On the application of CFD modeling for the prediction of the degree of mixing in a PWR during a boron dilution transient

In a Pressurized Water Reactor, negative reactivity is present in the core by means of Boric acid as a soluble neutron absorber in the coolant water. The main functions of the boric acid are to compensate for the fuel burn up and Xenon poisoning during normal operation and to provide the required sub-criticality of reactor shutdown during refueling and maintenance. During a so called Boron Dilution Transient (BDT), the borated coolant water is diluted by mixing with unborated water. The resulting decrease in the boron concentration leads to an insertion of positive reactivity in the core, which may lead to a reactivity excursion. The associated power peak may damage the fuel rods.

The most severe BDT involves scenarios in which a slug of unborated water has been formed in a cold leg in a stationary (Main Coolant Pumps are down) primary circuit. An inadvertent start-up of the MCP of the affected loop causes the transport of the unborated slug into the Reactor Pressure Vessel (RPV). The resulting positive reactivity insertion in the core is governed by the degree of mixing of the unborated slug and the borated water in the cold leg, the downcomer and the lower plenum. This mixing of borated and unborated water is an important process, because it mitigates and determines the degree of reactivity insertion.

The objective of the present study is to develop a validated Computational Fluid Dynamics (CFD) model for the prediction of the boron concentration distribution in the RPV as function of time during a BDT. This CFD model has been validated using the measurement data from the Rossendorf coolant mixing model (ROCOM) experiment. The ROCOM test facility represents the primary cooling system of a KONVOI type of PWR (1300 MWel). The linear scale of the ROCOM experimental facility is 1:5. The RPV is connected with four circulation loops. Various experiments on boron dilution scenarios have been performed. For the pump start-up experiments the following boundary conditions were varied: the length of the pump ramp, the final mass flow rate of the loop with the start–up pump, the volume of the deborated slug and initial position in the cold leg and the status of the unaffected loops.

A detailed CFD model of the RPV including the inlet nozzles, the downcomer, and the lower plenum has been developed to predict the mixing of deborated and borated water in this RPV. The validation of the model has been achieved by comparison of the calculated and measured relative boron concentration at the core inlet plane as function of time. For code validation a slug mixing experiment with 14 s ramp length, 185 m³/h final flow rate, 4 m³ slug volume and 10 m initial slug position was taken. The unaffected loops were open. In spite of the complicated spatial, temporal, and geometrical aspects of the flow in the RPV, the agreement between the calculated and the experimental data is good. The minimal relative boron concentration measured at the core inlet is 64% and the calculated minimum value is 60%.