Post on 07-Mar-2018
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE1
Road Safety Audit training course
Road safety analysis
Lárus Ágústsson, Road safety auditor, COWI A/S
Content
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE2
Road safety analysis
1. Overview of analysis methods and tools
2. Breaking the accident chain
3. Importance of good crash data
4. Identification of high risk sections
5. Measures
6. Exercise on accident analysis
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE3
Different approaches
Road safety analysis
RoadSafetyImpactAssess-ment(RIA)
RoadSafetyAudit
(RSA)
RoadSafetyInspec-tion
(RSI)
Crash-proneMana-gement
(BSM)
Net-workSafetyMana-gement
(NSM)
Pro-Active (Prevention)
New Schemes
Re-Active (Cure)
Existing Roads
Breaking the accident
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE4
Road safety analysis
› Contributing factors
› human
› road and its environment
› vehicles
› Often multiple factors
› 19% road factors interact with human factors
› Break the accident chain
› eliminate the risk of collision/accident leading to serious casualties
› reduce consequences
Accumulations of accidents produce accident-prone locations
ROAD SAFETY AUDTING COURSE
Road safety analysis
31 OCTOBER 20145
Treatment of accident-prone locations
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE6
The steps of identification and treatment of accident prone locations
Identify crash prone locations
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE7
The steps of Treatment of accident prone locations
› Need good source of data, e.g. accident database
› In Denmark two methods:
› Accident model method - more than expected - mainly overall roads, needs long statistics
› and through concentration of accidents and/or the accident frequency/rate
Structured and rigorous approach to identifying sites involves consideration of accident numbers and rates
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE8
Road safety analysis
› The accident frequency (in terms of the number of casualty accidents per unit length of road (or per junction) per time period)
› The accident rate (in terms of casualty accidents per hundred million vehicle kilometers)
› Weighting accidents by severity to produce an index value (e.g. fatal accidents score 5, serious (hospitalised) accidents score 3 and minor (non-hospitalised) accidents score 1)
› For identifying hazardous areas the number of crashes per square kilometer or the number of crashes per kilometer of road could be used.
Methods for identification of black spots
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE9
Road safety analysis
Methodology Advantages Disadvantages
Accident frequency (density)
-Accident/km/year
- Accident/year
Simplicity
Sites with high number of accidents detected
Bias towards high traffic volume sites
Does not take into account accident severity
Does not take into account random nature of accidents
Accident rate
- Accident/MVeh km per year
Takes into account traffic
Takes account of exposure to accident risk and may be used to compare roads with different traffic volumes
Is the must widely used identification criterion, which facilitates comparisons
Traffic volume must be known
Does not take into account random nature of accidents
Bias towards low-traffic roads
Does not take into account accident severity
Assumes a linear relationship between traffic volume and accidents
Often results in priority being given to sites on lightly-trafficked roads with only few accidents
Combined criteria
- e.g. accident frequency and rate
Reduces short coming of individual criteria Traffic volume must be known
Does not take into account random nature of accidents
Methods for identification of black spots
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE10
Road safety analysis
Methodology Advantages Disadvantages
Weighting accidents by severity
Useful technique to identify the sites with the more severe crashes
Careful not to put too high a weighting on fatal crashes, as there are often other factors involved (such as age of victim, and time taken to get to hospital)
Critical accident rate
- compare with average of group with similar characteristics
Take into account random nature of accidents
Takes into account traffic
Complexity of method
Does not take into account accident severity
Assumes a linear relationship between traffic volume and accidents
Equivalent property damage only index
- factors on seriousness
Take into account accident severity
Simplicity
Does not take into account traffic exposure
Does not take into account random nature of accidents
Bias towards high-speed sites (rural)
Relative severity index
- each accident compared similar conditions weighted by costs
Take into account accident severity
Reduces the influence of exogenous (other) variables having impact on severity
Development of cost grid may be complex
Does not take into account traffic exposure
Does not take into account random nature of accidents
Bias towards high-speed sites (rural)
Methods for identification of black spots
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE11
Road safety analysis
Methodology Advantages Disadvantages
Accident prediction models
- compare site with similar average sites
Improves accuracy of the estimated potential for improvement
Relative complexity
Does not take into account random nature of accidents
Emerical bayesian methods
- statistical method
Take into account random nature of accidents
Improves accuracy of the estimated potential for improvement
Relative complexity
Other considerations are:
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE12
Road safety analysis
› Treat personal views with caution – no-one has reliable and comprehensive knowledge of where the worst accident sites are
› Try and resist the pressure to “do something” that arises when one bad accident makes headlines in the media
› Be cautious about treating sites that look dangerous but seem to have no accidents – are you sure you will improve the situation?
