Safer Level CrossingsSafer Level Crossings

Chris Wullems (QUT)Chris Wullems (QUT)

Peter Furnell (DoT Victoria) Peter Furnell (DoT Victoria)

Rail CRC Showcase - 24Rail CRC Showcase - 24thth August 2010 August 2010

IntroductionIntroduction

R3.122 Affordable Level Crossings Project – R3.122 Affordable Level Crossings Project – Stage 2Stage 2 Commenced in March 2011Commenced in March 2011

Participants:Participants:

Affordable Level CrossingsAffordable Level Crossings

Some figures on level crossing incidents…Some figures on level crossing incidents… Cost of incidents on level crossings to rail industry Cost of incidents on level crossings to rail industry

and communityand community $116,279,817 per annum (2010 $) $116,279,817 per annum (2010 $) (Tooth & Balmford, RISSB (Tooth & Balmford, RISSB

Railway Level Crossing Incident Costing Model 2010)Railway Level Crossing Incident Costing Model 2010)

Indicative costs only. Based on historical averages for Indicative costs only. Based on historical averages for number of incidents and casualties.number of incidents and casualties.

Does not include suicidesDoes not include suicides 8838 public level crossings in Australia as of 2009 8838 public level crossings in Australia as of 2009

(RISSB Level Crossing Stocktake 2009)(RISSB Level Crossing Stocktake 2009)

33% with active protection 33% with active protection 67% with passive protection67% with passive protection

Cost of installing active protection at all passive Cost of installing active protection at all passive level crossings in Australia is estimated to be level crossings in Australia is estimated to be between $1.2 billion and $1.8 billion (excluding between $1.2 billion and $1.8 billion (excluding maintenance costs) maintenance costs) (Cairney, 2003)(Cairney, 2003)

Affordable Level CrossingsAffordable Level Crossings

Collectively, passive crossings represent a significant Collectively, passive crossings represent a significant safety issuesafety issue

Why Affordable Level Crossings?Why Affordable Level Crossings? They cost at least 25% of the cost of traditional crossing They cost at least 25% of the cost of traditional crossing

technologiestechnologies For a given investment, more crossings can be treatedFor a given investment, more crossings can be treated

Greater safety benefit for same investment used to treat Greater safety benefit for same investment used to treat crossings using traditional technologiescrossings using traditional technologies

Not intended to be a replacement for traditional Not intended to be a replacement for traditional technologies on high-exposure crossingstechnologies on high-exposure crossings

Target environmentTarget environment Crossings on a single-track line with Crossings on a single-track line with Relatively low vehicle and rail trafficRelatively low vehicle and rail traffic Little passenger train services if anyLittle passenger train services if any

Affordable Level Crossings (2) Affordable Level Crossings (2)

Key IssuesKey Issues Can low-cost level crossings provide better safety benefits Can low-cost level crossings provide better safety benefits

for the network than the current approach of incremental for the network than the current approach of incremental upgrades?upgrades?

Reliability issues of low-cost technologies Reliability issues of low-cost technologies Do low cost warning devices need to have the same Do low cost warning devices need to have the same

reliability as traditional warning devices?reliability as traditional warning devices? Issue of liability should an accident occur and it is determined Issue of liability should an accident occur and it is determined

that the same accident would not have occurred on a that the same accident would not have occurred on a traditional high-integrity level crossingtraditional high-integrity level crossing

How to evaluate and compare systems build using How to evaluate and compare systems build using alternative low-cost technologies? alternative low-cost technologies?

