Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - McSAFE (High-Performance Monte Carlo Methods for SAFEty Demonstration- From Proof of Concept to realistic Safety Analysis and Industry-like Applications)

Teaser

Monte Carlo methods are increasingly use for nuclear reactor calculations. MC-capabilities have largely increased thanks to cheap and huge computer power. Prediction of local safety parameters in full core simulations still requires improvements to get sufficient statistical...

Summary

Monte Carlo methods are increasingly use for nuclear reactor calculations. MC-capabilities have largely increased thanks to cheap and huge computer power. Prediction of local safety parameters in full core simulations still requires improvements to get sufficient statistical accuracy and acceptable CPU-time.
The overall objective of the McSAFE project is to move the Monte Carlo coupled solution to become valuable and widespread numerical tools for realistic core design, safety analysis and industry-like applications of LWRs of generation II and III.
As Monte Carlo codes are very flexible with respect to reactor geometry and composition, the results of the project will also be useful for future types of nuclear reactors as well as to research reactors with complicated geometries.
Main targets of the project are advanced depletion, optimal coupling of Monte Carlo codes to thermal-hydraulic and thermo-mechanic solvers, time-dependent Monte Carlo and methods for massively parallel simulations.
This developments will enable the prediction of safety parameters with less conservatism and hence to improve the operational flexibility.
The analysis of transients i.e. the dynamic core behavior, extensions of Monte Carlo e.g. the behaviour of prompt and delay neutrons including their precursors is needed. in addition, modeling of control rod movements is important.
Another important pillar of the McSAFE project that the multi-physics coupling developments rely on the open-source SALOME platform, which is the basis of the European NURESIM Platform for reactor analysis developed during the NURESIM, NURISP and NURESAFE projects.
Finally, one important objective of McSAFE is to increase the confidence in high-fidelity tools by performing an extensive validation using plant data of operating reactors (VVER, and Western type PWER) as well as of experiments (SPERT III E REA). This step is very important for the acceptability of the new codes by the stakeholders (Regulators, utilities, manufacturers).

Work performed

Within the McSAFE project, the work scheduled for the first 18 months were achieved mainly in time. Selected achievements of each technical work package are given hereafter:

For an optimal depletion of full-cores taking into account TH-feedbacks, selected achievements are given:
• Development of a numerically stable MC-depletion simulations which are affected by many parameters e.g. time step length, number of criticality MC-calculations per time step, number of active/inactive cycles at each criticality calculation, number of neutron histories per criticality cycle. The new method is based on the use of a diffusion solver and a simplified depletion feedback that allow to carry out a large number of coupled calculation in short time.
• Assessment of scalability and stability of MC-codes for depletion calculations to identify limitations e.g. memory requirements and scalability in a HPC-cluster.
• Development of a collision-based domain decomposition (CDD) method to facilitate full-core calculations reducing the memory needs in view of the large number of nuclides.
In the Work Package devoted to the multi-physics coupling developments, the following achievements can be highlighted:
• Monte Carlo solvers (SERPENT, TRIPOLI and MONK) were coupled with a thermal hydraulic solver using the ICOCO-method and the MEDCoupling libraries.
• Furthermore, SERPENT, SUBCHANFLOW and TRANSURANUS were coupled to each other taking advantage of the ICOCO-methodology.
• Implementation of an internal coupling of SERPENT/SUBCHANFLOW as a back-up solution (based on MP-interface).
• Testing of coupled codes solving different problems for cores with square and hexagonal fuel assemblies.
• Benchmark problems for VVER and Western type PWR were defined to compare the different codes and the two coupling approaches.
The main achievements of the WP4 devoted to time-dependent Monte Carlo codes are:
• Major extensions of the Monte Carlo codes (SERPENT, TRIPOLI and MCNP) were added to the codes in order to be capable of solve transient problems coupled with a thermal hydraulic solver.
• The dynamic version of the MC-codes was tested by solving an academic problem (minicore). Work hast been started to simulate the SPERT-III E core for a real validation of the new dynamic version of the involved codes. Present results are very promising.
• To identify the bottlenecks and challenges of dynamic MC-codes in solving large problems, a rod ejection in a mini-core -consisting of a 3x3 Fuel Assembly problem- was solved.
• Investigations related to the assessment of the variance in dynamic Monte Carlo simulations has been started (SERPENT, TRIPOLI) by performing pin-by-pin simulations of a PWR fuel assembly. Different options to reduce the variance will analyzed in the future.
• Finally, studies about the parallelization of MC-TH-codes to perform time-dependent simulations has been started and will play a major role in the second reporting period.
The validation work is perform in WP5 and it will be done mainly during the second reporting period using plant data (PWR and VVER-1000) as well as SPERT IIIE-experimental data.

Final results

The progress achieved in McSAFE considered as beyond state-of-the-art is listed below:
a) development of methods for stable depletion calculation, fission source acceleration and improved convergence behavior of MC/TH-solvers,
b) Collision based domain decomposition for MC-codes to reduce the memory requirement of large depletion problems,
c) flexible methods based on ICOCO for neutronics, thermal-hydraulics and thermo-mechanics coupling using different geometries, and
d) dynamic capability of Monte Carlo codes paving the way for unique transient analysis.

The potential impacts of main outcomes of the McSAFE-projects can be summarized as follows:
• Impact on innovation capacity and integration of new knowledge within EU in the field of reactor safety: deliver novel numerical tools not available before for transient analysis; provide validated numerical simulation tools which may serve as reference solutions to deterministic codes in the frame of safety demonstration
• Impact on interaction with civil society and stakeholders: MC-based McSAFE-tools can be used by nuclear community (regulators, utilities, manufacturers) to reduce conservatism and better predict safety margins.
• Impact on reinforcing the European leadership’s nuclear engineering by providing advanced methods to better assess the safety of NPP and make NPP operation more flexible while keeping high safety standards.
• Economic impact: EU-users will profit from the extended capabilities of the tools, which may lead to a decrease of conservatism of engineering factors and to better fuel exploitation and fuel cycle economy in general.
• Environmental impact: The McSAFE-tools contribute to better predict safety features of core designs and to reduce probability of accidents leading to a release of radioactivity, as well as to the EU goals pf low carbon economy.
• The societal impact of the project is reflected by the fact that the risks of potential accidents is reduced by the use of novel McSAFE tools under development. In addition, the dissemination of the gained knowledge to the young generation and society will increase the general understanding about the nuclear technology and related risks.

Website & more info

More info: http://www.mcsafe-h2020.eu.