Extension of the X2 VVER-1000 benchmark by a control rod cluster ejection exercise

•VVER-1000 control rod ejection benchmark.•Detailed benchmark definition.•Suitable for high-fidelity code systems.•Code-to-code comparison. This paper proposes a new benchmark for VVER-1000 control rod ejection transient. The benchmark is designed for code-to-code comparison and its purpose is testing and verifications of time-dependent solvers, core models and cross section generation methodology. Four solutions obtained using different combinations of three lattice transport, two nodal neutronics, and two thermo-hydraulic codes are analyzed, demonstrating good agreement in predicting transient reactivity and power peaks and highlighting the impact of gas gap modeling and DNBR correlation on simulation results.