cross-industry freight derailment working group … rssb | cross-industry freight derailment working...

Cross-Industry Freight Derailment Working Group

Interim Progress Report

October 2017

Copyright© RAIL SAFETY AND STANDARDS BOARD LTD. 2017 ALL RIGHTS RESERVED

This publication may be reproduced free of charge for research, private study or for internal circulation within an organisation. This is subject to it being reproduced and referenced accurately and not being used in a misleading context. The material must be acknowledged as the copyright of Rail Safety and Standards Board and the title of the publication specified accordingly. For any other use of the material please apply to RSSB's Head of System Safety for permission. Any additional queries can be directed to [email protected]. This publication can be accessed by authorised audiences, via the SPARK website: www.sparkrail.org.

Written by: Mike Laskowski, Programme Manager, System Safety, RSSB

Lynne Collis, System Safety Engineer, RSSB

Published: October 2017

Cover image courtesy of RAIB report 21/2014 Derailment at Primrose Hill/Camden Road West Junction 15 October 2013; noted in that report as courtesy Network Rail.

Issue Version Date

First draft Draft v0.3 July 2017

Final draft addressing ORR comments Draft v0.7 3 October 2017

Final version for publication V1.0 18 October 2017

mailto:[email protected]

http://www.sparkrail.org

1 Introduction.......................................................................................1

1.1 Background ......................................................................................................................1

1.2 The Cross-Industry Freight Derailment Working Group ..................................1

1.3 Method ..............................................................................................................................2

2 Progress report ..................................................................................3

2.1 Work stream 1: Risk Assessment - Update of the risk assessment. .............3

2.1.1 Risk assessment process ................................................................................................................... 32.1.2 Background to the update............................................................................................................... 52.1.3 Update method ................................................................................................................................... 52.1.4 Results ..................................................................................................................................................... 62.1.5 Next steps .............................................................................................................................................. 6

2.2 Work stream 2: Frequency Report - Identification of changes to the railway over last 10 years and in the future that could change our understanding of the way in which these types of derailment can occur and the way they are modelled/assessed ....................................................................6

2.3 Work stream 3: Wavelengths - Assess the potential benefits that could be gained from adding an additional wavelength track twist measurement/criteria. .........................................................................................................8

2.4 Work stream 4: Guidance Review - Review of existing loading practice and the guidance provided for vehicle testing..........................................9

2.5 Work stream 5: Simulation Studies - Assess the need and feasibility of testing and/or computer simulations of existing wagons to establish limiting offset loading conditions and related wheel unloading limits..........10

2.6 Work stream 6: GOTCHA - Analysis of GOTCHA measurements to establish existing wagon uneven load profile distributions ..........................11

2.7 Work stream 7: Container Loads - Identifying the scale of asymmetric container loading and controls in place to reduce and manage such loading .............................................................................................. 12

2.8 Work stream 8: RAIB Recommendations - Recommendations Mapping.......................................................................................... 13

3 Summary of Next Steps..............................................................14

3.1 Work stream 1: Risk Assessment ........................................................................... 14

3.2 Work stream 2: Frequency Report ........................................................................ 14

3.3 Work stream 3: Wavelengths ................................................................................. 14

3.4 Work stream 4: Guidance Review ......................................................................... 14

3.5 Work stream 5: Simulation Studies ...................................................................... 15

3.6 Work stream 6: Analysis of GOTCHA measurements.................................... 15

3.7 Work stream 7: Container Loads........................................................................... 15

3.8 Work stream 8 - RAIB Recommendations Mapping ...................................... 15

4 References .......................................................................................16

5 Abbreviations..................................................................................18

Appendices ............................................................................................. 19

Appendix A: List of group members...........................................19

Appendix B: Relevant RAIB recommendations......................20

Cross-Industry Freight Derailment Working Group: Second Progress Report

1 Introduction

1.1 BackgroundOn 5 December 2014 Ian Prosser, HM Chief Inspector of Railways, sent a letter to several key companies in the rail industry [1] requesting that the industry work collectively to address concerns about freight train derailment relating to the combinational effects of track condition, vehicle sensitivity to track geometry and asymmetric loading of containers. Specifically, the ORR invited the infrastructure manager and the freight industry to work together with RSSB and others, to:

1 Define the current track, vehicle and load system

2 Define how these three elements interact

3 Describe the current risk controls in place

4 Detail the gaps in those risk control systems

5 Specify the measures required in the short, medium and long term to reduce those risk gaps so far as is reasonably practicable

6 Implement those measures

1.2 The Cross-Industry Freight Derailment Working GroupIn response to the ORR letter [1], Freight Technical Committee proposed a cross-industry working group should be set up to oversee the work needed. The cross-industry freight derailment working group (XIFDWG) was therefore set up, facilitated by RSSB, with representatives from: Network Rail, freight operators, Interfleet (now SNC Lavalin), Huddersfield University, Lloyds Register Rail (now Ricardo Rail), ORR and RSSB. A full list of current members is included in Appendix A.

