14/1/19
Golfballs on the Moor: Building the Fylingdales Ballistic Missile Early Warning System
Graham Spinardi, University of Edinburgh
Working Paper (to be revised for publication in Contemporary British History)
Introduction
At the beginning of the 1960s three large golf ball like structures appeared on Fylingsdale Moor, just inland from Whitby in the North Yorkshire Moors National Park. These radomes housed huge radar dishes designed to scan the horizon and provide early warning of Soviet nuclear-armed ballistic missiles, providing the third and final link in the American Ballistic Missile Early Warning System (BMEWS).
The ‘golf balls’ were replaced in the early 1990s by a ‘pyramid’ using phased array radar technology, and in early 2003 the UK agreed to a US request to update Fylingdales as part of its plan to deploy a ballistic missile defence system. In many regards the situation now thus has parallels with that of the late 1950s when the US first sought permission to base a BMEWS site in the UK. Then, as now, there were doubts not only about whether the technology would actually work, but also whether it was of significant benefit to the UK.
Based on documents available in the UK National Archives this paper provides for the first time a detailed account of the deliberations over the building of the Fylingdales golf balls. This is the story of one of the key technologies of the nuclear age, and yet one that could not be realistically tested. Ironically, the only confirmation that BMEWS worked technically would have come when deterrence itself had failed.
Building BMEWS in the UK: The Deliberations
The development of BMEWS stemmed from growing concern over the vulnerability of the US to nuclear-armed ballistic missiles. The US was developing such missiles and was well aware that the Soviet Union sought a similar capability. MIT’s Lincoln laboratory was commissioned to carry out studies during 1953 and 1954 on ballistic missile detection and these ‘suggested that radar was the only sensor technology that offered the near-term possibility of developing a warning system against missiles.’[1] The Laboratory began design work in 1955 and formed the Systems Research Group to assess the challenges and compare different designs.[2] Further study was carried out on the most promising design – ‘detection radars scanning several pencil beams in azimuth at fixed elevations and an associated pencil-beam tracking radar’.[3] This approach was recommended to, and accepted by, the Department of Defense and Air Force. Contracts to build BMEWS were awarded to RCA in January 1958. Four objectives were defined: 1. 15 minute warning of mass ICBM attack on US; 2. Reliability of 0.9999; 3. maximum false alarm rate of one during three month period; 4. inherent flexibility and growth potential.[4] The first two BMEWS sites were to be at Thule, Greenland and Clear, Alaska.
The Lincoln Lab’s technology demonstrator, built at Millstone Hill, about twenty miles from the Lab, was nearing completion when the Soviet satellite Sputnik was launched on October 4, 1957. Although not fully ready, the radar was successful in detecting Sputnik, thus providing confirmation that sufficient advances had been made to accomplish the task of ICBM detection.[5]
To get its performance Millstone Hill used a very large tracking antenna (a dish 84 feet in diameter), high power klystron transmitters, and state-of-the-art sensitivity receivers. However, such a tracking type of radar was considered unsuitable for the main BMEWS stations because of the large area of the sky that had to be watched. Instead, both Thule and Clear used very large fixed antennas, each covering a particular azimuth and elevation. Thule had four of these antennas, described as ‘larger than a football field turned on end’, while Clear had three.[6] In addition, each site would eventually also have a tracking radar based on the Millstone Hill system.
Thule’s four AN/FPS-50 radars provided total azimuth coverage of 160 degrees, each radar covering 40 degrees from 297 degrees to 097. Clear had three AN/FPS-50 radars to give coverage of 120 degrees from 240 to 360 degrees. Both used fixed Torus reflectors 165 feet high and 400 feet long. Two elevations were scanned. The lower ‘fan’ was scanned by dual pencil beams, 1 degree by 1 degree, 20 degrees apart, which moved across the 40 degree sector at an elevation of 3 1/2 degrees. Single beams at 7 degrees elevation formed the upper fan. The performance claimed was 90% probability of detection of missile at range of 2,600 miles.[7]
More than one BMEWS was required to provide complete radar coverage against Soviet based ICBMs because the Earth’s curvature limited the radars’ ‘line-of-sight’. Although Thule and Clear provided coverage of most of the Soviet missile threat, there was a small gap and to fill this a third site was sought somewhere in Northern Europe.
