The costs of cutting steel
5 Nov 2014|
A young woman arc welding part of an anti-tank gun in a munitions factory in South Australia in 1943.

In 2013, the early replacement of the Anzac frigates was proposed as a way to bridge the shipbuilding ‘valley of death’. The idea was to continue building AWD hulls and equip them with a combat system based on the Anzac upgrade now underway.

Even starting tomorrow, it’s doubtful that such a scheme would preserve more than a subset of white-collar jobs at the shipyards. Yet the idea survives, and has grown into an ambition for a continuous-build program involving as many as twenty vessels. The number of hulls has grown because a continuous-build program has to be fed by the taxpayer, either by routinely retiring vessels before their economic life or by building a larger fleet than the RAN needs.

But the imperative for domestic shipbuilding is weak and the costs would be high, so why does the idea persist? Two factors are at play: lobbying by incumbent firms and the politics of South Australian jobs. The latter is especially potent given the likely foreign build of the Collins replacement. As for the expense of jettisoning recently-upgraded Anzacs, the consoling thought is that we could bolster regional maritime capabilities by gifting vessels to Indonesia and potentially selling others to New Zealand.

What would the proposal cost? That depends on many factors, including the reduced life and/or increased number of the surface combatants. Then there’s the competiveness of local shipyards.

The head of DMO recently testified to a Senate inquiry that ‘a productive yard, building continuously’ could have built the first AWD in 3 million work-hours, and that 4.7 million had been allowed for the local build’s cold start, yet the current estimate for the first vessel stood at 9.3 million. Using the per-capita annual cost of labour at ASC ($132k) and the estimate of 2,000 annual hours per worker from 2013 Submarine Skill Plan, the additional cost of the first vessel can be estimated (excluding shipyard overheads and material costs for rework):

Work-hours Labour cost
Continuous production 3.0 million $198 million
Originally planned 4.7 million $310 million
Now expected 9.3 million $615 million

Thus, on the basis of labour costs alone, we’ve spent at least $305 million more than planned, and $417 million more than best-practice for the first vessel. I say ‘at least’ because Australia has high wages. The US Bureau of Labor Statistics ranks Australia 6th out of 34 OECD countries in hourly compensation for manufacturing, and 7th out of 16 for the sub-category including shipbuilding.

It’s hoped that a continuous-build program would lift Australia’s shipyard productivity. Although higher productivity won’t reduce wage rates (and will do the opposite if wages are linked to productivity), it’s still worth asking how much improvement is realistically possible.

The Anzac program is often held up as a paragon of efficiency, with ten vessels delivered for $9.3 billion in 2012–13 dollars. Lacking better data, labour use can be estimated by multiplying its 125-month duration by the reported 1,223 prime contractor and 1,337 subcontractor personnel in 1994–95. Using the former as a minimum, this implies between 2.5 and 5.3 million hours per vessel. Noting that an Anzac is only 60% the displacement of an AWD, this is hardly impressive if a ‘productive yard’ can build an AWD in 3 million hours.

In comparison, Denmark has recently built three frigates at a cost of US$325 million per vessel (PDF) or around US$383 million including the value of reused equipment from retired vessels. Labour use was 700,000 hours per hull. As the table below shows, the contrast with the cost of our AWD and Anzac programs is startling.

And the Danes didn’t cut corners; their vessels are larger and better equipped than our Anzacs—including with SM-2 missiles.

  Unit cost ($A million) Displacement
(tonnes)
Cost per kg ($A/kg) Labour (hours) Crew
Danish Frigate 435 5,452 80 700,000 101
Anzac Frigate 932 3,600 259 >2,500,000 163
AWD 2,700 6,250 432 *6,200,000 202

*Average figure assuming that the second two vessels are each built using half the labour of the first.

How did they do it? To start with, the vessels were built in an efficient civil yard by an established workforce. In addition, lessons from civil design were adopted to make the vessels easy to build and cost-effective to maintain—hence the relatively small crew. And Denmark has access to low-wage module construction on the other side of the Baltic.

Australia’s naval shipyards would undoubtedly benefit from more efficient designs and construction techniques—and the Danes are already working with Canadian shipyards to help them to this end. Nonetheless, we’ll never have a commercial steel shipbuilding sector to leverage. And we’re poorly placed to access low-wage yards for module construction. If we build ships in Australia, we’ll pay a premium.

A practical alternative would be to build frigate hulls offshore and fit them out locally (as with the Canberra-class LHD). Doing so would allow us to focus on strategically relevant, high value-add areas such as systems integration rather than cutting steel.

Mark Thomson is senior analyst for defence economics at ASPI. Image courtesy of Flickr user State Library of South Australia.