KALININ-3 Coolant Transient Benchmark

Switching off of one of the four operating main circulation pumps at nominal reactor power

Over several years, considerable effort and progress was made in various countries and organisations in incorporating full three-dimensional (3-D) reactor core models into system transient codes. The coupled thermal-hydraulic (TH) and neutron kinetics (NK) code systems allow performing of a best-estimate calculation of interactions between the core behaviour and plant dynamics. Several benchmarks have been developed to verify and validate the capability of the coupled codes in order to analyse complex transients with coupled core-plant interactions for different types of reactors.

The OECD Nuclear Energy Agency (NEA) completed the VVER-1000 Coolant transient benchmark (V1000CT-1) and (V1000CT-2) for evaluating coupled TH system NK codes by simulating transients at the Bulgarian nuclear power plant Kozloduy Unit 6. The available real plant experimental data made these benchmark problems very valuable.

This benchmark is a continuation of the above activities and it defines a coupled code problem for further validation of thermal-hydraulics system codes for application to Russian-designed VVER-1000 reactors based on actual plant data from the Russian nuclear power plant Kalinin Unit 3 (Kalinin-3). The selected transient 'switching off of one main circulation pump (MCP)' was performed at nominal power and lead to asymmetric core conditions with broad ranges of the parameter changes. The experimental data was very well documented. Measurements were carried out with quite a high frequency and their uncertainties were known for almost all measured parameters. This fact allowed for the application of the studied transient, not only for validation purposes, but also for uncertainty analysis as a part of the Benchmark for Uncertainty Analysis in Modelling for Light Water Reactors (UAM-LWR).

The report provided the specifications for the international benchmark problem, coupled VVER-1000 coolant transient (KALININ-3). The specification report had been jointly prepared by leading specialists of the All-Russian Research Institute Nuclear Power Plant Operation (VNIIAES), the Russian Research Centre "Kurchatov Institute' (KIAE),  Gesellschaft für Anlagen und Reaktorsicherheit, mbH (GRS) and Pennsylvania State University (PSU).

The specification covered four exercises: point kinetics model inputs, transient core calculations, transient coupled calculations and uncertainty analysis. In addition, a CD-ROM was also made available with detailed data for the transient boundary conditions, decay heat values as a function of time and cross-section libraries.

In December 2008, the Nuclear Science Committee (NSC) Bureau had expressed support for the coupled Kalinin-3 benchmark problem to become an international standard problem for validation of the best-estimate safety codes. In its February 2009 meeting, the Working Party on Scientific Issues of Reactor Systems (WPRS) discussed the proposal and endorsed it as it is of particular importance for the last phase of the Uncertainty Analysis in Modelling (UAM) activities.

Publications and reports
Background and purpose of the benchmark

Under the guidance of the NEA, many benchmarks have been performed concerning the application of coupled 3-D TH/NK codes. Some of them have utilised code-to-code comparisons, others have compared code predictions with real measured data.

Most transients in a VVER reactor can be properly analysed with a system thermal-hydraulics code, with simplified neutron kinetics models (point kinetics). A few specific transients require more advanced modeling for neutron kinetics for a proper description. A coupled thermal-hydraulics 3-D neutron kinetics code would be the right tool for such tasks.

The proposed benchmark problem has already been analysed by the coupled system code ATHLET-BIPR-VVER. This allowed a better fixing of the Benchmark Specifications. However, within the present context the results of participants will be compared against the measurements. Interesting additional problems have to be solved in order to perform correctly the comparisons. This experience is incorporated in the text of the specification.

The reference problem chosen for simulation is the MCP #1 switching off at nominal power when the other three main coolant pumps are in operation, which is a real transient of an operating VVER-1000 power plant. This event is characterized by rapid rearrangement of the coolant flow through the reactor pressure vessel resulting in a coolant temperature change, which is spatially dependent. This leads to insertion of spatially distributed positive reactivity due to the modeled feedback mechanisms and a non-symmetric power distribution. Simulation of the transient requires evaluation of core response from a multi-dimensional perspective (coupled 3-D neutronics/core thermal-hydraulics) supplemented by a one-dimensional (1-D) simulation of the remainder of the reactor coolant system. The purpose of this benchmark is four-fold:

  • To verify the capability of system codes to analyse complex transients with coupled core-plant interactions and complicated fluid mixing phenomena.
  • To fully test the 3-D neutronics/thermal-hydraulic coupling.
  • To evaluate discrepancies between predictions of the coupled codes in best-estimate transient simulations with measured data.
  • To perform uncertainty analysis having at disposal not only the measured values but also their accuracy


The benchmark included a set of input data for the NPP Kalinin-3 and consisted of four exercises.

Exercise 1 – Point kinetics plant simulation

The purpose of this exercise was to test the primary and secondary system model responses. Compatible point kinetics model inputs which preserved the axial and radial power distribution, CR #10 and #9 reactivity obtained using a 3-D code neutronics model and a complete system description were provided.

Exercise 2 – Coupled 3-D neutronics/core thermal-hydraulics response evaluation

The purpose of this exercise was to only model the core and the vessel. Inlet and outlet core transient boundary conditions were provided by the benchmark team on the basis of calculations performed with the ATHLET-BIPR-VVER coupled code system. Alternatively, participants could apply the measured data. The hot full power (HFP) state (Exercise number 2a) of the core was required for comparison.

Exercise 3

This exercise combined elements of the first two exercises of the benchmark and represented an analysis of the transient in its entirety. For participants that had already taken part in the Kozloduy-6 NEA/OECD Benchmark, it was suggested to start directly with this exercise. As a preliminary step for the latter participants, it was recommended to perform steady state core calculations at hot zero power (HZP) state (exercise 3a), HFP (exercise 3b) and deliver the results for comparisons. Exercise 3a and exercise 3b ensured and checked out the correct application of the cross section libraries, the core loading and the core design geometry.

Exercise 4 – Performing of uncertainty analysis for the purpose of Phase III

System phase of the OECD benchmark for uncertainty analysis in best-estimate modelling (UAM) for design, operation and safety analysis of LWRs. The aim and the specification of the exercise was described in a separate volume, which depicted the state of the art of the results and requirements identified after performing UAM phases I and II. The specification document (edition 1) that covered exercises 1-3 of the OECD Kalinin-3 VVER-1000 coupled code benchmark and the corresponding experimental database and report was made available to participants.

Related data package

Data related to the Kalinin-3 benchmark can be requested from the NEA Data Bank: https://www.oecd-nea.org/tools/abstract/detail/NEA-1848/


Members' working area  (password protected | reminder)

Mailing list archive

Uncertainty analysis in modelling (UAM) list archive