Two questions are frequently raised about the process that selected Australia’s preferred partner for the design and build of the future submarine.

The first is why the Defence Department didn’t pursue a ‘son of Collins’, that is, an evolution of the successful Collins design, with the questioner’s underlying assumption being that evolving an existing design should be cheaper and faster than starting an entirely new design. The second is why the Swedish shipbuilder Saab, which had acquired Kockums—the company that designed the Collins—wasn’t invited to participate in the competitive evaluation process (CEP), given that it (other than the Japanese) was the only entity with demonstrated experience in designing and building large conventional submarines.

It’s worth revisiting these questions because they’re relevant to the Collins life-of-type extension (LOTE) program, which is the key to Defence maintaining an effective submarine capability throughout the long transition ^[1] to the Attack-class submarine.

The two questions are distinct but related. The Australian National Audit Office’s 2017 report ^[2] on the CEP is silent on why Saab wasn’t offered the opportunity to participate in the competition. The report simply says that Defence determined that the three entities that were invited to participate (TKMS of Germany, the Japanese government, and the ultimately successful DCNS of France) were the only ones that met Defence’s requirement that the future submarine be ‘designed and built by a proven submarine designer with recent experience in designing and building diesel-electric submarines’. The report doesn’t assess whether that was an appropriate requirement or explain why Saab didn’t meet it.

But there was a more fulsome discussion at Senate estimates ^[3] in February 2015, only two weeks after then Prime Minister Tony Abbott announced the CEP, followed by Defence’s written response ^[4] to the Senate’s questions on notice. On the first issue, Defence testified that a study into the possibility of evolving the Collins ‘demonstrated that the design effort involved would be similar to a new design’. Ultimately Defence concluded that an evolved Collins ‘would not provide a beneficial, nor a low cost and low risk solution for the Future Submarine’.

On the second question, Defence officials repeatedly argued ^[3] (pages 109 and 129, for example) that while the French and Germans had not designed or built a large conventional submarine, they along with the Japanese had demonstrated continuous submarine design and build activity.

In contrast, the Swedes, despite previously designing and building submarines both large and small, hadn’t completed a full design and build program since 1996–97. Even though they were well into the design of their own A26 class, had been involved in the build of the Collins into the 2000s, had an extensive record of major upgrade activities (including inserting entirely new ‘plug’ sections into submarines), and employed over 3,000 naval and submarine engineers ^[4], that hiatus was judged to be an unacceptable risk.

Whether or not you’re convinced by Defence’s reasoning to exclude the Swedes (for me, it is one of Defence’s most bizarre capability decisions), its argument is clear—submarine design, even modifications to an existing design, is difficult, and deep expertise is needed to undertake it successfully. There are no easy wins in submarine design, and mistakes create major cost, schedule and capability risks.

Which brings us to the Collins LOTE ^[5]. To recap, the Collins submarines were meant to be progressively withdrawn from service every two years from 2026. Since it’s been clear for some time that the future submarine wouldn’t enter operational service until the 2030s, again on a two-yearly cycle, some of the Collins fleet would need to undergo a LOTE to avoid a capability gap. In essence, the LOTE is the mitigation strategy to address the schedule risk in the future submarine program.

It’s also been clear that the LOTE would be based on an additional full-cycle docking, which means taking a Collins boat out of the water for two years of deep maintenance and upgrades so it can keep operating for the next 10 years. What hasn’t been clear is how many of the six Collins would need to undergo a LOTE and what its scope would be.

As the future submarine schedule has developed, with the first Attack-class boat now not expected to be operational until 2034, Defence’s Senate estimates testimony has moved from saying one to three LOTEs would be required to five. Theoretically, doing all six Collins could mitigate a further two-year slide in the program, but beyond that total submarine numbers could fall below six.

Defence has also started to reveal the scope of the LOTE—see here ^[6] (page 31) and here ^[7] (pages 17–22). In addition to all the usual maintenance and obsolescence management of a full-cycle docking, Defence wants to replace the Collins’ main motor, diesel generators, and electrical conversion and distribution system with new hardware made by the suppliers for the future submarine.

Interestingly, Defence has also said that these are three of the five most important systems on the future submarine. It is also looking at mast and sensor updates (for example, replacing periscopes with modern digital optronics ^[8] masts) as well as combat system updates.

In short, the LOTE concept is starting to look a lot like a son of Collins ^[9]—which Defence told the Senate in 2015 wasn’t worth the cost and risk involved. This poses serious questions about Defence’s risk-mitigation strategy for the submarine transition.

First, the Collins maintenance cycle is a finely tuned process. Between full-cycle, mid-cycle and intermediate dockings, two of the six boats are almost always out of the water in deep maintenance. If the amount of work required for the LOTE exceeds the two-year window for full-cycle dockings, Defence will have to choose between having more than two boats out of the water—with a consequent impact on the number of boats available—or deferring LOTEs and continuing to operate 30-plus-year-old boats with mounting obsolescence and reliability issues. Can all the regular maintenance plus replacement of three of the submarine’s five major systems fit in that two-year window?

Second, this potentially poses second-order and third-order effects on the submarine transition. For example, to train the much larger uniformed workforce needed to operate the larger and more numerous future submarines, Defence needs boats in the water. If submarine numbers or availability falls, submariner numbers won’t increase at the rate required to transition to the eventual 12 boat future fleet.

Third, who has the design expertise to replace three of the five major systems on the Collins? ASC, which maintains the boats, is the design authority for the Collins, and has started early design work ^[7] on the LOTE. But it has never done anything like this, and—as the government’s shipbuilding program ramps up—ASC has been losing its engineering workforce.

Since the LOTE will use systems made by the future submarine’s suppliers, then presumably Naval Group, the future submarine’s designer, will need to be involved. Defence may also need to involve Saab, which inherited Kockums’ Collins design pedigree, but Saab could have its hands full should it win the submarine ^[10] programs ^[11] it is bidding for in Europe. And if Defence’s original assessment that getting the Swedes to evolve the Collins wasn’t worth the risk, how confident should we be that this is a safer approach?

It may be that the LOTE will ultimately be a straightforward affair despite Defence’s testimony that even small changes to submarine designs can have great consequences. And commonality of key systems between the Collins and the future submarine is probably a good thing.

But the prospect of Defence pursuing something like a son of Collins to mitigate the risks involved in designing and delivering a new submarine from scratch does give pause for thought.