XIFDWG reports into the National Freight Safety Group (NFSG) [2].

RSSB | Cross-Industry Freight Derailment Working Group: Second Progress Report 1

2

1.3 MethodThe work of the cross-industry group is being undertaken following a structured process in accordance with the industry guidance on safety related decision making and risk management, Taking Safe Decisions [3]. Taking Safe Decisions is built around key regulatory requirements that duty holders must comply with, in particular the Common Safety Method for Monitoring and the Common Safety Method on Risk Evaluation and Assessment (CSM RA)[4,5].

Figure 1 - The Taking Safe Decisions Risk Management Framework

According to that process when a safety concern has been identified, analysis is required to understand the nature of the problem, and identify options that can be considered to reduce any identified risk. Appropriate application of risk acceptance criteria is necessary in order to understand whether there are any measures which are necessary.

Following the identification of options to progress, any necessary changes to the railway and its operation are made, and necessary monitoring arrangements put in place to monitor the impact of the change. This process is consistent with the ORR request for industry activity.

RSSB | Cross-Industry Freight Derailment Working Group: Second Progress Report

2 Progress report

This 2nd Interim Progress Report is an update on XIFDWG’s programme of work since the previous interim report was issued in February 2016 [6]. It should therefore be read in conjunction with that previous report.

However, for this edition, the report has been re-structured to better align with the XIFDWG programme of work which consists of eight work streams. The work streams are described below:

2.1 Work stream 1: Risk Assessment - Update of the risk assessment.2.1.1 Risk assessment process

The risk assessment work stream has been undertaken as a series of key steps, illustrated in Figure 2, mapped to the Taking Safe Decisions risk management framework.

Figure 2 - Risk management steps mapped to Taking Safe Decisions


4

1 Freight derailment events are monitored via the Safety Management Information System, (SMIS), hosted by RSSB. These were used in the initial risk assessment included in the RSSB/TRL report [7], analysed by the XIFDWG.

2 The initial risk analysis was analysed by the XIFDWG as part of the group’s deepening understanding of freight derailment risk.

3 Bowtie risk workshops were undertaken to identify hazards and risk control measures, and to assess the strength of the risk controls. The Bowtie Risk Analysis report was published in October 2016. [8]. The report details the methodology used and reports on the results of the analysis.

4 The risk controls identified as not functioning well today were then analysed further in risk prioritisation workshops. Here, the improvements needed were described. The effort and safety benefit of each improvement was then assessed according to criteria from Taking Safe Decisions. [3].

5 The risk improvements were ranked according to effort and safety benefit, and a top 10 roadmap was derived.

6 The risk analysis in the RSSB/TRL report [7] was updated, as described below. RSSB will develop the final report on the risk assessment for publication in December 2017.

Steps 7, and 8 are to be progressed next:

7 It is proposed that Step 7 will involve the duty holder cost benefit analysis of the relevant top 10 risk improvements.

8 Following implementation, Step 8 will include analysis of the residual risk to ensure that the risk improvement measures have been effective.

The bowtie risk analysis report was published in October 2016. [8]. The relationships between the top ranked controls are shown in the Top ten roadmap diagram below.


Figure 3 - Top 10 roadmap

2.1.2 Background to the update

The update to the risk analysis addresses recommendation 2 of the RAIB report into the Camden Road West Junction derailment (see Appendix B).

2.1.3 Update method

Research was undertaken into SMIS to provide an updated list of in-scope derailments. These derailments were then analysed for frequency of contributory factors, to permit analysis of the safety benefit of addressing each of these.

Risk analysis was undertaken, using the RSSB Safety Risk Model (SRM) data for the relevant risk figures. This was augmented by sensitivity analysis adapted from the SRM secondary collision event trees to provide an upper bound risk figure.

Upper bound costs were taken from significant derailment costs obtained from Network Rail, as an input to upper bound initial justified safety spend per contributory factor.


6

2.1.4 Results

The draft results were presented to the XIFDWG meeting of 18 September 2017, and will be published in the updated risk assessment report [18].

2.1.5 Next steps

The results, once validated via RSSB quality assurance processes, will be provided to the duty holders for use in their safety cost-benefit analyses (CBA) of the relevant top 10 improvement measures shown in Figure 3, as part of the safety decision making process regarding implementation of these improvements.

2.2 Work stream 2: Frequency Report - Identification of changes to the railway over last 10 years and in the future that could change our understanding of the way in which these types of derailment can occur and the way they are modelled/assessedThis report was published in June 2016 [9] and presents the results of an analysis to determine the derailment frequency of freight wagons on GB railways between 2000 and 2014. Certain wagons are known to have larger numbers of derailments than others. However, unless the number of derailments is normalised by fleet size and by exposure to the risk, it is not possible to be certain whether a wagon is more prone to derail because of inherent design features or simply because it encounters severe track faults more often. This report attempts to address this question by normalising by the miles run for individual wagon fleets.

