1 Copyright © 2009 by ASME Proceedings of the17th International Conference on Nuclear Engineering ICONE-17 July 12-16, 2009, Brussels, Belgium ICONE17-75099 ANALYSIS OF THE QUENCH-14 BUNDLE TEST WITH M5 ® CLADDING Jonathan C. Birchley, Bernd Jaeckel, Timothy J. Haste Paul Scherrer Institut, CH-5232, Villigen, Switzerland Martin Steinbrueck, Juri Stuckert Forschungszentrum Karlsruhe GmbH P.O. Box 3640, 76021 Karlsruhe, Germany Correspondence to J. Birchley: TEL +41 56 310 2724, FAX +41 56 310 2199; E-mail jonathan.birchley@psi.ch ABSTRACT The QUENCH experimental programme at Forschungszentrum Karlsruhe (FZK) investigates phenomena associated with reflood of a degrading core under postulated severe accident conditions, but where the geometry is still mainly rod-like and degradation is still at an early phase. The QUENCH test bundle is electrically heated and consists of 21 fuel rod simulators with a total length of approximately 2.5 m. The cladding and grid spacers are identical to those used in Pressurized Water Reactors (PWR) whereas the fuel is represented by ZrO 2 pellets. Experiment QUENCH-14 was successfully performed at FZK in July 2008 and is the first in this programme where Zr- Nb alloy M5 ® is used as the fuel rod simulator cladding. QUENCH-14 was otherwise essentially the same as experiment QUENCH-06, which was the subject of the CSNI ISP-45 exercise. It is also the first of three experiments in the QUENCH-ACM series, recently launched to examine the effect of advanced cladding materials on oxidation and quenching under otherwise similar conditions. Pre- and post-test analyses were performed at PSI using a local version of SCDAP/RELAP5 and MELCOR 1.8.6, using input models which had already been benchmarked against QUENCH-06 data. Preliminary pre-test calculations with both codes and alternative correlations for the oxidation kinetics indicated that the planned test protocol would achieve the desired objective of exhibiting whatever effects might arise from the change in cladding-material in the course of a transient similar to QUENCH-06. Several correlations were implemented in the models, namely Cathcart-Pawel, Urbanic-Heidrick, Leistikow-Schanz and Prater-Courtright for Zircaloy-4 (Zry-4), and additionally a new candidate correlation for M5 ® based on recent separate-effects tests performed at FZK on M5 ® cladding samples. Analyses of the QUENCH-14 data demonstrate strengths and limitations of the various models. Some tentative recommendations are made concerning choice of correlation and effect of cladding material. 1. INTRODUCTION The QUENCH programme is being performed at the Forschungszentrum Karlsruhe/Germany (FZK) to investigate the effectiveness of water injection as a means of reflooding and quenching a core, following a beyond-design-basis accident with temperatures above 2000 K and possibly some early phase degradation. Among the topics of concern is the hydrogen generation due to contact between the overheated cladding and the flowing steam. Fourteen experiments have been carried out under a range of flooding/cooling conditions and bundle configurations, thus creating a strong database for model development and code improvement in the field of severe accident simulation [1]. One of the ultimate goals of QUENCH is to identify the limits (temperature, injection rate etc.) for which successful reflood and quench can be achieved. Almost all the experiments to date were performed with Zry-4 as the cladding material. Other cladding materials based on zirconium-niobium alloys are being increasingly adopted for PWR fuel, by virtue of their improved resistance to corrosion during operation, for example M5 ® by AREVA and Zirlo ® by Westinghouse. In contrast to the extensive database available for Zry-4 oxidation, data for the more recently adopted cladding materials are comparatively scarce. For that reason FZK has recently launched the QUENCH-ACM series [2] in order to investigate the impact of alternative claddings on high- temperature reflood and quench. In parallel FZK is also