The NBN Strategic Review: its dodgy accounting, and omissions (Guest Post)

By James Chisholm (james at chisholm dot id dot au)

Vindication is a bitter-sweet pill to swallow. We, the Australian taxpayers, have finally taken delivery of the Strategic Review into the National Broadband Network, which has shown what many of us have been saying since the Coalition announced their Fibre-to-the-Node alternative back in March 2013 – that there is no way that the Coalition can possibly provide 25mbps connections to all households by 2016. I’ve previously written about why I believe FTTN to be a completely inappropriate network architecture for this country, and many of the issues I raised in that post aren’t addressed in the Strategic Review. It did take me a while to get through, but I made it through the whole document. What stunned me was the way so much critical information was redacted from the document, and how much vital information was left out to make FTTP look like an inferior solution.

My provisional analysis is this – that the “multi-technology-mix” we are going to get now from NBNCo is a dogs breakfast of outdated technology for 76% of the country, with a lucky 24% who will get to get the next generation of Fibre-to-the-Premises connections, which we know are already capable of 1000mbps throughput (and were scheduled to have this capability made available this month). I call this MTM a dogs breakfast for several reasons, the primary one being that upload speeds are not addressed at all by this mix. Neither are fundamental details about the FTTN network, such as how many nodes are going to be implemented. Additional reasons are that the HFC networks are demonstrably congested in many locations already with only a small number of subscribers, that FTTN at 25mbps is (as I’ve pointed out with my other blog post) a totally inappropriate solution for this country, and that this technology mix puts control of the network firmly back into the hands of Telstra.

It comes as no surprise that, given the sheer volume of ex-Telstra employees that Malcolm Turnbull has appointed to NBN Co, that they are hell-bent on sending as much of the public funds back into the hands of Telstra, but to do it in such an overt manner is quite frankly disgusting. I, like many of my contemporaries, have a rather low opinion of Telstra – having previously worked for them, I’ve never before, or since, encountered a company with such incredible contempt for their own customers, nor worked with a company with such incredible incompetence. To get 6 phone lines installed for a former employers new SHDSL connection, took 13 visits by Telstra technicians in 2011. Personally I would like it to be placed on record the level of Telstra shareholdings of all senior NBN Co staff.

Anecdotal evidence of demonstrated small business needs for higher upload speeds, which will not be provided by either HFC or FTTN, is all well and good (and I wrote 6 paragraphs about personal tales of former employers and customers of mine, I’ve redacted this to keep this shorter) it’s straying further away from the analysis I’ve been meaning to do. So let’s get to that, shall we?

“The national broadband network policy was released many months ago by myself and Malcolm [Turnbull]. No-one has been able to question the costings. It is absolutely bulletproof.” Tony Abbott, 6 September, 2013

First of all, there is the cost blowout to the LNP bulletproof, 100% costed policy. After crying for months about Labor party waste, we now have a strategic review which revises the figure upward by 25% on the cost of a Multi-Technology-Mix network. The review also claims a cost blowout for a full FTTP network to be $72bn, whilst redacting the associated costs to date for the FTTP rollout.

Assuming that we believe these figures (and we really can’t, due to the fact the review has already been outed for fudging the figures; see this) one has to accept certain caveats to them – the actual rollout is now only going to occur to 70% of the population (24% FTTP, 46% FTTN), versus to 100% with the labor plan. So that’s really an unfair comparison – $41bn for a 70% rollout versus $72bn for a 100% rollout.

A more appropriate comparison is to scale the cost of the FTTP rollout down to 70% and then we can crunch the numbers slightly better – $41bn vs $50.4bn is more of an “apples vs apples” comparison of the figures.

Additionally, the strategic review also doesn’t mention how much it’s going to pay for:
a) the Telstra Copper network
b) the Telstra HFC network
c) the Optus HFC network

These networks aren’t going to be given to them for free. Telstra CEO David Thodey has stated Telstra expects to be paid the $11bn it was already contractually going to get for decomissioning the copper, as well as additional fees to access the copper they were supposed to decomission and there will be an additional cost to acquire or lease the HFC network for the remaining 30%. I’ve seen estimates that the two Telstra networks will cost $30-50bn on top of the $41bn for the FTTN – which would mean that the MTM network will far exceed the cost of rolling out FTTP to the whole country!!

