk; and
(d) the availability and suitability of ways to eliminate or minimise the risk; and
(e) after assessing the extent of the risk and the available ways of eliminating or minimising
the risk, the cost associated with available ways of eliminating or minimising the risk, including
whether the cost is grossly disproportionate to the risk.
The Act also defines “level crossings” to include both where “a railway line crosses a road on
the same level” or where “the public is permitted to cross a railway line on the same level”. The
latter can therefore include crossings that are only accessible by people walking or cycling.
Hence both vehicle and pedestrian/cyclist crossings are addressed by the Act.
When undertaking an LCSIA the assessor identifies what improvements are required to achieve a
‘Low’ and ‘Medium-Low’ risk score at each crossing assessed. Where half-arm barriers at
vehicle crossings or automatic gates at pedestrian crossings do not achieve this low level of risk,
then the recommendation is either closure or grade separation of the crossing. In many locations
it is neither ‘reasonably practicable’ to close the crossing or grade separate, in most cases due to
cost. Closure of the crossing may also deter walking and cycling trips due to the increased travel
distances, which is undesirable. Hence in many situations a barrier system is reasonably
practicable compared with grade separation or closure.
When Criterion 1 is not achieved then Criterion 2 needs to be meet. Criterion 2 requires that the
risk score for the modified crossing is lower than the existing level crossing after opening and ten
years in the future. For example, if the train frequency is increased at a crossing by 50%, then
improvements should be made so that the LCSS in the after situation is lower than the existing
crossing with fewer trains. Where a barrier treatment is already in place then only closure or
grade separation can normally reduce the risk to the required level.
LCSIA PROCESS
The following section is a summary of the process detailed in the Level Crossing Risk
Assessment Guide (2). Figure 1 shows the general process for undertaking LCSIA’s at new and
modified pedestrian (and cycle) and vehicle crossings. It commences with a meeting with
KiwiRail to explain what is proposed and ends with an LCSIA report being submitted to
KiwiRail for approval and discussion.
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
Turner, Cook, Bosher 5
Figure 1: LCSIA Process for new and modified existing level crossings
184
185
Turner, Cook, Bosher 6
The LCSIA report provides both the RCA and KiwiRail with an understanding of the road and
rail safety risks at the site and what changes should (and could) be made to make the crossing as
safe as reasonably practicable.
Risk Assessment Methodology
The elements of the LCSIA methodology include:
Selecting the appropriate team members (including an accredited LCSIA Assessor) to
undertake the assessment
Site visit with a local roading engineer and KiwiRail representatives (must include a
locomotive engineer) to understand the risks and road user behavior
A site visit and assessment of current conditions at the level crossing (for existing
crossings) and surrounding road network
Review of the proposed ‘change in use’ activity to the level crossing, whether they be
physical changes or changes in traffic volumes e.g. a new residential development nearby
Complete the Site Specific Safety Score (SSSS) Assessment- looks at matters like
number of cyclists and vulnerable pedestrians and proportion of distracted pedestrians
Identify and specify any safety issues at the existing crossing (including maintenance
issues) and with the proposed design / upgrade
Calculation of the LCSS for the existing crossing conditions and the change in use
(whereby the risk value of the 10 year forecast increase in users is assessed on the
crossing with no upgrades employed).
Development of a list of improvements / modifications of the proposed design to cater for
the increase in users at the crossing. Undertake a LCSS assessment for the proposed
design.
Undertake a LCSS assessment for the future risk score of the crossing, based on ten year
forecast increase in crossing users. Includes any necessary increased levels of controls at
the crossing to cater for increase in users.
A recommendation on the necessary changes required at the crossing in order to achieve
Criterion 1 or Criterion 2 after opening and ten years in the future.
The following sections include more detail on these processes. An LCSIA must be led by a
KiwiRail accredited assessor. For level crossings that involve a cycleway or shared paths it is
also strongly recommended that a designer of walking and cycling crossings/facilities also be
involved. The New Zealand “Railway Crossing Pedestrian/Cycling Guidelines” (5) provide
more details on designing crossings for pedestrians and cyclists.
