As part of a planned cross-cutting intervention relating to NNB Generation Company Ltd’s (NNB GenCo) safety functional requirement notes (SFRN), a team of seven Office for Nuclear Regulation (ONR) specialist inspectors examined (by sampling) the adequacy of the substantiation of the safety classification results presented in the SFRNs at NNB GenCo’s London office, the Qube. SFRNs are key design documents, providing the classification of safety features and will be part of the substantiation process supporting the next revision of the Hinkley Point C pre-construction safety report (PCSR).
A similar cross-cutting intervention was carried out in August 2014 that focused on the substantiation of the balance of nuclear island SFRN. This concluded that: NNB GenCo did not provide adequate evidence of sufficient substantiation of the safety classification; timescales for production of additional justification were not in line with ONR’s expectation; and insufficient evidence of adequate substantiation was considered a risk to ONR consenting to the start of construction (first nuclear safety concrete). A level 3 ONR issue was therefore raised, number 2703, and the need for a repeat intervention was recommended.
The purpose of this intervention was to:
ONR specialist inspectors considered the adequacy of the substantiation through review of relevant NNB GenCo and Responsible Designer documentation and discussion with NNB GenCo. This intervention is part of a planned series of interventions outlined in the ONR project task sheet for safety categorisation and classification.
The adequacy of the substantiation of the safety classification will inform ONR’s judgement on NNB GenCo’s progress with preparations for proceeding with construction of Hinkley Point C.
Not applicable as a safety system inspection was not carried out.
For the systems and safety features sampled, no major concerns were identified regarding the resulting class of the safety features. It should be noted that a more detailed assessment of the classification is planned once the RC1.1 SFRNs and SCCLs have been produced and shared with ONR.
The traceability of the substantiation of the classification of safety features (as presented in the SFRNs) to the supporting documentation was not always clear or explicit. It was harder still to identify a clear documented basis for the categorisation of the safety functions. Notwithstanding the limited visibility, NNB GenCo was able to provide evidence to support the substantiation of the SFRNs.
There were a significant number of safety features, and it was not always clear why there were significantly more for a given system at Hinkley Point C than for Flamanville 3. Most additional safety features appeared to relate to monitoring functions, which are not identified as explicit safety features for Flamanville 3.
In terms of this intervention, the spreadsheet developed by NNB GenCo, for the purpose of assisting the efficiency of the intervention, helped identify the relevant substantiation. However, given the effort involved in providing this visibility, for the six systems that were within scope of this intervention, a significant amount of resource would be required to cover all systems.
NNB GenCo provided acceptable evidence of substantiation of support safety features. However, this was limited to events within the design basis and it records the status at the end of 2012. For other events I consider the document needs expanding or an alternative document is required, and it needs bringing up-to-date for all safety features. NNB GenCo had also recognised this and is planning improvements in this area. Furthermore, I consider it would be beneficial if a similar document was produced for control and instrumentation systems. NNB GenCo is planning to produce additional control and instrumentation requirements documents that it claims should address this point. I will require oversight of these deliverables to ascertain if they meet ONR’s expectations in this area.
Given the complexity of the classification approach, I advised NNB GenCo that it should consider commissioning an independent review of its classification approach against that implemented at Sizewell B, which appears significantly less complex, and that someone with Sizewell B experience should carry out this review. I also advised NNB GenCo to consider the long term position in that maintaining such a complex classification system could be a significant burden. Notwithstanding this, it was clear from the intervention that the approach adopted is wider than just for classification. The SFRNs, safety features and component classification are embedded into the wider design processes and they play a key role in identifying interfaces and claims on systems. It would therefore be extremely difficult for NNB GenCo to move away from its current approach. These wider aspects are not clear in the documentation and I advised NNB GenCo that this should be emphasised.
A number of detailed points were raised during the intervention that will be followed up by relevant ONR specialist inspectors during future regulatory interactions.
Timescales for the adequate completion of the supporting analysis remain unclear, although it was noted that the Responsible Designer is currently revising the classification design quality plan. This is anticipated to be issued in quarter 1 2016.
Regarding concerns raised at the previous intervention, NNB GenCo demonstrated it had a better understanding of safety categorisation and classification, and better evidence of substantiation was provided, although visibility of the substantiation remains an issue. No concerns regarding NNB GenCo’s capability were raised as part of this intervention. The risk of ONR not consenting to the start of construction (first nuclear safety concrete) due to the adequacy of the classification has reduced as a result of the intervention.
This was a useful and timely intervention. It was very well supported and contributed to by NNB GenCo and the Responsible Designer, including Areva.
Although there remains some uncertainty regarding the adequacy of the substantiation prior to the second consent, and this will remain until ONR level 3 issue 2703 has been adequately addressed, NNB GenCo demonstrated evidence of substantiation and that progress has been made. However, its visibility needs to be improved.
No new issues, above the extant level 3 issue, were raised as a result of this intervention. The areas highlighted in this intervention record are adequately captured by this existing issue and will be taken forward with NNB GenCo through the regular classification level 4 meetings.