› Beware of accident clusters around police stations – it may just reflect the fact that crashes are more likely to be reported if the police are nearby.
For safety assessment the following information should at least be collected:
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE13
› Location of accident (location by geographic coordinates, road/street name/number, chainage or distance from known point)
› Time of accident (year, month, day, time of day)
› Involved parties (people, vehicles, animals, road side objects)
› Result of collision - number of casualties in each category (fatalities, severe injuries, light injuries)
› Road feature (road type, junction, section, etc.)
› Environmental conditions (road surface condition, weather, light)
› Reason for collision (why or how) – direction of road users, collision type, error, and reason for error.
14
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE
Road Safety Analysis – Police reports
› Example – Police report
› Location
› Short deskription
› Accident type no. involved
› Information concerning involved parts
› Geo coding/localization
Illustration of combined method
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE15
Road safety analysis
SECTIONS JUNCTIONS
Calculate accident density using accident data
ACCfrequency=accident/year/km ACCfrequency=accident/year
Rank junctions and sections according to density
Sections e.g. more than 4 or 5 accidents with casualties within glider a)
Junctions e.g. choose those with more than 4 or 5 accidents with casualties per year a)
Choose amount of junctions and sections which are possible to assess and e.g. sort these according to frequency/rate
𝐴𝐶𝐶𝑟𝑎𝑡𝑒 =𝐴𝐶𝐶𝑟𝑒𝑔
𝐴𝐷𝑇 ∙ 365 ∙ 𝑦𝑒𝑎𝑟 ∙ Length (km)∙ 106 𝐴𝐶𝐶𝑟𝑎𝑡𝑒 =
𝐴𝐶𝐶𝑟𝑒𝑔
ADT𝑖𝑛𝑐𝑜𝑚𝑖𝑛𝑔 ∙ 365 ∙ no. year∙ 106
Total ADT is sum of incoming traffic from each leg 1𝑛 𝐴𝐷𝑇
2
Select accident prone locations
Note a): If the list becomes very long a higher threshold may be chosen.
Analyse accidents
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE16
The steps of Treatment of accident prone locations
› Is there a pattern,
› Suggest how a problem could be solved
› Need information about each accident
Road Safety Analysis
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE17
The steps of Treatment of accident prone locations
› Stick diagram
Summary analysis
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE18
The steps of Treatment of accident prone locations
Collision type
Casualty crashes Damage only crashes
Casualties
3-yr total
Year 3-yr total
% Av. annual
3-yr total
Fat Ser Sli All
‘97 ‘98 ‘99
Head on 2 0 0 2 10 0 2 1 5 8
Rear end 0 0 0 0 0 0 0 0 0
Side 0 1 0 1 5 1 0 1 2 3
Overtake 0 0 0 0 0 0 0 0 0
Turning 0 0 0 0 0 0 0 0 0
Single vehicle rollover
0 1 1 2 10 1 0 1 16 17
Single vehicle collision
0 1 0 1 5 2 0 0 1 1
Hit pedestrian 2 2 2 6 30 4 1 1 6
Hit cyclist 0 0 0 0 0 0 0 0
Other 3 5 0 8 40 3 1 35 14 50
Total 7 10 3 20 100 6.7 7 7 39 39 85
Night 4 20 4 KSI Index 0.54
Day 16 80 3
Wet 3 15 1
Dry 17 85 6
Accident analysis
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE19
The steps of Treatment of accident prone locations
Gl. Sorøvej
Møllevangen
3
1
2
4
5
6
7
8
9
10
11
List of problems and hypothesis
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE20
Road safety analysis
Slagelse Landevej
Gl. Sorøvej
Møllevangen
3
1
2
4
5
6
7
8
9
10
11
Hypothesis:
Local road lack of give away
To high speed in the village
Lack of recognition of the junction
Glider method (Accident density for a road section should be calculated, using glider of e.g. 500m)
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE21
Road safety analysis
Suggest countermeasures
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE22
The steps of treatment of accident prone locations
› Suggest appropriate measures to prevent
further accidents based on e.g. the pattern
identified above
› Based on the analysis, appropriate measures
should be found to prevent further accidents.
