Broadband Delivery UK and the 1 Million Virgin Media Premises Overbuild

New research has estimated that between 950,000 and 1.1 million premises, which were already being catered for by Virgin Media’s cable broadband network, have been overbuilt by networks (e.g. Openreach (BT)) supported through the Government’s Broadband Delivery UK project.

At present the national BDUK programme, which is fuelled by £1.6bn+ of public investment, is aiming to help extend the reach of “superfast broadband” (24Mbps+) networks to 95% of the United Kingdom by the end of 2017 (they’re currently at nearly 94%) and possibly 98% by around 2020.

It’s important to recognise that the first c.70-75% of UK coverage (estimate) was largely achieved by commercial deployments via Openreach (BT) and Virgin Media (plus some smaller altnets), while most of the final 25-30% is benefiting from the work being supported by BDUK and their suppliers (Openreach, Gigaclear, Airband etc.).

Under EU state aid rules this public subsidy shouldn’t be used in areas where an existing superfast network is already present – assuming the network has identified itself as part of a local Open Market Review (OMR) or other accepted process – because it would waste tax payers money and risk distorting competition.

Over the years we have of course seen examples where the OMR process doesn’t work quite as intended or councils are slow to adapt when new network operators deploy into the area, which can result in a clash between those supported by BDUK and those using purely private investment. However that’s another story and today we’re focused on the known Virgin Media overbuild.

The EU Position on UK State Aid for Broadband

Only projects that are confined to “NGA white areas“, where no operator is willing to invest in NGA infrastructure without State aid in the next three years, are eligible to receive aid pursuant to the measure under examination.

What is more, where a broadband network already exists, the measure requires that a “step change” be achieved; the public intervention must result in significantly better broadband capacity and thus service availability and the selected bidder must carry out significant new investments in the existing broadband networks.

Thus, in line with paragraph 51 of Broadband Guidelines, such an open infrastructure brings significant new capabilities to the market as it provides further multiplication of distribution nodes, shortens “the last mile” to end-users and allows for competition between operators which will provide access to final customers.

In the rules there is some recognition that in complicated network environments a limited degree of overbuild may exist and even be unavoidable (e.g. at the edges of a network and on wireless platforms). On top of that any new state aid supported network must deliver a “step change” in performance (e.g. the UK requires that “download speeds have to be at least doubled … and substantially higher upload speeds need to be provided“).

On this point we should highlight that the majority of BDUK supported deployments have involved the use of ‘up to’ 80Mbps capable FTTC (VDSL2) lines, which are of course a lot slower than the ultrafast cable connections that Virgin Media typically sell (currently up to 350Mbps).

Despite this it’s long been known that BDUK supported projects have overbuilt some of Virgin Media’s network, which in the real-world would be difficult to avoid without leaving an unacceptably large number of premises isolated (stuck on slow connections) between the two operators. For example, Openreach may build a new FTTC street cabinet that serves 200 lines but find that 50 of those serve premises within Virgin’s area.

Back in 2014 Virgin’s Head of Public Affairs, Daniel Butler, told ISPreview.co.uk that they were working with councils to ensure proper use of state aid (here). “There is no need or value for the taxpayer in using public funds to build networks where they already exist,” said Butler. Fast forward to the end of 2015 and Virgin began calling on the EU to tackle the “unacceptable risk of overbuild of existing networks” (here), but little happened.

Scale of the Overbuild

One of the biggest question marks in this whole affair has of course been the issue of scale and related funding. Just how many of Virgin Media’s premises have actually been overbuilt by BDUK supported deployments and precisely how is the allocation of state aid impacted when such a situation occurs? A lot of assumptions have been made about this over the years and many are often incorrect, so we’ve set about trying to clarify.

Andrew Ferguson of Thinkbroadband, which is just about the only reliable and independent source of such detailed coverage data, was able to harness their vast database and analysis system to produce a figure for the overlap. Overall the prediction is that the overbuild impacts around 950,000 to 1.1 million premises across the UK (we picked c.1 million in the headline as a middle ground), although more work on this needs to be done and even the coverage data provided by ISPs won’t always be 100% accurate.

Andrew Ferguson told ISPreview.co.uk:

“Analysis of the broadband roll-outs is a very dynamic time consuming job and while it is possible to roll the clock back for even deeper analysis, without access to all the corresponding Open Market Review data and invoices the local authorities are paying, it is very difficult to pin down precise figures and it is highly likely that different authorities are approaching the overlap issue differently while remaining within the BDUK framework.

The time and effort required to pin down the precise costing for an individual premise that was or was not overlapped when scaled up to a project the size of the BDUK scheme has the potential to keep a number of people employed full time and thus waste money that could be spent on delivering more better broadband.

At the end of the day the BDUK project is not perfect, but how many other Westminster/Local Government projects ever go through phases where savings are being reused to deliver more?”

The c.1 million figure might actually be higher, although some premises have been excluded for key reasons. For example, the figure excludes existing BDUK areas that have later been overbuilt by Virgin Media (c.100,000) as these are less contentious and another c.100,000 that are not superfast from FTTC (VDSL2) but are from cable etc. As the deployment is on-going then the figures remain fluid and thus perfect precision should not be expected.

A Question of Public Funding

Now to the crucial bit. In August 2017 BDUK reported that some that their public investment had so far helped 4,551,226 extra premises across the United Kingdom to be put within reach of a fixed “superfast broadband” (24Mbps+) network (here), although what a lot of people don’t realise is that this figure already excludes the c.1 million overbuild of Virgin Media (otherwise it would be closer to 5.5m) because those aren’t able to use the public subsidy.

A BDUK Spokesperson told ISPreview.co.uk:

“Only infrastructure that is required to target non-superfast premises is eligible for subsidy. If there is overspill to other premises then these are not eligible for subsidy, and they are not counted towards the contractual superfast delivery targets. However, these premises are included in the gainshare calculation, so we benefit from any additional take-up resulting from them.”

Simple enough, although it does raise an obvious question with respect to how you cope with the cost vs public subsidy of a street cabinet that overlaps into Virgin Media’s estate. For example, if a cabinet that gets built and put live costs £30,000 and serves 200 premises, then using a simplified approach – plus no overlap with Virgin – we could put that as a basic £150 per premise intervention cost.

However if the situation is changed and 50 of those 200 premises are already within Virgin Media’s network, then the cabinet itself will still cost the same to deploy (i.e. how does the public investment get separated?). Once again we put this to BDUK and were informed that it’s handled via the gap funding calculation.

A BDUK Spokesperson said:

“In the example you’ve given, the supplier could only claim gap funding for the 150 premises with no overlap. Either the supplier would need to seek more funding for each of the 150 premises (so the subsidy would be spread across 150 rather than 200) or alternatively the premises would not be included in the bid at all.

Note that the supplier also has to provide their own funding – so in the example, they would have to pay their normal commercial contribution for all 200 premises passed.”

No doubt some people will still find fault with the approach that BDUK has taken, particularly if we consider that the new infrastructure might not have been built into a Virgin Media area in the first place without such support, although the complicated reality of modern networks makes it very difficult to completely avoid overbuild. At least the public funding is being balanced to restrict its use in such situations.

NOTE: Gainshare or Clawback is the contract mechanism that returns public funding as take-up of the service rises, which supports reinvestment so that more areas can be upgraded with superfast broadband. This is the reason why BDUK has recently been able to set a higher coverage aspiration of 98% for around 2020 (here).