Molten salt reactors have been selected as one of the promising candidate Generation IV reactor technologies, due to the advantages of inherent safety and high economic efficiency. The small modular molten salt reactor (SM-MSR), which utilizes low-enriched uranium and thorium fuels, is regarded as a wise development path to speed deployment time. Uncertainty and sensibility analysis of accidents possess a great guidance in nuclear reactor design and safety analysis that can be performed to obtain the safety boundary and through sensitivity analysis, thereafter to obtain the correlation of the accident consequence and input parameters. Reactivity insertion transient accident represents a type of hypothetical accidents of SM-MSR, and the study of reactivity insertion transient accident can offer useful information to improve physics thermohydraulic and structure designs.

This study aims to investigate the uncertainty and sensibility of MSR reactivity insertion accident and provide supports for the design and safety analysis of the small modular molten salt reactor.

RELAP5-TMSR code was employed to establish a transient behavior analysis model for SM-MSR, and the model consisted of four coupled parts, including the primary circuit, 2nd circuit, air cooling system modules and passive residual heat removal system. Then, propagation of input uncertainty approach on the basis of Monte Carlo methods was employed to analyze the uncertainty of reactivity insertion transient accident consequence. Uncertain parameters for the reactivity insertion transient accident were selected by the phenomena identification and ranking table (PIRT). Subsequencely, a list of input parameters along with their associated density functions was adopted by using a probabilistic methodology to establish the code run times and sets of uncertain input parameters that was propagated through the RELAP5-TMSR code, and then obtain the upper and lower uncertainty bands of the reactivity insertion transient consequence. Finally, the sensibility of input parameters was analyzed by performing Multiple Linear Regression (MLR) method, and the F-test was used to assess whether the MLR models comply with statistical laws. If the linear model was strong collinear, a significance test of the semi-partial correlation coefficient (SPC) was used for the ranking of input uncertainty parameters, otherwise, the standardized regression coefficient (SRC) would be used for the significance test.

The uncertainty analysis results show that the maximum fuel salt temperature of SM-MSR is 727.4 ℃ which is lower than the acceptance criteria (800 ℃). Through statistical analysis, the maximum value of reactor outlet fuel salt temperature is normally distributed.

The molten salt reactor has good safety characteristics, and the 5 important parameters are density of fuel salt, local resistance coefficient of reactor core, reactor power, local resistance coefficient of primary circuit and reactor shutdown margin.