Existing Conditions at the level crossing
A site visit must be undertaken by an accredited assessor for each new or modified level crossing
to assess the existing conditions, review proposed changes to the crossing, undertake a safety
review and assess the Site Specific Safety Score. If an ALCAM assessment is not available for
the crossing (i.e. there is no record in the ALCAM Level Crossing Management (LXM)
database) then a separate ALCAM field survey also needs to be undertaken and a new ALCAM
record produced. A local roading engineer and a locomotive engineer from KiwiRail are also
normally involved in the site visit. They can raise safety concerns at the site and also provide a
risk score out of 5 for each existing and proposed level crossing arrangement.
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
Turner, Cook, Bosher 7
Proposed and Modified Designs at New / Upgraded Level Crossing
The LCSIA should ideally inform the design process and recommend what level and type of
control devices that are required to meet Criterion 1 and/or Criterion 2. In many cases the
applicant (i.e. the organisation or development that is increasing road user demands, increasing
train demands, or making physical changes to the crossing) provides a proposed upgrade to the
crossing. This proposed design can also be assessed as one of the potential upgrade options.
The LCSIA assessment team can also specify any additional protection measures that the
applicant should consider to further reduce the risk at the crossing, e.g. automatic gates at a
pedestrian/cyclist crossing rather than a maze with flashing lights and bells, where this is needed
to achieve Criterion 1 and/or Criterion 2. This is called the proposed design.
Components of the Level Crossing Safety Score (LCSS)
The risk of pedestrian and motor vehicle crashes are assessed using the Level Crossing Safety
Score (LCSS). The maximum score (60 points) signifies a very unsafe crossing. This score
consists of the following components:
ALCAM Score (30 points),
Crash and incident history (10 points),
A site-specific safety score (SSSS) (10 points), and;
Locomotive and local roading engineers’ risk assessment (10 points).
Separate assessments are undertaken for the individual vehicle and pedestrian crossings. Based
on these scores, the crossing is placed into risk bands as shown in Figure 2, which correspond to
a five-level risk description, ranging from HIGH to LOW
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
Turner, Cook, Bosher 8
Figure 2: Level crossing safety score risk bands (2)
The following sections explain how the individual components which make up the overall LCSS
are derived. The overall ranking of the crossing is based on the sum of the four components.
ALCAM (Maximum 30 points)
The ALCAM model was first released in 2000 (6) and has been modified a number of times
since then based on an improved understanding of level crossing safety. The second release of
the vehicle model was in 2004 and the pedestrian hazard index model was released in 2005. In
2007 New Zealand adopted the ALCAM model, which is also used in all the Australian States
and Territories The most recent update in the model was in 2014. See the ALCAM in Detail (6)
and ALCAM Road/Pedestrian Model Crossing Assessment Handbook (7) for more details on
ALCAM.
The ALCAM risk score and associated risk bands are extracted from the LXM database, which
includes scores and risk bands for all public and most private vehicle and pedestrian level
crossings in Australia and in New Zealand There are five ALCAM risk bands (low, medium
low, medium, medium-high and high) that are linked to the ALCAM score, with high risk
crossings having the highest scores. The ALCAM scoring values for road and pedestrian
crossings differ. The LCSIA guidelines (2) provide tables on how the ALCAM scores are
converted to a LCSS score out of 30.
The ALCAM score is updated using the most recent vehicular, pedestrian, cyclist and train
counts. Any site characteristics observed that may have changed since the initial ALCAM
assessment was completed, can be used to update the site characteristics in LXM. The updated
ALCAM score and risk band should then be used as the new baseline condition (Updated
Existing), as all the original ALCAM assessments were undertaken more than five years ago (as
far back as 2008).
ALCAM assessments can be undertaken within LXM (under proposals mode) for the Change in
Use, Proposed Design and Future Score. If the future ALCAM risk score / band is expected to
increase, then the SRT can assess higher levels of control to reduce the risk to an acceptable
level.
