Reduction Factors for Estimating the Probability of Failure of Mechanical Damage Due to External Interference Available to Purchase

Quantified risk assessments (QRAs) are widely used in the UK to assess the significance of the risk posed by major accident hazard pipelines on the population and infrastructure in the vicinity of the pipeline. A QRA requires the calculation of the frequency of failures and the consequences of failures. One of the main causes of failures in onshore pipelines is mechanical damage due to external interference, such as a dent, a gouge, or a dent and gouge. In the published literature, two methods have been used to calculate the probability of failure due to external interference: • historical failure data and • limit state functions combined with historical data (i.e. structural reliability-based methods). Structural reliability-based methods are mathematically complicated, compared to using historical failure data, but have several advantages, e.g. extrapolation beyond the limited historical data, and the identification of trends that may not be apparent in the historical data. In view of this complexity, proposed supplements to the UK pipeline design codes IGE/TD/1 (natural gas) and PD 8010 (all substances) — on the application of QRAs to proposed developments in the vicinity of major accident hazard pipelines — include simple ‘reduction factors’ for use in ‘screening’ risk assessments. These ‘reduction factors’ are based on a comprehensive parametric study using a structural reliability-based model to calculate the probability of failure due to mechanical damage, defined as: gouges, and dents and gouges. The two ‘reduction factors’ are expressed in terms of the design factor and wall thickness of the pipeline. It is shown that, through appropriate normalisation, the effects of diameter, grade and toughness are secondary. Reasonably accurate, but conservative, estimates of the probability of failure can be obtained using these ‘reduction factors’. The proposed methodology is considerably simpler than a structural reliability-based analysis. The development and verification of these ‘reduction factors’ is described in this paper.