Broadband News

Ofcom consults on price impact of BT USO proposal

The state of broadband roll-outs and what speeds people get across the UK is set to remain a major data collection and analysis for a few years more yet, with the Universal Service Obligation becoming the big driver in public policy once the existing BDUK projects finish, and subsequent to this we envisage tracking the success or not of and numerous full fibre roll-outs will be a big thing too.

Ofcom has arrived at an estimate of 785,000 premises in 2021 falling within the specification of the USO where that specification is based on:

  • BT proposal download speed of a minimum 10 Mbps
  • BT proposal upload speed of a minimum of 1 Mbps
  • BT proposal Delays due to contention and latency to be minimised
  • BT proposal Minimal use of satellite (estimate 0.3%) with fixed line tech covering 99% of UK premises

The 785,000 is only slightly below the premise count of 888,860  we announced for premises below 10 Mbps download speeds the other day,  but based on the BT proposal most people would say that ADSL2+ is excluded since getting above 1 Mbps connection speeds for the upload is relatively rare, and additionally offers no obvious upgrade path to faster options in a few years. Therefore if we were to add an additional analysis model to our data where we want a minimum 10 Mbps download speed and upload is above 1 Mbps our USO model number would increase, perhaps we should call this our USO BT Proposal figure.

There will be some premises that will not receive 24 Mbit/s but will have access to speeds that meet the proposed service specification of at least 10 Mbit/s. Based on BDUK data, in June 2016 around 10%, or 2.9m premises, did not have access to 24 Mbit/s. Comparing this to our Connected Nations data for the same time, where 2.6m premises did not have access to the 10 Mbit/s specification in BT’s proposal, this implies that around 90% of those that did not have access to 24 Mbit/s also did not have access to 10 Mbit/s. We have assumed this ratio of availability of 24 Mbit/s and 10 Mbit/s remains broadly constant, so based on our forecast that by 2020/21 873,000 premises would not have access to 24 Mbit/s, at this time some 785,000 premises would also not have access to 10 Mbit/s and would therefore be qualifying premises in March 2021.

Ofcom method for determining 2021 USO intervention premise count

How Ofcom arrived at the 785,000 premise figure is interesting, since it relies on a projection from the coverage data around April/May 2016 and the assertion that the ratio of sub 10 Mbps premises will remain constant. It is possible that it will, but based on what we see from weekly updates to our coverage data we do not believe it will. The phase II projects are doing a lot more infill than the phase I project, i.e. often a further kilometre closer to premises than the original VDSL2 cabinets, so the profile of speeds people had previously is different and the areas where native GEA-FTTP and Gigaclear FTTP has been deployed is very different to the VDSL2 of the phase I BDUK projects. One missing element from our quote is an estimated premise count of 29.09 million premises across the UK in 2021. The 2.6 million premises without access to the 10 Mbit specification we assume takes into account the 1 Mbps upload figure, since the Ofcom Connection Nations report reported a figure of 1.4 million premises (released Dec 2016, data Apr/May 2016, we recorded a figure of 1.32 million premises for May 2016). So time to fire up some more analysis and submit our findings officially to Ofcom rather than rely on them reading and digesting all the data we make public.

What we do know is that if the BDUK and commercial roll-outs do hit 97% superfast coverage and the premise count is 29.09 million the worst case figure needing USO work will be 872,000 premises.

Some snippets that need to be understood as we now look at the cost of the BT USO proposal to UK broadband customers are as follows:

    • A Committed Information Rate of 1.5 Mbps - this is pretty much mid range for what providers currently budget for when scaling their backhaul and core networks.
    • The cost recovery for the fixed wireless access and satellite services that are expected to comprise the final 1% do not feature in the cost recovery, satellite services are expected to sell at current retail prices and fixed wireless (FWA) is presumed to have its own cost recovery built in.
    • Existing wholesale access requirements do not apply to BT for fixed wireless access and satellite services they sell.
    • The proposed cost recovery would not affect those who take a standard telephone line with no broadband unless it was via a full LLU service (which is rare), most voice only sales seem to be via the WLR product range.

