The Connecting Devon and Somerset BDUK project (CDS) has seen a fair bit of arguing and complaints about the project not delivering and now that 2016 has ended and we have had a week or two to look at services delivered in December 2016 and are now able to share our analysis of what has happened in the two counties.
The Phase I project had a core contract goal of 277,000 premises benefiting from superfast broadband at speeds over 24 Mbps, and our analysis indicates that BT are 20,000 premises short of this target. Though as we are still seeing cabinets going live this may reduce some more over the next few weeks, and there are FTTP areas in our data which appear to be under active build and others waiting for the first work to start. A Council meeting on 10th January had BT predicting it would fall short of the 277,000 target by 1,700 premises. This report highlights that there is FTTP left to build and that issues such as wayleaves and highway notifications are one of the reasons for the delay, some of the missing premises may also be accounted for by Exchange Only (EO) conversions and we have yet to see evidence of this happening.
One feature of the last couple of months has been infill cabinets appearing along with more FTTP, so its a case of wait and see how much more of this is to be delivered. We believe that some tactical deployment of vectoring and changes to the Openreach DLM (namely allowing a lower target noise margin of 3dB) may be deployed in parts of Devon and Somerset to improve the reach and speed of VDSL2. We will be analysing this and if this makes a significant change to the speeds achieved will endeavour to create additional nuances in the model we utilise, we could model the effect based on the theoretical effects but we prefer to use massed observation to inform decisions used in the model.
| thinkbroadband analysis of Superfast Broadband Coverage for Connecting Devon and Somerset as of 13th January 2017 | |||||||
|---|---|---|---|---|---|---|---|
| Area | % fibre based | % superfast Over 24 Mbps (change since August 2013) |
% superfast 30 Mbps or faster |
% Ultrafast | % Openreach FTTP | % Under 2 Mbps USC |
% Under 10 Mbps USO |
| Combined Area | 93.3% | 86.9% (+33.5) | 85.8% | 27.6% | 0.67% | 1.7% | 6.7% |
| CDS Delivery (*) | 99.6% | 86.6% | 84.3% | 4.4% | 2% | 6.1% | 10.1% |
| Devon | 91.6% | 82.1% (+44.6) | 80.8% | 19.4% | 1.26% | 3% | 10.3% |
| Torbay | 95.9% | 94.9% (+17.2) | 94.3% | 46.5% | 0% | 0% | 0.5% |
| Plymouth | 98.5% | 98.3% (+1.4) | 98.1% | 90.8% | 0.22% | 0 | 0.5% |
| Somerset | 92.7% | 84.3% (+43.7) | 82.9% | 4.9% | 0.39% | 1.9% | 8.7% |
| North Somerset | 93.4% | 89.5% (+30.3) | 88.1% | 21.4% | 0.34% | 0.3% | 2.7% |
| Bath and North East Somerset | 93% | 89.5% (+16.3) | 88.7% | 31.9% | 0.57% | 0.5% | 3.8% |
(*) The change figure since August 2013 is not shown as we don’t track that data set with the same historical access as we do with the local authorities, though as we believe the first cabinet went live under the project in the summer of 2013 the figure could be shown as 0%.
The combined area is the six local authority areas combined into a single figure and thus represents the coverage both commercial and gap funded. The CDS Delivery figure represents those cabinets and FTTP areas that we believe have been delivered via the CDS Phase I project, there is likely to be more FTTP in build and there may be additional cabinets we have not become aware of that are to be delivered yet.
