For most of the past decade, Australia’s six Collins class submarines have provided the nation with a fine underwater warfare capability. Designed to meet operational requirements beyond the capabilities of other conventional submarines, the government of the day has been able to rely on the availability of several highly capable submarines.  

But it hasn’t always been that way. In fact, it was nearly 20 years after the delivery to the navy of the first of type, HMAS Collins, before the navy and government could reliably plan around the availability of submarines that were fit for purpose. For a long time, more than $10 billion of investment had provided the nation with very little payback in terms of a credible defence capability. Submarines spent years out of commission and availability across the fleet dwindled—sometimes to near zero.

The story of how the Collins class was finally brought up to a high standard of reliability is the subject of my new ASPI report, Nobody wins unless everybody wins: the Coles review  into the sustainment of Australia’s Collins class submarines. 

The problems with the Collins class began early. The building and delivery of the boats became highly politicised and the negative aspects that can reasonably be expected from any complex project—cost overruns, schedule slippages and technical problems that take time and effort to resolve—became consistent headline fodder. ‘Dud subs’ was a label that stuck. 

That perception could have been put to bed once all the bugs had been ironed out and the submarines successfully put into use. After all, the original project problems with the F-111 were also well publicised and the subject of political point-scoring and public discontent with the cost. But once in service, the fuss died down, and the F-111 became something of a favourite at air shows and public displays—to the point of there being quite a bit of grumbling when they were retired. 

Instead, in the years following the delivery of the last of the Collins class in 2003 and the resolution of most of the remaining technical issues, a new set of negative headlines started to appear. Partly through publicly available sources and partly through leaks from disaffected people within the submarine service and industrial support base, the parlous state of the fleet was revealed. A series of studies and reviews over the years—as many as 18 by some counts—identified issues and made recommendations but nothing seemed to make a lasting impression.  

Within the world of submarine management, relationships between the operator (the navy), the contractor (ASC, a government-owned enterprise based in Adelaide owned by the Department of Finance) and the group within the Department of Defence responsible for project management (the Defence Materiel Organisation—DMO) had become highly dysfunctional and occasionally outright hostile. 

By 2011 it was clearly that something drastic had to be done. The long lead time for the Collins class replacement meant that discussions for the successor class needed to start and early decisions about the direction of that project would be required. But it simply wasn’t possible to get traction on another very expensive submarine project when the existing fleet was essentially moribund—it would seem a case of throwing good money after bad. 

The answer was for the government to sponsor yet another review. Many observers at the time (including me) were sceptical that another such exercise was unlikely to do much more than rearrange the proverbial deckchairs. In my ASPI role at the time I spoke to many of the key individuals involved. Opinions varied on many points of detail, but there was a unanimous view across the entire enterprise that it was someone else’s fault. Many meetings were called, with the usual outcome being that the warring parties would have another meeting at some future time. In between those meetings, metaphorical wagons were circled and arguments marshalled for future stoushes.  

Many of the key players are quoted in my new report, and their views are often expressed quite candidly. That’s because at heart the woes of the submarine fleet were due to organisational issues and the breakdown of trust between the parties. The maintenance people at ASC felt that they were being micromanaged by the DMO to the point of not being able to do what they needed to do. The DMO felt that ASC were taking advantage of a contract that paid them whether or not the submarines were being maintained efficiently. The navy, ultimately the customer for submarine availability, didn’t have a clear understanding of what it could expect from a fleet of six submarines—and didn’t have a clear statement of its requirements in any case. 

The review team that was put together was headed by John Coles, a British naval architect and engineer who had already established a track record for sorting submarine maintenance problems in the United Kingdom. The team was formidable—it wasn’t a consultancy group of generalists, but a small group of career professionals with hard-won specialist knowledge in submarine sustainment. That was to prove a crucial factor in the success of the review—the deep understanding of the business within the review team meant that nobody could pull the wool over their eyes.   

It didn’t take Coles and his team long to work out that the Collins class submarines were fundamentally sound but the system (being generous with the term) put in place to support them wasn’t fit for purpose. And they also found that everyone wanted to fix the problem but that nobody agreed on what the causes of the problem were. 