By its own admission on page 19 of the strategic review, the MTM network is going to be outdated within 5 years of completion. Once it is outdated, it will need to be replaced with FTTP. Simon Hackett’s explanatory blog post about why he feels HFC is the right way to progress the rollout faster even admits that FTTP is the ultimate endgame. And this is the crux of the stupidity of this MTM – why bother rolling out last millennium’s technology when the limitations of it’s shared bandwidth potential are well and truly known, and cannot take us through the majority of the 21st century?

A big issue for me is the sheer number of nodes and footprint that each node will service – will it be 70,000 nodes for the country? 80,000? If the goal is to get everyone on 100mbps by 2019, it’s probably going to need to be more like 100,000 nodes. The Strategic Review does not specify how many nodes they think will be required, nor the maximum copper loop length. It also doesn’t factor in possible objections (in the form of lawsuits) by councils, replacement costs of equipment, specifications of these nodes and how well they are going to cope with the 47 degree days we’ve had in recent summers, or what is going to happen in Queensland when it floods and thousands of powered nodes end up underwater!

From a personal perspective, as most of you probably can guess, my household is a high-tech household. Judging by my routers IP address table, there are 22 devices on my home network at the moment. I’m lucky that when we moved a few months ago, we’re really quite close to the exchange – approximately 400m, as the crow flies – and hence my naked ADSL2+ syncs at 21mbps and I get about 17mbps throughput. However, when I’m playing online games on my PC, my eldest daughter is uploading a youtube video on her laptop, and my wife’s xbox-one decides it’s going to download a 2.2gb update for a new game (which is something that you get no choice in, and can’t be stopped without turning off the power), all of us suffer. So much of our entertainment these days comes from the internet – even our phone line is VoIP based so as to reduce our costs – and bandwidth needs are going to increase, not decrease. My daughter hopes to pursue a career in animation – she has already won an award at school for it – however I’ve had to explain to her that the reason her youtube uploads take several hours is as a direct result of this countries poor communications infrastructure. The NBN was supposed to alleviate these issus and drive this country forward providing a 21st century capable network – sadly, at this rate, it won’t happen.

I also do wonder if, when Tony Abbott said earlier this year “We are absolutely confident 25 megs is going to be enough – more than enough – for the average household.” if he actually understands the difference between megs and gigs? When referring to megs in the speed context, we are talking megabits per second of connectivity – and when talking about gigs, we are talking about gigabytes of data. To download a 1 gig patch for a new Xbox One game, on a 21 meg connection, takes 25 minutes. Put that connection speed up to 100megs and the download takes 5 minutes. Similarly, if we are wanting to download an entire game for the new generation of consoles, that’s currently around 40 gigs of download. On the 21meg connection I have, it would take approximately 30hrs. On a 25meg connection, about 24hrs. 100meg would reduce that to 6hrs – a speed which is approaching acceptable – but nowhere near the amazing speed of FTTP – which, once upgraded to 1000mbps (also known as a gigabit connection) would have that whole 40gb game downloaded in the time it took the 1gb patch to download on the 21 meg connection! It is a truly massive difference – and FTTP is so close to being able to do 1000mbps already!

In light of this, it seems completely and utterly stupid to spend $41bn to provide an outdated technology which must be replaced with FTTP in 2025. We cannot afford to waste this opportunity to do this network properly now, in 2013. If the cost-benefit analysis being commissioned looks at only the short term costs of this network over 10 years, then it will show that the MTM mix is correct – but if it truly is being done with Australia’s best interests in mind, then the terms of reference will look at the cost of the NBN over 20 years, including the cost and time to upgrade everything to FTTP.

Knowing that FTTP is the ultimate goal, to spend $41bn between 2013 and 2021 rolling out a network that will need to be replaced in 2025 with a full FTTP network is not cost effective. As the saying goes, measure twice, cut once – if the cost of rolling out a full FTTP solution by 2024 is $72bn, do it properly, and don’t waste this opportunity by rollling out an outdated technology that needs to be replaced fully 5 years after it’s completed.