At government level the UK was supportive of American interest in siting BMEWS No 3 in the UK. Initially the north of Scotland was considered because this would have given the best coverage from the US point of view.[8] In October 1957 ‘preliminary siting investigations were made by a US/UK team’.[9] An official request was made in December 1957:
I refer to recent informal conversations between officials of the Air Ministry and officers of the US Third Air Force regarding the proposed establishment of a missile detection facility in the United Kingdom. In this connection I wish to confirm to you the desire of the United States Government to obtain permission for the entry to a United States survey party in the latter part of January, 1958 for the purpose of locating possible sites for a ballistic missile detection facility.[10]
Three sites in Scotland were of particular interest. The one considered most promising was near Fochabers and coincidently called Millstone Hill; the other two were about 20 miles south-west of Wick and near Ballantrae on the south-west coast.[11] During 1958 these, and a number of other sites, including one on the Yorkshire coast near Bridlington, were surveyed for suitability. At this stage, however, Scottish siting seemed most likely, with surveys done in England ‘in order to allay the susceptibilities of the people of North East Scotland’.[12]
Towards the end of 1958, however, interest shifted further south as it was clear that such a location would be best for providing early warning for the UK. In particular, it was argued that if BMEWS was to form a useful part of a defensive anti-ballistic missile (ABM) system for the UK then ‘it is apparent that a BMEW station sited in Scotland would not meet our operational requirement.’[13] Sites were surveyed in Lincolnshire and Norfolk, with the Stanford Practical Training Area, near Thetford in Norfolk, appearing to offer the most promise.[14] However, there was a major obstacle to the use of Stanford – the War Office, which vigorously opposed the loss of what it considered ‘by far the most valuable training area in the UK’.[15] As an alternative, Sir Harold Pyman at the War Office, suggested: ‘I wonder if you have looked at Fylingdales in Yorkshire which is shortly to be released by the Army?’[16]
In March 1959 Fylingdales became the favoured location. It had now been realised that siting in densely-populated Norfolk would have been problematic in any case because of the radio interference caused by the BMEWS radars.[17] On April 3, 1959 the American Embassy in London handed an Aide Memoire to Sir Frederick Hoyer Millar ‘which expressed an interest in the Flyingdales Moor site in Yorkshire as a possible area for a Ballistic Missile Early Warning System station, and which stated that the United States expected to submit a draft agreement concerning the establishment of such BMEWS station for detailed consideration by the United Kingdom.’[18]
Further US-UK discussion led to an agreement in principle, endorsed by Minister of Defence Duncan Sandys and US Deputy Secretary of Defence Quarles, ‘that the station should serve both the UK and the US’.[19] The outcome was summarised in May 1959: ‘at high level discussions between the UK and US the merits of using the station to give early warning both for the UK and US had been appreciated, and to this end the Americans were prepared to concede some loss of cover by siting the station further south in order to provide cover for the UK.’[20] Scotland would in that case no longer have been ideal because a ‘site in Scotland would give no warning of a missile aimed at a UK target south of the line Cardiff/Hull.’[21]
Although the UK government was keen to cooperate with the US over BMEWS, there remained two issues to be worked through. First, to what extent did BMEWS meet UK military requirements? Second, how would the costs of a Fylingdales BMEWS be shared? The first of these matters was largely a matter for discussion within the RAF, given that it was the service whose air defence role also encompassed defence against missiles. As it happened, an Air Staff Requirement (ASR 2149) for ballistic missile early warning had just been promulgated:
The United Kingdom operational requirement for early warning against BM has been stated in ASR 2149, dated 11th September, 1958, which briefly is as follows:- (a) To provide maximum early warning on the approach of ballistic missiles launched from Russia or satellites against the United Kingdom. (b) To be able to compute the ground impact point to an accuracy of ±20 miles. (c) To provide sufficient capacity to compute ground impact points of approximately 10 missiles within the coverage at the same time. (d) To provide maximum immunity to false alarms. (e) To be designed in such a way that it can form the first stage of an active defence system.[22]
Thus discussion initially centred on whether BMEWS would meet British requirements as set out in ASR 2149. The proposed Fylingdales BMEWS was to comprise three AN/FPS-49 radars, operating automatically, two of them scanning at a low angle (2 1/2 º to the horizontal) in the direction of Russia from 340º true to 135º true with the third radar used for tracking (ie positively identifying and calculating missile trajectories). One scanning radar was to scan from 0 to 75 degrees, while the other would cover 75 to 135 degrees.[23] This configuration differed from the other two BMEWS sites, where only one AN/FPS-49 tracking radar was used, while the scanning was done by large fixed antenna positioned to cover a set area of the horizon. The discussions concluded that ‘although there were arguments for and against the choice of scanning radars versus tracking radars, it was the general opinion that scanning radars would match the US requirements best, though not the British. Because of this and the lesser cost of the works aspects of using tracking radars the US were prepared to agree to a station using three tracking radars.’[24]
The system’s specifications were described as follows: ‘The frequencies will be in the band 405-445 mcs, the peak power per equipment will be 5 MW per beam and the average power 270 KW per beam. The tracker aerial is an 84 ft. dish, weighing with its mounting 75 tons.’[25] In fact, by the time they were built each radar had a peak power of 10 megawatts operating at a frequency of 425 mc giving a range performance of 2500nm. The pulse repetition rate was 27 pps and the pulse width 2000 µsec. The 84 feet diameter dish was fed by a five-horn monopulse system using circular polarisation.[26]
The echo data which were to be automatically fed to data take-off (DTO) equipment were the target’s range, elevation, and bearing/azimuth, and the rate of change of target range, elevation, and bearing/azimuth. If an echo was seen on two successive scans the DTO would start processing it as a potential missile threat, but one hit followed by three misses would automatically cancel the echo as a potential target. This information would be available to the UK and also transmitted to the US Air Force Defense Command, Colorado Springs (NORAD), where it would be collated with information from the other two BMEWS stations.
