Wednesday 5 September 2012

NBN History and Delays.....or prudence?

Ok, so it's been a while since my last post. I've been busy beavering away around forums and Tech sites trying to spread the word on the NBN (would you believe some people don't even know what it is?...). One thing I've noticed over these past few weeks is the argument from the Coalition and from supporters- "The NBN is drastically behind schedule- they were supposed to have 150 000 connections by now! This just shows how much of a waste it is and how much longer it is going to take."

Is this true? There's no question that it is indeed a long time, in anyones' book, since the NBN as we know it, was started. But why? For this, first I would suggest you read my new page NBN History. There is alot of backstory of the NBN and it is necessary to understand before we continue, where we came from.

Regulation- Bane and Saviour


Such as life in Australia that we have many, MANY, MANY regulations/laws for just about everything. NBNCo. needs legislation to make it exist and give it power, albeit funded by the government (those funding requirements need legislation too). The power OF NBNCo. must be regulated, to ensure it doesn't get above itself, even with the funding restrictions imposed by government. Regulations must be put into place surrounding how NBNCo. will offer services to RSP's (Retail Service Providers as they are in an NBN world) and how much they can charge for this- enter the ACCC. Legislation must also be passed to separate Telstra, seeing as the government can't just force a company to separate just because they use publicly bought infrastructure and have paid to use it (shareholders); the ACCC has to approve this too.

The upshot of all this boils down to- time. These agreements with Telstra took time. These regulation discussions and negotiations took time. This new legislation passing parliament (with a hostile Opposition too) took time. And add to that a hung parliament with the election of Gillard in mid-2010, meaning each and every piece of legislation must be gone over with a fine tooth comb before the Independents/Greens could force its' passage through parliament with the government, due to a completely hostile Coalition on the issue.

The discussions with Telstra started in late 2009, but they aren't completed until mid-2011 and they aren't signed into existence (with agreement from shareholders) until December 2011 or enacted until March 2012....only 6 months ago! The legislation to create, give power to and manage NBNCo. isn't passed and completed until 2010, well over a year after NBNCo's birth. And the regulations surrounding NBNCo's WBA (Wholesale Broadband Access) are still ongoing, due to constant objections from the big Telco's who are barging for more money, more restrictions on pricing and shorter contract terms.

The agreements with Telstra are crucial- they allow NBNCo. to sign a lease for accessing Telstra ducts, pits and exchange space for the next 35 years, at a cost of around $5.5 Billion. This is essential, so that NBNCo. DON'T have to dig holes and rip up driveways and paths (sound familiar....) and it saves them a bunch of money and time. They also allow NBNCo. to migrate customers off the Telstra copper and HFC (cable internet), 18 months after those areas are provided with live FTTH, ensuring the customer uptake is high without exception and maintaining NBNCo's business case which will allow it to service its' loans and thereby not costing the taxpayer by using bonds to borrow the capital. This part comes at a cost of $4 Billion over 8 years. The final piece is about USO (Universal Service Obligations) of Telstra to the government to maintain Payphones, Emergency (000) Services, traffic lights etc. This doesn't directly impact NBNCo, but will in the future. This amounts to about $1.5 Billion. Total agreement cost, $11 Billion. But, the agreement has a clause- If Telstra has to sign this agreement, Optus must also sign a binding agreement to migrate customers off their HFC (cable internet).....more on this next post....

That hurdle is overcome and, as stated, is finally signed and sealed in December 2011 and implemented in March 2012 by Telstra shareholders (99.5% agree in fact....). However, in the mean time, there is an interim agreement that allows NBNCo to use Telstra ducts etc. to enable it to start trials.

Now, this is where things start to get a bit dicey in terms of timing. Do we count the schedule, released by NBNCo. in mid-2010, as part of the original Corporate Plan, as the beginning of the construction of the NBN? After all, they're putting in fibre to premises, signing satellite contracts, running wireless backhaul and planning towers. Or do we count it AFTER the trials have finished? After all, you can't very well rollout infrastructure for millions of customers without first trialling it in real-world scenarios? What if something doesn't work properly, or takes longer than expected? This is the point of trials.

