Coolant mixing under natural circulation conditions following a hypothetical SB LOCA event

Funded from 01.01.2002 to 31.03.2003 by the Nuclear Special Committee "Plant engineering" of VGB PowerTech (Germany) under the registration number: SA"AT" 29/01.

Background

Partial depletion of the primary circuit during a hypothetical small break loss of coolant accident can lead to the interruption of one-phase flow natural circulation. In this case, the decay heat is removed from the core in the reflux-condenser mode. For the scenario of a hot leg side leak and hot leg safety injection thermal hydraulics analyses using the system code ATHLET showed, that weakly borated condensate can accumulate in particular in the pump loop seal of those two loops, which do not receive safety injection. According to these ATHLET-calculations, one-phase flow is maintained in the remaining two loops at high residual heat conditions because of the entrainment of safety injection coolant into the steam generators. After refilling of the primary circuit, natural circulation in the two stagnant loops simultaneously re-establishes and the de-borated slugs are shifted towards the reactor pressure vessel. Mixing in the downcomer and the lower plenum is an important phenomenon mitigating the reactivity insertion into the core in this postulated scenario. For that reason, the mixing processes were investigated at the ROCOM test facility.

Boundary conditions for the ROCOM experiments

Based on the ATHLET-calculations, a volume flow rate of 5 % of the nominal rate was set in the loops running in one-phase flow. The volume flow rate in the two restarting loops increases from zero to 6 %. In these two loops, de-borated slugs of 7.2 m3 were assumed corresponding to the volume of the whole loop seal. Experiments were carried out with variation of the density difference between de-borated slug and ambient coolant due to differences in boron concentration and temperature. Two of them are presented here, the first without density difference and the second with a relative density difference of 2 %. The density of the de-borated slugs was reduced in comparison to the ambient coolant by adding alcohol. Further, in each experiment, the de-borated slugs are tracered with salt, what changes the electrical conductivity. The measured conductivity values (Link: working principle of the wire mesh sensors) are transformed into a dimensionless mixing scalar by relating the local conductivity to the conductivity of the initial slug. So, this dimensionless mixing scalar represents the normalized de-boration theta, characterizing the amount of the initial perturbation (slug) detected at a certain position. These values can be easily transformed into boron concentration data using boundary values for the initial boron content in the reactor and in the slug.

Results

Fig. 1 shows the time records of the perturbation theta at the sensors in the upper and lower downcomer in the experiment without density differences. The figure depicts the unwrapped view of the time dependent theta distribution in the middle of the downcomer. Red arrows indicate the positions of the loops with the slugs, black arrows the two others. First of all, the time records of the perturbation at the upper sensor demonstrates, that both slugs enter the vessel at the same time. The transportation time is about 45 s and the deviation is less than 1 s, what underlines the quality of the pump control system. Further, it is clearly to be seen, that the perturbation is restricted to a sector of the vessel. The sector corresponds nearly to the fraction of the flow rate of the two loops with the slugs. A certain amount of mixing takes place with the coolant from the two other loops on the way from the upper to the lower sensor, as can be concluded from the visualization at the lower downcomer sensor. The maximum values measured at both sensors are 90.4 % and 86.1 %, respectively. That means, the perturbation is only slightly attenuated on the way through the downcomer. The shape of the perturbation in the downcomer is created by the typical velocity distribution in the downcomer, as it was found in velocity measurements. There is a minimum of the velocity directly below the four inlet nozzles, while in the middle between two inlet nozzles a maximum of the velocity is observed.

With no density difference, the weakly borated coolant almost perpendicularly flows down in the downcomer and a maximum of 64 % of the initial perturbation is detected in the core entry section below the loops where the slugs were formed (Fig. 2, the arrows indicate the positions of the loops with the slugs).

Now, the experiment was repeated with identical boundary conditions, only the density of the slugs was reduced by 2 %. This density difference impedes the down flow of the de-borated condensate. It stratifies on top of the heavier coolant and flows around the core barrel in the upper downcomer. The transfer to the core is delayed in this way. At the opposite side of the downcomer, that coolant with lower density is entrained by the heavier water (see Fig. 3). Consequently, the maximum values of theta reached at the sensor in the lower downcomer and at the core inlet (Fig. 2) are lower in comparison to the experiment without density difference. For the investigated density difference of 2 %, a value of 31 % only of the initial under-boration was measured at the core entrance. In addition to that, the position of the maximum mixing scalar theta is shifted towards the core entrance cross section part opposite to the two loops that provided the slugs.

The variation of the density difference significantly changes the mixing behavior.


Fig. 2: Distribution of the perturbation in the core inlet plane at the time of maximum (maximum: bold number in %, stepwidth: difference between two isolines)
Fig. 1: Time record of the perturbation at the sensors in the downcomer (unwrapped view) in the experiment without density differences


Fig. 3: Time record of the perturbation at the sensors in the downcomer (unwrapped view) in the experiment with a density difference of 2%

Publications

Contact

S. Kliem