Validation and verification of the coupled neutron kinetic/thermal hydraulic system code DYN3D/ATHLET. (October 2015)

• Record Type: Journal Article
• Authors: Kozmenkov, Y.; Kliem, S.; Rohde, U.
• Is Part Of: Annals of nuclear energy. Volume 84(2015:Oct.)
• Page Start: 153
• Page End: 165
• Publication Date: 2015-10
• Subjects: Coupled code -- Thermal-hydraulics -- Three-dimensional neutron kinetics -- Validation -- Verification -- Benchmarks; Nuclear energy -- Periodicals; Nuclear engineering -- Periodicals; 621.4805
• DOI: 10.1016/j.anucene.2014.12.012
• Languages: English
• ISSNs: 0306-4549

Abstract

Highlights: General description of the coupled code DYN3D/ATHLET developed in HZDR (Helmholtz-Zentrum Dresden-Rossendorf) and used to simulate steady-state and transient behaviour of LWRs is given. Nearly 20 real plant transients and dynamic benchmarks for 5 LWR designs were simulated to evaluate the coupling interface between two integrated codes and interaction of DYN3D model of 3-D neutron kinetics with plant components modelled by ATHLET. An overview of DYN3D/ATHLET validation and verification activities is given with references to published results. The results of two most recently performed validation (Kalinin-3) and verification (AER-7) tasks are presented in a separate chapter in more detail. The validation/verification status of DYN3D/ATHLET was noticeably increased. Abstract: One of the most intensively developing areas in the LWR multi-physics is a coupling of different best estimate 3-D neutron kinetic (BIPR, DYN3D, KIKO3D, NEM, PARCS, etc.) and thermal hydraulic (ATHLET, CATHARE, RELAP5, etc.) codes. Resulting coupled code systems have advanced capabilities of modeling both steady-state spatial distributions of the core power and their evolutions during different kinds of reactor transients. They are also highly useful in the analyses of possible reactor instabilities. Initial steady-state core power distributions can be disturbed by changes in the reactor loop mass flow rates and/or temperatures, by relocations of the low-temperature/diluted-boron water slugs within the primary system or by movements of control rods. The coupled code used for LWR simulations in HZDR is DYN3D/ATHLET, which includes the 3-D core neutron kinetic and thermal hydraulic model of own development – DYN3D. The paper reports major capabilities of DYN3D as well as different ways of its coupling with the thermal hydraulic code ATHLET (external, internal and parallel), but mainly focuses on the validation and verification of the coupled code DYN3D/ATHLET. In the course of DYN3D/ATHLET validation/verification nearly 20 real plant transients and dynamic benchmarks have been simulated and analysed. The LWR types covered by these tasks are: VVER-440 (Bohunice-3, Greifswald-5 and Loviisa-1 units), VVER-1000 (Balakovo-1, Balakovo-4, Kalinin-3, Kozloduy-6, Saporozhye-6 and Temelin-2 units), B&W PWR (TMI-1) and BWR/4 MK-1 (Peach Bottom-2). For each reactor unit a computational model was developed according with the benchmark specifications. The simulated tasks describe different scenarios of increasing complexity, including the transients initiated by the main steam header or main steam line breaks, switching off/on of main circulation pumps, turbine trip and generator load drop. Some of the transients are characterized by a strongly asymmetric behavior of the primary system (e.g. caused by a steam line break), and the processes of coolant mixing in the lower and upper reactor plenums as well as in the downcomer are important for these cases. The coupled code DYN3D/ATHLET models the primary coolant mixing in two ways – whether by an appropriate nodalization of the mixing area or by using two specific models for mixing in the lower plenum. The first of the lower plenum mixing models is a part of the coupling interface at the core inlet plane, while the second one is the analytical coolant mixing model developed for the downcomer and the lower plenum regions of VVER-440. The results of DYN3D/ATHLET simulations were assessed both against available measured data and calculations performed with similar multi-physics codes. The paper includes an overview of the simulated problems and the most representative results of DYN3D/ATHLET validation/verification for all coupling modes. The obtained experience of code validation provides a better understanding of reactor transients with a strong interaction between neutron kinetics and thermal hydraulics, helping to improve computation models. This experience is also useful for the present and further activities in coupling of DYN3D with other best estimate codes, like CFX, TRANSURANUS, etc.