The role of investigation - before and after the accident

Size: px

Start display at page:

Download "The role of investigation - before and after the accident"

Jeffrey Johnson
5 years ago
Views:

1 The role of investigation - before and after the accident IOSH Railway Group Conference 27 November Simon French, Deputy Chief Inspector

2 Potters Bar, May 2002 Cause: points failure Outcome: derailment (97 mph) 7 fatalities, 11 injuries

3 Grayrigg, Feb 2007 Cause: points failure Outcome: derailment (95 mph) 1 fatality, 28 serious injuries

4 Potters Bar and Grayrigg some common factors A perception that the risk associated with stretcher bar failures was low An incomplete understanding of the performance of stretcher bars in service Limited process for the analysis of design, maintenance and inspection requirements Hidden weaknesses in risk control measures (eg. over-reliance on inspection and regular maintenance activities)

5 How can the railway industry better understand how its systems, sub-systems and components fail in service? 5

6 This is a gizmo (ie a system, sub-system, component or management system that plays a role in our business) 6

Top-down An approach based on a high level understanding of the risk presented by hazardous events How often do hazardous events occur and how bad are they?

7 Top-down An approach based on a high level understanding of the risk presented by hazardous events How often do hazardous events occur and how bad are they? How often has a failure of our gizmo been implicated? What does our experience to date tell us about the contribution of our gizmo to the overall risk profile of our business? 7

8 Bottom-up (1) An approach based on understanding the ways that things and people can fail and the potential consequences How is our gizmo performing in the real world? How does our gizmo fail, and how often does it do so? Based on experience to data, and all that we know about our gizmo, what other failure modes can be imagined, and how likely is each? What would be the consequences of failure of our gizmo, and what arrangements are in place to manage these? 8

9 Bottom-up (2) How does our gizmo interact with other gizmos? How can these interactions cause system failures? 9

10 So what s best? Both top-down and bottom-up analysis is needed to manage risk effectively. A good understanding of risk must be informed by both: - a good understanding of your risk profile; and - an understanding of how your gizmos can break and what will happen when they do 10

11 Top down Effective risk management Bottom up

12 Techniques for bottom-up analysis Precursor modelling Hazard and operability study Stress testing Hazard logging Failure mode and effects analysis Analysis of real-life performance Investigation of system failures and accidents 12

13 Have things moved on since Grayrigg? The industry is generally good at recording its accidents, using this data to assess its risk, and setting priorities for the management of risk ( top-down approach) The industry s understanding of risk is now informed by precursor modelling (eg RSSB s PIM) RAIB s investigations often reveal that more could have be done to understand hidden risk by applying bottomup analytical techniques to identify the ways that failures of systems/sub-systems, components and people can combine to cause catastrophic outcomes

14 The benefits of accident investigation Good accident investigations reveal how combinations of factors combined to create a dangerous event Accident investigations shine a searchlight into particular corners of the railway industry and therefore provide valuable intelligence Accident investigations demonstrate to those involved, those affected and the wider industry that action is being taken and lessons will be learnt

15 What does a good investigation look like? Good guidance on railway industry investigations is available from RSSB

final safety check) Poorly located car stop marker (relative to

16 Finding the hidden gaps - operations PTI accident at Brentwood (Jan 2011) Driver s over-reliance on door interlock (rather than a final safety check) Poorly located car stop marker (relative to DOO monitor) RAIB report 19/2011 Need for better risk assessment of PTI

similar crossings RAIB report 17/2013 Level crossing risk management process did not

17 Finding the hidden gaps - legacy design Fatal accident at Beech Hill (Dec 2012) The light output from the red wigwag light units was lower than specified at this, and many similar crossings RAIB report 17/2013 Level crossing risk management process did not adequately recognise and deal with the effect of sunlight on the visibility of crossing equipment

this was not identified or addressed Competence management systems did not

18 Finding the hidden gaps - human performance Collision at Norwich (July 2013) The driver was prone to lapses in concentration despite repeat indications, this was not identified or addressed Competence management systems did not sufficiently encompass non-technical skills or the management of fatigue RAIB report 09/2014

19 Conclusions Accident investigation is a key element of effective safety management An effective investigation will: o ensure that lessons are learnt and translated into meaningful actions o reveal gaps and areas of weakness in your control measures The railway industry has moved a long way in developing its own skills in this area The techniques of bottom-up analysis can be applied to the identification of precursors to accidents that have yet to happen

20 A closing thought Why not investigate an accident before it occurs? Many accidents could have been avoided had someone had asked the what if? question before the event

HSE health inspection initiative in construction webinar. 17 November 2014

HSE health inspection initiative in construction webinar 17 November 2014 Health and and Safety Executive Think health HSE inspection initiative 2014 Ian Strudley HM Principal Specialist Inspector Head