Improvements to the physical control type of a level crossing site will generally have a greater
impact at reducing the ALCAM risk score than other measures. For a pedestrian crossing these
are usually automatic gates (including emergency egress), and flashing lights & bells. As
opposed to signage, lighting or adjacent controls (i.e. adjacent road controls). For a road
crossing this means the installation of half arm barriers.
Another useful output from the LXM database is the expected return period for injury and fatal
crashes at each vehicle level crossings. This can be calculated for an existing and proposed
crossing upgrade with and without the change in use. This is possible as the vehicle model is
based on a crash prediction model (Peabody Dimmick). Validation of the crash model
predictions does show that the majority of fatalities at level crossings do occur at the ‘medium-
high’ and ‘high’ risk crossings. This information is normally presented for each option in the
LCSIA report.
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
Turner, Cook, Bosher 9
The pedestrian ALCAM model is a hazard index and so a direct relationship between pedestrian
crashes and risk is not expected. However recent analysis by the authors has shown there is a
fairly strong relationship between fatal pedestrian (and cycle) crashes and the ALCAM crossing
score. It was found that 48% (12) and 64% (16) of the fatal pedestrian and cycle crashes in New
Zealand occurred at the top 50 and 100 risk rated sites respectively. Only 8% (2 fatalities)
occurred at the bottom 200 sites. Based on these findings a sudeo return period has been
developed for fatal crashes at pedestrian crossings. The lowest return periods being one fatal
every 20-40 years.
Crash and Incident History Analysis (Maximum 10 points)
This score is based on the number of crashes in the New Zealand CAS (crash analysis system
the nationwide road crash database) and the number of incidents reported in the KiwiRail IRIS
(crash and incident) database (mainly those recorded by locomotive engineers). It can also be
supplemented by other data collected locally. When there is a train vs vehicle / pedestrian fatal
crash at the crossing, 10/10 points would be automatically scored.
The scoring system is weighted for certain incident types, based on severity, e.g. an injury crash
scores higher than a near miss, or a crash involving a truck scores higher than a light vehicle.
The IRIS data is most likely to yield more results than CAS (due to capturing incidents not
involving a motor vehicle), and therefore is the stronger guide on the score for this category.
Further details on how to score each element are found in the LCSIA Guidelines.
Site-Specific Safety Score (SSSS) (10 points)
This site-based score aims to analyse elements of the layout that are not well covered or missing
from the ALCAM risk rating. Scoring systems are available for both urban and peri-urban road
and pedestrian crossings (where speed limits are posted at 70 km/h and below) and rural road
level crossings (where posted speed limits are > 70 km/h). Some of the factors considered under
the SSSS include:
Evidence of distracted pedestrians and cyclists, which is becoming a more major issue
Grounding out which can occur when the crossing is on a hump in the road
Large flange gaps (gap between rail and asphalt) especially at acute angle crossings
Proportion of vulnerable users (school children) and cyclists.
It is acknowledged, that there are some subjective ratings required within the SSSS, with some
scoring narratives for individual sections enabling different interpretations by different assessors.
The assessor should describe the rationale for their risk ratings carefully and what information
was used to establish the rating of each element. Further details on the scoring are provided
within the LCSIA Guideline (2).
Locomotive and RCA Engineers’ Risk Assessment (10 points)
This risk score reflects the level of death and serious injury crash risk that locomotive engineers
(train drivers) and RCA engineers give to each railway crossing compared with other crossings
they encounter regularly within their jurisdiction. Where possible this relative risk score should
be determined by a number of different practitioners involved with the crossing. In the case of
locomotive engineers this may be the opinion of several drivers that use each line. In the case of
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
Turner, Cook, Bosher 10
the local roading engineers, they should also consider the experience of the public (including
drivers, pedestrians and cyclists), either through surveys or through an interest group
representative (e.g. for motor vehicles and cyclists).
The combined opinion score of the two engineers is weighted 2:1 in favour of the locomotive
engineers, as they generally have more interactions with the crossing than the local roading
engineer, who may only infrequently encounter the crossing.
