2. At the heart of safety: high standards and leadership

In addition to difficult situations arising from plant (see Chapter 5) (see Chapter 9) the main safety risks include fire, natural hazards (see Chapter 6) and misinterpretation of plant status by the operator (see Three Mile Island).

While remaining humble and remembering that this indicator is far from reflecting the full picture, we can still be proud of the good results achieved for automatic reactor trips in France: with 15 unplanned automatic trips, as 2022, the second-best result in the history of the French fleet (14 in 2020), the top three results having been seen over the last four years. In the wake of the pandemic and the stress corrosion cracking (SCC) issue, this progressive return to normal availability is welcome news, for both electricity production and nuclear safety.

In the UK, 2023 was spent consolidating performance, with acceptable indicators, and for the second year running, generation was not impacted by any significant technical issues. The successful defuelling of the first unit at Hunterston B is worth highlighting.

The spirit of openness between the two fleets, towards the World Association of Nuclear Operators (WANO) and internationally, is also to be commended in both countries.

A major multi-year industrial-scale programme to resolve SCC in pressurised water reactors (PWR) has been initiated, with significant resources allocated for inspections, repairs and welding. The choice of replacing the safety injection system pipes in the 1300 MWe P’4 plants as a preventive measure is yielding results. Schedules and worksites are easier to manage this way, rather than carrying out ‘condition- based’ repairs, with radiation exposure levels being controlled. No SCC was found at Sizewell B during inspections in 2023.

SCC must be handled carefully because it is a complex, multi-factorial problem. This is illustrated by the deep cracks in some welds that were detected in early 2023. These welds had not been considered among the most at risk, but they had been repaired during the construction phase. Expert analysis to better understand and quantify the parameters involved is still ongoing. It is necessary to accelerate the installation of thermocouples on certain pipes, which has already been done at Sizewell B and is planned to take place in France.

This issue has been managed with an outstanding sense of accountability and commitment. However, this multifaceted organisation can be a source of complexity. In particular, EDF experts must be fully involved in the preparation of advice and strategies; they should not merely be seen as internal suppliers invited to answer individual questions to compile safety cases for the ASN, sometimes within overly demanding timescales. In keeping with the Group’s spirit of inclusion, Framatome’s experts could also have been better included in these studies.

When the time is right, an analysis of the practical experience gained will be useful. This issue is a reminder of the relevance of conducting wide- spectrum inspections which are conducive to detecting unforeseen defects (see Chapter 5), and to consider new methods of repair that have already been approved in other countries, e.g. overlay repairs, clamps, etc.

Throughout the year, I found the plants that we visited to be in a satisfactory condition, with an improving trend.

Progress was visible in the materiel condition of the pumping stations that we visited in both fleets. The action plans targeted to fight corrosion are producing positive results. Efforts must be continued as the ideal conditions have yet to be achieved, and the fight against corrosion for coastal plants is never-ending.

The attention given to on-site power supplies is even more crucial since the grids may prove to be more vulnerable in the future (see Chapter 9). For Nuclear Operations, the situation is good at Sizewell B, being equipped with four diesel generators. However, for certain AGR sites equipped with gas turbines, the WANO Safety System Performance (SPP5) indicator for emergency generation used to be sub-standard, and the initial positive effects of a fleet intervention are now visible. In France, diesel generators are not always operated in line with standard diesel industry practices (see report 2021), as a result of failing to give the issue the priority it deserves.

Difficulties in performing work remain, whether in terms of work permits, technical specifications, planning, spare parts, documentation, or technical capabilities. Bringing all these things together is an exhausting task, which affects efficiency and spanner time. The latter is considered to be one of the priority workstreams by the Group’s chairman.

Industrial asset maintenance is an area where sites are having difficulties, in particular with their plant defect backlogs (list of work requests). In France, although this figure has been falling for several years, it is still too high. While the backlogs appear to be decreasing in the UK, progress is slow as processes are not being effectively used, and equipment obsolescence remains problematic. Both fleets have ‘Cinderella’ equipment which, though not specifically safety-classified, contribute to nuclear safety. Budgetary constraints are a factor, and, in France, this is compounded by the burden of the fourth ten-yearly outage programmes (see Chapter 5).

