If you had to pick a country in the world where it would be most expensive to deploy a Fibre-to-the-Premises (FTTP) network, Australia is always going to be very close to the top of that list.
The world’s first FTTP networks were deployed in cities like Tokyo, Seoul and Hong Kong, where almost everybody lives in big apartment buildings in very high density areas – that means that if an operator deploys 1km of fibre down a street it can connect thousands of premises and do it at very low cost.
Indeed, if you look at the world’s leading FTTP market of Hong Kong, you will find that operators such as Hong Kong Broadband Network have deployed FTTP on a cost per premises of below US$150 – hence why 90% of people in Hong Kong have access to FTTP.
By contrast, Australia’s cities are vast, sprawling outwards for many kilometres and, apart from in our relatively small CBD areas, the majority of Aussies live in standalone dwellings with decent sized front gardens that set the house a fair distance back from the road – these are key in driving up our costs.
The stats tell their own story, Singapore – which has a national FTTP network – has population density of 7,987 people per square kilometre and Hong Kong 6,442 people per square kilometre. Here in Australia we have the 14th-lowest population density in the world, with fewer than three people per square kilometre.
There is simply no getting away from the fact that delivering universal FTTP in Australia is always going to be more expensive than almost anywhere else in the world. Conditions here do not lend themselves to a cheap and fast FTTP deployment to every single premises and that is not going to change.
While costs may come down with new deployment processes and delivery methods, the costs of deploying alternate technologies will also improve, so it's highly unlikely that fibre to the premises will ever be cheaper than the other technologies nbn is deploying, nor anywhere near as quick to deploy.
There are thousands of areas on our suburban fringes where a kilometre of new fibre deployment might only actually connect a couple dozen premises – and these are areas that are less than an hour from a major capital city. This inevitably drives up per premises costs.
Furthermore, you then have to look at the type of premises that we have to connect, remembering that we have to make access to the nbn™ network available to all; we cannot pick and choose who we connect based on how much it will cost to connect them.
This means that, while some private operators such as Verizon in the US are able to reject an application for an FTTP connection because of the high cost involved – thereby helping to keep their overall FTTP connection cost per premises down, we do not have this ability.
Verizon has been able to keep its overall cost per premises for FTTP down by focusing on deploying its network to high-margin, low-cost metropolitan areas and ignoring more outer suburban and rural areas where its cost to connect would be much higher – nbn cannot adopt this model.
As a result, over the last few years nbn has had connection costs for individual premises that have run to tens of thousands of dollars – and these are not isolated cases by any means – inevitably driving higher costs for our overall FTTP deployment.
Remember, we cannot simply shove these premises onto our Sky Muster™ satellite service. Capacity on Sky Muster™ must be preserved for our rural and remote Australians – and we can’t just shove a problem premises onto Fixed Wireless either, as there is unlikely to be coverage available.
You also have to look at how we are connecting premises to our FTTP network in comparison to how operators in other countries are connecting their premises – this is another big reason for why our per premises costs are higher.
In many Asia Pacific markets, and even in the US, operators are able to deploy their FTTP networks aerially, connecting premises to the network via overhead cables that run from the telegraph pole to the exterior of the premises – this is the cheapest way to deliver FTTP.
However, in Australia the vast majority of our fibre is deployed via underground ducting – effectively plastic tunnels – which means that the fibre is safe and secure for the long term, but which is considerably more expensive than going aerially – and takes a lot more time to do.
This is borne out in our costs to connect, where we spend an average of $1,500 going from ‘pit to premises’ outside a property, $2,150 for the local distribution network component and a further $750 in duct leasing costs to Telstra – a component few other operators would have to build into their cost per premises.
Other operators, such as Chorus in New Zealand, have chosen to often deploy their fibre via micro-trenching across people’s front lawns – where the fibre is buried in a very shallow trench – and have even run fibre down people’s fences and then into the property.
These means are much quicker and cheaper than running fibre through a properly constructed underground duct, but often leave homeowners unhappy and this method does not protect the very fragile fibre as well as if it is safely inside a protected duct.
We are not alone in the world in facing these problems – operators around the world are struggling with the same issue.
Just look at Google Fiber in the US, where they began their FTTP deployment with high hopes of connecting millions of premises right across America – but they soon found that building FTTP was expensive and time consuming and so ceased the deployment last year and are now focusing on delivering Fixed Wireless networks instead.
We have seen the same thing happen in France, where their plan was originally to deploy FTTP to the entire country by 2030 and have already connected several millions of FTTP premises in their major cities such as Paris and Lyon – but the full reality of the cost of deploying FTTP outside of the big cities has now dawned on the French Government.
That’s why the French Auditor General has now recommended that the country move away from its all-FTTP goal and instead embrace a ‘mix of technologies’ to deliver better broadband to every premises in the country by 2030.
This is exactly why it also makes sense for us to be flexible in our approach and have a range of technologies at our disposal and we then use those technologies in the area most suitable. By doing this we can not only get end users connected much faster but we can also do so at a lower cost.
Peter Ryan is Chief Network Engineering Officer at nbn.