The 2016 Defence White Paper (DWP) announced that the Future Submarine will be delivered under a ‘rolling acquisition program’. Initially we assumed this meant that submarines would be produced continuously under the same sort of scheme envisaged for surface combatants ^[1]. We weren’t alone in drawing such a conclusion; we’ve spoken with parliamentarians, defence journalists and industry people who’ve done likewise. But it’s since been pointed out to us that we may be incorrect, and on a close re-reading of the DWP we’re now not so sure. The relevant passage in the DWP appears in para 4.28:

‘To ensure no capability gap and the ability to progress development of a replacement submarine in the 2050s, the Government has decided to implement a rolling acquisition program for Australia’s submarine fleet.’

Note that there’s no concrete commitment to replace the submarines in the 2050s, only to have the ability to do so. In doing so, the government appears to have left open the option of leaving a gap between the end of the construction of the first twelve boats and the commencement of their replacement. That wouldn’t necessarily be a bad thing. As we’ve previously pointed out in the context of surface vessels, continuous production for a smallish fleet makes for uneconomically short vessel lifetimes.

But what about the use of the term ‘rolling’, surely that’s an unambiguous commitment to a never ending production schedule? Well, not necessarily. The term used in the context of the surface combatant fleet in the DWP is ‘continuous’ rather than ‘rolling’. We presume that White Paper’s authors intend the distinction to be meaningful rather than decorative. If so, the use of ‘rolling’ in this context is spectacularly misleading. As recently as 2013, here’s how Defence defined the term in one of its own publications ^[2]:

‘Rolling build program is the term used in this report to describe an ongoing shipbuilding project. This is where ships are built at a steady cadence… Unlike the start and stop nature of pre-defined build projects, these projects are set up to run for an indefinite period. Ideally, the project will be designed to build ships at a certain interval, for a certain life of type so that when the last ship is built, the first requires replacement.’

Nonetheless, as best we can infer, the term ‘rolling acquisition’ leaves open the options of both continuous and non-continuous build programs, while allowing the DWP to make the following reassuring but ultimately vague claim (para 4.122):

‘A rolling program of acquiring submarines will provide long-term planning certainty for Australian industry, allowing those Australian companies involved in the submarine program to invest in the capabilities needed to support their involvement in the construction and sustainment activities.’

If the goal really is to provide ‘certainty for Australian industry’, the government should clarify exactly what it means by the term ‘rolling acquisition’. As explained below, the choice between a continuous and non-continuous program entails multi-billion dollar consequences.

Building on our 2012 analysis ^[3] of scheduling the submarine replacement, we’ve been looking in detail at the transition of the Collins class to the Future Submarine, including the options of continuous and non-continuous build strategies. One of the critical factors is the production interval between new boats. According to the DWP, the Future Submarine program will deliver the new vessels between the ‘early 2030s’ and the ‘late 2040s to 2050’, which implies an average production interval of close to 18 months. Our detailed analysis will appear in the 2016 Defence Budget Brief along with an examination of the Future Frigate program. The key conclusions are as follows:

• Some of the Collins fleet will need to have their service life extended by three or more years to avoid gap in submarine availability.
• Because of the Collins maintenance cycle, a one-year or two-year production interval would allow a smooth transition from the Collins to the Future Submarine in terms of vessel availability, maintenance and crewing. In contrast, an 18 month production interval would greatly complicate vessel availability, maintenance and crewing during the more than decade-long transition.
• For a twelve vessel fleet, continuous production will result in a short life-of-type; 12 years for a one-year production interval, 18 years for an 18-month interval, and 24 years for a two-year interval. The need for frequent recapitalisation will be exceedingly costly. For example, assuming a unit cost of $2.5 billion per vessel (consistent with the Integrated Investment Plan), a continuous program with a one-year production interval will cost $2.5 billion per year, an 18 month production interval will cost $1.67 billion per year and a two-year production interval will cost $1.25 billion per year.
• A non-continuous production scheme can reduce capital costs by retaining vessels in service for longer, so amortising their cost over a longer time. The resultant savings are independent of the initial production interval. A 28 year life-of-type results in an average annual capital cost of $1.07 billion, and a 34 year life-of-type (a not too unusual figure ^[4], and which some of the Collins will likely deliver) would cost only $883 million a year.

But keeping in mind that the Future Submarine program is about keeping Australia safe rather than keeping industry happy, the most important factor to consider is how quickly we can field an expanded fleet of ‘regionally superior submarines’. Assuming, consistent with the DWP, that the first vessel is delivered in 2032, a one-year production interval would deliver the 12th boat in mid-2043. An 18 month production interval would deliver the 12th boat at the start of 2049 and a two-year production interval would deliver the 12th boat in mid-2054. Can it really be the case that we need to double the size of our submarine fleet at a cost of tens of billions of dollars, but have luxury of waiting until almost the middle of the century? It’s hard to reconcile the apparent sense of gravity with the extraordinary lack of urgency.

In terms of continuous build options, there’s no sweet spot. The shorter the production interval; the higher the annual capital cost. The longer the production interval; the greater the delay to achieving a twelve vessel fleet. Only by jettisoning a continuous build program can we achieve a timely buildup of capability and avoid exorbitant costs.