It now appears that spent fuel may be stored in interim dry stores for longer than was expected. The safety analysis requires knowledge of the characteristics and behaviour of materials under these storage conditions. Despite much R&D, questions about material properties remain open. Two key issues are residual water after drying and subsequent radiolytic generation, and the possibility of cladding embrittlement caused by thermal transients during the cask drying process and its impact on the cooling of stored fuel. By Maurice Dallongeville, Hervé Issard, Elisa Leoni and Aravinda Zeachandirin
More and more utilities are choosing to move spent fuel from pools to dry storage systems.
Safety analysis of long-stored and cooled spent fuel requires understanding of the characteristics and behaviour of materials under these storage conditions. The main R&D gaps, identified by the US Nuclear Regulatory Commission (ML120580143) and the Pacific Northwest National Laboratory (PNNL 21360), concern canister integrity, physical and chemical behaviour of fuel rods, and corrosion products under irradiation and ageing.
Addressing the main technical data gaps in extending dry storage beyond the initial forty-year period will provide the necessary arguments to justify licence extensions. This paper covers two important subjects of direct concern to AREVA TN: gas and hydrogen build-up from residual water, through corrosion and radiolysis; and fuel integrity, especially hydride effects on high burn-up fuel rods. These phenomena are connected and mutually dependent: high temperatures are needed to improve drying, while excessive heating may modify the cladding microstructure (for example, recovery of irradiation damage and hydride reorientation). One of the limiting factors in reaching higher burn-up is corrosion and the accumulation of hydrides in fuel cladding.
Residual water and drying
The process of loading spent fuel consists of submerging the transport cask in the cask pit or the storage fuel pool while spent fuel is transferred into the cask basket. The water is drained and the cask is dried to eliminate residual amounts of water in the cavity and in damaged rods inside the fuel assemblies.
The presence of water during transport and storage would increase the risk of potential oxidants and the subsequent degradation of the canister system internals. Radiolysis or corrosion may cause an unsafe accumulation of hydrogen during the storage period.
Each cask vendor is then required by nuclear safety authorities to guarantee that an adequate dryness has been achieved. A detailed description of different drying techniques and the dryness criteria are out of the scope of this paper. In transport casks, hydrogen accumulation is prevented by the short transport duration. In long-term storage, an accurate analysis of residual water is necessary to ensure that the packages do not accumulate an unsafe concentration of water, hydrogen or oxidant gases.
Drying techniques are usually effective in removing free water and also ensure effective removal of physisorbed water if sufficient heat is provided. In contrast, chemisorbed water is difficult to remove. Chemisorbed water compounds (the chemical forms hydrates of different oxides and several metal hydroxyls) may be formed as a result of water reacting with fuel pellets (due to cladding failures) or water corrosion of the cladding or cask materials. Four categories of hydrated compounds are found: uranium oxide hydrates, zirconium oxide hydrates, corrosion products (for example, hydrates of aluminium) and crud compounds (for example, hydrates of iron).
It is generally accepted that an unknown amount of water can remain in a spent fuel cask even after effective drying. The hydrated compounds (as hydroxides or oxyhydroxides) in the native oxide or corrosion products may be decomposed to produce water during transport and storage conditions, increasing the risk of corrosion and hydrogen build-up by radiolysis. So it is assumed that when these compounds may release water at normal storage temperatures, they need to be dehydrated as a preliminary step. This is not necessary when the dehydration is at temperatures which are higher than design storage temperatures. An analysis of water desorption kinetics is necessary and the stability under irradiation of the fully dried solids, hydroxides and oxydroxides potentially present in transport and storage casks must be proved.
In a reactor, a thick porous ZrO2 corrosion layer is formed on the fuel cladding. Water may be trapped in the pores and hydrolysis of zirconium oxides will generate hydroxides of Zr that include ZrO2, xH2O Zr(OH)4 and ZrO(OH)2. Drying techniques must remove all the water trapped in the pores of the ZrO2 corrosion layer.
ZrO2 powder has been the main focus of water desorption kinetics studies. Very limited data exist on the stability of Zr oxides, hydroxides and hydrates formed under storage conditions, and in particular the ZrO2 corrosion layers formed in nuclear plants. N.G. Petrick, among others, has shown that bound water desorbs in three steps as function of temperature (‘Interfacial Energy Transfer during Gamma Radiolysis of Water on the Surface of ZrO2 and Some Other Oxides’, J. Phys. Chem. B 105(25), 5935-5944 (2001)). Moreover, in the same paper, Petrick also shows a correlation between radiolytic hydrogen generation and the different forms of absorbed water.
Experiments have been carried out to evaluate the desorption kinetics of water from a 25 – 28µm thick ZrO2 layer formed on a Zy-4 (zircaloy) fuel cladding. Preliminary results (Fig. 3) suggest that most of the H2O contained into the ZrO2 layer desorbs at low temperatures (100°C < T < 300°C), and a smaller amount desorbs at higher temperatures.
