Australia will buy between three and five nuclear submarines of the US Virginia class then switch to a new British design, according to a bunch of news reports that came from Washington and London this week.

Why not just Virginias? Why not just a single stream of identical submarines that would all share just one support establishment for maintenance, updates and training?

The Royal Australian Navy has already rejected the idea of operating a second diesel-electric submarine class as too complicated, so why do we want what looks like the migraine headache of running two nuclear classes?

More details will surely appear on Monday when Anthony Albanese stands with President Joe Biden and Prime Minister Rishi Sunak to formally announce the program.

In the meantime, here's a stab at working out what's going on.

First, the Virginias in the deal are obviously stop-gaps; our ultimate target is an Anglo-Australian building program that will briskly produce subs for both the Royal Navy and Royal Australian Navy.

The next Royal Navy design, Submersible Ship Nuclear (Replacement), is not ready, nor is the British submarine industry, which has its hands full with the national priority of building four giant ballistic-missile subs - doomsday weapons for nuclear war. And Australia doesn't have a nuclear submarine industry at all yet.

So the first SSNR probably can't be completed before 2037 even if an infusion of Australian money pays for acceleration of development and construction.

We can't wait that long. While the aggression of Xi Jinping's China is increasing the risk of war, our six diesel-electric subs of the Collins class are too few and too incapable.

So the Virginias that reports say we will "buy" in the early 2030s will be an urgent first step in building up the navy. If "buy" means taking delivery early next decade, then Defence Minister Richard Marles has done a fine job.

The overwhelming issue in Australia's pursuit of nuclear submarines is to get steel in the water as soon as we can. The fastest way is to pay to expand the US industry so it can build extra subs directly for us or as replacements for second-hand boats that we buy from the US Navy's fleet.

The number of Virginias for Australia, "up to five", is being kept flexible probably because our strategic circumstances could worsen even more by the 2030s and because SSNR could suffer delays.

Why, then, not just stick with buying made-in-USA Virginias until we have at least eight of them?

I hope that feeding Adelaide with jobs is not part of the answer to that question, but I fear that it is. Quite possibly, the US was not willing to share Virginia production technology with us, or maybe it would not take the risk of turning the Virginia program into a joint effort with a bunch of newbies.

If so, we may imagine our government being unable to stomach the idea of facing down South Australian voters, to whom it has promised a nuclear submarine industry, leading it to look for a long-term solution from the more cooperative British instead.

To justify local construction, the government will push the invalid idea that Australia needs to build its submarines if it is to properly operate them - even though we've had great satisfaction with fully imported warships before and even though almost all our military aircraft were built overseas. Just read the manual.

Nor is it true that Adelaide must be mobilised to add capacity for the three-nation alliance. Just think: if the US Navy needed to build faster for itself, would it come to Australia begging for Adelaide's help?

MORE AGE OF THE DRAGON:

Still, shifting to SSNR, wherever it is built, will offer advantages, even if they don't look big enough for the complexity that will come with running it in addition to the Virginia class.

SSNR will be a more powerful weapon than the Virginia class, which has been under construction in successively updated batches since 1998. SSNR is likely to be harder to detect, better at detecting, and designed to work with uncrewed subs that move ahead of it into danger zones.

Everyone who talks about nuclear submarines points out that British boats need fewer crew members than US ones, a supposedly great advantage for Australia, which struggles to get enough people even for the Collins class. But for the whole fleet we'd be talking about a difference of only about 1000 people - in a population by then exceeding 30 million.

It's high time we lost patience with the managerial failure of our armed forces in recruitment.

Expect a few measures to take some of the pain out of the two-class migraine. One will be that our Virginias will at first have plenty of American sailors aboard.

Also, you can bet that as little of the support establishment as possible will be in Australia. We'll have to be able to do at least minor maintenance and some of the training, but our submarines and submariners will be spending plenty of time in the US.

Our navy has US-built MH-60R helicopters that, for support purposes, are basically treated as part of the US Navy's fleet. We'll probably take the same approach with Virginias.

If our Virginias are new or almost new, there's even a possibility that we would sell them back to the US in the 2050s as our SSNR fleet reached targeted numbers.

But how many submarines we will need in 30 years is hardly worth speculating about. At this stage, we just need steel in the water.