Management options for Fukushima corium

Highlights

• Options for corium management are reprocessing or packaging for direct disposal.

• Direct disposal may have cost advantages but presents novel safety challenges for both operational and post-closure phases.

• Concerns include criticality management and modelling of long-term dissolution of materials.

Abstract

The loss of core cooling for units 1–3 during the accident at Fukushima Dai-ichi caused major fuel damage. Although full details are not yet available, fuel melting produced corium within the reactor pressure vessels that has, to an unknown degree, melted through into the primary containment. The present priority is cooling the damaged reactors and managing contaminated water, but planning of longer term decommissioning has already begun. Management of highly damaged fuel and corium will be of primary concern, with the main options being recovery for reprocessing or packaging for direct disposal. Although the latter option may have significant cost advantages, it presents some novel safety challenges for both operational and post-closure phases. Concerns include criticality management and modelling of long-term dissolution of materials having highly variable composition. Further R&D is required to fill knowledge gaps – of which the most sensitive may involve determination of the extent to which small “hot particles” of corium have been produced.

Introduction

The 2011 earthquake off the pacific coast of Japan generated a devastating tsunami that triggered an unprecedented series of reactor severe accidents at the Fukushima Dai-ichi nuclear power plant (denoted here as “1F”). An overview of the progress of this incident and its consequences are described in detail elsewhere (Hatamura et al., 2012, Kurokawa et al., 2012). This paper focuses entirely on planning the decommissioning of the damaged reactors and, in particular, management of the corium produced as a result of core melting in units 1–3. The main technical issues involved are outlined in Fig. 1 although, as discussed below, socio-political and communication issues must also be taken into account before major actions are implemented.

Worldwide, there have been a number of accidents involving reactor core damage, most of which had little radiological significance (McKinley et al., 2011). Three-Mile Island (TMI) is probably most relevant for corium management, with defueling completed in 1990 (USNRC, 2009), after which the corium was transported to the Idaho National laboratory where it sits on a concrete plinth awaiting final disposal (IAEA., 1991, IAEA., 1992, EPRI, 1990, EPRI, 1992). Other reactors that suffered major core damage were either simply sealed, e.g. Windscale, UK; Chernobyl, Ukraine or decommissioned, with damaged fuel either reprocessed, e.g. Lucens, Switzerland (ENSI, 2012) or stored for eventual direct disposal, e.g. SRE (Sodium Reactor Experiment), USA. In addition, reactor severe accident experiments have been conducted for decades to study a wide range of phenomena. These include fuel rod dryout and degradation, e.g. Steinbrück et al., 2010, Tóth et al., 2010, in-vessel (RPV) retention and cooling of corium (Bechta et al., 2001, Kang et al., 2006), vapour explosions (Kim et al., 2010, Magallon and Huhtiniemi, 2001) ex-vessel corium spreading (Cognet et al., 2001, Journeau et al., 2003) and corium/concrete interactions and coolability (Journeau et al., 2009, Lomperski and Farmer, 2007). This paper considers management of the corium waste rather than accident phenomena and progression.

Loss of instrumentation and hydrogen explosions have obscured the extent of core damage. Even now, high radiation fields and contamination limit our ability to inspect and characterise the reactor cores and corium debris. Severe accident codes, e.g. MELCOR and MAAP, have used available data to produce the core damage estimates shown in Table 1 (JAEA, 2014).

The three BWR units contained a total of 1496 fuel assemblies, with 32 of them MOX in unit 3. Each assembly has 60 zirconium alloy-clad fuel rods. The fuel burnup histories vary between reactors, the original location of fuel assemblies in the core and location along the fuel rods (due to differing fuel loading patterns). The 1F reactors discharge fuel after 4 cycles with a burnup of 39.5 MW d kg⁻¹ (GW d/t_HM) at which point, the remaining three quarters of the load would have burnups of about 10, 20 and 30 MW d kg⁻¹ (NEI, 2012).

Whilst there is yet no direct evidence that corium reached the primary containment of any of the three reactor units, as illustrated in Fig. 2 (TEPCO, 2014), accident progression simulations generated by reactor severe accident codes clearly indicate corium breach for at least unit one (Yamanaka et al., 2014, Gauntt et al., 2012). The extent to which melt through has occurred is, however, unknown and probably varies significantly between reactors. In addition, there is considerable uncertainty in what little data is available, for example a recent press release has suggested possibly more melt through in unit 3 than had been previously reported (TEPCO, 2014). More recently, preliminary data from a cosmic-ray muon radiography installation at unit one suggests that most or all of the core has melted and relocated (IRID, 2015). Though this measurement technique has low spatial resolution, it can remotely map the disposition of reactor internals using the density difference between reactor fuel and structural materials (Miyadera et al., 2013, Takamatsu et al., 2015).