Trials and Timing- Subjective or Objective


And here I suppose is the crux of the issue. NBNCo. are a government company who has a duty to provide the Australian people with the best, most cost-effective and fastest build possible. To go ahead, full steam after nothing more than on paper planning would be extremely reckless- Australia is a vast country and many factors in geography, existing infrastructure, housing types and so on could effect the rollout. These need to be analysed in a real-world rollout. Hence the trials. The first FTTH trials started to rollout in Tasmania, where a State NBN had already been planned and given to Aurora Energy and NBNCo. had tied up a deal to incorporate that into the NBN at large. That saw the first fibre trials rolled out in 2010 by Aurora  and after legislation was passed in early 2010, saw that come under NBNCo. jurisdiction. The trials were completed and the first services tested in mid-2011, modest as they were (several thousand premises were connected, a few hundred used the service). Several thousand are now connected since commercial services have been turned on, with tens of thousands now passed.

The next fibre trials, on mainland Australia were in Armidale, Kiama, Townsville, Wilunga and Brunswick. These covered varying geographies and societal differences. They were conducted from September 2011. Again, initial takeups of the service were small, but have now risen to many tens of % in several of the sites after commercial service was turned on, including to nearly 35% in Kiama. While this seems small in the overall scheme of things, takeups of 15% overseas in similar rollouts are considered excellent. Don't forget too, takeup is largely irrelevant because of the Telstra deal- all those on Telstra copper (most people) will be migrated automatically to the NBN after 18 months of receiving fibre in their area and the FAN being active. Therefore even small takeups of 20% now, will still result regardless, in 70-80% takeups in 2 years time.

In the meantime, satellite deals for NBNCo's "Interim Satellite Services", which see 6/1 services supplied over existing commercial satellites, were signed in May 2011 with Optus and IPStar and offered from July 1 2011. These services have much cheaper prices than normal commercial services (see SkyMesh and compare NBN Satellite with IPStar satellite plans) and much higher speeds, with much higher downloads. And the contention is managed on these satellite slots, meaning these NBN customers continue to receive higher speeds when in peak times, on the same satellites. Once NBNCo. launch their own satellites, by 2015, there will be 90Gbps shared between 300 000 premises. This seems low; after all, if everyone were to download at full speed the entire time, it would result in 450Kbps maximum speed. However, the satellite area is so vast, the likelihood of even half these services downloading at full speed (12/1) at the same time is negligible. However, NBNCo. are also only predicting some 120 000 people to use these satellite services (see the new Corporate Plan), making contention even easier to manage.

The second mainland fibre trial sites partially skipped "trial" phase and instead switched on commercial services as soon as connected. These were done at 15 other locations and expanded 4 of the original locations in the trials, resulting in some 65 000 connections by April of 2012. However, reported takeup was only some 3700 actual services by then, on fibre.

Wireless trials were conducted from April this year, with a few hundred premises outside the fibre footprint of the trials in Armidale, being connected via NBNCo's fixed wireless LTE solution, for current speeds of 12/1. This was expanded to Toowoomba and Tamworth in June. There are a little over one thousand people on the NBNCo. wireless plans.

All these trials, as you can see, have been planeed, executed and spaced out according to the best ways to tackle the various geographies of Australia. The trials are now totally completed. Full commercial rollout of the NBN services, announced as part of the 3 Year Rollout plan in very late March this year (just after the Telstra agreement signed in December took effect in March), were to begin in July. Some 750 000 connections will be started or completed by the end of this year and 3.5 Million will be started or completed by 2015.

Why did I just go through, what may seem like, a giant advertisement for the work NBNCo. have done? Because it gives an indication of the sort of groundwork that has to be done for a rollout on this scale. It cannot be simply rolled out as soon as the numbers are crunched. Real-world numbers must be seen and compared to ensure the best outcome. And this is exactly what has happened as a result....

Blowout? or Blow hard?


So here we come to the point- is NBNCo. on time, on budget and prudent in their efficiency? Most of the mainstream media would have us think not- the term "blowout" was used in conjunction with the new Corporate Plan (2012-2015) announcement more times than I've seen it used in the 5 years previous....

The Australian: "Extra $3.2bn needed to fund NBN" (Google the Headline to get around the paywall..)

None of those are particularly positive for NBNCo. or the government. But let's have a look a bit more closely about what the new Corporate Plan actually says....

