» ISP News » 

First BT Trial of Future G.Fast Copper Broadband Hits Nearly 720Mbps

Tuesday, September 2nd, 2014 (3:06 pm) - Score 9,196

The next generation of hybrid-fibre G.fast (aka – FTTC2 / ITU G.9700) and FTTdp broadband technology managed to achieve an aggregate downstream / upstream speed of nearly 720Mbps (Megabits per second) during BTOpenreach’s first UK trial with Huawei, which took place at several homes near the operators Adastral Park engineering HQ in Ipswich (England).

The highly experimental trial, which was announced last year (here) and had been designed to prove the physics of G.fast transmission on an actual copper plant, was the first one to be conducted by BT outside of a lab environment and involved three short copper lines (homes).

Thankfully ISPreview.co.uk has been able to get our hands on an official report from BT’s on-going trial, which reveals quite a lot about its potential capabilities. But the hardware involved is still based off early prototypes, which are most certainly not standards complaint. The first proper kit from Broadcom, Huawei’s HiSilicon, Lantiq and Sckipio isn’t due to surface until the very end of 2014 or early 2015.

What is G.fast?

At present BT’s mass-market Fibre-to-the-Cabinet (FTTC) solution can deliver speeds of up to 80Mbps using 17MHz (radio frequency spectrum), which works by replacing the existing copper cable between street cabinets and your local telephone exchange with a fibre optic line. The final run (between cabinets and homes/offices) is then managed by VDSL2 technology over your existing copper line, which works like ADSL but is faster over short distances (VDSL2 works best at distances of sub-400 metres but it can operate at up to around 2000m with significantly slower speeds).

By comparison G.fast works in a similar way to VDSL2 but it’s been designed for considerably shorter runs of copper cable (ideally below 250m but there may be benefits up to around 400m) and uses 17 – 106MHz (note: the specification allows for potentially even higher frequencies in the future), which means that to work at its best it needs to be paired with a Fibre to the Distribution Point (FTTdp) / Fibre to the Remote Node (FTTRN) style solution (i.e. shortening the run of remaining copper lines by replacing more of it with fibre optic).

In simple terms FTTdp/rN takes the street cabinet out of the equation and instead brings the fibre optic cable all the way to a smaller and low powered ‘remote node’, which could be positioned underground or on a telegraph pole nearer to your home. Note: this isn’t to say that G.fast couldn’t be run from a street cabinet too, but BT also needed to test FTTdp/rN and it makes sense to combine them (note: BT is currently about to trial FTTrN in several larger areas using existing FTTC technology).

fttrn network diagram v1 ispreview edited

Openreach’s G.fast Trial Results

The operators G.fast trial was fed by the Portman Road telephone exchange (Ipswich) via an Integrated Layer 2 Switch with XG-PON1 link to trial DPU and GE connectivity to Adastral. The trial used existing duct structures and distribution point (DP) locations, but created a separate overlay to prevent harm to BT’s existing plant.

Dedicated copper lines, of lengths 19 metres (house 1), 47 metres (house 2) and 66 metres (house 3), were fed into the customer premises where they terminated on the G.fast NTE + end-user hardware (router etc.). Several different frequency plans were tested, both with and without vectoring enabled (this is a technology that’s designed to remove crosstalk interference from the lines).

The performance returned was extremely impressive, as can be seen below, albeit not yet representative of a larger real-world environment using final hardware.


It’s worth noting that BT sees upstream speeds as a way to differentiate themselves from Virgin Media’s cable platform, which often dominate in terms of download performance but not uploads (here). G.fast appears to be a good fit for that approach because it can deliver significant benefits to both the upstream and downstream side, provided you live close enough (note: most premises reside within around 400m of their nearest street cabinet / dp).

The report also took a closer look at the 3rd home on the longest 66 metre line to gauge how performance changed between the different frequency plans and vectoring (note: we anticipate that g.fast will be deployed with vectoring enabled by default).


Overall BT considers that the first, if admittedly extremely small, field trial was broadly successful and has “significant potential“. Indeed some sources hint that they may even jump the plan to roll-out Vectoring on existing FTTC lines and skip right to G.fast, although we have yet to see any firm proof of this.

It should be separately noted that the Phase 2 FTTC Vectoring trials are already underway. In any case G.fast is still some years from a full commercial deployment (maybe starting in the 2016-18 window for the UK), assuming it happens at all.

Elsewhere we note that the results from BT’s test appear to be broadly in agreement with one of Alcatel-Lucent’s recent trials, which similarly demonstrated that 650Mbps was possible at around 50m with vectoring enabled (AL’s test was probably setup in a not dissimilarly way to BT).

gfast and vectoring 2.0 fttc vdsl2 speeds

At this point it’s wise to stress that trials like this do not yet reflect a real-world environment and the report agrees, “BT continues to believe that G.Fast has significant potential, though it has yet to be proven in operational deployment in a final standards compliant form. We are currently focused on driving the ITU, BBF and ETSI standards to completion.”

A BTOpenreach Spokesman told ISPreview.co.uk:

We are currently trialling G.Fast prototype kit in our research labs at Adastral Park. We’re watching this technology with interest given it’s capable of theoretical speeds of up to 1Gbps. We intend to trial the fully standardised kit once it’s available from the vendors.

Given it’s an emerging technology we’ve made no decision around whether or how it might be deployed in our network in the future.”

Just one of several potential problems, aside of the hefty upgrade cost (depending upon how they choose to do it), is that the performance of a G.fast line could suffer if it needs to coexist with VDSL2, which currently requires 17MHz of the spectrum. A recent example from Swisscom, which is highlighted over at Fast Net News, notes how vectored FTTC/VDSL2 can deliver a bit over 115Mbps (25Mbps upload) at up to 100 and 200 metres from the street cabinet, which rises to 700Mbps and 380Mbps (aggregate DL+UL speed) respectively for G.fast with access to the full spectrum.

But if G.fast has to coexist with VDSL2 then the speed falls to 580Mbps at 100m and 280Mbps at 200m (aggregate). Stretch to 400 metres and FTTC delivers a little less at 95Mbps (22Mbps upload), while G.fast on its own could do 140Mbps, but add in VDSL and suddenly G.fast can only deliver 40Mbps (i.e. worse than the current FTTC/VDSL setup). This largely reinforces why many studies describe G.fast as being of primary benefit to sub-250 metre copper lines. In other words, it could be very fast for some but many people wouldn’t see a huge benefit unless FTTdp is adopted more widely.

