AUKUS Submarines: U.S. and U.K. Partners Show the Realities Facing Australia's Small Fleet

In 2056, 30 years from now, when Australia has eight nuclear powered submarines (SSNs), we will have two deployable submarines consistently available. That’s assuming all goes well with AUKUS’ “optimal pathway” and we do in fact have eight SSNs in service.

The problem with nuclear submarines is shared by all submarines: they are hugely complex electro-mechanical devices and they operate in extraordinarily demanding environments – both physical and threat.

However, nuclear submarines are even more complex than their diesel electric counterparts, compounding the maintenance and availability challenges. As the inaugural head of Australia’s Submarine Agency, Vice Admiral Jonathan Mead has said, nuclear submarines are “more complicated than the Space Shuttle or the International Space Station”.

Complex machines like modern submarines are not available for operations all the time as they comprise many systems and sub-systems that need maintenance. In addition to the everyday maintenance performed dockside, submarines require periodic overhauls that take them out of the water for one, two or even more years, depending on the particular overhaul and the capacity of their supporting industrial base.

Moreover, there is a vicious cycle of submarine availability and submariner availability—if submarines aren’t available, submariners can’t go to sea to train and qualify. But if you don’t have sufficient number of qualified submariners across the necessary critical skills, then you can’t take submarines to sea. So availability and deployability is a complex, interrelated challenge of both people and materiel that we can only skim over lightly here.

Australia’s six Collins class submarines have proven to be very capable diesel-electric boats. From a maintenance and availability perspective, the program overcame enormous challenges early in the fleet’s life, but only with concerted political attention and external intervention, notably through the Coles review.  

In the early 2010s Collins’ availability had crashed. The Coles review introduced a ‘usage and upkeep cycle’ that sought to achieve world’s best practice in availability. Under this cycle, two of the six boats are in longer term maintenance – in activities like full cycle dockings that take boats out of the water and pull them apart. The remaining four are available to the Chief of Navy. Three of these are consistently available for tasking while one is in shorter term maintenance. However, only two of them are consistently deployable. Defence has provided evidence at parliamentary committees that it has achieved this model (although there is no public data that would allow us to confirm actual submarine availability or days at sea).

In short, Australia’s experience is that when things are working well, and the Navy, the Department of Defence and defence industry are working well together, the 3 to 1 benchmark is achievable—but not more.

The Collins submarines are now ageing and about to be start going through the long-awaited forthcoming Life of Type Extension program (LOTE) to keep them operating through the transition to the new AUKUS fleet. Each submarine’s LOTE is meant to be conducted in a standard two-year Full Cycle Docking (FCD) slot. However, we are already seeing full-cycle dockings take longer than the planned two years, so each LOTE is almost certain to also take longer than planned due to the age of the boats and technical risks around replacing many of their major systems. This will likely result in decreased availability across the fleet, challenging achievement of the two from six deployability model.

The Collins submarine problems were compounded by submariner retention and recruitment difficulties that meant that even if a submarine itself was available, the Navy may not have had a complete crew available to operate it. These challenges were addressed through various workforce measures, including targeted retention bonuses.

However, the need to syphon personnel off into the SSN training pipeline means that those challenges could well return to the Collins class. Six or more years before Australian receives its first SSN, the Chief of Navy has stated that already more sailors are entering the nuclear submarine pipeline than the Collins pipeline, raising questions about the long term viability of the Collins workforce.

If there’s any lesson from the Collins experience, it’s that submarine capability, particularly in small forces, is fragile.

While the RAN has generally managed to achieve 3-to-1 outcome, a review of our AUKUS partners shows that they struggle to achieve this ratio in their SSN fleets. The situation in the US Navy has been covered in detail by the US’ independent accountability agencies, the Congressional Budget Office and the Congressional Research Service, agencies that consistently demonstrate the value of transparency in government – and the good use that can be made of relevant information routinely provided (by law) by the Pentagon to such agencies.

We should note that different submarine forces have different missions; a global power like the US employs its submarines differently from a navy with a coastal defence focus. This is a function of missions and technology.  A coastal defence focus is not the sweet spot for large, nuclear powered submarines.

Public sources state that the USN SSNs conduct long deployments of around six months roughly every 18 months (e.g. this from the US Naval institute). They do long deployments because of the USN’s global operating profile; they operate around the world far from home. SSNs accompany carrier strike groups that themselves deploy for long periods of time. They conduct surveillance in strategic locations or wait to employ their cruise missiles in combat operations. They trail other nations’ SSBN, the submarines that carry intercontinental ballistic missile, ready to destroy them should nuclear war start. Since the Australian Government hasn’t said what our SSNs will be doing, we don’t know what their mission profile will be. More on that later.

Six months every 18 months sounds like the 3-to-1 model. However a report by the Center for Strategic and Budgetary Assessments (whose authors include a former SSN engineer officer) uses a model that adds another six months of ‘pre-overseas movement’, in other words six months of deployment every 24 months, or more like 4-to-1. But neither of those ratios includes time out for deep maintenance.

