PSA simulations

Published: 9 October 2020| Version 1 | DOI: 10.17632/4nmvpnst8h.1
Contributor:
JOAO MARCIO LIMA DO NASCIMENTO

Description

This data base is the result of testing an unavailability model for Emergency Diesel Generators (EDG) of a Nuclear Power Plant Onsite Power System (Described in Figure 1). The unavailability model (presented on Mathematical_development.pdf) innovates for considering post maintenance tests degradation. Traditional models consider only successful surveillance tests startups. The hypothesis is that post maintenance tests degradation may lead to EDG unavailability values higher than the defined in the plant design basis. The model is based on two parameters Nadd and Npmt which, through Eq(12) and Eq(25) (of Mathematical_development.pdf), provide mean unavailability values corresponding to those parameters. Three sets of simulations are presented: Table 2 – EDG mean unavailability considering Nadd 0.pdf for Npmt values varying (1, 2, 4, 6, 8 & 10) during 10 Wonking Cycles Table 3 – EDG mean unavailability considering Nadd 1.pdf for Npmt values varying (1, 2, 4, 6, 8 & 10) during 10 Wonking Cycles Table 4 – EDG mean unavailability considering Nadd 2.pdf for Npmt values varying (1, 2, 4, 6, 8 & 10) during 10 Wonking Cycles Figure 2 shows the mean unavailability behavior for different values of Npmt along a ten working cycles period. Using mean unavailability from Tables 2, 3 & 4, PSA simulations to determine CDF for different values of Nadd (0, 1 & 2) and Npmt (1, 2, 4, 6, 8 & 10) for ten working cycles. Results are presented on Table 5 (Nadd 0) Table 6 (Nadd 1) and Table 7 (Nadd 2). Figure 3 shows the CDF behavior for different values of Npmt along a ten working cycles period. Folders Nadd 0, Nadd 1 and Nadd 2 contain files with CDF and the minimal cut sets of each simulation. The file PSA_Results. xlsm presents the numerical analysis of the minimal cutsets for all simulation data. Through this analysis is possible to verify the Fussel- Vessely importances for the real Onsite Power configuration and for an hypotetical configuration with only the Emergency 1 subsystem avaialble.

Files

Institutions

Universidade Federal do Rio de Janeiro

Categories

Computer Simulation, Probabilistic Analysis

License