Scotland - access to superfast broadband

There are frequent references in forum postings to the proportion of the UK's population who live sufficiently close to a cabinet to be able to get superfast broadband (SFBB) - either 24 Mbps or 30 Mbps - if the cabinet has FTTC. I am not sure what the source of the statistics is and it has always been clear that the estimates are not representative of different regions or countries within the UK.

Some researchers at the University of Edinburgh have carried out a detailed investigation by postcode and local authority sector for Scotland - see

http://idea.ed.ac.uk/data/PostcodeAreas/Scottish_pos...

The cut-off is 1.2 km from the cabinet for 30 Mbps. In addition, the study allows for exchange-only lines. There are two sources of potential error:

(a) It is not possible to take account of the quality of local (copper) network or of lines that do not follow the most direct route to a cabinet. This means that the "true" proportions of the population not able to get SFBB are higher than the estimates suggest.

(b) Details of VM coverage are confidential, so that there will be some areas that can get SFBB from VM that lie outside the cutoff for FTTC.

It seems likely that the poor state of the copper network outweighs the availability of VM service outside FTTC areas.

The estimated percentages of the population not able to get SFBB vary from a minimum of 14% in East Ayrshire, East Lothian & Fife to 30%+ in 9 local authorities. It would be interesting to learn what similar investigations come up with for other countries/regions of the UK.

There is a serious point here. The estimates suggest that the stated targets for the BDUK Step Change and HIE projects in Scotland for coverage of SFBB (even allowing for an adjustment to the 24 Mbps threshold) cannot be met other by some creative reinterpretation of the statistics. If the pattern in urban and suburban areas of Scotland - e.g. local authorities such as Glasgow City (26%) or West Lothian (21%) - is replicated in the rest of the UK, it is equally impossible to meet the DCMS targets for large parts of England.

Edited by deleted (Sun 02-Mar-14 17:48:08)

Link not working so cannot see how they have done the figures.

Interesting if they have line to cabinet and cab location info as to the source.

Also have they verified method with speeds actually available?

The question really is whether the Scottish project will use fttp in some areas

Fixed version of link.

Thank you for fixing the link. I am not sure what went wrong.

In answer to the questions raised by Andrew:

A. Yes, they had information (though not always complete) on cabinet location and line lengths. Most of the analysis is based on postcodes with census household figures mapped to postcodes. There are links which provide more details including on exchange only lines.

B. No verification of line speeds, but this is difficult to do even with Thinkbroadband's data. My view is that, at least at the individual level, one gets a biased sample of people carrying out speed checks, so separating out who has what type of connection from what speed each type of connection can deliver at postcode level.

C. I don't see why FTTP would make any significant difference. In practice, FTTP is a substitute for FTTC either for exchange-only lines or for very small cabinets - i.e. when the fixed costs of installing or upgrading cabinets cannot be justified. The calculations assume that every cabinet is enabled for FTTC. It is not clear what the installation charges will be for FTTP beyond the range of FTTC coverage but the installation cost of FTTPoD would be more than £4,000 for > 1.2 km. FTTP might be ok for those with short exchange-only lines but the numbers who would benefit are small.

Remember that the goal is supposed to be "affordable" SFBB. BDUK's definition is that this means < £100 installation cost and < £25 monthly cost. No FTTP provision can meet this - or not without heavy cross-subsidies. For example, the charge for BT Infinity 4 is £50 pcm. If the new FTTPoD charges reflect the costs of installing the service, then making FTTP available at an installation charge of £100 would involve a huge subsidy for longer lines.

http://idea.ed.ac.uk/data/PostcodeAreas/Scottish_pos...

You had a "full stop" included at the end of the link . It's a common mistake.

my view is that 30 meg is very optimisitic at that distance and probably not correct i would suggest around half tof that at that distance - anything much greather that 15 m/bps at 900 metres fromt he physical location of the cab as the copper runs would be expection rather than the rule -- the post code to match file only give the central location of each individual postcide served by the cabinet it does not provide the geographic location of the cabinet - where distance is actually determined from

Edited by deleted (Sun 02-Mar-14 16:05:46)

My recollection is that they were trying to be very cautious, allowing for the potential offered by vectoring and other improvements. Note also that their calculations assume that all EO lines within 1200 m of the exchange will be transferred to cabinets, which isn't part of the current projects unless there is a lot of money left over.

If, as a very rough approximately, 750 m from the cabinet or exchange is used as the cut-off for 30 Mbps and the relative population density with distance from cabinet corresponds approximately to the shape in the fibre guide, then the share of households in Scotland not able to get 30 Mbps at the end of the Step Change program would increase to about 36% - some way from the target figure which is supposedly only 5% by 2017-18.

Inevitably, the figures could be improved by more detailed mapping data, etc. But the big picture is that the official view that all but 5% of the population in Scotland will have access to SFBB by the end of the current BDUK program is entirely wrong. Instead we should be talking about the final 20% or even the final 33%. These are not just rural households - the reported calculations show that 55% of the households without access to SFBB live in what are classified as urban areas (with separate categories for small towns and rural areas).

Something else follows from this. It is frequently argued that the current model - gap funding of FTTC - may be tough on those who get left out but the vast majority of the UK get access to SFBB at a relatively low cost in terms of public support (mostly from the BBC licence fee and local councils). Instead, we see that numbers provided with access to SFBB from the non-commercial part of the program may be quite small, given that the commercial phase is supposed to cover > 50%.

One might ask: what are the actual improvements in broadband speed relative to, say, spending a smaller sum of money on upgrading every exchange to 21 CN / ADSL 2+? In Scotland, according to SamKnows, there are 282 (out 1070) exchanges with ADSL 2+, though I suspect that estimate is too low now. How much would it cost to upgrade all of the remaining exchanges to a minimum of ADSL 2+? Probably less than £50 million. Then, take the weighted average speed improvement, allow for time (since this could be done much quickly than FTTC), and you finish up with a benefit to cost ratio that is much higher than for the BDUK program.

I would expect if the same study was done for the whole uk the numbers would possibly get worse.
I'd imagine when the entire commercial roll out completes there will be a MAJOR shouting match about who did the original estimates and how did they get it so wrong.
People talk about bt fudging the numbers I'm order to push the fttc roll out as the way to achieve the government targets but surely there will come a day when it becomes clear that this is in fact the case.

Census areas are very difficult to map to postcodes, and in general premises are not spread evenly over either a postcode area or a Census Output Area in rural areas.
Also as a matter of protocol reading the T&Cs for the thinkbroadband map suggests that the researchers need permission to use the results in this way as they are not for individual consumption.