How It All Started

Between 2006 and late June 2008, most broadband consumers in Winterbourne Stoke received a slow (900kbps to 1400kbps), but generally reliable, broadband service from a number of different ISPs:

* BT Broadband
* Virgin
* AOL
* Zen
* Orange
* Waitrose
* Talk Talk

A few villagers had longer term phone and broadband problems, some of which had been reported to their ISPs and/or BT if it was a phone fault. These had remained unresolved. Following the replacement of power poles and lines in the village in June 2008 by Southern Electric, one villager noticed a major and recurring problem with his broadband connection and reported it to his ISP on 2nd July 2008. We did not know if the replacement of the power poles and lines was a cause of the problems subsequently identified, or simply a coincidence of timing.

The symptoms of the problem were:

* a loss of synchronisation between the villager’s broadband router and the BT equipment at the Shrewton exchange – this happened every night between midnight and 04:00 in the morning;

* The BT equipment sensed the loss of synchronisation and reduced the bRAS/IP Profile (the maximum speed at which the broadband will operate) each night;

* The bRAS/IP Profile continued to drop until it reached the minimum possible speed of 135kbps – roughly 10% of the original speed;

* When reset by BT Wholesale to the original level, the bRAS/IP Profile continued to drop;

* As a consequence of the low speeds, many internet services were unavailable or unusable.

The problem was reported (to BT Broadband) a second time in mid July and the problem was escalated and passed to BT Wholesale. Two visits by BT Openreach failed to find the cause and none of the 'repairs' achieved anything positive. During one of these visits, the BT Openreach engineer noted that a likely cause of the problem was the poor quality of the underground telephone line between the BT exchange in the next village and the BT distribution cabinet in Winterbourne Stoke. This was an aluminium cable (not copper) and prone to developing poor joints.

Lack of progress led to two things, checks were made around the village to see if others were suffering similar problems and also, in an email discussing concerns about rural broadband as a national issue, to raise the issue with Ian Livingston, the CEO of BT. It soon became apparent that virtually every broadband connection in Winterbourne Stoke was detecting the nightly fault and that about half the connections were being degraded severely.

Research around the village quickly showed that the likely cause of this problem was “Repetitive (or random) Electrical Impulse Noise” (REIN), although it took some time to convince BT to send out a specialist REIN engineer – who came during the day – at a time when over a month’s monitoring of the line by BT and the villagers showed the REIN would not occur. REIN can be caused by a variety of faulty electrical equipment including street lights, pumps, switches, Sky TV boxes, etc.

Villagers made every effort to find the source of this problem themselves – by looking for electrical devices that might be switching on or off in the early hours of the morning. We were even able to persuade Salisbury District Council to turn off the electrical power to a local sewage plant as an experiment to see if this was the cause. The Highways Authority were also contacted in case there was a possibility that traffic cameras on the A303 might be causing the effect, and the street lighting engineers were asked to look at their equipment. None of this activity has identified the source of the REIN.

After collecting further data on a daily basis and passing this to BT’s Technical Management Centre (TMC) in Leeds (a very, very helpful group of people), BT Openreach were eventually persuaded to send out a two-man REIN team late at night on 12 September 2008. They were able to witness the REIN event for themselves, both on their own equipment and on a second broadband connection which showed an identical fault at exactly the same time. These engineers investigated further and discovered that there were REIN events at roughly 6-hourly intervals, 24/7. Most of these were too small to switch broadband routers off, but those after midnight seemed particularly strong and likely to do so. Since then, it has been noticed that routers do lose synchronisation with the exchange, particularly in wet or windy weather.

More villagers reported their own faults to their ISPs and attempted to get the problem escalated to BT Wholesale, with limited success, but it appears to have taken over 3 months for BT Wholesale to begin to link these faults as elements of one and the same problem. Needless to say, despite several further visits by engineers to the village, to look at various broadband setups, the problem was not resolved.

Throughout September and October 2008, various experiments were conducted in order to eliminate the REIN problem – at the suggestion of the BT TMC and with the help of one of the villagers. These included:

* Removing bell-wires from broadband faceplates;
* Fitting new Cat 5 modem cables;
* Fitting ferrite cores to cables;
* Fitting the BT iPlate (a new device said to help in such circumstances).

One thing the experiments demonstrated quite clearly, was that if the REIN were to be eliminated, most villagers could receive broadband at a speed of 1,400 to 2,500 kbps; significantly higher than most have previously had.

By mid-September we were advised that “BT Wholesale had costed the replacement of the cable at £180,000 and that these costs would have to be met by the ISPs and that this simply wouldn’t happen as the ISPs would never recoup the costs from the number of connections in the village”.

By the time BT gave up on trying to identify and fix the problem in December 2008, the only suggestion they had made by way of a “solution” was to cap the speed of affected lines to 500 kbps. In other words, to reduce the speed to one third of the June 2008 levels. This would have effectively 'concealed' the REIN event from the consumers, but done nothing to improve the situation, so we felt this proposal was entirely unacceptable