A Comparison of Modifications to MELCOR versions 1.8.2 and 1.8.6 for ITER Safety Analysis

During the Engineering Design Activity of the International Thermonuclear Experimental Reactor (ITER), the MELCOR 1.8.2 code was selected as one of several codes to be used to perform ITER safety analyses [1]. MELCOR was chosen because it has the capability of predicting coolant pressure, temperature, mass flow rate, and radionuclide and aerosol transport in nuclear facilities and reactor cooling systems. MELCOR can also predict structural temperatures (e.g. first wall, blanket, divertor, and vacuum vessel) resulting from energy produced by radioactive decay heat and/or chemical reactions (oxidation). The Idaho National Laboratory (INL) Fusion Safety Program (FSP) made fusion specific modifications to the MELCOR 1.8.2 code [2-6], including models for water freezing, air condensation, beryllium, carbon, and tungsten oxidation in steam and air environments, flow boiling in coolant loops, and radiation in enclosures, that allowed MELCOR to assess the thermal hydraulic response of ITER cooling systems and the transport of radionuclides as aerosols during accident conditions. Recently, the ITER International Organization (IO) used a “pedigreed” version of MELCOR 1.8.2 [7] to perform accident analyses for ITER’s “Rapport Preliminaire de Surete” (Report Preliminary on Safety - RPrS). The MELCOR thermal-hydraulics code [8] is currently under development at the Sandia National Laboratory (SNL) formore » the US Nuclear Regulatory Commission (NRC). MELCOR is used to model the progression of severe accidents in light water fission reactors. Because MELCOR has undergone many improvements between version 1.8.2 and 1.8.6, the INL FSP decided to introduce fusion modifications into MELCOR 1.8.6, and thereby produce a version of MELCOR 1.8.6 with similar capabilities to the pedigreed version of MELCOR 1.8.2 used for the ITER RPrS. We have applied this version of MELCOR 1.8.6 to the same set of problems used in the MELCOR 1.8.2 pedigree analysis [7]. Section 2 describes a non-regression analysis that involves comparing the results from the modified version of MELCOR 1.8.6 against those predicted by the original, unmodified version of MELCOR 1.8.6. The purpose of this non-regression analysis is to demonstrate that the modifications made to the MELCOR 1.8.6 code do not drastically alter the intended functions of the MELCOR base code, and if they do to explain why the departure occurs and if the change is needed. The analysis was performed using the demonstration problem that came with the MELCOR 1.8.6 source distribution. Section 3 makes a comparison between the pedigreed version of MELCOR 1.8.2 and the new modified version of MELCOR 1.8.6 on a set of accident problems used by the ITER Joint Central Team (JCT) in the Generic Site Safety Report (GSSR) [9]. Finally, in section 4, the two versions of the code are compared on a series of developmental test problems described in the change documents [2-4, 6].« less