[en] A new statistical test with applications to censored lifetime data has been found. This test makes no assumptions about the dependence structure underlying the censoring. Here we describe the background to the test and give an algorithm to implement it. A number of numerical examples are given to illustrate how the algorithm may be applied

[en] A model for availability growth is developed to capture the effect of systemic risk prior to construction of a complex system. The model has been motivated by new generation offshore wind farms where investment decisions need to be taken before test and operational data are available. We develop a generic model to capture the systemic risks arising from innovation in evolutionary system designs. By modelling the impact of major and minor interventions to mitigate weaknesses and to improve the failure and restoration processes of subassemblies, we are able to measure the growth in availability performance of the system. We describe the choices made in modelling our particular industrial setting using an example for a typical UK Round III offshore wind farm. We obtain point estimates of the expected availability having populated the simulated model using appropriate judgemental and empirical data. We show the relative impact of modelling systemic risk on system availability performance in comparison with estimates obtained from typical system availability modelling assumptions used in offshore wind applications. While modelling growth in availability is necessary for meaningful decision support in developing complex systems such as offshore wind farms, we also discuss the relative value of explicitly articulating epistemic uncertainties. - Highlights: • A new model is developed for system availability growth. • The model is motivated by and applied to the offshore wind industry. • The general model is applicable to systems where availability performance growth is important. • Systemic risks on system performance are aggregated from subassembly level modelling. • Availability-informed capability can be predicted under different intervention scenarios.

[en] This paper reports a sensitivity analysis of the Cognitive Reliability and Error Analysis Method for Human Reliability Analysis. We consider three different aspects: the difference between the outputs of the Basic and Extended methods, on the same HRA scenario; the variability in outputs through the choices made for common performance conditions (CPCs); and the variability in outputs through the assignment of choices for cognitive function failures (CFFs). We discuss the problem of interpreting categories when applying the method, compare its quantitative structure to that of first generation methods and discuss also how dependence is modelled with the approach. We show that the control mode intervals used in the Basic method are too narrow to be consistent with the Extended method. This motivates a new screening method that gives improved accuracy with respect to the Basic method, in the sense that (on average) halves the uncertainty associated with the Basic method. We make some observations on the design of a screening method that are generally applicable in Risk Analysis. Finally, we propose a new method of combining CPC weights with nominal probabilities so that the calculated probabilities are always in range (i.e. between 0 and 1), while satisfying sensible properties that are consistent with the overall CREAM method

[en] In the past, standard reliability and risk approaches have sufficed to identify the dominant causes of failure in forensic analyses, and the dominant risk contributors for proactive risk investigations. These techniques are particularly applicable when individual or even simple common failure events of a similar type dominate the analysis. However, nowadays due to increased understanding of the 'simple' mechanisms and the increasing complexity of the systems we build, failures in highly dependable systems arise from unexpected interactions between subsystems and the external and internal environment. Engineering data analysis is the process of data collection and investigation from a variety of perspectives, alternatively dissecting it into its underlying (yet often unknown) patterns; this process is becoming ever more necessary as systems become more complex. Some of the techniques employed are slicing the data sets according to known underlying variables, or overlaying data gathered from different perspectives, or imbedding data into previously established logical or phenomenological structures. This paper addresses the issues involved in visualizing patterns in data sets by providing examples of interesting maps from the past, indicating some of the maps currently in use, and speculating on how these visual maps might be developed further and used in the future to discover problems in complex systems before they lead to failure. Guidance is proposed as to how to explore and map data from different technical perspectives in order to evoke potentially significant patterns from reliability data. The techniques presented have been developed by combining approaches to common cause failure (CCF) classification with multidimensional scaling (MDS) to produce a new method for exploratory engineering data mapping

[en] The availability of a system under a given failure/repair process is a function of time which can be determined through a set of integral equations and usually calculated numerically. We focus here on the issue of carrying out sensitivity analysis of availability to determine the influence of the input parameters. The main purpose is to study the sensitivity of the system availability with respect to the changes in the main parameters. In the simplest case that the failure repair process is (continuous time/discrete state) Markovian, explicit formulae are well known. Unfortunately, in more general cases availability is often a complicated function of the parameters without closed form solution. Thus, the computation of sensitivity measures would be time-consuming or even infeasible. In this paper, we show how Sobol and other related sensitivity measures can be cheaply computed to measure how changes in the model inputs (failure/repair times) influence the outputs (availability measure). We use a Bayesian framework, called the Bayesian analysis of computer code output (BACCO) which is based on using the Gaussian process as an emulator (i.e., an approximation) of complex models/functions. This approach allows effective sensitivity analysis to be achieved by using far smaller numbers of model runs than other methods. The emulator-based sensitivity measure is used to examine the influence of the failure and repair densities' parameters on the system availability. We discuss how to apply the methods practically in the reliability context, considering in particular the selection of parameters and prior distributions and how we can ensure these may be considered independent—one of the key assumptions of the Sobol approach. The method is illustrated on several examples, and we discuss the further implications of the technique for reliability and maintenance analysis

