Reduced-moderation water reactors – the future of LWRs?

As a future energy source, the concept of reduced-moderation water reactors (RMWRs) is being investigated by the Japan Atomic Energy Research Institute (JAERI) in conjunction with the Japan Atomic Power Company (JAPC) under technical co-operation with the Japanese LWR vendors Hitachi, Toshiba and Mitsubishi Heavy Industries. Based on LWR technology, RMWRs are being developed to use uranium resources more efficiently, use plutonium as a fuel, and reduce production of radioactive waste through a high burn-up/long operation cycle. To accomplish these aims it is essential to have a high conversion ratio of around 1.0. By maintaining plutonium quality multiple recycling will be possible, thereby contributing to the reduction of radioactive waste production.

To obtain such a high conversion ratio the neutron energy in the core should be increased by reducing the water to fuel volume ratio. The RMWR core design should consequently be different from conventional LWRs. The water volume in the RMWR core can be considerably reduced by introducing tight-lattice fuel assemblies. Increasing the core void fraction for a BWR-type reactor also contributes to the reduction of neutron moderation in the core. For a PWR-type reactor the heavy water coolant can also reduce neutron moderation. The resultant neutron spectrum is similar to that in FBRs, and much harder than that in LWRs (see Figure 1).

Due to the safety requirements for LWRs, another design goal is to establish negative void reactivity coefficients for the RMWRs. This can be achieved by:

•Designing the core to be short and/or flat in order to increase neutron leakage while maintaining the total core volume at the same value.

•Introducing axial and/or radial depleted uranium blankets, which also increases the conversion ratio.

•Inserting streaming channels into the core.

Three BWR-type and two PWR-type core designs have been developed. To confirm the feasibility of these RMWR core designs further research and development studies is being carried out. The table on page 26 summarises the main dimensions and characteristics of the RMWR designs.

BWR-type designs

The three BWR-type core designs have a high conversion ratio of around 1.1, a high burn-up of 60GWd/t along with long operation cycles of about two years and a simple core without blanket region with the similar square fuel assemblies convertible to ABWRs.

High conversion ratio core

To achieve a high conversion ratio, the core consists of hexagonal fuel assemblies with triangular tight-lattice configuration (see Figure 2 below). A Y-shaped control rod is introduced for every three fuel assemblies.

The pitch of the assembly is about 220mm. The diameter of the fuel rods is 14.5mm and the gap between rods is kept at 1.3mm, which is considered to be a tentative limit with regard to structure and heat removal. The core average void fraction is significantly increased to 70%. As a result, the effective moderator-to-fuel-volume ratio (Vm/Vf) is reduced to about 0.17.

To obtain negative void reactivity coefficients, each core layer is shortened to about 0.2m high. An internal depleted UO2 blanket region is positioned between two core layers (Figure 3). With the upper and lower blankets, the total core region has the five-layer structure with a height of 1.21m.

The fuel rods consist of MOX pellets and Zircaloy cladding. There is a radial distribution in Pu of five steps in the assembly to reduce the local peaking to under 1.04.

The core calculation was performed by the three dimensional void-power iteration method, treating each assembly separately. The conversion ratio is calculated to be 1.1 at the middle of the equilibrium cycle. The void reactivity coefficient is evaluated as -1.0x10-4dk/k/%void. The number of assemblies necessary for 1100MWe output is 924. The core outer diameter is 7.2m, though the size is within the fabrication capacity of the reactor pressure vessel. The core average discharge burn-up is 45GWd/t and the operation cycle length is 14 months.

The maximum linear power density is 55.8kW/m, which is within the limit for the previously used 7x7 type fuel assembly for BWRs. The minimum critical power ratio (MCPR) is calculated to be 1.3 under the normal operational conditions.

Another design with the reduced gap width of 1.0mm is also under investigation to reduce the core diameter.

High burn-up and long operation cycle core

Due to the high conversion ratio in RMWRs the burn-up reactivity is small. Therefore high burn-up and a long operation cycle can be achieved simultaneously.

To reduce the void reactivity coefficient the core is shortened and a void tube assembly is introduced. The void tube assembly (see Figure 4) consists of a hexagonal diving bell-type void tube and a heat source.

The void tube is filled with steam generated in a hafnium heat source located just below the tube. The heat is generated by the gamma rays due to neutron capture reactions. The flow area between the void tube and the channel box is filled with two-phase flow. When the core power is increased, the void fraction in the void tube assembly is also increased due to the gamma heating. This results in the increase of neutron leakage through the void tube assemblies, and hence the void reactivity coefficient becomes negative.

The core configuration with the void tube assemblies is shown in Figure 5. In the present design, the core consists of 252 large hexagonal fuel assemblies and 61 void tube assemblies. The fuel assembly consists of 469 fuel rods, which are 11.9mm in diameter and have a triangular lattice arrangement with a gap width of 1.3mm. The Vm/Vf is reduced to about 0.5 and the core average void fraction is set at around 60% to reduce the water volume.

Simple core without blanket

If blankets are not used and the conversion ratio is more than 1.0, there is no need to separate uranium and plutonium in the reprocessing. The reprocessing process might therefore be significantly simplified, with only the removal process of the fission products being necessary. Since the conversion ratio is more than 1.0 the spent fuel contains more plutonium than the fresh fuel, suggesting that the MOX fuel will be simply fabricated by adding depleted UO2 to the MOX material obtained from the reprocessing.

The fuel rods are arranged in a square lattice configuration (Figure 6). Cross-shaped control rods are introduced as shown.

Negative void reactivity coefficients are attained by shortening the core to about 2.6m. Since the core length is not reduced significantly the fissile Pu content can be lower than other BWR-type designs resulting in a higher conversion ratio and lower void reactivity coefficient.

PWR-type designs

Two types of PWR core with heavy water and light water moderation have been developed.

Heavy water cooled core

For the heavy water cooled PWR-type reactor a radial and axial heterogeneous core is employed to attain a conversion ratio greater than 1. The triangular tight-lattice fuel assembly is also employed to reduce the core water to fuel volume ratio. The core consists of seed fuel assemblies with or without an internal blanket, and blanket fuel assemblies. The fuel assembly with an internal blanket has been employed for negative void reactivity coefficients and the fuel assembly without an internal blanket has been introduced to alleviate the axial neutron decoupling effect (Figures 7 and 8). The loading pattern of the blanket fuel assemblies is designed to give a flat distribution of the assembly power (Figure 9).

Light water cooled core

The light water PWR-type RMWR has seed fuel rods surrounded by blanket fuel rods (Figure 10). The seed fuel region has 15 layers and the blanket fuel region has 4.5 layers to achieve negative void reactivity coefficients. The void reactivity coefficient is further reduced by scattering 36 zirconium hydride (ZrH1.7) pins in the surrounding blanket region.

Since the seed-blanket assembly concept is still at a preliminary stage, the design goal of a high conversion ratio greater than 1 has not been fully satisfied. However, the conversion ratio is expected to exceed unity by considering the contribution of the radial blanket. Also, the void reactivity coefficient is negative through the whole operation cycle under the 3-batch refueling scheme except at the beginning and end of its life. The void reactivity coefficient is expected to become negative during the whole operation cycle by using the shuffling scheme shown in Figure 11. These two effects were not considered in the present evaluation.