'

Turnbull fooling nobody but himself on NBN Co FttP savings

It wasn't too long ago that the Coalition was claiming that it would cost AU$3,600 per premises to build an FttP NBN. With NBN Co savings dropping per-premises costs towards AU$1,000, there's a valid argument that the Strategic Review and cost-benefit analysis no longer represent FttP's costs correctly.

After a fortnight in which around 120 pages of evidence has emerged suggesting that NBN Co has succeeded in reducing the cost of fibre-to-the-premises (FttP) installs, it defies belief that Australian Communications Minister Malcolm Turnbull would nonetheless stand in front of the Corporate Turnaround Conference and claim that no such savings has been achieved.

"There is no evidence FttP or fixed-wireless costs per premise are declining," Turnbull said in a speech dated September 18, apparently choosing to ignore the revelation that implementation of the Labor-era "Project Fox" — a process-improvement program stewarded by former CEO Mike Quigley and enthusiastically embraced by current CEO Bill Morrow — had been reported to have reduced the average cost of a fibre install by around AU$315.

Those figures backed up earlier reports that a trial of new technologies, construction, and project-management methods using the NBN Co-derived Render network rollout methodology had reduced the cost of many FttP rollouts in Melton, Victoria, by up to 50 percent, and shortened installation times by up to 61 percent.

NBN Co challenged the report, with its head of corporate affairs tweeting that there had been "no such trial" and issuing an entire press release claiming that the Melton results had delivered "no such conclusions". However, additional figures subsequently obtained by ZDNet confirm not only that Melton was not an anomaly, but also that Turnbull clearly remains misinformed about the work being done within NBN Co.

NBN Co is proving itself to be an effective learning organisation that is, as you would expect, actively running trials of new technologies and techniques for improving the cost effectiveness of its rollout. The seeds of the change were planted well before the last election, and built into a revised NBN Co Corporate Plan v13 that was presented to the former NBN Co board, but buried after Turnbull discredited the claims of FttP savings and pushed out the entire board.

Those well-meaning employees must get some pleasure in noting that NBN Co's work in Melton was able to boost rollout speeds by giving contractors a more continuous pipeline of work, and using new techniques to deliver significantly faster civil works. Boring, for example, was completed 154 percent faster than the Victorian average, while trenching was sped up by 250 percent, and cable splicing was completed 160 percent faster.

Workers were able to install two pits per day in Melton rather than 0.6 pits across Victoria, and proved able to install 12 of the smaller, slimmer multiports per day compared to 4.1 multiports across the state. Credit this to the five-minute multiport installs in Melton, compared to 15 minutes for a traditional multiport.

Underground cable was being hauled in Melton at 1,195 metres per day compared with 692 metres per day on average across the state. Teams were able to install three Fibre Distribution Hub cabinets per day, compared with 0.8 cabinets per day previously. And the smaller-diameter cable, to which NBN Co has transitioned, is 10 to 20 percent faster to deploy because it is lighter, easier to carry, and can be installed by fewer people.

While NBN Co initially treated the Melton results as an anomaly, its further argument — that its crews had actually reduced the cost of FttP by even more in other sites — was corroborated by the follow-up figures, which spelled out the nature of the average AU$351 savings found in the numerous sites where NBN Co has trialled them.

Turnbull still says that FttP isn't getting cheaper — but what, then, are we to think when presented with evidence that simply redesigning the fibre loop reduced the average cost of connecting 3,657 premises in Victoria Park, WA, from AU$1,239 to AU$1,034? That the costs of connecting 4,900 premises in rural Maitland, NSW, had dropped from AU$1,844 to AU$1,287 per premises?

Would Turnbull have been so quick to discount the projected cost savings — which had been duly incorporated into an updated version 13 of the NBN Co Corporate Plan 2013-16 that he discredited, and has, unsurprisingly, not released — if there had been proof last September that the rollout to 3,625 mostly new homes in Toowoomba, Queensland, could be slimmed down from AU$1,776 per premises to AU$1,363?

Or that 4,689 mostly new homes in Penrith, NSW, could be connected for AU$1,189 per premises using optimised placement of Fibre Distribution Hubs, compared with AU$1,480 before?

Remember that the government recently committed AU$150 million to get Telstra to connect 200,000 fibre-to-the-node (FttN) homes — an average of AU$750 per premises. That's a big saving on AU$1,844 per premises, but not such a big difference, compared with AU$1,287.

Objective analysis must consider how these sorts of savings would have significantly changed the conclusions of the recent cost-benefit analysis (CBA) and the previous NBN Strategic Review, which highballed the cost of FttP and lowballed the cost of FttN in order to produce a preordained outcome that surprised nobody. Vertigan's later cost-benefit analysis built on those figures, and in so doing became the proverbial fruit of the poisoned tree.

Despite repeatedly conceding that an all-fibre outcome would be the ideal NBN scenario, Turnbull has repeatedly cited the higher cost of FttP as justification for shifting strategies. Even in the face of evidence suggesting that FttP can be rolled out more economically, however, he is clearly holding firm to his views — because to do otherwise would be to force compromise on his multi-technology mix (MTM).

Given that Turnbull has committed AU$150 million to a 206,000-premises FttN pilot — which works out as nothing more than an approximate AU$750 per premises subsidy to Telstra to deliver services on a network that the government still doesn't own — common sense would suggest that cheaper FttP would make the relative financial case far more appealing.

Morrow may want to organise a briefing with Turnbull rather urgently; having confirmed both that FttP cost savings are possible and that the eventual mix of technologies is still open to change , he may soon end up in the uncomfortable position of contradicting his main stakeholder minister and pushing to maximise the use of FttP within the AU$29.5 billion allocated by the government.

However it pans out now, the revelation of these rollout improvements also raises questions about the chain of events just after the 2013 election, when the NBN Co board was presented with a plan to cut what was believed to be AU$4.5 billion worth of costs from the fibre rollout, in large part due to the savings that have subsequently been validated.

Demonstrated savings of AU$315 per property, applied across a 11 million-premises FttP rollout, would deliver AU$3.465 billion in cost savings. Add in other process improvements in other parts of NBN Co, and you're not that far off the AU$4.5 billion savings proposal presented to the former board and incorporated into that revised Corporate Plan v13 — the findings and implications from which Turnbull continues to ignore to this day.

That was the kick-off to a series of policy decisions that have sent the entire telecoms industry topsy-turvy as Turnbull defies logic and common sense — not to mention his own previous pronouncements — by actively seeking to buy and maintain Telstra's copper network, cut off competitive threats from rival TPG at the knees, and disincentivise telcos from investing in their own networks, all to protect the most significantly anti-competitive telecoms policy Australia has ever seen.

Turnbull used to refer to NBN Co as a "repugnant monopoly", but his determination to stay on-message, and to deliver FttN at all costs — even in the face of evidence from his own policy implementation organisation — is taking the discussion to a whole new level. And, to top it all off, Turnbull still isn't guaranteeing that his version of the NBN will even work. Such claims may be his way of hedging his bets, but for someone who was elected on claims that he would fix the NBN and finish it his way, they do raise more questions than they answer.

What do you think? What can we take from the optimised FttP results that have come to light? How would they have affected the NBN Strategic Review and CBA? Will NBN Co ultimately use them to boost the penetration of FttP after all?