by Peter Yule
requirements, technical risks and engineering suitability of the
designs, logistics support, Australian industry involvement, pro-
duction methods, company viability and costs. The teams’ reports
were then used by the Tender Evaluation Board to make an overall
assessment of the proposals.7
Evaluating the designs was a new and complex challenge for
the Australian navy and one that many thought was beyond
its competence. Terry Roach, who was then director of subma-
rine policy and who oversaw the navy’s interests in the project,
remembers Admiral Sandy Woodward, head of the Royal Navy’s
submarine squadron, telling him that: ‘Australia should think
carefully about the consequences of not buying the British subma-
rine. How would you judge them?’ To which Roach responded:
‘We will judge them by the lessons you have taught us.’ This typi-
fied the Australian submarine community’s confidence that exper-
tise gained from their British training, operational experience and
the success of the Oberon weapons update gave them the ability
to choose Australia’s new submarine.
It has often been claimed that many of the principal actors
came to the decision-making process with strong prejudices. In
particular the alleged anti-British and pro-German sentiment of
senior members of the submarine project team was said to have
B U T H O W W I L L Y O U J U D G E T H E M ?
63
greatly affected the evaluation.8 Those involved naturally brought
their own experience to the process and argue that their views
were based on their experience and their professional judgments
of the quality of the bid.9 For example, Harry Dalrymple, who
was in charge of naval design and a significant decision maker in
the process, recalls that his initial preference was for IKL/HDW
on engineering grounds – he had visited them and was impressed
by IKL/HDW’s design processes, competence and track record.
Dalrymple considered their experience of building in other coun-
tries was an important point for the project. He supported the
British Type 2400 design as a safeguard against risk and cost.
Other senior officers such as Bill Rourke and Graham White
were also impressed by the German design and manufacturing
processes, and White considered that the Germans’ engineering
was superior to the other contenders. Bill Rourke was not con-
vinced that Kockums could design a large submarine, because he
had seen one of the small Swedish boats being lifted from the
water by a crane.10
Rourke’s evaluation plan was designed to temper such sub-
jective viewpoints, but inevitably the views of senior officers car-
ried great weight and some refinement was needed to balance the
weighting of the various judgments leading to a decision. Oscar
Hughes, who was in charge of naval production at the time, recalls
that at one evaluation meeting Kockums was about to be excluded
because its design had insufficient space for Mark 48 torpedoes
and Harpoon missiles, but it was realised that Kockums’ design
was based on inaccurate information provided by the navy and it
had to be allowed to adjust the design.
Hughes says that this incident led to the adoption of a new
approach, with problems in a proposal being categorised as criti-
cal, important or less important.11 The Tender Evaluation Board
Report followed this procedure meticulously, producing at the end
of each section a matrix combining the proposals and their perfor-
mance against the relevant criteria. Nonetheless, this process did
not necessarily clarify the reasons for decisions in the minds of all
participants, with even some of those who participated in the final
report finding that judgments they had made in their particular
section were not necessarily reflected in the ultimate choice of the
definition study participants.
64
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
Each proposal was judged against the relevant criteria and an
assessment was made of the cost and risk to remedy each unsatis-
factory aspect. This approach was inevitable to a degree, because
each bid reflected company design practice and none followed
closely the ship’s characteristics outlined in the tender request.
For instance, the navy had specified freshwater cooling systems,
mostly inside the pressure hull, primarily to reduce the risk of
flooding. Yet most proposals featured external seawater cooling
systems.
Inevitably, as the cost and risk of correcting each deficiency
was considered, proposals the navy was prepared to accept moved
further away from any submarine that already existed. Thus the
requirement that the new submarine had to be based on a boat
soon to be in service with a foreign navy was gradually worked out
of the process. Oscar Hughes, noting the increasing irrelevance of
these conditions, privately concluded that they should be dropped
from the evaluation criteria. To Hughes the ‘in-service’ condition
was asking for an outdated design, as any boat that met it would
date from the 1970s, while the navy was looking for a submarine
for the 1990s and beyond. Significantly, the government was never
officially told that the nature of the new submarine project had
fundamentally changed – but from this time on Australia was
committed to developing and building a new class of submarine,
with all the risks that this entailed.12
The Italian proposal provided little of the data required for
evaluation, failed to consider production in Australia and was
quickly discarded.