Ultimately the reality will only become clear once BT can test with standards compliant hardware and that seems unlikely to happen until sometime in H1 next year.

Share with Twitter
Share with Linkedin
Share with Facebook
Share with Reddit
Share with Pinterest
By Mark Jackson
Mark is a professional technology writer, IT consultant and computer engineer from Dorset (England), he also founded ISPreview in 1999 and enjoys analysing the latest telecoms and broadband developments. Find me on Twitter, , Facebook and Linkedin.
Leave a Comment
89 Responses
  1. Raindrops says:

    So every 20M or so of copper loses about 100Mb……. Good luck BT LOL

    1. No Clue says:

      It would appear around 100Mb loss or thereabouts looking at both the charts.

      I see the comment has upset some a pity stating what is obvious riles some people so much.

      Oh and before anyone that is thought challenged asks yes i did enjoy my neigh on month long holiday out in Nassau. No need to welcome me back 🙂

    2. Raindrops says:

      Welcome back, hope you had a nice holiday.

    3. No Clue says:

      Thanks yep was a nice break from things.

  2. four_eyes says:

    Another failure of BT openreach clowns the speeds don’t add up end of… take your vdsl rubbish and stick it where the sun don’t shine

    1. Curious says:

      “take your vdsl rubbish and stick it where the sun don’t shine”

      What do you want instead?

  3. X66yh says:

    Is it not time to remove any ability to comment on news items if instead of a debate about the pros and cons all we seem to get is posters doling out abuse.

    1. Raindrops says:

      I fail to see who i have abused with my comment. If you mean i have abused BT then i suggest you and those with a vested interest in BT take your own advice and never abuse B4RN, Virgin or other suppliers.

    2. Abuse Alert says:

      Is there some reason why some posters cannot be polite in their replies, do they really think they are being clever?

      ‘stick it where the sun don’t shine’

      ‘BT employees do not think the just randomly speak.’

      ‘NO frigging dumb ass line rental either lol’

      Not intelligent comments on what should be a professional forum.

    3. Raindrops says:

      Again i fail to see how any of that is being abusive towards any person on this site.

  4. MikeW says:

    Ta Mark. More interesting stuff…

    On the problem where VDSL2 and G.fast must be deployed in a mixed environment, I saw recently that Adtran had come up with a Frequency-Division-Vectoring concept that overcomes the issue.

    What it does exactly, and how effective it is, is hard to judge. It seems to be PR-ware so far.

    It is also proprietary, and has no backing from a standard (as yet), so might go nowhere.

  5. FibreFred says:

    Sounds promising!

    On another note, can we ban the troll again?

  6. FibreFred says:

    I’m about 20m from the pole so it looks good to me, plus this is just the start, its just experimental kit/software I’m sure they can push the speeds higher.

    1. MikeW says:

      It seems the average is around 25m of drop wire from the DP, min 5m and max 80m. It probably looks good to a lot of people.

      As ever, the tech looks good, but we’ll need to see where the finances point.

    2. Raindrops says:

      LMAO you said 200M here
      Make up your mind or did the previous follow up BT link i provided help

    3. MikeW says:

      Short memory Raindrops. You’re too quick to rant, and too slow on the reading. Or the uptake.

      The 200M was a guess for FTTRN, and pointedly included some of the D-side. This page is about FTTdp. Two different technologies; two different deployments.

      Do keep up

    4. MikeW says:

      But thanks for showing everyone that you don’t actually bother to read their responses before turning on the vitriol.

    5. Raindrops says:

      “Short memory Raindrops. You’re too quick to rant, and too slow on the reading. Or the uptake.”…… “This page is about FTTdp.”

      Now, now if you do not like personal abuse do not start it oh and on the contrary i read the news item.
      “Note: this isn’t to say that G.fast couldn’t be run from a street cabinet too, but BT also needed to test FTTdp/rN and it makes sense to combine them”

      “But thanks for showing everyone that you don’t actually bother to read their responses before turning on the vitriol.”

      Thanks for the personal attack which all just because you hate anything negative towards BT Awwwwww poor BT.

      Also take your own advice if you do not like what i have to say…….

      Ill also repeat the reply of….
      “Im sorry you find anything i have to say in response not to your satisfaction, i would like to help by suggesting perhaps the easy way to avid me upsetting you is to not respond to anything i have to say. We will both be happy then.”

      Perhaps you should not reply.

    6. MikeW says:

      Seriously? All you can think about is how much you hate the people who appear, in your eyes, to say anything even vaguely positive about poor BT?

      This is a (pretty hot exclusive) report on real-life G.fast – one of the first we (as Joe Public) have had the privilege to read, in the entire world, without attending telco-specialist conferences. Mark should feel proud about ferreting this out.

      The story here is about the technology, not BT. Yes, the trial was run by BT researchers, using UK-standard copper lines. Those aspects make us adapt the results to our own situation better than, say, if the research were done in Berlin or Shanghai. Other than that, though, it has very little to do with BT, more to do with Huawei as vendor & whoever the chipset maker was, and a lot to do with international co-operation & standardisation.

      Will this ever get deployed here? Who knows – the numbers will dictate the answer to that, in maybe 4, 6, 8 years time – and only then will we know how far from properties this kind of thing will get deployed, if at all.

    7. Raindrops says:

      1) I do not hate any one.

      2) The story is about a BT trial, look at the title of the story.

      3) as for ” it has very little to do with BT, more to do with Huawei as vendor & whoever the chipset maker was” using that logic we may as well say NO BT story is about BT but who manufactured the copper they use. NO story about Virgin is about Virgin but who made their coax, and NO story about Sky, Talk Talk is about them but who made the LLU equipment. If you want to deal with stories about companies and their tech in that manner.

      4) And finally, your problem is my opening remark, and much as you take issue with that its true. approx 100Mb loss per 20M of copper cable. Its a fact BTs own diagram shows. If you do not like it, that is your problem not mine. Its not abuse to any individual its a simply observation from the story. Whether you like me adding “Good luck BT” or not is also irrelevant. It was not rude to any individual no matter how much you take it that way.

