Broadband News

How is the rural versus urban broadband divide looking?

There is a common complaint when we talk about the levels of broadband coverage across the UK and that is that averages hide the reality, but when tracking towards national goals if quoting a single figure or talking for just 30 seconds on radio it is impossible to convey the full picture. Our labs.thinkbroadband.com/local stats site lets people drill down to the constituency and district council levels and even further down to the individual postcode level, but we thought given the latest news on broadband spending in the it is time to once more publish our rural versus urban data.

The table below shows the currently levels for the different rural and urban classifications as used by ONS, and then three aggregate figures are included, GB Urban and GB Rural make up the full GB area, but a deeper rural classification that is sometimes called the final 10% or the most deeply rural areas of the UK is also present as GB Deep Rural.

thinkbroadband analysis of Superfast, USO and Full Fibre Broadband Coverage for the rural and urban areas of Great Britain and Northern Ireland
In descending order of superfast coverage - figures 11th September 2017
(change since 14th September 2016)
Area% superfast
24 Mbps or faster
% superfast
30 Mbps or faster
% Ultrafast
100 Mbps or faster

% Full Fibre
FTTH/FTTP

% Under 10 Mbps download

GB Urban (74.8% of premises)

98.6% (+1.6) 98.4% 66.1% (+2.6) 1.4% 0.5%
GB Urban - less sparse 98.6% (+1.6) 98.4% 66.1% (+2.5) 1.4% 0.6%
Belfast Metropolitan 98.4% (+0.9) 98% 72.7% (+5.3) 0.1% 0.5%
GB Urban - area sparse 98.3% (+2) 98.1% 64.4% (+3) 0.1% 0.3%
GB Town and Fringe - area sparse 97.7% (+3.1) 97.4% 41.4% (+4.7) 0.8% 0.4%
Derry Urban 97.5% (+1.7) 96.9% 76.5% (+13.6) 0% 1.7%
NI Large Town 95.4% (+2.2) 94.5% 0.2% (+0.2) 0.2% 1%
NI Intermediate Settlement 94.6% (+3.3) 94% 0.4% (+0.2) 0.1% 2.2%
NI Medium Town 94.5% (+2.2) 93.8% 0.3% (+0.3) 0.3% 1.6%
NI Small Town 94.4% (+2.3) 93.2% 0% (=) 0% 1.7%
GB Town and Fringe - less sparse 92.8% (+4.3) 92.2% 15.5% (+1.3) 1.5% 3.2%
NI Village 91.1% (+6.8) 89.7% 0.6% (+0.6) 0.6%  4.3%
GB Rural (23% premises) 83.8% (+6.2) 82.5% 15.7% (+2.4) 2.5% 9.8%
GB Village - less sparse 78.5% (+8.7) 76.4% 8.4% (+2.3) 3.2% 13.9%
GB Deep Rural (10.8% premises) 72.1% (+8.6) 69.9% 8.3% (+2.6) 3.8% 18.2%
GB Hamlet - area sparse 58.5% (+11) 57% 9% (+2.9) 8.5% 32%
GB Hamlet - less sparse 54% (+9.2) 50.8% 9% (+8) 4.1% 29.8%
NI Small Village 43.8% (+5.6) 41% 2.2% (+1.1) 1.7% 40.2%

The good news is that in the last 12 months the gap between urban and rural areas has been closing up, but clearly still a good deal more work to do, hence the continued announcements from local authorities of project extensions and the gainshare news from the weekend. The figures also help to highlight why when people from villages complain that 90% and 95% looks totally impossible it is very often down to the reality that villages and hamlets are still well behind the larger towns.

The good news for rural areas is that full fibre (FTTP) is growing rapidly as the following bar charts that show Sept 2016 and Sept 2017 illustrate, though if your concern is just ultrafast downloads and not the technology itself the presence of Virgin Media in urban areas means they are still king for ultrafast coverage and with DOCSIS 3.1 cable on the way over the next few years those download speeds could be 500 Mbps to 1000 Mbps.

Chart of FTTP availability across GB and NI in Sept 2016
Click image for larger version

Chart of Full Fibre (FTTP) availability in rural and urban areas of GB and NI in September 2016

Chart of FTTP availability across GB and NI in Sept 2017
Click image for larger version

Chart of Full Fibre (FTTP) availability in rural and urban areas of GB and NI in September 2017

The next few months as the UK races to the overall 95% target will of course see things change, but as is obvious from the chart the urban areas will be well ahead still and the size of the UK is such that even with a 98% superfast ambition it would be possible to reach that goal without any further roll-out in the GB Hamlet - area sparse grouping which comprises some 174,500 premises.

One sneaking suspicion we have had for some time, is that some people believed the BDUK roll-outs were to deliver 90% coverage specifically to the rural areas i.e. in those 6.5 million premises that make up the GB rural 23% of premises that the aim was to deliver superfast broadband coverage to 5.85 million, when the reality was that the 90% goal was always the combined result of commercial + BDUK footprint.

Comments

I live in rural N. Ireland about 6 miles outside of Omagh and according to the Openreach checker we are now "in scope" for a FTTP solution after years of sitting at "seeking solutions" so hopefully things are moving in the right direction.