However, there were strong doubts about the value of the BMEWS data that would be available to the UK, particularly as judged against Air Staff Requirement 2149: ‘If the ASR does truly represent Air Staff requirements, No 3 BMEWS as proposed would not meet them. The extent to which it does, the extent to which it could be made to meet them and the price to pay must all be carefully assessed.’[27] The two key areas of performance of concern were the number of targets that could be tracked and the amount of warning time provided.
Potential targets could be tracked for 7 to 200 seconds by the tracking radar to compute the following data on each missile track: two velocity components of reentry angle, predicted ground impact coordinates, launch point coordinates, data quality, time to go to impact, and the alarm level. Tracking was to be done serially so it was unlikely that more than four targets could be tracked per minute. This was less than considered desirable: ‘A realistic tracking requirement for the UK station would therefore appear to be in the order of 10 tracks per minute. However, as the actual capacity will be approximately four tracks per minute there will obviously be a conflict of priorities between tracking ICBMs aimed at the US and identifying IRBMs launched against the United Kingdom.’[28]
From the point of view of ABM potential, there was particular concern about the ability to track on an average only four or so missiles per minute. The early warning information would be too vague to be trustworthy unless corroborated by tracking information, and thus the value of the station for providing warning and evidence of attack would be ‘severely limited by the low capacity of the tracking radar’.[29] While the US was primarily concerned with ICBMs, the value to the UK lay in the ability of BMEWS to detect and track IRBMs. In this regard there were British doubts about the utility of BMEWS. It was felt that to ‘meet the UK and US requirements satisfactorily, and to identify both IRBM and ICBM in adequate numbers, the tracking capacity provided would have to be much greater than at present planned.’[30]
Not only was the tracking capacity considered inadequate, but there was also scepticism about the utility of such short warning times. It was estimated that the BMEWS at Fylingdales would provide ‘between 3 1/2 and 17 minutes of warning of the impact of a ballistic missile, depending on the type of missile, the launch point and the trajectory.’ Few, if any, UK nuclear delivery systems could be launched with such little warning. According to an early study of BMEWS: ‘Given strategic warning, the ‘V’ Force can be kept on dispersals at 15-minute readiness for up to a week. A few bombers might become airborne given 3 1/2 minutes warning. Thor could not be launched in this time. The reaction time of Blue Streak, when it is fully operational in the period 1965-70, will be much shorter than that of Thor, but it is unrealistic to suppose that any political decision to launch Blue Streak on the strength of radar warning could be obtained and transmitted in such a period. Moreover, Blue Streak is, of course, designed to ride out the initial nuclear impact.’[31] The conclusion was that ‘the direct value of the system to the UK would not be great. We could achieve broadly the same results with a telephone link to NORAD, if the Americans agreed to provide one.’[32]
This criticism was expressed publicly in an article by Chapman Pincher in the Daily Express on 26th November, 1959. ‘This is not nearly enough time to get the RAF’s bombers airborne and would be of little help to the civil population. …To me it seems that we are to pay £7,000,000 for the privilege of having time to say to the Americans: “It’s been nice knowing you.”’[33]
Alongside such concerns, there was also irritation over the feeling that the Americans were being very pushy, demanding early agreement without supplying the information that the RAF considered necessary. On 10 June 1959 the Deputy Chief of the Air Staff made this opinion clear to the Secretary of State:
We are by no means certain that the scheme, as proposed, will necessarily be worthy of our support, nor are we certain that we can afford it in terms of money or manpower, bearing in mind the operational advantages we shall get from it. … Any delay there has been to date has been caused by lack of information, largely of a technical character, from the Americans, and by the fact that we have not received an official proposal until very recently. Until we receive this information and an official approach, we could obviously not give the Americans any assurances; what the Americans perhaps do not realise fully is that such a site in England raises a great many more problems than one in Greenland or Alaska.[34]
The next day AMSO backed up this feeling, noting that: ‘I share DCAS’s misgivings about the way in which these negotiations are moving. We seem to be on the verge of being committed to a project of uncertain value to us on the basis of an Agreement we have only just received in draft and to a time-scale of planning and construction which is unrealistic.’ AMSO pointed out that the ‘possibility that the Station could be so sited as to be of value to UK defence only emerged last autumn, and although a good deal of investigation has been done at Staff level, a lot remains to be done before we can be reasonably sure that the project can be implemented or is worth implementing from the British point of view.’ Moreover, it was his opinion that the proposal ‘ bristles with problems – technical and constructional as well as political – and we should be more certain than we are at present that they can be overcome before we make an agreement.’[35]