Advertisements

Guest Post: Why it is pure folly to implement a Fibre-to-the-Node network in Australia

By James Chisholm (james at chisholm dot id dot au)

On one hand, I can see why Malcolm Turnbull wants to press forward with his inferior technology plan for the future… he firmly believes that the National Broadband Network was a major reason people voted for the Coalition in the recent election. Sadly, as has been proven by the recent campaign on www.change.org/nbn as launched by Nick Paine, over a quarter of a million of Australian voters have disagreed with him – the overwhelming majority of people who voted Liberal did so despite the inferior NBN plan. I don’t claim to be an expert on xDSL technology; just a geek who has worked in IT since 1995, and has a more than average knowledge of tech issues, but someone needs to point out to the Coalition that enough people care about this issue to sign a petition which they so arrogantly dismissed. They need only look at the NBN forum on Australian broadband website Whirlpool to find hundreds upon hundreds of informed opinions debating the merits and deficiencies of both proposals.

In case you are unaware of the differences between the two technologies, there’s a very good website called nbnmyths.wordpress.com which I’ve yoinked this summary from:

• FTTN is a short-term “stop gap” using old technology

• Most countries that have installed FTTN are now replacing it with FTTP (i.e.: To the same system as the Labor NBN plan)

• The investment in FTTN would be largely wasted when the inevitable upgrade to FTTP is required

• FTTN would be almost as expensive to implement in Australia as FTTP

• FTTN’s Speed-to-Price ratio is poor

• FTTN delivers vastly different performance levels depending on location

• FTTN delivers very low upload speeds

The detailed explanation goes into exactly why it’s not a good fit for this country. I strongly recommend you read it, and then follow up that reading with the ex-CTO for British Telecom, who explains here why deploying FTTN was a huge mistake for BT.

Australia currently sits at a very low place on the world internet speed rankings (as of January, we were ranked 40th) and has the opportunity to take leaps and bounds by implementing the right kind of National Broadband Network – such as NBN Co has been rolling out to date. However, the Coalition FTTN network is no substitute for this – even the manufacturer’s claim Fibre to the Node is a stop-gap technology till a Fibre to the Premises network is achievable.

I personally read the coalition plan for FTTN on the day it was announced and my co-workers distinctly recall me swearing profusely, in amongst shrieks of “what lies!”. I work in a software development company and our proxy server cached the request to download the PDF document from the Liberal website, such that when all 30 of us went to read the document, it only appeared to their server as one request. To suggest that each of us haven’t actually read the document is not only insulting but ignores the fact that their document was summarised into hundreds of news articles in the days and weeks following the Coalition announcement. As the internet would say, tl;dr  – Too Long; Didn’t Read – there several pages of complete crap which were purely anti-labor propaganda in what was supposed to be a policy document!

I hope Malcolm Turnbull realises that Telstra weren’t lying in 2003 when group managing director of regulatory strategy, Tony Warren, told a Senate enquiry that the copper network has a maximum viable life of 15 years left. The Coalition’s “fraudband” Fibre to the Node network will never achieve the speeds that Malcolm Turnbull and Tony Abbott are promising. Malcolm Turnbull in particular keeps pointing out technology trials conducted in lab conditions which use much higher grade copper wiring than is out in the wild in Australia, over shorter distances than most people will ever be able to get to a “node” or exchange. Two of the countries which Malcolm Turnbull regularly refers to as having successfully rolled out FTTN technologies are New Zealand and the UK. What he fails to mention, without fail, is that both of these countries are now looking at replacing FTTN with FTTP solutions, as are many of the other countries which have implemented FTTN. It remains really unclear as to why you would choose a technology which has not only been implemented but is now being superseded!

Turnbull is claiming that the copper will eventually deliver 100mbps by 2019 – however, the technical trials which that speed was achieved in was over a 400m distance on a much higher grade copper than Telstra have ever deployed, and only ever in lab situations. Considering that the Nodes are due to serve a radius of up to 2km, it’s impossible to believe that everyone will be capable of getting 100mbps from this technology which the manufacturer’s state will only work to a distance of 400m. As a side note, it’s interesting in that Alcatel-Lucent report on VDSL2 vectoring that they state “While fiber-to-the-home (FTTH) remains the ultimate goal” which implies that even the VDSL2 manufacturers acknowledge that it is an inferior technological solution to the Labor FTTP NBN plan!