1- An extra 3.9% in Capex (Capital Expenditure) or $1.4 Billion
2- An extra $3.2 Billion in Opex (Operating Expenditure)
3- A slight dip in revenues of $0.6 Billion
4- Internal Rate of Return raising from 7% to 7.1%
5- 9 Month Start delay, compressing to 6 month end delay (Jun 2021 instead of Dec 2020)

Now, this all seems quite odd- NBNCo. appear to be spending more to build the network, more to operate the network, taking longer to build it AND receive less from customers, yet they will produce MORE return on investment?? Firstly, here's the mistake often overlooked by the media- the IRR of 7.1% is over 30 years. The Capex, Opex and Revenue forecasts are over 10 years. (till 2021)

Second, the start date was pushed back, but that doesn't actually mean it's taking longer to build. The total build time is actual 9 years 3 months (Mar 2012- Jun 2021) but it was supposed to begin in July 2011 and end in Dec 2020 (9.5 years). Primarily because of the Telstra Definitive agreement vote being pushed back to December 2011 and the agreement not becoming effective until March 2012, meaning NBNCo. could not access detailed information on ducting, current infrastructure condition and other detailed work offered through the Telstra deal, the start date for commercial (or volume) rollout was pushed back from Jul 2011 to Mar 2012.

3rd- let's have a closer look into that extra Capex and Opex:

Capex increases for 2 main separate reasons:

1- Build Drop- This is the method used by NBNCo. in predetermining and pre-building the lengths and connections of the fibre being rolled out through every street. Originally, NBNCo. were using a Demand Drop system, ie. The main local fibre would be run past all premises and then when the customer wanted (or was migrated to) an NBN connection, an NBN crew would then come out and connect the premises connection device (PCD) to the street fibre (the drop). Then the team would install the NTD and connect it to the PCD to have a full service operational by the end of the day for the customer to ring up an RSP and be connected.

Now however, NBNCo. have chosen the "Build Drop" method. This is where the drop to each premises and the PCD are installed AS the local fibre is rolled down the street past each premises. This means, when a customer wants an NBN service, 2 technicians simply come out with an NTD and connect it to the PCD and the customer is up and running. This technique saves on double work requirements (treading the same streets twice) and is expected to save NBNCo. money and time in the long run, while bringing forward some extra costs (speeding up customer connection time artificially with the Build Drop).

2- The Optus Agreement- This Agreement (which a clause built into the Telstra agreement required) was not assumed in the first Corporate Plan. This agreement has seen much debate, as it essentially pays Optus $800 Million to migrate its' internet customers off their HFC (cable internet) instead of letting NBNCo. slowly remove customers as natural speed improvements tempted customers to change to the NBN. It is a lot of money to pay Optus, especially when they are likely not to actually compete with the NBN anyway, however, the ACCC believed it was for the good of the NBN business case (and anything that increases the business case for the NBN increases the likelihood the NBN stays intact- the best outcome for the consumer) that this deal should be allowed. This requires NBNCo. to accelerate rollout in these areas to ensure it meets the guidelines of the agreement (all Optus HFC areas to be connected by 2018) thereby requiring NBNCo. to move forward Capex to cover this.

Several other increases in Capex included:

- Greenfields- NBNCo. must connect Greenfields even outside the current fibre footprint now, raising Capex costs to get fibre to these areas in the next few years.

- Network Distances- More accurate network data, available after the Telstra agreement came into affect, has caused the amount of fibre required to be increased after this data was analysed.

- MDU's (Multiple Dwelling Units)- NBNCo. had underestimated the Capex required to connect all MDU's with fibre

- Higher Labour costs.....don't think this needs much explanation.

Opex increased for several reasons:

1- Once again, the Optus deal, which will result in more customers being online on the NBN faster than originally planned, meaning higher Capex to deal with this (but also correspondingly higher Revenues).

2- The Telstra deal gave NBNCo. access to more dark fibre and exchanges than predicted, meaning Opex goes up for using these systems, but Capex goes down for not having to provision them separately.

3- Small extra requirements for IT and network interactions between NBNCo. and Telstra as well as with other RSP's.

