5
views
0
recommends
+1 Recommend
0 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found

      Xenon instability study of large core Monte Carlo calculations Translated title: Monte-Carlo-Rechnungen zur Xenon-Instabilität von großen Kernen

      research-article
      1 , * , , 2
      Kerntechnik
      Carl Hanser Verlag

      Read this article at

      ScienceOpenPublisher
      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          One of the goals of neutronic calculations of large cores may be self-consistent distribution of equilibrium xenon through the reactor core. In deterministic calculations such self consistency is relatively simply achieved with the help of additional outer iterations by xenon, which can increase several times solution run time. But in stochastic calculation of large cores such increase is utterly undesirable, since even without these outer iterations it demands modeling of billion of histories, which in case of complicated large core may take about a day of 100 processors work. In addition the unavoidable statistical uncertainty here plays role of transient process, which excites xenon oscillations. In this work the rise of such oscillations and the way of their overcoming with the help of hybrid stochastic/deterministic calculation is studied. It is proposed to make at first single static Monte Carlo calculation of given core and to receive multi-group mesh cell characteristics for future use in operative code. This one will evaluate xenon distribution through the core, which will be equilibrium for deterministic solution and substantially close to equilibrium Monte Carlo solution, paid with enormous computing cost.

          Kurzfassung

          Ein Ziel der Berechnung der Neutronenkinetik von großen Kernen ist die Berechnung einer in sich stimmigen Gleichgewichtsverteilung des Xenons im Reaktorkern. Bei deterministischen Rechnungen wird dies mit zusätzlichen Iterationen für das Xenon erreicht, die aber die Gesamtrechenzeit entscheidend vergrößern. Bei stochastischen Rechnungen soll dieser Anstieg der Rechenzeit unbedingt vermieden werden, da Rechnungen großer Kerne schon ohne diese Zusatziterationen bis zu 100 Tagen Rechenzeit benötigen können. Zusätzlich treten bei der Berechnung von Transienten nicht vermeidbare statistische Unsicherheiten auf, die die Größenordnung der eigentlichen Xenonoszillationen übersteigen. In diesem Beitrag werden diese Oszillationen untersucht und es wird eine Methode entwickelt, diese durch eine kombinierte stochastisch/deterministische Analyse zu überwinden. Es wird vorgeschlagen zuerst einfache stationäre Monte-Carlo-Berechnungen des zu untersuchenden Kerns durchzuführen und damit Mehrgruppen-Charakteristika der Gitterzellen für die weitere Nutzung zu erzeugen. Damit können dann die Xenonverteilungen im Kern bestimmt werden, die zum einen den deterministisch berechneten Gleichgewichtslösungen entsprechen und zum anderen den Monte-Carlo-Gleichgewichtslösungen am nächsten kommen. Dieses Vorgehen spart enorme Rechenzeiten.

          Related collections

          Most cited references4

          • Record: found
          • Abstract: not found
          • Article: not found

          The status of MCU-5

            Bookmark
            • Record: found
            • Abstract: not found
            • Article: not found

            Preventing xenon oscillations in Monte Carlo burnup calculations by enforcing equilibrium xenon distribution

              Bookmark
              • Record: found
              • Abstract: found
              • Article: found

              Applying full multigroup cell characteristics from MCU code to finite difference calculations of neutron field in VVER core

              Up to now core calculations with Monte Carlo provided only average cross-sections of mesh cells for further use either in finite difference calculations or as benchmark ones for approximate spectral algorithms. Now MCU code is capable to handle functions, which may be interpreted as average diffusion coefficients. Subsequently the results of finite difference calculations with cells characteristic sets obtained in such a way can be compared with Monte Carlo results as benchmarks, giving reliable information on quality of production code under consideration. As an example of such analysis, the results of mesh calculations with 1-, 2-, 4-, 8- and 12 neutron groups of some model VVER fuel assembly are presented in comparison with the exact Monte Carlo solution. As a second example, an analysis is presented of water gap approximate enlargement between fuel assemblies, allowing VVER core region be covered by regular mesh.
                Bookmark

                Author and article information

                Journal
                kt
                Kerntechnik
                Carl Hanser Verlag
                0932-3902
                2195-8580
                26 August 2016
                : 81
                : 4
                : 363-366
                Affiliations
                1 National Research Nuclear University “MEPHi”, Kashirskoye sh., 31, Moscow, 115409, Russian Federation
                2 Nuclear Research Centre “Kurchatov Institute”, 1, Akademika Kurchatova pl., Moscow, 123182, Russian Federtion
                Author notes
                Article
                KT110712
                10.3139/124.110712
                8f284ff9-7d4b-4acc-8c8c-dbcaa6bbe7c1
                © 2016, Carl Hanser Verlag, München
                History
                : 9 February 2016
                Page count
                References: 6, Pages: 4
                Categories
                Technical Contributions/Fachbeiträge

                Materials technology,Materials for energy,Nuclear physics
                Materials technology, Materials for energy, Nuclear physics

                Comments

                Comment on this article