Skip Navigation

Welsh Government talks of plugging BT gaps
Tuesday 17 February 2015 09:35:56 by Andrew Ferguson

The Welsh Government has started the process of tendering for firms to partner to deliver superfast broadband to 100% of the country. The tender is described as being in the range of £1 million to £3 million and will have a duration of seven years.

"It is expected that the current Superfast Cymru contract with BT will not provide Superfast Broadband services to 100 % of premises in Wales by 30.6.2016 due to combinations of the following BT-reported issues in some locations:

  1. Legacy network architecture, such as ‘exchange-only lines’ preventing standard product deployment.
  2. Relatively low forecast revenue density for standard product development, such as business parks where investment may be more involved than in residential areas.
  3. Relatively high-cost of deploying non-standard products, some of which may exceed cost per premise expectations.

The Welsh Government accepts these issues are real and to address the resulting gaps in service provision. To this end, the Welsh Government has decided to establish a project to procure a service/s that:

  1. Provides Superfast Broadband services to these areas (as identified by the latest Open Market Review).
  2. Meets the necessary and sufficient requirements for compliance with the EC State Aid Decision as interpreted by latest BDUK Guidance.
  3. May use additional grant aid funding over and above that secured for Superfast Cymru.
  4. Promotes maximum possible competition from across the market and the various technical solutions.

This procurement will cover infill requirements for business and industrial parks (not residential) only and will be divided into 2 by geography (South Wales and North Wales).

Key part of tender document

Reading the description it sounds like BT have had an epic fail and are about to fail on their Superfast Cymru contract, but it must be emphasised that the contracted target was 96% coverage with fibre based broadband. So there was always going to be 4% who had no boost in speeds available to them and also those in areas with FTTC where speeds were less than superfast. The performance of FTTC means that the expected coverage at superfast speeds based on our analysis is in the 90 to 91% region.

Looking at the tender, if it is meant to help the 4% who got nothing from the existing BT contract then it needs to help some 6,000 premises and £3 million of funding equates to just £500 of funding per business premise. If this is about getting the final 9 to 10% up to superfast speeds the public money is just £200/premise.

Update 3pm The original estimates on gap funding have been adjusted to take into account that this funding is for businesses, where generally businesses account for 1 in 10 premises, though may be higher in some areas, e.g. rural farms and people working from home can skew the 1 in 10 approximation. Essentially though it means the funding has to deliver for fewer premises increasing the amount available to around £200 to £500 per business. Another difficulty of just trying to address business premises is that any economy of scale is lost and the geographic spread of the premises is much wider. Certainly compared to traditional estimates for the most rural areas £500 does not go very far, particularly when we know Virgin Media is looking to spend around £750 per property passed in urban areas, which means likely bidders will be fixed wireless or satellite based solutions.

Maybe there is a different story, that cost savings in the current BT project are looking like they may be able to extend beyond 96% fibre based coverage and that the number needing intervention is going to be much smaller and for the tender to make sense this suggests a figure of under 1,000 premises. Though if the number of premises is so low why is this contract being given a seven year life span?

The uncertainty perhaps explains why in other parts of the UK firms show initial interest and then hardly anyone bids for the work once they get to find out some more detail.

Update 6pm The Welsh Government has requested we include a statement 'in order to clarify some of the points and inaccuracies which have been made'.

The article on thinkbroadband has made some highly inaccurate assumptions about a current procurement tender which has been issued by the Welsh Government. We would like to clarify some points.

“Superfast Cymru, the partnership between the Welsh Government and BT, is making excellent progress towards achieving the aim of having 96 per cent of premises being able to access fast fibre broadband by 2016. That has always been the aim. The Welsh Government is currently out to tender on covering some of those areas which are not included in Superfast Cymru for a variety of reasons. This was always going to be the case. This is in addition to Superfast Cymru which is an ambitious programme and the largest partnership of its kind. Our partners in Superfast Cymru, BT, are making excellent progress in often difficult conditions in bringing superfast broadband across Wales. It has already been credited by Ofcom as having an impact on improving access to broadband in Wales.

“To suggest that this planned procurement means there has been any failure by Superfast Cymru or BT is an entirely inaccurate assumption.

Welsh Government spokesperson

We have included the full un-edited statement above but feel it is a little unfair as we did highlight that the original contract was for a 96% fibre based target, so there was going to be premises missing out. The EU tender did talk of BT issues rather than just a project to look at areas that were always outside the scope of the 96% project. We have asked the Government to confirm if they feel the 90 to 91% at superfast speeds estimate is incorrect and will happily publish what they state as an accurate figure.


Posted by ValueforMoney about 1 year ago
'Savings' according to the NAO are excess modelled costs. For BDUK contracts is will good to know where 38% of the original contract value ends, this was equating Cabinets subsidies in excess of £40k plus.
In Wales the 3,000 cabinet programme had a generic cabinet subsidy of £60,000. The report from the Welsh audit office is due in April.
Posted by andrew (Favicon staff member) about 1 year ago
@ValueForMoney you MUST take into account the FTTP roll-out too which is more than some other projects and may skew the results.
Posted by Gadget about 1 year ago
Long timescales could mean a focus on providing and running the service post deployment for a decent length of time
Posted by andrew (Favicon staff member) about 1 year ago
Interesting - so less of a gap funded but more of a subsidy for providing a service. Something the core BDUK process has gone to great lengths to avoid, i.e. repeat of Digital Region
Posted by TheEulerID about 1 year ago
It may be relevant that the BDUK "clawback" period is seven years.

BDUK certainly takes into account the operation costs (and revenues) for the seven year period against which gap-funding bids are made. After this period, it's all meant to be financially self-sustaining, including infrastructure renewal.

