BT announced that it was bringing out a new range of home phones with much improved energy efficiency. The claim is that ‘the new handsets boast power units designed specifically to consume around half the power of previous units’. BT said that almost all its extensive home phone range would contain the new energy-saving technology by mid-2008. Its press release gave very precise figures for the amount of CO2 saved – comparing the savings if all home phones incorporated the new technology to taking ‘57,000 cars off the road for a year’.Read More
In Denmark and Germany, large numbers of individuals own shares in local wind farms. If the government encouraged this in the UK, a large part of the local opposition would disappear. Onshore wind farms in windy locations are good investments which could form an effective part of many people’s pension plans. One of the few co-operatively owned wind farms in the country has almost finished raising its funds. Investors have put up £3m to buy two existing turbines in the Fens. Locally owned wind farms should be encouraged as a cost effective means of cutting emissions.Read More
BT uses over half of 1% of the UK’s electricity and is the single largest purchaser of green electricity in the UK. It buys over 10% of the country’s total supply of renewable electricity. It now seeks to develop wind turbines on some of its own sites. It intends to invest in about 120 2MW turbines to produce about a quarter of its own electricity or between 0.1 and 0.2% of the UK’s total need. This is an impressively large plan. The cost is about £250m. The financial return will depend on how much of the electricity replaces power BT would have bought from other suppliers and how much is ‘exported’. Assuming very little is used by BT itself, the return will be approximately £50m a year, yielding a return of about 20% on the initial investment. These figures assume that BT gets a yield of about 28% of the rated capacity of the turbines, which is about the UK average.
These figures depend entirely on finding sites. I think that BT may well have substantial difficulties finding as many 120 places where it can capture enough wind to average 28%. Perhaps more importantly, at many of those sites which do have enough wind, I think it will have problems getting connections to the local distribution network. Two of the three initial sites identified by BT are in the Scottish Islands. Although a typical 2MW turbine is not a huge generator to add to the local network, the islands have quite limited electricity needs. Scottish and Southern may not easily be able to add these turbines to their network.
When I asked BT whether it had approached the local distribution companies to check on this point, I was not given an affirmative answer. This raises the possibility that BT announced these plans before detailed consideration of whether its aspirations are technically feasible. So it may be a great idea to erest wind turbines, but it looks like it will be much more difficult than BT realises. Companies like Ecotricity have been developing wind turbines on industrial sites for years. Though planning permission is easier, there are still huge obstacles to overcome. BT needs Ecotricity's expertise immediately, but it will still struggle to meet its aspirations to grow its wind power capacity.
BT’s green credentials are well established. It is the largest commercial buyer of renewable electricity in the UK, emphasises the importance of carbon reduction across the organisation, and pushes voice and video conferencing at an unconvinced customer base. In any international ranking, BT’s sustainability measures get high marks. But BT has the same problem as many other organisations: its server farms are growing in number and size. The increased power consumption in its data centres explains why the organisation’s electricity demand is growing. Eventually, its brand image will suffer as critics suggest that its public stance on green issues is not matched by its internal behaviour.
BT’s electricity use is about half a percent of the UK’s total, and its server farms represent over 10% of its energy consumption. BT says that data centre use is rising at 40% a year, and the company’s emphasis on growing video businesses, such as BT Vision, is likely to increase data storage and transmission demands into the foreseeable future.
BT’s response has been to attack the power use of the server with radical measures that set best practice elsewhere in the world. Its new data centres use fresh air cooling, not air conditioning, and the company runs its machines at much higher temperatures than used to be considered possible. Since cooling servers uses at least as much power as running them, this is an important step. The second major innovation is to run the farms on DC power, cutting the very significant losses in the multiple AC to DC conversions in a conventional centre. Better ‘loading’ of the computers helps as well. A well-utilised machine uses only a little more power than an intermittently under-employed server. BT claims that these measures can reduce the typical power consumption of a server farm by 60%.
Across the world, data centre energy consumption is becoming a bigger issue. The world has about 26m servers pumping out data day and night. Estimates suggest that they use about 2% of all electricity produced and global growth is probably around 15% a year. BT’s innovations may be a useful model for others to follow. But the unfortunate fact is that at current growth rates the maximum efficiency gains will be wiped out in less than four years.
In an intriguing trend, some companies are dealing with apparently unquenchable growth in data traffic by beginning to move away from thousands of servers based on PC technology towards huge single computers with lower total energy costs. Who said the mainframe was dead?Read More