This study reviews several aspects of the APEX test facility concerning its capability to accurately simulate the plant conditions of the AP600 – in particular, mass loss quantification through the simulated break geometry and its effects on the primary system depressurization transient.

Brief introductions of several critical flow models have been reviewed and some derivations of said model equations have been given. The models discussed included the Homogeneous Equilibrium Model, Equilibrium Rate Model, and Henry-Fauske Subcooled Model. These correlations were then compared to the measured break flow data from the APEX Test facility. It was clearly demonstrated that the equilibrium flow models (i.e., HEM and ERM) were inadequate in predicting the data. This discovery is, of course, a result of the facility’s break nozzle geometry having a length-to-diameter ratio (L/D) significantly less than 40, which has been shown to be required to attain equilibrium flow conditions. The Relaxation Length Model also demonstrated this inadequacy even though it contains a non-equilibrium parameter. The predictions of the HEM, ERM and RLM under-predicted the data by at least an order-of-magnitude. The Orifice Equation Model over-predicted the initial break flow, but in spite of its simplicity, its predictions were of the same order-of-magnitude as the measured data. To reasonably predict the initial mass flow rate from the simulated break, the Henry-Fauske Subcooled Model should be used with a simple assumption of saturated liquid at a given system pressure. If actual subcooled system conditions are used, the HFSM yields slightly higher flow rates.

The measured break flow data from the APEX Test facility has been demonstrated to be consistent and repeatable. Due to the design of the BAMS, an initial delay exists within the measured flow rates. This delay has been demonstrated to limit accurate measurements of initial break flow rates using a pressurizer liquid level-depression rate for simulated break diameters greater than one inch. This limitation results from the fact that the subcooled blowdown transient occurs much more quickly for the simulated two inch break than for the simulated one inch and one-half inch breaks. However, it is the author’s opinion that this delay has little or no negative effect on the BAMS’ ability to accurately measure the flow rates for the entire depressurization transient.

In addition to initial break flow data assessments and comparisons, it was desired to accurately quantify and predict the mass lost from the primary cooling system through the break. To do this, a constant critical mass flux model was defined and compared to the time-dependent integrated mass lost via the break nozzle. The model predicted mass losses based upon initial system conditions, and it was shown to predict the measured data very well. Due to its repeatability, the BAMS data was presented in a non-dimensional form on a single plot, and all of the test data was shown to be reasonably predicted by the model to the time of ADS-1 actuation.

A phenomenon related to the BAMS measurement delay occurs during tests with simulated two-inch and larger break sizes. This phenomenon is a depression in the Break Separator’s initial liquid level. Because the liquid level is determined using a Differential Pressure transducer, the magnitude of the liquid level depression can not be accurately determined during the initial blowdown. Although the duration of this phenomenon is quite short and its impact on the overall transient is negligible, it is recommended that improved measurement techniques be used for liquid level measurements in the separators.