    8. MikeW says:

      There’s little point in arguing with you about what the story is about, but I will ask you where Sky, TalkTalk and Virgin have their research departments? Where do they get involved in developing the technology and the standards that become the core of future deployments around the world? This trial is from a very different part of BT than retail, wholesale or openreach.

      You are also wrong on this: “my problem” is not about your “opening remark” (100 instead of 70 is close enough). It is everything about the way you treat other people, other comments, on here. We rarely get a thoughtful, considered comment from you: it is almost always a rude, arrogant, aggressive put down.

      One aspect I complained about earlier was that you didn’t read other comments (and to prove the point, you misread that, and asserted that, on the contrary, you did read the article) remains true. It is barely worth the effort of typing a response to you when I know what the quality of your response is going to be like.

    9. Raindrops says:

      The story is about BT the title of the news item makes that very clear. “BT TRIAL”

      My opening remark says ABOUT 100Mb loss. Which it is. Any loss in this trial is also likely to be worse in a real world situation. When it is not separate from other BT plant.

      Ill also repeat if you have any issue with any reply i make or its quality, do not engage me in conversation. It is a lot more simple then reporting every post.

    10. MikeW says:

      Oooh. A bit touchy after your original reply got deleted? Think I reported it, just because it went?

      That’s about as accurate as accusing me of being so many other personas on here, and that we’re all BT employees. Quite funny, but it is tedious to have to regularly wade through the bile on article after article.

    11. No Clue says:

      The only person that appears touchy is the person that has time to report posts and continually reply to a person they claim they would not reply to.

      That is far more funny than anything else.

    12. Raindrops says:

      And it still makes no difference to the original remark of “So every 20M or so of copper loses about 100Mb” LOL

  7. Comic Assassin says:

    I can see this working well with FTTrN but thats the only way this is working for a vast amount of properties like mine 600meters from my FTTC. This allows BT to keep their Tax………..i mean line rental/up keep……well what ever they call the Tax!

    1. Steve Jones says:

      That “tax” you describe will have to be paid whether the line is copper or fibre. It has to cover such things as cost of the capital invested, maintenance, ducts, rates, buildings, return to those investing money and so on. The wholesale cost is £6.45 per month. Anything more than that is either VAT or added by the SP at retail level. Even if the entire copper network was swapped out by fibre. Even if that swap-out was by another company altogether, those costs will have to be paid. Indeed, if the £25bn required for national roll-out was to be spent, then then costs will be much higher than for the copper network simply because of the cost of capital and providing a return to those who provide the capital (banks, shareholders and bond holders don’t provide money out of an act of charity).

      So, if you want broadband without paying for a fixed line (whether copper or fibre),then I suggest you subscribe to a wireless operator. Of course, they have their own costs to recover, but good luck anyway.

  8. Comic Assassin says:

    To expand on my comment it does depend on how many FTTrN BT are going to install in an area for FTTC2 to be of benefit, other wise this is another FTTC situation with headline speeds that only some customers will see.

  9. NilSatisOptimum says:

    Regarding the first two comments, I’m all for free speech etc, however could you not have some editorial control, not banning or deleting, but perhaps if way off topic or generally and anally abusive have it hidden with a option to read if I wish.

    1. Raindrops says:

      It was on topic look at BTs chart it loses approx 100Mb every 20M of copper.

  10. Steve Jones says:

    It’s just a cost/service/reliability issue. By the time fibre gets as far as the local telephone pole, it may be just as cost effective to swap out the drop wire with fibre. It may be a bit different with estates like mine which use direct buried wire from footboxes as replacing that would involve a lot more work.

    However, as usual, the real issue is whether there’s a viable self-sustaining market at this level. It’s quite possible that many people will be happy enough with a service that supports two or three HD streams if the alternatives costs a lot more. For that, you don’t need gigabit speeds.

    For the record, I’m not convinced g.fast has a great commercial future, simply because of the economics. However, I’m always rather impressed how far copper can be pushed. Not only is there the elegance of transmission line theory but what is one of the great wonders of modern technology – digital signal processing. (Not that DSP doesn’t apply to fibre as well).

  11. zemadeiran says:

    Hi Guy’s,

    In regards to line rental, who and where does the money go? Should it not go directly into Openreach’s account? After all they are responsible for maintaining the last mile surely?

    If someone is in the know about line rental then please do let us all know.
    According to this wikipedia article: http://en.wikipedia.org/wiki/Telecommunications_in_the_United_Kingdom there were 35 million lines in 2002 which without doubt have dropped due to mobile usage etc.

    If we say that there are now 25M lines at £10 line rental per month,

    £250M pm x 12 = £3Billion

    Now the question really is, how much does it cost per year to maintain 25M copper lines?

    Regards to all

    1. Steve Jones says:

      I’m not sure where the figure of £10 per month comes from. Openreach actually get paid £86.11 per year (£7.17 a month) for what is called “MPF”, which is the basic copper loop facility used by LLU operators. Of course that excludes VAT, which would make the total £8.60 per month, but as that goes to the government it’s not Openreach revenue. That would yield about £2.15bn per year if there are 25m copper pairs. N


      (There are some other wholesale products, like WLR which are used by companies which use wholesale voice services too, but that has only a slightly extra cost and relates to

      As for what the costs cover, I would suggest looking at the Ofcom reports (which sets out the principles on which they set the prices). However, bear in mind that the cost is not purely maintenance. It also includes rates (all those buildings, lines, ducts, poles and soon attract that), rent (most of OR’s buildings are leased), renewal of infrastructure (different lifetimes, depending on the type of items), financing costs and, shock horror, a rate of return based on a (cost based) estimation of the value of the network. That’s to say the owners (shareholders – which includes me by the way – are entitled to a relatively low, but at least secure, return on the assets which they bought off the government).

      Needless to say, this stuff is subject to a lot of lobbying by both sides. In general, the Ofcom mandated prices are lower than equivalent charges in comparable countries in the EU.

    2. Raindrops says:

      He is actually closer with his £10 per month estimate…

      The latest (newest price) and the cheapest it has been for a while being £91.05 ex.vat and £109.26 inc.vat, which equates to £9.10 and half pence per month.
      Which when using zemadeiran calculation works out to £2.7315 billion.

      Your link is what LLU providers pay Openreach, that link is what BT Retail and any NON-LLU pay.

