Best estimate analysis method for the loss of Residual Heat Removal (loss-of-RHR) event during the mid-loop operation is being conducted along the Code Scaling, Applicability and Uncertainty (CSAU) evaluation methodology. The analysis method uses RELAP5/MOD3.2 as a best estimate analysis code. One of the important processes in the CSAU methodology is the development of the Phenomena Identification and Ranking Table (PIRT) which identifies thermal-hydraulic phenomena during the event and ranks the identified phenomena from the view point of influence on the safety evaluation parameters. The safety parameters for evaluation are Reactor Coolant System (RCS) pressure and reactor vessel water level. The PIRT for the reflux cooling of the loss-of-RHR event during the mid-loop operation was developed based on existing integral test results, plant analysis results and related papers considering influence on coolant distribution, non-condensible gas distribution and heat transfer. Referenced integral tests are ROSA-IV/LSTF, BETHSY, PKL and IIST. Uncertainty of RELAP5/MOD3.2 physical models related to high ranked phenomena identified in the PIRT for the reflux cooling is quantified using the related experimental data for application to PWR plant statistical analysis based on the developed verification matrix. Uncertainty quantified models are void model, horizontal stratified flow criteria and SG condensation heat transfer. These models are related to the following phenomena respectively. Void model (interfacial friction factor in bubbly and slug flow regimes): - Two phase expansion in core and upper plenum due to core boiling. - Two phase flow to Steam Generator (SG) inlet plenum and U-tubes. Horizontal stratified flow criterion: - Stratification of flow in hot leg. - Water transportation from hot leg to SG by steam flow. SG condensation heat transfer model: - Heat transfer in SG U-tube under presence of non-condensable gas. Distribution of model parameter multiplier which represents model uncertainty was obtained by either experiment analysis by RELAP5 or comparison of separate RELAP5 model prediction to experimental data. Mean value and standard deviation are calculated for distribution of model parameter multiplier.

This content is only available via PDF.
You do not currently have access to this content.