NBN FTTN's power bill: $89m a year?

Back-of-the-envelope calculations on how much will it cost to power the National Broadband Network's nodes

[[xref:https://www.youtube.com/watch?v=QTG4TBxVOZ4&sns=tw|Malcolm Turnbull last month announced an expanded FTTN trial]].

Malcolm Turnbull last month announced an expanded FTTN trial.

The Coalition government's decision to switch the fixed line element of the National Broadband Network from being primarily fibre-to-the-premises (FTTP) to a mix of technologies, notably fibre-to-the-node for brownfields, has not been without its critics.

At least some of that criticism has centred on the street-side cabinets that will be needed under a FTTN deployment. They will be condemned as "hideous" and vulnerable to vandalism, and critics have alleged that powering them will rack up a substantial electricity bill.

The aesthetics of the cabinets may be subjective and whether they will suffer at the hands of vandals remains to be seen. However, it is possible to do some rough calculations on the cost of electricity thanks to information provided by NBN Co in response to questioning by former communications minister Stephen Conroy at a May hearing of the Senate's NBN committee.

NBN Co COO Greg Adcock was unable to provide Conroy with a detailed answer on the electricity costs of hooking up a test node in Umina to the electricity grid, but NBN Co later provided the committee with a detailed breakdown.

The nodes are powered by a device mounted on the closest power pole, with an underground cable is run from that device to the node. Hooking up a node to the electricity grid costs $6250. According to NBN Co, annual electricity costs run at around $1404 plus an $85 service charge, for a total of $1489 per cabinet per year.

An FAQ on the Coalition's broadband policy posted on communications minister Malcolm Turnbull's website states that "50,000 nodes would be needed (many of them in the basement of apartment blocks)", though the Coalition has also thrown round the figure of 60,000 nodes.

"The exact number will be dependent on detailed planning and designs which have not yet been finalised," an NBN Co spokesperson said.

That would put the annual power bill for the nodes at between $74.5 million and $89.3 million (and electrical installation costs at between $312.5 million and $375 million). Of course, in the real world, things will be substantially more complex than these back-of-the-envelope calculations.

NBN Co is currently in the middles of conducting several FTTN trials.

Late last month NBN Co revealed plans for a 1000-node test FTTN deployment in conjunction with Telstra. Under the agreement, which is reportedly worth $150 million for Telstra, FTTN will be rolled out in Belmont, Boolaroo, Gorokan, Hamilton and Morisset in NSW; and in Bribie Island, Bundaberg, Caboolture, Gympie and Warner in Queensland.

The trial will "deliver valuable insights into how to build a sustainable and consistent program of work that allows the industry to ramp up and deploy the FTTN element of the NBN at scale," according to NBN Co CEO Bill Morrow.

The FTTN contract with Telstra "will ensure initial rollout of FTTN focuses on areas categorised as ‘underserved’ in the Government’s MyBroadband broadband quality study," a statement issued by Turnbull's office said at the time of the announcement.

The large scale test builds on the smaller NBN Co-led FTTN test deployments in Umina, NSW, and in the Melbourne suburb of Epping.

In Senate Estimates hearings in May, NBN Co COO Greg Adcock revealed that although the Umina trial was "progressing well", the Epping trial "has slowed down a bit while we work with the utility there to find a power solution, and we are working through that".

In a limited FTTN test in Umina conducted earlier this year, NBN Co achieved download speeds of 105 megabits per second and upload speeds of 45Mbps.

The FTTN trials follow the revised statement of expectations for NBN Co that was sent to the government-owned company in April by Turnbull and finance minister Mathias Cormann.

That SOE backed 'Scenario 6' outlined in the strategic review of the NBN rollout prepared by NBN Co shortly after the Coalition's electoral victory last year. 'Scenario 6' is an 'optimised multi-technology mix'.

Instead of Labor's NBN blueprint, which involved rolling fibre to the doorstep of most homes and businesses in Australia, the NBN will instead employ FTTN for some 32 per cent of premises scheduled for fixed line NBN connections, fibre-to-the-basement for multi-dwelling-units (such as apartment blocks), hybrid fibre-coaxial in areas covered by existing HFC network, and fibre-to-the-premises for greenfields developments.

Read More:

Join the Computerworld newsletter!

Error: Please check your email address.

Tags Networkingfibre to the nodenational broadband networkNational Broadband Network (NBN)nbn cofibre to the node (FTTN)broadband

More about BillIslandNBN CoTelstra CorporationTelstra Corporation


Abel Adamski


With those in power, Australia has a limited future anyhow, so what does it matter anymore.
The damage that will be done will be close to practically impossible to undo.

Emigration increasingly appears a better option for those with Assets or skills well rewarded elsewhere, but with no or a limited future here



Even 60,000 nodes seems wildly optimistic given that Telstra (Greg Winn Sydney media briefing 05/12/2008) originally estimated 50,000 nodes for a 90% solution (customers in capital cities to be no further than 800m from a node, while customers outside of capital cities to be within 1500m) to deliver only 12Mbps while Turnbull guaranteed 25Mbps (I think he mentioned a maximum 500m distance at one point).
He also said they had 900 engineers working for three years to put together the design and it would take five years, working 24/7 to build it.



This is the fully costed and ready to go NBN that the coalition promised before the election, yet "the Coalition has also thrown round the figure of 60,000 nodes." so much for fully costed and ready to go. Also no mention of what they will have to pay telstra to purchase the rotting copper network NOR an annual cost to remediate this copper. FTTN is a joke and a huge backward step for Australias telecommunications. Mr Turnbull should note that NZ dropped their FTTN roll out half way through and moved to the much superior FTTP solution. The BT FTTN that he modelling our FTTN on has been a disgraceful failure and will only be worse in a country whose density of population is far less than that of Briton. But wait no mention of the other test site in Epping Vic, which hasnt even been able to proceed because they couldnt power the node? what a FTTNNNN disgrace



I don't believe the cost of powering the node to be "total of $1489 per cabinet per year" that would mean it only consumes about 700 watts, FttN nodes typically draw much more especially as to get any real reliability they will have active cooling for the Australian climate. This will put them in the 2,500-5,000 watts and even more to power 200-400 ports, the consumption of enterprise grade VDSL2+ cards use much more than 2 watts a port. The early vdsl2 chipsets had GPU graphics level cpu's to do the complex maths and these used over 30 watts a port. Cooling would be 1 to 2 kw/h as the chips would not like operating at over 55 celsius and the cabinets would be a hot box easily going over 70 internally in the direct sun light. This would put the anual bill closer to $5, 000 and the co2 figures for 30k to 60k cabinets, huge. Not cooling them would result in much higher failure rates, and possibly reduced line speeds and greater errors, bringing on TCP's slow start that will kill performance further. The monitoring systems in each cabinet will use additional power and they may need to implement some redundancy allowing for failed ports to be automatically switched out and spares switched in. For this level of scale redundancy would be the cheaper way to go, or else you will be servicing these nodes way to often.

Comments are now closed

Cybercrime reporting site launches in Australia