Broadband News

New estimate for the cost of rolling out G.fast to 10 million premises

G.fast which is seen as an evil technology by some and a god send by accountants who don't want to commit to the costs of a widespread FTTP roll-out is set to be rolled out to some 10 million UK premises by Openreach by the end of 2020. That is the plan as things stand today, but given the political and economic climate this could all change, e.g. full Openreach split or a recession that stalls any industry investment.

7490-gfast-map-thumb.png
Click image for full size version

We produced a model that showed where we thought Openreach might deploy G.fast pods back in June with two scenarios suggesting between 28,000 and 38,000 cabinets would gain a pod or two.

The interesting part today is that Dave Burstein of DSLPrime has taken our figures and based on some estimates of the cost of a G.fast DSLAM has arrived at a cost of around $300m a year for the four years of the roll-out.

96 ports of DSLAM at $250 is $24,000. Multiplied by 40,000 cabinets you come to $960,000,000. Add 25% for overhead and cost overruns raises that to ~1,200,000,000, about 1% of BT's sales in the four years. A telco looking for a higher figure than $250 for a press release would come much higher. A vendor looking to bid low enough to guarantee the contract would probably come in much lower. Both could find senior economists who will confirm their figures for a fee.

The US dollar to Sterling conversion rate is vary variable at present but at the conversion rate while writing this item $300m is the equivalent of around £229m.

The Openreach investment plans also include 2 million premises of FTTP, which if you work to a figure of £500 per premise passed cost works out at a cost of £250m each year for four years. With FTTP the costs are going to be much more variable depending on the exact nature of the roll-out needed for each property, but the ball park figures illustrate the difference involved in adding another stepping stone before the inevitable full rollout of FTTP.

Comments

Adding just 25% for all the costs of deployment and other contingencies strikes me as rather optimistic, even if the amount of planning and installation work is vastly lower than is is for the original cabinet.

That said, £200-250m a year is clearly fundable, but just how much incremental revenue it would generate is debatable. How many would bother upgrading?

  • TheEulerID
  • about 1 year ago

I should add much of the pressure for FTTP is from MPs with significant rural populations. None of the proposals for the ownership of OR make the slightest difference to the fundamental commercial and economic issues in those areas. Indeed, full separation would expose the issues more clearly with none of those misleading claims about BT Sport competing with BB investment (if anything the former stimulates demand for the latter).

  • TheEulerID
  • about 1 year ago

£300m per year for G.Fast plus £250m per year for FTTP... Definitely an increase in investment compared to FTTC, and within the bounds of BT's announced spend of £6bn over 3 years (for Openreach and EE).

Demand? Maybe 10-15% of the premises covered? 1m-1.5m premises, so maybe 30-50 ports per pod.

  • WWWombat
  • about 1 year ago

The G.fast on its own comes in lower than the current FTTC project which shouldn't be surprising. That project required civils to build new cabinets and get fibre and power to them while this project can very much piggy back on that.

  • CarlThomas
  • about 1 year ago

As another thought G.fast from the cabinet is not another stepping stone towards FTTP. It is purely there to avoid placing fibre deeper and taking steps towards FTTP.

G.fast from cabinet will both deliver somewhat higher, though far from awesome, speeds to those closer to the cabinet and may allow for backhaul of deeper nodes via copper.

The original FTTC deployment was a stepping stone, this isn't. Contributes nothing to future-proofing the network, purely about achieving the premises passed at >100Mb as cheaply as possible.

  • CarlThomas
  • about 1 year ago

I'm not convinced about the costs for FTTP either. These will be considerably lower for BT than anyone else as they have pre-existing ducting and poles. It would be very strange for it to cost them as much as the Cityfibre / TalkTalk / Sky project given that uses very little pre-existing infrastructure.

A fair chunk of the FTTP investment will also be on new build. Cheaper than brownfield and way cheaper than brownfield and building new ducts in existing highways.

  • CarlThomas
  • about 1 year ago

Hi Broadband Watchers.
I feel the G/fast will be targeted at the TBB Red Post Code Sections and will be demand driven and only used where it is cost effective and the coverage will be the whole CAB addresses that are over the EU 30 MEG target. The post Codes will be opened as required determined on staff levels and fibre crews.

  • Blackmamba
  • about 1 year ago

So there in lies the crunch, what is the cost of doing FTTP now that FTTC has brought aggregation nodes and thus the GPON fibres closer?

If lower or close to £191 then suggests zero benefit financially to doing G.fast, though speed of roll-out may still make G.fast preferable depending on what the external pressures are.

  • andrew
  • thinkbroadband staff
  • about 1 year ago

Yes, surely the overall cost of FTTP will now be much reduced when you take into account A: All the aggregations nodes as Andrew says and B: 10 million G-Fast lines reducing the overall number of lines by a third or so.

  • ahockings
  • about 1 year ago

Hi Broadband Watchers.
Could it be that the coverage of 10 million customers offered by the CEO of Openreach be a Red Herring and not stating that he is going further into the network,thus keeping his options open. I would think there is not a large difference in price on the high % of jobs ( post codes locations). These jobs may use off the shelf method and will be completed by a dedicated group.

  • Blackmamba
  • about 1 year ago

The right 2 million premises can come in significantly below the £500 per premises passed.

Costs in a number of areas will, of course, be way higher than this.