A total of 41 track twist derailments occurred between 2000 and 2014 involving 20 types of wagon, of which 5 types have since been withdrawn from service. Taking the wagon fleets which have experienced track twist derailments as a whole, such derailments occur on average at a rate of approximately 1 derailment for every 61 million fleet miles run. However, these are not distributed evenly and many vehicle types have experienced no derailments in the period studied. For the total GB wagon fleet the derailment rate is much lower (it was not calculated in this study due to the complexities of doing so across all wagon types).


With a particular focus on intermodal container wagons, the data for this period shows that FEA wagons had the second highest number of track twist derailments with 6 in total. ZBA wagons had the highest with ten. However, once these are normalised by fleet mileage, FEA wagons, with one derailment every 94 million fleet miles, have a lower derailment frequency than 13 other wagon types. The two highest derailment frequencies amongst wagons still operating on the network were 11 and 15 million miles per derailment for IFB and FAA wagons respectively, around 8 times higher than the frequency for FEA wagons. Wagons with a higher derailment frequency than FEA wagons included a number of other intermodal container wagons such as FCA, FIA, FCA, IFB and FAA. Conversely, there are a number of examples of container wagons, also with Y-series bogies, which have run similar fleet mileages without derailing, such as FSA wagons. As these will experience similar traffic and loading conditions, it is reasonable to conclude that there are detail differences in the design of these wagons which make one type more prone to derailments than the other.

A total of 12 cyclic top derailments occurred between 2000 and 2014 involving 10 types of wagon, of which only 1 type has since been withdrawn from service. Cyclic top derailments occur on average, for the wagon fleets which have experienced derailments, at a rate of approximately 1 derailment for every 127 million fleet miles run, significantly lower than the rate for track twist derailments. Again, the derailment rate for the whole fleet, including those wagons which have not derailed, is much lower but was not calculated for the reasons described above. Of the wagons involved in these derailments, FEA and FSA container wagons had the lowest derailment rates with one derailment every 562 and 250 million miles respectively. The remaining 8 wagon types all had derailment frequencies worse than 1 in 150 million fleet miles.

Two axle wagons with UIC double-link or BR long link suspensions with taperleaf springs feature prominently amongst those wagons with the very highest derailment frequencies.

Freight wagon derailments are increasingly rare events. Many of the wagons types discussed in this report only experienced one derailment within the 15-year period studied. As a result the ranking of wagons within the results could be significantly altered by a single additional derailment to one wagon type.


8

2.3 Work stream 3: Wavelengths - Assess the potential benefits that could be gained from adding an additional wavelength track twist measurement/criteria.The report on the work carried out by Network Rail was issued to the XIFDWG in June 2017 [10] and presents the work to assess any opportunities in measuring track twist over an alternative base length to the existing 3m measurement.

An assessment was made to understand the occurrence of track twist using a range of alternative base lengths (1.8 m up to 18 m). To enable a fair comparison against the maintenance limits, a twist gradient value was used for each twist threshold requiring an action (i.e. from ‘Planned Maintenance’ to ‘Block The Line’). A total of 126 miles were analysed over a freight dominant route between Felixstowe Beach and Ipswich. Results indicated that zero track twists were identified when assessing with a base length greater than 3 m. All longer base length track twists that were close to the Intervention Limit were related to entering and exiting curve transitions. Intervention Limit alerts were observed when using a base length <3 m, which was a direct result of increasing the sensitivity of measurement. Moving to a <3.0 m twist measurement would significantly increase the number of Intervention Limit alerts and hence have a significant impact on maintenance.

The occurrence of locations containing combined track twist base lengths was completed on an additional 124 miles of track geometry data. To add a further level of scrutiny to the work, the ORE (Office of Research and Experiments) track twist limits were also used. Zero combination track twist measurements were identified when assessed over a combined total of 250 miles of track geometry data. One location with a potential combination was identified and related directly to a curve transition.

To understand what this meant from a risk perspective, a parameter study was completed to assess each alternative base length against existing requirements for derailment risk. Worst case derailment quotients and wheel unloading values were assessed, for each twist base length, and indicated that derailment risk fell well within the existing, allowable acceptance limit. Larger base length twist measurements predicted derailment quotients marginally outside the existing limit. However, zero track twists were identified using these longer base length measurements. In all circumstances, the degree of


wheel unloading remained significantly lower than the acceptance limit, indicating that derailment would not occur.

The Network Rail report complemented earlier findings of a study carried out by The Institute of Railway Research (IRR) at the University of Huddersfield (UoH) into the effects of twist wavelength on wheel unloading [11]. Both reports concluded that there is currently no evidence to suggest than an alternative track twist measurement should be implemented.

