This feasibility study on core control using only the control rods is conducted with the TOPAZE algorithm - implemented in the CRONOS2 core calculation code - in its two versions: version 1 (minimization of the two-dimensional peak, imposed axial offset) and version 2 (minimization extended to three-dimensional, without imposed axial offset). The sensitivity analysis on the power peaks was carried out on the variations of the axial height of the burnable poisons and the type of grey or black control rod clusters. It is demonstrated that the reduction in the number of rod cluster controls allows a correct smoothing of the reactivity over the whole cycle, except for the end of cycle when control rods are moved upward.

For load follow feasibility studies, several approaches, based on simulations performed with MISRITME have been evaluated: variation of the primary flow rate for axial offset control; use of a program, with temperature decreasing with the power; and finally, coupling of a temperature range, centered on a reference temperature with a negative gradient, to the French N4 reactor control mode Dispositif de Manoeuvrabilité Accrue: X (DMAX). It is shown that the return to equilibrium following a low threshold of 40% induces an additional penalty between 15 and 20% on the power peak. Solutions are suggested to globally reduce these peaks, which appear during all operating conditions.

Two types of reactivity-induced accidents linked to clusters are studied: the removal of a rod cluster control assembly (RCCA) at full power (class III) and the ejection of a RCCA (class IV). It is also shown that ejection at zero power, with a released reactivity of 1.86 $, does not cause major damage to cladding and fuel. However, at full power, with the assumptions made, a partial melting of the pellet occurs without however creating fuel dispersion in the coolant.