Abstract
Probabilistic Safety Analysis (PSA) is a useful and widely utilized tool in the priorization of safety improvements. However, carrying out a PSA is not quite straightforward because of a number of uncertainties related to analysed systems and used methods. One of the most problematic areas is accident sequences, including human contribution. The effect of human behaviour on a technical system is generally difficult to identify and model. Therefore the quantitative assessment of risks remains deficient and imprecise.
This paper discusses the modelling of the human in PSAs on a general level. The topic is divided into two subcategories. These categories are the accident sequence models and the human internal behaviour models. Pros and cons of several widely implemented PSA models are discussed in respect of the use in these purposes. Furthermore, activities in the Nordic countries round the subject and needs for the future development of human modelling are described briefly.
This paper discusses the modelling of the human in PSAs on a general level. The topic is divided into two subcategories. These categories are the accident sequence models and the human internal behaviour models. Pros and cons of several widely implemented PSA models are discussed in respect of the use in these purposes. Furthermore, activities in the Nordic countries round the subject and needs for the future development of human modelling are described briefly.
Original language | English |
---|---|
Pages (from-to) | 277-294 |
Journal | Reliability Engineering and System Safety |
Volume | 22 |
Issue number | 1-4 |
DOIs | |
Publication status | Published - 1988 |
MoE publication type | D1 Article in a trade journal |