Over the past few months’ we’ve seen the UK Government step-up market engagement exercises as part of work to develop a framework for their £5bn investment, which aims to ensure that “every home” can access “gigabit-capable” broadband by the end of 2025. As a result we now have a better idea of what this will look like.
At present it’s not unreasonable to assume commercial deployments will ensure that speeds of 1000Mbps+ (1Gbps+) should become available to around 70% of premises by the end of 2025, although aspects like overbuild between rival networks and uncertainty over long-term roll-out plans make it difficult to be certain of where the final figure is likely to land. Official reports tend to talk of 70-80% as an expectation.
Initially most of this will come from Virgin Media’s upgrade of their existing Hybrid Fibre Coax (HFC) and Full Fibre network to the latest DOCSIS 3.1 standard (covers over 50% of UK but on-going expansion could yet take them to c.60%). After that we have Openreach’s plan to build Fibre-to-the-Premises (FTTP) out to 20 million premises by the mid-to-late 2020s (costing c.£12bn), but much of that will overbuild Virgin.
On top of that there’s a growing mass of alternative network providers, which seem to be attracting significant investment and have some equally big deployment plans (2020 Full Fibre Build Progress). Headline examples include Cityfibre’s aim for around 8 million premises after 2025 and Hyperoptic’s aspiration for 5 million by 2024, among many others. We expect a fair bit of overbuild there too.
The Gigabit Broadband Framework (F20)
Last year the Government committed £5bn to help those in the hardest to reach final 20% (F20 – mostly rural) of the UK – about 6 million premises – gain access to “gigabit-capable” services using an “outside-in” approach to deployment (here). Since then a lot of market engagements have taken place with ISPs and, as a result, we’re now starting to build-up a picture of how the Framework for this will work.
At this stage we should stress that the following details are still very TENTATIVE and will no doubt change a bit before the final framework is agreed in the very near future. But broadly speaking the Government’s (DCMS) Building Digital UK team are expected to split the F20 project into three mechanisms. Naturally some of this will be similar to existing schemes, which have been proven to help.
BDUK’s New F20 Model
1. General Demand Side Interventions (i.e. providing fibre to public sector sites / buildings, which can later also be harnessed by network providers to reach other nearby properties). This also feeds into no.2 below because vouchers are also a type of demand side intervention.
2. Voucher Scheme for Premises (i.e. effectively a continuation of the current £200m Rural Gigabit Connectivity (RGC) programme that is due to end in March 2021, albeit possibly with much more funding and bigger vouchers. Such vouchers will also become available to a moderately wider area than the original RGC scheme, albeit still based around that F20 focus). A new website will make access to these even easier.
3. Direct Supply Side Intervention (i.e. this will gobble up most of that £5bn – we think about £4bn of the total – and on the surface will be similar to the original gap funded BDUK model, including the usual state aid approval and open procurement requirements). Suppliers will thus bid on contracts to deploy gigabit broadband in areas where the market has failed to do so itself, but the detail of how this works is different.
Both 1 and 2 above should be very familiar as the Government have been using targeted investments like that for a while via their various Voucher and Local Full Fibre Networks (LFFN) programmes. But where things start to get interesting is with no. 3, which on the surface sounds very similar to BDUK’s original superfast broadband programme but there will be some key differences this time around.
By the looks of it BDUK will adopt a bundle strategy for no.3, which seems to split areas up into bundle sizes of c.3,300 premises for procurement (naturally larger operators are calling for bigger bundles, to ensure economies of scale, and longer-term contracts). Bundles will be categorised to identify the areas’ most in need and to also factor the cost of delivery (a weighted combination of 11 models is used to estimate this).
At present the draft bundles are currently composed of premises grouped by Openreach (BT) exchange area – just for design purposes – and will of course exclude locations where existing “superfast” upgrades are already taking place or planned (the use of “superfast” here may be confusing as a lot of related contracts are now actually being delivered using gigabit FTTP).