| thinkbroadband analysis of Superfast Broadband Coverage in Devon and Somerset Constituencies as of 13th January 2017 In ascending order of superfast coverage |
|||||||
|---|---|---|---|---|---|---|---|
| Area | % fibre based | % superfast Over 24 Mbps (change since August 2013) |
% superfast 30 Mbps or faster |
% Ultrafast | % Openreach FTTP | % Under 2 Mbps USC |
% Under 10 Mbps USO |
| Central Devon | 89.7% | 72.3% (+67.2) | 70.7% | 1.4% | 1.44% | 5.7% | 18.7% |
| Torridge & West Devon | 86.5% | 72.5% (+44.9) | 70.9% | 0.5% | 0.48% | 5.7% | 18.8% |
| Tiverton and Honiton | 88.2% | 74.8% (+60) | 72.7% | 0.3% | 0.27% | 4.2 | 13.9% |
| Somerton and Frome | 91.1% | 78.3% (+44) | 76.7% | 23.3% | 0.15% | 3.6% | 14.4% |
| North East Somerset | 89.7% | 83.1% (+22.2) | 81.8% | 26.3% | 0.15% | 1.1% | 7.3% |
| Totnes | 92.2% | 83.6% (+37.1) | 82.1% | 7.9% | 0% | 2% | 7.3% |
| North Devon | 92.5% | 83.7% (+48.9) | 82% | 0% | 0% | 3.5% | 10.1% |
| Wells | 92.7% | 84% (+34.7) | 82.9% | 0% | 0.02% | 1.2% | 7.6% |
| Bridgwater and West Somerset | 92.5% | 84.2% (+42.3) | 83.1% | 0.8% | 0.74% | 2.3% | 8.8% |
| Taunton Deane | 90.9% | 84.9% (+23.8) | 83.1% | 0.8% | 0.73% | 1.5% | 7.6% |
| East Devon | 92.8% | 86.8% (+34.8) | 85.3% | 7.8% | 0.47% | 1.3% | 5.2% |
| Newton Abbot | 94% | 88.5% (+21.4) | 87.8% | 41.9% | 0.13% | 1.3% | 3.5% |
| Weston-Super-Mare | 92.2% | 88.7% (+30) | 87.2% | 1.3% | 0.49% | 0.2% | 1.9% |
| Yeovil | 96.2% | 89.8% (+50.9) | 87.9% | 0.2% | 0.2% | 0.9% | 4.5% |
| North Somerset | 94.7% | 90.3% (+24.9) | 89.1% | 43.5% | 0.06% | 0.4% | 3.6% |
| South West Devon | 95.3% | 91.2% (+15.8) | 90.3% | 58.3% | 0.57% | 0.5% | 5.2% |
| Bath | 95.7% | 95.2% (+7.3) | 94.7% | 36.8% | 0.58% | 0% | 0.5% |
| Torbay | 96.3% | 95.7% (+11.4) | 95.4% | 55.9% | 0% | 0% | 0.3 |
| Exeter | 97.8% | 97.6% (+1.6) | 97.6% | 93% | 6.4% | 0% | 0.7% |
| Plymouth Sutton and Devonport | 98.2% | 98% (+0.6) | 98% | 93.2% | 0% | 0% | 0.6% |
| Plymouth Moor View | 99.1% | 98.9% (+0.6) | 98.9% | 89.4% | 0.08% | 0.1% | 0.4% |
August 2013 was chosen as the comparison date for the constituencies as this was the date we believe the project started to deliver improvements and while no specific constituency level targets are set in any BDUK projects, they are a nice sub-division or roughly equal premise numbers to make for comparison across areas, and for those not benefiting yet how much pestering they need to do for their MP.
The Phase II contracts awarded to Gigaclear should add around 3.8% to the superfast and ultrafast totals, and if delivered to the right areas have the ability to eliminate all the sub 2 Mbps areas. The AirBand contract covering Exmoor and Dartmoor which is meant to bring superfast broadband to 5,800 premises and with delivery still underway and take-up sitting at around 4.4% it will take sometime for us to understand what AirBand is delivering and exactly which areas are benefiting or not (NOTE: This is not an indication of a bias against fixed wireless, but is identical to how we are approaching vectoring and a 3dB margin from Openreah, i.e. gather evidence to see if the product label is correct).
Update Sunday 15th January To highlight that Keri Dunton while presenting at the council meeting on 10th January, indicated that the figure of 1,700 that BT were reporting as the likely shortfall had grown to 11,000. Keri Dunton does go on to say that since Christmas another 1,000 have been knocked of the total and as she talked on 10th the shortfall was 10,000, hence why we believe that it is likely the 1,700 shortfall refers to a March 2017 estimate and the 11,000 (or 10,000) short is the current state of play. With we believe around 30 cabinets still in the process of going live from the project and the FTTP that is in build we are expecting the shortfall to drop even more, hence why the original item used the 1,700 shortfall and still does, rather than muddle things with ever changing figures.
Update Monday 23rd January We are constantly changing the data set behind the model, but have concentrated on Devon and Somerset in the last week to ensure our model is as up to date as possible both in terms of coverage and number of premises. This means we can report a shortfall of 12,500 premises on the phase I contract target at this time, which seems to confirm the 11,000 and 10,000 figures.
Test your broadband speed
Follow us on X for the latest broadband news 
You’ve put the increases in the second table, but only put the combined total in the first table. Is there a way to get the increase in combined coverage?
