Lessons learned from the quench-11 training exercise

16 organizations in 12 countries are participating in a RELAP/SCDAPSIM training exercise based on the Quench 11 experiment performed at Karlsruhe (Germany) in 2005. This exercise is being conducted in parallel to an International Standard Problem (ISP). Both the ISP and the RELAP/SCDAPSIM training exercise included a 'semi-blind' portion that was completed in the fall of 2006 and an 'open' portion that is to be completed in the summer of 2007. The RELAP/SCDAPSIM training exercise is coordinated by Innovative Systems Software with support by the International SCDAP Development and Training Program (SDTP). The Quench-11 experiment is based on an electrically heated fuel rod bundle representative of a PWR design. The bundle was subjected to a boil down transient, heat-up, and quenching with peak temperatures exceeding the melting point of the Zircaloy cladding. This experiment was chosen by the European Union as an International Benchmark exercise to compare the effectiveness of quenching models in the severe accident computer codes used today for accident analysis. This paper briefly describes (a) RELAP/SCDAPSIM/MOD3.4, (b) the Quench facility and experiments used in the training exercise, and (c) the training guidelines provided to the participants followed by a more detailed description of the lessons learned from the initial 'semi-blind' portion. The representative results demonstrate that good analysts can still have a difficult time predicting the thermal hydraulic response of a relative simple transient in a complex system