Furthermore, the copper, which Telstra were being paid $11bn to decommission as part of the Labor Fibre to the Premises National Broadband Network, has been made so vital to the Coalition plan – and Malcolm Turnbull has stated he expects Telstra to give it over for no additional remuneration. Telstra’s CEO, on the other hand, David Thodey, has stated that he fully expects the contract with NBN Co to be paid out in full as well as expecting additional remuneration over and above what they are contracted to get from NBN Co already, to prolong the life of the copper – which will, in turn, make this slower FTTN network more expensive than putting in place a full FTTP network.

The Labor FTTP option is being deployed currently at speeds of 100mbps and is upgradeable to 1000mbps (1gbps) and is not subject to slowdowns during wet weather (when the Telstra pits fill with water), and won’t eventually degrade like the copper will. Most of the Telstra copper network is degraded to the point where it will not sustain a high speed signal – and if you doubt this, see here for a few examples. I have just moved house to a place which is 300m from the local telephone exchange, and in theory I should be getting 24mbps ADSL2 – yet I’m getting a lower speed than when I was living 3.2km from the same exchange. It took me 54 minutes to upload a 91 second HD movie of my daughter to YouTube on Sunday night. This same copper is supposed to provide me with a VDSL2+ connection at speeds up to 100mbps?

The initial Coalition plan is to migrate the copper ADSL2+ networks, theoretically capable of going at 24mbps download speeds, to VDSL at 25mbps (and upload speeds still stuck at the same, 1mbps). That will then be upgradeable in the future to a maximum of 50mbps and again to 100mbps by 2019. Whereas those in areas which have already had Fibre to the Premises NBN connections installed can get 100mbps today!

As most of Australia has known for the last few years, climate change (which the Coalition would have you believe is a myth!) has lead to some of the most extreme weather this country has ever seen. We have had record high temperatures, record floods, record droughts… what’s going to happen to Queensland when, invariably, as they have for the last 3 years during summer, they have major floods? All of a sudden you have all this mains-power running to all these cabinets and everything inside them will short out. What will the insurance premium be on these? Will an insurer even provide a policy for such a device in such an area – knowing full well it will be very likely to suffer catastrophic water damage? Have replacement costs been factored into the coalitions plans?

Another thing which needs to be highlighted is that Malcolm Turnbull is happy to invest in companies overseas who are replacing FTTN networks with FTTP networks, but doesn’t want the country to get the same kind of advantage he’s personally investing in! This is hypocrisy of a terrible scale.

Then there is the inevitable lawsuits from councils who don’t want the ugly and noisy FTTN cabinets every 2-3km throughout the suburb – see here for an example of what they look like overseas – and if anyone has a memory back to the Optus cable rollout of last millennia, there were many councils which not only balked at having the cables run from the power lines, but sued Optus to prevent the rollout occurring. I also feel it necessary to mention that they require power and air-conditioning to operate – not that the Coalition even cares in the slightest about the environment – and if (or should we say, when) a car veers off the road and knocks over a cabinet, all the houses around will lose their internet and phone until it is replaced.

I’ve completely avoided the issue of the Foxtel and Optus cable networks up till here – which are to be incorporated into the Coalition NBN network rollout as another technology option in addition to FTTN. Arguably, another reason for the Coalitions plan to leave these 20 year old networks in place is that the advent of IPTV, a service which provides HD quality TV shows for as little as $10 per month, will drastically eat into the cash-cow which Foxtel has over cable TV. I must confess I was personally disappointed when we moved house recently that my wife was talked back into getting Foxtel re-connected (via satellite) for $30 a month (half price, but only for 6 months) – I had even chosen an IPTV service which provided as much content and the ability to watch it in multiple locations, unlike Foxtel. By preserving the cable networks, the Foxtel bottom line is boosted for many further years… and let’s not even touch on the issue that cable internet is a shared medium and how badly it slows down during peak hours. I personally have witnessed a Telstra Bigpond cable connection go from 100mbps down to about 4mbps around 4pm due to congestion at my former employer – sadly, I haven’t kept the speedtest.net results from a year ago when this occurred – but this technology whilst technically delivering a 100mbps connection, will not provide 100mbps of throughput to all the houses it deploys to.

