by James Jinks
Negotiating terms that governed the supply of the missiles and related equipment were relatively simple, although talks on aspects of the warhead and re-entry body were ‘made all the heavier because of the bevy of lawyers on the US side whose main aim is to protect themselves by reflecting all conceivable legislative requirements into the agreement’.197 However, in general, there were ‘no insuperable problems’.198 An agreement was quickly assembled consisting of fifteen articles and a number of confidential secret annexes setting out the terms, conditions and arrangements for the supply of Polaris missiles (minus the warhead) and equipment, and measures for governing development work carried out by the US on behalf of the UK. It granted the UK use of the Atlantic Missile Range at Cape Kennedy, now Cape Canaveral, for test launches, and set out a financial procedure for payments consisting of twenty-four equal instalments of $729,167 payable quarterly into a special Polaris Trust Fund.199 It also outlined the institutional links between the Polaris Executive and the SPO, principally a commitment to establish a Joint Steering Task Group (JSTG) to meet four times each year in London and Washington to monitor the programme. Significantly, the agreement made no provision for termination.200
During the negotiations the Navy argued that in order to build Polaris submarines it required information about the latest US nuclear reactors. But the final agreement did not ‘authorize the sale of, or transmittal of information concerning the nuclear propulsion plants of United States submarines’. The problem was Admiral Rickover. As we have seen, having fulfilled his obligations with respect to HMS Dreadnought, Rickover was unwilling to provide any further assistance to the British.201 Rickover also had a difficult relationship with the SPO. He had repeatedly tried and failed to expand his role in the US Polaris programme, but he had been deliberately shut out, confined to supplying the reactors for the US Navy’s Polaris submarines.202 The immediate impact of this restriction on the UK programme was limited. Although US technical information would have been useful, Mackenzie had already concluded that the British nuclear reactor based on the Dounreay prototype should power the Royal Navy’s Polaris submarines. Although there were important legal and technical reasons for doing so, Galantin had also advised him to ‘go for our own design of nuclear reactor in order to insure that Admiral Rickover had no grounds for interfering in our own Polaris programme’.203
After some minor amendments, the Polaris Sales Agreement was signed in Washington on 6 April 1963 and placed before Parliament on 9 April 1963. It was the ultimate piece of that ‘great prize’ Macmillan had spoken about in 1958 when Anglo-American nuclear collaboration had been restored.204 Two hours after the agreement was signed, the Polaris Executive Special Projects Royal Navy (SPRN) team landed in Washington and left the Americans ‘impressed by the nice timing’.205 Serious work could now begin. But the Sales Agreement was a political document; how it would work at the operating level was now for the Polaris Executive and the SPO to determine. As Captain Peter La Niece, Mackenzie’s representative in Washington, put it, it ‘was really only like signing adoption papers. What this lusty new obligation really meant took a lot longer to work out.’206
CONSTRUCTING POLARIS
Following the signing of the Polaris Sales Agreement, the Polaris Executive began to bring together all the various components of what was known throughout Whitehall as the British Naval Ballistic Missile System (BNBMS): the submarines, warheads, re-entry bodies, communication facilities, navigation facilities, maintenance support and logistics systems, all of which needed to come together at the right time to deliver a credible deterrent. The timescale was so short that all major activities had to start concurrently, requiring the systematic breakdown of the project into systems and sub-systems. By this time a total of 275 naval officers and civilian staff were now working full time on the programme, fifteen in Washington, twenty-six in London and the remainder in Bath. The shipbuilders, Vickers and Cammell Laird, had started to tackle what was to become one of the most highly organized efforts of planning ever undertaken by the British shipbuilding industry, defining precisely what was needed in terms of works, drawings, orders, programmes, manpower and costs, and upgrading facilities, recruiting staff and coordinating the hundreds of British and American firms involved in constructing the submarines and the many complex components that went inside them.207 In Scotland, the Admiralty embarked on a ‘gigantic effort … at breakneck speed’ to settle the requirements for the support facilities, and began to construct what would likewise become one of the largest works projects ever undertaken by the Admiralty.208 In the Ministry of Aviation and the Atomic Weapons Establishment at Aldermaston, clarity on the type of missile allowed work on the ‘one – and vital bit of the POLARIS weapon system which is British made’, the nuclear warhead, to proceed with all dispatch.209
Normally, a new submarine design evolves out of a series of studies into requirements and desired capabilities, which are then traded off against costs.210 With the Polaris submarines, the staff requirements (the main features, functions and concept of operations of the submarine), were completed using what was described as an ‘abbreviated procedure’ in order to have them completed by the end of February 1963.211 These requirements specified a submarine that could patrol undetected for fifty-six days and fire its missiles within one hour’s notice while operating under dual NATO – national control. The submarine had to be able to defend itself, mount a counter should it come under attack, maintain simultaneous, continuous and instantaneous communications with shore installations, and possess the ability to change the target allocation of the missiles within five minutes, all while achieving a reliability factor of 99 per cent.