The whole story was a fascinating one to pull together. I’d tracked the facts and figures over the years, often reporting them here on the Strategist. And I’d been buttonholed at ASPI and Defence functions by enough people anxious to explain why those other guys were to blame to know that there were deep cultural divides at play.

The new ASPI publication explains how the ‘Coles review’ (as it became universally known) solved the problem. There was no magic formula: the answer was to work out in detail what needed to be done, who was best placed to do it, and then make sure that the delegated authorities and accountability mechanisms were in place to let them get on with it. The title of my report comes from Coles’ approach of making everyone a winner by enabling them to kick the goals they needed to kick while others did the same. 

The improvement that followed the review and its implementation was swift and dramatic. The review did a comparative study of six other submarine-operating nations and found that Australia’s performance was by far the worst. In a couple of years, that had turned around to the point where Collins submarine availability exceeded the international benchmark, to the relief and delight of everyone involved. 

There’s much more to the story than this brief overview, and there are lessons that go well beyond the narrow task of submarine sustainment. I hope the report gets widely read within the departmental and industrial organisations responsible for managing defence projects and delivered capabilities.

In particular, I hope those charged with the even more demanding task of preparing for Australia to get into the business of supporting, operating and one day building nuclear-powered submarines appreciate what they are getting into. As the experience with the Collins class showed, buying the hardware—however expensive and complicated that process can be—is only the beginning.

While we await the AUKUS partners’ announcement of their plans for Australia to acquire nuclear-powered submarines (SSNs), the public discourse on the agreement has become wildly out of sync with reality.

Charges that Canberra’s SSN program might promote conflict with Beijing or start a nuclear arms race need to be evaluated far more critically. Key to understanding the impetus for the trilateral agreement are considerations driven by the regional dynamics of military acquisitions as well as international agreements on the control of nuclear material.

Uncritically accepting the positions put forth by Beijing and Moscow would not only mean Australia abrogating its sovereign right to make decisions about its own national defence but also be detrimental to the interests of Southeast Asian nations. Acting Prime Minister and Defence Minister Richard Marles and opposition leader Peter Dutton both told parliament yesterday that acquiring the submarines was about maintaining peace in the region rather than going to war.

Submarines have been a regular presence across the Indo-Pacific for decades. Indonesia, one of the more vocal critics of Australia’s plans, operated the largest submarine fleet in the region in the 1960s and 1970s with a force of 12 Whiskey-class diesel-electric submarines obtained from the Soviet Union. Today, its submarine force numbers four, following the tragic loss of KRI Nanggala in 2021. The Royal Malaysian Navy operates two Scorpene-class submarines. Singapore has four submarines and has placed orders for an additional four modern boats. Vietnam has six modern Kilo-class submarines purchased from Russia. Thailand and the Philippines are actively pursuing submarines, although Bangkok’s planned purchase of Chinese Yuan-class boats appears to be foundering. So, in maritime Southeast Asia, only Cambodia and Timor-Leste aren’t operating or actively pursuing submarines as part of their maritime strategies.

It seems safe to assume that none of these states consulted Canberra while developing their submarine programs. However, both Malaysia and Indonesia have voiced their opinions that Australia’s pursuit of SSNs risks a regional arms race—while remaining painfully silent on the subject of an unprecedented naval expansion to the north being carried out by the People’s Liberation Army Navy. Somehow submarines that don’t even exist yet have caused more regional consternation than a scores-strong fleet that is not only nuclear-propelled but armed with nuclear weapons as well.

It is the height of hypocrisy for the Chinese government to hyperventilate over a nuclear-powered submarine program in Australia when the PLA Navy has commissioned at least 15 nuclear-powered boats in the past two decades, several of which carry nuclear-tipped intercontinental ballistic missiles. The Chinese sub fleet is projected to grow by at least 10 by 2030. Under Australia’s plan, the Royal Australian Navy may only start receiving its new submarines by that point. In the meantime, Australia’s ageing Collins-class boats will likely need to serve well past their intended decommissioning dates. In fact, one of the key questions that the AUKUS partners are expected to answer next week is how Australia will avoid a submarine capability gap.