I can't see that seven years for implementation would be remotely acceptable.
Posted by TheEulerID about 1 year ago
There's some information on the BDUK 7 year period in the latest NAO report (section 5.5).

Incidentally, for those cynical about the slow take-up of BDUK-provided superfast, the same section says:-

"the rate of take-up of superfast broadband is currently nearly 5 times higher than modelled in the contract assumptions"

The implication is that BT's bid assumptions were deeply conservative (if this pattern continues).
Posted by WWWombat about 1 year ago

In the analysis of £50 per premise, you are assuming the money is to help all of the 4%.

The last line of the quoted tender tells you otherwise - this is for business premises only.

Ordinarily, that would be around 10% of the total number of premises, but it might be skewed one way or the other when you look at the final 4%.
Posted by andrew (Favicon staff member) about 1 year ago
Missed the business only - changes range to around £200 to £500 then, but also means you are looking either at lone premises or a small cluster of SME. Making it a difficult sub-division to service due to geographic spread.

Still not a lot of money for the final hardest to reach areas.
Posted by TheEulerID about 1 year ago

Whilst that's not a lot of money, one very important thing to note is this is an SME product (whilst BDUK was primarily residential). As such, the revenue model might look very different as it might not be constrained to the same pricing. When assessing gap funding, it's the difference between revenue and costs (inc relevant capital-related) which matter.
However, that said, it's still a modest amount.
Posted by TheEulerID about 1 year ago
I also ought to mention that an adjusted revenue model might be appropriate for the "last 5%" problem in general. Of course that's weasel words. It simply means that those in the last 5% might be expected to pay more, at least for some period of time. The loss of universal fixed pricing (at least in the OR network) might not be popular, but personally I think it's worthy of consideration. An extra £5 per month over 15 years is £900.
Posted by Blackmamba about 1 year ago
Hi Broadband Watchers.
Has anybody heard what the clawback rate in £ per line over the 20% if in contract because we have a tie cable full (100) and the extra pairs are going to be paid for by Openreach. The clawback should be then (80 times £xxx).
Posted by TheEulerID about 1 year ago

I'd be surprised if the clawback is per-cabinet. Isn't is more likely it's by BDUK project? As far as expansion like this goes, then I'd expect it all to be picked by OR as it must be self-funding at that point.
In any event, I'd be a bit surprised if that 20% takeup figure coincided exactly with 100 pairs. Most likely that threshold is hit a bit before.
Posted by kijoma about 1 year ago
"Meets the necessary and sufficient requirements for compliance with the EC State Aid Decision as interpreted by latest BDUK Guidance."

so that is fixed wireless off the agenda straight away unless the EU/BDUK have again moved the goalposts?
Posted by Blackmamba about 1 year ago
Hi Broadband Watchers.
As the contract (Surrey) is ending in aprox 6 weeks we are trying to get the money that is left over plus the clawback on the 620 cabs to be spent on the post codes (6000 lines) that are under the 15 meg target if possible.
Posted by TheEulerID about 1 year ago

From the linked EU document. The key word would appear to be "advanced".

"NGA networks currently comprise fibre-based access networks (e.g. FTTB, FTTH, FTTC/VDSL), advanced upgraded cable networks (HFC/DOCSIS 3.0), and certain advanced wireless access networks. As technology evolves, this enumeration might change in the future."
Posted by TheEulerID about 1 year ago
Regarding the 18:00 hrs update, I suspect it was sharply worded as the (unnecessary in my view) "epic fail" sentence grabbed the eye, even though it was qualified.

On the subject of SF, rather than "fibre based" coverage, the Superfast Cymru page defines a process for validating profiles of lines (see link), so they should have that information. However, in the assembly it has been stated the average achievable will be over 60mbps (so you might x-check that).
Posted by TheEulerID about 1 year ago
This is a Welsh Government release quoting 61mbps as the average speed that can be delivered using the Superfact Cymru infrastructure as of 11th December 2011.

It should be possible to check this against the TB estimation methos. Not the same as an official SF coverage figure, but a correlation check.
Posted by Blackmamba about 1 year ago
Hi Theeulerid
That's for the link.
I can know prove my point over targets in Surrey.
Posted by andrew (Favicon staff member) about 1 year ago
To be honest an average of 61 Mbps seems pretty amazing when maximum sync is 76 Mbps.

I know our FTTC estimates generally come out to the lower end of the A/B ranges for areas that have been checked by hand, but am happy with that as it means we will be right when take-up is high.

The actual speed profile for BT Consumer can be see at and illustrates how tough to get a 61 Mbps average as seen by the end-user.
Posted by TheEulerID about 1 year ago

A few things to note. Those are throughput tests whilst I suspect (but don't know) that the project are using line data rate. At 61mbps sync you's only expect 57mbps throughput. The project is also reporting line capability and doesn't take account of sub-optimal domestic wiring (or anything sub-optimal about the way the test was run).

So I'm sure the 61mbps figure is with everything absolutely optimal.
Posted by andrew (Favicon staff member) about 1 year ago
@TheEulerID and would happily fight the corner that says they should actually say that is the case categorically or the 61 figure will get repeated as fact elsewhere and add fuel to the fire for the critics of the project.

User experience is what matters at the end of the day, since they are the one electing to buy the service.
Posted by TheEulerID about 1 year ago

True, but it's what happens when statistics are bandied around without them being properly qualified. I had a career in measurement and performance of major IT systems. The complexities are immense.

As far as user experience goes, I think the vast majority of the population only care if the services they use are functional and reliable. Does video stream cleanly; can they work from home. Does it take hours to upload a video clip.
You must be logged in to post comments. Click here to login.