  • pipcoo
  • 9 days ago

A most useful piece of work. The kernel of the problem is exposed for all to see:- "even with a 98% superfast ambition it would be possible to reach that goal without any further roll out in the 174,000 premises in GB Hamlet sparse area"

Ofcom needs to devote more attention to this end of the market. I lived in a hamlet where it was clear nothing would ever happen. The exchange concerned now has superfast, the hamlet is still under 5Mbps and looks like being there for many moons to come

Rural will eventually be solved by pressure of complaints. Hamlets need a more vigorous Ofcom

  • 961a
  • 9 days ago

The final paragraph is a little misleading. BT was going to do 66% which included a significant amount of FTTP. BT has not done £2.5bn of work but £1.5bn worth - less than 50k cabinets before capacity cabs are added.
The commercial rollout in Wales and Northern Ireland was less than 50% and in Scotland less than 60%. This still excludes all the in-fill needed in BT areas where FTTC does not support superfast.
BDUK in doing 4.5m premises had had to pass more than 6m premises to do so.
The state is doing more, BT has done less and we have big gaps to fill.

  • ValueforMoney
  • 9 days ago

So @valueformoney are you saying BT has done less than 66% in terms of superfast roll-out using commercial money? If so what percentage has BT delivered, and all the other commercial projects?

You appear to forget that the BDUK projects are only paying for the 4.5 million superfast premises, those additional premises don't count towards being invoiced for delivering superfast.

If anything the pressures to reach 90% may have compounded any decisions to switch from P to C in some areas, to ensure timelines held.

  • andrew
  • thinkbroadband staff
  • 9 days ago

I live in a village 2 miles from the exchange, so I think I can be called "rural". Last year we went onto fibre with a box 100yds away. Believe it or not we regularly get 74 Mbs on Infinity2. Don't think I can complain about that.

  • ndevongull
  • 9 days ago

@VFM

Where do you get that figure that of premises passed by FTTC in BDUK only 75% can get 24mbps or more? Even in Scotland, where you might expect the ratio to be worst dye to geographical spread, it's 87%.

Given approximately 95% coverage at SF then that 1.5m you estimate has been passed yet can't get 24mbps essentially means every property in the BDUK areas must be on an enabled cabinet. That's clearly not the case.

So go ahead - from where have you got that 6 million properties passed figure?

Bear in mind that there have been many infill cabinets added.

  • TheEulerID
  • 9 days ago

Incidentally, Northern Ireland is a rotten example to choose as the intervention scheme that started there coincided with the commercial roll-out. In NI it's very difficult to distinguish what is commercial and what is subsidised as it was done at much the same time and goes some way to explain why the subsidy cost per cabinet was lower.

Also, BT was not going to be doing a lot more FTTP in the commercial rollout - the plans for 2m FTTP were dropped in favour of about 18m FTTC for financial reasons.

  • TheEulerID
  • 9 days ago

I put in an FOI request to BDUK at the end of 2015 the answer to which indicated about 500,000 or 15% of premises passed at that time would not get 24Mbps. Since then there will be have a certain amount of infill to lower this figure.

  • gerarda
  • 9 days ago

@gerarda

Thanks. An initial figure of 85% makes sense and is roughly in line with those Scottish numbers and, as you way, the infill will increase that further.

  • TheEulerID
  • 9 days ago

Ofcom's Infrastructure report (2013) has a graph that suggests about 90% of FTTC lines get 24Mbps+.

A graph from the 2014 report suggests that 90% of FTTC lines got 30Mbps+.

Those will be predominantly commercial results. I wouldn't be surprised for the BDUK projects to have a lower percentage - and I'd accept the 85%. I'd equally expect that infill will improve this somewhat.

  • WWWombat
  • 8 days ago

You sure that graph was 90% of FTTC lines, i.e. cabinets delivered in 2013 or 90% of the line population for the UK overall as the network stood at that time?

A big difference.

As for BDUK vs commercial commercial seems to be at 95% superfast, bduk Scotland 91.2%, HIE 80.6%, Wales 91.8, NI 84.1

In short a wide variation and could do an overall for BDUK vs commercial but out of date in a few days and only settles 1 small question some have

  • andrew
  • thinkbroadband staff
  • 8 days ago

  • brianhe
  • 8 days ago

Not sure what classification I come under as an isolated rural premise. On 4km line from exchange, cab near exchange which is too far away, now in fill cab installed, which is also too far away, so fttc passed twice. After years of promises, Superfast Scotland have now admitted they have no plans.

  • brianhe
  • 8 days ago

Not sure what classification I come under as an isolated rural premise. On 4km line from exchange, cab near exchange which is too far away, now in fill cab installed, which is also too far away, so fttc passed twice. After years of promises, Superfast Scotland have now admitted they have no plans.

  • brianhe
  • 8 days ago

GB Hamlet - area sparse would be the underlying area i.e. most rural category.

  • andrew
  • thinkbroadband staff
  • 8 days ago

brianhe

"After years of promises, Superfast Scotland admit they have no plans"

"our commitment to make sure that 100% of the UK can get affordable, fast and reliable broadband by 2020" Digital Matt Hancock

Dream on

  • 961a
  • 8 days ago

Just a thought... would be nice to see the percentage change in "under 10 Mbps" column. Not sure if it is something that can be easily done, but it would give us an indication on how much the improvement works benefit those with slowest speeds, and how much goes to the easy part of the network upgrade.