This could be civil works measures such as:
› improving junctions
› creating pedestrian facilities
› speed reducing measures
› etc.
› The costs of the measures and the expected
reduction of accidents from the selected
measures should be assessed for each black
spot
Suggest countermeasures
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE23
The steps of treatment of accident prone locations
› trouble solving table
Crash type Countermeasures
Hit pedestrian or cyclist
Speed management Wider, paved shoulders Pedestrian/vehicle segregation - footways in built-up areas - shoulders with intermittent kerbs Pedestrian crossing facilities Pedestrian guardrail or other protection Enforcement of speed limits
Loss of control
Better delineation: - bigger or better road signs - marker posts - road markings (centreline and raised profile edge lines) - rumble strips - road studs Speed management Improved alignment Superelevation Safety barrier
Skidding
Restoring surface texture Resurfacing Improve drainage
Collision with roadside objects
Remove objects Better delineation Break-away posts Safety barrier
Head-on collisions ‘No-overtaking’ road markings Hatch marking in the centre of the road Road signs Climbing lanes
Crashes at night
Better delineation Reflective road markings Reflective signs Street lighting in built-up areas and at junctions
Poor visibility
Trim or remove vegetation Improved sightlines Realignment
Poor driving behaviour or lane discipline
Police enforcement Road markings (lane lines) Hatch markings reinforced with rumble strips Median barriers Access control (kerbs, barriers) Climbing lanes
Failure to give way at junctions Improved signing and marking Re-align minor road so that it joins major road at 90
0
Replace 4-way junction with two 3-way junctions Deflect entry to the left (roundabouts) Rumble strips Speed management
Crashes at bridges Replacement of bridge railings with high-containment, crashworthy parapet Shielding of parapet ends with guardrail Guardrail on embanked approaches Provision of footways on the outside of the bridge.
Suggest countermeasures
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE24
The steps of treatment of accident prone locations
› trouble solving table
Crash type Countermeasures
Hit pedestrian or cyclist
Speed management Wider, paved shoulders Pedestrian/vehicle segregation - footways in built-up areas - shoulders with intermittent kerbs Pedestrian crossing facilities Pedestrian guardrail or other protection Enforcement of speed limits
Loss of control
Better delineation: - bigger or better road signs - marker posts - road markings (centreline and raised profile edge lines) - rumble strips - road studs Speed management Improved alignment Superelevation Safety barrier
Skidding
Restoring surface texture Resurfacing Improve drainage
Collision with roadside objects
Remove objects Better delineation Break-away posts Safety barrier
Head-on collisions ‘No-overtaking’ road markings Hatch marking in the centre of the road Road signs Climbing lanes
Crashes at night
Better delineation Reflective road markings Reflective signs Street lighting in built-up areas and at junctions
Poor visibility
Trim or remove vegetation Improved sightlines Realignment
Poor driving behaviour or lane discipline
Police enforcement Road markings (lane lines) Hatch markings reinforced with rumble strips Median barriers Access control (kerbs, barriers) Climbing lanes
Failure to give way at junctions Improved signing and marking Re-align minor road so that it joins major road at 90
0
Replace 4-way junction with two 3-way junctions Deflect entry to the left (roundabouts) Rumble strips Speed management
Crashes at bridges Replacement of bridge railings with high-containment, crashworthy parapet Shielding of parapet ends with guardrail Guardrail on embanked approaches Provision of footways on the outside of the bridge.
