Conference Agenda

Subchannel (SC) analysis has historically supported core design numerical simulations for a wide variety of concepts encompassing thermal and fast reactors. The suitability of SC codes to heavy liquid metal coolants and extreme operating conditions have been the object of a work package in the framework of the IAEA Coordinated Research Project ‘Benchmark of Transition from Forced to Natural Circulation Experiment with Heavy Liquid Metal Loop (NACIE)-CRP-I31038’, based on the experimental data provided by the NACIE-UP loop located at the ENEA.

The facility features 19, electrically heated, Fuel Pins Simulator (FPS) arranged with a triangular pitch and spaced by a wire wrap, instrumented with 67 thermocouples. The thermal hydraulic problem in the FPS assembly has been simulated by eleven institutions with nine different SC codes. The focus of the simulations is on the steady states of both forced and natural circulation conditions, as well as the in-between transition. In this paper two extreme cases ahave been analysed, one with all pins heated (ADP10) and one with only the seven central pins active (ADP06). ADP10 test is more representative of a condition which could be found in power reactors, while ADP06 test is a challenging power-profile condition, which allows for unprecedented physics insights, both cases can be used for validation of the SC codes .

The comparison demonstrates that SC codes can reliably capture the temperature profile within a wire-wrapped pin assembly, though it also highlights a need for modelling improvements of the wire effect with extreme intra-bundle temperature gradients.

TerraPower is developing a subchannel analysis code, Mongoose++, to support the design of the Natrium^® demonstration reactor. Development of this code was initially motivated by the need to predict core-wide duct temperature distributions to calculate reactivity feedback from radial expansion and assembly bowing. Such calculations require resolution of local coolant flow and temperature distributions within an assembly as well as global inter-assembly heat transfer effects. While Computational Fluid Dynamics (CFD) remains computationally prohibitive for routine design and analysis calculations at the core-wide scale, intermediate fidelity subchannel methods are well-suited for these tasks and have a proven record in the industry for licensing calculations.

Mongoose++ is written in C++ and traces its lineage to the legacy COBRA series of subchannel codes. Mongoose++ utilizes a similar formulation of the subchannel conservation equations but with several advancements to support the specific needs of the Natrium^® project. The subchannel equations are discretized using the finite volume method on a staggered mesh and solved iteratively using a variant of the SIMPLE algorithm. Each iteration of the SIMPLE algorithm is parallelized across assemblies. While more costly than traditional axial-marching schemes, the SIMPLE algorithm is more robust when modeling assemblies with significant buoyancy-induced flow redistribution and localized flow recirculation; such conditions may arise in SFR non-fuel assemblies during off-normal operating conditions at reduced flow.  

This paper provides a detailed overview of the design and implementation of Mongoose++, discussing current capabilities and planned developments. Benchmark comparisons against legacy experimental data and recent CFD calculations are presented.

Recent efforts have focused on at establishing high-fidelity, multi-physics safety analysis methodologies to assess realistic safety margins. However, these approaches still exhibit conservatism by employing conservative initial conditions, such as assuming a hot rod with maximum power for the whole reactor core. This study presents the development of a coupled code for pin-wise reactor core analysis, coupling the subchannel analysis code CUPID, neutron kinetics code MASTER, and fuel performance code GIFT. The objective is to generate accurate pin-wise fuel rod conditions during normal operation, providing more realistic initial conditions for safety analysis. The coupled code performs thermal-hydraulic analysis of the reactor core, accounting for fuel deformation, and simulates realistic power distributions with reactivity feedback from coolant and fuel temperatures. Coupling between the codes was achieved using socket communication and dynamic link library (DLL). A practical simulation was performed on an OPR1000 reactor core during the first cycle, and key results from the steady-state simulation were evaluated. The impact of the fuel performance code was also examined by comparing the results of the coupled CUPID/MASTER and CUPID/MASTER/GIFT codes. Finally, the effect of pin-wise initial conditions on safety analysis was investigated for a reactivity-initiated accident (RIA) scenario with results compared between conservative initial conditions and pin-wise initial conditions.

Hitachi has been developing subchannel analysis codes to predict thermal-hydraulic characteristics of newly designed BWR (Boiling Water Reactor) fuel bundles. The steady-state subchannel analysis code SILFEED (Simulation of Liquid Film Evaporation, Entrainment, and Deposition) with an updated film flow model has been mainly utilized for mechanistic film dryout predictions of various fuel bundle designs. Next-generation fuel bundles, such as Hitachi's RBWR (Resource-Renewable BWR) and GNF3, feature tight lattice configurations, axially varying water rod, and partial-length rods. For these designs with complex geometry, accurate evaluation of thermal hydraulic behavior for each subchannel is required.

To address these demands, Hitachi is developing a new subchannel analysis code based on a transient two-fluid three-field model enabling more advanced evaluations of void fraction and pressure drop by directly solving the void fraction. While its current capabilities are limited to steady-state and transient void fraction and pressure drop, future developments aim to extend its capability to fuel temperature and critical power predictions.

The first step of the code validation, we performed calculations of void fraction and pressure drop in NUPEC 8×8 bundle and φ5.2 to 10.1 mm tube, and compared with experimental data. The results demonstrate that the code achieves prediction accuracy of void fraction within ±15% and pressure drop also within ±15% under BWR operating conditions, which are comparable to those of other subchannel analysis codes.

Accurate modelling of the annular film flow regime in BWR fuel assemblies is of paramount importance for the prediction of the minimum film thickness, dryout location and duration, and crud deposition. Recent decades have seen the introduction of complex assembly structures such as part-length rods and spacer grids, whose complex effects on the flow must also be modelled appropriately for accurate estimation of safety parameters. This paper presents the initial steps towards an improved modelling package for churn-turbulent and annular flow at a subchannel scale. The subchannel code CTF, which uses a two-phase three-field approach, is modified to implement updated models for the droplet entrainment and deposition rates and a new solver (SCARF) has been developed, verified, and applied to partition the subchannel flow rate of the liquid field from CTF during annular flow into distinct films on adjacent rods. This method enables the assessment of film depletion on the sides of each rod and preserves the relative isolation of the films. The results of the two codes are then compared and show that the SCARF produces similar estimates of global flow parameters significant differences at the subchannel scale are observed. The effect is exacerbated for non-uniform radial power profiles. Future research will expand on this by incorporating new models for turbulence and void drift. Additionally, these models will consider how the distribution of film flow within the assembly affects inter-channel and inter-field transfer rates.

Plate-fuel reactors are one of the most common types of research reactors. The thin fuel plates have characteristics such as enhanced thermal conductivity and heat transfer properties, optimized neutron flux, and fuel performance, which makes them an attractive alternative. CTF is a state-of-the-art sub-channel code used for reactor thermal-hydraulics, initially developed for rod bundles and core analysis. In the current version of CTF, the heat generative solid structures are limited to cylindrical shaped rods or tubes. This work aims to develop a new heat structure model in CTF for the plate-type fuel geometry expanding the current capabilities of the code. This new feature extends the applicability of CTF to research reactors, mini- and micro reactors utilizing this fuel design. This development supports the growing demand for accurate thermal-hydraulic modeling and simulation of diverse fuel types and configurations to support the feasibility and safety analysis of small modular reactors. The proposed paper details the implementation process of the plate-fuel heat structure in CTF together with a detailed verification and validation process based on the analytical solution and the experimental available data, respectively. The verification process includes heat conduction within the plates and the convective heat transfer between the plate and the coolant. Thermal expansion and fuel performance models have been analyzed and they are currently suggested as a future improvement of the code.