Abstract
The aim of this report has been to investigate weather the decision analysis methodology fulfils the demands from the industrial decision-makers concerning the most cost-effective fire protection alternative. The report also contains a sensitivity analysis to demonstrate which parameters that has great importance to the
validity of the model.
A fictive example has been defined; Termoplast AB. A decision analysis has then been carried out with estimated inputs. These inputs have some uncertainties part of the estimation that has been done and part of reality factors. To show how these uncertainties effect the model a deterministic and probabilistic sensitivity analysis has been carried out. Observe that the benefit in this report is the same as a minimisation of the risk cost
consisting of insurance premiums, annual cost of fire protection and expected values of deductibles.
The decision analysis shows that the most cost-effective alternative is sprinkler with a deductible level of 100 000 Skr. What is of more importance is that the sensitivity analysis shows that the estimated losses (both direct and indirect) and statistical basic data have great importance to the validity of the model. Further more the insurance premium and the annual cost of fire protection have no or very little importance.
The second part of the report consists of a summary and an identification of problems concerning the underlying models to predict both losses and occurrence of fires. Finally a suggestion of some future research areas has been given.
validity of the model.
A fictive example has been defined; Termoplast AB. A decision analysis has then been carried out with estimated inputs. These inputs have some uncertainties part of the estimation that has been done and part of reality factors. To show how these uncertainties effect the model a deterministic and probabilistic sensitivity analysis has been carried out. Observe that the benefit in this report is the same as a minimisation of the risk cost
consisting of insurance premiums, annual cost of fire protection and expected values of deductibles.
The decision analysis shows that the most cost-effective alternative is sprinkler with a deductible level of 100 000 Skr. What is of more importance is that the sensitivity analysis shows that the estimated losses (both direct and indirect) and statistical basic data have great importance to the validity of the model. Further more the insurance premium and the annual cost of fire protection have no or very little importance.
The second part of the report consists of a summary and an identification of problems concerning the underlying models to predict both losses and occurrence of fires. Finally a suggestion of some future research areas has been given.
| Original language | Swedish |
|---|---|
| Publisher | Department of Fire Safety Engineering and Systems Safety, Lund University |
| Number of pages | 110 |
| Volume | 3098 |
| Publication status | Published - 1997 |
| Externally published | Yes |
Publication series
| Name | LUTVDG/TVBB--3098--SE |
|---|---|
| Volume | 3098 |
| ISSN (Print) | 1102-8246 |
Subject classification (UKÄ)
- Other Civil Engineering
- Building Technologies
Free keywords
- Cost-benefit
- decision analysis
- risk assessment
- thermal damage
- non-thermal damage
- indirect losses
- computer models
- industrial fire protection
- decision-making