      That is the price for most of us that have a BT line and not a LLU line, as also detailed and pointed to by the BT wholesale website pay BT for a BT line…
      Specifically part 3 of Wholesale Service Provider Services.

      SO his off the cuff £3 billion guesstimate was pretty close to being spot on.

      That cost does not cover things line ducts as you mention either, that is quite clearly posted as other charges on the openreach site. I sincerely doubt it covers poles either or could be argued it does in any way considering those are not regularly replaced. In fact if that was the case people without a pole should be paying less using your logic or at the least a different rate.

      Even if you allowed for all that i doubt it would cost the £2 million plus per month they are making based on the calculations of 25 Million lines.

      Sorry but no he has a fair point and you did not brush that under the rug very well.

    3. Steve Jones says:


      The original question was about the maintenance costs of copper lines. That surely implies the basic loop. That is the MPF product and is effectively the common denominator in that it’s used by LLU operators (a very large proportion of the market) and is effectively included in the WLR product. I mentioned WLR, but as that provides the voice service too (and BT are, of course, the main customer), it’s not particularly relevant to a question which said “how much does it cost per year to maintain 25M copper lines?”. There will, of course, be extra costs associated with the voice equipment at exchanges for WLR.

      As for including VAT, that’s just plain daft. It forms no part of the income to Openreach. It’s simply the government using industry as tax collectors. It’s particularly inappropriate in the light of a question effectively asking for how much it actually costs to maintain 25m lines.

      Rather than this tiresome game, I would suggest those interested in such things look at the Ofcom documents and consultative documents which sets out in considerable detail the basis calculating MPF and WLR costs. The following is one which is relevant.


    4. Steve Jones says:

      Just to note, that company turnover is normally held to exclude VAT, as per the following link. Not that it’s particularly relevant. Either total sales value net of VAT is used, or it’s the total turnover including VAT and the VAT is simply considered an operating cost. It amounts to the same thing, but the former is surely the sensible way to view things as it means that numbers aren’t changed when governments change the rate of VAT.


    5. No Clue says:

      NO his question was….
      “In regards to line rental, who and where does the money go?”

      And you have not explained it, in fact no body can explain it because it is a mystery. Well not a mystery really because for a start several million goes in bosses pockets as “bonuses”.

      As for VAT thats entirely relevant given his question and the fact we all pay VAT on our line rental. You can not exclud it from one set of figures but include it for another.

      There is no explanation on where around 2.5 billion line rental goes in a year. NO WAY in hell is that to maintain lines.

      If it took that kind of money to maintain lines then where is the new FTTC network getting money for its maintenance????

      That must be either A) running at massive loss or B) having basically zero maintenance. That now covers around 90% of the countrys premises…. How are they “maintaining” that with such small income from the product The take up for FTTC is only around 20%.. NO way near enough to pay to maintain things. Even if it cost HALF as much to maintain as the copper side of the network and that copper side needs 2+ billion it still would not be making enough cash to maintain itself.

      Line rental is a rip off, every one knows it, just BT and its share holders like a dodgy bank will not admit it.

    6. Raindrops says:

      I still do not understand why he wants to quote the price LLUs pay Openreach for a line, apart from it trying to suit his agenda cos its cheaper by a quid or so.

      Even arguing its because its a large part of the market makes little sense considering BT Retail are the largest and the are not LLU they are BT wholesale and that Openreach pricing is what i quoted.

      Mind you if he wants to do that it brings up another query. Why is it or rather was it so expensive to convert a LLU line back to BT???


      It looks like it took Ofcom 8 years to crack down on that little rip off and get it down from £70 to a more reasonable but still expensive to effectively unplug and pluging in a cable £30.

      As for voice equipment i imagine these charges are designed to pay for that IN PART…

      Given the title is “calling features” and i assume some equipment is needed for those features. Otherwise i imagine the line rental pays for the equipment the line is plugged into. If it does not there does not appear to be any other prices listed for whatever he thinks costs more money.

      The rest of what you had to say @NO clue i pretty much agree upon, there is no explanation where the billions each year on line rental is going. The FTTC thing is also a good point, if it costs BILLIONS each year to maintain things how is that paying for it self with such small uptake vs customer base.

    7. Steve Jones says:

      Read the damn Ofcom report on where the money goes and the principle on which it is calculated.

      As for the FTTC network, there’s a completely separate wholesale charge for that. That is what will pay for the FTTC cabinets, fibre terminations at the exchange, maintenance and so on.

      As for VAT, I’m just astonished by your logic. VAT does not go to Openreach, it’s merely collected by them. It forms no part of the income to OR. If you don’t want to pay VAT, then discuss that with the government. No companies do not get income from VAT, they merely have the unwelcome duty of collecting it for the government. VAT forms no part of the Ofcom calculation of costs (or income) for the very good reason it’s irrelevant.

    8. Steve Jones says:

      Apologies for the grammar slips. I was in a hurry and didn’t re-read what I’d written.

      However, referencing the point about FTTC. It’s very likely that it is not yet recovering its full costs through the wholesale charges yet. However, telecommunications investments are relatively long term and the business plan will have been based on various business assumptions. It is very common for new products to operate at a loss in the short term with the an eye to the longer term. The mobile operators did this as did all the LLU operators. In the case of companies with “market power”, the ability to cross-subside from from the relevant products (e.g. LLU and WLR) to compete “downstream” are strictly limited and subject to competition law. In other words, if OR were to subsidise FTTC through overcharging on LLU & WLR (difficult as they are highly regulated) and there was no prospect of FTTC becoming viable, there would be the potential for actions under competition law. However, it’s notable that, far from undercharging for FTTC, Talk Talk commissioned a report claiming FTTC charges were set at far too high a level. (Although if you read the report, you find there are some heroic assumptions made about market penetration rates and the like).

      It’s also true that the majority of shared infrastructure costs in the distribution network (ducting, poles and so on) will be attributed to copper lines and not the FTTC fibre (or any other form of fibre for that matter). In general the vast majority of the infrastructure will be on the D side anyway, simply because it’s far more efficient to run to a concentration point than the part of the network that “fans out” to several hundred properties.

      Some extremely bright (and well paid) Ofcom employees along with consultants spend many man years working on this stuff in great detail using lots of arcane accountancy. All the customers for LLU and WLR are consulted on this, and you can guarantee that they will have a lot to say.

