Present status of reactor physics in the United States and Japan-IV. 5. Preliminary Optimization Studies of ADS Target and Buffer Design

Accelerator-driven systems (ADSs) have been considered in several countries to treat spent fuel and to generate power. In these systems, a high-power particle accelerator produces energetic protons that interact with a heavy-metal target to produce neutrons. The source neutrons are generated by direct impingement of the accelerator proton beam onto a target material in a process called spallation. The spallation neutrons are subsequently multiplied in the surrounding subcritical blanket. The spallation target and the subcritical blanket are coupled through a buffer region, which helps the spallation neutrons to diffuse into the blanket and reduces the damage to fuel by high-energy neutrons. The target and buffer designs should be performed in conjunction with the blanket design. For example, to minimize the accelerator power required to produce a desired power with a given blanket design, the target position and the buffer thickness need to be optimized such that the fission neutron production per proton is maximized. At the same time, the damage to fuel at the buffer-blanket interface and the power peaking need to be considered in determining these key design variables. In this paper, we present the preliminary results of target and buffer design studies, focusing on maximizing the source importance in a lead-bismuth eutectic (LBE)-cooled ATW system. The effects of target position and buffer thickness on the source importance were studied using the beginning of equilibrium cycle configuration of an 840-MW(thermal) LBE-cooled ATW design. An R-Z model was used in this study; the blanket region was divided into three enrichment zones and 25 depletion regions. In this configuration, the LBE target region is 169.6 cm high and 8.5 cm in radius and is surrounded by a 28.6-cm-thick LBE buffer. For a 1-GeV proton beam impinging on the top surface of the LBE target with a uniform radial distribution, system performances were calculated using the MCNPX code, which was created by combining the LAHET high-energy physics code and the MCNP Monte Carlo code. The transition energy between LAHET physics and neutron transport using evaluated nuclear data was set to 20 MeV. The effects of buffer thickness were investigated by changing the optical thickness (through LBE density variation). Target position effects were studied by moving the top surface of target along the axial axis. The neutron balance components in the transmuter and the source multiplication factor are summarized in Table I. The nuclear interaction component of production denotes the source neutrons generated by spallation. The source multiplication factor represents the ratio of the fission neutrons to the sum of fission and spallation source neutrons. Table II summarizes the net neutron currents at the interface between the buffer and blanket as functions of energy and axial position. These results show that the number of spallation neutrons produced per proton is relatively insensitive to the buffer thickness or the target position (provided the target/buffer region is thick enough). However, the number of fission neutrons varies more sensitively because of the variations in the transmission of spallation neutrons into the blanket and the reflection characteristics of the buffer. As shown in Table I, the source multiplication factor initially increases and then decreases as the optical thickness of the buffer decreases; one can see by interpolation that the source multiplication factor attains a maximum value when the optical thickness is reduced to ∼80% of the reference value. This behavior seems to be due to two competing effects: the leakage of high energy spallation neutrons into the blanket and the leakage of fission neutrons out of the blanket. As shown in Table II, more high-energy (>20 MeV) neutrons leak into the blanket from the buffer as the optical thickness of the buffer decreases. These high energy neutrons produce a relatively large number of neutrons by nuclear interaction or fission. On the other hand, the leakage of low-energy neutrons from the blanket into the buffer also increases. The fraction of fission neutrons leaking out of the blanket increases from 30.7 to 31.6% when the optical thickness is reduced to 25% (Table I). These low-energy neutrons leaking into the buffer subsequently leak out of the system or into the reflectors and do not contribute to the fission of actinide isotopes. Consequently, the fission neutrons initially increase and then decrease as the optical thickness of the buffer decreases. The source multiplication factor decreases when the target is moved away from the middle portion of the blanket, as expected, since more spallation neutrons leak into the lower or upper reflectors. In particular, in case 5 where the top surface of the target is close to the bottom of the blanket, the high-energy neutrons leaking into the blanket are reduced to less than half of the base case value. One can see by interpolation that the source multiplication factor attains a maximum value when the top of the target is located ∼25 cm above the blanket midplane. In summary, the source multiplication factor reaches its maximum when the buffer thickness is ∼25 cm and the top of target is located ∼25 cm above the blanket mid-plane. Therefore, to ensure negative temperature coefficients (related with thermal expansion of the LBE target/buffer), it is desirable to design the buffer slightly thinner than 25 cm. If a windowless target is employed, similar reasons dictate locating the top of the target slightly higher than 25 cm above the blanket mid-plane. Further studies need to be performed to evaluate changes in buffer geometry and power peaking. The impact of high-energy neutrons that leak through the buffer on the damage rate is a key consideration. (authors)