Revenue decreases for 1 primary reason- The agreement NBNCo. made with RSP's in mid 2011 for a rebate on the first 150Mbps of CVC, before a POI reaches 30 000 connected customers. This erodes NBNCo's initial revenue takeup in each of the CSA's (Connectivity Serving Areas) until a critical limit is hit and all CVC has to be paid for. This can be seen on page 62 of the Corporate plan- you can see the CVC earnings are pushed back significantly, taking nearly 4 years to come back to initial predictions. This lowers overall revenue earned in the first 10 years (covered by the new Corporate Plan).

All this is pretty dry, but the point about all these things is, the major changes to the costs and thereby the explanation for the "blowout" (I don't like that word- I'd hardly call 4% or $900 Million a blowout when BHP just dumped $50 Billion on an investment a few weeks ago....) is quite clear and it is, primarily, due to changes in scope. Both the Optus deal (not part of the original Corporate Plan) and the New Developments requirements to connect all houses over 100 premises, rather than over 500, as NBNCo. originally budgeted for, soaks up 50% of the cost increases. Add the build drop for another 20-30% of the cost now, but less costs later. The 20-25% left (or 1% overall) in increased costs shows actually how close NBNCo. assumed and predicted many costs to be in its' original Plan. I don't know about you, but I think if my custom built house was to come in 1% over budget + 3% over budget for additions I made to it, I'd be pretty happy with that.... (remember that Opex increase is a move forward of costs from the Optus agreement. Those costs would always have been required, they're just required sooner rather than later when there is more Revenue to cover them)

NBNCo. appear to be doing a decent job of designing and rolling out a network over the past few years, that has constantly had its goal posts moved. Their original predictions about the costs within the scope they were given, was within about 2%. And the scope changes have added around 10% extra cost of 10 years, but much of that is clawed back through higher revenues and no longer required labour for the Demand Drop.

Is takeup relevant?


Well, I guess that depends on what you believe. Does it matter that people aren't jumping left, right and centre onto the NBN straight away? (although in some cases they are- see Kiama and Willunga takeups of 25% and 38% respectively) Not really- the Telstra deal, which the Coalition regularly overlook in explaining how uncaring people regard the NBN, requires that any FSA (Fibre Serving Area) that has full and active connections for its' coverage have ALL premises within its' footprint migrated to the NBN, off Telstra copper or HFC and now Optus HFC as well, within 18 months of commissioning.

This means, at an average of 12 months from beginning work on an FSA to completion (some will take 16 months, some will take 9 months) that within 2.5 years of a FSA being started, all ~40 000 premises connected in that FSA will be on the NBN. That's a 100% takeup rate of all fibre connected premises (a 70-80% total takeup rate once wireless only premises and vacant premises are taken into account, as they are in the Corporate Plan). There are dozens of FSA's finished each year during the rollout. This means, after an initial delay, there will be some 40 000 premises connected every few weeks. Regardless.

Don't agree with me about any of this? Comment away, or, come join me and others debating it over in the Whirlpool forums (Fighting the NBN FUD, 2012-2015 NBN Corporate Plan or the Coalition NBN Position are 3 threads many, including myself, frequent readily). Many analysts and tech journalists have said since the new Corporate Plan has emerged that now that NBNCo. is in full rollout mode, their predictions of takeup and spending should be judged from here on out. I think this is fairly reasonable- CEO of NBNCo. Mike Quigley said himself, along with Senator Stephen Conroy, that this is the first Operational Corporate Plan. The original was a Corporate Plan for a company starting up. But this new Corporate plan should be used to judge their performance over the next few years (new Plans will be released each year regardless).

So, NBNCo. have vaulted most of their hurdles and are now on the beginning of the 200m back straight. They're due to pickup speed over the next few months and years and we'll soon see, by June next year in fact, whether NBNCo. are capable of delivering. And if they are, what is going to happen at the next election in late 2013.....






3 comments:

  1. well that is an great article i had a great time while reading this article.
    glue removal

    ReplyDelete
  2. nice reads mate, check out my site http://www.nbncompared.com.au it compares nbn prices and plans from all ISPs, well not all yet, but im working on it

    ReplyDelete
  3. NBN is facing a lot of issues. Attacking Turnbull is like attacking the NBN service as a whole. What we need to do is to be positive about these turn of events. Let us keep our hopes high that internet connections will be better after the installation of new satellites.

    ReplyDelete