These results seem to confirm the desorption kinetics observed in Petrick’s work above; detailed results will be published elsewhere. It is interesting to observe that a small quantity of hydrogen desorbs at high temperatures (T > 450°C). It likely comes from the oxide layer. Desorption experiments at constant temperature show that a thermal treatment of about 600 minutes at 150°C is sufficient to eliminate most of the water contained into the ZrO2 layer, as shown in Fig. 4.
Mechanical behaviour of fuel rods
Initially, the damage incurred by spent fuel assembly metallic components, and especially fuel rods, is mainly in-reactor creep, outer corrosion, and through-thickness hydriding of the zirconium-based cladding material. Methodologies based on analytical and numerical models are under development to predict fuel behaviour in transport conditions and ensure its retrievability during or after the dry storage period. This is closely linked to cladding integrity.
One of the main difficulties is characterising the mechanical properties of the irradiated cladding, which may change due to the thermal transients during the cask drying process and the cooling of stored fuel. Another is determining the hazardous phenomena arising during dry storage and transport.
In a first approximation, fuel rod parameters are: geometry, cladding material, type of fuel assembly and reactor design, axial profile of temperature and outer corrosion, fuel rod cooling rate in various circumstances, internal gas pressure and cladding circumferential stress, and fuel rod irradiation parameters including burn-up, fast fluence, and hydrogen concentration (linked to the burn-up).
For fuel rods to be transported after discharge and wet storage (but not after dry storage), elementary mechanical thresholds, defined for a given cladding material, are:
- A limit of -40°C in temperature, which is the lowest value in transport and storage conditions that can possibly be required by safety authorities in the frame of the current regulations.
- A room temperature threshold (20°C), which is a common value in performing cladding ring rupture tests, used to characterise the embrittlement of irradiated claddings (consequently these tests, although conservative in temperature, do not fully cover the domain between -40°C and +20°C).
- The brittle-to-ductile transition depending on hydride orientation in used fuel (the threshold is generally found between about 20°C and 200°C), and below which the brittle fracture risk of irradiated claddings increases.
- The dissolution of hydrides during heating for a given hydrogen concentration (for example: transition towards 395°C for 200ppm of hydrogen in stress-relieved annealed Zy-4 cladding) which depends on the burn-up.
- The precipitation of hydrides during cooling for a given hydrogen concentration (for example: transition towards 330°C for 200ppm of hydrogen in SRA Zy-4 cladding), which depends on the burn-up.
- The beginning of metallurgical transformation of claddings (around 400-420°C) and the recovery of irradiation damage which starts in the same temperature range (350-420°C).
- The thermal creep rupture risk limit for given duration and thermal history, for which the acceptable maximum stress varies roughly inversely with temperature.
For fuel rods to be transported after dry storage, additional considerations are mainly the thresholds of hydride reorientation (defined separately in circumferential stress and temperature for a given cladding material), increasing the brittle fracture risks, or its consequences (ductility degradation). The reason for defining two types of thresholds is that embrittlement does not depend only on radial hydride content, but also on the total hydrogen content in the cladding and on the length and continuity of hydrides, which in turn depends on thermodynamic conditions during storage and cooling.
These thresholds depend on the cladding material and on the irradiation history. The respective thresholds for hydride reorientation compared with those for ductility degradation are different for PWR and BWR cladding materials.
In practice, the temperature recovery effects of irradiation damage and the absence of hydride reorientation with fast-enough cooling should leave fuel rod claddings with enough ductility for transport and handling.
Another type of risk to consider is delayed hydride cracking, a phenomenon sometimes arising in Candu and BWR fuel. This phenomenon requires a stress concentration and temperature allowing for a solid solution of hydrogen. When the stress intensity factor threshold is reached, a regime of stable crack growth settles and is characterised by hydrogen diffusion in the stress concentration zone; precipitation and rupture of hydrides occurs as soon as crack length becomes critical. This unlikely in light water reactor fuel assembly transport.
In normal conditions of transport (0.3m drop), the iodine-induced stress corrosion cracking (SCC) mechanism is unlikely because of the low level of fuel cladding stress. In accident conditions (9m drop), the fuel cladding stress level is higher but only for very short durations. Release seems unlikely.
Risk management principles
The main parameters of spent fuel assemblies being transported or stored in a canister are evolutions of the temperature and circumferential stress of fuel rods. Consequently, despite the numerous phenomena to be accounted for, it should be possible to analyse them simultaneously to develop risk management principles and allowable margins in temperature and stress.
As an example, the history of a fuel rod from a spent fuel assembly depends strongly on the coupling between internal gas pressure and average cladding temperature. The internal pressure at room temperature serves as a reference to quantify the amount of gas (initial filling of helium and fission gas released in service), which does not evolve later in the first approximation for current used fuel transport and storage. At thermal equilibrium of the cask cavity, the internal pressure results mainly (in the absence of significant cladding thermal creep) from the ideal gas law in internal free volumes, considering the previous reference pressure at RT and the average cladding temperature weighted with axial gas-volume distribution.
