Sunday, 15 September 2013

Turnbull's Fantasy: The Copper Fairy will come in the night and fix everything.

One of the consolations of losing the All-Fibre NBN (for the 93%) will be watching Turnbull's political career unwind.

Will his leaving politics be gracious and dignified like Julia Gillard or the mother of all dummy-spits? With his sharp tongue and unlimited belief in his ability and infallibility, I'm looking forward to the dummy-spit.

Turnbull delivered for the Libs in the election campaign with a "blocking" move: the NBN was removed as an election issue, regardless of what the real Coalition plan is. There is probably nobody else that could have done this, which for me, raises deep questions about Turnbull's integrity and character. He was very happy to destroy the current NBN for a political end, can anything he says now be trusted?

This tweet by RickW (on the Tooth Fairy theme) reminded me of one of the "drop dead" project risks of a (Telstra) Copper Node Plan: First, stocktake the assets and audit the records/database to see if the plan is still even possible.

I've never seen an estimate for this additional cost of the Turnbull Node Plan, either in time or money.

It's somehow magically assumed by the all-encompassing $900/port and "will take no time at all". Just ask Turnbull, as if he's ever going to provide any actual numbers (put your fireproof suit on, if you do ask).

Turnbull would like us to think the opposite of what's true, the "counterfactual" spell:

  • he's experienced in dealing with Telstra, not just as a customer, but as a regulator or peer acquiring their assets when he doesn't, and
    • [Telstra play "hard-ball", ask Senator Conroy]
  • he has Big Project, even Telecommunications, experience, when the most he's done is push paper and sign cheques for an antiquated email provider, small by current standards. 
    • [Ask BHP about the surprises in Big Projects. They are hard and sometimes even the best and brightest have to walk away. Size Matters, Experience is necessary.]

We already know there are two large significant and uncosted parts of the Turnbull Node Plan: remediation and reconfiguring the Telstra copper Customer Access Network (CAN), the last 800m that he desperately needs to work.

In 2005 and 2008 Telstra went on record with real plans for a Fibre to the Node network where their first priority was remediating the Telstra Copper network, including removing all pair-gain systems (8500 RIM/CMUX's out there) and "bridge-taps". In 2007 we learned in a Telstra response to the G9/FANOC proposal, they needed to reconfigure the (Telstra) copper to make nodes, not exchanges, the centre of each "distribution area", reducing the number of nodes by 30%-40%.

In 2005, 12Mbps at 1500m to Metro-only customers was an appropriate solution. It seems Turnbull and the Liberal Party didn't get the memo that it is now 2013 and both demand and technology have moved on many-fold. The "window of opportunity", where an FTTN could be economic and technically feasible, has long shut.

Before anyone can move building an FTTN, an audit of the whole line records database is required (20-50 million records) to reach acceptable, not perfect, "data quality", a "ground-truth" survey to get the database in sync with Reality and an assessment of the condition of every single metre of the 100-200 million pair-kilometres of the Telstra copper network.

To "fix the worst first", you first need to identify the worst and then define a test to "fix, replace or go Fibre?". There are at least 200-250,000 cable-km in those 800m passing 75% of fixed-lines Turnbull wants to thrown down the FTTN well? Compare this to the 208,000 cable-km in the whole FTTP rollout of NBN Co.

To survey the last 800m in one year, Turnbull needs to examine and test 1,000 cable-km every day. How big a team does he need? Thousands? How good will the results be? Variable. (If you rush, the results will be incomplete and/or inaccurate. If your intent is complete, correct records then rushing makes that impossible.).  Where will he find that number of spare, skilled and competent contractors? Not in Australia.

What's that 100,000 "man"-years going to cost Turnbull? $10 billion a time, because it has to be repeated. Remembering that he's not included either the cost or delay in his (public) plans...

The problem for Turnbull is, that unlike Fibre, the Telstra Copper Access Network is in a constant state of flux. As soon as an audit and "reality-check" are done, they are out of date. Guaranteed.

Turnbull has to pay twice for a complete database audit, ground-truth survey and line test:

  • Once before planning can get underway, and
  • immediately prior to cutting the Telstra copper in every Distribution Area.
    • The second audit and survey/test is hidden in the cost of works.
    • Starting from better quality data, the cost is 25%-50% of the first audit.

