ATHLET-calculations of natural circulation experiments at the PANDA test facility

F. Schäfer

Within the NACUSP project the thermohydraulic code ATHLET was used for the calculation of natural circulation experiments at the PANDA test facility. PANDA is a large-scale thermohydraulic test facility located at Paul Scherrer Institute Villigen, Switzerland (PSI). The test facility was designed to investigate the system behaviour of Light Water Reactors (LWR) and to study containment phenomena. For the NACUSP project the test facility was modified and only a few parts of the facility were used for the tests. For the experiments the RPV is connected to the isolation condensor (IC). The steam produced in the reactor core flows to the IC and the condensate flows back to the downcomer (DC). The subcooling at the core inlet can be controlled by the Auxiliary Water System (AWS). For different natural circulation experiments the core inlet hydraulic resistance coefficient can be varied by adjusting the gap between the lower edge of the shroud and the bottom of the RPV.


Two experiments were defined as a benchmark to validate thermohydraulic computer codes. The experiments were performed at a pressure of 5 bar, appr. 900 kW core power and with different core inlet hydraulic resistance coefficients (k=7 and k=500). The calculations were performed with the code ATHLET. The input dataset models all main parts of the PANDA configuration used for the natural circulation experiments. The ATHLET model consists of the lower plenum, the core section with 115 electrical heater elements, the riser, upper plenum, downcomer and the upper part of the RPV. The Isolation Condensor is not modeled. Therefore a bypass is connected to the upper part of the RPV and to the downcomer, modeling the IC-feed and drain lines by fills with constant mass flows and given enthalpy (only drain line). A valve at the lower end of the downcomer (VLV) is used to model different k-factors of the core inlet. The cross section of this valve can be changed to adjust the core inlet flow resistance.

The transient calculations start from a steady state calculation. To simulate the system behaviour with respect to natural circulation stability a short disturbance of the drain line mass flow was used in order to stimulate oscillations in the loop. The response of the system can be seen in the riser mass flow.


Although the oscillations were damped in the ATHLET calculations, a oscillation period and also a decay ratio can be calculated from the DC velocity. The ATHLET calculations predict stable behaviour for both tests and no limit-cycle oscillations occur.


The results of the experiments and also the ATHLET calculations show that the oscillation period decreases with increasing DC velocity. Both ATHLET calculations show a good agreement with the experimental results.

More about the ATHLET-calculations and the experiments:

ATHLET calculations for PANDA benchmark (PDF 348 kb)