Australia's broadband debate is beginning to take on elements of farce.
For quite some time on Friday, April 19, Malcolm Turnbull, the opposition spokesman for communications, quizzed National Broadband Company (NBN Co) CEO Mike Quigley and CTO Greg McLaren about the feasibility of using VDSL for in-building distribution in blocks of flats, aka multi-dwelling-units (MDUs).
For someone who professes technical agnosticism, the exchange sounded very much like someone seeking a role for the technology he champions. Turnbull also sounded very keen to get NBN Co to devote resources to estimating the costs, and therefore savings, if VDSL in the basement were used in suitable apartments, instead of fibre to the premises (FTTP).
Ever-helpful, Vulture South has decided to try and shed some light on the question.
From the outset, it would appear that Turnbull is barking up a very spindly tree indeed. According to the Australian Bureau of Statistics, around four percent of Australia's housing stock comprises high-rise units. That gives an upper boundary of 300,000 dwellings which could theoretically be served by fibre-to-the-basement.
Except: most of those dwellings will be in buildings too small to suit the VDSL solution. As NBN Co's Ralph Steffens (COO) told the committee, the average “multi-dwelling unit” (MDU) in Australia is a mere nine dwellings. Most don't count as high-rise.
Even in the high-rise stock mentioned, the ABS definition starts at four stories. Such buildings can be seen in practically every suburb in Australia, and are probably – again – too small to have the two conveniences that fibre-to-the-basement requires: a basement, and centralised copper distribution via an MDF (main distribution frame).
Australia has been building these “low rise high rise” for decades – since at least the late Victorian era – but it's also true that there's been a boom in genuine high-rise. Meriton alone claims 50,000 dwellings, although the number of dwellings per development would be needed to tell us how many individual buildings are involved.
Let's assume for this discussion that one-third of the total “high rise”, as defined by the ABS, is suitable for a fibre-to-the-basement installation. That's 100,000 dwellings, more or less.
And let's be very harsh with our numbers: I'm going to assume that the cost of fibre-connecting an existing MDU is twice that of a house – $4,800. On the other hand, delivering fibre to the basement will cost around the same as a house – $2,400. I'll add $100 for the VDSL port (industry sources tell me this is a reasonable rate for small switches. They don't much like selling 24 or 48 port devices and don't try to price them down).
For simplicity, this model assumes an average 48 dwellings per building, as a handy match with the number of ports in a common low-density VDSL node. That assumes the rollout covers just over 2,000 tower blocks.
On those assumptions:
- Cost of FTTP to 100,000 high-rise dwellings: $AU480 million
- Cost of FTTN to 100,000 high-rise dwellings in 2,000 buildings: $AU30 million
- Saving: $AU450 million
Sure, $AU450 million is a decent wad of money, but in the context of the $AU37.4 billion that NBN Co says it still believes is a reasonable estimate for building the network, it's trivial – 1.2 percent.
Not only that, but this model ignores other matters – the basement kit will incur additional operational costs, for example, not only because it consumes electricity, but because it introduces a non-standard element into the network which will need its own management and maintenance procedures (the latter is much harder to attach a financial estimate to).
And, as was pointed out later in the same proceedings, the basement VDSL would prevent any of its users from signing onto a gigabit service in the future. There's an entirely unassessable opportunity cost in the potential lost sales.
NBN Co gave serious answers to Malcolm Turnbull's questions about FTTN in high-rise buildings. It would have been just as valid to say that they haven't conducted a rigorous analysis of the possible savings, because it would be a waste of time and money.
And yet, The Register is mystified by the entire line of questioning. Every poll suggests that by September, Turnbull will be communications minister, at which point he will be mandating FTTN wherever it's not already too late to halt the FTTP rollout.
Obsessively nit-picking at how best to serve a couple of percent of the total rollout seems oddly out of context. Almost as if Turnbull is starting to wonder whether FTTN is the red-hot great idea it's cracked up to be. ®