UPDATE BT Explain Unusual Broadband Speed Estimates on UK FTTC Lines

An interesting question popped up during BTWholesale’s last Customer Service Forum, which occurred after an ISP raised concerns that BT’s estimate for speeds on its ‘up to’ 80Mbps capable Fibre-to-the-Cabinet (FTTC) broadband lines were changing after a period of time. This made it difficult for them to determine whether particular users had a speed fault.

The performance of FTTC broadband lines on BT’s network will always vary a little due to the “last mile” style run of copper cable between street cabinets and premises, which is prone to interference (e.g. signal degradation over distance). On top of that they can also be affected by network congestion and Traffic Management measures among other things. But once your speed has settled it shouldn’t vary too aggressively, although you do have to keep an eye on it in order to establish the typical trend for your line.

Never the less ISPreview.co.uk has heard various reports over the past few months about more unusual behaviour and a question posed at the recent BTWholesale Forum by PlusNet’s resident fault analyst, Ian Boydon, appeared to touch on a number of similar areas.

Ian Boydon, PlusNets Fault Analyst, said:

“We have noticed recently that the BT estimate for Fibre speeds on lines is changing after a period of time. I am assuming that this is due to contention, as more connections are added at the cabinets? The problem is, it makes it difficult for us to determine whether or not a customer has a speed issue or not.

We have seen examples recently where a customer has raised a speed fault with us and after a week or so of investigation, the BT estimate has been reduced and we then have to tell the customer there is no problem as their speed is within range of the new estimate. As you can imagine, this causes a lot of dissatisfaction with our customers.”

BTWholesale’s response was unsurprising and rather general with the operator referencing how “Line Length, Line Quality and Cable Fill can affect the estimates given by the speed checker“, which is all well known. But interestingly they also stated that “Speed estimates assume an average local line length of 50 metres, however this is an average and local line length can sometimes be up to 1km“.

ISPreview.co.uk has heard of lines from the cabinet that can run beyond 1km and thus requested further clarification from BT, not least with regards to precisely what this covered and whether it would be possible to know what the MEDIAN distance is. Fifty metres is quite a short run and if this is being assumed as an “average” for all FTTC speed estimates then those on longer lines might be getting a more optimistic prediction than could be deemed realistic.

In the pursuit of additional information we contacted Thinkbroadband, which collects a wealth of real-world speed test and line data, to see how BT’s connection speed estimates might compare against predicted FTTC line lengths using actual real-world speeds. TB estimates that 90% of the UK cabinet-to-premise line lengths are under 1km (i.e. 10% should be above) and, using test data from BTInfinity customers, they suggested an average (mean) length of 450m, with a median of 550m.

So far we’ve been unable to gain a clarification from BT concerning their response to PlusNet.

UPDATE 10:23am

An additional couple of points have been made to us that are worth mentioning. Firstly, TB has seen some users with lines that stretch up to 2km from the cabinet. Secondly, a longer term problem for FTTC is that speeds may drop as crosstalk (interference) begins to have an impact (TB suggest this could drop performance up to 10-20%); this is of course one reason why BTOpenreach are currently testing Vectoring technology as a possible solution for the future (here).