The French design based on the Rubis class nuclear submarine
offered an impressive performance in several areas, exceeding the
requirements significantly in range, battery endurance, deep diving
depth and indiscretion rate.13 However, there were some serious
problems with the design. The arrangements for stowing, han-
dling and discharging the weapons were designed around French
equipment and naval practices and were incapable of handling
the Australian navy’s American weapons; the design’s noise per-
formance was little better than the Oberons; and it did not provide
adequate electrical power or space for a combat system of the type
required. Further, the logistics planning to support the vessel was
considered inadequate and it appeared that the French had not
understood the requirement to give detailed plans for building in
B U T H O W W I L L Y O U J U D G E T H E M ?
65
Australia.14 These problems were considered so serious that the
French design was deemed to be unsatisfactory.
The German Thyssen TR 1700A also exceeded expectations
in some areas. Slightly enlarged from the Argentinean version, the
design’s battery endurance was excellent and, although no detailed
construction plan was provided, Thyssen claimed that the Aus-
tralian proportion of construction would reach 80 per cent by the
sixth boat.
However, the TR 1700A was noisier than some types of
submarine already operating in the Asia-Pacific area and an exten-
sive redesign, involving high cost and risk, would be required to
overcome this problem.15 The board concluded that it looked like
the first design of a company with limited independent research
and development capacity and rated it as unsatisfactory.16
British company Vickers offered two closely related designs:
the Type 2400A, a version of the Upholder class then being devel-
oped for the Royal Navy, and the Upholder itself as the Type
2400B for ‘cost’ – in other words, an alternative should the pro-
curement of an entirely new submarine design prove to be too
expensive. This was useful for comparing the performance of an
existing design with the required ship’s characteristics.
The evaluation found that the standard Upholder had critical
performance deficiencies in endurance, range, indiscretion rate
and deep diving depth. It was designed to patrol the Iceland–
Faeroes Gap in the North Atlantic and had only 60 per cent
of the required range and endurance. The British boat had only
two diesel engines and generator sets, resulting in slow battery
recharging. The Oberons had a similar limitation, and many
involved in the Australian project regarded the Upholder as lit-
tle better than the Oberons. The board concluded that selecting
the Upholder would ‘eliminate the majority of forward operat-
ing areas’ used by the navy and they rated it unsuitable for the
project.17
The Type 2400A was a redesign of the Upholder class stretched
by 10 metres to accommodate a third diesel engine. Both Rod
Fayle and Eoin Asker [later to be the submarine project director
but during this period submarine liaison officer in Britain] saw
the Type 2400A as inefficiently designed – ‘clunky’, in Asker’s
words. Fayle thought that Vickers had used many components
from nuclear submarines in the Type 2400A and that these were
inefficient in a conventional submarine.
66
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
The board noted that the 2400A failed to meet the require-
ments for submerged endurance by as much as 30 per cent
because it lacked sufficient battery storage.18 Electrical generat-
ing capacity was inadequate and during snorting both converters
would be required to keep systems functioning while the batteries
recharged.19 An overload would cut power to lighting, commu-
nications, hydraulics, air conditioning and the combat system.
Despite its size, the Type 2400A lacked space to carry the weapon
load the navy wanted.
While considerable redesign would be necessary to accommo-
date the four additional weapons required, the evaluation board
considered that the technology of the Type 2400A was of a high
level and acknowledged the value of continuing Australia’s close
ties to the Royal Navy and Britain’s military research and devel-
opment program. The 2400A was rated overall as a ‘marginal’
proposal.
The Kockums proposal, the Type 471, was also a design for a
large submarine of 2215 tons. It was the only design apart from
the Type 2400A that could fit the large bow sonar that was part
of the Rockwell combat system. With a high volume allocated
to batteries the assessment indicated that the battery endurance
requirement could be exceeded by 75 per cent. It was also capable
of exceeding the deep diving depth by 18 per cent. On the other
hand, at periscope depth the Type 471 could reach only 75 per cent
of the specified speed before the propeller started to cavitate.20
Noting that the design ‘has not undertaken any tank testing’,
the board thought that the cavitation problem could be recti-
fied by such testing, perhaps in conjunction with a new propeller
design.21 A number of other problems were identified, including
the use of seawater cooling systems and unsuitable high-pressure
air distribution.22
It was the ‘novelty’ of the Swedish proposal and Kockums’ lack
of experience with large submarines that most worried the board.
