Under the International Cooperation Plan of Rostechnadzor for 2019, a delegation of JSC VO “Safety” took part in a workshop arranged jointly by the Nuclear Energy Agency of the Organisation for Economic Co-operation and Development (OECD NEA) and Society of Nuclear Installations and Reactors Safety of Germany (GRS). The event was held in the GRS Headquarters in the town of Garching (FRG) in the period of July 23 to 27 of 2019.

About 40 specialists representing European, US, South Korean, and Russian technical support organizations as well as research and design institutes took part in the workshop. The participants discussed the results of process simulations for NPPs with VVER-1000 type reactors. Unit 2 of the Rostov NPP at the stage of the power start-up was used as a basis for the simulation. Particular attention was paid to the problems related to validation of multi-physical models that describe the NPP unit.

The workshop held in GRS was considered as a report about the activities performed by the Expert Group on Multi-Physics Experiment and Benchmark Validation (EGMPEBV) of the Nuclear Energy Agency of the Organisation for Economic Co-operation and Development (OECD NEA). It was the second workshop after a kick-off workshop which was arranged by Italian technical support organization Nuclear and Industrial Engineering (N.I.N.E) in the town of Lucca (Italy) in 2018.

Historically, one of the goals of the EGMPEBV working group is to promote the development and improvement of safety analysis calculation techniques for water-water reactors, in particular, VVER type. As computing technologies mature, software tools are also in progress by using more advanced computing techniques. Today, different physical phenomena may be simulated jointly by the same software tool (hereinafter, software): thermal hydrodynamics of coolant, three-dimensional kinetics, fuel behavior, etc. Several presentations given at the workshop were dedicated to this issue and demonstrated how several software codes may be consolidated in a single one. Developers of such software tend to improve the accuracy of calculations, and experimental data are required in this case for the software verification. However, the experimental base for the software verification shrinks, and the base accumulated does not often cover all the essential needs for the verification. In this respect, the benchmark developed on the basis of the experimental data from Unit 2 of the Rostov NPP (hereinafter referred to as Rostov-2 benchmark) is considered as a huge step forward as the I&C of Rostov NPP Unit 2 allowed one to get more data than same systems at other NPP units.

The discussion covered the following topics:

• Reference specification and contents of the specification itself;
• Cross-section library, its format and contents;
• Experimental and instrumentation data;
• Missing information and clarifications required;
• Schedule of control measures and scope of exercises performed.

GRS also hosted a workshop on the validation of multi-physical models covering the following:

• Issues of validation of respective software that implements traditional multi-physical approaches to the simulation of phenomena;
• Issues of assuring required data for the simulation of phenomena with the help of advanced techniques and multi-physical tools;
• Comparison of capabilities of types of software that implement single physical phenomena with the multi-physical approaches;
• Issues of verification and validation (V&V) and uncertainty quantification (UQ) in the existing types of software that implement traditional multi-physical approaches to the simulation of phenomena;
• Issues that occur during verification and validation (V&V) and uncertainty quantification (UQ) in the use of new (novel) multi-physical software;
• Demand for data to validate new (novel) multi-physical software;
• Issues of the optimum use and preservation of obsolete experimental data;
• Issues of verification experiments to be held at new facilities;
• Possibility of use of present-day scientific data and artificial intelligence for the validation;
• Preparation of benchmarks and development of a methodology for validation of a multi-physical model;
• Issues of international and national joint activities related to the validation of models.

The discussion also covered the use of existing software tools and their combinations, such as ATHLET and SKETCH-N, ATHLET, Serpent, and DYN3D, a Swiss multi-physics simulation platform, a software tool of the US Nuclear Regulatory Commission (US NRC), etc.

The workshop organizers paid particular attention to the ATHLET software and its advanced capacities. In particular, there was a discussion about a so-called plug-in to combine ATHLET with other types of software.

Mr. Kostadin Ivanov from North Carolina State University, USA, made a specific presentation about the BEPU2020 Conference which will be held in 2020 and it will be dedicated to all the problems above.