The Collins Class Submarine Story

by Peter Yule

  was now an extensive change of personnel. On 4 August 1986

  Oscar Hughes was promoted to Rear Admiral and his position

  renamed submarine project director. Graham White’s replace-

  ment, Captain Geoff Rose, an experienced submarine engineer,

  took the position of project manager, reporting to Hughes.

  Control of price remained a central concern during the project

  definition phase. Kim Beazley had an extensive program for buy-

  ing military equipment. The government was not prepared to

  spend proportionately more on defence, so Beazley refined defence

  policy, focusing on the defence of Australia and its closer region,

  and ensured this was reflected in the priority given equipment

  T H E P R O J E C T D E F I N I T I O N S T U D Y 1 9 8 5 – 8 6

  97

  programs. He was strongly supported by the Labor Party,

  although some questioned the place in this strategy of large sub-

  marines capable of deploying to the north Pacific. Beazley later

  said that a box called ‘deterrence’ had to be attached to the policy

  to accommodate the submarines as well as F-111 bombers and the

  army’s special forces. But there was always the Australian building

  program to keep caucus firmly behind the submarines.

  The groundwork for the policy was Paul Dibb’s Review of

  Australia’s Defence Capabilities, released in early 1986. Although

  Beazley told him the submarine project was the one area he could

  not touch, Dibb nonetheless looked at the project’s finances, with

  the estimated price now at about $2.6 billion. Dibb noted that the

  six proposed new submarines would be the equivalent of nine or

  10 Oberons and would allow three separate areas to be patrolled

  simultaneously. Since Australia’s submarine fleet was already the

  most powerful in the region, the additional capability provided a

  margin available for trade-off if project costs increased: ‘Should

  there be predictions of a substantial cost escalation in the sub-

  marine project due to local construction problems, then options

  involving some lesser capabilities could be considered.’20

  In the event, the financial situation for military spending wors-

  ened drastically. In May 1986 the Treasurer, Paul Keating, made

  his famous comment that Australia’s balance of payments situa-

  tion was threatening to turn the country into a ‘banana republic’.

  Viewed at first as a political gaffe, it was in reality the beginning

  of his campaign to force the government to tighten fiscal policy

  by reducing expenditure across all portfolios. The result for the

  military was that the projected increases for future budgets were

  halved, with little real growth for the foreseeable future.

  As the definition study drew to a close, Beazley told Oscar

  Hughes that he could not get the submarine project through the

  cabinet if the price started with a ‘4’. Hughes argued that any

  reduction in capability should be the last option, and sought sav-

  ings in other areas. For this reason the contingency funding against

  future problems was limited to $50 million whereas, on a par with

  similar projects, it should have been around 15 to 20 per cent of

  the estimated price – about $600 million.

  In mid-1986 Hughes visited Anaheim and talked with Rick

  Neilson about Rockwell’s work on the combat system. When

  Neilson spoke of his efforts to improve the combat system design,

  98

  T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y

  Hughes told him: ‘That’s all very well but I can’t afford any more

  money.’21 The directive was authoritative and, for the remainder

  of the definition study, Neilson and his colleague John Dikkenberg

  told Rockwell to cut costs. They were so persistent that Rockwell

  reduced the number of control consoles in its design to reach a bid

  price of around $460 million. Nonetheless, Neilson and Dikken-

  berg remained impressed by the development of the Rockwell

  system and the full-scale control room mock-ups, which allowed

  them to see how the system would function.

  From August 1985 Neilson’s colleague from SWSC, Andrew

  Johnson, was also in Anaheim representing Computer Sciences in

  Rockwell’s consortium. Having been one of the leaders in devel-

  oping the combat system concept, Johnson had a good idea of

  how it was supposed to work. However, he found that Rockwell

  knew little about submarine combat systems and that it had

  entered the Australian competition to gain an entry into this mar-

  ket. That was all too apparent to another consortium member,

  Singer Librascope, which owned the expertise and had no desire

  to create a competitor. Rockwell also had difficulty working with

  Thomson CSF, the French sonar makers, which also feared for

  its intellectual property. Well into the project the problem of link-

  ing the components together with software written for Rockwell’s

  data bus was looking no closer to solution.

