The gas-cooled fast reactor (GFR), which has been studied since the 1970s, is a high-temperature fast spectrum reactor capable of using a closed all-FR fuel cycle (NEA, 2006b) (Fig. 13.25) . It combines a more sustainable use of uranium

13.25 GFR System Layout with Supercritical-CO2 Indirect Cycle (Hejzalar et al., 2006).

resources and waste minimization with high efficiency electricity generation. If He is used as a coolant, the outlet temperature reaches around 850 °C, requiring an entirely ceramic core. However, this also means the co-generation of high-quality process heat. As well as being identified by GIF as one of the designs to develop, the GFR is also one of three fast reactor designs selected for development to the demonstration stage within the European Sustainable Nuclear Industry Initiative (ESNII), see Table 13.16. Unlike the SFR design, there is less operational experience with GFR design, which means it will take longer to develop. GIF plans to complete a viability assessment by 2012. GFRs will therefore need to be introduced gradually with the move to a fully fast reactor cycle only probable after the turn of the next century.

600 MWth 48% 490/850 °C at 90 bar 100 MWth/m[26] UPuC/SiC (70/30%) with about 20% Pu content 50/40/10% Self-sufficient 5% FIMA; 60 dpa

Table 13.16 GFR characteristics defined by GIF (Foley and Knight, 2009)

The use of a gas coolant (such as He in GFRs) has several advantages (van Rooijen, 2009):

• chemical compatibility with water, obviating the need for an intermediate coolant loop

• good chemical compatibility with structural materials

• negligible activation of coolant

• since gas coolants are transparent, fuel shuffling operations and inspection are easier

• since gas coolants cannot change phase in the core, the potential of reactivity swings in the case of an accident is reduced

• significant reduction of the void coefficient in comparison with SFR systems

• a harder neutron spectrum, which increases the breeding potential of the reactor

• the potential for a larger coolant fraction in the core without an unacceptable increase in parasitic capture

There are also some disadvantages, resulting from the very poor specific heat capacity of gases in comparison to liquid coolants (Bomboni, 2009):

• the need for artificial roughening of the cladding to maintain acceptable cladding temperature, resulting in an increased pressure drop over the core, and necessitating a higher pumping power

• the need to keep the coolant at high pressure compared to liquid coolants (e. g. 7 MPa is needed for He-based coolant systems)

• the risk of significant vibration of the fuel pins due to the high coolant flow velocity

• difficulty in extracting the decay heat from the high power density core, particularly following a depressurization event (an essential element in passive safety systems identified by GIF)

He is the most promising gas coolant. As with the VHTR design, He allows the potential use of a direct Brayton cycle with high efficiency (around 50%). As a backup option, an indirect cycle using a secondary circuit with supercritical CO2 (25 MPa, 650 °C) could be used (Fig. 13.25b), achieving a cycle with a similar efficiency (van Rooijen, 2009; Hejzlar et al., 2006).

Although no definitive design has yet been agreed, a good example of a GFR system could be the plate-type GCFR 2400 MWft ‘E’ proposed by CEA (Richard et al., 2006). The main design parameters are summarized in Table 13.17. An overview of the core layout is shown in Fig. 13.26 , the geometry of fuel assembly shown in Fig. 13.27 and Fig. 13.28, and the main fuel plate characteristics set out in Table 13.18. The materials composing the core (except for HMs) are Si, C, He and Zr, which minimize parasitic absorptions. A small fraction (1.5% by volume) of the core is composed of a liner, which functions as a sort of catcher for

Table 13.17 Main core parameters of GCFR 2400 ‘E’ (Richard et al., 2006; Girardin et al., 2006)

Thermal power (MWth) 2400

Power density (kW/l) 100

Specific power (W/gHM) 40

Height/diameter ratio 0.63

Theoretical breeding gain 0.0

Fissile height (mm) 2300

No of fuel assemblies 162+120

No of control rods 24

No of reflector assemblies (mixture of Zr3Si2, SiC and He) 168

No of Nominal coolant pressure (MPa) 7.0

Helium inlet temperature (°C) 480

Helium outlet temperature (°C) 850

Maximum clad temperature (°C) 985

Maximum fuel temperature (°C) 1860

Coolant volumetric fraction (%) 30.8

Structural material volumetric fraction (%) 20.8

Helium pressure drop through the core (bar) 1.6

Average coolant speed through the core (m/s) 85

# Inert 1 central

О Fi ssile 162 zone 1 О Fi ssile 120 zone 2

# Control / AU 6 + 12 « AU SAC 6

О Reflector 168 О Neutronic shielding 164

13.26 GCFR 2400 MWth ‘E’ core (Girardin et al., 2006).

the volatile FPs (Girardin et al., 2006). A definitive choice of liner materials still needs to be made.

An important feature is the high height to diameter (H/D) ratio in comparison with typical FR values (‘pancake cores’) and the ‘zero’ breeding gain. A higher H/D ratio reduces leakages and improves the neutron economy (Bomboni et al.,

2008a). This allows for very high irradiation levels, a relatively small fissile inventory, large flexibility in the choice of fuel composition and the option of inserting dedicated targets for transmutation without significant reduction of core performance. Since the GFR design aims at a ‘self-sustainable’ cycle, i. e. a

Table 13.18 Main fuel plate characteristics (Richard et al. 2006) (U, Pu)C fraction (%vol) 53.1 SiC (%vol) 16.1 Helium (%vol) 27.7 Liner (%vol) 3.1 Pu enrichment (%at) 18 Density (g/cc) (U, Pu)C — TD 13.6 (U, Pu)C — HM theoretical density 12.9 (U, Pu)C 8 5% TD SiC 3.16 Thermal conductivity of (U, Pu)C (W/mK) 19.6

production of fissile material that is equal to its consumption, an optimal H/D has to be established, which is sufficiently high to sustain the cycle without external addition of fissile material and/or the presence of fertile blankets but, at the same time, is not too high for thermal-fluid-dynamics reasons (van Rooijen, 2009). A breeding blanket of depleted uranium (DU) is not envisaged, because it could pose proliferation risks.