A paper summarising the results of the two projects is currently being prepared by Network Rail and IRR.

2.4 Work stream 4: Guidance Review - Review of existing loading practice and the guidance provided for vehicle testingCurrently this work stream is on hold, pending completion of Work stream 5: Simulation Studies - Assess the need and feasibility of testing and / or computer simulations of existing wagons to establish limiting offset loading conditions and related wheel unloading limits.

It is envisaged that a review of the existing loading practice and the guidance provided for vehicle testing will form part of a wider review of GMRT2141 by the Rolling Stock Standards Committee.

At the September 2017 meeting of XIFDWG it was agreed that the Steering Group for Workstream 5 Project T1119 - Simulating Offset Loading of Container Wagons on Twisted Track will make recommendations as to the next steps at their November 2017 meeting.


10

2.5 Work stream 5: Simulation Studies - Assess the need and feasibility of testing and/or computer simulations of existing wagons to establish limiting offset loading conditions and related wheel unloading limitsThis work is being progressed by The Institute of Railway Research (IRR) at the University of Huddersfield (UoH) as part of the RSSB Industry Research Programme through Project T1119 - Simulating Offset Loading of Container Wagons on Twisted Track.

The project aims to understand the impact that uneven loading of containers has on the likelihood of derailment, when travelling on twisted track. It is using accepted simulation methods commonly applied in the rail industry. This project is carrying out VAMPIRE® simulations in two work packages. The first work package (WP1) will simulate thousands of combinations of unevenly loaded containers to the test criteria outlined in appendices A and C of GMRT2141 (Resistance of Railway Vehicles to Derailment and Roll-Over)[12], including derailment case studies. The second work package (WP2) will validate real track conditions where incidents have occurred, on a subset of the scenarios simulated in WP1. If conclusive, changes to the GMRT2141 standard, to include eccentric loading scenarios, would be a key enabler for reducing freight derailment risk across the network.

The initial results of this work were presented at the September 2017 meeting of XIFDWG. It was agreed that these should be considered by the T1119 Project Steering Group and recommendations for next steps to be agreed at their November 2017 meeting.


2.6 Work stream 6: GOTCHA - Analysis of GOTCHA measurements to establish existing wagon uneven load profile distributionsThis work is being progressed by Network Rail.

The primary function of GOTCHA is to act as a wheel impact load detection system, reducing the risk of broken rails and track damage arising from defective wheelsets. Network Rail purchased the system for this purpose and has now installed it at 27 locations on the network.

However, as GOTCHA records individual wheel forces, the data can be used to calculate vehicle imbalance (measured as imbalance in the force applied by different wheels on the rails). This work stream is evaluating the capability to assess whether it can be used to help the industry manage the risk that asymmetrically loaded or defective vehicles contribute to the risk of flange-climb derailment.

The most recent statistical analysis of GOTCHA data has shown that the data should be capable of being used to classify each vehicle passage into one of three categories:

highly likely to have an imbalance above a predefined limit

highly unlikely to have an imbalance above a predefined limit

where (because of uncertainty in the measurement) it is not possible to tell whether the recorded imbalance is above or below a predefined limit.

This appears to be applicable for both lateral and longitudinal loading. The statistical model requires validation from measurement of vehicles and loads on the network. Additional validation has taken place to compare the same train passage over multiple GOTCHA sites. This has provided sufficient confidence to allow data to be used on a “for information” basis and for initial estimates on incidence of imbalanced loading to be produced.

Report design and testing is underway. In parallel, Network Rail are working to improve data provision internally to Network Rail and to automate some elements of the report generation to allow regular reporting to take place. This will provide the industry with both train specific data and to identify patterns and to monitor trends in the incidence of asymmetry to inform risk management.

The preliminary results were presented to the September 2017 XIFDWG meeting. A formal report is currently in preparation.


12

The work to date has focused on identifying lateral and longitudinal imbalances. Evaluation of GOTCHA’s ability to detect vehicle defects will commence once the data provision has improved.

2.7 Work stream 7: Container Loads - Identifying the scale of asymmetric container loading and controls in place to reduce and manage such loadingFollowing the XIFDWG meeting in September 2017 it was decided to rename this work stream from ‘Examine the rules and regulations around the loading and verification of containers as a wider transport issue’ to ‘Identifying the scale of asymmetric container loading and controls in place to reduce and manage such loading’ to more accurately reflect the core issues it is addressing.

This work stream aims to identify the scale of asymmetric loading, and controls in place to reduce & manage such loading including the rules and regulations around the loading and verification of containers as a wider transport issue and how well they are implemented.