  Singer Librascope had expected to build and integrate the

  system and write the system software but not to reveal source

  codes as this would threaten its market position. This undermined

  Rockwell’s strategy, so it proposed to Johnson that CSA should

  write the software, thus creating a system that Rockwell could

  market. CSA had entered the consortium to perform a compara-

  tively modest $40 million task of assembling the Australian end

  of the project and producing things such as land-based training

  systems. Writing all the software would be worth around $100

  million and placed CSA at the centre of the system’s development,

  but at the cost of reducing or removing access to Singer Libra-

  scope’s expertise. Johnson agreed and, he says, ‘within that deci-

  sion the Greek tragedy starts’.22

  Oscar Hughes also visited Signaal in mid-1986 and Mick

  Millington recalls him reinforcing the line of financial discipline,

  telling the Dutch company’s executives: ‘If you come in at one dol-

  lar over $500 million we’re not going to talk to you any more.’23

  T H E P R O J E C T D E F I N I T I O N S T U D Y 1 9 8 5 – 8 6

  99

  Signaal’s financial manager was absent from the meeting and was

  unimpressed by the imposition of a cost ceiling. Unlike their coun-

  terparts in the US, the liaison team was unable to convince Signaal

  to heed the warning and the company continued a single-

  minded pursuit of the specifications, designing a system it priced

  at $520 million. Then, over lunch with senior defence bureaucrat

  Malcolm McIntosh, Signaal’s Canberra representative, Altingh

  von Geuzau, learned that Rockwell had reduced the quality and

  cost of its bid. The Dutch followed suit, tendering their bid a week

  late at a more modest cost of $480 million.24

  Having changed its early conservative course, Signaal devel-

  oped the combat system of Millington’s aspirations. Further,

  Millington had come to respect Signaal’s abilities and had no

  doubt that it could deliver the system it promised. ‘With the Dutch

  and Germans a handshake is
a firm commitment, it is solid stuff,

  while my experience with American companies is that the hand-

  shake is only the beginning of negotiations for contract amend-

  ments.’ However, Signaal’s late change of direction meant that it

  was unable to adequately document its new system so that when

  the evaluation began it was difficult to match the documentation

  with the specifications.

  In Sweden, Kockums continued to develop its design appar-

  ently uninhibited by costs. Greg Stuart noted that the company

  had little commercial sense, as it had only designed for the Swedish

  navy in an arrangement that was closely consultative. He found

  that Kockums’ approach was very different to any in his experi-

  ence, as it was forced to constantly evolve its own designs because

  of Sweden’s isolation from NATO. Stuart, who has the reputation

  of being an engineering perfectionist, concluded that in subma-

  rine building there was ‘the right way, the wrong way and the

  Swedish way – the Swedish way is not necessarily wrong, but it is

  different’.25

  Olle Holmdahl headed the massive design effort by Kockums,

  where the company contained its costs within the $10 million

  provided by the Commonwealth only by discounting the design-

  hour rate. The effort of the 50-strong design team was focused on

  the central design issues by allowing them to select the best for-

  eign equipment in specialist areas. This was familiar practice for

  Kockums but generally unusual in submarine design, where most

  builders preferred their own nation’s components. The Swedish

  100

  T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y

  design featured Italian masts and hydraulic systems, British hull

  valves, air compressors and torpedo discharge system, the French

  Jeumont Schneider electric motor and German Varta batteries.

  Having left the preliminary design review convinced the cheap-

  est bid compliant with the navy’s requirements would win, the

  AMS team saw no benefit in offering more than was outlined in the

  ship’s characteristics.26 The German submarine was redesigned to

  reduce battery capacity by 10 per cent. Further savings of about

  $150 million were found by adopting German weapons-handling

  arrangements, with the bow being designed around the new MAK

  weapons discharge system.

  The Germans were confident that their research into fuel cells

  for air-independent propulsion would provide their future subma-

  rine designs with superior submerged performance. It is possible,

  however, that their enthusiasm for fuel cell propulsion led them to

  underestimate the Australian navy’s desire for a large conventional

  battery capacity.

  The two combat system tenders were submitted in October

  1986. The submarine bids were to be submitted on 11 November,

  but ASC decided to pre-empt the opposition and lodged its bid

  with great fanfare a day earlier. Geoff Davis contributed to the

  stunt by having the massive volume of documents delivered in

  metal containers by Wormald security vans. The result was that

  the Swedish bid received favourable press coverage, while the

  German submission the next day was virtually ignored. It was

  a little thing but it set the psychological atmosphere for the events

  that followed.

  C H A P T E R 10

  Debating the laws of physics: picking

  winners 1987

  In early 1987 conservative politics in Australia was splintered by

  maverick Queensland Premier, Joh Bjelke Petersen, and his bizarre

  attempt to enter federal politics. With the opposition in disarray

  and a federal election due by the end of the year, there was obvious

  temptation for the Labor Party to go to the polls sooner rather

  than later. The project team was under great pressure to meet Kim

  Beazley’s wish and have a contract signed before the election.