Beijing’s disingenuous talking points have also surfaced in Russian circles. Russian officials have implied that an Australian SSN  program will weaken the international non-proliferation regime and contribute to an acceleration of the imagined arms race. In fact, the Treaty on the Non-Proliferation of Nuclear Weapons doesn’t restrict non-nuclear-weapon states from developing ships or submarines powered by nuclear reactors; nor does it have provisions for overseeing the nuclear material committed to such a program.

It is equally incredible that the Russian government would criticise a state for embarking on an SSN program when the Russian Navy already operates a significant fleet of SSNs itself and has leased some of them to India. Having repeatedly threatened to use nuclear weapons as part of its illegal invasion of Ukraine, Moscow should be perhaps the last capitol on earth to criticise Australia for seeking a completely legal nuclear-propulsion capability for its own navy.

Australia has taken great pains to assuage concerns across Southeast Asia, reiterating its commitment to the Nuclear Non-Proliferation Treaty and attempting to correct misperceptions about the nature of AUKUS—namely, that it is a technology-sharing mechanism, not an operational network or alliance.

Technology sharing among partners is normal across the globe, including among the states that are the biggest critics of the AUKUS partnership. It appears, rather, that those states using their navies to impose their will upon their neighbours, from Russian naval strikes against Ukraine to Chinese ships enforcing illegal claims at sea, are effectively demanding that the world ignore their aggression and remain vulnerable to their military expansions.

With a full grasp of the facts, it’s plain that accusing Australia of contributing to an arms race is absurd, particularly when an aspiring regional hegemon already has 11 times Australia’s submarine strength in a navy seven times bigger than the RAN. An arms race requires at least two competitors, but there’s only one state in the Indo-Pacific that’s sprinting toward expanding its naval power—and that is China.

Asia has enjoyed a long period of relative peace and prosperity that is increasingly jeopardised by China’s intimidation campaign at sea and by its threats of violent invasion of Taiwan. Some regional powers, such as India, Australia, Japan and the United States, are eager to deter China from undertaking a catastrophic war of aggression that would upend the world economy. The measures required to deter China—which include developing modern military capabilities commensurate with those already possessed by Beijing—will necessarily generate some discomfort in the short term as China attempts to intimidate states into inaction. However, rather than adopt Beijing’s talking points, regional states might well consider what the regional order would look like should war not be deterred.

AUKUS holds out the promise of an extension of the Asian peace and a reduction in the likelihood that China will use its rapid and massive military expansion to resort to naked aggression. The AUKUS initiatives are designed to foster peace and stability in the region by maintaining the long-term regional military balance, which is surely worth enduring a bit of short-term discomfort.

You can have a cheap submarine capability, or you can have a safe and effective one. There’s no third option. Australia has chosen the second option; consequently, its submarine capability costs a lot. But how much exactly? We’re all now very familiar with the $89 billion price tag on the future submarine, but that’s a number that stretches far off into the distance, and it doesn’t cover all aspects of the capability.

Perhaps a more useful way to answer the question is to focus on the annual spending on Australia’s submarine capability. Parts of the answer are easy to find, but as with all areas of Australian Defence Force capability, it’s difficult to develop a comprehensive number. It’s particularly hard if we’re trying to include all elements of capability—or the nine fundamental inputs to capability, to use the Defence Department’s term—such as facilities. It’s hard to know how much of the $3,188 million spent annually on estate maintenance and garrison support should be attributed to submarines, for example. Or how to identify indirect costs such as educating future submarine officers at the Australian Defence Force Academy. So we’ll limit our analysis to direct costs that can be attributed to submarines with reasonable confidence.

The sustainment cost of the navy’s six Collins-class submarines is the easy bit. That’s spelled out clearly in Defence’s budget statements and annual report. It’s averaged $574 million per year over the past 10 years but has been increasing. It’s predicted to be $671 million this year.

However, that number doesn’t include some key elements. One is fuel. A full tank of diesel will cost several hundred thousand dollars. Another cost that isn’t included is even bigger, namely the uniformed personnel operating and maintaining the submarines. While each submarine has a crew of around 55, a much larger number than six lots of 55 is needed to have a robust, sustainable workforce. The navy has done well in increasing the number of its submariners over recent years and Defence informed the Senate earlier this year that that total had reached 881, although there were still some shortfalls. ASPI analysis (page 70) concluded that the average cost of each permanent ADF member was $160,000 a year five years ago, but submariners receive special allowances and retention bonuses so we could be looking at $250,000. Overall, the cost of the uniformed submarine workforce is probably more than $225 million.