Suggestions for countermeasure
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE25
Road safety analysis
Measures that can be recommended to solve the road safety problem
Gl. Sorøvej
Møllevangen
3
1
2
4
5
6
7
8
9
10
11
Suggestions for countermeasure
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE26
Road safety analysis
Gl. Sorøvej
Møllevangen
Slagelse Landevej
Measures that can be
recommended to solve the road
safety problem:
› Roundabout
› Left turning lanes
› Stop signs at local road
› Speed limit down from 70 km/h
› Raised pedestrian crossing
› Middle Islands on main road
› Raised junction
› Speed humps on main road
› (Closure of local road)
› By-pass
› Combination
Accident effects
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE27
Road safety analysis
PIARC road safety handbook
Accident effects
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE28
Road safety analysis
PIARC road safety handbook
Accident effects
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE29
Road safety analysis
PIARC road safety handbook
Calculate cost benefit ratio
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE30
The steps of Treatment of accident prone locations
E.g. First Year Benefit or Cost Effectiveness
› FYB if accident costs are introduced:
› FYB= (acc reduction * acc cost)/cost
› Value of Fatality = 70 * GDP/Capita
› Value of Serious Injury = 17 * GDP/Capita
› The projects with the highest first year return should be ranked first.
› Otherwise a cost-effectiveness figure could be used:
› CE= accident saved/cost
› Thus the black spots where you save most per invested money.
Prioritise accident prone locations for improvement
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE31
The steps of Treatment of accident prone locations
› Based on the lists, the projects for improvements should be selected.
› Generally the projects will be selected according to the priority list, until the allocated budget has been spent.
› Some projects ranked high may need further investigations and will be delayed.
› While some accident prone locations with lower priority may be low-cost and simple and may be handled during normal maintenance work.
Implementation
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE32
Road safety analysis
› Detailed design
› Road Safety Audit
› Tender process
› Contract
› Supervision
Evaluate effects
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE33
The steps of Treatment of accident prone locations
› By monitoring the performance – did it work?
› Whether the number of accidents has been reduced or other types of accidents have increased.
› The most feasible method for performing accident prone identification, analysis and improvements should be studied…… for the future
› Then a new year starts….
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE34
Road safety analysis
Issues to consider:
How is accident analysis done in your country?
What effective measures are used?
How are measures prioritised?
Exercise on accident analysis at accident prone location
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE35
Road safety analysis
› Base:
› Accidents – summary
› Summary base reports
› Base map
› Situation:
› A signalized 4-leg junction in the outskirts of city limits is accident-prone and the road authorities would like to find out what the problems are and what kind of measures could improve road safety in the junction.
› Task:
› Elaborate a conflict-diagram
› Analyse the situation with regard to road safety
› List problems
› Provide potential solutions.
Accident summary
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE36
Road safety analysis
Summary base reports
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE37
Road safety analysis
Accident No. 1 2 3 4> <4 5
Accident ID
Date
Week day, time
Road no.
Chainage (km)
Road 2 no.
691711
09 Apr 10
Sun, 0147
000-0122-0
086/0265
030-0527-0
744513
05 Jun 11
Tue, 11:40
000-0122-0
086/0265
343-0528-0
771216
27 Jun 12
Thu, 13:23
000-0122-0
086/0265
343-0528-0
773094
20 Jul 12
Sat, 20:04
000-0122-0
086/0265
030-0527-0
778456
12 Sep 02
Thu, 18:17
000-0122-0
086/0265
343-0528-0
Condition
Visibility
Light
Weather
Street light
Accident type
Dry
Clear
Dark
Dry
On
Person damage
Wet
Clear
Day light
Rain
Off
Damage only
Dry
Clear
Day light
Dry
Off
Damage only
Dry
Clear
Day light
Dry
Off
Damage only
Dry
Clear
Day light
Dry
Off
Person damage
Accident situation
660 Accident with left turn in of "approaching vehicle"
410 Accident with left turn in front of approaching vehicle
241 Head-on collision in part 2´s driving lane
140 Rear-end accident
410 Accident with left turn in front of approaching vehicle
Element no.
Element type
Speed (km/h)
Age
Alcohol (0/00)
Killed
Serious injured
Light injured
1
Car
70
17
0
1
1
2
Car
90
21
0
1
Car
5
84
0
2
Car
50
66
0
1
Car
5
49
0
2
Car
60
49
0
1
Car
65
34
0
2
Car
0
25
0
3
Ambulance
0
36
0
1
Car
50
49
0
1
2
Car
60
41
0
Summary base reports
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE38
Road safety analysis
Accident No. 6 7 8 9 10 11
Accident ID
Date
Week day, time
Road no.
Chainage (km)
Road 2 no.
