Analysis of a feed and bleed procedure sensitivity study, performed with the SIPACT simulator, on a French 900 MWe NPP

Abstract

An extensive sensitivity study related to a feed and bleed procedure was carried out with the SIPACT simulator. This work was performed independently from the real procedures (events- or symptoms-oriented), which are developed in the plant environment. We only look here at the combined effects of bleeding and its time of initiation, with the determining effect of an initial mass inventory and a residual power. The basic scenario we analysed is related to a total loss of feed water without any bleed action. In the calculations, full bleed of the primary system was initiated at certain time points by opening the three relief valves (for example, pressurizer full, start of high pressure injection system (HPIS), minimum of mass in the reactor vessel). Calculations were done for about 30 different time points for actuation of bleed to obtain a complete view of the process. A comparison between the different calculation results was first based on the minimum coolant mass, and then improved using ‘the safety time margin’ criterion based on physical parameters. This criterion provides a good indication of when to apply the bleed so as to effectively reduce core uncovery risk as far as possible. This approach also provides information on the degree of uncertainty, which is acceptable in the thermal–hydraulic code for such a calculation, so as to avoid core uncovery.