Uncertainty Analysis Of The Effect Of Plant State Parameter On LBLOCA In AP1000

As the newly-designed third-generation nuclear power plant,AP1000has many engineering design characteristics contributing toreactor safety, for example, the introduction of the concept of the passivesafety system. As a result, security system configuration is simplified, andthe possibility of man-made error is greatly reduced. Besides, in1998, the10CFR50.46rule allowed the use of realistic physical models to analyzeloss-of-coolant accident (LOCA), of which the execution of variousrealistic LOCA transient simulations is applied, where the effect of bothmodel and input uncertainties are ranged and propagated throughout thetransient. Parameter uncertainty analysis is developed by using thereasonable method to establish the response relations between inputparameter uncertainties and output uncertainties. The application ofparameter uncertainty analysis makes the simulation of plant state moreaccuracy and improves the plant economy with reasonable securityassurance.To investigate the design features of AP1000, RELAP5was appliedto model the NSSS of AP1000and it was utilized to systematicallyperform the analysis of LBLOCA. On this basis, the conservative nuclearpower plant geometric model is build by the way of sensitivity analysis ofmodel node number, and then, after the sensitivity analysis of power plantstate parameters, the calculation result is that the envelope LBLOCA accident condition and PCT envelope value.The conservative nuclear power plant geometric model is used toanalyze the effect of plant state parameter uncertainty on LBLOCA inAP1000NPP. The analysis results in this paper indicate that the randomsampling statistical analysis method, sensitivity analysis numericalmethod and traditional error propagation analysis method could providequite large PCT safety margin, which is much helpful for choosing suiTab.uncertainty analysis method to improve the plant economy. Additionally,the random sampling statistical analysis method applying mathematicalstatistics theory makes the largest safety margin due to the reducing of theconservation. Comparing with the traditional conservative boundingparameter analysis method, the random sampling method could providethe PCT margin of100K, while the other two methods can only provide50-60K.