XIFDWG suggested in the past that the existing guidelines such as the IMO Guidelines for packing of cargo transport units (CTU) are adequate [13], however, the problem is they are not always applied and are very difficult to enforce. XIFDWG would like enforcing authorities with responsibilities for duty holders involved in packing and transporting CTUs to investigate options to promulgate and enforce these packing guidelines. This is a multimodal transport system problem, and so the risk associated with carrying these unevenly loaded containers is not exclusive to rail. ORR has on-going liaison at a working level with HSE, MCA and DVSA, and has raised asymmetric loading as an issue. RSSB, supported by ORR is exploring options to raise the issue at a higher level in DFT and also via PACTS - Parliamentary Advisory Council for Transport Safety.

A key part of this work stream is RSSB research project T1112 - Quantifying Offset Loading of Container Wagons was undertaken to identify the proportion of eccentrically packed containers transported within the United Kingdom. The investigation was able to gain a good understanding of the extent and frequency of asymmetrically loaded containers but was unable to determine information about the movement of these containers (i.e. origin,


destination, mode of transport) as information on modal split is confidential and not available in the project. Following review of the outputs, some further work was requested from the supplier to address further commentary and improve the presentation of the dataset in an updated final report made available to the XIFDWG for their September 2017 meeting.

Research project T1119 - Simulating Offset Loading of Container Wagons on Twisted Track currently being undertaken within Work stream 5, is investigating what the ‘safe parameters’ are for containers being transported by rail. Following this there is a requirement to revisit the outputs from T1112 to consider the risk of derailment as a consequence of combinations of packing eccentricities and track twist.

The output of this additional work may, in turn, require a further assessment of the container packing eccentricities to refine the risk to the rail infrastructure through the provision of the additional data that this report has not been able to provide. Providing this data will require at least a further year of data collection and will require the cooperation of the major shipping companies in the provision of the required data.

2.8 Work stream 8: RAIB Recommendations - Recommendations MappingThe Rail Accident Investigation Branch (RAIB) has made a number of recommendations to various industry parties in relation to a number of freight train derailments that have recently occurred [Appendix B]. The Rolling Stock and Infrastructure standards committees proposed that the cross-industry group (XIFDWG) should consider these recommendations in its work.

Work streams underway as part of the XIFDWG programme of work are providing information that supports those to whom RAIB recommendations are addressed, in closing them.

The XIFDWG has considered the recommendations and, in undertaking the task set for it by the ORR, is providing any relevant information to support members in addressing their particular recommendations.

However, for those particular organisations, rather than the XIFDWG, to address the recommendations with the regulator. ORR is pursuing progress of these recommendations with the duty holders through its RIAB recommendation handling process.

Appendix B lists RAIB recommendations relevant to the activities of the group.


14

3 Summary of Next Steps

3.1 Work stream 1: Risk AssessmentThe results of the RRSB update to the TRL/RSSB risk analysis report are being validated via RSSB quality assurance processes and the final report is being developed. A draft of the final report will be available in time for the November 2017 XIFDWG, which will then be finalised via correspondence and published in December 2017. The results are being provided to the duty holders for use in their safety cost-benefit analyses (CBA) of the relevant top 10 improvement measures shown in Figure 3. RSSB has offered to provide technical assistance to those organisations that request it.

3.2 Work stream 2: Frequency ReportThis work is complete and the report published.

3.3 Work stream 3: WavelengthsThe primary work on the Network Rail report is complete. Network Rail are currently finalising some further clarifications requested from ORR before formal sign-off of the report. They are aiming to formally close this work out (i.e. with ORR clarifications addressed) at the next XIFDWG meeting in November.

The University of Huddersfield IRR report is complete and published.

A joint summary paper of the two reports is being produced by Network Rail and IRR and is expected to be finalised in time for the next XIFDWG meeting in November 2017.

3.4 Work stream 4: Guidance Review This work stream is currently being progressed. Following the September XIFDWG IRR presentation of the preliminary results for RSSB R&D Project T1119 - Simulation Studies (under Work stream 5) the project steering group is expected to make recommendations as to the next steps with reference to any potential new thresholds at their next meeting in November 2017. These will then be tabled at the next XIFDWG meeting also planned for November 2017.


3.5 Work stream 5: Simulation StudiesThe preliminary results were presented at the September 2017 meeting of XIFDWG. The results will be considered at the November 2017 meeting of the project steering group with recommendations expected at o next steps regarding setting of potential new thresholds.

3.6 Work stream 6: Analysis of GOTCHA measurementsPreliminary findings were presented at the September 2017 XIFDWG meeting.

Further report design, system testing and data capture and analysis is continuing. Network Rail are working to improve data provision and to automate some elements of the report generation to allow regular reporting to take place. This will provide the industry (FOCs) with both train specific data and to identify patterns and to monitor trends in the incidence of asymmetry.

3.7 Work stream 7: Container LoadsThe final report for RSSB R&D project T1112 – Quantifying Offset Loading of Container Wagons is complete and published. There is no further work currently planned under this work stream, unless it is specifically requested by the steering group for RSSB R&D Project T1119 - Simulation Studies (under Workstream 5).