  • hvis42
  • 8 days ago

Change in USO gets a call out in the monthly summary in words, but can say 12 months ago we reported 4% for UK and down to 2.9%, so a drop of 315,000 in the last 12 months

  • andrew
  • thinkbroadband staff
  • 8 days ago

@961a Scotland has its R100 ambition, but no flesh on the bones as yet.

On the Matt Hancock statement, was referring to the USO at that point I believe where 10 Mbps is often referred to as high speed or fast broadband.

If I was script writing I'd have ensured it was worded differently, as some outlets are using that quote in a way that may make some think he means 100% superfast.

  • andrew
  • thinkbroadband staff
  • 8 days ago

@EulerID - the c6m passed comes from a physical counts of premises attached to the some 23k cabinets installed to last December. It includes those accidentally passing Virginmedia and those who too far or who cannot get 24mbps.

  • ValueforMoney
  • 8 days ago

Andrew

"If I was script writing..."

But you're not a politician. I'd say that was exactly why it was written the way it was

As has been said several times before, those in these hamlets and elsewhere who are on speeds less than 2Mbps at which point the whole thing is useless care not what the definition is as long as someone gets off their butt and provides useable broadband.

  • 961a
  • 8 days ago

Andrew /Euler Based on a OR system size of 26.5m (not 28.8m) BT can claim they passed and connect 66-67% of premises or perhaps 18m, using c47k cabinets. How many of these can get >24Mbps? Using 87% reduces it to 15.6m.
BT did less and invested less. The state did more, but we have learned it is cheaper than portrayed and demand is greater. The resources are there to do a more complete job.
BDUK pay all costs presented, the full costs are paid over fewer customers. They do not pay less when overbuilding Virginmedia, but it may get fixed once the clawback is payback post 2023,(less the early release of £130m) unless some other plan emerges.

  • ValueforMoney
  • 8 days ago

The reason I asked was a hunch that there might be a regional difference in this.

Does the sub 10Mbps percentage decline proportionally to the size of the problem, or is there another divide there as well? I would not be surprised if sub 10Mbps figure was more or less the same in areas not receiving BDUK funding, which in turn would prove the Ofcom shibboleth "market and competition will sort out the urban areas" wrong.

  • hvis42
  • 8 days ago

What is this OR system size of 26.5m? The UK has more premises than that, and as the Openreach network via USO should be available to all UK premises using a subset to do your maths smacks of picking and choosing to get an answer that fits a theory.

As for doing more, you do realise that this is exactly what this news item is about, doing more and that it has not cost as much as originally thought.

On the pay less in Virgin Media areas, care to share an invoice showing that you say is true? Would make for a court case.

  • andrew
  • thinkbroadband staff
  • 8 days ago

@961a and Andrew From what I have seen the USO is not in the BDUK plans or at least not in the Suffolk ones. Lots of postcodes where the cabinet has been enabled but speeds of less than 10Mbps are achievable as showing as upgraded ie job done. This was not something that was showing a few months ago, where these postcodes would have been shown as still awaiting an upgrade

  • gerarda
  • 8 days ago

@Gerarda councils have not had any USO responsibility devolved to them, i.e. USO is still up for grabs

As for Suffolk infill is actively on going, as for postcodes showing as superfast available but speeds of less than 10 Mbps, feel free to share those postcodes for checking

Remember Openreach saying cabinet is enabled does not mean superfast subsidy has been paid, the actual speed available will be the decider.

  • andrew
  • thinkbroadband staff
  • 8 days ago

@Andrew
That 90% figure is of FTTC lines, yes. Both graphs.

@brianhe
ONS definitions for "hamlet" actually refer to "hamlet and isolated dwellings", so you will be in one of those two classifications (sparse or less-sparse).

The definition of "sparse" comes from looking at a much wider area. ONS simplistically say "This broadly reflects the surrounding 30km being characterised by low population density."

There's a good map of E+W on this ONS page, showing the sparse areas:
https://goo.gl/N4auwC

  • WWWombat
  • 8 days ago

@brianhe
Sorry, don't have the equivalent map for Scotland.

@vfm / @euler
The last time I had figures for the SF vs non-SF split of BDUK was for December 2015.
At that time, it was reported as 4m total passed in BDUK vs 3.5m with SF speeds.

That's a figure of 87.5%.

I can't believe that, in work done since then, it has needed 2m premises passed to cover 1m with SF speeds.

  • WWWombat
  • 8 days ago

@vfm
You're right that BDUK just pays the bills presented, no matter whether premises covered are VM or not.

However, the value-for-money calculation (to decide whether to upgrade a cab in the first place) only uses the non-VM, non-superfast count as a denominator.

Too many VM => cabinet not included at all.

Finally, though, the BDUK contracts do require that BT end up reaching the threshold of premises capable of getting superfast speeds ... and that total too can only come from premises that were non-SF, non-VM beforehand.

Otherwise BT have to add cabs at their own expense.

  • WWWombat
  • 8 days ago

And here is the kicker - which projects has BT failed to deliver the contracted number of superfast premises?

