Abstract:
A literature review of current phase separation publications was conducted. Data
sets were collected and compiled into a Two-Phase Flow Separation Database.
Examination of this database indicating a need for further investigation into the liquid
entertainment phenomena for smaller hot-leg to branch diameters and intermittent flow
regimes. A detailed analysis to the prototypic phase separation process is presented and
the associated phenomena are identified. Appropriate scaling criteria were employed for
the design of a scaled test facility. Geometry and the flow conditions of the test facility
were determined accordingly to Wu et. al (1998).
A series of phase separation tests conducted at the Air-water Test Loop for
Advanced Thermal-hydraulic Studies (ATLATS) and Advanced Plan Experiment
(APEX) has been completed. Results show that the criteria developed by Smoglie (1984)
used in RELAP5, reasonably predicts the onset of liquid entrainment. However, the
steady-state entrainment correlation in RELAP5 significantly underpredicts primary
coolant removal rates. This discrepancy is due to the effects of downstream boundary
conditions and pool entrainment and carry-over from the reactor vessel. Due to pool
entrainment, entrainment through the branch continues when the reactor vessel mixture
level drops below the bottom of the hot-leg. This investigation shows that RELAP5 is
non-conservative when predicting coolant removal rates due to steady state liquid
entrainment in a horizontal mainline with a vertical branch for stratified, stratified-wavy,
transition, and stepped hot-leg flow regimes.