Furthermore, Sweden was not part of NATO and this raised
concerns that the United States might refuse to supply informa-
tion about American weapons and systems. Indeed, the US Navy
had advised Admiral Rourke in May 1984 that it was unlikely
the United States would release technology to Sweden. Kock-
ums’ response was to propose sub-contracting the development
of the weapons handling system and torpedo tubes to the British
B U T H O W W I L L Y O U J U D G E T H E M ?
67
company Strachan & Henshaw. Although Harry Dalrymple
argued the Kockums’ boat was unsuitable for further consider-
ation, the majority of the board considered it to be a marginal
proposal.23
The proposal for the Walrus was virtually the same as that
being developed for the Royal Netherlands Navy, but without the
Dutch combat system. Despite its reputation, the Walrus had only
75 per cent of the required battery endurance and its indiscretion
rate at 10 knots was more than 20 per cent below that required.
Submerged speed was inadequate, as the Walrus had only a slightly
larger main motor than the smaller Zwaardvis class.24 The Walrus
had only four torpedo tubes, where six tubes had been specified, so
a redesign was proposed for the definition phase. Air-conditioning
and refrigeration capacities were also well below requirements.
Although the board acknowledged that the Walrus was well-
designed and very manoeuvrable with good stability, it needed
considerable redesign to meet the required ship’s characteristics
and the board concluded this would involve substantial techni-
cal risk and cost. It assessed both the Dutch navy version and
that being offered to Australia and judged both designs to be fair
contenders.
The other German design, the IKL Type 2000, generally met
most of the navy’s expectations. It was considered to be an excep-
tionally quiet submarine, with well-insulated machinery spaces
and substantially exceeding the speed at which the onset of cavita-
tion was expected. The proposal featured an outstanding logistics
support plan, promising 80 per cent availability, a great improve-
ment on that of the Oberons and exceeding the requirement by
around 18 per cent. The board was advised that some of IKL’s per-
formance figures were overly optimistic, and revised downwards
the German’s claims for endurance and range by about 8 per cent
while increasing the claimed indiscretion rate by 6 per cent, so
that the boat no longer met the requirement.25
The board was satisfied that IKL could rectify all the short-
comings it had identified and do so for less than the cost of the
changes it had stipulated for the others. The board was impressed
by the overall quality of the German bid and rated it a very strong
proposal.
Thus the evaluation ass
essed one proposal as very strong but
of the remainder only one was rated fair and two were marginal.
68
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
Even the favoured IKL design failed to meet all of the required
ship’s characteristics. Considerable additional effort would be
required during the definition study to design the submarine that
the navy wanted. The problem confronting the Tender Evaluation
Board was to identify the companies that could best undertake
that task.
The invitation to express interest in supplying the combat
systems for the new submarines attracted five acceptable pro-
posals. These were from the Dutch firm of Hollandse Signaal-
apparaten (Signaal); a German-British consortium of Krupp Atlas
Elektronik and Ferranti Computer Systems; the British company
Plessey; the French company Sintra Alcatel; and a consortium
led by the American Rockwell Corporation, consisting of Singer
Librascope, French sonar manufacturer Thomson CSF and the
partially Australian-owned software company Computer Sciences
of Australia.
While the initial evaluation of the submarine designs left the
final decision up in the air, the choice of combat system seemed to
be far clearer.
There were some important limitations to all the combat sys-
tem proposals. The request was for designs based on a distributed
architecture. Yet there was no widely accepted definition of what
constituted distributed processing, merely a generally accepted set
of characteristics: that particular functions were not dependent on
a single processor, that system databases were replicated around
a network, and that there was automatic backup around the net-
work in the event of a partial failure.26
The requirement also called for the system to be written in
Ada, an untested software language defined by American defence
development guidelines since no functional specifications existed.
Harry Dalrymple recognised that the selection of systems language
was crucial because it would drive the future costs and support
requirements of the combat system, so, after an evaluation of pro-
gramming languages by DSTO, each supplier was asked to provide
prices for using Ada or alternative programming languages.
Faced by conceptual novelty and technological risk, none of
the tenderers used a distributed architecture and two chose not to
