No abstract

No abstract

This paper deals with the development and validation of a coupled RELAP5/PARCS model of the Swedish Ringhals-3 pressurized water reactor against a Loss of Feedwater transient, which occurred on August 16, 2005. At first, the stand-alone RELAP5 and PARCS models are presented. All the 157 fuel assemblies are modeled in individually in both codes. The model is furthermore able to handle possible asymmetrical conditions of the flow velocity and temperature fields between the loops. On the neutronic side, the dependence of the material constants on history effects, burnup, and instantaneous conditions is accounted for, and the full heterogeneity of the core is thus taken into account. The reflectors are also explicitly represented. The coupling between the two codes is touched upon, with emphasis on the mapping between the hydrodynamic/heat structures and the neutronic nodes. The transient was initiated by a malfunction of the feedwater valve at the 2nd steam generator. Consequently, the turbines were tripped and, because of the low level in the SG-2 the reactor was scrammed. Activation of the auxiliary feedwater provided proper amount of cooling from the secondary side, resulting in safe shutdown conditions. Capabilities of the RELAP5 code were more challenged in this transient, where the influences of the feedback from the neutron kinetic side were also taken into account in the analysis. The calculated values of the parameters show good agreement with the measured data.

A Main Steam Line Break (MSLB) is an important transient for Pressurized Water Reactors (PWR) due to the strong positive reactivity introduced by the over-cooling of the core. Since this effect is stronger when the Moderator Temperature Coefficient (MTC) has a large amplitude, a conservative result will be obtained for a high burnup of the fuel due to the more negative MTC late in the cycle. The calculations have been performed at a cycle burnup of 12.9742 GWd/tHM. The Swedish Ringhals-3 PWR is a three loop Westinghouse design, currently with a thermal power of 3000 MW. The PARCS model has 157 fuel assemblies of 8 different types. Four different types of reflector are used. The cross sections, and kinetic data were obtained from CASMO-4 calculations, using a cross section interface developed at the department. There are 24 axial nodes, and 2×2 radial nodes for each assembly. The transient option for calculating the effect of poisoning was used. The PARCS model has been validated against steady-state measurements from Ringhals-3 of the Relative Power Fraction (RPF) and of the core criticality. The RELAP5 model has 157 channels for the core which means that there is a one to one correspondence between the thermal hydraulics model and the neutronics model. There is eight axial nodes. Originally, the intention was to have 24 axial nodes but this proved not to work because of some limitation in RELAP5. There is currently no mixing between the different channels in the core. The feedwater, and turbines are modelled as boundary conditions. The stand-alone RELAP5 model has been validated against steady state measurements from Ringhals-3. A number of different cases were considered. In the first case, both the isolation of the feedwater for the broken loop, and all the control rods were assumed to work properly. For the second case one of the control rods was assumed to be stuck. The stuck rod was located in the fuel assembly with the highest power. This rod has also one of the highest rod worths. In the final case, the feedwater control valve for the broken loop was fully open. None of the cases led to any recriticality. The increase in power for each fuel assembly was also investigated. With the control rod located in the assembly with the highest power, the maximum power increase before scram turned out to be about 25% compared to the initial power.

No abstract