The performance of subcritical nuclear reactors driven by proton and ion beams (accelerator driven system- ADS) is analyzed and the advantage of ion beams is substantiated. The conditions which maximize the power production and the energy gain G (defined as the ratio of the power produced to the power spent for the beam acceleration), ensuring in the same time a safe exploitation are identified.
With a proper choice of the target G of 20-30 can be obtained.
The particle fluence and the energy released are obtained through simulation with Geant4. The power spent to accelerate the beam is calculated by scaling from the data about the accelerator efficiency for a reference particle.
Cylindrical targets with rods of solid fuel (metallic alloy, oxide, carbide) in a bath of coolant are considered. The most significant influence on the energy released demonstrates the material used for the converter. The use of light materials increases the energy released especially for light ions at low energy. The best results are obtained with Be converter with length 100-120 cm.
The value of the criticality coefficient keff must be chosen as high as possible to maximize the power produced, but low enough to ensure a safe functioning of the reactor. The reactivity changes during various accident scenarios were analyzed in order to identify possible positive reactivity insertions. A value of 0.985 for keff ensures enough safety margin.
The results obtained with protons and ions beams from deuteron to 20Ne and energies from
0.2 to 2 AGeV, accelerated in a linac and interacting in U-Pu-Zr target with Be converter and keff 0.985 are presented. The beam intensity is 1.25‧1016, and the linac efficiency 0.18 for protons (values taken from the European Spallation Source project).
The optimal energy for proton is 1.5 GeV, with a G of 10 and net power Pnet of 150 MW. G 2-3 times higher can be realized with ion beams. With a beam of 7Li with energy 0.25 AGeV one gets the same Pnet but with G of 18, and necessitates an accelerator 2.6 times shorter. At intermediate accelerator length beams of Li or Be are the best option (G 20-25, Pnet 350-400 MW). In an accelerator with the same length as for 1.5 GeV proton it is preferable to accelerate ions with higher mass (12C, 16O, 20Ne) getting a G~ 30 and Pnet~1 GW.

Dr. Mihaela Paraipan has completed her PhD from Politechnica University Bucharest, Romania.
She is working in present as senior researcher at the Joint Institute for Nuclear Research, Dubna , and she has more than 20 articles in peer reviewed journals.