The QUENCH experiments are to investigate the hydrogen source term resulting from the water or steam injection into an uncovered core of a Light-Water Reactor (LWR). The QUENCH-07 experiment was to investigate the effect of B 4 C absorber rod failure on fuel bundle degradation as well as the B 4 C oxidation. Complementary to the planned PHEBUS FPT3 experiment, QUENCH-07 was expected to particularly provide information on the formation of gaseous reaction products during the absorber rod degradation and B 4 C oxidation, in particular of H 2 , CO, CO 2 and CH 4 , and on the impact of control rod degradation on surrounding fuel rods. The test bundle is made up of 20 fuel rod simulators with a length of approximately 2.5 m. The fuel rod simulators are heated over a length of 1024 mm, the central rod is unheated and in this experiment was made of an absorber rod with B 4 C pellets and stainless steel cladding, and of a Zircaloy-4 guide tube. The steel to B 4 C mass ratio of 3.5 is identical to that in the future PHEBUS FPT3 experiment. Heating is carried out electrically using 6-mm-diameter tungsten heating elements, which are installed in the center of the rods and which are surrounded by annular ZrO 2 pellets simulating the UO 2 fuel pellets. The Zircaloy-4 rod cladding is identical to that used in LWRs: 10.75 (OD) x 0.725 mm. The test section is instrumented with thermocouples (TC) that are attached to the cladding, the shroud, and the cooling jackets at elevations between -250 mm and 1350 mm. Besides, three TCs are embedded in a groove of the absorber rod cladding inside the central rod, and centerline TCs were mounted inside three of the four corner rods. The hydrogen is analyzed by three different instruments: two mass spectrometers and a hydrogen analyzer "Caldos 7 G" with the principle of measurement based on heat conduction of the sampling gases. QUENCH-07 performed at the Forschungszentrum Karlsruhe on 25 July, 2001 was the first experiment with an absorber rod in the bundle. Until the onset of cooldown 3 g/s of superheated steam and 3 g/s of argon as carrier gas enter the test bundle at the bottom and exit at the top together with the gases that are produced in the reactions of zirconium, boron carbide, and stainless steel with steam. Cooling was initiated by injecting 15 g/s of saturated steam at the bottom and switching the argon flow to the bundle head. The control rod failed at ∼1585 K (2036 s) shortly before the QUENCH-07 test bundle was conditioned (oxidized) at -1723 K for 15 minutes. During the conditioning phase CO, CO 2 , H 2 , metaboric and orthoboric acids were detected by the mass spectrometer together with very small quantities of methane. In the subsequent transient with heatup rates of 0.35-0.45 K/s temperature escalations started at the top of the heated zone and above. It was during this phase that most of the cladding thermocouples failed. The rod cladding as well as the shroud failed at the end of the transient phase at temperatures of around ∼2000 K. Up to the 650 mm level cooling was immediate. Above 650 mm delayed cooling occurred. The maximum measured rod temperature was ∼2320 K at the 950 mm level. This temperature as well as the maximum temperatures at the elevations above were measured in the cooldown phase. At the elevations above the heated zone the steam injection of 15 g/s triggered a temperature excursion with subsequent heatup rates of more than 40 K/s. At the lower elevations, however, the cooling effect dominated due to the increase in the coolant flow. Associated with the high temperatures at the upper elevations during the period shortly after beginning of the cooldown phase an increased release of all gaseous reaction products was observed. The evaluation of the mass spectrometer data as well as of the post-test bundle status results in ∼180 g of hydrogen release in total, most of which was during the cooldown, i.e. 120 g. In addition, there were large increases in the generation of CO and C