Nice to see you living up to the quote from September 2013:
“For those convinced that the final speed targets will never be met, and be simply announced as job done, you can be sure we will be keeping an eye on the data from our speed test and tracking the changes for many years to come so that we can look back at the pattern of changes and equate the improvements to coverage and take-up levels.”
Local authority level change figures are in the table now.
My cabinet went live nov 30th placed order 1st december and still shows as awaiting activation on the cds site wwstal cabinet 4 connection was full 80/20 now dropped below estimate 80/64.7 down due to capacity issues
https://www.connectingdevonandsomerset.co.uk/where-when/?cab=%7BWWSTAL%7D%7BP4%7D info on the cabinet concerned above
We have it mapped as live https://labs.thinkbroadband.com/local/broadband-map#13/50.9739/-2.3936/openreach/
Our postcode lookup is usually even more up to date than the map, as the tile rendering is not updated as often as postcode lookup/
thanks Andrew the results on the maps are mainly mine on plusnet and o2 for my postcode
Oddity noticed:
As you indicated that Gigaclear will only add 3.8%, I checked the tender… Their 4 lots apparently have 53k black premises, and they’ll reach 35k of them. A good proportion.
But then I realised the total of NGA black premises from the 6 lots is only 77k. That’s 7.5% … suggesting BT and Airband will have reached 92.5%.
That’s much higher than the coverage you are suggesting!
I went back to look at the documents used for the tender in 2015. That said there were 178,000 NGA black premises.
That’s over 100,000 difference!
Did you account for the lot that has not been announced yet?
The 277,000 is the contracted amount for phase I superfast, so should be the set in stone figure.
As an ‘intelligent’ layman these statistics confuse me. I am paying Zen for superfast broadband from cabinet Pilton2. With approx. 1kM of ‘copper’, my speed is around 17Mbps. I am told this is due to a ‘poor quality, life expired, aluminium’ cable. These situations do not seem to be reflected in the over-optimistic published statistics. Why?
@newhouse
How can you tell whether your experience (1 in 7 million) is reflected in the stats, or not? How can 1 result confirm a bias of optimism or pessimism?
Some stats are based on aggregate knowledge of “electrical length” rather than physical length. This attenuation *does* take into account the presence of aluminium vs copper.
TBB stats, I think, come from a pessimistic set of distance stats, weighted by actual speed test results. Those actual speeds obviously take into account existing aluminium.
@Andrew
Yes – 53k black premises in the 4 lots announced so far; 77k black premises in all 6 lots.
The 6 lots don’t include Dartmoor or Exmoor, which might hold some of the difference. I’ll have to re-check the scope of the old consultation, but I think the moors had already been pulled out for the separate FWA project.
@Newhouse One way to know is look up what we say for your postcode at https://labs.thinkbroadband.com/local/postcode-search which shows speed test results and what we think is possible.
You can see a visual summary at https://labs.thinkbroadband.com/local/broadband-map#14/51.1622/-2.5883/openreach/ where we highlight postcodes based on what we believe is possible
An example of the inability of CDS to manage their rollout information is Edingworth 4, live in September, shown as ‘Coming Soon’.
Is Devon the toughest county apart from H&I – a lot of 1-3 cabs with low numbers and long distances? Salute!
Tough counties?
Would have thought you could add North Yorkshire and Lincolnshire to that. Low density population too … but Devon has the “luxury” of having a large proportion of its population in Plymouth, Exeter and Torquay/Torbay.
North Yorkshire has 2 national parks in its remit too, matching Devon.
I tried to correlate the IA numbers, but I realised I should have used “white” in earlier comments, not “black”. Whoops!
That changes the whole basis of numbers that I was looking at. There’s still an oddity, but less so
…
The old OMR (early 2015) showed 119k premises as NGA white, and a further 31k labelled “at risk”.
6k of the NGA white went into the Airband project.
2k (presumably NGA white) went into a Gigaclear project for the Blackdown Hills), so are presumably lost from the new IA.
The latest tender had 77k NGA white premises:
– 53k in the 4 Gigaclear lots
– 24k in the other 2 lots.
That means 34k of the original NGA white premises have gone missing.
The tender also lists 76k premises as “at risk”, which is 45k more than previously.
deveon having to be next to cornwall which has very extensive Fibre coverage due to its previous procurement !!! so that may have something to do with it – Keeping up with the Jones’s