On 16 September 2021, the Australian government announced that it would acquire a nuclear-powered submarine capability with support from the UK and the US as the first measure of business under the AUKUS technology-sharing partnership. At the same time, it announced that it had established a taskforce that would devote 18 months to determining the ‘optimal pathway’ to establishing this new capability.

The taskforce has its work cut out for it, and the signing of an initial nuclear information sharing agreement only two months after AUKUS was announced suggests things are moving fast. Nevertheless, this new enterprise will be a massive undertaking and probably the largest and most complex endeavour Australia has embarked upon. The challenges, costs and risks will be enormous. It’s likely to be at least two decades and tens of billions of dollars in sunk costs before Australia has a useful nuclear-powered military capability.

Many commentators have suggested that the work of the taskforce is primarily about making a recommendation on the choice of submarine—either the US Virginia class or the UK Astute class. That’s misleading on two counts. First, the most important decision isn’t so much about the submarine, but about the strategic partner most able to work with us on our new SSN capability. Second, Australia will need to make many choices—about the strategic partner, about the submarine design, about the build strategy, about the schedule, and more. Those choices will involve hard prioritisation decisions about what’s most important. Is it capability, schedule, Australian industry content or something else?

A new ASPI report, released today, examines the decision space available to the government.

The most important decision is the choice of our primary strategic partner. While both the US and the UK will need to provide us with assistance regardless of which submarine design we choose, there’s no point picking a boat if its parent nation doesn’t have the capacity to assist us with all of the fundamental inputs to capability needed to deliver military effects, or its industrial base doesn’t have the capacity to deliver. While we shouldn’t pre-empt the work of the taskforce, initial analysis suggests that the US has more capacity to assist us.

Even once we choose a partner, we still have some difficult choices about the submarine design. Do we prioritise schedule and build our partner’s current design, with the result that we’re left with an orphaned and outdated fleet? Do we wait to get into step with our partner’s next class, exacerbating the risk of a capability gap? Or do we start with the current design and transition later to the partner’s future design, with the result that we have multiple classes of boat in our small fleet?

Another area of choice is the amount of modification we do to the design. While some modifications will be necessary due to Australia’s regulatory and safety regimes—unless we recognise our partner’s regulatory approaches as fit for purpose and accommodate ours to theirs—others will be discretionary. Every effort must be made to limit the changes, whether they’re motivated by capability or Australian industry content, as every change drives cost and schedule risk regardless of how well intentioned it is.

A fundamental choice is the build strategy. The government has stated that the SSNs will be built in Adelaide; however, it hasn’t committed to a continuous build. A continuous-build approach (one driven by a schedule designed to replace the first boat after around 30 years with no break in production) is appealing to Australian industry and current and future workers but would face many challenges—boats would be delivered on an inefficient three- or even four-year drumbeat, driving up cost and increasing the capability gap. Alternatively, an ‘economic’ approach that focused on the most efficient possible approach would deliver capability faster but require massive annual spending and produce the prospect of a ‘valley of death’ at the end of production. Either approach replicates a nuclear-submarine production capability in Australia as our own sovereign cottage-industry version of what the UK and the US already have.

Underpinning all of those choices is the issue of schedule. We’re facing the looming spectre of a submarine capability gap as the Collins-class fleet ages out. The government’s stated schedule, delivering at least the first boat by the late 2030s, is feasible, if optimistic, if we build all the boats in Australia. But that may be too late to avoid a capability gap.

An alternative that could accelerate delivery would be to open up the aperture of what ‘built in Adelaide’ should look like. One approach we consider in our report is to build the initial boats overseas to accelerate delivery but also to train Australian workers on a mature production line to avoid a ‘cold start’ to local production.

But there are also options that approach this enterprise as the embodiment of our AUKUS and alliance partnerships rather than as a large, but traditional, construction project. This could involve a ‘JSF’ (Joint Strike Fighter) approach of feeding components into a joint submarine enterprise that Australia enters with our primary enterprise partner. Such a joint enterprise would span more than just construction and enable Australia to be a ‘first line’ sustainment hub for our AUKUS partners’ submarines as well as for our own. When we consider that just the maintenance of Australia’s SSNs will be likely to require more workforce than the Attack-class build and Collins-class full-cycle dockings combined, there are ways to deliver submarines faster, sustain sovereign capability, contribute to our partnerships and still create jobs that don’t involve assembling submarines here.

And this can’t be stated clearly or often enough: successful transition isn’t about delivering just boats but all of the elements of the capability. So, the choice of boat and build approaches aren’t just matters of capability or industry but must be informed by broader inputs to capability. Perhaps the most important of those is the challenge of how the navy ramps up its uniformed workforce; solving that problem is just as crucial as delivering boats on time. Again, AUKUS has much to offer here.

Finally, we’ve provided an estimate of the cost of the enterprise. Such an exercise is inherently hazardous at this stage of the process, with so many assumptions still open and untested. The government has been open in stating that the SSN program will cost more than the Attack-class program, which would have cost around $56–57 billion in current-day constant dollars.

We agree: at an absolute minimum, an eight-boat SSN program will cost around $70 billion in constant dollars (or $116 billion in out-turned dollars, which account for inflation).

However, it’s highly likely that it will cost substantially more once the cost drivers are more clearly understood. Those include both the US and UK moving to bigger submarine designs, our choice of build strategy, and the broader support system and infrastructure needed to operate nuclear submarines. To channel Donald Rumsfeld, there are things we know we don’t know, and things we don’t know we don’t know; both will drive up the estimate.