    9. Raindrops says:

      LOL so FTTC is not yet recovering its costs. So now explain the line rental charge as obviously that recovered its cost years ago and is now nothing more than a exercise in greed. The ofcom report mention nothing about where 2.5 billion goes.

    10. Steve Jones says:


      I suggest you take up your complaints about the cost of WLR and LLU with Ofcom. It is, after all, they that set them. As you clearly have doubts that they are setting them correctly, then go complain to them.

      As far as the costs of copper having been recovered many years ago, then that completely ignores the fact that there are ongoing costs in maintaining the system. Such things as rates to be paid, IT support systems, software licensing, hardware maintenance cost, fuel for vans, staff costs. Also, everything in the network has an economic lifespan. That’s vans, cabinets, poles, ducts, frames, cable and so on. Some of those are very long, some not. The period of time for each is just one of the critical factors used by Ofcom.

      Another point is that Ofcom do permit a rate or return considered appropriate to a regulated utility on the estimated value of the BT access network. As the BT OR network represents far and away the largest asset of the company (and hence capital tied up), then even at a relatively modest regulated rate of return, the numbers add up. This may be considered controversial, but remember shareholders bought these assets off the government. As the state has seen fit to impose an intrusive regulatory regime, then they are duty bound to allow a fair rate of return. The alternative is simply to allow BT to operate freely within the constraints of competition laws. Similar regimes are applied to water and power distribution. It would be different if BT had been set up as a not-for-profit company (like Network Rail), but it was not. The assets were sold and if the state sees fit to intervene, then that has to be done in an equitable manner.

      As an alternative, the US incumbent operators are permitted to build next generation networks with what is called “regulatory forbearance”. For good or bad, this is not what is happening in the EU.

      In short, Openreach will (subject to being considered as efficiently run), continue to create a return commensurate with the value of the capital employed in the network. Or at least it will until such time as it becomes irrelevant due to some change of circumstance.

      Other parts of BT (BT Wholesale and BT Retail) are regulated differently due to the different level of competition in those segments. Openreach is very tightly regulated indeed.

    11. Raindrops says:

      “I suggest you take up your complaints about the cost of WLR and LLU with Ofcom. It is, after all, they that set them. As you clearly have doubts that they are setting them correctly, then go complain to them.”

      Indeed it is and as demonstrated by the LLU to WLR conversion link its takes them around 8 years best case to crack down on ridiculous pricing.

      “As far as the costs of copper having been recovered many years ago, then that completely ignores the fact that there are ongoing costs in maintaining the system.”

      What costs? and what maintaining? Nothing has been done to my phone line or telephone pole for years to “maintain” it. I have never seen a BT man at a cabinet maintaining things or up a pole to check things. The only time they touch the PHONE LINES which is what line rental is spose to be for according to you is when there is a fault, even then some faults they charge you to fix or apply a call out fee. Vans are not bought they are leased, fuel increases do not equate to the £1 or so increase the last two increase to line rental has had either (that equates to around 50 million a month based on 25 million subs). Ive heard of fuel being expensive but 50 Million a month… Maybe they should try different fuel LOL. Likewise again the last 2 increase which have happened in less than a year certainly is not for new computer systems or software, unless you are now going to try and claim BT buy staff new desktops every year?

      Even allowing for all that and agreeing there are ongoing costs. Then as No clue mentions, that still does not explain how the FTTC parts of the network are maintained if line rental is a separate cost, because FTTC can not in any way be making enough to “maintain” the 90% of premises it passes. When it only has 20% take up.

      Its quite clear from the last few increases which just so happen to occur around the time of Infinity “special” offers or when they have bought some rights for some sport nobody is interested in on BT Sport the price has gone up what is is being used on now….. To subsidise other things. Sky have done that for years and BT now are trying it on thereself, the difference is Sky unlike BT have no issues admitting charges may subsidise other things.

      You are right about one thing, technically it is Ofcom at fault. I personally do not blame BT for filling their pockets with every penny they can, thats the point of business, what i have a problem with is crap like you are talking making out they are only making a modest amount. Intersting you should say they are only making a regulated fixed amount and you still think that is a good thing to have shares in. Not very logical, but your whole argument they do not make much from line rental was not either.

    12. Steve Jones says:


      The business case for FTTC was not based on 100% take up. Just what the target penetration rate is, I suspect, commercially confidential but there is a clue in that many of the BDUK contracts reportedly have claw-back clauses at a 20% threshold. Thinkbroadband also reported that BT are operating a 12 year payback period for the commercial rollout of FTTC, and 15 years for BDUK contracts.


      As far as maintenance is concerned, then that includes faults of course. I’ve had two in recent years. The first involved digging up the pavement as, when the network was installed locally (in the early 1970s) it used direct burial, and a Y joint had to be remade. As this involved three OR visits (one to ascertain a local fault, another to use specialist equipment to trace the buried cable and a third to remake the connection). On top of that, there were two contractor visits to dig the hole in the pavement and another to fill it in. This sort of thing doesn’t come cheap given the cost of labour in this country, so I suspect it will take a fair time before those costs are recovered. Of course, it’s just anecdote, as is yours, but just as valid. There will be more detailed financial records, which Ofcom will be able to audit, but which will not be available to the public as clearly they are commercially confidential.

      Also much maintenance occurs out-of-sight. As mentioned, Ofcom have tables of replacement periods for different sorts of assets. They also have access to commercially confidential information and to comparable operators.

      Both WLR and LLU wholesale rental rates have been substantially reduced over the past few years reflecting continued cost cutting by BT. Now the pressure to do so may have been imposed, but it’s a clear sign that regulatory pressure works. OR finances are produced separately, and Ofcom will be easily able to see if the rate of return on the capital employed is “fair”.

      That retail line rental prices have gone up (from all service providers) is nothing to do with increase costs from OR. It’s because SPs are clearly shifting costs so they can offer tempting headline costs for broadband (and promotional offers). That line rental is in the “small print” is something that the ASA ought to be looking at.

    13. Raindrops says:

      How long it takes to pay back is irrelevant if you are not making money on a product you can not pay it back. Take up is 20%. There is no way you can you can be making money on a take up so small when you have spent money running it past 90% of premises. If you had a chain of 100 shops and only 2 made money (or 20%) that does not and will not pay for the other 98 shops costs. Thats just simple math and business sense.