  • andrew
  • thinkbroadband staff
  • 8 days ago

@brianhe / @andrew

It seems that Statistics Scotland uses a slightly different measure from the sparse vs less-sparse that ONS uses: remote vs accessible. Alongside an 8-fold rural-urban classification.

The remote distinction appears to be whether the place is within a 30 minute car drive of a town of 10,000+ population.
http://www.gov.scot/Topics/Statistics/About/Methodology/UrbanRuralClassification

I guess that might be reasonably equivalent to the 30km threshold for ONS.

Here's the 2014 map:
http://www.gov.scot/Publications/2014/11/2763/2#map2.2

  • WWWombat
  • 8 days ago

@Andrew - the better broadband to suffolk site probably deliberately does not say these postcodes can get superfast. It describes them as "already upgraded" or "existing coverage area"

See http://www.betterbroadbandsuffolk.com/LineCheck.aspx?strPostcode=ip300ah and the neighbouring postcodes for example.

  • gerarda
  • 8 days ago

Yeap aware of the difference in Scotland and did adjust them to try and line up with the England/Wales ONS stuff. The NI stuff was too different to roll into a single UK set of figures.

  • andrew
  • thinkbroadband staff
  • 8 days ago

@VFM

"the c6m passed comes from a physical counts of premises attached to the some 23k cabinets installed to last December. It includes those accidentally passing Virginmedia and those who too far or who cannot get 24mbps."

So you've gone round and counted them physically? Or perhaps you mean you have detailed configuration information not available to the rest of us? What I think is more likely is applied some number (260 works) as the average number of lines per cab and multiplied it by 23,000.

So in other words, make believe and completely at odds with reported stats.

  • TheEulerID
  • 7 days ago

@VFM as to overbuilds with VM, then what's the source for that number then and just how many is it? As far as I recall, only limited overlap was allowed and it surely doesn't account for the difference between 87% and 75% or else VM would surely have been screaming blue murder.

  • TheEulerID
  • 7 days ago

@VFM
Rather than repeatedly making unsubstantiated allegations supported largely by hearsay, extrapolation and guesswork, why not just submit your dossier of "evidence" to the authorities so that they can investigate them properly?

Otherwise your posts amount to little more than trolling, which I would imagine is an unhelpful image for an industry consultant to be projecting.

  • New_Londoner
  • 7 days ago

@gerarda The website like some of the others are vague, but given distance to the cabinet more like 7 Mbps looks likely hence us saying

https://labs.thinkbroadband.com/local/postcode-search-IP30%200AH

  • andrew
  • thinkbroadband staff
  • 7 days ago

TheEulerID; New_Londoner - Vodafone have submitted an annex via the WLA consultation response annex 2. It includes the evidence available, including the premise and cabinet counts. It should be the substantiation you have been yearning. It would good to see BT publish a full numerated response to this submission.

VM have gone to the EU and their submission to the CMS select committee outlines the issue pretty well.

Of the £463m capital deferral owed, only £130m has been released, the rest is resting in BT's accounts. LA investment account balances are an extra.

  • ValueforMoney
  • 7 days ago

Brianhe

I know it is not what you want to hear but! If you are a single isolated property 4km from the exchange all the relevant authorities will be assuming that Satellite is what will provide your greater than 10Mb service.

I know and appreciate that some properties(especially in Scotland)cannot get line of sight but this has not yet hit home to the decision makers. Also for those that can get satellite, latency etc can mean some services are still not what you would like. However they would say that all (non leisure) essential service would be accessible and that fulfils their promise.

  • jumpmum
  • 7 days ago

And the other money will be released once projects decide what to do, did you not notice the bit about £200m of the £645m already committed, so even DCMS release shows there is more available to do more. Or do you disagree on the £645m

  • andrew
  • thinkbroadband staff
  • 7 days ago

Hi Broadband Watchers.
Could it be that imperial measurements are not been used to the basic mile E.G. 1760 yards to a mile I use this method as I have the (GPS position of the FTTC )and the postal run (Post Code GPS position) and the customer GPS position from this you can get the approx Sync Of the address. The Exchange has its GPS position so it is easy to locate the GPS for the FTTC/GFast.

  • Blackmamba
  • 7 days ago

Andrew, the release of the remaining funds is under BT's control. It is sitting in BT's accounts as a Deferral. £130m was released early by BT to be spent with BT. You may find that the remainder has not been released because it might need to be competed for. I cannot find a reference in the new state aid doc on the use of clawback above £130m, but there is something stopping it being re-applied directly with BT.

Projects like CDS cannot plan because much of this information is withheld to a time convenient for BT to declare. The gaming of costs is deeply sub-optimal for all.

  • ValueforMoney
  • 7 days ago

@blackmamba Nothing to do with imperial measurements at all, never have and never will use imperial measurements in any analysis we do and very much doubt anyone else but you is too.

  • andrew
  • thinkbroadband staff
  • 7 days ago

Projects like CDS CANNOT plan for reinvestment until it knows which premises are actually going to be covered in phase II, and am sure they can ask for guidance from Westminster on the amounts they are entitled or work it our for themselves or pay an expensive consultant to do it.

  • andrew
  • thinkbroadband staff
  • 7 days ago

Just putting this out there....

