7. Nuclear fuel and reactivity control: the heart of nuclear safety

The place where fission occurs and the power density is concentrated, nuclear fuel represents the heart of a reactor. Its design, manufacturing and operation requires a high level of expertise.

In the UK, the Nuclear Safety Group (NSG) is strong and works closely with the control room teams. Refuelling at the Torness and Heysham 2 plants now happens with the reactors shut down and depressurised, instead of on-load, which changes how the core is controlled because its reactivity is higher after batch refuelling. Performance indicators and the committees for reactivity management have not changed for several years, and Nuclear Operations is reviewing the situation in search of greater effectiveness: I will be following this initiative closely as certain events require feedback.

Exploring avenues of progress must also be an engineering goal. Being able to cool the reactor after an automatic trip without risking it going critical again, even in the absence of a boron injection, would contribute to a higher level of intrinsic nuclear safety. In this context, the actual behaviour of the reactor could be assessed by removing conservatisms. Following the advances made in the last twenty years, it would also be pertinent to further explore methods to prevent the risk of injecting water plugs into the core, e.g., a slow start-up of the primary pumps.

I met with the Group’s main players in the nuclear fuel field: R&D, Technical Division (DT), Core Management & Engineering Group (GECC), Nuclear Fuel Division (DCN), Framatome, IECCs, NSG, and nuclear fuel and fuel route departments. Expertise, career development and networking remain robust. The responsibilities, training cycles, and qualifications are clear. We must ensure that the engineering overload does not hinder mobility to the detriment of integrated career paths. In view of the development of a UK pressurised water reactor (PWR) fleet, it would be beneficial to include Nuclear Operations staff in some French organisations and forums.

The teams recognise the professional level of relations and the quality of technical discussions with Framatome, as both a supplier and subsidiary. However, the contractual aspect sometimes slows down the flow of information, and trust can suffer as a result. I was able to visit the Framatome fuel engineering teams and those at the Romans- sur-Isère plant. The skills, product strategy and international scope impressed me. The investment allocated to upgrading the plants, which was much needed, is clearly visible.

The R&D site at Renardières has testing facilities that combine experimentation and modelling in a remarkable way. On a national level, strong public investment in test loops and reactors is essential, and the French Alternative Energies & Atomic Energy Commission (CEA) must ensure the scientific infrastructure upstream of the nuclear sector.

The nuclear fuel used in PWRs has never been more reliable than in 2023: only two unloaded assembly (AC) failures were detected, which is equivalent to a failure rate of 0.032%, a record performance. This historic performance is the result of product optimisation by Framatome and Westinghouse, and progress in their operation, including the prevention of foreign materials. In the UK, the PWR at Sizewell B has not detected any fuel failures for more than 15 years (in France, around a third of the reactors have been in this same situation for at least ten years).

Nuclear fuel has been affected by technical issues in the past: fretting, rapid development of pinholes (PPP), CRUDs, cladding corrosion, grid spring failure due to stress corrosion cracking (SCC), weld failure, fuel assembly deformation, rise in flux at the extremities of the MOX (mixture of uranium and plutonium oxides) fissile column, etc. Spread over time and without any serious consequence for nuclear safety, these issues have been gradually resolved. The most recent issue involved the corrosion of M5 alloy cladded fuel elements, which had affected foreign reactors and for which the operating experience was not sufficiently taken into consideration (see 2021 report). Although the finer details of the causes need to be understood, increasing the iron concentration seems to be a solution to the problem and is being applied to new fuel. Analysis confirmed that this corrosion causes very little hydriding, which embrittles the cladding, though its consequences on nuclear safety remain limited. Progress must also continue in incorporating operating experience from the Taishan EPR (see Chapter 4).

For 50 years, the general design of nuclear fuel is the same, and yet so much has changed:

As a result, technological developments have created real physical nuclear safety margins.

AGR fuel is manufactured at Springfield by Westinghouse. A contract ensures its procurement up to the end of operation of the last AGRs.

After recurrent fuel assembly failures, even requiring the power of one reactor to be limited, the issue seemed to have been resolved and no fuel assembly failures had been detected for three years. However, seven failures were detected in 2023, with five on the same site. Examinations to understand the cause will be performed at Sellafield once the decay heat of the fuel assemblies has decreased. In the meantime, the power of the relevant reactors will be gradually reduced at the end of each refuelling cycle.

The defuelling and refuelling activities at Torness and Heysham B have stabilised after considerable disruption to the organisation; they are now carried out every three to four months with the reactor shut down and depressurised instead of on-load.

The Group has taken the strategic step of aligning itself more closely with the products used internationally. Roadmaps include, for example, the adoption of intermediate mixing grids (GIM), as well as Framatome’s new-generation fuel, GAIA.