Need for lifecycle evaluation schemeNeed for lifecycle evaluation scheme

Value of solving these issues for industryValue of solving these issues for industry

R3.122 Affordable Level R3.122 Affordable Level Crossings – Stage 2 (1)Crossings – Stage 2 (1)

Project objectivesProject objectives To estimate lifecycle costs of LCLCWDs and To estimate lifecycle costs of LCLCWDs and

conventional warning devicesconventional warning devices Investigate human factors of unavailability – Investigate human factors of unavailability –

frequent / prolonged right-side failurefrequent / prolonged right-side failure To trial candidate LCLCWDs for a period of 12 To trial candidate LCLCWDs for a period of 12

months – shadow modemonths – shadow mode To determine whether adoption of LCLCWDs can To determine whether adoption of LCLCWDs can

provide a better safety outcome for Australia than provide a better safety outcome for Australia than the current incremental upgrade approachthe current incremental upgrade approach

R3.122 Affordable Level R3.122 Affordable Level Crossings – Stage 2 (2)Crossings – Stage 2 (2)

Expected outcomesExpected outcomes Set of requirements for LCLCWDs with safety and Set of requirements for LCLCWDs with safety and

availability targetsavailability targets Risk assessment model Risk assessment model Human reliability assessment modelHuman reliability assessment model

Lifecycle assessment criteriaLifecycle assessment criteria Where cost savings can be madeWhere cost savings can be made

Trial resultsTrial results Comparative performance and operational data Comparative performance and operational data

(reliability, availability, maintainability)(reliability, availability, maintainability) Results from human factors studyResults from human factors study

Effectiveness of various measures to improve Effectiveness of various measures to improve performance of road users at level crossings that are performance of road users at level crossings that are unavailable (effective communication of crossing state)unavailable (effective communication of crossing state)

Making the case for LCLCWDs (1)Making the case for LCLCWDs (1)

Safety is a key issueSafety is a key issue Types of failureTypes of failure

Wrong-side (dangerous failure)Wrong-side (dangerous failure) Right-side (fail-to-safe)Right-side (fail-to-safe)

Preliminary risk assessmentPreliminary risk assessment LimitationsLimitations

Based on averaged annualized data – statistical uncertainty is a Based on averaged annualized data – statistical uncertainty is a consequence of limited number of occurrencesconsequence of limited number of occurrences

ObjectivesObjectives To provide an indication of the magnitude of risk passive level To provide an indication of the magnitude of risk passive level

crossings pose to employees, passengers and the general public;crossings pose to employees, passengers and the general public; To evaluates the viability of low cost level crossing warning devices To evaluates the viability of low cost level crossing warning devices

(LCLCWDs) as a risk mitigation option(LCLCWDs) as a risk mitigation option

Assumes RX5 road user interface (AS1742.7-2007)Assumes RX5 road user interface (AS1742.7-2007) Eliminates the need to estimate the change in risk from a road-user Eliminates the need to estimate the change in risk from a road-user

human factors perspective and constrains the comparison to the human factors perspective and constrains the comparison to the underlying technologyunderlying technology

Making the case for LCLCWDs (2)Making the case for LCLCWDs (2)

Obstacles to adoptionObstacles to adoption Concern as to whether rail operators would be liable Concern as to whether rail operators would be liable

should an accident occur at a crossing with a should an accident occur at a crossing with a LCLCWD?LCLCWD?

where the use of a conventional high-integrity warning device where the use of a conventional high-integrity warning device would have most likely prevented the same accidentwould have most likely prevented the same accident

Concern as to whether adoption of LCLCWDs would Concern as to whether adoption of LCLCWDs would be prohibited by Rail Safety Act?be prohibited by Rail Safety Act?