So between the additional remuneration required to Telstra, the cost (and wasted time) of lawsuits from Councils, it’s really difficult to see why Malcolm Turnbull would want to spend at least $30bn rolling out an inferior solution, which will need to be completely replaced with the original $44bn solution. If we’re spending billions of dollars of money to roll out a high speed national network – do it once, do it right, and don’t put all the power back into Telstra’s hands – wasn’t that why Telstra was forced to undergo structural separation in the first place – because they were abusing their monopoly power?  The NBN was originally conceived as a wrench to prise  Telstra’s monopoly grip on premier networking services away from them – hence why they are receiving such a large amount of remuneration for the decommissioning of the copper network. Here’s a rather paranoid thought – what if, under the FTTN plan, Telstra, as the monopoly provider of copper from node to premise, were to turn around and say we’ll only fix your line problem if you switch to our service on a 24month contract? Or will only provide 100mbps capable copper connections to their own customers? Without a Universal Service Obligation (which is currently in place for ADSL) they may well be within their rights to only maintain the copper for their direct customers. The potential for Telstra to put in place anti-competitive behaviours which provide them with untold advantages over this network is staggering – and I would not be surprised to discover that they have already analysed to great extent what the potential for getting a larger cash-cow from this FTTN network is.

Yet another thing to consider in this folly of a fibre-to-the-node plan is the number of nodes/cabinets being deployed. The coalition plan is for approximately 60,000 cabinets to be deployed nationwide – there is some significant speculation on the aforementioned Whirlpool forums about the density of each cabinet’s reach. It’s hard to project a radius from a cabinet as housing doesn’t necessarily neatly fit all that well into a circle, however if the average node cabinet costs $50000 (I’m guessing at this price, by the way) to deploy and the number of cabinets required grows by only 10000 then that blows the cost of the rollout out by a factor of $5bn. To provide all houses with a 25mbps capable connection will probably require more like 90000 cabinets, especially in lower density areas.

Malcolm Turnbull has also stated that a cost-benefit analysis will be done on the NBN project – however I suspect, strongly, that the terms of reference for this cost-benefit analysis will be very limited (possibly as far as two election terms) and make ridiculous assumptions, such as obtaining the Telstra copper at no cost. Malcolm Turnbull has repeatedly stated that the FTTP project will end up costing $90bn, and most recently, in his attack on all the constituents of Australia who have signed the change.org petition, increased that to $100bn – with no justification as to why. A fully informed cost-benefit analysis needs to be forward looking with a scope to the implementation and operational costs over the course of the next 20 years; accounting for how much of the Telstra copper network is already stuffed and barely capable of lasting more than 10 years.

Simon Hackett, CEO and Founder of Internode, has recently made several publicly available speeches on the problems with FTTN and how to build a FTTP network on an FTTN budget. Simon has run one of the most successful ISPs Australia has seen, without the benefit of having a behemoth of a cash cow behind him like Telstra, and has done a brilliant job of it. I would STRONGLY urge Malcolm Turnbull and the rest of the coalition to consult with Australian industry leaders such as Simon, and get their input as to how best provide Australia with a world class National Broadband Network and bring this great country into the future with a future-proof and upgradeable technology path.

I fully expect that Malcolm Turnbull, if he ever does read my comments, will call me out on my pricing assumptions and number of required nodes. However, I stand by my belief – that a Fibre to the Node National Broadband Network is completely unsuitable for this country, for all of the reasons I’ve listed, and more, and truly hope that the terms of reference for the cost-benefit analysis factor in a longer period of time than the next two election cycles, and that if the Coalition proceed with the folly of their FTTN plan that they are held to account at the 2016 election – as I highly doubt that there is any chance with a FTTN network that they can provide the guaranteed minimum of 25mbps as stated by Tony Abbott during the policy launch, and that it will end up costing far more than Labor’s FTTP rollout will.