Illustration of US Navy nuclear submarine usage cycle. Source: CSBA.

The Congressional Budget Office has written that Virginia-class SSNs undergo an extended docking selected restricted availability (EDSRA) about every six years. The first five EDSRAs were meant to average 450 day (about 15 months) but actually took 760 (over two years). Consequently, the USN revised the target to 19.7 months. When we add that time to the six year, we get three or four six-month deployments every seven-and-a-half or eight years. That gets us to a 4-to-1 or even 5-to-1 deployment model.

In early 2025 when the CBO modelled the impact on USN submarine capability of transferring SSNs to Australia it concluded that ‘the loss of those submarines translates to a loss of 65 operational years for the SSN force from 2032 to 2060 in Scenario 1 and a loss of 102 operational years over that period in Scenario 2. Those losses result in 28 fewer SSN deployments in Scenario 1 and 43 fewer deployments in Scenario 2.’

The analysis was based on the assumption that Virginia class submarines go on 14 deployments over the course of a 33-year service life.’ 14 six-month deployments over 33 years is firmly in the 4-to-1 or 5-to-1 territory.

Whether you are closer to 4-to-1 or 5-to-1 depends on the effectiveness of your industrial base. All the indicators are that the USN’s industrial base is not in good shape.

US attack submarine maintenance delays have increased over the last 15 years with around 1/3 – 34% in 2024 – of its SSNs in depot maintenance or idle. As the US Congressional Budget Office notes “the increase since FY2012 in the number of SSNs in depot maintenance or idle has substantially reduced the number of SSNs operationally ready at any given moment, reducing the SSN force’s capacity for meeting day to day mission demands and potentially putting increased operational pressure on SSNs that are operationally ready.” This strains crews  who are required to stretch deployments for longer than anticipated, and it also increases the maintenance load required for boats that operate for longer than planned, which worsens existing recruitment and retention troubles that the US Navy submarine force faces.  It creates a systemic problem.

To give a more granular flavour of the effect of the maintenance backlog on the US submarine fleet, here’s an excerpt from the 1 January 2026 Congressional Research Service’s report on the issues:

“In fiscal (year) 2016, because of idle time for subs awaiting maintenance—on boats which have exceeded their operational limits and were no longer allowed to submerge under the water until they underwent maintenance—the Navy lost about 360 days of operations.

In FY21, the fleet lost nearly 1,500 days to idle time—the equivalent of taking four submarines out of the fleet.

Additionally,…in FY21 the fleet lost the equivalent of 3.5 submarines to repair periods that ran longer than planned.

“That’s about seven and a half SSNs that I cannot use last year because of awaiting maintenance or maintenance delay,”’ said Rear Admiral Jablon, commander of the US Pacific Fleet’s submarine force.

The USN says that industry best practice is 20% in depot maintenance and none idle at any given moment, and that is its target. Unfortunately, the target has proven aspirational for a decade now.

Unfortunately, some individual submarines are experiencing much longer periods out of service in maintenance or awaiting maintenance. USS Boise is the evil poster child here – a Los Angeles Class SSN that has been out of service since 2015. Despite having a $US1.2 billion overhaul contract awarded to Huntington Ingalls in 2024, it now may be scrapped instead of returning to the fleet in 2029.

Also maintenance labour hours for Virginia-class submarines are proving higher than the hours required for their predecessor– the Los Angeles-class – despite the Virginias originally being designed to require less maintenance. And even more interestingly, the number of days in the shipyard for Virgina class submarines has been rising as each boat enters maintenance, according to Congressional Budget Office analysis.

In summary, the USN is doing somewhat better than 5-to-1 but even if it managed to unplug the bottleneck in its maintenance system, it still can’t do better than 4-to-1. That’s just the nature of this extraordinarily complex beast.

The UK’s Royal Navy nuclear submarine fleet is considerably more troubled than the US’. And it is a UK design that Australia will be getting for its mature SSN fleet under the AUKUS optimal pathway. UK submarine production challenges have been well-documented, but its maintenance system is also failing.

We are not aware of any public sources that state a target availability or deployability ratio or model for RN SSNs. What we can say is that based on the thorough reporting of the Navy Lookout site, actual availability has been extremely low. The RN is currently completing an extended transition from seven Trafalgar-class SSNs to seven Astute-class SSNs and over the past several years has had between five and seven submarines in commission. However the number it has been able to get to sea has been far below that.

In October 2025, Navy Lookout wrote, ‘there have been several periods in the last few years when the Royal Navy has not had a single SSN at sea. Currently, only one of the six boats in commission is operational, and four of them are at very low readiness.’

In July 2025, zero Royal Navy SSNs were at sea. In November 2024, three of six were at sea, however this follow a long period of limited availability. Moreover, ‘at sea’ does not mean deployable or on operations. In October 2024, a single SSN put to sea, which was the first since July 2024; in other words, not a single SSN had put to sea in three months.