3.8 Work stream 8 - RAIB Recommendations MappingThis work is complete with the ORR now pursuing progress of the relevant recommendations directly with the duty holders, through its RIAB recommendation handling process.


16

4 References

1 Letter from Ian Prosser to Network Rail, Freightliner, GBRF, DB Schenker, DRS, Colas, RSSB, RAIB and HSE, titled ‘Recent freight derailment: The interaction of track, vehicles and freight container loads, and potential areas for improvement’, 5 December 2014.

2 National freight safety Group (NFSG) https://www.rssb.co.uk/groups-and-committees/rssb-board/industry-safety-groups-supported-by-rssb/national-freight-safety-group

3 Taking Safe Decisions, http://www.rssb.co.uk/risk-analysis-and-safety-reporting/risk-analysis/taking-safe-decisions , RSSB.

4 ORR Guidance on the application of Commission Regulation (EU) 402/2013 March 2015 - Common Safety Method for Monitoring and the Common Safety Method on Risk Evaluation and Assessment, http://www.orr.gov.uk/rail/health-and-safety/health-and-safety-laws/european-railway-safety-legislation/csm-for-risk-evaluation-and-assessment

5 Common safety methods for risk assessment, http://www.era.europa.eu//Core-Activities/Safety/Safety-Management-System/Pages/Risk-Assessment.aspx

6 RSSB, Cross-Industry Freight Derailment Working Group Interim Progress Report, Version 3.4, February 2016, https://www.rssb.co.uk/Library/risk-analysis-and-safety-reporting/2016-02-cross-industry-freight-derailment-working-group-interim-progress-report.pdf

7 PPR691 ‘Potential risk to road and rail transport associated with asymmetric loading of containers’, RSSB and TRL 2014

8 RSSB, Cross-Industry Working Group on Freight Derailment Bowtie Risk Analysis Report, October 2016, https://www.rssb.co.uk/Library/risk-analysis-and-safety-reporting/2016-10-FreightDerailmentBowtieRiskAnalysisReport.pdf

9 IRR, Report on Calculation of Freight Vehicle Track Twist and Cyclic Top Derailment Frequency, https://www.sparkrail.org/Lists/Records/DispForm.aspx?ID=24150

10 Network Rail, Alternative base length Track Twist Measurement Report, June 2016 ‘a’

11 IRR 110/139 The Effects of Twist Wavelength on Wheel Unloading

12 GM/RT 2141 Resistance of Railway Vehicles to Derailment and Roll-Over, RSSB, 2009


https://www.rssb.co.uk/groups-and-committees/rssb-board/industry-safety-groups-supported-by-rssb/national-freight-safety-group



http://www.rssb.co.uk/risk-analysis-and-safety-reporting/risk-analysis/taking-safe-decisions

http://www.rssb.co.uk/risk-analysis-and-safety-reporting/risk-analysis/taking-safe-decisions

http://www.orr.gov.uk/rail/health-and-safety/health-and-safety-laws/european-railway-safety-legislation/csm-for-risk-evaluation-and-assessment



http://www.era.europa.eu//Core-Activities/Safety/Safety-Management-System/Pages/Risk-Assessment.aspx

http://www.era.europa.eu//Core-Activities/Safety/Safety-Management-System/Pages/Risk-Assessment.aspx

https://www.rssb.co.uk/Library/risk-analysis-and-safety-reporting/2016-02-cross-industry-freight-derailment-working-group-interim-progress-report.pdf

https://www.rssb.co.uk/Library/risk-analysis-and-safety-reporting/2016-02-cross-industry-freight-derailment-working-group-interim-progress-report.pdf

https://www.rssb.co.uk/Library/risk-analysis-and-safety-reporting/2016-10-FreightDerailmentBowtieRiskAnalysisReport.pdf



https://www.sparkrail.org/Lists/Records/DispForm.aspx?ID=24150

https://www.sparkrail.org/Lists/Records/DispForm.aspx?ID=24150

13 IMO / ILO / UN ECE Code of practice for packing cargo transport units. www.unece.org/fileadmin/DAM/trans/doc/2014/itc/id_07_CTU_Code_January_2014.pdf

14 RAIB report 21/2014 Derailment at Primrose Hill / Camden Road West Junction, 15 October 2013

15 RAIB report 11/2015 Freight train derailment at Angerstein Junction, 2 April 2014

16 RAIB Report 02/2013 Freight train derailment at Reading West Junction, 28 January 2012

17 RAIB Report 20/2014 Freight train derailment near Gloucester 15 October 2013

18 RSSB Updated risk assessment report, 2017 (in progress)


18

5 Abbreviations

Acronym Definition

ACCTRAF Actual Traffic Data

HSE Health and Safety Executive

IMO International Maritime Organisation

GOTCHAGOTCHA is the Wheel Impact Load Detection

system installed on NR infrastructure

IRRInstitute of Rail Research, University of

Huddersfield

MCA Maritime and Coastguard Agency

NFSG National Freight Safety Group

ORR Office of Rail and Road

RAIB Rail Accident Investigation Branch

SOLAS Safety of Life at Sea (IMO Regulations)

