Due to obvious deficiencies in the calculations with previous APROS versions a new solution model for the thermal stratification of APROS code was developed [23]. The old method used upwind solution for the enthalpy. Due to numeric diffusion the code lost information about the stratified layer. The new higher order numeric method uses information from three consecutive nodes to solve the transported liquid enthalpy. The new enthalpy solution contains a special weight function, which is calculated from liquid enthalpies of the three nodes. The experiments GDE-41 and GDE-43 were recalculated with the new model (Fig. 4). The model was also tested separately with a standalone PSIS [24]. The calculation results were good. The new model eliminated significantly the numerical diffusion and restricted the spreading of the thermal front.

FIG. 4. Temperature distribution in the GDE-43 experiment vs. APROS calculation with the old and new models at time 4000 s.

Lappeenranta University of Technology in co-operation with VTT Energy has performed totally 24 experiments to investigate Passive Safety Injection system with PACTEL facility during past eight years. The experiments in PACTEL have provided valuable information about PSIS behaviour in SBLOCA’s. The PSIS system worked as planned when the CMT was equipped with flow distributor (sparger). The main source of disturbances was condensation, which took place when stem met cold water. The experiment data has been and will be used to further develop Finnish thermal hydraulic analyses code APROS. The computer simulations have reproduced the measured transient behavior with good accuracy despite of smoothing of the thermal front generated by numerical diffusion. The code has been improved by developing a new solution model for the thermal stratification, which largely eliminated the effect of numerical diffusion.

[4]