30,085 cabinets with BDUK projects average 139 superfast premises per cabinet (FTTP is not included in this)

51,894 commercial live VDSL2 cabinets average 481 superfast premises per cabinet (FTTP is not included in this)

Additional cabinets for capacity are not included in these figures.

  • andrew
  • thinkbroadband staff
  • 7 days ago

@VFM

It's wholly untrue that money in the deferral fund is under BT's control. The deferral "fund" is simply a recognition of a future liability either to repay the money or for local BDUK projects to use as reinvestment.

If BT just went ahead and invested that money it would not reduce the deferral fund by one penny. The liability would remain. Local politicians would be very happy - they'd see the roll-out happening knowing at the end of the contracted period they'd get that gainshare money without the need to reinvest it. BT shareholders - they'd be less than happy...

  • TheEulerID
  • 7 days ago

@EulerID LOL - you are funny. The monies are under BT's control as they are owed to Gov/LA but not usable by them until BT agrees to release it for future procurement. In the mean time BT is free to use that cash flow from subsidies to pay for Football.

  • ValueforMoney
  • 7 days ago

Andrew, useful - but BDUK cabs, 30k x 139 is less than the superfast passed figures claimed by BDUK, so that needs checking.Last Christmas the average was 219, 28.9k cabs against 6.36m passed (all, not superfast).
BT has passed nearly 52k x 480 > 24m premises hmm Dec2016 I counted 48.9k cabs against 19.3m - or an average of 391premises (all, not superfast) Interested to see your 480.
Thanks

  • ValueforMoney
  • 7 days ago

Did you not see the bit where I said FTTP is NOT included comparing with all (not just superfast) does you no favours when trying to query the data. Who says BDUK figures are 100% correct anyway...

As for seeing the 480 you can pay like anyone else who needs raw data access, or wants to commission a report.

  • andrew
  • thinkbroadband staff
  • 7 days ago

An additional variable is that I've removed the superfast premises where Virgin Media is available, but that is reducing the amount BDUK delivered since the cable roll-out is visiting some areas where cabinets have previously been delivered with gap funding.

  • andrew
  • thinkbroadband staff
  • 7 days ago

Andrew, FTTP makes no measurable difference. So you have an estimate for Virginnedia overbuild, can you publish? This still does not reconcile to the BDUK number. 139 is too small, the cabinets number look ok.
If you provide a total number for commercial why not a total count for BDUK's work? It is a bit mis-leading.

The 480 may include Virginmedia so the difference is interesting 4.5m, that makes sense.

  • ValueforMoney
  • 7 days ago

So what is the BDUK number?

And how much FTTP means no measurable difference?

Throwing away what might be 200,000 premises sounds like someone trying to make things fit their own narrative rather than constantly trying to move with the times.

  • andrew
  • thinkbroadband staff
  • 7 days ago

@VFM
I found the document after some hunting. Much better when data is sourced. Links are nice, so there's one at the end.

Looking past the gratuitous brickbats at a competitor, there are some reasonable gripes (like the paucity of audited BDUK public numbers, including premises passed), some unreasonable (not having detailed access to OR BB investment breakdowns). Lots of questions (e.g. 6.3m premises passed yet 5.5m in the intervention area). More reading, more observations to follow.

https://www.ofcom.org.uk/__data/assets/pdf_file/0021/105249/Vodafone-annex-2.pdf

  • TheEulerID
  • 7 days ago

@VFM

"but not usable by them until BT agrees to release it for future procurement."

Wrong. The accelerated clawback exercise was created to allow BDUK projects to reinvest the money quicker. It requires a new contract in place (and BDUK projects are report it needs a new an agreed deployment plan - many intervention areas been amended & local priorities change) but that's it.

However, if what you expect is BT will hand cash back before term end or a new contract, that won't happen. The framework was designed to reinvest with the contractor within the contract period.

  • TheEulerID
  • 7 days ago

Now having a chance to read the Vodafone submission, perhaps they should buy some consultancy from someone a body that is tracking the roll-outs, it would be good to get some payback for the 28 working weekends so far this year.

  • andrew
  • thinkbroadband staff
  • 7 days ago

Hi Broadband Watchers.
You may find the information on Ofcom. Post Code DBase which classes premises that pay tax on the building in that post Code group not the ones on the old DP records which are way out. Ect names changed. As these records are postal they can be measured either metric or imperia between post Codes l. I have checked via ZENs incorrect and BT records incorrect Ofcom correct. I have a feeling that Surrey is using OFComs results. Ofcom states 30 meg as Superfast and Surrey 24 Superfast.

  • Blackmamba
  • 7 days ago

@Andrew

That Vodafone report has some questionable numbers and claims. For example, the note under table 11 says

"not all premises connected to a cabinet receive a superfast broadband signal, approximately 70% receive superfast broadband speeds"

The estimate that only 70% of lines on enabled cabinet can get >24mbps strikes me as far too low.

As for those 28 working weekends, we know you love it really. All those stats and data cleansing/reconciling exercises. Heaven.

  • TheEulerID
  • 7 days ago

Well it should be obvious to anyone in the industry that I have a lot of analysis available and its not based on what are sometimes questionable premise counts from BT

  • andrew
  • thinkbroadband staff
  • 7 days ago

@Andrew

I would trust an impartial source rather more than those with vested interests. I used to have to do a lot of mangling of stats and one thing I learnt long ago is never to trust any of it without some form of independent cross-checking. Also, the same term gets used many different ways with different definitions and (mostly) people repeat numbers without question and without even doing a sanity check.