The Collins sustainment cost includes full- and mid-cycle dockings in which the submarines are taken out of the water for overhaul, but it doesn’t include the cost of major upgrades that are installed during those dockings. Those are listed separately as acquisition projects in the SEA 1439 portfolio of projects. Here’s where numbers can get tricky. Since the Defence budget statements include only the department’s top 30 projects by planned spending for the year, we don’t know projects’ spending if they don’t make the cut-off, which was $97 million this year (although it’s been lower at around $50–60 million in recent years). If a project doesn’t ever make the top 30 it’s hard to know that it even exists, let alone what it’s spending.

Projects currently running include SEA 1439 Phase 6, which is upgrading the Collins’ sonars. It spent $133 million in 2020–21 but doesn’t make the top 30 this year. There’s also SEA 1439 Phase 5B2, which is an upgrade of the Collins’ communications and electronic warfare suites. It spent $64 million in 2018–19 and $70 million in 2019–20. There are other phases of SEA 1439, but we can’t see what they are spending. This year the budget statements also include Joint Project 9013, which is upgrading the Collins’ satellite communications system with predicted spending for the year of $101 million. We’ve collated the available data in our Cost of Defence database. Let’s lump all of the Collins upgrade projects together and call it $300 million a year.

With around $670 million for sustainment, $225 million for workforce and $300 million for upgrade projects, the Collins class’s direct costs are in the order of $1.2 billion per year.

But since we’re looking at the cost of Australia’s submarine capability, not the Collins per se, we also need to include the cost of the future submarine. Annual spending on SEA 1000, the project designing and building the Attack-class submarine, has been steadily growing. Even though we’re still several years away from starting construction, spending hit $719 million last year and is planned to pass $1 billion this year.

So the direct cost of our submarine capability is around $2.2 billion this year. That’s about 5% of Defence’s $44.6 billion budget. That’s the cost. The fundamental question is whether that presents value. That’s hard to address in one post. Under the Collins’ usage and upkeep cycle, that spend results in two deployable boats. But in light of the long transit distances to where our submarines operate, that model can’t ensure one boat on station. On the other hand, even the prospect of encountering one capable submarine can generate significant uncertainty for any adversary.

The annual spend is going to increase considerably—and quickly. In the near term, as the future submarine project ramps up and transitions from design to construction, it’s reasonable to assume that its annual cash flow will pass $2 billion by the middle of this decade. It’s important to realise that now that we’re in the era of continuous naval shipbuilding, those costs will never ramp down; we will always be designing and building submarines, so we will be paying something like $2 billion per year (in current dollars) until submarines go out of fashion.

The current round of Collins upgrade projects will wind down, but they will be replaced by the Collins life-of-type-extension program in the second half of this decade. While the LOTE is based on a full-cycle docking, its scope is going to be much larger, including new main motors, diesel generators and electrical distribution systems, so its costs will be much higher. How much higher isn’t clear, although the 2020 force structure plan has a funding line for the LOTE of $3.5–6.0 billion (page 45). It’s hard to see its cost being less than the current program of upgrades, so it’s likely we’ll still be spending at least $300 million per year once the first LOTE installation starts in 2026. The number of submariners will need to keep growing in anticipation of the arrival of the Attack class.

By the second half of the decade, we’re likely to be looking at around $3.5 billion per year for our submarine capability—which will be the same as it is now. That’s still some years before the first Attack class is deployable in 2034 on the current schedule.

In the longer term, costs will continue to grow as the Attack class enters service. Its increased sustainment cost will a major driver. The new submarines will be 40–50% bigger than the Collins and the fleet will be twice as big, so the sustainment cost will be at the very least three times that of the Collins—or around $2 billion (in today’s dollars) by the time all 12 boats are in service in the mid-2050s. Defence’s own numbers support this. It has informed the Senate that the sustainment cost of the Attack class will be $50 billion in current dollars; for a 25-year life of type, that works out at $2 billion per year.

