Quantitative evaluation of thermal-hydraulic code uncertainties is a necessary step in the code assessment process, especially if best-estimate codes are utilized for licensing purposes. With the goal of quantifying code accuracy, researchers in the past developed a methodology based on the fast Fourier transform (FFT) that consisted of qualitative and quantitative code assessment. Here, the FFT-based method is applied to International Atomic Energy Agency (IAEA)-Standard Problem Exercise (SPE)-4 test results with pre- and posttest code calculations of the IAEA-SPE-4 experiment. Four system codes (ATHLET, CATHARE, MELCOR, and RELAP5) are used for calculations of the experiment, performed at the PMK-2 facility, which simulated a cold-leg break in a WER-440 plant. The results show that the posttest calculations had better accuracy than did the pretest calculations. None of the best three pre- and posttest calculations were able to predict core dryout, which was the most important phenomenon observed during the test. The results obtained can give an objective indication of the capability of the aforementioned codes in predicting relevant variables characterizing the transient (too few experimental parameters may limit full application of the FFT-based methodology).