  The assessment of the four bids involved over 300 people in

  more than 40 specialist teams, with literally tonnes of documents

  to be assessed. The documents supplied with the Signaal bid alone

  made a pile seven metres high.

  By early January 1987 the individual working groups had

  concluded their evaluations of the submarine design and these

  were consolidated in the report of the submarine evaluation

  team. Harry Dalrymple, one of the principal signatories of the

  report, found the assessment a challenge. His team had devel-

  oped expertise in submarine design principles while supporting

  and improving the Oberons, but had no experience of origi-

  nal design. BHP Engineering was contracted to provide gen-

  eral guidance through to the contracting stage and, somewhat

  101

  102

  T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y

  ironically, the British Admiralty Research Establishment and the

  Netherlands Ship Model Basin were both approached to provide

  modelling research to assess the propulsion performance of the

  designs.

  The results of the evaluation were startling – the German pro-

  posal now barely came up to scratch. Their design, though the

  clear winner of the earlier contest, was now rated as only a ‘fair’

  response ‘which would seriously jeopardise the ability of this sub-

  marine to fulfil the specified mission requirements’. It had ‘an

  assessed inability to fulfil major performance requirements with-

  out significant redesign’, a condition that ‘borders on a CRITICAL

  [ sic] deficiency’.1

  The German boat was more than three metres shorter than the

  Swedish design and displaced 275 tonnes less on the surface. The

  evaluation team considered that the smaller pressure hull, with

  reduced battery capacity and fuel load, limited its capability. The

  Australian evaluators considered the Type 2000 could stay sub-

  merged for only two-thirds of the time required, would have a

  covert transit range 15 per cent less than asked for, and be able

  to stay on station for less than half the specified time. While the

  Swedish boat could not quite meet some of the navy’s criteria it

  failed only narrowly, being three per cent down on the covert tran-

  sit range and seven per cent short of the time required submerged

  on batteries. It fully complied with the required patrol length.2

  Surprising as they were, these results became controversial

  when the way they were calculated became public. To measure

  the competing designs against a common baseline, the evaluation

  team recalculated much of the data submitted by the two design-

  ers. The result of this process was that many of the German figures

  were revised downwards, while the Swedish figures were revised

  upwards.

  The estimated covert transit range of the IKL design was

  reduced by about 25 per cent while that for the Kockums design

  was increased by five per cent. The long-range and short-range

  patrol endurance figures claimed for the German boat were

  decreased by around 70 and 55 per cent respectively, while the

  Swedish figures were increase
d by five and 10 per cent respec-

  tively. The indiscretion ratio for the German boat in transit was

  reduced by 16 per cent, while the Kockums design increased eight

  P I C K I N G W I N N E R S 1 9 8 7

  103

  per cent. On patrol, the indiscretion ratio claimed by the Germans

  was reduced by 50 per cent, while the Swedes were penalised only

  14 per cent. The only area in which the German boat benefited

  from the recalculations was that its claimed submerged endurance

  at a speed of 21 knots was increased by 33 per cent while that of

  the Swedish boat went up some 20 per cent.3

  Mark Gairey explained the reasoning of the evaluation team

  that led to these results:

  Bearing in mind that both companies are designing to meet

  the same requirements, you would expect them to come up

  with a fairly similar solution, which in many ways they did.

  The designs were similar in dimensions . . . yet the

  performance they were offering was different. The Germans

  were claiming that they were meeting or exceeding everything

  we asked for, while the Swedes said, ‘We meet most of the

  requirements, but not all of them . . .’ But when you sat

  down and looked at the fact that the Swedes had more fuel, a

  bigger battery, a bigger engine, a bigger motor, but they said

  they were offering less performance, something didn’t make

  sense. The laws of physics say that if the bodies you are

  pushing through the water are more or less the same then if

  you put more power and everything else, you should get

  more performance and not less. That made us go away and

  do a whole bunch of work to try to understand why. And

  basically it boiled down to the assumptions being made by

  the designers. Basically the Germans were taking the most

  optimistic approach they could, even to the extent of saying

  things like, you can’t actually operate the galley and you

  can’t do this and you can’t do that, or you’d be using too

  much power, while the Swedes assumed normal operating

  conditions.4

  As Greg Stuart put it: ‘With the German submarine, if the crew

  were in bed, the combat system turned off, the lights off, etc then

  it could do what it claimed – but it could not do it in a real-

  istic operational mode.’ This was because, at speeds less than

  seven knots, propulsion demanded only a low drain on the bat-