The circumferential cladding stress coming from this high internal temperature is in practice directly proportional to it because the cladding thickness-to-diameter ratio of fuel rods varies only marginally between PWR or BWR and the lattice array. Consequently, used fuel rod cladding temperature / average stress evolution is linear in a first approximation. The potential rupture risks of spent fuel rods can be appreciated by comparing the temperature and average stress evolution with known damage or rupture risk limits, expressed also in temperature and stress, and possibly depending on the cooling rate.
Transport of spent fuel
AREVA TN and International Nuclear Services started the Fuel Integrity Project (FIP) in early 2000s, aiming to develop a methodology to evaluate, as a safety requirement, the nature and the extent of LWR fuel assembly damage when a package is accidentally dropped.
AREVA TN has experience acquired from drop tests of fresh fuel assemblies. Now a mechanical testing programme, including fresh and used fuel rod samples, has been planned and carried out by the two companies. As a result, the FIP methodology, developed from test results analysis and structured in flow charts, gives guidelines in studying the effects of accidentally dropping packaging loaded with fresh or used LWR fuel assemblies.
Applying the FIP methodology leads to the following output, used as criticality hypotheses for the safety analysis: whether fuel rods have ruptured, their number, their location, the associated amount of fuel material released, and the extent to which the fuel array suffered deformation and axial sliding. The FIP methodology allows a first evaluation of the main damage and rupture risks of spent LWR fuel rods in transport after discharge and wet storage.
Complementing the FIP methodology are further developments to integrate the possible embrittlement of used fuel rods. An initial literature review by AREVA TN indicates that, for transport after discharge, the main additional risk for used fuel rods without defects (coming from in-reactor irradiation) is probably linked to the presence of radial hydrides. Risks of DHC and SCC seem unlikely, due to the very short duration of possible high stresses in the cladding during accidental drops.
The next step is defining a domain of ductility for the dry transport of used fuel assemblies from French PWRs for all used cladding materials. A further step will widen this analysis to spent fuel assemblies from other reactors.
Rods without defects transported after long-term dry storage potentially have higher damage or risks than rods transported directly after discharge and wet storage. These additional studies have yet to consider the possibility of safe transport of used fuel rods after long-term storage. Nevertheless, the approach could still be based on the analysis of cladding stress and temperature histories, which involve various phenomena like maximum temperature during drying, cooling rate and irradiation damage recovery.
The implications of extended dry storage
Improving understanding of the characteristics and behaviour of long-stored and slowly-cooled spent fuel requires analytical tests, which are easy to manage and have piecemeal answers for the short term questions (that is, in the frame of transport or short-term storage). At the same time it requires integrated long-term realistic tests in representative long-term storage conditions of fuel and casks.
Verifications and analysis are carried out in international and national research programmes, as shown by Donghak Kook et al (‘Review of Spent Fuel Integrity Evaluation for Dry Storage’, NEAT, 45 (1) (2013).
Extension of interim dry storage of used fuel requires safety analysis, and it is necessary to characterise the irradiated cladding mechanical properties that may change due to thermal transients during the cask drying process and the cooling of stored fuel.
AREVA TN has developed contributions for the better understanding of the evolution of used fuel with the issue of residual water after drying possibly leading to subsequent radiolytic generation and the potential transition from ductile to brittle cladding behaviour.
In a dry storage system, the fuel surfaces and the attached corrosion products (hydroxides and oxyhydroxides) formed on metallic surfaces are subject to gamma radiation. R&D programmes are under way to explore their responses and the subsequent radiolytic hydrogen generation.
The risks of damage in transporting fuel rods after discharge and wet storage are now well understood and predictable, but still need to include the hypothetical embrittlement of cladding. After dry storage, additional risks include cladding embrittlement induced by hydride reorientation. For extended dry storage, the consideration of the additional risk of further helium or hydrogen release from materials seems unjustified.
Several national R&D programmes and international teams such as those developed by the US Department of Energy (High Burn-Up Dry Storage Cask Research And Development Programme) and the Electric Power Research Institute (Extended Storage Collaboration Program) are addressing these gaps. However, these are necessarily long-term tests and are still far from bringing validated responses. Fuel storage must continue to be managed very prudently to be able to face any unexpected fuel reconfiguration.
About the authors
Maurice Dallongeville (firstname.lastname@example.org), Hervé Issard and Aravinda Zeachandirin, AREVA TN, and Elisa Leoni, AREVA NP E&P, both at 1 rue des hérons, 78180 Montigny-le-Bretonneux, France
A version of this paper was presented as ‘Fuel behaviour in transport after dry storage, a key issue for the management of used nuclear fuel : an industry view,’ at Top Fuel 2013, Charlotte, North Carolina, September 15-19, Article 8575, 557-563 (2013).