TRAC-PF1/MOD-1 analysis of Loss-Of-Flow Test L9-4

Los Alamos National Laboratory is developing the Transient Reactor Analysis Code (TRAC) to provide advanced best-estimate predictions of postulated accidents in pressurized water reactors (PWRs) and for many thermal-hydraulic experimental facilities. As part of our independent assessment of code version TRAC-PF1/MOD1, we analyzed Loss-of-Fluid Test (LOFT) L9-4 and compared the test data to the calculated results. This was an anticipated-transient-without-scram test in which the pumps were tripped, the steam generator main feedwater discontinued, and the main steam-outlet valve closed. This data comparison is the first extensive test of TRAC's reactor-kinetics models. The comparisons show that TRAC can calculate the power generation within a nuclear reactor if the program is supplied with adequate reactor-kinetics input specifications. The data comparisons also indicate that TRAC calculated the thermal-hydraulic parameters within LOFT well with only minor discrepancies. A number of models within TRAC-PF1/MOD1 were verified for the first time. They include the reactor-kinetics models, the trip-activated time-step controls, and the LOFT pump-coastdown calculations. In general, the final input description is adequate to analyze the experiment. The calculations indicate the importance and difficulty of obtaining accurate and applicable reactor-kinetics input data. They also indicate the need to include the effects of xenon-poisoning buildup in the analysis