Assessments of the kinetic and dynamic transient behavior of sub-critical systems (ADS) in comparison to critical reactor systems

Abstract

The neutron kinetic and the reactor dynamic behavior of Accelerator Driven Systems (ADS) is significantly different from those of conventional power reactor systems currently in use for the production of power. It is the objective of this study to examine and to demonstrate the intrinsic differences of the kinetic and dynamic behavior of accelerator driven systems to typical plant transient initiators in comparison to the known, kinetic and dynamic behavior of critical thermal and fast reactor systems. It will be shown that in sub-critical assemblies, changes in reactivity or in the external neutron source strength lead to an asymptotic power level essentially described by the instantaneous power change (i.e. prompt jump). Shutdown of ADS operating at high levels of sub-criticality, (i.e. k eff ∼0.99), without the support of reactivity control systems (such as control or safety rods), may be problematic in case the ability of cooling of the core should be impaired (i.e. loss of coolant flow). In addition, the dynamic behavior of sub-critical systems to typical plant transients such as protected or unprotected loss of flow (LOF) or heat sink (LOH) transients are not necessarily substantially different from the plant dynamic behavior of critical systems if the reactivity feedback coefficients of the ADS design are unfavorable. As expected, the state of sub-criticality and the temperature feedback coefficients, such as Doppler and coolant temperature coefficient, play dominant roles in determining the course and direction of plant transients. Should the combination of these safety coefficients be very unfavorable, not much additional margin in safety may be gained by making a critical system only sub-critical (i.e. k eff∼0.95). A careful optimization procedure between the selected operating level of sub-criticality, the safety reactivity coefficients and the possible need for additional reactivity control systems seems, therefore, advisable during the early design phase of any ADS systems in order to assure a benign transient response of the particular ADS design under investigation to typical plant transient initiators.