Of course, the 2050s are a long way away. But there’s likely to be a 10-year period from the mid-2030s to the mid-2040s when we’ll be operating both the Collins and Attack submarines simultaneously, and that will bring significant overheads.

The future fleet will likely need more than twice as many submariners since the crew of each Attack-class boat will be bigger. Plus, overheads will increase—for example, through the establishment of a second submarine base on the east coast (see page 120 of the 2020 force structure plan).

So, between the operating cost of the future fleet and the ongoing cost of continuous shipbuilding, we could easily be looking at $5 billion per year in current dollars once we reach the mature Attack-class fleet. That’s more than twice as much as we’re paying now, but it should result in at least twice as many boats with greater capability available for operations, meaning more boats will be on station for longer.

The following extract is from ASPI’s latest special report, Submarines: Your questions answered, which attempts to answer the many questions that Australians pose when it comes to the design, acquisition, cost, operational service and strategic implications of submarines.

There’s a common rule of thumb in militaries known as ‘the rule of threes’: for every three platforms you have, you can deploy one. This is how the Australian Army is structured, and it broadly applies to ships and aircraft as well. Militaries can surge beyond this for short periods, but that’s not sustainable in the longer term due to deferred maintenance, lack of training, burnout of personnel and so on.

The rule of threes is reflected in the Royal Australian Navy’s current requirement for submarine availability. Of the six Collins-class submarines, two are in deep maintenance and upgrade. Four are available for service, with three available for tasking and one in shorter term maintenance. Of those four, only two are deployable.

The RAN’s submarine requirement: two from six

Source: John Coles, Study into the business of sustaining Australia’s strategic Collins class submarine capability: beyond benchmark—May 2016, Australian Government, Canberra, 2016, page 5.

That doesn’t mean two will always be in the area of operations. Since the likely areas for submarine operations are far from Australia, a large percentage of any deployment is spent in transit. The exact percentage depends on the distance, the kind of operation and so on, but the net result is that it’s unlikely that the navy can sustain one submarine on station in an area such as the Malacca Strait or the South China Sea for an extended period. If, in a future conflict, the operational concept isn’t to maintain a submarine on station but to conduct offensive operations at a time of our own choosing, it might be possible to surge to two submarines in the area of operations, but that couldn’t be sustained.

The requirement for 12 submarines first appeared in the 2009 defence white paper. It wasn’t driven by the Department of Defence but appears to have come from then Prime Minister Kevin Rudd. Since then, it has remained a constant in Defence’s strategic planning documents.

With 12 submarines, it’s reasonable to assume that four would consistently be available for operations, with greater capacity to surge. That would be likely to allow the navy to keep one submarine on station at a long distance from Australia while simultaneously supporting some presence closer to home.

There are factors that influence presence and time on station other than just numbers of submarines. Increasing the endurance of the submarine could potentially have a large impact. The Collins is reputed to have around 50 days’ endurance. If, for example, a 10-day transit is required to reach the area of operations, a Collins boat effectively has only 30 days on station. If the future submarine has 70 days’ endurance (a 40% increase), that would result in 50 days on station in the same scenario (a 67% increase). Put another way, two Attack-class submarines would provide greater presence than three Collins. Also, if the pump-jet propulsor that’s to be used in the Attack class can increase its transit speed without consuming more fuel, that could reduce transit times and further increase time on station.

Since nuclear-propelled submarines have much greater transit speed than conventional submarines and endurance constrained only by the crew’s resilience, a smaller number of nuclear submarines would provide greater presence than 12 conventional submarines.

Another factor is the Attack-class submarine’s pump-jet propulsion system. While Defence has been tight-lipped about the pump-jet’s performance, it could potentially result in greater time on station by reducing transit times.

Reductions in the amount of time spent in maintenance would also improve availability. However, the Collins submarines are now already meeting and exceeding international benchmarks, so it could be difficult to achieve significant improvements beyond the current availability requirement and thus escape the rule of threes.

The switch from creating a ‘Son of Collins’ to making a ‘son of Collins’ is a conundrum of Australia’s submarine saga.

The Defence Department abandoned the option of building a second generation of the Collins-class submarines long ago. And our partner in building the Collins, Sweden, wasn’t even considered in the contest (between France, Germany and Japan) to create the new submarine.

