full opinion/analysis by Renai LeMay
10 July 2013
Image: Office of Malcolm Turnbull
opinion/analysis Those broadband speed freaks holding out hope that the Coalition’s pledge to provide ‘fibre on demand’ services will save them from life in the slow lane in a fibre to the node future need to take a cold shower and wake up to reality. ‘Fibre on demand’ is nothing but an fluffy ephemeral dream which has no chance of becoming reality in the short- to medium-term under the Coalition’s National Broadband Network vision.
In its rival NBN policy document published in mid-April (PDF), the Coalition included several innocuous-looking paragraphs about a relatively new concept in terms of Australia’s National Broadband Network project. Termed ‘fibre on demand’, as it was outlined in the document, the idea sounded entirely rational and logical at the time.
“NBN Co will,” the document promised, “provide for fibre on demand at individual premises as soon as possible where fibre does not extend to the premise and this is technically feasible and commercially viable.” It further sets out some basic terms about the relationship between NBN Co, retail ISPs and end users with respect to fibre on demand, noting that NBN Co will retain ultimate ownership and responsibility for the infrastructure.
Although it was largely new to the Australian NBN debate at the time, the concept of fibre on demand is not unique globally. In fact, like so much else in the Coalition’s NBN policy, the fibre on demand aspect of it appears to have been pretty much entirely ported directly from the UK, where OpenReach, the infrastructure arm of incumbent telco BT, has recently started offering this type of service to residents and businesses, as an offshoot of its successful nationwide fibre to the node (called ‘fibre to the cabinet’ in the UK) rollout.
The concept of ‘fibre on demand’ is a simple one. BT in the UK, and Shadow Communications Minister Malcolm Turnbull in Australia, believe that fibre to the node-style rollouts, which see fibre deployed from telephone exchanges to neighbourhood street cabinets (with the last few hundred metres connected by the existing copper cable) are capable of providing fast enough speeds to satisfy the short to medium-term broadband needs of almost every resident and business in each country.
However, both are also cognizant that there are some specific types of businesses, non-profit organisations and even individuals who greatly desire the exponentially greater broadband capacity provided by fibre rollouts all the way to their premises. The theory behind ‘fibre on demand’ is that those who desire fibre all the way to their premise should be able to pay an extra fee to have it connected.
Taken on it’s own, this concept is not a controversial one. After all, it largely mirrors the existing broadband infrastructure situation in most countries, where copper, hybrid fibre networks (cable) and even mobile broadband solutions provide for most broadband needs, but large businesses and non-profit organisations such as hospitals and schools are able to pay extra to have faster connections extended to their premises. It’s a pretty fundamental concept for capitalist social democracies: Everyone gets the basics, but those who need or want more can get it — for a price.
However, in Australia the issue has recently become one of the most contentious political footballs in the already poisonous game of claim and counter-claim which represents our National Broadband Network debate.
In what appears to have been either the result of a centrally issued memo by the office of then-Communications Minister Stephen Conroy on the issue or a centrally generated advertising campaign, it has now become common for Labor politicians right around Australia to take the high end of BT’s fibre on demand costs in the UK and extrapolate them into a compulsory $5,000 figure for fibre on demand services on the NBN in Australia. We’re seeing outlandish advertising campaigns by Labor and very public speeches pushing the idea that if Australians don’t pay a $5,000 fee, under the Coalition’s NBN they will be “left on the old, slow copper network”, coupled with ‘worst of the worst’-style photos of Telstra’s copper network in disrepair. The Coalition has been rightfully outraged by this claim, with Turnbull labelling it as “classic spin” and putting the average cost much lower, at just several thousand dollars if those connecting to the NBN are within a reasonable distance from their local neighbourhood node.
Now, I don’t want to get into another debate about the costings involved in fibre on demand. We’ve already had that debate several times on Delimiter, and though it was a highly entertaining debate, to explore that issue again would be to overlook the bigger picture when it comes to fibre on demand.
What I want to do in this article is argue that the Coalition is wrong to even lay out fibre to the demand as a possibility to the Australian public at all. There is no doubt that the concept is a tantalising one; after all, who wouldn’t consider paying a little extra for fibre on demand, given the massive service delivery benefit it will bring? But in reality, it simply will not be possible for a Coalition Government to offer fibre on demand options to Australians any time within the short to medium-term; consequently the Coalition should simply delete this option from its rival NBN policy entirely.
To make this argument, what we need to do is compare the closest example we have of a combination FTTN/fibre on demand rollout — BT’s deployment in the UK — with what we know of the current NBN rollout in Australia, and its likely future. Then we need to look at other associated factors such as NBN’s current ability to do so-called user pays ‘fibre extension’ services in Australia, and what they might tell us about its ability to deliver fibre on demand.
