Sensitivity analysis of QRA to simplifying criteria used on the probability of initiating events
The purpose of this thesis is to assess the sensitivity of the results of quantitative risk analysis (QRA). The first step was to reproduce a QRA for the MILES S.A plant, a simulated reference company. In carrying out the analysis, all the required steps were followed as described in Instruction 14...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Fecha de publicación: | 2022 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/381123 |
| Acceso en línea: | https://hdl.handle.net/2117/381123 |
| Access Level: | acceso abierto |
| Palabra clave: | Industrial accidents Industrial safety QRA SAFETY SENSITIVITY ANALYSIS LUP Seguretat en el treball Treball -- Accidents Àrees temàtiques de la UPC::Enginyeria química |
| Sumario: | The purpose of this thesis is to assess the sensitivity of the results of quantitative risk analysis (QRA). The first step was to reproduce a QRA for the MILES S.A plant, a simulated reference company. In carrying out the analysis, all the required steps were followed as described in Instruction 14 in force in Catalonia, regulating the conduct of the QRA. After having identified and described the primary events, a frequency value equal to that given in the BEVI manual was attributed to each type of scenario identified and then it was adapted to the process conditions. The evolution of primary events was assessed using the method of quantified event trees, necessary to link the primary event with the possible accidentalscenarios that can be generated from them. In addition, with this method the final accidental scenario frequency of occurrence was also evaluated as the frequency of the initial accidental scenario multiplied by the probability product on each branch of the event tree. The next step was to describe the accident scenarios identified in terms of the physical variables that characterise these scenarios (overpressure, thermal radiation, concentration). The physical variables obtained were used as input to the Probit equations required to calculate the lethal concentrations associated with both the substance and the final accident scenario under consideration. Once all the necessary data had been collected, the risk around the plant was calculated, represented by isorisk curves (curves joining all points with the same risk value). The sensitivity of the results was assessed by changing one of the input parameters of the QRA, the frequency of the primary event, which is a determining factor in calculating the risk associated to a given accidental scenario. The change that influenced the variation in the final results was assessed by examining how the isorisk curves changed in diameter, with particular reference to the 10-6 value curve representing the limit between acceptable and non-acceptable risk. |
|---|