General Safety Review
In a similar approach to a safety audit, the LCSIA assessor identifies any safety issues at the
existing crossing, in relation to the interaction of level crossing users with the rail infrastructure.
The LCSIA safety review is not considered to be a replacement for formal road safety audit. The
general safety review is to include all safety issues ranging from missing or damaged signs
through to concerns with the layout of the level crossing and surrounding road. For example,
queues of vehicles waiting to turn right into a major access or side-road that extend across the
level crossing may not have been identified through the ALCAM site inspection process but have
significant safety implications on the crossing.
SUMMARY OF LCSIA UNDERTAKEN TO DATE AND LESSONS LEARNT
After nearly three years of the new process (October 2016 to mid-2019), a lot has been learnt
about level crossing safety and it is noted that no two level crossings are the same. In mid-2018,
progress began on a second edition of the LCSIA Guidelines, which was put into practice from
2019 onwards. A number of alterations to some of the scoring procedures were changed to try an
evoke movement in the LCSS values between each of the four assessment phases. By mid-2019,
upwards of 200 individual crossing points have had LCSIAs conducted on them. The sites
varied from quiet rural roads to busy urban pedestrian crossings. Each assessment was unique
with varying characteristics and scores. This section discusses some of the key lessons learnt
from undertaking these assessments.
A key element of the LCSIA assessment is the ALCAM score. An analysis undertaken for
KiwiRail of over 30 level crossings in Christchurch and Wellington which identified several
faults with the New Zealand ALCAM data. Therefore, the first step in any LCSIA is to check the
validity of the recorded ALCAM data and correct it where possible, to produce an ‘updated
existing’ ALCAM score. An extra benefit of the LCSIA process is that it will improve the
quality of the ALCAM database (LXM) over time. KiwiRail have also taken other steps to
improve the quality of the ALCAM data and to keep it up to date, including resurveying high-
risk crossing across New Zealand.
An ALCAM score was not always available for all crossings, especially pedestrian crossings in
peri-urban areas, even where pedestrians were already crossing the tracks. In such situations an
ALCAM survey must be undertaken. The collection of ALCAM data often created delays due to
the limited number of ALCAM certified practitioners. Recently this has been addressed by
increasing the pool of ALCAM surveyors through new training courses.
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
Turner, Cook, Bosher 11
An analysis of pedestrian data at level crossings in the ALCAM LXM database showed that
pedestrian volumes were only recorded (or estimated from partial day counts) for crossings in
Auckland. Everywhere else in New Zealand a default volume of 100 pedestrians and cyclists per
day was adopted. Many of the Auckland volumes were out of date, i.e. four or more years old.
Hence at most pedestrian crossings there is a requirement to collect the actual pedestrian and
cycle volumes to improve the ALCAM database, as the ALCAM risk score is an exposure-based
model, heavily reliant on the volume of users at the crossing. In addition to the counts
themselves, other important information like the proportion of school children, cyclists and
disabled users is collected. This additional information is used in the development of an
ALCAM scores for each crossing. The preferred data collection approach recommended is at
least one full day count (7am to 7pm) collected from video. Ideally traffic count data should also
be updated when doing a LCSIA.
Having a video of each crossing allows the assessor to also look at the behavior of pedestrians
and pick-up any distraction issues such as pedestrians using their cellphones or cyclists with
headphones, which is a factor considered in the wider LCSIA. KiwiRail have a count program
that is collecting additional pedestrian and cycle count data at sites where it is suspected that
there are greater than 100 users per day, to help prioritise sites for treatment nationally and fill
this gap in the ALCAM data. This data and counts conducted at low volumes crossings, will be
used to create a range of default values for different cities and surrounding land uses, that can be
applied to all crossings instead of a blanket 100 users. For example, this will see some crossings
increase to 200 or 300 users per day, while others reduce down to 20 or 50 users per day.
In situations where there is a parallel or crossing cycleway, it is important to understand the
origin and destination or desired line of the pedestrians and cyclist crossing the level crossing, to
understand the risks at each site. For example, if there is a strong diagonal movement across the
crossing then cyclists or pedestrians may choose to cross the railway line at an angle rather than
at the correct closing place. This information should be obtained at the same time as pedestrian
volumes data is being collected.
