Natural hazards can lead to Technological accidental scenarios (so called NaTech) with severe consequences if preventive and mitigative safety barriers of safety-critical systems, such as Nuclear Power Plants (NPPs) and chemical facilities, are not properly designed accounting for their performance degradation due to the natural event occurrence, which needs to be integrated within a risk assessment and management framework. Such performance degradation is, however, typically assessed by expert judgement, without considering the increasing frequency of occurrence and severity of such events, also in view of climate change. Also, to now, none has been proposing a methodology that identifies the set of improvements needed on the present design of the safety barriers of a given system to optimally manage the risk arising from such NaTech scenarios. To overcome these limitations, in this thesis, two novel advanced methods are proposed to: 1) Identify the safety barriers whose performance degradation is most critical and on which an investment to gather more information about their performance is rationally justified. 2) Identify the improvements on the safety barriers design that bring the largest benefit in terms of consequences mitigation with the lowest cost and, on the basis of these, provide the retrofitting actions to improve the safety barriers design. With respect to the first issue, we propose a novel sensitivity analysis framework based on the calculation of a set of sensitivity measures, namely the Beta, the Conditional Value at Risk (CVaR) and the Value of Information (VoI), that are used to prioritize the safety barriers with respect to the need of: • Accounting for performance degradation during an accidental scenario. • Planning additional investments for further characterization of the safety barriers performance. With respect to the second issue, we propose a robust Multi-Objective Optimization (MOO) framework that searches for the optimal performance parameters of the barriers while relying on a phenomenological dynamic model that realistically mimics the response of the system when impacted by the NaTech scenario. The frameworks have been tested on a case study of literature that consists in a chemical facility, in which there are three tanks of flammable substances, equipped with five safety barriers and exposed to the risk of NaTech scenarios triggered by floods and earthquakes. The results obtained provide the analyst with a more complete risk picture of the system in case of NaTech scenarios, while also allowing informed decisions about the allocation of investments to better characterize the safety barriers degradation due to natural events and to improve their performance.
Le calamità naturali possono portare a incidenti industriali (chiamati scenari NaTech) con conseguenze gravi se le barriere di sicurezza mitigative e preventive dei sistemi critici, come le centrali nucleari e gli impianti chimici, non sono progettate opportunamente considerando il loro degrado prestazionale a causa dell’evento naturale, che deve essere integrato in un framework di valutazione e gestione del rischio. Questo degrado prestazionale è, tuttavia, tipicamente valutato basandosi sul giudizio di esperti, senza considerare l’aumento della frequenza di accadimento e della gravità di questi eventi, anche in vista dei cambiamenti climatici. Inoltre, finora nessuno ha proposto una metodologia che identifichi il set di miglioramenti nel design attuale delle barriere di sicurezza di un dato sistema necessari per gestire il rischio derivante da questi scenari NaTech in maniera ottimale. Per superare queste limitazioni, in questa tesi vengono proposti due nuovi metodi avanzati per: 1) Identificare le barriere di sicurezza il cui degrado prestazionale è critico e sulle quali un investimento per raccogliere ulteriori informazioni riguardo le loro prestazioni è razionalmente giustificato. 2) Identificare i miglioramenti nel design delle barriere di sicurezza che portano i benefici maggiori in termini di mitigazione al prezzo più basso e, sulle basi di questi, fornire le azioni di retrofitting necessarie a migliorare il design delle barriere di sicurezza. Riguardo il primo problema, proponiamo un nuovo framework di analisi di sensitività basato sul calcolo di un set di misure di sensitività, in particolare la misura Beta, il Conditional Value at Risk (CVaR) e il Value of Information (VoI), che sono usate per classificare le barriere di sicurezza rispetto alla necessità di: • Considerare il loro degrado prestazionale durante uno scenario accidentale. • Pianificare ulteriori investimenti per caratterizzare meglio le loro prestazioni. Riguardo al secondo problema, proponiamo un framework di Ottimizzazione Multi-Obiettivo (MOO) robusta per identificare i parametri prestazionali ottimali delle barriere di sicurezza, sfruttando un modello fenomenologico dinamico in grado di riprodurre realisticamente la risposta del sistema all’impatto dello scenario NaTech. Questi framework sono stati testati su un caso studio presente in letteratura, che riguarda un impianto chimico, nel quale sono presenti tre serbatoi contenenti sostanze infiammabili, protetti con cinque barriere di sicurezza ed esposti al rischio di scenari NaTech causati da terremoti ed alluvioni. I risultati ottenuti forniscono all’analista un’immagine più completa del rischio del sistema in caso di scenario NaTech, permettendo anche decisioni informate riguardo l’allocazione di investimenti per la caratterizzazione del degrado delle barriere di sicurezza a causa di eventi naturali e per il miglioramento del loro design.
Advanced methods for the performance assessment and optimization of safety barriers to NaTech scenarios
MARCHETTI, STEFANO
2021/2022
Abstract
Natural hazards can lead to Technological accidental scenarios (so called NaTech) with severe consequences if preventive and mitigative safety barriers of safety-critical systems, such as Nuclear Power Plants (NPPs) and chemical facilities, are not properly designed accounting for their performance degradation due to the natural event occurrence, which needs to be integrated within a risk assessment and management framework. Such performance degradation is, however, typically assessed by expert judgement, without considering the increasing frequency of occurrence and severity of such events, also in view of climate change. Also, to now, none has been proposing a methodology that identifies the set of improvements needed on the present design of the safety barriers of a given system to optimally manage the risk arising from such NaTech scenarios. To overcome these limitations, in this thesis, two novel advanced methods are proposed to: 1) Identify the safety barriers whose performance degradation is most critical and on which an investment to gather more information about their performance is rationally justified. 2) Identify the improvements on the safety barriers design that bring the largest benefit in terms of consequences mitigation with the lowest cost and, on the basis of these, provide the retrofitting actions to improve the safety barriers design. With respect to the first issue, we propose a novel sensitivity analysis framework based on the calculation of a set of sensitivity measures, namely the Beta, the Conditional Value at Risk (CVaR) and the Value of Information (VoI), that are used to prioritize the safety barriers with respect to the need of: • Accounting for performance degradation during an accidental scenario. • Planning additional investments for further characterization of the safety barriers performance. With respect to the second issue, we propose a robust Multi-Objective Optimization (MOO) framework that searches for the optimal performance parameters of the barriers while relying on a phenomenological dynamic model that realistically mimics the response of the system when impacted by the NaTech scenario. The frameworks have been tested on a case study of literature that consists in a chemical facility, in which there are three tanks of flammable substances, equipped with five safety barriers and exposed to the risk of NaTech scenarios triggered by floods and earthquakes. The results obtained provide the analyst with a more complete risk picture of the system in case of NaTech scenarios, while also allowing informed decisions about the allocation of investments to better characterize the safety barriers degradation due to natural events and to improve their performance.File | Dimensione | Formato | |
---|---|---|---|
Tesi Marchetti Stefano.pdf
non accessibile
Descrizione: Tesi
Dimensione
6.1 MB
Formato
Adobe PDF
|
6.1 MB | Adobe PDF | Visualizza/Apri |
Executive Summary Marchetti.pdf
non accessibile
Descrizione: Executive Summary ENG
Dimensione
1.04 MB
Formato
Adobe PDF
|
1.04 MB | Adobe PDF | Visualizza/Apri |
Executive Summary Marchetti Italiano.pdf
non accessibile
Descrizione: Executive Summary ITA
Dimensione
955.11 kB
Formato
Adobe PDF
|
955.11 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/189675