# Investigation of solution methods suitable for modelling steam collapse and pressure peaks using RELAP5

[Examensarbete på avancerad nivå]

In this report an investigation of solution methods suitable for modeling steam collapse and pressure peaks using RELAP5 (modification 3.3, patch 03) is presented. Simulations using RELAP5 can deviate much from the reality as steam collapse occurs in the system. Steam collapse results in high and sudden pressure peaks and from a safety point of view it is essential that the pressure amplitudes are overestimated in the simulations. The uncertainty in the modeling results using RELAP5 can cause problems, for example when dynamic loads for nuclear power plants are calculated. There are today no clear instructions on which settings should be used in RELAP5 during such conditions or thorough investigation to validate the deviation of the results from the actual values. This has imposed problems in validating that load calculations using RELAP5 when steam collapse occurs are correct. The purpose of this project is therefore to develop a solution method suitable for RELAP5 simulations when steam collapse occurs to better simulate the actual force on the system. In the future, this developed solution method can be used when no experimental data is available and this work can be used as a reference as how to handle certain types of steam collapse and as a validation that this solution method does not result in an underestimation of the pressure peaks. Experimental data is available for three different setups, where one of the experiments was performed at the auxiliary feed water system at Ringhals 1. The method used for obtaining a suitable solution method, is to compare the experimental data to simulations performed using different solution approaches in RELAP5. Three main changes to the solution approach are investigated: activation of the equilibrium solution and homogenous solution and deactivation of the critical flow model. In addition to these changes some further solution approaches, not anticipated to affect the results in a higher degree, are investigated. The results showed that one solution approach were more suitable to use as steam collapse occurs. The equilibrium solution should be used, as the nonequilibrium solution resulted in a clear underestimation of the pressure peaks. The critical flow model should be deactivated to estimate the pressure amplitudes more correct. The homogenous solution should not be used, since it was observed that RELAP5 was unable use the homogenous solution in the volumes and changed to the nonhomogenous solution as steam collapse occurs. The recommended solution method is hence to activate the equilibrium solution and deactivate the critical flow model.

**Nyckelord: **RELAP5, steam collapse, cavitation, pressure peaks, simulation, modelling