As already highlighted, some of the existing data in ALCAM requires updating. Information that
needs updating often includes pedestrian volumes, train speeds, train numbers, vehicle AADT,
passenger numbers, and carriage numbers. Train speeds in particular have a large impact on the
ALCAM score and are often inaccurate in ALCAM. In some places temporary speed changes
are in effect due to the need for track maintenance. However, in ALCAM the standard train
speeds should be used as this reflects the ongoing risk at the crossing.
Rural road level crossings require an understanding of the road users, especially heavy vehicle
types (logging/dairy) and the nearby activity. For example, nearby side roads that attract a lot of
heavy vehicles could result in queuing at or over the level crossing. It is also important to
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
Turner, Cook, Bosher 12
consider the high-speed environment and horizontal and vertical alignments of crossings. For
example, a loss of control vehicle on a crossing approach curved could potentially taking out the
barrier arm components.
In urban areas it is important to be aware of nearby railway stations, reverse tracking and
shunting activities that can increase risk especially of a second train coming when there are two
or more tracks. Second train coming can be an issue were one train, potentially at nearby
stations or in shunting yards, triggers the crossing bells, lights and barriers, and a pedestrian
attempts to cross but does not see the second train approaching the crossing. KiwiRail is looking
at visual and verbal warnings to alert pedestrians and motorists to another train approaching.
Our experience with developing improvement options for level crossings involving footpaths,
cycleways and shared paths is that it can be challenging to fit in the necessary infrastructure. In
some places mazes and automatic gates cannot be easily be fitted in within the road and rail
designations and land purchase is required. Consideration also needs to be given to stacking
room for the maze or automatic gates at busy cycleways. This was a key issue at one of the
crossings assessed due to a nearby sideroad. There are also benefits at cycleways of having
automatic gates rather than mazes as during normal operation the cyclists can travel straight
through the crossing when there are automatic gates while at mazes they have to wind around the
alignment at all times, even when the mazes are larger to accommodate cyclists.
Difficulties can also occur when a vehicle and pedestrian/cycle crossing have different levels of
control. The most common being where the vehicle crossing has a barrier and the pedestrian
crossing does not. In several situations the cyclists are required to use the vehicle crossing but
we know that they may choose to use the adjacent uncontrolled pedestrian crossing to avoid the
barrier when they believe they can beat the train. This is particularly common where the desire
line is to cross at the pedestrian crossing on the opposite side of the road to where the cyclists
would normal cross to link up with a sideroad.
Another situation where this imbalance may occur is on rural cycleways, where there is an
adjoining local road that has not barrier control while the cycleway has a maze or automatic
gates. In these situations it is important that a fenced-off cycleway for cyclists leads into the
crossing, to funnel the cyclists to not bypass the maze or automatic gates by using the
uncontrolled vehicle crossing.
The general safety assessment side of LCSIA is also important for picking up unique situations at
crossings, noting that as a model, ALCAM cannot pick up all safety issues. For example, there
are some driveways near an urban level crossing where drivers can exit onto the level crossing
from behind the barrier arms. In such circumstances it can also be hard for the driver to see the
flashing lights, and hence the only warning they get is the bells, which they may or may not hear
depending on whether there is other noise in the vehicle. Ideally in these situations drivers
should be banned from turning onto the crossing, or alternatively a set of flashing lights should
face the exiting driver. While these matters may not always impact on the ALCAM score, they
should never-the-less be addressed during and LCSIA assessment and followed up in the safety
audit.
The site meeting with locomotive engineers and other KiwiRail signaling and maintenance staff
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
Turner, Cook, Bosher 13
has often provided many key insight into the crossing that would otherwise have gone unnoticed
by the LCSIA Assessor, especially if they had travelled from out of town to conduct the
assessment. It has also provided a different perspective to the crossing, where most assessments
are conducted by assessors who have a road safety background and not necessarily fully
knowledgeable in all things rail. This discussion has been a key facet to the success of the
LCSIA process.
