Periodic Reporting for period 2 - CAMIVVER (Codes And Methods Improvements for VVER comprehensive safety assessment)
Reporting period: 2022-03-01 to 2023-08-31
The European nuclear fleet is currently going through LTO upgrades and that remains an important concern for the European Community. The codes and methods continuous update needed for answering the regulatory requirements for reactors LTO is the basis of CAMIVVER project, for the improvement of codes and methods for VVER comprehensive safety assessment in support to other activities carried out concerning VVER fuel development and qualification. CAMIVVER project is oriented to VVER-1000 reactor type.
The CAMIVVER project activities have been built to reach four main objectives:
- The improvement of scientific computer codes, models, and methods to be used at an industrial level.
- The promotion of 3D neutronics-thermalhydraulics coupled calculations to improve the safety assessment by a better representation of the physical phenomena.
- The promotion of the use of advanced mathematical methods for the assessment of uncertainty propagation within numerical simulations.
- The integration of the VVER specificities, to challenge the robustness of codes and methods and their validation strategy.
The CAMIVVER project activities have been built to reach four main objectives:
- The improvement of scientific computer codes, models, and methods to be used at an industrial level.
- The promotion of 3D neutronics-thermalhydraulics coupled calculations to improve the safety assessment by a better representation of the physical phenomena.
- The promotion of the use of advanced mathematical methods for the assessment of uncertainty propagation within numerical simulations.
- The integration of the VVER specificities, to challenge the robustness of codes and methods and their validation strategy.
Apart from management activities (WP2), the programme was distributed over 5 technical Work packages (WP3 to WP7) plus one for the project communication, dissemination and exploitation activities (WP8).
Collecting all necessary reference data has been conducted within WP3.
WP4 aimed to make a step forward in the framework of multi-parameter neutron data libraries generation for LWR (in particular for VVER) using the new code APOLLO3® from CEA. A proof of concept (PoC) of a multi-parameter neutron data library generator (NEMESI) has been developed. 2D assembly calculations have been carried out with NEMESI and compared against APOLLO2, SERPENT2, and TRIPOLI-4® computed results using an automatic comparison tool developed for this specific purpose. PWR and VVER MPOs for WP5 and WP7 activities have been produced via NEMESI and VVER assembly calculation schemes have been proposed. Analysis of the advanced features (new self-shielding model, new neutronic data, 2D vs. 3D treatment, etc.) available in the APOLLO3® lattice code has been performed. Results obtained on 2D and 3D configurations have open discussions on improved reflector models and axial interfaces, respectively.
WP5 aimed to open the discussion on a core multi-physics approach based on neutronics and thermal-hydraulics coupled Best-Estimate simulations. Discussions on multi-physics resolutions are internationally available. The activities carried out in WP5 allowed to bring the industrial view and needs to the international level. The actions proposed aimed at providing test cases representative of PWR and VVER cases and boundary conditions to assess:
- performances of APOLLO3® core solvers using its internal multi-1D thermo-hydraulic library and comparing its results against other state-of-the-art closed channel codes;
- a newly proposed APOLLO3®/CATHARE3 coupling and benchmarking against existing 3D multi-physic High Fidelity models based on Serpent/SCF coupling;
- the suitability of a framework for verifying the multi-parameter neutron data libraries produced in WP4.
Considerable work has been done on the definition and the assessment of the reference test cases and core boundary conditions.
WP6 had the objective to improve CFD modelling and validation for VVER applications, especially mixing in primary vessel. It consisted in three tasks: building a CFD model of the primary vessel of Kozloduy-6; performing a CFD transient calculation of a mixing experiment; and finally, running multiple times this transient with different inlet conditions to assess the propagation of their uncertainties through the model. Thanks to the high-performance capabilities of calculations available, by increasing the number of cells in the models, it has been possible to refine the boundary layers or model fine geometric elements otherwise treated with a simplified geometry. Even for the complex elements still represented with a simplified geometry, the associated physical models have been improved in the codes what ensures a better accuracy than before. A benchmark was conducted over the 3 tasks to ensure the consistency between several CFD codes and methods. The aim was to demonstrate the robustness of CFD models and the high fidelity of the results they provide despite their complexity. In the first two tasks, the benchmark of the codes consisted in comparing the results to actual data and code-to-code comparisons. In the last task, the propagation of the uncertainties through the CFD models was estimated. Each partner of WP6 has been able to set up a calculation with its own methods and practices and provide results comparable to the reference data.
WP7 objectives aimed to improve thermal-hydraulics modelling of VVER plants, especially: Challenge robustness and validation of CATHARE3 in the context of VVER reactors, and perform significant upgrades of RELAP5 models, including switching to TRACE models. The CATHARE3 code includes advanced capabilities for 3D modeling and is based on three-field equations that allow a better simulation of two-phase flows. CATHARE3 relies on a strong validation regarding Western PWR but had not been tested to VVER specificities. To improve the models’ descriptions, the activity has been organized in several phases with increased complexity: 1) development of thermal-hydraulics-code models of VVER primary and secondary circuits and performing steady-state benchmarks to check models consistency; 2) simulation of Kozloduy-6 Main Coolant Pump start-up transient (derived from OECD/NEA VVER-1000 Coolant Transient Benchmark); 3) modeling and comparing results over an “SB LOCA + SG line break” transient; and 4) modeling and comparing results of a main steam line break (MSLB) transient typical chosen for evaluating 3D methods between thermal-hydraulics and neutronics codes.
