Benchmark definition

An approach similar to the OECD/NRC Boiling Water Reactor Full-size Fine-mesh Bundle Tests (BFBT) Benchmark was adopted by dividing the benchmark activity into phases. Each phase consisted of several exercises. It was desirable also to plan the benchmark specification to accept as many potential numerical approaches as possible.

The Nuclear Power Engineering Corporation (NUPEC) Pressurised Water Reactor (PWR) Sub-channel and Bundle Tests (PSBT) Benchmark consisted of two phases, and each phase contained three exercises. Phase I was the Void Distribution Benchmark and Phase II was the Departure from Nucleate Boiling (DNB) Benchmark. Participants could choose either of the two phases and any of the exercises within the phases to take part and contribute. The preliminary indications showed that a sufficient number of participants attempted both phases with different numerical approaches.

In addition to the measured experimental data and the relevant boundary conditions, the detailed geometrical data of mock-up assemblies, spacers and the test loop were included as far as possible in the specification to allow a wide range of numerical modelling. Below are the corresponding exercises of the two phases.

Phase I: Void Distribution Benchmark

Exercise I-1: Steady-State Single Sub-channel Benchmark

The goal of this exercise was to benchmark the sub-channel, mesoscopic and microscopic numerical approaches. The experimental data included computed tomography (CT) scan measurements of the sub-channel averaged void fraction in four sub-channel types: typical central, central with a guide tube, side and corner. Also, graphical images of the void distribution within the typical central sub-channels and central sub-channel with guide tube were made available. The test cases were selected at PWE-rated conditions. Different types of single sub-channel test assemblies were used to investigate the effect of geometry on the phenomenon concerned.

Exercise I-2: Steady-State Bundle Benchmark

This exercise was designed for benchmarking mesoscopic numerical approaches. The experimental data includes X-ray densitometer measurements of the void fraction (chordal averaged over the four central sub-channels) at three axial elevations along the bundle length and graphical images of the bundle void distribution. The test cases for this exercise were chosen at PWR rated conditions and deviations of quality from the rated conditions.

Exercise I-3: Transient Bundle Benchmark

The NUPEC PSBT Database included simulation of four representative transients of PWRs; power increase, flow reduction, depressurisation and temperature increase. All four transients were selected as benchmark cases. The experimental data included time histories of X-ray densitometer measurements of the void fraction (chordal averaged over the four central sub-channels) at three axial elevations along the bundle length for four transient scenarios: power increase, flow reduction, depressurisation and temperature increase. Exercise three of Phase I was designed for benchmarking sub-channel numerical approaches.

Exercise I-4: Pressure Drop Benchmark

This exercise was designed for performing code-to-code comparisons concerning axial pressure drop. Although no empirical data was available, code results were to be compared with relevant graphical data.

Phase II: DNB Benchmark

Exercise II-1: Steady-State Fluid Temperature Benchmark

The exercise was designed to assess the thermal-hydraulic code capabilities of predicting the exit coolant temperature. The experimental data included measurements of the sub-channel-averaged fluid outlet temperatures.

Exercise II-2: Steady-State DNB Benchmark

The goal was to assess the thermal-hydraulic code capabilities of correct prediction of DNB along rod bundles. The experimental data includes the power at which DNB occurs and the corresponding locations in the bundle.

Exercise II-3: Transient DNB Benchmark

The exercise was designed to enhance the development of truly mechanistic models for DNB prediction during the four postulated transients in PWRs. The experimental data included the time histories of the boundary conditions and the DNB time detected for four transient scenarios: power increase, flow reduction, depressurisation and temperature increase.