The use of GIM improves cooling by mixing the veins of water more efficiently over the length of the fuel rods and enhances nuclear safety. In the transitory situation, it will be necessary to justify operation with fuel assemblies with and without GIM. Certain conservative assumptions could be reassessed to substantiate the use of mixed cores considering the transitional nature of the situation in view of improving nuclear safety.

The GAIA skeleton boasts greater rigidity, having been designed to reduce fuel assembly deformation. Its stronger grids must prevent buckling under the heaviest loads, including extreme earthquake conditions. The friction springs are pressed into the metal strap rather than being welded, which eliminates spring failure phenomena. The lower end cap will provide better protection against foreign material and fretting, etc.

In the long-term view, the first enhanced accident-tolerant fuel (E-ATF) assemblies will be tested in the reactor. A chromium coating on the cladding is designed to make it more robust in accident conditions. Chromium could also be added to fuel pellets in order to better trap fission gases. However, its industrial suitability needs to be assessed, as it slows down recycling rates.

Defining fuel loading plans, conducting physical tests, interpreting neutron flux maps and calculating system protection parameters all rely on rigorous procedures, operational calculation tools and strong knowledge in reactor physics. In France, this is the role of the GECC within the Operations Engineering Unit (UNIE), together with the IECCs on sites. In the UK, the NSG site teams fulfil this role. The career paths and skills are strong, with clearly defined and recognised roles. The Technical Division (DT) and R&D support the GECC to develop generic tools and deal with unusual cases.

The GECC resources need to be carefully managed, as the definition and justification of fuel loading plans has become increasingly sophisticated. While these generally follow a standard plan, characteristic of the reactor series and its fuel (e.g. MOX), adaptations are often necessary. Several factors have led to the multiplication of bespoke solutions: technical difficulties, MOX delivery problems, the requirement for additional safety demonstrations from the VD4s, and changes in outage placements (inducing changes in the cycle duration) due to the Covid pandemic, to VD extensions and to stress corrosion cracking. The generic safety case for fuel load variations, which allows the loading of plus or minus four or eight new fuel assemblies, is, from this point of view, a success.

Safety margins in the core are carefully monitored. Some believe that conservatisms are “piling up” and “that only France takes into account pellet-clad interactions and water gaps”. These actual phenomena cannot be ignored unless it is agreed that they are covered by the margins. I believe it is now time to re-examine the chain of assumptions and conservatisms, with a view to re-legitimising the balance of safety justifications and sharing it between all stakeholders.

The fuel handling teams that we met at the sites are skilled. Initial training is strong at the Operations & Engineering Training Department

(UFPI), the CETIC, and the manufacturers. Mentoring is in place across the board. However, the skill pools are small, the current teams have no head-count margins, and the GPECs are stretched.

The fuel and maintenance departments remain siloed, and the quality of maintenance is inconsistent. Bringing some of the maintenance work in-house helps staff take better ownership of the equipment, maintains the level of competences, helps the supervision of contractors, and improves responsiveness to unexpected problems.

While incidents still occur, for which operating experience is shared, progress is visible: there were only two events in 2023 (damage to S holes, a historic all-time low), remote inspections are more efficient (but the deployment of HD cameras is slow), and fuel loading/unloading schedules are respected.

I noted that an operations team at Flamanville 3 has been responsible for the fuel handling and lifting equipment since the construction phase. They are familiarising themselves with the equipment and preparing for operations. The Hinkley Point C (HPC) project is following the same path and its team regularly visits the Flamanville site. I suggest that the EPR2 project seek inspiration from these good practices and broaden its scope to other systems.

On a national level, the maintenance programmes are now updated via meetings called ‘plateau flash’, bringing together the corporate services, engineering teams and manufacturers; this is a good practice. Obsolescence, however, is affecting many components and an inventory of these is currently being established, which is a welcome initiative. Although I met some very committed staff and progress has been made, problems resolved and reliability improved, it is nevertheless a case-by-case response. Resources are lacking in this area, and some fuel handling equipment still bears the brunt. It is necessary to invest more, develop a long-term vision and commit to a refurbishment plan.

Discussions have been initiated on the future of downstream fuel cycle facilities (La Hague and Melox plants), their upgrade, and their replacement. The need for this has been recognised at the highest levels of industry and the French government. In the long term, this shift confirms the strategy of closing the fuel cycle, which will also require a fast reactor development plan.

The assessment of storage requirements for spent fuel has been refined according to different scenarios, and several options are at the engineering study or industrialisation stage: densification of the La Hague pools, interim dry storage, EDF centralised pool, increasing the size of MOX fuel loads, ‘MOX-ing’ of additional reactors, etc.

In the UK, AGR spent fuel is stored at Sellafield, while PWR spent fuel from Sizewell B is stored onsite in its pond and in a dry storage facility, which will gradually achieve the expected transfer rates.