Development of PIRT and Verification of RELAP5 Void Model for Application to the Loss-of-RHR Event During Mid-Loop Operation

The loss of RHR during mid-loop operation in PWR is relatively high risk event. More confident analysis of the event is desirable to develop better counter measures and increase plant safety. The analysis methodology with statistical method using a best estimate analysis code to increase confidence of analysis result is under development. The method employs the RELAP5/MOD3.2 code as a best estimate code and is being developed along the CSAU methodology. One of the most important steps in the CSAU methodology is development of PIRT (Phenomena Identification and Ranking Table) for the event. The PIRT is developed for the loss of RHR event during mid-loop operation with mitigation measure of reflux cooling and gravity injection from RWST and important models of the RELAP5/MOD3.2 related to high ranked phenomena are identified. Verification matrix is also developed for the important models. One of the important models identified is void model. This model affects two phase water level of the reactor vessel and how much water is transported with vapor from reactor vessel. Verification of void model is especially focused on low power and low pressure conditions which are characteristics of the loss of RHR event under mid-loop operation. Prediction error of void model was quantified for both heated rod bundle channel and non-heated channels. Experiment with rod bundle core geometry under low power and low pressure conditions used for verification analysis is the THETIS experiment. The experiment was performed under quasi-steady condition. Two phase level under specified collapsed level was measured with varying power and pressure. Analysis results with pressure 0.5 to 1.0 MPa predict two phase level within 10% error. Void prediction analyses with non-heated channels were conducted against both steam-water experiment and air-water experiment with various pressure and hydraulic diameter. Most of data are predicted within 30% error.Copyright © 2008 by ASME