HMS Anson, the Astute class submarine that came to Australia’s HMAS Stirling naval base a few weeks ago and has now departed, probably for the Middle East, has been the single UK attack submarine available for deployment.

The crisis in the Royal Navy and UK industry’s nuclear enterprise is systemic. Former director of UK nuclear policy Rear Admiral Philip Mathias has said that, “The UK is no longer capable of managing a nuclear submarine programme.”

“Dreadnought is late, Astute class submarine delivery is getting later, there is a massive backlog in Astute class maintenance and refitting, which continues to get worse, and SSN-Aukus is a submarine which is not going to deliver what the UK or Australia needs in terms of capability or timescale.

“Performance across all aspects of the programme continues to get worse in every dimension. This is an unprecedented situation in the nuclear submarine age. It is a catastrophic failure of succession and leadership planning.”

Put another way, it’s quite possible to do worse than a 5-to-1 ratio. And the implications for Australia’s AUKUS plan of relying on a UK designed submarine with major UK components are extremely concerning given the lived experience of the UK’s own submarine enterprise in the last decade.

Of course, this entire analysis has assumed Australia obtains its eighth SSN sometime in the late 2050s. It’s a letter to our grandchildren about what we hope to leave them.

Before then under the optimal pathway, Australia will operate 1-5 Virginia Class submarines alongside some remaining Collins submarines, transitioning to a mixed fleet of Virginia class submarines and SSN-AUKUS boats that will total less than eight boats.

It is a courageous assumption to think that as an SSN neophyte Australia will operate its nuclear submarines at availability and deployability levels exceeding its two AUKUS partner Navies who have decades of experience with nuclear attack submarines, particularly when Australia will largely be reliant on the same stressed industrial bases for components that they are. So it is entirely realistic to expect that our best case is around a 4-to-1 ratio.

That means that until the RAN has three Virginia submarines (sometime in the late 2030s under the optimal pathway), we should not expect the Chief of Navy to be able to consistently deploy more than one nuclear submarine. Deploying two SSNs at least some of the time will likely require a fleet of six or more boats; that is not going be achieved until around 2050. And consistently deploying two will likely not occur until well into the 2050s.

That’s probably a best case scenario. Australia is at the start of its nuclear submarine learning curve and plans to operate two different types of these complex machines simultaneously. Small fleets of eight submarines are challenging even when they’re a single type. Operating two types of nuclear attack submarines designed by two different nations with two different supply chain and training pipelines adds to this challenge and is likely to moderate any aspirational levels of availability form the AUKUS fleet.

Of course it is possible to surge capability in time of crisis. But there are clear limitations to this. First, submarines in deep maintenance are out of the water with key components removed. Second, surging simply defers maintenance and training. In the short term this may be viable, but we are seeing that modern conflicts are not short due to the resilience of state and non-state actors. Considering AUKUS was designed to counter China, the state with the most robust industrial capacity in the world, it would be rash to assume that any conflict with China would be resolved before our surged SSNs needed their deferred maintenance and their crews rest and rehabilitation.

Mathematics, analysis and paying attention to the hard-won lived experience of our AUKUS partners matters more than cheerleading and faith when it comes to building, operating and maintaining the most complex machines humans have yet manufactured. Maybe that’s why Ewen Levick raised the basic value for money and opportunity cost issue around AUKUS, and the implications this endeavour has for what other military power Australia can afford in an era when asymmetric capabilities are proving so effective.

This issue once again shows the importance of transparency that can only be delivered by requiring bureaucracies to disclose information and by funding independent entities with the time and mandate to analyse that information.

Fortunately, for those interested in these issues, even in the open source arena there is a large amount of data and analysis available from both our AUKUS partners’ governments and their think tank communities. Key elements were used in this overview, notably from the US Congressional Research Service, the Congressional Budget Office and the Center for Strategic and Budgetary Analysis think tank. In the United Kingdom, there is the Infrastructure and Projects Authority and the (unpaid) Navy Lookout think tank. Australia is fortunate in having recent analysis of the Collins experience by Andrew Davies published by the Australian Strategic Policy Institute. 

We shouldn’t rely on the assertions of current or recently departed politicians and officials and their assurances that everything is on track and going ‘full steam ahead.’ Our mantra must always be ‘show us the data.’ Unfortunately, compared with the UK and US sources, transparency in Australia is limited by the reluctance of ministers and officials to release even unclassified information to allow external, independent assessment.

Marcus Hellyer is Head of Research at Strategic Analysis Australia.

Michael Shoebridge is founder and director of Strategic Analysis Australia. From 2018 until September 2022, he was the Director of the Defence, Strategy and National Security Program at the Australian Strategic Policy Institute (ASPI) in Canberra.

This article appeared originally at Strategic Analysis Australia (SAS).