FINAL REMARKS/SUMMARY
The new LCSIA process has led to a step change in managing the safety of level crossings in
New Zealand. It is doing this in several ways.
Firstly, it brings together not only the ALCAM risk scores but also incident and crash data from
the site along with the safety opinions of locomotive and local roading engineers. The
locomotive and local roading engineers having the own experience of using the crossing and also
having access to the public and other professional views on the safety of the crossing. It also
adds a more detailed review of several of the roadway and pedestrians/cyclist related impacts of
the crossing, beyond that provided in ALCAM. Specifically looking at network impacts like
traffic queuing across the line and observations of pedestrian distraction.
The LCSIA process also looks at broader safety issues that might impact on the crossing, such as
marking and signage condition. The opinions of the locomotive and local roading engineers in
identifying these issues is important, as they have a better understanding of how the crossing
operates. A key reason for doing this assessment is to capture key issues that need to be
considered as part of the safety audit process.
The LCSIA process also provides an opportunity to review and update the ALCAM data for each
crossing, which in many cases may now be out of date. Of particular importance in the
assessments is the pedestrian, cyclist and motor vehicle volumes that change over time, but are
very important in ranking the risk of the crossing.
One area that is going to receive further attention by KiwiRail in the short to medium term are
details around the trains usage of each rail line, including numbers of trains per day, length of
trains and train speeds. The data for many railway lines is out of date and needs updating. When
undergoing an LCSIA pay careful attention to this information and seek to update where
necessary directly from KiwiRail.
‘Second train coming’ is a key issue at level crossings with two or more tracks, and especially on
metro lines where train numbers are high. This issue occurs when two or more trains cross the
level crossing on a single closure period, i.e. where the bells, flashing lights and barriers are
activated. In many places the pedestrians and cyclists are not controlled by a barrier and may
choose to cross after the first train has gone through the crossing (or when a train is waiting at a
station or part of a shunting movement), and the concern is they will not see the second train that
is approaching as it could be obscured by the first train. This illegal movement may in part be
due to the frustration resulting from the flashing lights, bells and barriers being applied for long
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
Turner, Cook, Bosher 14
periods in peak periods on metro lines. In such circumstances the use of second train coming
signage and verbal warnings is recommended, if automatic gates are not planned for installation.
Careful attention needs to be paid to the design of new pedestrian crossings and especially
crossing of shared use paths (pedestrians and cyclists). Firstly, it can be difficult in some places
to design mazes and automatic gates while keeping the crossing as close to 90 degrees as
possible. When making design compromises it is important to consult the design guidance
around pedestrians and cyclists at level crossings provided in KiwiRail (2017a). It is also
important to make illegal movements (such as diagonal crossings and footpath riding) across the
level crossing difficult through use of planting, fencing, traffic islands and hoops. This is
particularly important where the shortcut movement is a lot quicker than the correct crossing
movement.
REFERENCES
1. Sperry, B, Naik, B and Warner, J Current Issues in Highway-Railway Grade Crossing
Hazard Ranking and Project Development, 96th TRB Annual Meeting Washington DC,
USA, 2017
2. KiwiRail, Level Crossing Risk Assessment (LCSIA), Interim Version 2, July 2017.
http://www.kiwirail.co.nz/uploads/Publications/LevelCrossingRiskAssessmentGuide.pdf
3. Railways Act 2005???????????????
4. Health and Safety at Work Act 2015???????????????
5. KiwiRail, Rail Crossing Pedestrian/Cycle Design Guidance, Interim Version 2, May
2017 http://www.kiwirail.co.nz/uploads/Publications/RailXing-PedBikeXingDesign-
RevisedGuide.pdf
6. ALCAM Committee, ALCAM in Detail, Australia, May 2015 (Available from ALCAM
Committee)
7. ALCAM Committee, ALCAM Applicable to Road Model and Pedestrian Model
Croasing Assessment Handbook, Australia, July 2014 (Available from ALCAM
Committee)
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.