A recent report from BDUK uses an example from Staffordshire to illustrate how such bundles may look. The darkest red areas are those most likely to benefit from intervention with a high proportion of eligible premises. There are some complexities to how these are defined but, as that is subject to change, we won’t go into it now. All are based on a model of delivery that ideally takes 2-3 years to complete.
Example of Bundles in Staffordshire (Moorlands)
As a quick example. One of the bundles above, a little further to the North West of Ashbourne, is a red area and comprised of about 3,700 premises, which includes 3,250 that are deemed to be within the final 20% (i.e. 88% of the total). But by the sounds of it bundles may also come in different sizes (some are 1,000 premises and a few have already gone up to 8,000).
The F20 programme is, as previously reported, also taking a somewhat technology neutral approach (e.g. FTTP, wireless broadband / 5G or hybrid fibre coax are all potentially viable as delivery solutions). However, this doesn’t mean to say that all premises identified by the authority for intervention will receive 1Gbps speeds, indeed like past contracts there’s a lower real-world requirement in the detail.
F20 Speed Expectations
All premises identified by the Authority in the Intervention Area must be provided with:
○ Gigabit Capable Connectivity with a normally available download speed of at least 500Mbps and upload speed of 200Mbps.
○ In limited circumstances, Ultrafast Capable Connectivity with a normally available download speed of at least 50Mbps and 20Mbps upload.
In fairness many physical limitations (1Gbps router ports that only do c.970Mbps, slow WiFi, network congestion etc.) and restrictions on the performance of remote online servers mean that even the best 1Gbps connections today often can’t take full advantage of such speeds. However, most people, particularly in F20 areas, would be more than happy with speeds of 500Mbps and uploads of 200Mbps, if they could get those.
Nevertheless some consumers may be confused by the dilution of “gigabit-capable,” although we saw something similar with BDUK’s earlier “superfast broadband” scheme and this is thus fairly normal (e.g. the 2016 state aid agreement promoted 30Mbps+ speeds, but the contracted peak-time minimum requirement was for a connection that could deliver 15Mbps+ some 90% of the time).
Arguably a bigger query is the “limited circumstance” exception above, where “ultrafast capable” seems to have become 50Mbps+ (20Mbps upload). Most of the industry and EU tends to define “ultrafast” as 100Mbps+, although Ofcom prefers the somewhat wonky definition of 300Mbps+. In either case that’s well above 50Mbps and we’re not yet clear on what those “limited circumstances” would be.
Timelines
At present BDUK is currently still developing their approach and a new Dynamic Purchasing System (DPS) is expected to help handle this. However, the current expectation is that the target operating model for all this could be ready during the Autumn 2020 (we’re likely to hear about this at the next budget in October 2020).
After that the tentative plan suggests that the first suppliers (e.g. BT etc.) could be accepted onto the related DPS system in January 2021, although it’s worth remembering that the procurement and tender phase of such programmes can go on for a long time and we might not see any contracts being signed until around October 2021 (it’s not uncommon for delays to strike big projects, so it could be later).
In other words, and much like we reported last year (here), the first deployments might not then begin until the very end of 2021 or early 2022 and, once again, that’s assuming no delays. As usual any operators taking part in this will need to provide some details to confirm their financial eligibility and security, such as audited accounts for the previous three years. But otherwise the process looks as if it will be fairly open.
A wholesale requirement also seems set to play a role, much like it did with the earlier BDUK programme. Otherwise the procurement and delivery timelines for all this remain ambitious and could put additional strain on existing problem areas, such as the shortage of skilled fibre engineers.
Suffice to say that achieving that 2025 target on time still looks very shaky, although we don’t yet know whether the UK Government’s recent acquisition of OneWeb’s LEO broadband satellites will factor into this at some point (here). SpaceX also have a similar constellation. We’ve yet to see how any of those space-based platforms actually pan out and what the real-world performance or affordability is like.