By 1962, the Admiralty’s naval architects were ready to exploit the knowledge they had gained from designing and building Dreadnought and Valiant. The fact that the Valiant design was so far advanced was crucial to the success of the programme.212 By using it the architects were able to produce one for a Polaris submarine in a remarkably short space of time and avoid ‘the evil of everything new at once’.213 But it would be a mistake to think of the Polaris submarine as a simple cut and shunt job. This was not just a case of cutting an existing British submarine design in two and inserting a self-contained missile compartment in the middle, which would give ‘a false impression of the magnitude of the task’.214 The design ‘was entirely new’ and involved ‘the installation of a highly sophisticated and closely integrated weapons system of US design and manufacture in a British built nuclear submarine of advanced design’ that at that stage was still unproven.215 This was a ‘stupendous, mammoth task’, but it was completed on time and submitted to the Admiralty Board for approval on 6 February 1964, ‘a quite remarkable achievement’.216
With a submerged displacement of around 8550 tonnes (surfaced 7750 tonnes), the submarine was 43 metres longer than ‘Valiant’, only half of which was directly attributable to the missile compartment; the rest was required for other associated changes.217 In total, around 50–60 per cent of the hull construction, propulsion machinery and hull services were the same as in ‘Valiant’.218 The design also differed considerably from US Polaris submarines. The most obvious superficial alteration was the relocation of the hydroplanes (horizontal rudders) from the conning tower (as in US submarines) to the upper part of the forward hull. The challenges involved in grafting the Polaris weapon system onto the ‘Valiant’ design were considerable. In nearly every boundary between the US equipment and the UK submarine, piping, wiring and power supplies had to be integrated.219 To facilitate the fitting of US equipment, the designers decided that the frames, decks and bulkheads in the missile compartment should follow those of the ‘Lafayette’ class submarine. But it was not an exact ‘Chinese copy’. The hull diameter of the missile compartment had to be reduced by 3 inches in order to make it fit with the ‘Valiant’ hull and there were some minor changes at the top and bottom of each missile tube.
Mackenzie knew that ‘If our Polaris submarines are to mean anything
at all they must be as near 100% perfect as possible.’220 This placed a great responsibility on the shipbuilders. If they failed to construct the submarines on time, the government would be unable to maintain its deterrent policy. Mackenzie was initially unimpressed with the way both shipyards, in particular Vickers, approached the complex task of managing the construction programme. As far as he was concerned both shipbuilders ‘had no conception of what they were taking on in building, fitting out, and testing and tuning of an SSBN.’221 Vickers, he said, was ‘virtually still working in the nineteenth century’ and had yet ‘to move into the twentieth century in their methods of organizing themselves to address such a complicated task’.222 It took ‘months and months of argument and persuasion’ before both shipyards ‘committed the resources necessary for detailed planning, which was absolutely essential if the SSBNs were to be built on time’.223
Before physical construction of the submarines could start ‘both shipyards embarked on a task which involved an expansion of facilities, an increase in staff, a change in the structure of staffing, and a profound impact on the standards of work and the nature of management operations’.224 In 1963, the workforce at Vickers numbered 3100 men. By 1967, this had risen by 45 per cent to 4500.225 The most dramatic change was in the numbers of support staff. In 1963 there were 800. By 1967 there were 2400 involved in expanded planning, testing and quality control organizations.226 The ratio of qualified to unqualified staff increased from 1:5 to 2.5:4.227 At Cammell Laird, employment also reached new highs, averaging 8280 between 1960 and 1964 and reaching a peak of 11,400 by 1969. Polaris dragged ‘the shipbuilders … into the modern age of submarine construction’.228
Towards the end of 1963, the new First Lord of the Admiralty, George Jellicoe, steered the Navy through what the Flag Officer Submarines, Admiral Sir Horace Law, later called ‘one of the most painful episodes’ in the entire Polaris programme.229 The US S5W reactor first fitted in USS Skipjack and used by the Royal Navy in Dreadnought contained valves, piping and fittings that were constructed out of stainless steel. Those responsible for designing the British prototype plant at Dounreay concluded that stainless steel was very liable to stress corrosion and they did not want to use it in situations where it might come into contact with sea water. Despite warnings from Rickover the designers decided to use a different material, a chrome-moly steel and a material known as Inconel.230 This was a costly mistake. In November, as Dounreay was nearing completion, engineers discovered serious flaws in some of the welds in certain pipes and valves of the British prototype nuclear plant. The successful installation and proving of the plant in Valiant was ‘seen to be absolutely critical’ as it was going to power the Polaris submarines.231 In total there were 1658 affected welds and installation work had already started in both Valiant and Warspite. None of them could ‘be regarded as trustworthy’ and by December Rolls-Royce had ceased all work.232 Engineers were unable to find a ‘scientific or engineering explanation’ and ‘no remedy’ was ‘available or in sight’.233 The only solution to this serious, ‘not calamitous but … serious enough’ problem was to replace the defective parts in Dounreay, Valiant and Warspite with a stainless steel alternative.234
The technical capability to design and produce such components in the UK did exist, but it was impossible for UK industry to produce what was needed without causing significant delays to both the Valiant and Warspite construction programme. This had important implications for the Polaris programme as there would be, Le Fanu argued, ‘less and less time in which to prove at sea … the operation of the nuclear propulsion plant on which the SSBNs are utterly dependent’.235 Using a UK-sourced alternative in Valiant and Warspite was out of the question as ‘the probability of delay to SSBN.01’ was judged to ‘be very strong’.236 ‘The only feasible alternative,’ argued Mackenzie, was ‘to acquire the stainless steel pipes and valves from the only other source, the United States.’237 But that meant going back to Rickover.
