The RELAP5-3D thermal-hydraulic code was assessed against Japanese Safety Experiment Loop (SEL) and Heat Transfer Loop (HTL) tests. These tests were chosen because the phenomena present are applicable to analyses of Russian RBMK reactor designs. The assessment cases included parallel channel flow fluctuation tests at reduced and normal water levels, a channel inlet pipe rupture test, and a high power, density wave oscillation test. The results showed that RELAP5-3D has the capability to adequately represent these RBMK-related phenomena.
iv v
SUMMARYThis report documents the assessment of the RELAP5-3D code against tests conducted in the Japanese Safety Experiment Loop (SEL) and Heat Transfer Loop (HTL). These tests provide data on selected phenomena that are relevant to RBMK safety analyses. The focus of the present work was in two areas. First, validation of the RELAP5-3D code was conducted using test data for which previous RELAP5/MOD3.2 assessment results are available. These assessment calculations were repeated with RELAP5-3D to determine whether the results were comparable to those previously obtained. Second, code validation calculations were conducted using RELAP5-3D for the SEL High Power Flow Instability Test, a case for which RELAP5/MOD3.2 failed to reproduce the response of the test facility.The assessment cases covered five phenomena identified as applicable to RBMK designs: The major conclusion was that RELAP5-3D has the capabilities to properly represent RBMK phenomena. The results showed that RELAP5-3D was capable of representing CCFL, break flow rates, manometer oscillations, flow pattern-induced oscillations, and other phenomena applicable to RBMK transient analyses. In most cases, the RELAP5-3D predictions were nearly identical to previous assessment results obtained using RELAP5/MOD3.2. The results also demonstrated that RELAP5-3D has the capability to predict the high power instability, or density wave oscillations, although it was necessary to adjust the channel pressure loss distribution to obtain this result. This adjustment was justified based on inconsistency and incompleteness of available data.The CHF model needs improvement; however, the successful prediction of the density wave oscillation response was not strongly dependent upon the precise prediction of dryout in the channel. Results of nodalization studies, which investigated multiple parallel paths in the high power channel, demonstrated that the fluid was well mixed and therefore a single-pipe (or single path) hydraulic model was adequate to represent the channel flow conditions. vi vii
CONTENTS
Abstract