1 / 38

Why Risk Models Should be Parameterised

Why Risk Models Should be Parameterised. William Marsh, william@dcs.qmul.ac.uk Risk Assessment and Decision Analysis Research Group. Acknowledgements. Joint work with George Bearfield Rail Safety and Standards Board (RSSB), London. Aims. Introduce idea of a ‘ parameterised risk model ’

hua
Download Presentation

Why Risk Models Should be Parameterised

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Why Risk Models Should be Parameterised William Marsh, william@dcs.qmul.ac.uk Risk Assessment and Decision Analysis Research Group

  2. Acknowledgements • Joint work with George BearfieldRail Safety and Standards Board (RSSB), London

  3. Aims • Introduce idea of a ‘parameterised risk model’ • Explain how a Bayesian Network is used to represent a parameterised risk model • Argue that a parameterised risk model is • Clearer • More useful

  4. Outline • Background • Risk analysis using fault and event trees • Bayesian networks • An example parameterised risk model • Using parameterised risk model

  5. no 80% no 95% no 95% yes 20% yes 5% no 75% yes 5% no 95% yes 25% yes 5% Fault and Event Trees Outcome • Quantitive Risk Analysis Hazardous event OR Events AND AND Base event

  6. RSSB’s Safety Risk Model • 110 hazardous events • Fault and event trees • Data from past incidents • UK rail network • Average • Used to monitor risk for rail users and workers • Informs safety decision making

  7. Mild Normal Severe 70% 20% 10% Conditional Probability Table Yes No 80% 20% Bayesian Networks • Uncertainvariables • Probabilistic dependencies Bayes’ Theorem Incline Speed Fall

  8. Bayesian Networks • Uncertainvariables • Probabilistic dependencies • Efficient inference algorithms Bayes’ Theorem Mild Normal Severe 0% 0% 100% Incline Speed Fall Yes No 60% 40%

  9. Example Parameterised Risk Model

  10. Falls on Stairs • Falls on stairs common accident • 500 falls on stairs / year (2001) • Influenced by • stair design & maintenance • the users’ age, gender, physical fitness and behaviour • Injuries • Non fatal: bruises, bone fractures and sprains … • Fatal injuries: fractures to the skull, trunk, lower limbs

  11. Fault Tree Lose Footing OR GATE 2 AND AND Misstep GATE 3 GATE 4 TripHazard Imbalance Slip Inattention

  12. Events and Outcomes Break LoseFooting Holds Falls sideways Vertical yes Forward-short forward no Forward-long drops yes Backward-short backward no Backward-long holds Startled

  13. Events and Outcomes Break LoseFooting Holds Falls sideways Vertical yes Forward-short forward no Forward-long drops yes Backward-short backward no Backward-long holds Startled

  14. Can the Model be Generalised? • Logic of accidents same (nearly) but numbers vary with design • Reuse logic • Estimating probabilities once only

  15. Factors – Risk Model Parameters • Factors with discrete values

  16. Factors to Base Events • Base event probabilities depend on factors

  17. Factors to Events • Probabilities of event branches depend on factors • … also on earlier events

  18. FT Bayesian Network

  19. Event Tree Bayesian Network

  20. Accident Injury Score (AIS) • Harm from accident

  21. Complete Bayesian Network AgenaRisk see: http://www.agenarisk.com/

  22. Explicit Factors make Clearer Models • Are there factors in the fault or event tree? Age Break LoseFooting Holds Falls sideways Vertical yes Forward-short forward no Forward-long yes Backward-short backward no drops Backward-long yes Forward-short forward no Forward-long yes Backward-short backward no holds Backward-long Startled

  23. Using the Parameterised Model Reuse of the model Modelling multiple scenarios

  24. Prior probability distribution Using the Parameterised Model • Observe (some) factors

  25. Using the Parameterised Model • Suppose 3 stairs • Value of each observed factor

  26. Results – Outcome • Probability distribution • Outcome of a ‘stair descent’ • Hidden ‘nothing happens’ outcome

  27. Results – Accident Injury Score

  28. System Risk • University has many stairs in different buildings • How to assess the total risk? • Solution 1 • Used parameterised model for each stairs • Aggregate results • Solution 2 • Model ‘scenario’ in the Bayesian Network • Scenario: each state has shared characteristics e.g. geographical area

  29. Scenario • Each value is a ‘scenario’ for which we wish to estimate risk Could be each staircase

  30. Imprecise Scenarios • Imagine three departments • Factors do not have single value • Probability distribution over factor values

  31. Departments Proportion of events in each scenario Exposure • Some scenarios more common • Distribution of ‘stair descents’

  32. Using the System Model • Use 1 • Select a scenario • … like the parameterised model • Scaled by total system events

  33. Using the System Model • Use 2 • Whole system risk, • … weighted by exposure for each scenario

  34. Parameterised Risk Models in Practice Improving Safety Decision Making

  35. Better Safety Decision Making • Safety benefits of improvements • Existing models only support system-wide improvements • Detection of local excess risk • E.g. poor maintenance in one area • Requires risk distribution (not average) • … variations in equipment type and condition • … procedural and staffing variations

  36. Risk Profile: Sector and Network

  37. Derailment Event tree Investigation found the cause to be: ‘the poor condition of points 2182A at the time of the incident, and that this resulted from inappropriate adjustment and from insufficient maintenance ….’ Factors Fault tree

  38. Summary • Parameterised ET + FT • Using Bayesian Networks • Factors made explicit • Clearer and more compact • Reuse of risk model • Risk profiles • Guide changes to reduce risk • Challenge of including more causes Thank You

More Related