How to do modern pipeline risk assessment?

Pipeline Risk Management Chapter 1 highlights

• 1.1 Risk assessment at-a-glance
• 1.2 Risk: Theory and application
• 1.2.1 The Need for Formality
• 1.2.2 Complexity
• 1.2.3 Intelligent Simplification
• 1.2.4 Classical QRA versus
  Physics-based Models
• 1.2.5 Statistical Modeling
• 1.3 The Risk Assessment Process
• 1.3.1 Fix the Obvious
• 1.3.2 Using this Manual
• 1.3.3 Quickly getting answers
• 1.4 Pipeline Risk Assessment:
  Example 2
• 1.5 Values Shown are
  Samples Only

Implementing a Risk Assessment
The overall steps for assessment of pipeline risk under a modern risk assessment methodology, and consistent with this guideline, may be summarized as follows:

1. Define ‘failure’ and level of conservatism for the risk assessment
2. Exposure:  Estimate exposures from each threat at all points along the pipeline
   1. Degradation rate from time-dependent failure mechanisms
   2. Event rate from time-independent failure mechanisms
3. Mitigation:  Estimate combined effect of all mitigations at all points along the pipeline
   1. Identify all mitigation measures
   2. Rate effectiveness of each
   3. Use probabilistic summation to show aggregated effectiveness
4. Resistance:  Estimate the amount of resistance at all points along the pipeline
   1. Produce best estimate of current pipe strength by using pipe specifications with the “governing” information from:
      1. Pipe wall implied by last pressure test
      2. Pipe wall implied by last NDE inspection (including ILI, bell hole exam, etc)
      3. Pipe wall implied by assumption of leak-free at current NOP
      4. Possible pipe wall weaknesses (including era-of-manufacture/construction issues and threat interactions)
   2. Estimate pipe’s resistance to failure from each threat
5. Dynamically segment the pipeline based on collected data and estimates
6. Probability of Failure (PoF):  Calculate PoF from each threat for each segment PoF Triad:  combine Exposure, Mitigation, Resistance
   1. Produce time-to-failure (TTF) and then PoF estimates for time-dependent failure mechanisms
   2. Produce PoF estimates for time-independent failure mechanisms
   3. Combine all PoF’s
7. Consequence of Failure (CoF):  Identify representative consequence scenarios for each segment
   1. Estimate magnitude of consequence from each scenario, in verifiable measurement units
   2. Estimate frequency of each scenario
   3. Produce representative ‘consequence per failure’ value
8. Risk:  Produce risk values for each segment, perhaps in units of ‘expected loss’ (EL) such as $/mile-year