Finally CAMIVVER highlight was the Final Workshop (WP8) held at KIT in July 2023. A total of sixty-four participants from ten different countries attended the workshop, including representatives from twenty organizations belonging to universities, research centers, industries and regulators. Three invited talks were also given by VVER experts from UJV, Studsvik Scandpower and Atomenergoservice LLC. Three Ph.D. works related to CAMIVVER and VVER modelling in general have been presented. Last work in WP8 consisted in writing a guidance report based on the lessons learned and best practices observed during the project.
Collecting all necessary reference data has been conducted within WP3.
WP4 aimed to make a step forward in the framework of multi-parameter neutron data libraries generation for LWR (in particular for VVER) using the new code APOLLO3® from CEA. A proof of concept (PoC) of a multi-parameter neutron data library generator (NEMESI) has been developed. 2D assembly calculations have been carried out with NEMESI and compared against APOLLO2, SERPENT2, and TRIPOLI-4® computed results using an automatic comparison tool developed for this specific purpose. PWR and VVER MPOs for WP5 and WP7 activities have been produced via NEMESI and VVER assembly calculation schemes have been proposed. Analysis of the advanced features (new self-shielding model, new neutronic data, 2D vs. 3D treatment, etc.) available in the APOLLO3® lattice code has been performed. Results obtained on 2D and 3D configurations have open discussions on improved reflector models and axial interfaces, respectively.
WP5 aimed to open the discussion on a core multi-physics approach based on neutronics and thermal-hydraulics coupled Best-Estimate simulations. Discussions on multi-physics resolutions are internationally available. The activities carried out in WP5 allowed to bring the industrial view and needs to the international level. The actions proposed aimed at providing test cases representative of PWR and VVER cases and boundary conditions to assess:
- performances of APOLLO3® core solvers using its internal multi-1D thermo-hydraulic library and comparing its results against other state-of-the-art closed channel codes;
- a newly proposed APOLLO3®/CATHARE3 coupling and benchmarking against existing 3D multi-physic High Fidelity models based on Serpent/SCF coupling;
- the suitability of a framework for verifying the multi-parameter neutron data libraries produced in WP4.
Considerable work has been done on the definition and the assessment of the reference test cases and core boundary conditions.
WP6 had the objective to improve CFD modelling and validation for VVER applications, especially mixing in primary vessel. It consisted in three tasks: building a CFD model of the primary vessel of Kozloduy-6; performing a CFD transient calculation of a mixing experiment; and finally, running multiple times this transient with different inlet conditions to assess the propagation of their uncertainties through the model. Thanks to the high-performance capabilities of calculations available, by increasing the number of cells in the models, it has been possible to refine the boundary layers or model fine geometric elements otherwise treated with a simplified geometry. Even for the complex elements still represented with a simplified geometry, the associated physical models have been improved in the codes what ensures a better accuracy than before. A benchmark was conducted over the 3 tasks to ensure the consistency between several CFD codes and methods. The aim was to demonstrate the robustness of CFD models and the high fidelity of the results they provide despite their complexity. In the first two tasks, the benchmark of the codes consisted in comparing the results to actual data and code-to-code comparisons. In the last task, the propagation of the uncertainties through the CFD models was estimated. Each partner of WP6 has been able to set up a calculation with its own methods and practices and provide results comparable to the reference data.
WP7 objectives aimed to improve thermal-hydraulics modelling of VVER plants, especially: Challenge robustness and validation of CATHARE3 in the context of VVER reactors, and perform significant upgrades of RELAP5 models, including switching to TRACE models. The CATHARE3 code includes advanced capabilities for 3D modeling and is based on three-field equations that allow a better simulation of two-phase flows. CATHARE3 relies on a strong validation regarding Western PWR but had not been tested to VVER specificities. To improve the models’ descriptions, the activity has been organized in several phases with increased complexity: 1) development of thermal-hydraulics-code models of VVER primary and secondary circuits and performing steady-state benchmarks to check models consistency; 2) simulation of Kozloduy-6 Main Coolant Pump start-up transient (derived from OECD/NEA VVER-1000 Coolant Transient Benchmark); 3) modeling and comparing results over an “SB LOCA + SG line break” transient; and 4) modeling and comparing results of a main steam line break (MSLB) transient typical chosen for evaluating 3D methods between thermal-hydraulics and neutronics codes.
Finally CAMIVVER highlight was the Final Workshop (WP8) held at KIT in July 2023. A total of sixty-four participants from ten different countries attended the workshop, including representatives from twenty organizations belonging to universities, research centers, industries and regulators. Three invited talks were also given by VVER experts from UJV, Studsvik Scandpower and Atomenergoservice LLC. Three Ph.D. works related to CAMIVVER and VVER modelling in general have been presented. Last work in WP8 consisted in writing a guidance report based on the lessons learned and best practices observed during the project.
A step forward in using new-generation European codes has been achieved. CAMIVVER has pushed the development and validation of new generation scientific computer codes as part of a lab to industry process. CAMIVVER allowed improving the robustness and qualification level of these codes for VVER applications but also for LWR in general. Reinforcement of the modeling capabilities of LWR has been obtained using improved physical models in the new codes, and the development of 3D multi-physics coupled methods that provide a better evaluation of local heterogeneities, allowing a more comprehensive assessment of the accidental events. Eventually CAMIVVER has provided recommendations about code development practices, numerical and experimental benchmark definitions, verification & validation practices, and uncertainties approaches.