From the information I’ve been able to glean, it appears as though BT first started deploying fibre to the node throughout Britain in January 2009, with a number of trials being conducted around the country that year, and commercial services launching 12 months later in January 2010. At the time, the platform was dubbed ‘BT Infinity’. The deployment of this kind of service can broadly be considered analogous to the 2005 plan outlined by then-Telstra chief executive Sol Trujillo to upgrade Telstra’s copper network to FTTN, in that the rollout is predominantly being conducted by an incumbent telco which already owns its own copper network and all associated infrastructure, and which already has tens of thousands of engineers in the field to help deploy new infrastructure.
That same year, 2010, BT’s infrastructure arm Openreach, announced it would deploy FTTN to some 19 million households across the UK, and in October 2010, UK regulator Ofcom announced that BT would be required to provide open access to its fibre infrastructure, in the same way that telcos such as Optus, iiNet and TPG access Telstra’s network in Australia.
In May 2012, OpenReach announced it had passed the 10 million premises mark, and the end of June 2014, Openreach expects to have completed its rollout, although it has also already announced that it will extend the rollout to new areas, beyond the two thirds of the UK that it had initially planned; and it seems easy to predict that some rollout work will progress indefinitely.
If you take this information in aggregate, what we see is that BT will have taken between four and five and a half years, in terms of its real rollout, to deploy fibre to the node to some 19 million premises around the UK — or about 3.4 million premises every year, perhaps up to 4.2 million, depending on when its rollout really started. And it has only been in the latter stages of its rollout — from May this year — that the telco was able to offer fibre on demand services to end user customers, and then only in some limited areas, ahead of a phased rollout across the UK. Even right now, if you want fibre on demand through BT in the UK, it’s not always guaranteed (far from it!) that you’re going to be able to get it.
In Australia, it is obvious that our own FTTP-based National Broadband Network rollout is proceeding astronomically slower than BT’s FTTN rollout in the UK. NBN Co was initially set up in April 2009. Since that time, the company announced last week (PDF), it has connected just 207,500 fibre premises, of which only 163,500 were in ‘brownfield’ areas — areas where Telstra already had copper cable.
The difference between the two rollouts could not be more stark. In the UK, BT has been able to deploy FTTN broadband services — which offer download speeds of up to 76Mbps and upload speeds of up to 20Mbps — to some 10 million premises. In Australia, because the Federal Government pursued a radically different policy, NBN Co has only been able to connect several hundred thousand premises to broadband.
The key difference here between the two rollouts is pretty much directly attributable to the fact that NBN Co started its rollout from a standing start, while BT was not. Openreach’s website states that it has access to some 25,000 engineers to conduct its rollout, backed by some 650 planning staff and existing access to BT’s network infrastructure — 200,000 manholes, more than 4.2 million telecommunications poles and so on. NBN Co has, over the past several years, accumulated some of these same advantages — such as access to Telstra’s existing infrastructure, tens of thousands of construction staff through its contractors and an administration/planning in-house staff of several thousand. But when it comes down to it, NBN Co just isn’t anywhere near the integrated organisation that an incumbent telco like BT (or, for that matter, Telstra) is.
Let’s assume that the Coalition wins the upcoming Federal Election and mandates that NBN Co switch its rollout style to a fibre to the node deployment. And let’s assume the absolute dream case scenario for that rollout — say, where NBN Co, with most of its setup under its belt and much of its workforce now across the complexities of fibre, is able to match BT’s speed. Remember, BT was able to stay on track to deploy something like three to four million FTTN premises each year.
Judging by the BT rollout, even if you assume the best case scenario, it’s going to take NBN Co around three to four years to get around to most of Australia’s around 12 to 13 million premises with FTTN.
And right now, NBN Co is not deploying fibre infrastructure at anywhere near those speeds. True, it is a fact that FTTP is significantly harder, and takes a lot more time to roll out than FTTN. However, remember, since April 2009, NBN Co has only finished deployed FTTP to around 207,500 premises, and that it’s currently projecting that its network will have hit some 3.8 million premises by mid-2016.
In fact, NBN Co is actually even having trouble meeting its existing goals. The company downgraded its mid-yearly construction targets in March, and even the veracity of the results issued several weeks ago have been publicly questioned, with many of those premises NBN Co has listed as completed not actually being able yet to connect to the network. The causes of the delays are very widely acknowledged; a lack of availability of the right skillsets in, and poor management of, NBN Co’s contracted construction workforce. Issues which BT, with its in-house, technically trained team, largely doesn’t face.
So what does this mean for fibre on demand?