If low-cost technology achieves risk to be eliminated or If low-cost technology achieves risk to be eliminated or reduced so far as is reasonably practicable (SFAIRP), then it reduced so far as is reasonably practicable (SFAIRP), then it would be utilized consistently with the Rail Safety Act. would be utilized consistently with the Rail Safety Act. (letter (letter from Victorian transport safety regulator to Victorian Railway Crossing Safety Steering from Victorian transport safety regulator to Victorian Railway Crossing Safety Steering Committee)Committee)

Making the case for LCLCWDs (3)Making the case for LCLCWDs (3)

In considering cost of eliminating or reducing risk In considering cost of eliminating or reducing risk SFAIRP, practitioners must demonstrate that the SFAIRP, practitioners must demonstrate that the likelihood of the risk eventuating is remote or that the likelihood of the risk eventuating is remote or that the cost is cost is grossly disproportionate grossly disproportionate to the safety benefit.to the safety benefit.(National guideline for the meaning of SFAIRP, NTC)(National guideline for the meaning of SFAIRP, NTC)

For risk assessment: gross disproportion < 0.1 benefit to cost For risk assessment: gross disproportion < 0.1 benefit to cost ratio (BCR) for risks to general public on level crossingsratio (BCR) for risks to general public on level crossings

Grade separation (elimination of risk) is grossly Grade separation (elimination of risk) is grossly disproportionate to safety benefit for a low-exposure level disproportionate to safety benefit for a low-exposure level crossing with passive controlscrossing with passive controls

Making the case for LCLCWDs (4)Making the case for LCLCWDs (4)

Reduction of risk SFAIRP applied to LCLCWDsReduction of risk SFAIRP applied to LCLCWDs Treatment of a population of crossings with a given budgetTreatment of a population of crossings with a given budget but... but... accidents and fatalities occur at a single level crossing…accidents and fatalities occur at a single level crossing… will safety regulators and law will safety regulators and law

courts accept the population treatment argument?courts accept the population treatment argument?

Making the case for LCLCWDs (5)Making the case for LCLCWDs (5)

Using statistics, a simple cause-Using statistics, a simple cause-consequence model was developedconsequence model was developed

Intent to measure improvement to the success Intent to measure improvement to the success of of the road user notices and makes a the road user notices and makes a controlled stop controlled stop gate with the installation of gate with the installation of LCLCWDs. LCLCWDs.

More comprehensive model would elaborate More comprehensive model would elaborate range of causes and consequences range of causes and consequences

Types of trains and vehicles involved and their Types of trains and vehicles involved and their cargo, collision speeds, condition of the road cargo, collision speeds, condition of the road surface, human factors, crossing topology and surface, human factors, crossing topology and environmental factors environmental factors

Would require access to incident reportsWould require access to incident reports Not necessary for preliminary analysisNot necessary for preliminary analysis

On average a collision in Queensland results in On average a collision in Queensland results in approximately 0.112 fatalities and 0.17 major approximately 0.112 fatalities and 0.17 major injuries to general public and 0.025 serious injuries to general public and 0.025 serious injuries for employeesinjuries for employees

Average of 7.8 collisions per year Average of 7.8 collisions per year (2006-2010)(2006-2010)

Near-miss occurrence statsNear-miss occurrence stats

RLX ALCAM dataRLX ALCAM data

Collision statsCollision stats

Preliminary risk assessment (1)Preliminary risk assessment (1)

Risk assessment involvedRisk assessment involved Hazard identification – passive level crossingHazard identification – passive level crossing

““Train to motor vehicle collision at level crossing”Train to motor vehicle collision at level crossing” Not all hazards and causes were developedNot all hazards and causes were developed Risk assessment focused on hazards that can be avoided Risk assessment focused on hazards that can be avoided

through treatment with low-cost active protectionthrough treatment with low-cost active protection

Analysis of potential losses associated with hazardAnalysis of potential losses associated with hazard Annualized averages of fatalities, major injuries on level Annualized averages of fatalities, major injuries on level

crossingscrossings Person Equivalent Fatalities (PEF) – values used:Person Equivalent Fatalities (PEF) – values used:

1 fatality = 10 major injuries1 fatality = 10 major injuries 1 major injury = 20 minor injuries1 major injury = 20 minor injuries

Analysis of mitigation options for hazardAnalysis of mitigation options for hazard 3 LCLCWD and 1 conventional warning option considered 3 LCLCWD and 1 conventional warning option considered Cost-benefit analysis for each optionCost-benefit analysis for each option

