Reports from the National Audit Office (NAO) and the House of Commons Public Accounts Committee (PAC) reveal that recent cost overruns in the nuclear weapons programme have the same underlying causes as repeated Ministry of Defence management failures dating back to the 1980s.
The reports look in detail at three case studies from the MOD’s many infrastructure upgrade projects: the primary build facility at Barrow-in-Furness, the core production capability facilities at Raynesway and project MENSA at AWE Burghfield. The NAO report was published in January, and the PAC report, which builds upon the NAO report was published in May.
Much of the press attention has understandably focussed on the scale of the cost overruns, which amounted to £1.35bn across the three projects. Perhaps more significant is the underlying message of the reports that the MOD has failed to learn the lessons of the past, and is repeating the same mistakes made during the transition from the Polaris missile nuclear weapons system to the current Trident fleet: beginning construction before designs are complete, giving insufficient oversight to contractors, and commercial arrangements where the MOD assumes too much risk and contractors are not properly incentivised to deliver.
The sites
The Primary Build Facility at Barrow-in-Furness will be used to manufacture components for, and construct, the PWR3 reactors for the Dreadnought submarines. The MoD are funding a £1.1bn site-wide redevelopment at the site, which is owned and run by BAE Systems. The two buildings that make up the Primary Build Facility are both two years behind schedule. The project was forecast to cost £111m when it was approved in 2012. At the time of the NAO report it was expected to cost £240m, which included £11m set aside to deal with potentially hazardous objects discovered in the ground during building work. However, during oral evidence to the PAC in March Stephen Lovegrove, the Permanent Secretary at the MoD, revealed that the sum set aside had increased to £30m.
The core production capability facilities at Rolls-Royce Raynesway in Derby is to manufacture nuclear reactor cores and fuel for the PWR3 reactors that will power the Dreadnought submarine fleet. The core manufacturing building has been built, and the process of installing machinery and equipment is expected to complete in December 2020. Construction of the fuel manufacturing building has yet to begin. The buildings were originally expected to be completed by May 2021, but at the time of the report the expected completion date was June 2026. Figures for the joint cost of the two buildings are not given in the report, but the overall budget of the core production capability project has risen from an expected £1.2bn in 2012 to over £1.8bn.
MENSA is the project at AWE Burghfield to construct a new warhead assembly facility. The project was approved in 2011, and was originally expected to be completed in 2017 and cost £734m. The current projected completion date for the project is 2023, and the budget has risen to £1.8bn. Construction work has now been completed and equipment and machinery is being installed in the facility. Due to the size of the overspend, project MENSA was ‘re-set’ in 2016 while the MoD considered their options, and the contract governing the project was revised. The 'Gravel Gertie' complex, where warheads are currently assembled and disassembled only has regulatory approval to be used until 2026. A recent FOI release states that AWE are confident of being able to bring MENSA into service before then.
Construction on both the core production capability facilities and MENSA were begun before designs for the buildings were finalised. The first phase of constructing the core manufacturing building at Raynesway was begun before the MoD had decided how the PWR3 reactor cores would be built, resulting in an additional £146m cost, including the cost of constructing an additional building. In July 2018 Stephen Lovegrove, the permanent secretary to the MoD, said that work on MENSA had begun when the design was only 10 or 20% settled. The NAO estimates that this resulted in a cost increase of at least £399m.
Another recurring problem identified in the reports is contracts where risks are not shared between the MoD and contractors, meaning that the contractors were not incentivised to control costs. The contract governing the Primary Build Facility at Barrow is run on an ‘ascertained cost’ basis, meaning that the risks of any cost increases fall on the MoD rather than BAE Systems. BAE Systems is also not liable for any damages relating to non-performance. Due to these contractual arrangements BAE Systems have been paid an additional £108m due to the design immaturity issue and £11m due to unforeseen events. The contract also pays BAE Systems ‘management fees’ at a proportion of the overall project cost, meaning that the problems have garnered BAE Systems an additional £10m in management fees. The MoD has tried to move away from contracts of this type, but Rolls-Royce have insisted on keeping their contract, which runs until 2056, running on an ‘ascertained cost’ basis.
The reports also raise questions regarding the MoD’s capacity to manage these projects, in part due to a shortage of staff. The MoD are struggling to fill the team which oversees the Rolls-Royce contract, with two positions in a team of 18 being vacant in June 2019. The Submarine Delivery Agency, which is responsible for the submarine programme within the MoD, was short of 250 staff in March 2020.
Will the same mistakes be made again?
Towards the end of their report, the NAO list examples of similar failures occurring in previous projects in the nuclear weapons programme. Issues around limited sharing of risks with contractors occurred in the 1987 refurbishment of the Atomic Weapons Research Establishment (AWRE, now known as AWE), and the 1994 Trident works programme. Both of these projects also suffered due to construction projects being started before plans were finalised, causing significant cost overruns. The AWRE refurbishment also suffered from oversight failures and a lack of an overall management plan. Similarly, the 2002 Devonport facilities upgrade saw the MoD needing to change its original 'hands off' approach in the face of spiralling costs.
The PAC report states that the MoD “cannot explain why its leadership has not learned from these experiences”. Given the tight timetables of myriad interlocking projects involved in upgrading the UK’s nuclear weapons, the prospect of the MoD being able to deliver the upgrades as planned appears slim.