Modeling of horizontal steam injection in a water pool

Abstract

A comprehensive understanding of the direct contact condensation (DCC) phenomena of steam discharged into a subcooled water pool is essential for the nuclear engineering field. DCC is the most rapid and efficient way to condense high-pressure and high-temperature steam with simple engineering systems. However, the pressure oscillation due to the rapid condensation and small length and time scales of turbulent two-phase flow make the investigation of DCC phenomena challenging either by experiments or with numerical simulations.This paper presents the computational fluid dynamics (CFD) simulations of the small-scale separate effect test facility (SEF-POOL) experiment of Lappeenranta-Lahti University of Technology (LUT University). In the SEF-POOL test, the steam was injected using the orifice of a diameter of 16 mm. All simulations were performed by employing the compressible two-phase solver of OpenFOAM which was based on the Eulerian–Eulerian two-fluid approach. The interfacial heat transfer between steam and water was modeled by using the Nusselt number formulation of Coste (2004). The Rayleigh–Taylor interfacial (RTI) area model of Pellegrini et al. (2015) was applied for the interfacial area modeling. In this work, the formation and collapse of the steam bubbles are studied using the extended pattern recognition based image analysis algorithm. The pattern recognition (PR) algorithm is based on the video material recorded during the SEF-INF2 experiment of the SEF-POOL test facility.The presented study show that if the two-phase interface is rough and smear into the subgrid scale, the interfacial area density calculation solely with the volume fraction gradient is inadequate for predicting rapid condensation rates. Results indicate that rapid and high interfacial accelerations in DCC trigger interfacial instabilities which increased the interface roughness and total interfacial area consequently. Thus, interfacial area modeling with the RTI increased the total interfacial heat transfer rate in the simulations. The achieved results exhibit that the selected modeling approach captured the rapid bubbling condensation oscillation mode of the SEF-POOL test. Presented results demonstrate that image analysis and pattern recognition have great potential in thermal-hydraulic research that can be beneficial for direct contact condensation and phase-interface dynamics model development.