The impact on the annual charges is therefore summarised as:

Range and (central estimate)Proposals for Annual Charges (£ - nominal)
Additional cost for network expansion £0.23 to £1.57
£0.71 to £3.80
£1.14 to £5.89

Ofcom is proposing that the additional cost would be added to MPF services if bought on their own, MPF with GEA services (only applied as one charge) and WLR with GEA. The WLR only and WLR with ADSL2+ (SMPF) customers appear to not be impacted.

LLU charge control proposals - MPF Rental
Central estimateAnnual charge with effect from 1st July 2017 (£)Proposals for Annual Charges (£ - nominal)
March 2017 proposals (central estimate) £85.29 £83.50 £82.28 £81.98
Additional cost for network expansion   £0.39 £1.19 £1.93
Illustrative impact   £83.89 £83.47 £83.91
GEA charge control proposals - GEA 40/10 rental when taken without MPF
Central estimateCurrent annual charge at 31 March 2017 (£)Proposals for Annual Charges (£ - nominal)
March 2017 proposals (central estimate) £88.80 £66.28 £57.00 £52.77
Additional cost for network expansion   £0.39 £1.19 £1.93
Illustrative impact   £66.67 £58.19 £54.70

Once combined with the already proposed price cuts the extra cost of the USO work if the BT proposal is adopted is lost and people at least at the wholesale level would still be looking at reduced prices, what happens at the retail level is another different ball game of course.

The Ofcom consultation runs until 27th September, and we hope that before then we will be able to work in the 1 Mbps minimum upload speed requirement into a new set of USO figures and get some idea of how this is reducing over time.

There are always pros and cons to different proposals but with time marching on if there are operators with large scale proposals now is the time to get their proposals on the table e.g. costing a full fibre USO for 99% and a similar FWA/Satellite mix for those where costs spiral out of control unless people are willing to undertake a work themselves.

Update 7:50pm The first run for a BT USO premises figure where we specify a minimum upload speed of over 1 Mbps, i.e. will exclude people who can get ADSL2+ at 16 Mbps but are only getting 0.8 Mbps upload speeds has revealed some interesting snippets so we are sharing a few insights:

  • Staffordshire currently at 92.8% superfast coverage over 24 Mbps, sees previous USO figure jump from 3.6% to 4% so now around 13,801 premises.
  • Pembrokeshire with 82.5% superfast coverage, but VDSL2 stretching to 92.3%, sees sees the old USO figure rise from 11.5% to 12% i.e. difference of 240 premises
  • City of Westminster which is 78.8% superfast, has a big jump from 0.2% to 21% i.e. with upload over 1 Mbps looking at needing to help 26,587

If ADSL2+ Annex M was allowed as part of the USO and it ticks the boxes then another variation is possible, i.e. for lines with download speeds above 16 Mbps one could say upload speeds of 1.5 Mbps to 2 Mbps are possible if Annex M is enabled. In short there are many technical variations that can play around with the premise counts, once we've done more processing we will publish a national and regional summary.


Nice work Andrew.

The update suggests that most of the /rural/ premises in the range of 10-24Mbps get that because of VDSL2, not because of ADSL2+. Is that a fair assessment?

Good to have your analysis sent on to Ofcom too ;)

  • WWWombat
  • over 2 years ago

Presumably your figures don't agree with Ofcom that 90% of sub-24Mbps premises are also sub-10Mbps?

Your /local figures suggest about 50% - 3.1% out of 6.6%

  • WWWombat
  • over 2 years ago

The answer is it seems to vary based on deployment scenarios in different areas.

18 months ago roll-outs where existing cabinets getting twins, now a much higher proportion are infill.

In Scotland lots of ADSL only exchanges have jumped from 8 Mbps maximum to 1 or 2 cabs with lots getting superfast and a tail of 10 to 24 further out and that has been something that has accelerated in the last 12 months.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

In theory can try and and roll clock back to recalculate our data for April/May 2016 again, i.e. apply 1 Mbps upload criteria but so much has changed its a day long exercise with little payback.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

Post a comment

Login Register