777603
18 Sep 12
Wed, 16:15
000-0122-0
086/0265
343-528-0
803008
10 Aug 13
Sun, 06:20
000-0122-0
086/0265
030-0528-0
807086
11 Sep 13
Thu, 16:15
000-0122-0
086/0265
030-0527-0
814714
22 Oct 13
Wed, 16:13
000-0122-0
086/0265
030-0527-0
816315
29 Oct 13
Wed, 08:29
000-0122-0
086/0265
030-0528-0
842647
21 May 14
Fri, 12:40
000-0122-0
086/0265
343-528-0
Condition
Visibility
Light
Weather
Street light
Accident type
Dry
Clear
Day light
Dry
Off
Damage only
Dry
Clear
Day light
Dry
Off
Damage only
Dry
Clear
Day light
Dry
Off
Extra
Dry
Clear
Day light
Dry
Off
Damage only
Dry
Clear
Day light
Dry
Off
Person damage
Dry
Clear
Day light
Dry
Off
Damage only
Accident situation
660 Accident with left turn in of "approaching vehicle"
032 Single accident while turning in junction, access, roundabout or similar
410 Accident with left turn in front of approaching vehicle
410 Accident with left turn in front of approaching vehicle
410 Accident with left turn in front of approaching vehicle
410 Accident with left turn in front of approaching vehicle
Element no.
Element type
Speed (km/h)
Age
Alcohol (0/00)
Killed
Serious injured
Light injured
1
Car
5
65
0
2
Van
10
21
0
1
Car
50
20
0.65
2 1
Van
0
30
0
2
Car
50
30
0
1
Bus
5
56
0
2
Moped
30
44
0
1
Car
10
21
0
1
2
Car
80
29
0
1
Car
15
81
0
2
Car
60
44
0
Summary base reports
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE39
Road safety analysis
Accident No. 12 13 14 15 16> <16
Accident ID
Date
Week day, time
Road no.
Chainage (km)
Road 2 no.
843009
22 May 14
Sat, 01:56
000-0122-0
086/0265
843651
02 Jun 14
Wed, 08:05
000-0122-0
086/0265
343-0528-0
848067
08 Jul 14
Thu, 12:50
000-0122-0
086/0265
343-0527-0
849935
24 Jul 14
Sat, 09:19
000-0122-0
086/0265
343-0528-0
852006
11 Aug 14
Wed, 16:30
000-0122-0
086/0265
343-0528-0
Condition
Visibility
Light
Weather
Street light
Accident type
Dark
Extra
Dry
Clear
Day light
Dry
Off
Damage only
Dry
Clear
Day light
Dry
Off
Damage only
Dry
Clear
Day light
Dry
Off
Damage only
Dry
Clear
Day light
Dry
Person damage
Accident situation 098 Other single accidents
410 Accident with left turn in front of approaching vehicle
510 Accident between crossing vehicles without turning and part2 from right
410 Accident with left turn in front of approaching vehicle
520 Accident between crossing vehicles without turning and part2 from left
Element no.
Element type
Speed (km/h)
Age
Alcohol (0/00)
Killed
Serious injured
Light injured
1
Car
0
2
1
Car
30
30
0
2
Car
50
34
0
1
Car
20
49
0
2
Car
70
76
0
1
Car
10
61
0
2
Car
70
34
0
1
Car
80
53
0
2
Car
70
72
0
1
3
Pole
Base map – North is upwards
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE40
Road safety analysis
Groups
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE41
Group 1
› Ivan Đurčević
› Ariana Hasani
› Samir Džaferović
› Darko Miceski
› Ivana Subotić
› Fatma Orhan
Group 2
› Dorian Stafa
› Spasoje Mićić
› Dalibor Milošević
› Emir Morina
› Željko Majstorović
Group 3
› Meri Lumi
› Gjynejt Mustafa
› Miroslav Mašić
› Valben Mustafa
› Salih Erdemci
Group 4
› Miroslav Đerić
› Kiril Karkalashev
› Olivera Stević Ledenčan
› Dražen Antolović
› Mustafa Işık
ROAD SAFETY AUDTING COURSE42
Road safety analysis
Example of solution to exercise to come
31 OCTOBER 2014
Solution – conflict diagram
31 OCTOBER 2014
ROAD SAFETY AUDTING COURSE43
Road safety analysis
› separate left turn phases in signal
or
› roundabout