[en] Current regulation in the UK and elsewhere specify upper and target risk limits for the operation of nuclear plant in terms of frequencies of various kinds of accidents and accidental releases per annum. 'As low as reasonably practicable' (ALARP) arguments are used to justify the acceptance or rejection of policies that lead to risk changes between these limits. We assess the suitability of cost-benefit analysis (CBA) and multi-attribute utility theory (MAUT) for performing ALARP ('as low as reasonably possible') assessments, in particular within the nuclear industry. Four problems stand out in current CBA applications to ALARP, concerning the determination of prices of safety gains or detriments, the valuation of group and individual risk, calculations using 'disproportionality', and the use of discounting to trade off risks through time. This last point has received less attention in the past but is important because of the growing interest in risk-informed regulation in which policies extend over several timeframes and distribute the risk unevenly over these, or in policies that lead to a non-uniform risk within a single timeframe (such as maintenance policies). We discuss the problems associated with giving quantitative support to such decisions. We argue that multi-attribute utility methods (MAUT) provide an alternative methodology to CBA which enable the four problems described above to be addressed in a more satisfactory way. Through sensitivity analysis MAUT can address the perceptions of all stakeholder groups, facilitating constructive discussion and elucidating the key points of disagreement. We also argue that by being explicitly subjective it provides an open, auditable and clear analysis in contrast to the illusory objectivity of CBA. CBA seeks to justify a decision by using a common basis for weights (prices), while MAUT recognizes that different parties may want to give different valuations. It then allows the analyst to explore the ways in which different parties might (or might not) come to the same conclusion even when weighting items differently. (author)

[en] In reliability modelling it is conventional to build sophisticated models of the probabilistic behaviour of the component lifetimes in a system in order to deduce information about the probabilistic behaviour of the system lifetime. Decision modelling of the reliability programme requires a priori, therefore, an even more sophisticated set of models in order to capture the evidence the decision maker believes may be obtained from different types of data acquisition. Bayes linear analysis is a methodology that uses expectation rather than probability as the fundamental expression of uncertainty. By working only with expected values, a simpler level of modelling is needed as compared to full probability models. In this paper we shall consider the Bayes linear approach to the estimation of a mean time to failure MTTF of a component. The model built will take account of the variance in our estimate of the MTTF, based on a variety of sources of information

[en] Classical approaches to estimating the rate of occurrence of events perform poorly when data are few. Maximum likelihood estimators result in overly optimistic point estimates of zero for situations where there have been no events. Alternative empirical-based approaches have been proposed based on median estimators or non-informative prior distributions. While these alternatives offer an improvement over point estimates of zero, they can be overly conservative. Empirical Bayes procedures offer an unbiased approach through pooling data across different hazards to support stronger statistical inference. This paper considers the application of Empirical Bayes to high consequence low-frequency events, where estimates are required for risk mitigation decision support such as as low as reasonably possible. A summary of empirical Bayes methods is given and the choices of estimation procedures to obtain interval estimates are discussed. The approaches illustrated within the case study are based on the estimation of the rate of occurrence of train derailments within the UK. The usefulness of empirical Bayes within this context is discussed

[en] Moment methods have been employed in decision analysis, partly to avoid the computational burden that decision models involving continuous probability distributions can suffer from. In the Bayes linear (BL) methodology prior judgements about uncertain quantities are specified using expectation (rather than probability) as the fundamental notion. BL provides a strong foundation for moment methods, rooted in work of De Finetti and Goldstein. The main objective of this paper is to discuss in what way expert assessments of moments can be combined, in a non-Bayesian way, to construct a prior assessment. We show that the linear pool can be justified in an analogous but technically different way to linear pools for probability assessments, and that this linear pool has a very convenient property: a linear pool of experts' assessments of moments is coherent if each of the experts has given coherent assessments. To determine the weights of the linear pool we give a method of performance based weighting analogous to Cooke's classical model and explore its properties. Finally, we compare its performance with the classical model on data gathered in applications of the classical model

[en] Probabilistic inversion is used to take expert uncertainty assessments about observable model outputs and build from them a distribution on the model parameters that captures the uncertainty expressed by the experts. In this paper we look at ways to use minimum information methods to do this, focussing in particular on the problem of ensuring consistency between expert assessments about differing variables, either as outputs from a single model or potentially as outputs along a chain of models. The paper shows how such a problem can be structured and then illustrates the method with two examples; one involving failure rates of equipment in series systems and the other atmospheric dispersion and deposition. - Highlights: • We look at a method to specify uncertainty distributions for coupled models. • We consider problems in which model parameters are unobservable quantities. • We use minimum information methods to avoid the problem of under-constraint. • We develop and implement a sequential approach to avoid over-constraint. • We give a method to assess the feasible region for multiple constraints