RCA para incidentes

Chapter 17

Some recent incidents and their implications

Introduction
Three Mile Island, Chernobyl, Bhopal, Mexico City, Zeebrugge and Kegworth are all place names which have become familiar to us through various media reports. Although they are many miles apart they share a common misfortune. They are places where major incidents and disasters have occurred; disasters in which many peoplelost their lives and the consequences of which are still being felt in terms of human pain and suffering. Many of the other costs of such disasters, social, environmental, economic and organizational, and their causes continue to be evaluated. Indeed causal analysis (see Chapter 4) is a well established part of accident investigation procedure and communication of causes of past incidents should,in theory, help prevent recurrence of similar incidents in the future. However, most major disasters provide us with ample evidence of the failures of organizations to learn from their own or other organizations’ previous experiences. For example, in the case of the Three Mile Island incident a similar accident had occurred some months before at the technologically similar Davis Besse plant(Embrey, 1989). In the Davis Besse incident, correct operator action had prevented an accident (the operator had responded to the meta-stable system state, see Chapter 5). The Zeebrugge enquiry also revealed that there had been several occasions prior to Zeebrugge when ferries had left port with their bow doors open, but without incident in these cases. These examples provide a salutary reminder thatorganizations should respond not only to consequences but to antecedent conditions.

Some recent incidents and their implications 291

This chapter will review several major accidents and incidents and, using such case histories, will highlight a number of key points relating to: 1. The importance of applying the methodology of reliability and risk assessment discussed in earlier chapters; 2.The need to design 'forgiving' systems and the importance of an integrated systems approach; The importance of communication in the prevention of 3. accidents and incidents; 4. The human, economic, social and environmental costs of disaster. It will also present a systems framework for accidents and incidents in line with previous chapters. The reader will be referred to additional references tosupport the case studies.

Systems models o accidents and incidents f
A systems approach to the analysis of accidents and incidents recognizes the inter-relatedness of the various components of the accident process (see Figure 7.1) and accepts multicausality (Cox and Cox, 1996). It is rarely the case that accidents occur as a result of machine and equipment (hardware) failures alone. More oftenthey occur as a result of a combination of organizational policy and procedures (software), human actions (liveware) and hardware failures. This approach is typified in the work of Embrey (1989) who has described a simple system-induced error model. This model is based on the hypothesis that all individuals have certain error tendencies. Furthermore, these error tendencies have to be combined witherror-inducing conditions (organizational time pressures and deadlines, occupational stress, distractions, etc.) for an error to result. For the error to give rise to a significant consequence for either safety or reliability, an unforgiving environment has to exist. An unforgiving environment prevents or reduces the likelihood of error recovery described in Chapter 5. It also involves ametastable (vulnerable) system state, which in combination with the unrecovered error leads to the undesirable consequence. In some cases the error may give rise to a vulnerable state which does not produce a significant consequence, Embrey (1989) has termed this a latent failure. It is interesting to note that many key actions with respect to safety have been consequence-driven and have ignored the...