Oh - and people operating computers often put the wrong information into systems and the god of entropy means it accumulates randomness over time.

  • TheEulerID
  • 7 days ago

EULARID - most of the sources quoted is BT data in the public domain and representations relied upon by BT at various times. The efficacy of copper gain kit varies by geography, the report quotes more than one number.

Andrew - so how much VM overbuild are you seeing? If your declaring c30k BDUK cabs with an average of 139 premises, then the overbuild is huge, not incidental. That Vodafone report includes an annex of urban subsidised build. I an certain you could do better.

  • ValueforMoney
  • 6 days ago

Suffice to say this, if a BDUK project was to declare a goal that we were confident was false we would be all over it like a rash. Two fold advantage, council is mis reporting and/or Openreach has not delivered.

  • andrew
  • thinkbroadband staff
  • 6 days ago

Hi Broadband Watchers.
I was under the impression in Surrey SCC BD UK had paid for a CaB with 100 ports/tie pairs to provide 80/20 to customers in the Cabs area only. It was the Responsibility of the ISPs to take up the subserdised 100 ports for their customers. When Openreach provided extra ports they were not subserdised by SCC.

  • Blackmamba
  • 6 days ago

Did this survey include Scotland? Specifically the Northern Highlands. Very poorly served up here. I am still struggling to achieve 6Mbps on a DSL Max exchange and that is good servcie for what it is.
No sign or chance of a Fibre service for us.

  • jpirie
  • 6 days ago

@VFM

Why you think that 30K BDUK cabinets at 139 SF premises per cabinet + 200K FTTP means there's a massive amount of VM overbuild, I've no idea. Andrew's figures appears to show just short of 4.4m premises, BDUK claim around 4.5m. That's a difference of 100K (a little over 2%) and Andrew has (many times) stated that TBB's speed estimation system is deliberately on the conservative side. That is on the low side (which matches my experience in two properties).

So how have you worked out that BDUK is overbuilding very substantial numbers of existing VM networks on that basis?

  • TheEulerID
  • 6 days ago

Yes as its a GB figure (with NI called out on their own) it does include the Highlands and those IPStream Max or worse exchanges.

Lots happening in Highlands http://labs.thinkbroadband.com/local/highland,S12000017 but with 10% of premises still on ADSL/ADSL2+ a lot more to do, and also line lengths mean even when cabinet enabled or added to EO area not everyone gets superfast....NOTE: A VDSL2 connection at 7 to 10 Mbps down while only marginally faster than ADSL at 6 Mbps should feel a lot better as the backhaul side of things is better, and DLM system behaves a lot nicer.

  • andrew
  • thinkbroadband staff
  • 6 days ago

@VFM

Oxera looked into the issue of BDUK overbuild of VM areas. Given that OR cabinets cover wide areas a certain amount of overlap was inevitable unless those premises on the OR cabinet not serviced by VM were to be left out of the scheme. The question is how many. The Vodafone report picks an example, but is otherwise rather lacking. We have also been told that in assessing cabinets, the value calculations were based on unserved (by VM) counts.

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/428381/The_UK_s_National_Broadband_Scheme_-_an_independent_evaluation.pdf

  • TheEulerID
  • 6 days ago

EulerID - Oxera report was for a specific purpose of renewing SA33671 and it failed to convince EU, so a new state aid measure was needed, but it provides some references but is not definitive.

Andrew called out for BDUK 139 premises average times 30k cabinets which is less than BDUK reported superfast. He indicated he had an number for VM overbuild. Reporting it may help build that Capital Deferral further so UK rural can benefit more fully as originally intended.

  • ValueforMoney
  • 5 days ago

@VFM so are you saying that only some of the findings in the Oxera report should be used and some disregarded even though they are still presented as facts but our outside the perceived specific purpose of the report? I can see that specifics and sweeping generalities should be considered by the informed reader as non-representative of the whole, but anything else begins to lean towards reader selection of only those items that support their argument

  • Gadget
  • 5 days ago

@VFM so are you saying that only some of the findings in the Oxera report should be used and some disregarded even though they are still presented as facts but our outside the perceived specific purpose of the report? I can see that specifics and sweeping generalities should be considered by the informed reader as non-representative of the whole, but anything else begins to lean towards reader selection of only those items that support their argument

  • Gadget
  • 5 days ago

On our figures looks like BDUK has delivered 4.42m million premises, and that includes allowance made for 1.1 million cable premises. Now some of those cable areas were rolled out to after the BDUK had previously gone there, so the 4.42m is another 20k to 30k higher, and you can then into a 1% variation which needs not much of a change to vanish.

If you want to declare failure with data that is just 1% out, then feel free but you'd get no support from me.

  • andrew
  • thinkbroadband staff
  • 5 days ago

The 30,085 is VDSL2 cabinets, FTTP has its structure but I don't measure numbers of splitters (why would I?) so a simplistic BDUK figure divided by 30,085 cabinets gives 151 premises per cabinet, and that is not correct due to things like FTTP but shows we are all in the ballpark.