There's also the small matter of doing all this again in 10 or 20 years when the Telstra Copper is finally retired and replaced with Fibre. That's written in the Coalition NBN Plan, they choose to not emphasise  it. That additional $10 billion audit and survey of all premises has to be added, along with the deliberate wastage or "CapEx Reuse", to the budget of the Coalition NBN Plan. What's another $10 billion for the taxpayers to cop?

I know that it took 6 months for a major bank to have it's 120,000 network devices audited/checked when new network contractors were hired. Their records were 60%-70% accurate and these had been used for billing. The number of sites was a tiny fraction of the 9 million premises Turnbull wants to cover.

One of the first delays for NBN Co was "simply" creating an accurate premises database. The street address of every service in Australia: this wasn't available before NBN Co created it. Wouldn't you expect the Telstra records to already have that? This is proof positive that as recently as 2011, the line record database was inaccurate and inadequate for any major project, especially for an FTTN.

But there's more...

As RickW points out, what Telstra is giving NBN Co to run Fibre for the current $11 billion is prepared ducts.

This is necessary for the Turnbull Node Plan rollout as well.

After the audit, survey and test, before the remediation and reconfiguration of the Telstra Copper Access Network, the pits, pipes and ducts must first be "prepared", but that's code for repaired and remediated. The Copper Access Network is two separate parts: the cables and the access-ways. If you can't run the new Fibre to the Node or fix and reconfigure the (Telstra) Copper distribution pairs, then you can't build your FTTN.

Turnbull might be happy he's duped the public with his uncosted "NBN Lite", but he's also completely ignorant about the real size, costs and schedule of any project touching connections to every premise in Australia.

Size Matters. "Getting to scale" is difficult and requires great skill, ability and experience.

This is why Google, Amazon and Apple have survived while hundreds of "wannbe" competitors have fallen by the way in the last 15 years: Execution at 'scale' is hard. Something Telstra knows and has warned of [slide 12&18] many times.

Turnbull and his advisers simply don't have the experience or ability to complete or control an FTTN project, let alone get it done on-time and on-budget. But worse, they don't even know they don't know.

It's the taxpayer that is going to pick-up the tab for Turnbull's inevitable failure and we will have to wear the decade or two wait for better Broadband. Australian Business Productivity, and hence Competitiveness, will suffer and we will get even further behind the rest of the world. The all-Fibre NBN is simply a catch-up, not an overtake.

If Turnbull's magical "Cost Benefit Analysis" doesn't uncover these drop-dead project risks and predictable schedule challenges, then we'll know it was just a predictable sham and smoke-screen.

The Turnbull Node Plan is based on the Copper Fairy coming in the night and magically transforming decades of neglect and decay into a bright, shining new network. Telstra knows this and will lead Turnbull a long and merry chase, extracting tens of billions in taxpayer dollars in the process.

Remember back to late 2001 when the World Trade Centre came down and Ansett Airlines were stopped from flying. QANTAS kept "negotiating" with them to wet-lease their fleet (planes, fuel, crew, facilities) until the clock was run out and Ansett lease payments were due, forcing the company into liquidation. Who saw that coming? [A: Everybody but the Ansett management, seemingly.]

Just as Ansett were blind-sided by QANTAS, Telstra will 'take' the Coalition government for everything they can, forcing Turnbull to be sacked and rightly pilloried for his failure. The recriminations and fireworks within the Liberal Party will be quite something to watch. Not useful or helpful to Australian voter and taxpayer, but at least some sort of recompense for destroying our future.

The two guaranteed outcomes of the Turnbull Node Plan: Telstra will do well out of it! And the taxpayers will pay billions of ongoing dollars more than the full fibre scheme would have cost.


  1. This new $10 billion for a copper audit must be added to whatever extra Telstra demands on top of the $11 billion it agreed with NBNCo.

    The reason Telstra must demand more than $11 billion is because retail revenues will be lower on FTTN than they would have been on FTTP. 99.25% of Telstra shareholders signed off on the $11 billion proposal with the guarantee of permanently large retail revenues from high speed fibre that simply cannot be equalled by copper.

    Telstra is required by law to maximise shareholder benefit from the asset it paid $26 billion to buy. I'm expecting a price tag in the ballpark of $30 billion, possibly negotiable down to $20 billion.

    FTTN will be very, very expensive on Telstra's copper.

    1. Francis,

      glad we agree. FTTN will be more expensive than people can guess.

      Thanks very much for your input. Always great.



Note: only a member of this blog may post a comment.