There are no indications OR are going to be using this technology to plug gaps in existing SFBB coverage for right now, and FTTdp is significantly more costly than this cabinet method as it needs entirely new backhaul and power.

Building to cabinet spreads cost between more premises than a new remote node, so way cheaper per premises passed.

  • CarlThomas
  • about 1 year ago

Where did you the $250 per port for a DSLAM port? It was less than $20 per port in 2013. I will get you the research reference. I concur with Carl, G.Fast can re-use spines, ag nodes and tij chambers.
On FTTP the emerging rules of thumb appear to be £350-£500 for urban-suburban per premise connected.
£800 for villages/hamlets per premise connected
£1000 for rural clusters per premise connected
We should consider FTTP as a transition from PST and bring costs lower, while gleaning efficiencies on top.

  • ValueforMoney
  • about 1 year ago

@valueformoney Follow the link in the story, pretty clear where the figure comes from.

Remember a G.fast DSL port is new silicon and much more advanced than a VDSL2 or ADSL2+ one.

  • andrew
  • thinkbroadband staff
  • about 1 year ago

@CarlThomas
"As another thought G.fast from the cabinet is not another stepping stone towards FTTP. It is purely there to avoid placing fibre deeper and taking steps towards FTTP."

This is a great point, G.FAST pods on cabinets do nothing/little to advance fibre deeper into network for those on longer lines.

  • craski
  • about 1 year ago

@CarlThomas
"G.fast from cabinet will both deliver somewhat higher, though far from awesome, speeds to those closer to the cabinet and may allow for backhaul of deeper nodes via copper."

Also not something I had thought of but another interesting point. I wonder if we'll see a new type of remote node developed to use G.FAST over copper back to the cabinet instead of pushing fibre deeper.

  • craski
  • about 1 year ago

Major cost is the vectoring engine. It's state of the art, 96 port G.fast vectoring engines aren't even available yet.

The line cards themselves are obviously pretty expensive too, but the power needed on the vectoring engine board is prodigious, with the price tag to match.

  • CarlThomas
  • about 1 year ago

@VFM the rule of thumb for FTTP passed premises is given by KCOM, which is averaging out at ~£400 per premises passed not connected.

  • themanstan
  • about 1 year ago

Important to remember that the g.fast rollout is supposed to continue until 2025.

So while the initial 2020 g.fast rollout might not get much fibre deeper into the network, it will help prove demand for faster speeds making further g.fast and FTTP roll outs easier to justify which will push fibre deeper.

  • gt94sss2
  • about 1 year ago

Given the costs KCOM are incurring and the costs mentioned here the cabinet-based G.fast project is looking like an extremely expensive PR offensive to keep politicians and regulator off BT's back.

If the G.fast kit is going to cost half per premises passed what FTTP would in areas without adding any upgrade options beyond, maybe, backhauling deeper nodes via a copper bundle the value is questionable.

Do we actually urgently need the relatively modest speed boost? As far as the speed competition goes who really cares about 100-300Mb when Virgin will be at a gigabit this decade?

  • CarlThomas
  • about 1 year ago

@craski
Yes, copper backhauling is being considered. From, say, a G.Fast node out to a deeper node, reducing the need for fibre to be sent out there. If the same copper can be used to supply power, it might have more traction.

On the face of it, though, the pods don't seem likely to be able to power much external hardware.

  • WWWombat
  • about 1 year ago

The simple fact is that it is far easier, quicker and cheaper for BT to adapt their existing network to G.fast.
Widespread deployment of FTTP will have to come from new players in the market but very little it being done to encourage them to invest.

  • chilting
  • about 1 year ago

But if OR was to start and/or encourage FTTPod, even subsidise it a little, then it would reduce their needed investment on FTTP. Depending on the cost, even getting 10,000 premises nationally to pay for their own FTTP would be a big help, maybe OR could do other FTTP work whilst doing a FTTPod installation, saving even more.

  • lincsat
  • about 1 year ago

10,000 premises paying for their own FTTP is nothing, in fact would represent just 3% of the existing FTTP foot print so not going to kick start a revolution.

Consider this at 2m premises a year looking at 15 years for 100% roll-out

  • andrew
  • thinkbroadband staff
  • about 1 year ago

TheManstan - Is FTTP passed, FTTdp with a customer paying for a drop wire off a manifold, or fibre from point on the pavement?

  • ValueforMoney
  • about 1 year ago

Andrew - so new silicon and vectoring engine. Anything in comparing the elusive operational costs on these complex objects v FTTP?

  • ValueforMoney
  • about 1 year ago

@vfm
A reminder that FTTdp is still not a sub-variant of FTTP where someone pays for a portion.

FTTdp is a sub-variant of FTTC, where the MSAN/DSLAM is placed close to the customer.

  • WWWombat
  • about 1 year ago

@valueformoney You are the person who surely has all that at your finger tips, given the analysis you run and get quoted at various places. Or are you saying your figures are just as much guesswork as anyone elses?

  • andrew
  • thinkbroadband staff
  • about 1 year ago

@VFM

KCOM's own newspieces show an engineer splicing on the front garden path... so a pavement point.

http://www.kcomhome.com/news/articles/more-hull-and-east-yorkshire-locations-to-go-lightning-fast-in-2014/

  • themanstan
  • about 1 year ago

Post a comment

Login Register