Yet today we are building a new version of the Collins through a life-of-type extension of the existing subs. After casting off the capital ‘S’ option we now clasp the small ‘s’ son of Collins.

The 2020 force structure plan says the cost of the son of Collins—extension plus sustainment—will be between $3.5 and $6 billion. In the way of subs, expect that $6 billion figure to grow. Insight Economics estimates the life extension for the Collins could cost $15 billion.

The sub option that didn’t fully surface is the Son of Collins, although it’s conning tower is visible as the son of Collins. Many factors fathered the decision not to do the Son.

First, politics, with its dimensions of dollars and debate, dithering and delay.

Second, the agonising process of turning the Collins from dud sub to beaut boat. The Collins sustainment was on Defence’s list of projects of concern for a record nine years.

Third, the quarrelsome marriage with Sweden. The legal battles over submarine intellectual property had divorce-court elements: a rerun of the relationship problems conducted as an argument about property and progeny.

Fourth, Defence’s fears about getting the expertise for the evolution to a next-generation boat. Subs need the right minds as well as lots of money.

On the politics of dollars and dithering, Labor’s defence policy platform when it won office in 2007 proclaimed that it’d accelerate work on Australia’s next generation of subs ‘ahead of the current timetable which schedules first pass approval for 2011’. Instead, we missed that target by five years. The Turnbull government did first pass in 2016.

Labor policy in 2007 thought ‘a developmental project involving the migration of evolved Collins class combat and ship control systems might be necessary’. By the 2009 defence white paper, Labor proclaimed the need for 12 new subs.

The stage was set for the Son of Collins. Yet zip happened. The first-pass window kept passing. The global financial crisis hit. Struggling to fix the Collins, Labor didn’t have the energy for the Son and adopted a son stop-gap.

If Labor didn’t act on a Son of Collins, the Liberals couldn’t or wouldn’t. The Libs made much noise about the Collins problems and Labor failures. When Tony Abbott won government in 2013, the Collins was more political pariah than the potential parent of the next-generation submarine. (For this chronology, see the parliamentary library’s new account of the subs story, building on its previous report from 2012.)

Beyond the politics, the conundrum centres on the thinking in Defence and the Royal Australian Navy. Why didn’t the navy want a Son of Collins? Why didn’t Defence put Sweden in the mix?

On those two questions, Marcus Hellyer (ASPI’s sage on the inner workings of the Defence mind) judges that excluding the Swedes ‘is one of Defence’s most bizarre capability decisions’. Bizarre, indeed.

Defence argued there’d been a hiatus in Sweden’s sub building and that gap posed an unacceptable risk. The claim that the game had moved far and fast was also deployed to attack the Son of Collins—developing what we had wouldn’t deliver what Defence said we needed: a brand new design.

As Hellyer writes: ‘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”.’

Australia had the intellectual property for the Collins, but seemed to doubt its intellectual and technical ability to create a Son of Collins. Defence feared we didn’t have the critical mass of expertise to design and build a new boat.

Naval sage James Goldrick emphasises an old line offering a difficult truth: ‘The greatest restriction on naval expansion is draughtsmen not money’.

Canberra worried that it had the money but not the minds. That informs the whispered response to the criticism that Australia should be running a competition between a Son of Collins and the French-designed Attack class. Defence fears it’ll be fiendishly difficult to get the skills and smarts to achieve just one boat design.

As Goldrick told me:

The French may have realised the potential benefit to themselves of this process earlier than anybody else—apart from the fact that their boat was the best, according to the final Australian evaluation.

What is happening all over the world is an increasing problem of continuity for the evolution of design because that requires there to be continued work. Almost nobody is building enough submarines, frequently enough, to be self-sustaining as a centre of design and enterprise.

Even if you are building continuously, if you have a big break in your design effort, it’s very difficult to recover, as the British and even the Americans and Russians have experienced.

Association with the Australian continuous build/batch upgrade scheme would help the French maintain critical mass and sustain their design skills.

So, the Son of Collins didn’t surface. But the stretching Attack-class timeline means we’re now committed to the son of Collins. The saga has many bizarre twists.