Office for Nuclear Regulation

Barrow – Inspection ID: 22-013

  • Site: Barrow
  • IR number: 22-013
  • Date: April/May 2022
  • LC numbers: 10, 23, 24

Executive summary

Purpose of Intervention

This inspection focussed on BAE Systems Marine Ltd (BAESML) arrangements for compliance with Licence Condition (LC) 10 (Training), LC23 (Operating rules) and LC24 (Operating instructions). This inspection formed part of ONR’s Integrated Intervention Strategy for 2022/23.  The Defence Nuclear Safety Regulator (DNSR) also participated in the inspection.

The purpose of this inspection was to gain confidence that BAESML’s arrangements for LC10, 23 and 24 meet regulatory expectations and that they are adequately implemented for Phase 7 activities (critical operations) for Astute boat 5 (Anson).

Interventions Carried Out by ONR

During the inspection, we assessed the adequacy of BAESML arrangements for LC10, LC23 and LC24 against ONR guidance and assessed the adequacy of their implementation against a sample of faults from the critical operations safety case.  To achieve this we:

  • Explored how conditions and limits necessary in the interests of safety are defined and implemented into training and Operating Instructions (OIs).
  • Explored the adequacy of training in providing the necessary competence to undertake critical operations.
  • Explored the adequacy of OIs in providing adequate task support.
  • Walked through a sample of safety important activities to ensure that conditions and limits are achievable and that the training and OIs provide adequate task support.

Explanation of Judgement if Safety System Not Judged to be Adequate


Key Findings, Inspector’s Opinions and Reasons for Judgements Made

Against our sample of faults, we judged that, in most cases, appropriate conditions and limits were identified in the safety case.  We did not identify any instances where we considered nuclear safety to be compromised.  However, we did find deficiencies against LC 23(1), including instances where conditions and limits that appeared to be required were not designated; instances where conditions and limits were confusing, introducing the potential for breaches; and instances where it was difficult to establish the basis for conditions and limits from the safety case provided.  Further, we found shortfalls with the presentation of information in the fault schedule, which we found difficult to interpret, as it did not identify the specific operational conditions and limits relevant against each fault.

BAESML continue to improve their use of Human Factors (HF) to support the production and implementation of the safety case, which we consider to be in line with relevant good practice (RGP).  However, we found significant deficiencies in the scope and application of the HF work completed in support of the critical operations case.  This requires improvement as BAESML continue to apply the practice in future safety case updates.

For LC10, we gained confidence that BAESML staff, who hold nuclear safety important roles, have adequate role profiles, which identify the required competencies (e.g. being a commissioning engineer).  For critical operations, BAESML, Royal Navy and Rolls Royce personnel will work collaboratively.  We saw evidence of activity specific training, which all relevant personnel will attend.  This included training on the scope and purpose of the critical operations and training on the safety case.  Further, BAESML explained that task-based training will be provided to individuals prior to tasks commencing.  None of the task-based training material was available for inspection.

BAESML arrangements identify a systematic approach to training (SAT), which is in line with RGP.  However, we found deficiencies in the application of the approach, particularly the analysis phase.  We considered that these gaps increase the potential for inadequate training needs being identified for specific activities.  We also identified a concern around capacity within the central training team to apply the SAT approach.

Linked to this, we found that the approach to training for critical operations had evolved organically.  BAESML were unable to demonstrate that a systematic approach had been applied to the identification of training needs, which made it difficult to establish if the approach being taken by the Test & Commissioning (T&C) function was adequate or complete.

For LC24, we found that OIs associated with activities on the submarine are approved via the Reactor Test Group (RTG), whereas OIs associated with activities on the dock side, are approved via the Docking and Nuclear Services Authorisation Group (DNSAG).

We found that the thread from the case to specific OIs was adequate for RTG and DNSAG OIs.  For the RTG OIs, we found that the conditions and limits were appropriately implemented and found that the general quality of OIs to be acceptable.  However, for the DNSAG procedures, we initially found instances where draft OIs incorrectly or incompletely implemented the safety case requirements.  Following the initial inspection, BAESML revised these OIs.  We found the revised OIs adequately implemented the safety case requirements and that the OIs were appropriate.

Across the RTG and DNSAG arrangements, we found instances where the guidance for writing OIs did not fully meet RGP for procedure writing.

Conclusion of Intervention

We considered that the issues identified with the derivation of conditions and limits and the HF work to be significant, but without prejudice to overall nuclear safety, and as such rate LC23 Amber.  BAESML already have a Level 3 Regulatory Issue associated with LC23 and are implementing a plan to address identified shortfalls.  I will work with BAESML to ensure that the scope of this extant issue adequately covers the findings of this inspection.

For LC10, we judged that the provision of training for undertaking critical operations to be broadly adequate.  However, we considered that the gap in arrangements and the lack of structured analysis to underpin the activity-specific training to be significant.  As such LC 10 is rated Amber.  A level 3 Regulatory Issue will be raised, targeted at improving capacity within central training team.  The scope and actions of this will be agreed with the Nominated Site Inspector and BAESML.

For LC24, I considered that the quality of the OIs was adequate and so rate LC24 as Green.  However, given findings about meeting RGP, we identified an observation for BAESML to undertake benchmarking against other licensee organisations to improve their arrangements.