Both fleets have difficulty in managing outage schedules. The START 2025 programme is beginning to produce some good results for the DPN, whilst the ‘Cornerstones’ initiative for Nuclear Operations has yet to fully demonstrate its effectiveness.

At the DPN, although the internalisation of some maintenance tasks remains inconsistent (see 2022 report), it is delivering real benefits: sites are now more capable of reacting rapidly to unforeseen events and ensuring technical supervision of contractors.

One of the challenges, for generation and nuclear safety, is regaining the sense of performance. There are examples where priority on generation has severely impacted nuclear safety, e.g., Davis-Besse, but the opposite is not a gauge of nuclear safety either. The effective management of activities, plant and risks is beneficial to both generation and nuclear safety.

The DPN’s move towards reinforcing the technical disciplines rather than focusing on processes is very positive. It is now necessary to aim for quick improvements.

While the engineering functions have some remarkable experts and are extremely active, they are still overloaded to the point where their ability to step back and assume a questioning attitude is challenged. Their desire to do ‘just what is strictly necessary’ is positive (see Chapter 5).

Some engineering units have cut back on immersive site familiarisation training for budgetary reasons. Field visits should be systematic, but this is not the case. This regularly leads to overly theoretical systems or modifications, blindly applying requirements without being connected to the reality of operations and without sufficient engineering judgement. I have seen absurd engineered structures designed to protect against high winds, or have heard an idea to install cooling units on the roof of buildings swept by sea spray.

Improving efficiency means resolving certain issues closer to the field, giving managers greater flexibility, and making decisions at the right level. The focus on field presence by the DPN and Nuclear Operations is essential. It needs to be considered as a normal aspect of team and activity management, and not just used as another means of reporting. This is why I am pleased to hear that the formal tracking of manager field visits has been replaced with discussions on in-field observations, followed up by concrete actions.

The senior managements of the DPN, the French regulator (ASN) and, in an increasing measure, the Institute for radiological protection and nuclear safety (IRSN) all agree on the principle of simplification, but its application within their organisations is slow and the ‘complexification machine’ always seems to be in operation.

Too many quality issues have led to a practice of reassurance through the continuous addition of requirements, authorisations, risk assessments, lines to instructions, and equipment and criteria to the general operating rules (RGE). It must be shown that improvements made to technical practices are sufficient and effective, whereas ‘proceduralising’ activities often only provides an illusion of control. Those in the field have said, “they’re doing their best to ensure we make a mistake” or “we don’t work anymore, we reassure”.

I am pleased that the DPN has decided to review the application of Directive DI 100 governing the declaration of significant events, in line with my recommendation of 2022. It relies on a better method of choosing what should be the subject of a significant nuclear safety investigation report (CRESS), and to focus on what really matters, including issues with competence and behaviour. I also appreciate UNIE’s efforts to spend less time on prescribing requirements and more time on providing support to the sites.

Simplification requires being more rigorous.

The RGE in France have grown significantly in volume and complexity over the years. The DPN opened a debate in 2023 to discuss the reformation of these rules so they more closely resemble those used internationally. I commend this approach, which is key to achieving simplified and safer operation of the plants. Sharing experience with Nuclear Operations could prove to be beneficial.

Without waiting for this reform to be implemented, I recommend that the current RGE be simplified in consultation with the ASN, and that we stop making things more complex in the first place. Are 16,000 safety criteria really necessary for periodic tests at Flamanville 3, including 160 per diesel generator? Nor do I believe that the management of external hazards requires the RASA to be included in the RGE (see Chapter 6).

The technical specifications (STE, Chapter 3 of the RGE) must be concise, state the essentials, and avoid trying to cover everything. It is not because something has an impact on nuclear safety or figures in the safety case, that it must be included in the STE or RGE, nor does an item need to be classified as safety-important equipment (EIP) because it may prove useful to nuclear safety. There are many things to know, to do, and to check in a nuclear power plant that are not destined to be included in the RGE, except to render them unusable and to lose sight of the key criteria and parameters that must be visible.