      Your example about an issue where the pavement has to be dug up is all very well and good except A) that is not typical and B) If it were typical across the land they would have to charge more.

      I doubt based on a street of 100 homes and them all paying £15.99 line rental per month which equates to approaching 20 grand a year. The rest of the year they probably basically had nothing done to their phone lines that year. So in that case what they had paid may have been spent on that repair, the other roads around them and across the country are not having work like that done regularly. If that were the case money would not be made. BT should be allowed to make money, i have no issue with that but billions in profit each year is ridiculous, that should not be allowed in any public sector industry which is supposed to be regulated.

      Most “maintenance” which occurs nowadays which is out of sight which you refer to is pressing buttons at a keyboard, for a whole raft of things now its either that or entirely automated, especially when it comes to the internet and what ISPs describe on their status pages as “ROUTINE maintenance”. Other stuff say replacing cards at the exchange along with similar items which are not done on a regular basis i imagine has been paid for well in advance with the £130 charge per line when you sign up to BT for a new phone line.

      Lets not pretend BT are some poor suffering child operating on a shoe string with tiny profit, to suggest so is ridiculous. They did not magically find 2.5 billion of their own money for FTTC by being a company that operates with tiny profit margins.

    14. Raindrops says:

      Eek that will teach me to watch TV and type at the same time MATHS fail by me rather than BT for a change that figure of 100 shops and 20% of them should be 20 not 2 shops DOH! Oh well another example where i am happy to admit a mistake.

    15. No Clue says:

      “Lets not pretend BT are some poor suffering child operating on a shoe string with tiny profit, to suggest so is ridiculous. They did not magically find 2.5 billion of their own money for FTTC by being a company that operates with tiny profit margins.”

      Exactly, even if you remove the arguments about how much profit Openreach make or even argue if they make a profit, or even if all the profit is being used to maintain things that statement right there tells you they must be making more than enough money somewhere.

      There is no way Openreach can be operating at barely above cost if you can spend 2.5 Billion on a rollout. NO other company including Sky and Virgin the nearest competitors have spent that type of money.

      The money came from somewhere and looking at all the pricing link its clear most of the money comes from copper line rental and installation of phone lines.

      Its no different to how BT Retail keep jacking up their line rental and call pricing to pay for “new” BT Sport content and stupid sainsbury vouchers.

      Line rental subsidises most areas of BT services just like TV subsidises most areas of Skys other products. Even an idiot can see that.

    16. Curious says:


      I’m not defending any one, or backing up anyone’s claims here. Nor am I a lover of BT.

      But as I’ve worked for quite a few different ISP’s over the past 8 or so years, I’ve come to understand pretty much exactly what it takes to maintain a telephone line.

      People go on about maintenance of a telephone line, and how much it costs.
      Well, that depends on the functionality, location and type of line (they have sometimes used aluminium cables to repair lines).

      I remember a few years ago, working for an un-named WLR customer (ISP)of Openreach, a ‘simple’ line fault took 5 months to fix, and almost £10,000!!
      There were multiple road closures needed (rental of traffic lights/diggers/cherry pickers/not to mention the amount of different engineers that were tasked on this job) and this is not a one off. I’ve seen a few repairs run into the £5/6 thousands. And this is just a few of the customers I’ve seen when working for these companies. How many more have had the same thing happen to them?

      10 repairs a year costing OR £50, £60 or even £70,000.
      Then there’s the equipment maintenance (MDF/HDF/Tie Pairs/Bar Pairs/electricity costs of running exchanges/cabinets).

      I can’t remember how many times I’ve seen new ducts needed, new spans of cable.

      Not to mention how many employees they have (engineers/works control/support for ISP’s etc. eViper escalations team, High Level Escalations team)

      Yes. there is lots of income from line rental (both WLR and LLU) and yes, they’re going to make profit (which telecommunications company in the world isn’t?) but the profit margins of line rental isn’t what you may think here.

    17. Raindrops says:

      BT make around 2-3 Billion a year in line rental…

      SO “10 repairs a year costing OR £50, £60 or even £70,000.
      Then there’s the equipment maintenance (MDF/HDF/Tie Pairs/Bar Pairs/electricity costs of running exchanges/cabinets).”

      Equates out to £700,000, even if we says its £700,000 FOR every other thing you mentioned in your post and times it by a factor of 10 (thats actually more than the amount of things you mentioned) thats still only £7 Million a year which is nowhere near the 2 BILLION plus they are taking a year in line rental.

      You can not (under any name) explain where 2+ Billion a year is going.

    18. No Clue says:

      None of that explains where billions is spent.

  12. Matthew Williams says:

    So am I reading this right that possible speeds are looking to be max of 400Mbps download and 100Mbps up in real world conditions?

    1. MikeW says:

      From the look of the first bar-chart, where the green bar probably represent the real-world deployment that we would see (co-existing with VDSL2), then aggregate speeds of 500-600Mbps seem to be the right order.

      However, that first barchart comes with the proviso “All results with 20dB target to allow comparison with un-vectored results”.

      In the second barchart (for the 66m line alone), it seems that the middle bar (or left-most yellow bar: 23-106MHz, 20dB margin, Vectored) corresponds to the 488Mbps speed of the first bar chart.

      In the second chart, there are then better results with a 10dB target (50% better, so presumably a speed of 732Mbps) and even better when allowed to use the extra 21MHz (22% better, so presumably 893Mbps).

      What does this “dB target” value mean? Is it an SNRM target, like today? Is it a target for how much crosstalk gets removed by vectoring? A noise floor target? Something else? What value should we choose to match “real life”?

      I don’t know the answer to any of those questions (though I’d guess at a vectoring target)

      As I have understood things previously, vectoring of the full 106MHz is taxing on the DSP power we have today – we need better hardware if we are to get the peak performance out of G.fast’s vectoring. Real-world performance on commercial kit in a few years is more likely to be able to get that performance than we can in trials today. It was the same thing in trialling vectoring for VDSL2.

    2. MikeW says:

      A follow-up to the question on “20dB target”:

      It does appear that this is a noise margin, like the SNRM we understand today.

  13. Ignitionnet says:

    The comments on this article may well plumb new depths. Excellent work.

  14. robert scriven says:

    Its great if you are near the cab, im 1200 metres away, on fttc i get 15 down and 700k up , what are the chances of openreach sticking a thingy on the pole which is 25 metres across the road? 😉

    1. FibreFred says:

      Who says its going in the cab?