SRM Safety Risk Model (RSSB)

XIFDWG Cross-industry Freight Derailment Working Group


Appendices

Appendix A: List of group members

George Bearfield, RSSB, Chair

Hugh O’Neill, RSSB

Neil Halliday, RSSB

Bridget Eickhoff, RSSB

Lynne Collis, RSSB

Mike Laskowski, RSSB

David Griffin, RSSB

Richard Thomas, ORR

Mark Burstow, Network Rail

Colin Newsome, Network Rail

Ian Coleman, Network Rail

Deanne Haseltine, Network Rail

Tim Gabb, Freightliner

Andy Martlew, DRS

Ian Dougherty, DB Cargo

David Barney, Private Wagon Owners Federation

Simon Blake, Aggregate Industries

Julian Stow, Institute of Rail Research

Gareth Tucker, Institute of Rail Research

Philip Rogers, SNC Lavalin

Joss Apps, Ricardo Rail


20

Appendix B: Relevant RAIB recommendations

Recommendations from RAIB report 21/2014 (Camden Road), [14] 2 & 3 are relevant to the XIFDWG.

2. The intent of this recommendation is for the key stakeholders in the railway industry to work together to assess the risk from asymmetric loading and to identify and adopt reasonably practicable control measures to mitigate that risk.

Freightliner and Network Rail should jointly request that RSSB:

a Researches the factors that may increase the probability of derailment when container wagons are asymmetrically loaded, and in particular:

i sensitivity to combinations of longitudinal and lateral offsets in loads that can reasonably be encountered in service;

ii the predicted performance of wagons with high torsional stiffness along their length (using the FEA type as an example); and

iii the effect of multiple twist faults, track twist over distances other than 3 metres (as commonly specified and measured by Network Rail) and lateral track irregularities.

b updates and amends as necessary the risk assessment contained within the RSSB and Transport Research Laboratory joint report (‘Potential risks to road and rail transport associated with asymmetric loading of containers’); this should take into account the results from the research referred to in a) and additional evidence presented in this investigation report; and

c works with industry stakeholders to use the outputs of a) and b) to identify, evaluate and promote adoption of any additional reasonably practicable mitigations capable of reducing the risk from asymmetric loading of wagons (paragraphs 128c, 130a, 130b and 131b).

3. The intent of this recommendation is to clarify the requirements for the design and acceptance of freight wagons, taking account of the possibility of asymmetric loading.

RSSB should amend Railway Group Standard ‘Resistance of Railway Vehicles to Derailment and Roll-Over’, GM/RT2141 to refer specifically to


asymmetric loading, including possible combinations of longitudinal and lateral load imbalance (paragraph 131a).

Recommendations from RAIB report 11/2015 (Angerstein Junction) [15] 1, 2, 4, 5 & 6 are relevant to the XIFDWG.

1 The intention of this recommendation is to prevent wagons operating on the network with unacceptable uneven retained loads after unloading.

Aggregate Industries, in consultation with relevant train operators, should review its processes for discharging aggregate hopper wagons, and for inspection of train loading and condition prior to despatch, to ensure that the risks arising from uneven residual loads are identified and effectively managed. Aggregate Industries should then implement appropriate control measures to mitigate this risk so far as is reasonably practicable (paragraphs 153(c), 153(e) and 154(a)).

2 The intention of this recommendation is to manage the contribution that diagonal wheel unloadings, due to twisted bogie frames or other defects, make to derailment risk. The RAIB notes that action taken in response to this recommendation could be informed by work undertaken as part of the railway industry’s response to the ORR’s letter of 5 December 2014 (paragraph 163).

RSSB, in conjunction with freight wagon operators, freight operating companies and entities in charge of maintenance for freight wagons, should review the extent to which diagonal wheel unloadings are present within freight wagon bogies that are operating on Network Rail infrastructure, and the contribution that this makes to derailment risk. This review should consider:

identifying the magnitude and prevalence of diagonal wheel unloadings caused by bogie frame twist (and other possible causes)

proposing criteria for acceptable levels of diagonal wheel unloading, or for bogie frame twist

proposing proportionate measures for identifying, and then managing, unacceptable diagonal wheel unloadings (paragraphs 153(d) and 155(b)).

4 The intention of this recommendation is to review whether the historic track twist measurement base (3 metres) is still a sufficient control for track twist risk applicable to current rolling stock. The RAIB notes that this


22

recommendation could be informed by the joint industry action taken in response to ORR’s letter of 5 December 2014 (paragraph 163).

