More like this




'On demand' fibre: Could it happen in Oz?

Trying to find the subsidy

A question posed by El Reg to Australian opposition communications spokesperson Malcolm Turnbull has opened up something of a can of worms.

It’s probably unfair to say that Turnbull “floated” or “supports” the idea of Australia imitating the Openreach “Fibre on Demand” model. He merely didn’t dismiss the notion out-of-hand.

But with the idea in the debate, what about the numbers?

The first thing to understand about the Openreach model is that in commercial terms, it’s brand-new. BT Openreach doesn’t actually have a finalised price sheet for the installation yet. Estimates range from £1,000 to £1,500 – up to $AUD2,250 – for a “typical” connection, plus a fixed £500 ($750) installation fee.

The big “if” is “if you live within 500 metres of the fibre” – that is, the cabinet at which your premises fibre will terminate.

The second big “if” is the availability of duct space. That is: there has to be room in the local ducts (now carrying copper) for BT to install the fibre. If there isn’t room for the ducts, then you’d better have deep pockets, because new ducts cost at least £46 per metre: say around $70,000 per kilometre.

If the BT Openreach pricing – sans duct costs – is accurate, then it seems to Vulture South that a level of commercial subsidy of some kind is involved.

The Register has sighted papers that put the cost of renting duct access from Telstra at $6.95 per metre (per year). Obviously, that price builds in a profit margin of some kind, but how much?

At £1,000 per 500 metres, BT is charging $3 per metre for a combination of duct access, fibre purchase (fibre is cheap, but it’s not free) – and the cost of actually pulling the fibre, which is also not free.

Hence the suspicion that BT is discounting the service, for commercial reasons. A couple of possibilities exist:

1. Long-term recovery from subscription fees – since the fibre is there forever, and presumably will have subscribers forever.

2. Cost equalisation – the first home in an area is heavily discounted, on the basis that some of the work becomes less expensive for later installations. For example, instead of pulling a single fibre strand to a premise, BT Openreach would (if it’s sensible) install multiple fibres. Later connections will cost BT Openreach less (the fibre’s closer to the home) but the subscriber will pay the same. Implications for Australia are difficult to draw without a full economic model to work on.

However, the model would depend on three things at least:

1. Sufficient users within a reasonable distance of their node

At (say) $1,500 to $2,000 for a 500-metre cable pull, there may well be households and businesses willing to pay to get access to a BYO fibre service. After all, NBN Co’s Jim Hassell told Senate Estimates in October 2012 that 44 percent of users connecting to the NBN had adopted the 100 Mbps service (well ahead of NBN Co’s own business plan).

2. Sufficient available, usable duct space

This is an unknown, since only Telstra – and NBN Co to a certain extent – can discuss which ducts could accommodate fibre without new duct work (which costs plenty, as BT Openreach’s pricing shows).

3. Telstra

The big if. Under the Coalition plan, the Telstra / NBN Co multi-billion-dollar duct access deal would have to be unwound – and then Telstra would have to be convinced to offer affordable fibre cable-pulling to end users.

Economic Value of Duct

For an Australian estimate, let’s ignore the new duct problem and assume, instead, that the real economic value of a duct as a sunk-cost is in the order of 20 cents per metre per year. NBN Co, for example, is paying $AU5 billion over 30 years ($167 million per year) for access to millions of kilometres of Telstra duct.

Since the owner of the duct is the same company as owns the fibre, it doesn’t matter much if its $150 notional annual value is ignored in the charge – because the expensive part of the installation is getting the fibre through the duct.

If you’re so efficient you can pull the fibre through at $5 per metre, if the fibre itself ships at $1, and if the average node-home distance is 750 metres (we’re less densely packed than in the UK), then the real cost of installation is $4,500.

The wholesale provider, or the retailer in the hope of recovery, or the government – someone would be footing the other $3,000.

Even if only five percent of households bought into the “DIY fibre” vision, the construction subsidy would amount to nearly a billion dollars.

Of course, customer subsidies are common in this industry – if you didn’t buy your smartphone off-the-shelf, you’re getting one now. But it’s a good thing to know they exist. It's also a good thing to realise that this story is based on some slightly-better-than-back-of-the-envelope calculations, and is in no way a critique of anything anyone has said about the NBN. ®

Sponsored: Ensuring business success by selecting the right application development platform