It would be worth questioning the rules governing the accumulation of group 2 action conditions. Developed to make sure the general level of nuclear safety does not deteriorate, they often prevent maintenance or equipment repair, which goes against the original objective.

In 2023, many dozens of controlled shutdowns still occurred because of a criterion in the RGE. While some shutdowns are necessary and indisputable, not all of them are. Sometimes they are accumulations that have no impact on safety, or minor deviations from criteria. Shutdown is not a trivial activity, and while doubt should benefit nuclear safety, a shutdown does not automatically assure this.

The RGE will never be able to explain everything about equipment availability. Though some criteria are not open to discussion (those to be incorporated into the RGE), engineering judgement, discernment and accountability remain essential: they must be given their rightful place.

Though safety is inherently reliant on demonstrations, the essential goal must be safety rather than its demonstration.

There are several reasons to avoid overusing these terms:

Independent nuclear safety oversight (INSO) teams are listened to by management and the technical disciplines with whom relations are effortless.

In the UK, the advice of Independent Nuclear Assurance (INA) is wide ranging, with the ensemble of the INA, the Nuclear Safety Group (NSG) and the Technical Safety & Security Manager (TSSM) covering a broad scope. However, the scope of INSO in France remains narrow and is evolving slowly (see 2022 report). Additionally, the nuclear safety engineers (IS) tend to interpret the RGE too literally, reading more into them than is actually written, thus being at risk of losing the original meaning. Their own judgement must also play a part.

Both fleets need to make efforts in the advanced planning of jobs and skills, including the promotion of subsequent career paths.

I believe that the combined peer reviews, which merge the Nuclear Inspectorate’s (IN) evaluations with WANO peer reviews, could be optimised, while maintaining the volume of in-field observations. Follow-up reviews should focus more on performance diagnosis to avoid leaving sites without assessment for four years. The Nuclear Inspectorate could supplement peer reviews with subject-based inspections.

In the UK, the INA at Nuclear Operations is independent in expressing its opinions and has adopted a more challenging approach. I appreciate the efforts made to ensure that weather-based hazards and cybersecurity risks are dealt with robustly.

Whilst the INSO teams remain too focused on checking processes and non-conformities, they are more and more involved in giving their advice in meetings, which is a positive step.

They are gradually expanding their scope of responsibility, for instance: the challenge of modifications at the DIPDE (In-Service Fleet Engineering Division), or the ‘Compass’ themes at the DIPNN (Engineering & New-Build Projects Directorate), but this is sometimes insufficient. I expect them to challenge the technical decisions when necessary and voice their opinion more often on the management of nuclear safety, nuclear safety culture, competences, and standards.

At Framatome, INSO is integrated into every unit. I met the INSO team at the Romans-sur-Isère site whose actions were robust and whose role greatly resembles that of the INSO teams at the plants. I appreciate the invested interest in skills shown by the INSO team at the Engineering & Technical Directorate (DTI), as well as their practice of conducting open-discussion interviews.

In the UK and France alike, relations between the Executive Directorates are healthy and dialogue is effective. Relations between the sites and their safety authority are robust. Relations between the ONR (Office for Nuclear Regulation) and EDF Energy are mature: the distribution of roles and the operator’s prime responsibility are well defined. In France, even though there is good general agreement between the national directorates, the practical application of this and management of priorities often still need to be developed.

Technical exchange seminars between engineers are a good practice and are currently being used by the EPR2 project and in the field of nuclear fuel. They should be made common practice to avoid exchanging information, essentially through questionnaires, and to avoid technical assessments behind closed doors. I also believe that, in return, experts should be given more access to the field so they can fully appreciate the reality.

The Operator (DPN, engineering, projects) must develop a nuclear safety strategy, demonstrate its convictions and back them up. Dare to think and say what you think! Leadership in nuclear safety must come, above all, from the Operator, who by assuming its prime responsibility, will be listened to all the more by the ASN. The safety options document for the fifth ten-yearly outages on the 900 MWe series (see Chapter 5) seems to be a well-thought-out strategy to be implemented across the board.