    2. Ignitionnet says:

      Pretty good given that’s your distribution point and g.fast is a fibre to the distribution point technology.

  15. robert scriven says:

    If this is the next thing after fttc, then it is brilliant, lets just hope they start with people who get 5-20mb who are 800-1500 metres away with poor aluminium and a fair distance from a cab, who have had poor broadband for years.

  16. Col says:

    I’m hoping that Openreach will look at the cheaper option of bonding all copper pairs from the cab to the DP.
    This can deliver up to 400 Mb/s for up to 15 km. It both terminates the bonded signals and the Dp acts as the gateway node for the subscriber ring.
    This system is non-disruptive to existing traditional POTS services and allows the continuity of mandatory and emergency services (E999/E911).

    1. Gadget says:

      doesn’t it also require lots of spare copper pairs?

  17. col says:

    It requires two twisted pairs from each house.

    1. Ignitionnet says:

      You said the same a year ago. I have no idea why you’d care how FTTdp were backhauled.

      Do you work for GTS perchance?

    2. TheFacts says:

      Many lines into houses have 2 pairs.

    3. No Clue says:

      NO the drop wire into some homes has 2 pairs, what is in that and what is available at the pole is not the same thing.

  18. MikeW says:

    I found an interesting analysis by TNO, that studies the cost of FTTH to central Amsterdam. They compare “full FTTH” and a hybrid-FTTH that is really “FTTdp w G.fast”

    In the analysis they have a detailed knowledge of where all the cabinets are, and know the number of ports of each, plus line lengths of 360k properties. They also know the location of all the “splice locations”, equivalent to our underground joint chambers.

    They then play with a number of scenarios, moving the node locations, in an attempt to minimise the copper length, and then further adjustments by setting a minimum number of ports on a node.
    – With no minimum port count, they’d use 10,400 nodes, getting 70% within 100m & 97% within 250m
    – With a minimum of 8 ports, they’d use 7,300 nodes
    – With a minimum of 32 ports, they’d use 4,900 nodes

    The relative costs? Around 250m euro for full FTTH, and 80m euro for the hybrid FTTdp with 10,400 nodes.


    Who knows how similar the UK data would be, but it is a fascinatingly detailed study.

    1. Steve Jones says:

      Thanks – and refreshing to see something that references a useful report and analysis rather than the knee-jerk reaction exhibited by some.

      I think the general, and surely uncontroversial, point is that costs increase the further fibre is pushed out into the network but so does attainable speed. It’s simply where is the best compromise point for what is supportable by the market (and any public subsidies). As the UK is not homogenous with regard to population density and other factors, then it’s highly likely a patchwork will appear as there seems zero prospect of any company (or government) being able to raise the £20bn or so required to do a 100% fibre roll-out.

    2. FibreFred says:

      Thanks for that

      There’s your answer to “might as well just fibre all the way” comments then

    3. Raindrops says:

      No it does not FTTH from that equates to 3x the cost, by the time BT have messed about with FTTC and then their FTTrn/FTTdp and other tweaks like G.Fast i imagine the cost saving compared to FTTH is is going to be significantly less, if significant at all.

  19. jeep says:

    TBH this seems a great idea,i am fortunate to have a semi decent fttc 24d 7up @500 metres fed by pole,the pole is less than 25m away from my house maybe if this tech were to be introduced the end user could opt for paying for the last run of fibre if they wished,which in return would generate some revenue ? .

    1. No Clue says:

      24Mb down seems pretty poor according to some for a 500M line. I personally believe that though, it mirrors other lines i have seen and how poor FTTC can perform.

    2. Daniel says:

      Why bother with upto speeds at 500M?

      FTTH can do 20km without loss of speed with GPON.
      FTTH can do 40km without sweating it with XG-PON.

      Just two random search for manufactures.

    3. TheFacts says:

      It’s all about funding, just need the government to come up with £25Mish.

    4. TheFacts says:

      Error! £25Billion approx.

    5. Raindrops says:

      That would make it 10x more expensive which is not shown via steves info above.

  20. zath says:

    This all looks very promising.
    That said, there are many locations where the current Infinity service is failing due to usage, even over short 200-400m runs, of old wiring that’s not been upgraded or replaced for decades. “Round the moon to meet the sun” routes of wiring created when new dwellings are built in old villages causes many points of resistance in the connection.
    My own 80mbps BTInfinity never managed speeds beyond 56mbps and now, after rewiring from a telegraph pole after a fallen tree broke the line and pole, and after an engineer “solved some problems” I reach just 49.9 mbps.
    I dream of 700mbps but 76-80mbps would be a start.
    I had understood the fibre to the house was a possibility, I assume that isn’t a long term option now.

  21. Daniel says:

    I don’t see how this is going to help drop either costs of the rollout or increase the average broadband speeds.

    G.Fast is technically limited to 250meters, by default, those with long lines or poor copper are going to get same results, regardless of the copper technology is put at the end of the Fibre trail.

    The problem with people and getting all wet over these speeds shouldn’t get excited over the sudden G.Fast miracle.

    Like many other xDSL technologies that have already been invented – that were not used.

    My country (Australia) is going the same route as you guys are, all for a measly suppose savings of something like $10 billion.

    1. Raindrops says:

      I guess some think a technology that loses around 5Mb every 1 Metre of copper is “progress” LOL

    2. TheFacts says:

      Like all new technology it is progress.

    3. MikeW says:


      The length of copper, and the quality of the copper, matters hugely to G.fast – so *the* most important factor for speed is: Where do you put those G.fast/FTTdp nodes?

      There’s a study of Amsterdam that asks exactly that question:

      In that, they come up with a model of 10,400 nodes that manage to get 70% of 360k premises within 100m of a node (from graph above, this is 500Mbps). That would cost roughly one-third of a full FTTH.

      The real saving for any individual town or country will depend on how those numbers change. That will dictate which way a telco chooses.

      From my understanding of the NBN, the question for Oz, much more than here, is whether the quality of the copper in the last 100m is good enough.

    4. Raindrops says:

      That one third saving does not calculate the original FTTC cost does it?

  22. MikeW says:


    You do like to keep dropping your approximation of “100Mbps per 20 metres” into the discussion here, don’t you? I count 5 or 6 times now. It is a bit like a dog with a stick, bringing it back to show master, hoping he’ll throw it again for you to comment some more.