The Chief Scientific Advisor at the Ministry of Defence, Solly Zuckerman, was sent to the United States to negotiate with Rickover. The two had first met in 1962, after Rickover had read an article Zuckerman had written on ‘Judgement and Control in Modern Warfare’ for the journal Foreign Affairs.238 Thereafter the scientist in effect replaced Mountbatten as Rickover’s preferred channel of communication with the Royal Navy and the British Government.239 Rickover, who in Zuckerman’s view had become ‘more paranoid as the years passed’ since the 1958 Agreement, agreed ‘somewhat reluctantly’ to supply the steel, by opening up US naval stores and allowing the British to purchase what they needed.240 ‘Unfortunately,’ as Zuckerman told Thorneycroft, ‘he did so very largely on the basis of the personal relationship he has now established with me.’241 And there was a price. Rickover saw an opportunity to use the fact that the British required his help to secure concessions in the ‘lengthy and difficult’ negotiations that had been taking place since 1960 over a memorandum of understanding permitting the use of Holy Loch in Scotland as a base for US Polaris submarines.242 Rickover informed Zuckerman that ‘The US helping us over our reactor troubles would be conditional on our coming to an agreement over the Memorandum of Understanding.’243
By the middle of January 1964 negotiations had come to ‘a complete halt’ due to Rickover’s refusal to allow British inspectors access to US submarines using Holy Loch so that they could conduct safety inspections and monitor radiation levels.244 Rickover had become very ‘bitter’ about the number of Foreign Office redrafts of the sections of the Holy Loch memorandum dealing with safety and inspections and he was ‘determined to scrutinize carefully all further changes and suggestions’.245 But it was apparent that there were other factors driving Rickover’s behaviour. One of the fundamental reasons he had agreed to the sale of the S5W plant for Dreadnought ‘was to prevent the British having continuous and direct access to the American project, because he believed that responding to their queries would disrupt American progress and deflect him from his singleminded determination to develop nuclear plants for the USN’.246 Yet here were the British asking for yet more help. ‘They ask for too much,’ complained Ted Rockwell, one of Rickover’s officials.247 On 22 January, Mackenzie warned Le Fanu that:
the most serious aspect of further delay in receipt of the stainless steel is that each week shortens by the same time the now marginal period for proving Valiant’s reactor plant, the pilot U.K. seagoing system, and incorporating the almost inevitable modifications which will prove necessary before reactor plant installation in SSBN.01 must start … The risk that further delays in receiving stainless steel from the US will be reflected in delays to SSBN.01 is therefore present from now on.248
Far more worrying was the fact that ‘both Valiant and Warspite in their final months at Barrow will clash with the initial stages of SSBN.01 testing and tuning’.249
Rickover still refused to provide what was needed.250 ‘If the safety clause in the Holy Loch Agreement is satisfactorily settled,’ Zuckerman repeated, ‘supplies of stainless steel will be forthcoming quite quickly.’251 Eventually, the Foreign Office conceded that Nassau and the Polaris programme had put the government ‘very much at the Americans’ mercy’ when negotiating the Holy Loch memorandum.252 When the Cabinet Committee responsible for the Holy Loch negotiations, GEN 836, eventually approved the memorandum on 24 January 1964, without the right to access certain areas of US submarines, it did so because ministers feared that Washington would ‘withhold certain vital equipment’.253 Rickover never wavered from the pursuit of US interests and the British government had little choice but to accept his demands to keep the Polaris programme on schedule.
The fallout was considerable. Although Rickover had got what he wanted, according to Vice Admiral Louis Le Bailly, the Washington Naval Attaché, Rickover was furious with how the whole affair had been handled and he ‘severed relations with the British embassy including my office’. Thereafter, Le Bailly wrote, Rickover’s ‘gall was well reserved fo
r the UK end of the Polaris affair’.254 Aside from conceding over Holy Loch, the British also accepted tighter security restrictions, including an assurance that ‘No information on this equipment will be published; nor will it be provided to foreign nationals.’255 This was accepted, but it put the government in an awkward position. It had always maintained that the Dounreay prototype was British designed and built. In reality, it owed a ‘very great deal to American information’ and was sometimes referred to by the Navy as the ‘Son of S5W’.256 Putting US-purchased auxiliary systems in both Valiant and Warspite meant that it was harder to sustain the argument that the end product was more British than the sum of its parts, as one official, Philip Newell, explained:
