Kinetics models for safety studies of accelerator driven systems

Abstract Safety analyses of Accelerator Driven Systems (ADSs) are mainly performed by codes developed in the past for critical reactors, a point-kinetics model for computing the transient power being often employed. It is shown that this model – that assumes time-independent neutron direct and adjoint flux shapes – may be inaccurate in the ADS case even if it is acceptable for a similar “critical” case. Although the material distribution remains almost unchanged, flux (power) shape variations may be significant in the first case due to external source related effects. An option for a more refined modelling of the neutron adjoint flux in ADS analyses is discussed. This option is shown to be rather complicated in the general case: it may give rise to involvement of several adjoint flux shapes and several sets of related point-kinetics parameters (reactivity, etc.). To improve the accuracy of the point-kinetics treatment, an extended point-kinetics model, which employs several flux (or power) shapes precomputed at steady-state conditions, is proposed in the paper. This approach may help to avoid the spatial kinetics treatment if no strong material movement occurs. The reactivity and other point-kinetics parameters are defined similarly to a critical reactor.