    And just like the dog, you could be wagging your tail in excitement at showing how bad this technology must be to lose so much speed every metre. You must feel like such a clever boy. LOLing and putting quotes around “progress” to show how deluded everyone is being.

    I think you must be enjoying yourself.

    Here’s the thing… The more accurate readings from the article actually live up to everyone’s expectations.

    The most accurate I can be about the speed drop (comparing the purple bars) is 130Mbps between the closest house and the farthest house. And those houses are 47 metres apart. So the drop seen amounts to 2.8Mbps per metre of copper.

    In the Alcatel graph, we can see that the expectation is for 750Mbps at 10 metres, and 150Mbps at 250 metres. A drop of 600Mbps over 240 metres total is, overall, 2.5 Mbps per metre of copper.

    The more accurate drop in speed is pretty damn close to what was expected. If you are happy that your “5Mbps per metre” counts as a good enough approximation, I’m certainly content that mine is better. It is certainly within the realms of experimental results using a first attempt at the G.fast spec in (almost certainly underpowered) generic FPGA hardware, as Mark was careful to indicate.

    Harder tests are yet to come for G.fast, but so is better hardware.

    The hidden data behind this is that speeds with a 10dB target are considerably higher. But as they’ve been given only for one house, we can’t calculate the drop. All we know is that, for the one distance of 66m, speeds are 50% higher than the first chart.

    1. TheFacts says:

      Why does this upset you, it’s the result of a trial many telcos are doing?

    2. TheFacts says:

      Above is reply to Raindrops.

    3. No Clue says:

      He already explained his statement several times. He said “ABOUT 100Mb” which it is.

    4. MikeW says:

      (S)He has said “about 100Mbps” over “20 metres” and then equated that to 5Mbps per metre. You are right – it is ABOUT – but only if you are being pretty generous. When the actual value is near half of Raindrops’ estimate, you have to ask if “about” is an adjective that cuts the mustard.

      I can say that it is “about 1Mbps per metre” – one fifth of Raindrops’ estimate – and still be closer than him/her.

    5. No Clue says:

      Nope he is about right. Here http://www.ispreview.co.uk/index.php/2014/09/first-bt-trial-future-g-fast-copper-broadband-hits-nearly-720mbps.html#comment-147247
      You suggest it is nearer 70Mb rather than his 100Mb.

      5(Mb loss per metre) x 20 (length in metres) = 100Mb So you could argue he is 30Mb out based on your earlier suggestion that its nearer 70Mb.

      Your example of…

      1(Mb loss per metre) x 20 (length in metres) = 20Mb which would make you based on your earlier 70Mb suggestion 50Mb out.

      He would be closer.

      Factor in this is a trial and as the story clearly states “The trial used existing duct structures and distribution point (DP) locations, but created a separate overlay to prevent harm to BT’s existing plant.” and im sure it will perform worse than the 70Mb loss per 20 Metres you estimated earlier.

      Put all that together and it is indeed AROUND 100Mb loss per 20 Metres of copper or around 5Mb per metre. In fact in a REAL life product id say that would probably be within 10-15Mb (either way better or worse) of being bang on correct.

    6. Raindrops says:

      Maths and what has already been stated and explained earlier in a news item based on his continual attempts to try to be right in another one is just going to confuse him.

Comments are closed.

Comments RSS Feed

Javascript must be enabled to post (most browsers do this automatically)

Privacy Notice: Please note that news comments are anonymous, which means that we do NOT require you to enter any real personal details to post a message. By clicking to submit a post you agree to storing your comment content, display name, IP, email and / or website details in our database, for as long as the post remains live.

Only the submitted name and comment will be displayed in public, while the rest will be kept private (we will never share this outside of ISPreview, regardless of whether the data is real or fake). This comment system uses submitted IP, email and website address data to spot abuse and spammers. All data is transferred via an encrypted (https secure) session.

NOTE 1: Sometimes your comment might not appear immediately due to site cache (this is cleared every few hours) or it may be caught by automated moderation / anti-spam.

NOTE 2: Comments that break our rules, spam, troll or post via known fake IP/proxy servers may be blocked or removed.
Cheapest Superfast ISPs
  • Vodafone £19.50 (*22.50)
    Speed 38Mbps, Unlimited
    Gift: None
  • NOW £20.00 (*32.00)
    Speed 36Mbps, Unlimited
    Gift: None
  • Hyperoptic £20.00 (*25.00)
    Speed 50Mbps, Unlimited
    Gift: Promo Code: HYPERFALL21
  • Shell Energy £21.99 (*30.99)
    Speed 35Mbps, Unlimited
    Gift: None
  • Plusnet £22.00 (*38.20)
    Speed 36Mbps, Unlimited
    Gift: £70 Reward Card
Large Availability | View All
Cheapest Ultrafast ISPs
  • Gigaclear £24.00 (*49.00)
    Speed: 300Mbps, Unlimited
    Gift: None
  • Vodafone £24.00 (*27.00)
    Speed: 100Mbps, Unlimited
    Gift: None
  • Community Fibre £25.00 (*27.50)
    Speed: 200Mbps, Unlimited
    Gift: None
  • Hyperoptic £25.00 (*35.00)
    Speed: 150Mbps, Unlimited
    Gift: Promo Code: HYPERFALL21
  • Virgin Media £28.00 (*52.00)
    Speed: 108Mbps, Unlimited
    Gift: None
Large Availability | View All
The Top 20 Category Tags
  1. FTTP (3570)
  2. BT (3023)
  3. Politics (1941)
  4. Building Digital UK (1929)
  5. FTTC (1888)
  6. Openreach (1837)
  7. Business (1693)
  8. Mobile Broadband (1480)
  9. Statistics (1410)
  10. FTTH (1365)
  11. 4G (1277)
  12. Fibre Optic (1174)
  13. Virgin Media (1173)
  14. Wireless Internet (1163)
  15. Ofcom Regulation (1149)
  16. Vodafone (846)
  17. EE (835)
  18. 5G (772)
  19. TalkTalk (769)
  20. Sky Broadband (747)
Helpful ISP Guides and Tips

Copyright © 1999 to Present - ISPreview.co.uk - All Rights Reserved - Terms , Privacy and Cookie Policy , Links , Website Rules , Contact