  • andrew
  • thinkbroadband staff
  • 5 days ago

Gadget - My opinon, Oxera was a report for the EU so the state aid measure could be extended. That was its only purpose. It was written only for that purpose. I support getting the job done, and the folk doing the work need applauding. But Andrew has now referenced a figure 1.1m of VM premises which may have been impacted by a state aid measure which referenced 'rural' at least 19 times and had clauses stating 'urban' needed a separate measure.
The state aid breeches can be set aside provided there is a plan to re-invest monies pursuing the original goal in rural.

  • ValueforMoney
  • 2 days ago

Gadget - Note; Oxera report did a useful service is adding detail and outlining the capital reconciliation process. They formally reported on the existence of the Investment Accounts where BT capital contributions may end being paid. These are not being reported on. This was helpful at a time when BT and indeed ex-Minister Vaizey were telling the CMS committee BT has been paid.

  • ValueforMoney
  • 2 days ago

Andrew; Thanks for sharing the 1.1m VM number. What is your confidence level? If we take 4.5m (BDUK) (151x30k cabs) and add the 1.1m and make some allowance for the distance based limitations of copper gain you will quickly exceed 6m premises 'passed' in the manner BT counts 18-19m premises.
This is helpful, as it better shows how the gap funding principal should be applied, and how the BT Capital Deferral needs to continue growing. The latter assumes there are limited balances in the LA Investment Accounts which continue to go unreported. The latter is a little unfair on BT.

  • ValueforMoney
  • 2 days ago

@ValueForMoney Confidence is that I've gone public with it. It seems many comments later that the BIG difference is you've been looking at passed at any speed rather than passed at superfast speeds.

If you want to fight the rural/urban I suggest you learn about the imprecision in defining those two words first. The public don't care, they just want better broadband.

  • andrew
  • thinkbroadband staff
  • 1 day ago

Would not believe any of it. Here in a rural part of NI the database on openreach and dsl checker is so far out of sync with reality how would they know what the figures are. I know of property's around here where people are told the can't get FTTC but can, and property's where the opposite is also the case. Hit and miss all over this area.

  • billdornan
  • 1 day ago

@VFM

Whilst BDUK will not count premises passed which already have cable access, any "overlap premises" which take-up VDSL2 are counted for the purposes of gainshare money so do return capital to public funds.

As for the way BDUK rolled out, that comes from the political priorities set out from the beginning to cover largest number of non-SF locations as fast as possible with the lowest cost. Somebody in a go-slow spot probably doesn't much care if they are officially "urban" or "rural".

Clearly the cabinet overlap with VM territory has worked to OR's advantage (and it's ISP customers).

  • TheEulerID
  • 1 day ago

@billdornan Well aware that even official things appear very odd in Northern Ireland for ordering, and this is reflected in our data with the largest gap between VDSL2 at any speed and superfast figures.

  • andrew
  • thinkbroadband staff
  • 1 day ago

5Mbps, 10Mbps, 100Mbps ........... that's all La La Land to me. Please can I have a steady 1.5Mbps, 2.0Mbps would be a dream! Speeds get so poor, that even a dial up service is tempting. Ever since Open Reach decided to upgrade exchanges to fibre, us poor schmuks are stuck on ADSL (not ADSL2) without any chance to improve this basic service. It was a joke years ago that we wouldn't see these superfast connections in our lifetime, it's not a joke anymore.

  • biglwcus
  • 1 day ago

TheEulerID - The Idea that VM overbuild of 1.1m contributes to clawback I am sure is a relief to VM! To date it has not been presented as such and this thanks to Andrew is the first acknowledgment of VM overbuild of such a scale.
WWWombat has confirmed BT get paid for the build, there is not deduction for overbuilding VM, and you say the numbers are included in the clawback calculation.
We need transparency so the intended benefit for once in a generation upgrades in rural are not impaired while BT's investment is correctly stated in the WLA price control.

  • ValueforMoney
  • 1 day ago

The overbuild aspect has not been hidden before, but its not being pulled out into figures of one syllable so that people accept the figure, or put another they don't need to use a calculator.

NOTE: Before you go quoting the 1.1 million everywhere you MUST add the caveat "Now some of those cable areas were rolled out to after the BDUK had previously gone there" - it might be 100k or more of the overlap.

Oh and Virgin Media while mainly urban is not always.

  • andrew
  • thinkbroadband staff
  • 1 day ago

@VFM
Given you've been so emphatic about the alleged misconduct related to this contract for some considerable time, I'm surprised that you keep asking so many detailed questions related to numbers, technology, contract terms etc. Surely these are all covered in your report?

Or is that really just a dodgy dossier, full of estimates, exprapolation and guesswork, supported by misinterpretation of accountancy and technology?

  • New_Londoner
  • 1 day ago

Extrapolation not exprapolation!

  • New_Londoner
  • 1 day ago

New-Londoner We must have coffee sometime. This is about a policy of predation, where BT Group attempts to eat that which is intended for Openreach engineering activity. Reversing it is tricky. You will have read the Vodafone dossier. You can download the ICBAN dossier at the same Ofcom site. None of these have a legal intent, only one of reversing an abuse of power insuring the monies end up where intended.
How is the £333m of the £463m Capital Deferral going to be released before 2023? That is a lot of engineering work! How we do locate the sum of the Investment account balances?

