Normal Accident Theory and Swiss Cheese Model

Only available on StudyMode
  • Topic: Three Mile Island accident, Nuclear safety, Chernobyl disaster
  • Pages : 8 (2169 words )
  • Download(s) : 676
  • Published : April 12, 2012
Open Document
Text Preview
1. Executive Summary
Normal accident theory and Swiss cheese model are influential models in studying system accident causation. This paper is going to help us to gain understanding of both models and to critically compare them. The first part of the study is an introduction of the both models. In the second part, Three Mile Island nuclear accident will be taken as an example to see how the models analyse causations of an accident. In this part, the fact of the accident will be presented first, then normal accident theory and Swiss cheese model will be invited to identify the causes of the accident respectively. The evaluation and comparison of the two models will be discussed in the final part. Both models conclude that accidents are unavoidable, while Swiss cheese model provides general methods to reduce accidents. Normal accident theory views technology itself as the main factor that contributes to accidents, however, Swiss cheese model has the opinion that accidents resulted from interaction of multiple factors from different levels in organization.

2. Introduction and Aims
System accident is regarded as unpredictable and inevitable in high-risk system because of the system’s complexity. Both normal accident theory and Swiss cheese model are accident theories to study accident causation. The study aims to understand how both accident models work and compare the two models to see what contributions and limitations of each of them. Three Mile Island accident is one of the most significant nuclear accidents in history. It is a typical accident in complex and high-hazard system. There have been plenty of studies to analyse the accident. Therefore, Three Mile Island accident will be a helpful case for us to study the models.

3. Accident Models Study
3.1 Introduction of Accident Models
Normal Accident Theory
Charles Perrow termed the system accidents as “normal accident”. He described systems by two important dimensions, interaction and coupling. Interaction can be divided to linear interaction and complex interaction. For systems with complex interaction, when unexpected and unplanned problem occur, they are not visible because there are many interrelated and interdependent elements in the system, it is hard to predict all of the possible failures. Coupling means how the interconnected components affect each other. For tightly coupled systems, if something happened to one component, it will directly influence other components around and connected to it, the chain reactions are easier to happen in tightly coupled system. Crossing the two dimensions, we can see four categories of systems. They are Linear Interaction, Loose Coupling; Complex Interaction, Loose Coupling; Linear Interaction, Tight Coupling; Complex Interaction, Tight Coupling. Perrow augured that those systems which are complex interaction, tight coupling are accident-prone. When a technology is with sufficiently complex interaction and tight coupling, accidents are inevitable and therefore in a normal sense.

Swiss Cheese Model
Swiss cheese model is an organizational model developed by James Reason to explain accident causation in complex technological systems. In Reason’s theory, high technology systems have many defensive layers to prevent accidents: some are engineered others rely on people. However, in real world, no defense is perfect, just like Swiss cheese slices. The holes in Swiss cheese are the imperfections and weakness of the defenses, although they are changing their location, opening and shutting all the time instead of staying statically. As long as the holes in all slices do not line up, the hazards which can go through one or several slices will be blocked by a slice after. In some rare circumstances, holes in all successive slices will line up, and allow hazards pass all defenses, the accidents will happen. There are two reasons to explain the holes arising: active failures and latent conditions. Active failures are unsafe...
tracking img