Preliminary risk assessment (2)Preliminary risk assessment (2)

Determining mitigated safety lossesDetermining mitigated safety losses 51% reduction of collisions from passive to flashing lights 51% reduction of collisions from passive to flashing lights

(estimates by Elvik, Høye, Vaa, & Sørensen, 2009)(estimates by Elvik, Høye, Vaa, & Sørensen, 2009)

Further 45% reduction with boom gatesFurther 45% reduction with boom gates Less effectiveness due to increased failure rateLess effectiveness due to increased failure rate

There is a safety benefit when failed to safe (not prolonged / frequent)There is a safety benefit when failed to safe (not prolonged / frequent) Wrong-side failure reduces safety benefitWrong-side failure reduces safety benefit

Increase of risk due to wrong-side failure estimated in Increase of risk due to wrong-side failure estimated in terms of person equivalent fatalities (PEF) terms of person equivalent fatalities (PEF)

Based on the assumption that a collision will occur when a vehicle Based on the assumption that a collision will occur when a vehicle and train are at the crossing in wrong-side failureand train are at the crossing in wrong-side failure

Monetary value of mitigated lossMonetary value of mitigated loss Value of Preventing a Fatality (VPF) $6,287,873 (2010 $) Value of Preventing a Fatality (VPF) $6,287,873 (2010 $)

(RISSB) (Tooth & Balmford, 2010)(RISSB) (Tooth & Balmford, 2010)

Costs per crossing accidentCosts per crossing accident

Preliminary risk assessment (3)Preliminary risk assessment (3)

Estimated costs per crossing accident (Tooth and Balmford 2010)

Costs and benefits of population treatment of level crossings with passive control over 25 years – Queensland

1616Gross disproportionGross disproportion

Costs include Costs include installation + ongoing installation + ongoing maintenancemaintenanceUpgrade budget Upgrade budget $10M/year$10M/year

Negative safety benefitNegative safety benefit

*Note that in order for option 1b to *Note that in order for option 1b to be accepted, the cost of both be accepted, the cost of both option 1c and option 2 would have option 1c and option 2 would have to be demonstrated to be grossly to be demonstrated to be grossly disproportionate to the safety disproportionate to the safety benefit.benefit.

Preliminary risk assessment (6)Preliminary risk assessment (6)

Refinement of risk assessment numbers and model as Refinement of risk assessment numbers and model as detailed figures become availabledetailed figures become available Set safety targetsSet safety targets

Human factors aspects will be investigated Human factors aspects will be investigated Aims to inform availability targetsAims to inform availability targets

Human Factors Aspects (1)Human Factors Aspects (1)

Project investigates the effect of frequent / prolonged Project investigates the effect of frequent / prolonged unavailability on road user behaviourunavailability on road user behaviour Right-side failure Right-side failure (road user cannot distinguish failure (road user cannot distinguish failure

mode from train approaching)mode from train approaching) Tail-ringingTail-ringing

Project aims to develop and evaluate various measures Project aims to develop and evaluate various measures to improve performance of road users when warning is to improve performance of road users when warning is unavailableunavailable Through simulation – CARRS-Q advanced driving Through simulation – CARRS-Q advanced driving

simulatorsimulator

Human Factors Aspects (2)Human Factors Aspects (2)

Project informs development of availability targets for Project informs development of availability targets for LCLCWDsLCLCWDs Using techniques including task analysisUsing techniques including task analysis Development of human reliability assessment models to Development of human reliability assessment models to

quantify human error at level crossingsquantify human error at level crossings

Questions?Questions?

For more information, please For more information, please contact:contact:

Chris Wullems (QUT) Chris Wullems (QUT) c.wullems@qut.edu.au

Peter Furnell (DoT Victoria) Peter Furnell (DoT Victoria) peter.furnell@transport.vic.gov.au