This article is by Michael Berry, an IT consultant currently working in the financial services industry. He’s also the director of the Progressive Democratic Party, which launched in January this year, with the principles of collaboration, honesty and duty. This article first appeared on the website of the PDP and is re-published here with permission.
opinion/analysis I’ve been an advocate of Labor’s plan for a National Broadband Network ever since it was announced in 2009. Four years later, some of the realities of creating a National Broadband Network have set in. There’s been a long negotiation with Telstra to get access to their infrastructure, setbacks with various network contractors and plenty of rework and re-planning as other issues occur along the way.
These delays have given the Shadow Communications Minister, Malcolm Turnbull, no end of material with which to claim that the NBN rollout is an expensive failure. You’d expect to hear plenty of anti-NBN bluster and rhetoric from the Opposition, but the Member for Wentworth has done far more than just loudly complain about the NBN Co’s failings. He’s spent the last few years researching in order to create the Coalition’s own alternative broadband policy.
At the Press Club in January Tony Abbott announced that the Coalition’s broadband plan would “… deliver superfast broadband for a fraction of the price and in a fraction of the time …”. This was a typical policy announcement, a few nice ideas, but no actual information on how this would be achieved. All we knew was the Coalition would be using fibre to the node to reduce the rollout costs.
When the Coalition launched its broadband policy in April I was pleasantly surprised to find that not only did they have an informative 18 page policy document but they also had prepared a further 36 pages of policy “background information”. After reading the full plan and accompanying information I can now say that the Coalition’s node-based solution is a well-researched plan which could be a viable alternative to Labor’s NBN. And as such it deserves some further scrutiny to see how it measures up against the already-under-way NBN build.
Many have criticised the Coalition’s plan simply on the basis that fibre is better than copper and therefore any plan that doesn’t involve connecting every household with fibre is bad. Just repeatedly chanting “fibre is the best, chuck out the rest” may be catchy, but it hardly justifies spending the tens of billions of government dollars the NBN will require for the next 8 or more years.
The competing NBN plans shouldn’t be measured on the merits of each technology but instead on how well the needs of the community will be met, how much the government will need to invest and what future benefits will be derived from that investment.
In looking to the future, some have asked why would we spend so much deploying fibre if it’s going to take 10 years to complete? Surely after that long fibre will be superseded by a cheaper, faster technology? Unlike much of consumer technology which has a yearly update cycle, telecommunications infrastructure has a much longer lifecycle. We’ve been using copper lines to communicate since the late 1800’s, likewise we can expect fibre to be used and improved for many decades to come.
In fact optical fibre was already being used to transfer data in the 70’s. In Australia, Telstra and Optus spent the mid 90’s laying hybrid fibre-coaxial networks so they could offer pay TV and high speed broadband services. Optical fibre now forms the backbone of telecommunications networks in countries all across the world. Just this year researchers at the University of Southampton have achieved a throughput of 73.7 terabits per second over fibre. That’s fast enough to download the full contents of Wikipedia in less than a second.
So yes, technology is changing and improving at a rapid pace. That doesn’t change the fact that fibre is the high-speed, high bandwidth transmission medium of choice. The Coalition has acknowledged the importance of fibre and will be replacing much of the current copper network with fibre. They won’t be connecting fibre to your home though, only as far as the node for the “brownfields” established suburbs. New “greenfields” suburbs will have fibre installed under either future government.
So now we must ask what exactly is a node and why does Malcolm Turnbull want to build 60,000 of them around the suburbs and towns of Australia?
In essence, a node is a miniature telephone exchange crammed into a powered street cabinet. Today if you happen to be lucky enough to live next door to a telephone exchange you should already get very good ADSL2+ speeds. Once the signal travels further down the copper phone lines, it becomes weaker and the connection speed slower. Hence the creation of the node which brings a mini-exchange closer to your home, thus resulting in a strong signal and faster broadband speeds.
Another technology, VDSL2, (Very fast DSL) can also be used to increase your connection speed. VDSL2 uses a mixture of techniques to both increase the signal and reduce the noise to allow connections up to 100mbps. The maximum speeds will only be for those within 500m of of the node, as you can see in the image below from BT which has capped it’s VDSL at 76mbps. Further improvements will soon increase this limit up beyond 100mbps, but only for those close enough to the node. That’s why the Coalition needs to build 60,000 nodes; otherwise their promise of a minimum 25mbps connection wouldn’t be possible.
This is where we begin to see the limitations of continuing to use copper wires for broadband. Even using the very latest high-speed transmission techniques, the highest speeds over copper are limited to short distances. Beyond the first 50-100m speeds drop drastically. So while placing a node near your home will make a significant improvement to almost any broadband connection, it’s a technology that will ultimately be replaced or limited to apartment buildings and other short distance applications.
The Coalition understands that ultimately our fixed telecommunications network will be almost all fibre with some wireless/satellite in remote regional areas. Their plan offers fibre to all greenfield suburbs and also offers older suburbs a “fibre on demand” model where the user pays for fibre connection to their premises. This appears to be a good compromise which allows only those that want a much faster connection to pay the extra cost. Unfortunately it’s an idea that’s unlikely to ever work in practice (paywalled link) and even if it did, the connection fee would be in the thousands for a fibre connection that would eventually need to be replaced in a future fibre-to-the-premises rollout.
So the question remains, if the future of telecommunication is almost certainly going to involve fibre all the way to your home, why bother building all these nodes? The answer is a combination of cost, timelines and utility.
