Page 119 - AERB
P. 119

in coolant temperature, fuel temperature and void are  (110% of FP), reactor power and core reactivity vs. time,
        also calculated.  The core calculations  have also been  and 3D power distributions were calculated. The results
        performed to find out the multiplication  factor, burn-  of TRIVAC were found to show the expected trend.
        up distribution, power peaking factors  and power
        distribution throughout the Cycle-26 reload. The void  7.7.5 OECD-NEA THAI-3 Project and Analysis of THAI
        distribution,  MCHFR and  control  rod  inventory have      Hydrogen Deflagration Tests
        also been estimated.
                                                                  As a part of the international collaborative OECD-
        7.7.3 Analysis of Reactivity Initiated Transients in a   NEA-CSNI THAI-3 project, analysis of the database
              VVER-1000 Reactor                               from a number of hydrogen deflagration experiments
                                                              performed in the THAI facility was carried out within
            Reactivity Initiated Transient (RIT) analysis due to  a theoretical framework. The peak pressures and
        ejection/withdrawal  of Control  Protection  and  System  temperatures obtained during the experiments were
        Absorber Rods (CPSARs) in a VVER-1000 reactor was  compared with the theoretical AICC (Adiabatic
        carried out using TRIKIN as part of Indo-Russian RPWG  Isochoric  Complete  Combustion)  estimates.  As  an
        bilateral benchmark. Simulating the above transients at  illustration, Fig. 7.22 shows the variation of calculated
        hot zero power and other power operating conditions  pressure ratio (burnt gas pressure / initial pressure) for
        (25%,  50%  and 100%  Full Power), the response of  different initial hydrogen and steam mole fractions. The
        core dynamics parameters  to the reactivity insertions  experimental trends were consistent with the theoretical
        were studied to demonstrate the ability of the design for  estimates and brought out the influence of important
        terminating this kind of transients.                  parameters like heat losses, combustion completeness
                                                              etc.  Dynamic  combustion  behaviour  was  analysed
        7.7.4 Analysis of Benchmark on Coupled Neutronics-    within  the framework of  calculated laminar burning
              Thermal Hydraulics Code System                  velocities and experimentally measured flame speeds.
                                                              The influence of direction of propagation, initial
            As an intra-DAE  Benchmark  on the Coupled        temperature and non-uniformity on flame propagation
        neutronics and System thermal hydraulics (ABCS),      was investigated. This analysis provided useful insights
        LOCA in a 540 MWe PHWR was identified for inter-      into the assessment of the static and dynamic effects of
        code comparison exercise. In the first phase, standalone   slow deflagrations and development of a methodology
        core neutronics (static) calculations  were performed   for  modelling  of  slow  deflagrations  in  hydrogen-air-
        using AERB in-house code REDAC (REactor Dynamics      steam mixtures.
        Analysis Code). Parameters  like
        reactivity device worths in different
        configurations   and    reactivity
        coefficients for changes in fuel and
        coolant states  were determined.
        The estimations were found to be in
        good agreement with other codes.

            The transient  phase  of the
        exercise was analysed  to evaluate
        the prediction capability of TRIVAC
        module of DRAGON code system.
        The problem involves localized
        perturbation  in  terms of defined
        changes in coolant density in one
        half  of the core leading  to power
        rise which will be arrested through
        reactor SCRAM. Results like
        SCRAM worth vs. time curve, time
        of occurrence of trip signals due to   Fig. 7.22: Calculated (AICC) Pressure Ratio for different initial Mixture Compositions
        period (less than 10 s) and power


                                                                                      AERB Annual Report 2019  91
   114   115   116   117   118   119   120   121   122   123   124