Figuring out the speed people are likely to get from a FTTC based roll-out, i.e. the bulk of BDUK spending in the UK is something that has vexed people for a long time and the Public Accounts Committee meeting and report has simply brought it even more sharply into focus.

We now have the situation where BT is on the record as saying it has no objections to map data showing where projects will be delivering being released, and then projects releasing their various phase maps, which both please and annoy to equal measure. Some Counties maps require almost forensic level image processing to recognise specific areas and very few counties release any information at the postcode level. As far as we are aware NONE of the county projects have identified exactly which cabinets will be enabled, even when postcode information is released, while not a problem for many as a postcode is served by a single cabinet, there are people with exchange only lines and no cabinet, or postcodes served by two cabinets.

We thought we might be able to do something, and possibly help answer the question about how much of an area will be superfast or not.

Distance versus speed from BT street cabinets

Distance versus speed from BT street cabinets
Credit to for providing their mapping system and API

The above map shows two cabinet locations (Binfield Heath in Oxfordshire to be exact), an area where residents raised money to get the two BT green street cabinets enabled, and we have then plotted all the postcodes that lie within a 2km radius of each cabinet, and used our table of distance versus speed for FTTC to estimate what speed range a postcode will fall into. We multiple the radial distance from the cabinet by square root of 2 to give an approximate road distance.

  • GREEN, Superfast with a speed >= 30 Mbps
  • YELLOW, speed is > 2 Mbps but less than 30 Mbps
  • RED, speed is <= 2 Mbps
  • BLACK, estimated distance of 3km and VDSL2 assumed to not work.

The problem now is that while it is easy enough to identify a postcode at a certain distance from a cabinet, identifying whether it is served by that particular cabinet is more problematic.  This is where local knowledge comes to the fore and some common sense, so we are very confident that the yellow and red postcodes in Caversham are served by different cabinets. Those yellow postcodes in Dunsden Green we are not so sure about. If we knew the location of all the cabinets in that map area we would have no problem plotting speed to the closest cabinet, but that also has its faults, as people are not always connected to the closest cabinet.

So while this system is not perfect, it has a degree of use and nothing stopping us from plotting the map using more colour to highlight the variation in speeds.

I may regret saying this, but if there are community projects who know where their BT cabinet is located and accept the caveats above we can produce data for other areas, not overlaid onto a map, but a basic list of postcodes with the expected speed then email me. Compared to the information coming from the councils and BT themselves we believe it would be useful.

I should point out no secret BT data was used to produce this map, and if BT really wanted to help inform the broadband speed debate we would like to invite them to provide us with lat/lng (or postcode) information for the 85,000 cabinets in the UK and a postcode to cabinet cross-reference. If BT were to do that we might accept a caveat of not releasing the cabinets exact location, though of course people could guess it based on the spread of speeds, or heaven forbid notice it as they walk/drive/jog by them in their local area.

10 Responses

  1. Chris Conder on 03 Oct 2013

    Well done, its about time someone used a bit of common sense. Its clear to see now why FTTC is not the answer to rural internet access, when so many are left out in the cold. If the residents had raised money and brought in gigaclear or another altnet they could have made a start on getting a network for the future. as it is, all they have done is help a minority go a bit faster. the same is happening all over the country. Its a scandal.

    • Somerset on 03 Oct 2013

      So it’s a scandal that residents of an area do not raise money for faster broadband? This is the problem, fast broadband is not that important to everyone and it’s difficult to get everyone involved, particularly where money is involved.

      It’s not helping a minority go faster…

      What’s the experience in Cornwall? Surely we can learn from there instead of guessing.

  2. gah789 on 03 Oct 2013

    This is an instructive exercise. It highlights the strengths and weaknesses of the FTTC upgrade approach. Leave out the concentration of postcodes in and around Caversham and what you have is rural suburbia, not rural rural. Even then about 10% get 30 Mbps.

    If this is at all typical, then there will be a great deal of disappointment about the actual achievements of the BDUK program because too many people live too far from cabinets for the targets to be met in terms of real line speeds rather than notional coverage. Even BT may be disappointed because take-up rates may be rather lower than they plan for.

    • gah789 on 03 Oct 2013

      My comment was garbled in transmission: The figures should say that 10% get 30 Mbps.

      • gah789 on 03 Oct 2013

        Hopeless: the system removes all text between less than and greater than signs.

        Final correction: 10% get less than 2 Mbps and only 30% greater than 30 Mbps.Pre

  3. andrew on 03 Oct 2013

    It is more complicated than that, which I covered in the article, i.e. there is another level of refinement needed, which is to identify all the cabinets in that area.

    As for helping a minority looking at the postcodes close to the cabinets and properties present, versus others probably served by the two cabinets (as opposed to just in the radius) I suspect the majority are in the green superfast area.

    If communities want to see how they stand as I said they can email, but they need to spot the cabinets themselves, as driving around on Google Streetview can mean you easily miss cabinets without local knowledge, e.g. where the old post office was.

  4. mervl on 03 Oct 2013

    May not be relevant, but wasn’t there available (perhaps “leaked”) for the commercial roll-out an Excel spreadsheet list by postcode of the expected “uplift” from ADSL to FTTC (VDSL2)? I remember seeing the estimated uplift for my postcode at 8x, which proved to be about right.

    Sure it’s going to need some surveying but why can’t BT make the info publicly available and updated as work progresses. Accountability. But I suppose in the nomansland between BDUK and the County Councils it’s no-one’s job. And that suits the bureaucrats and BT accountants just fine.

    • andrew on 04 Oct 2013

      Well aware of the leaked postcode to PCP list, but it does not tell you a PCP location.

      Inferring it might be possible, but inaccurate, as highest uplift does mean you are close to the cabinet, it might just a premise a long way from the exchange, so gets a big boost from FTTC.

  5. Iain on 15 Dec 2013

    There is a lookup that correlates postcodes to cabinets. It is at

    Put the postcode in and it tells you which cabinet serves that postcode.

    Sadly, it does not specify cabinet locations.

    • seb on 15 Dec 2013

      It does correlate and is useful information, but you can’t map a postcode to a cabinet accurately–you may find one house is on one cabinet and another is on a different one, even if thy are in the same postcode.

Leave your comment