High-speed broadband deployed via fibre-to-the-node technology is cheaper than fibre-to-the-premises in the majority of established suburbs. Based on construction costs for broadband deployments overseas the Coalition expects that FTTN will be 3 to 4 times cheaper than an FTTP solution. Not only is FTTN less expensive it’s also faster to deploy. Around 140 premises will be activated each time you build and connect a node.
While the current FTTP rollout remains (mostly) within budget it’s also drastically behind schedule with the current corporate plan being updated with substantially reduced targets compared to 2 years ago. The project was expected to start slowly, refine the rollout methods and then eventually ramp up to 6000 homes a day. Right now though it’s hard to see how the project will meet its target date of 2021.
So you can see why the Coalition would claim to be able deliver high speed broadband both cheaper and sooner, FTTN qualifies on both counts. So then what about utility? Make no mistake that while the Coalition plan is cheaper it still requires an estimated $29.5 billion in funding to complete. That’s a very significant investment for which you’d expect to have infrastructure that will be fit for purpose for many years to come.
Do we really need the 100/40 Mbps (and soon 1000/400 Mbps) connections the NBN fibre can offer? Or will broadband delivered through nodes be sufficient? At the end of 2012 the average Internet connection in Australia was 4.2 Mbps. That puts Australia at 41st place in the world. The Coalition’s promise of a minimum 25 Mbps would easily push us into the top 10 or even top 5 countries in the world. But how long can we expect to keep pushing copper lines in the search for faster broadband?
Even when armed with the knowledge of past and future broadband trends this is a difficult question. The Coalition doesn’t have an answer to this but they do hint at the possibility of future upgrade paths. On page 14 of the Coalition’s “Background Papers” we see the table below which suggests that a FTTN rollout would generate saving that could easily pay for a future fibre upgrade.
Net present cost, as demonstrated here is a method of comparing alternative investments. Typically you’d also include expected revenue, but from the basic cash flow model presented here you can see the concept.
It’s an intriguing idea that we could use FTTN in order to save for a future FTTP upgrade/replacement. The background papers go so far to suggest that after 10 years the savings accumulated could easily amount to $12.7 billion. While the timeframe of 10 years seems appropriate the model presented here is flawed.
Most of the arguments against using nodes seem to be centered around the idea that the copper network will soon degrade to a point where it will be rendered completely unusable. Telstra CEO David Thodey has already spoken out against the copper doomsayers indicating that Telstra’s copper network could last another 100 years. While that comment should be expected from someone that would like to sell its copper network to a future Coalition government, there can be no doubt that with enough maintenance the network will continue to remain fully operational.
The downside is that maintaining 60,000 nodes and connected copper lines requires significantly more “truck roll” (on site work) than a passive fibre network would. So the Coalition’s figure of $30 opex/year extra for FTTN is far too low. Indeed Verizon (USA) estimates that copper maintenance cost them $110 USD per year more than fibre. There are a number of reasons why a FTTN solution would cost more to maintain. Simon Hackett covers some of these reasons in his presentation at CommsDay Sydney 2013.
In addition to using a low FTTN opex cost, the Coalition’s model also uses a high fibre capex cost too. NBNCo expects the fibre connection cost per premises to be $2400 not the $3600 figure the Coalition is using. Now if we go back to the model and add an extra $110 for copper maintenance costs (ignoring the exchange rate and higher labour costs in Australia) and then we also lower the NBN fibre cost to $2400. The savings from delaying the fibre rollout vanish. In fact the FTTN and FTTP rollout now costs an additional $2 billion ($160 per household).
So we’ve established that a fibre-to-the-node solution would have slower download/upload speeds, it would need to be replaced after 10 or so years and it would ultimately be the more expensive option. So knowing that would I recommend using FTTN? Actually yes, I would.
While the Coalition views FTTN as a cheaper way to get out of their commitment to provide universal high-speed broadband, it can be put to much better use. Fibre is very expensive to deploy in some areas, as found in the New Zealand where Chorus is rolling out a national fibre network. They are experiencing costs up to $8000 NZD per premises in busy urban areas which make up around 10% of their expected coverage area. Likewise installing nodes in some areas will be a more costly proposition due to lack of suitable powered cabinet locations or copper wiring in need of remediation.
The solution to both these issues is to create a hybrid network. Part of the reason as to why BT have been so successful with FTTN in the UK is because they have developed very good models to make sure they first deploy FTTN to the areas where they stand to make the greatest ARPU (average revenue per user). Using similar logic we should be using a cost/benefit model in our network rollout. If it’s done right we could have the best of both worlds.
The overall cost would be cheaper because we could used targeted FTTN deployment to the most expensive 10%-30% of the network. Other simple changes like connecting apartments to the network using a fibre-to-the-basement deployment would save time and money. Further savings would result from the now reduced capex demands in what would have been the most expensive fibre rollout areas.
The overall number of NBN-connected premises would be much higher earlier in the project giving more Australians access to high-speed broadband. With that comes the added benefit of more customers on the NBN sooner. And more NBN Co customers means more revenue and a quicker return to profitability.
It’s well within reason to believe this could all be accomplished by 2020 leaving Australians with a now debt free NBN Co that’s able to produce a modest profit each year. It’s at this point where we can now return to the 10%-30% of the network still running FTTN and replace the nodes with FTTP. With contractors that now have a decade of experience laying fibre and funding that can come from the NBN Co we could knock over the remaining fibre rollout over the next 5 years.
No doubt this still won’t please the almost dogmatic fibre supporters or satisfy the cost cutting Coalition supporters, but the future of telecommunications networks is hybrid. So if Australians want a high-speed network of the future with a pricetag that we can afford today, then we need to make a choice. Not a choice between copper or fibre; but the choice to use both.
Image credit: Various