12 Failure rates and reliability
Obtaining failure rates for equipment and systems is essential for risk assessment. Failure rates can be fit to known distributions in order to model risk assessment for a particular system. Sometimes failure rates need to be computed from raw data. Use of a good data analysis package like pandas with python can perform a wide range of data analysis.
Reliability metrics
Determining how systems fail and devising metrics to characterize these systems is essential to risk assessment.
Weibull distribution
The Weibull distribution is a very important tool for risk assessment due to its versatility. Given that it is an exponential function with only a few parameters, curve fitting Weibull to failure data is flexible and can serve as a credible preliminary analysis tool when there is a lack of more precise data available. It is the Ferris Bueller of distributions. Having some working knowledge of the Weibull distribution and how to apply it, is a useful skill as part of risk assessment expertise. (There should be no surprise that python offers the capabilities to conduct reliability analysis with the Weibull function.)
Monte Carlo simulation
- Determining Reliability for Complex Systems
Part 1 – Analytical Techniques - Determining Reliability for Complex Systems
Part 2 – Simulation
Delphi techniques
Quantifying risk when there is a lack of data, such as operational cohorts or related historical data. The Delphi technique is a structured communication method, originally developed as a systematic, interactive forecasting method which relies on a panel of experts.