If we go back to the Coalition’s policy document, the core promise is that every Australian should have access to broadband at speeds of at least 25Mbps by 2016, and speeds of at least 50Mbps by 2019 to 90 percent of the fixed-line broadband footprint. What this very likely means in practice is that the Coalition plans to have its fibre to the node network finished by the conclusion of its second term in Government — by the end of 2019. And in fact it’s likely that the Coalition’s FTTN rollout won’t even really kick off until late 2014. “Rollout of fibre to the node does not begin at scale until late 2014,” the Coalition’s NBN background briefing document (PDF) states.
With NBN Co’s delays in mind, and considering the timeframe on which BT started to offer fibre on demand services — that is, not until its own FTTN rollout was more than halfway complete — it starts to seem very unlikely that NBN Co under a Coalition Government could realistically begin to offer fibre on demand services during its first three years in power, and even beginning to offer such services in the three years between 2017-19 could be quite a stretch. Put simply; it remains unlikely that NBN Co will have the resources to offer this kind of service until a large tranche, perhaps the the majority, of its network build is complete. New FTTN services must, of necessity, take priority over fibre on demand.
This conclusion is also bourne out further by the demonstrable experience which NBN Co has had in offering fibre on demand services in Australia already.
NBN Co first began deploying FTTP infrastructure in Tasmania in July 2009 in a number of small towns in the state. The first customers were connected a year later. However, although there was a degree of interest even that far back from residents in areas just outside the NBN’s footprint for the company’s so-called user-pays ‘Network Extensions’ approach, it has taken a great deal of time for NBN Co to agree to extend fibre upon resident demand. ZDNet reported in May this year that NBN Co had confirmed at that stage that it had completed a fibre extension in one case, and a further eight were under “commercial discussions”.
The difficulty to NBN Co of even organising this small number of fibre extension cases is reflective of the fact that organising a fibre extension on demand takes the company significantly outside its existing planning processes with respect to its network infrastructure. It must divert construction, planning and administration resources away from other areas of its operation towards an area which may offer it little to no return in investment in terms of its overall goal — getting as many Australians connected to fast broadband as soon as possible.
If you consider the case of fibre on demand compared to NBN Co’s existing ‘fibre extension’ process, the situation is likely to become even more complex. In the UK, BT encourages customers who are thinking of paying to upgrade to fibre on demand to band together in neighbourhood groups, as a group rollout will reduce the cost to individual premises. If one premise goes first, it is likely that it will absorb a large slice of the cost of extending fibre to that premises’ street — meaning it will be cheaper for those premises which come individually afterwards, and making the situation a little unequitable for those who go first.
It’s precisely these kinds of issues which has led BT to trial fibre on demand in a small number of exchange areas before launching the service more widely. In addition, in general, the disconnect between Openreach, retail ISPs and customers in the UK is “infamous” — and there have already been indications of similar issues in Australia.
If you extrapolate these dynamics a little to many thousands or perhaps tens of thousands (eventually, hundreds of thousands?) of cases of Australians wanting to sign up to a fibre on demand program through NBN Co, it’s not hard at all to see that such a program as a whole could become a collossal drain on resources, as the administration of the infrastructure upgrade process bounces back and forward between groups of residents and business owners, NBN Co, and the retail ISPs which will actually administer customers’ broadband plans. This burden must necessarily take second place to getting as much of Australia connected to FTTN in the first instance.
It’s also necessary to consider that the last leg of fibre rollouts — from nodes to premises — is the hardest.
Turnbull himself has acknowledged this issue publicly and repeatedly, as justification for the Coalition choosing a FTTN-based rollout style instead of FTTP as Labor has done. And internationally the same trend is witnessed. Opeanreach’s director of network investment, Mike Galvin told the Communications Day Crosstalk podcast recently:
“… it’s very expensive. And not to mention disruptive to the customers or to the businesses, to deploy fibre over that last drop. We found it was significantly more expensive and also took a lot longer – it had a much more complex provision process.”
Taken together, the evidence is overwhelming. It’s very nice in theory to think that the Coalition will be able to offer fibre on demand services under its FTTN NBN plan; the whole concept is a great sap to those Australians (and there are many of us) frustrated that the Coalition won’t support Labor’s more visionary all-fibre NBN policy.
However, the reality is that the Coalition and NBN Co itself will have its hands very much full over the next six years delivering the Coalition’s fibre to the node vision; and experience from the UK shows that fibre on demand options are probably not going to be made available for a very long time indeed, due to their complexity and the need to focus on the mainstream FTTN rollout first, with the limited resources which NBN Co has.
While it has its detractors, fibre to the node is a very solid technology which will deliver fundamental service improvements to many Australians. Fibre on demand may be a very nice icing on the cake; but it’ll have to wait while the main course is served. Don’t hold your breath to blow out the candles; they haven’t even yet been lit.