The concept of an accelerator driven subcritical reactor (ADSR) was elaborated initially as means to solve the problem of nuclear waste. The accent was put on the possibility to transmute the actinides from the spent nuclear fuel. In our works it was shown that ADSR can represent an efficient source of energy, able to ensure a safer exploitation and a deeper burning of the actinides in comparison with a fast reactor. Aspects related with the core geometry, the optimal value of the criticality coefficient keff, the material used for the converter, the fuel type and the level of enrichment were analyzed. A core with 0.985-0.988 and low enrichment ensures a deep actinide burning in one cycle (15-20 % from the initial mass of the actinides, in comparison with 6-7% achieved in a fast reactor). The fuel type and the peak factor (PF) of the power distribution limit the maximum power that can be achieved in ADSR. The nitride fuel gives the possibility to reach the highest beam intensities, allowing to achieve electrical power 750-800 MW for PF 2. The use of LBE itself as converter and a uniform distribution of enrichment, realizing a power distribution with PF ~2 is preferable in the case of proton beams. In the case of ion beams with energy below 0.5 AGeV, the use of a long Be converter and a non-uniform distribution of the enrichment maximize the energy released. For beam intensities above 1·1016, values of G higher than 15 can be achieved in ADSR. The optimal energy for proton beam is in the range 1-1.5 GeV. Ion beams starting with 7Li realize higher values of G. The most interesting results are obtained with 7Li. A beam of 7Li with energy 0.3 AGeV realizes the same electrical power as a beam of 1.5 GeV proton. That allows to diminish the dimension of the accelerator 2 times, correspondingly diminishing the building and maintenance costs.
The high values of the energy gain G that can be achieved in ADSR in comparison with the values of the energy gain predicted in fusion power plants (~3 in the case of magnetic plasma confinement, ~4 in the case of inertial confinement) demonstrates that ADSR can represent a more efficient source of energy.
Based on the analysis of possible experiments meant to check the efficiency of different beams, the design of an extended lead target LETASUR is presented.