  • ValueforMoney
  • about 23 hours ago

New_Londoner - accounting policy question. If 50% of FTTC costs are commons costs allocations as per WLA analysis, what is the thinking around allocating costs inversely to the distance served or the 'speed'. So 80/20Mbps FTTC is allocated more costs that a 40/10. It is expedient, but deeply peculiar as BT are now allocating costs based on the specific distance shortcomings of copper gain technology. Most of the link with cost is lost and the cost allocation becomes arbitrary. How does such a policy support a more fibre plan?

  • ValueforMoney
  • about 23 hours ago

New_Londoner - accounting policy question. If 50% of FTTC costs are commons costs allocations as per WLA analysis, what is the thinking around allocating costs inversely to the distance served or the 'speed'. So 80/20Mbps FTTC is allocated more costs that a 40/10. It is expedient, but deeply peculiar as BT are now allocating costs based on the specific distance shortcomings of copper gain technology. Most of the link with cost is lost and the cost allocation becomes arbitrary. How does such a policy support a more fibre plan?

  • ValueforMoney
  • about 23 hours ago

New_Londoner Another accounting question. Now 1.1m premise of VM is beginning to be acknowledged as over-built with subsidies meant for rural, and BT has been fully paid, what adjustments will occur? How does gap funding work in these cases? What is BT's capital contribution to passing these 1.1m premises? What system size is used to calculate the clawback?
Your insight will be most welcome.

  • ValueforMoney
  • about 23 hours ago

@VFM - to quote Andrew above "On our figures looks like BDUK has delivered 4.42m million premises, and that includes allowance made for 1.1 million cable premises. Now some of those cable areas were rolled out to after the BDUK had previously gone there"

So any decision or subsequent un-notified build made after the OMR is not (AFAIK) considered "overbuild" AS PART OF THE CONTRACT(my emphasis), neither is any area where the local authority has discounted input to the contract OMR phase.

  • Gadget
  • about 22 hours ago

Hi VFM.
Just sit back and think what I have said in the past over distance from the Cab where the recoded results will hit the two targets ( 24 SCC and EU 30).

  • Blackmamba
  • about 21 hours ago

"Now 1.1m premise of VM is beginning to be acknowledged as over-built with subsidies meant for rural"

So where in the DCMS figures of 4.5m premises delivered with £563m of Westminster money does the 1.1 million fit in.

It looks like they have already taken this into account in their figures.

  • andrew
  • thinkbroadband staff
  • about 21 hours ago

A; Some adjustment might be needed. For example;
1.) DCMS report 4.5m + a proportion of the 1.1m lowering the unit cost per premise passed while more accurately reporting the total effort.
2.) BT confirm their capital contribution or accrual for the overbuild.
3.) BT confirm or adjust the clawback measure or any additional measure to pay for what should be part of a commercial build.
It may be taken into account or it may not. In April 2015 the FD of BT said clawback would not exceed tens of millions. Overbuild of this level has been denied until now, so more change may be possible.

  • ValueforMoney
  • about 20 hours ago

@Gadget 1.1m Premise Overbuild in a final third programme! Reduce it by half and it is still too large.
Given Ofcom are crediting BT with £100 per premise passed (Annex 8 Fair Bet analysis) in the proposed WLA price control then all they need to do is confirm for all the overbuild BT has paid the capital contribution to support the proposed price control.
Do you want BT Group to game what should be invested in Openreach? It is not even in BT shareholder or BT customer interest for there to be any ambiguity on the matter.

  • ValueforMoney
  • about 20 hours ago

"It may be taken into account or it may not."

How can you say that? Do you actually think the APPROX 1.1m is to be subtracted from the 4.5m that DCMS has declared?

It HAS been taken into account, if there was a 1 million disparity between DCMS and my figures I'd be busy talking to the newspapers to expose a scandal.

Only way to be sure about things at this point is cease all roll-outs, employ a lot of people to hand check invoices + data and then report back.

  • andrew
  • thinkbroadband staff
  • about 20 hours ago

As for final third, if it was a final third product the footprint would be around ten million premises and target would be 100%, just remember final third was a handy soundbite for politicians, and also highlights that the project is not just rural, since by one measure just 10% of the UK is rural, by ONS definitions its 20-23% of the UK.

  • andrew
  • thinkbroadband staff
  • about 20 hours ago

@VFM

Yes, the way the political objectives for maximum coverage at the fastest speed with the minimum cost have meant that there has been inevitable overbuild on VM areas due to the cabinet overlap. True, some of those more urban areas would have eventually been incorporated in commercial roll-outs, but the politicians wanted to accelerate the process, and what arose was an inevitable results of their objectives.

Meanwhile, somewhere around 4.5m premises have potential access to SF and the fast takeup & reduced costs mean there's a whole lot of money still to invest.

  • TheEulerID
  • about 15 hours ago

Andrew, the 1.1m VM overbuild which is in addition to the 4.5m should mean an additional capital contribution from BT of c£75x1.1m or £82.5m if the gap funding principle is being applied.
This is in addition to the c£350m identified by the NAO, whose status is yet to be reported on.
Clawback from these additional customers should also be identifiable in the process.

  • ValueforMoney
  • 19 minutes ago

Post a comment

Login Register