Network Rail should liaise with RSSB to review whether the existing 3 metre measurement base used for identification of track twist is sufficient for managing the derailment risk applicable to rolling stock currently operating on Network Rail infrastructure. If found to be inadequate or insufficient, Network Rail should:

update its process for assessing track twist by the inclusion of additional and/or alternative measurement bases

implement a time-bound plan to apply the new process to all of its infrastructure (paragraphs 153(b) and 155(a)).

5 The intention of this recommendation is to encourage use of available monitoring data from wheel impact load detection systems, such as GOTCHA, to inform rolling stock maintenance.

Network Rail should review the potential to use wheel impact load detection system data to provide information about possible defects, such as uneven wheel loading or uneven load distribution, relating to specific wagons. The review should include consideration of how this information could be used to improve control of overall derailment risk (such as identifying the need for entities in charge of maintenance to check the condition of suspect wagons and take appropriate remedial action). Network Rail should seek inputs from relevant entities in charge of maintenance as part of the review. If justified by the review, Network Rail should implement track side and reporting processes needed for collecting

6 The intention of this recommendation is to ensure that the distribution of loads in wagons, including partly loaded wagons, is controlled in a manner compatible with wagon and track characteristics. The RAIB notes that action taken in response to this recommendation could be informed by work undertaken as part of the railway industry’s response to the ORR’s letter of 5 December 2014 (paragraph 163).

RSSB, in consultation with industry, should review the risks associated with the uneven loading of wagons, with particular reference to partial loads, and propose any necessary mitigation, so that the extent of permitted load imbalance is effectively controlled (paragraphs 153(c), 154(a), 154(b) and 155(b).


Recommendations from RAIB Report 02/2013 Freight train derailment at Reading West Junction [16] 3 is relevant to the XIFDWG.

3. The intention of this recommendation is for inter-modal freight terminal operators to develop requirements and investigate introducing a suitable monitoring system, for use during routine container and train handling, to prevent freight container wagons entering traffic with a side-to-side wheel load imbalance. The system could be based on the measurement of individual or side-to-side wheel loads prior to the train entering traffic or the identification of freight container load offsets during lifting.

Freightliner should develop requirements for a system to monitor and prevent load offsets from containers resulting in wagons with a side-to-side wheel load imbalance entering traffic from its terminals. The system should be considered when terminal equipment is planned to be installed or upgraded, and where practicable the system should be implemented (paragraphs 108a, 108a.ii and 108a.iii).

This recommendation may also be applicable to other operators of inter-modal freight terminals.

Recommendations from RAIB Report 20/2014 Freight train derailment near Gloucester [17] 1. These are not relevant to the XIFDWG current work but 5 & 7 may be relevant to a later phase of work of the XIFDWG on cyclic top.

5. The intent of the recommendation is to ensure that when a vehicle’s dynamic behaviour is assessed to identify whether its ride performance is compatible with the railway infrastructure in Great Britain (this may include infrastructure that does not comply with Technical Specifications for Interoperability), the susceptibility of its ride performance to track geometry with cyclic top is included in this assessment.

RSSB, in conjunction with Rolling Stock Standards Committee, should carry out a review to identify how a vehicle’s response to regular changes in vertical track geometry should be assessed (a cyclic top assessment). RSSB should then propose changes to the standards which are used assess the compatibility of vehicle’s ride performance with the railway infrastructure in Great Britain (at present this is Railway Group Standard GM/RT2141),

1 Cyclic Top (Gloucester Recs 5, 6, 7) is recognised as an issue within scope of the XIFDWG but is a separate issue from that of offset loading. The XIFDWG agreed to start by concentrating on the track twist, vehicle faults and the loading issue. These recommendations are included with reference to vehicle testing.


24

which will implement the cyclic top assessment identified by the review. The proposed changes to the standards, as agreed by Rolling Stock Standards Committee, should then be implemented by RSSB by means of a time bound programme (paragraphs 194b.i, 194b.ii and 195c).

7. The intent of the recommendation is to highlight the risk that a wagon may be susceptible to riding problems if it is designed with a bogie centre spacing distance that is the same as a wavelength commonly associated with cyclic top track defects.

RSSB, in conjunction with Rolling Stock Standards Committee, should propose that guidance on the design of freight wagons in document GM/GN2688 is amended, to explain that as well as two-axle wagons, if a wagon is designed with a bogie centre spacing that matches a wavelength commonly associated with cyclic top, it may be susceptible to poor ride on jointed track and cyclic top (paragraph 196c).


[email protected]

http://www.rssb.co.uk

RSSB

Floor 4, The Helicon

1 South Place

London

EC2M 2RB

http://www.rssb.co.uk

mailto:[email protected]

cross-industry freight derailment working group … rssb | cross-industry freight derailment working...

Documents