A new select committee report has given the UK's smart meter roll-out a thumbs up. But is the project doomed to failure?
Computer Weekly has spoken to experts who have raised concerns over the economic impact and benefits, the technical infrastructure and the IT project management behind the UK's smart meter programme.
When engineering consultant Mott MacDonald calculated the cost of the smart meter programme, the firm stated the net present value of the programme would be £4.0bn in the red, according to Alex Henney, an economist and advisor for the electricity industry, who gave evidence to the Smart Meter Select Committee.
Yet the civil service has put the net present value at £4.9bn.
“The civil service went to town to tweak the numbers," said Henney. "It's a political freak to go from -£4bn to +£4.9bn in four years. If that does not ring a bell, then you can believe pigs can fly.”
More expensive than Continental counterparts
He said the UK programme is twice as expensive as the Spanish and Italian smart meter programmes.
“There are two obvious differences. Firstly, the Italians and Spanish rely on a powerline network [for data communications], which is simpler and cheaper than wireless technology.”
Italy and Spain also use a central distributed network operator (DNO) to roll out the smart meters, which, he claims, is simpler than relying on electricity suppliers.
“We have devised the most complex roll-out in the world,” Henney said. The UK approach, which relies on suppliers rolling out meters, will need an extra large database to collate information on who owns the meter, incurring costs and introducing errors and complexity.
It's a political freak to go from -£4.0bn to +£4.9bn in 4 years. If that does not ring a bell, then you can believe pigs can fly.”
The UK's smart meter project will enable consumers to see how much energy they use. Henney said the programme will require an energy usage display in each household that costs £25 each. That may not seem like much, but over the course of 43 million households, it represents over £1.1bn.
Henney believes many people will throw these displays away and they are unlikely to have any impact on people's usage patterns.
"The average residential consumption is 4,000KwH compared with 16,000 KwH in Norway where home heating exclusively uses electricity," he said. "The average UK electricity consumption has not increased very much. In the case of gas, consumption has gone down."
The EU wants member states to provide 80% of all households in member states with smart meters. A new Ernst & Young study for the German Federal Ministry of Information and Technology has not recommended smart meters in Germany.
No major environmental boost
From Henney's analysis, smart meters will not give the UK a major environmental boost as many homes now use efficient condenser boilers for heating and energy efficient lighting. He believes the government's premise that people will manage their energy consumption is flawed.
In other parts of the world the largest gains in energy reduction have not come from smart meter roll-outs, but from targeted measures that reduce peak consumption.
"In California, the main demand side response is not coming from real-time smart meters, but from electricity suppliers directly controlling air conditioners," he added.
Ross Anderson, professor in security engineering at the University of Cambridge Computer Laboratory, warns that the energy industry has no reason to lower energy usage.
"There is no prospect it will meet its energy saving goals as the meters will be controlled by the retailers whose interest is to maximise sales volumes," Anderson said.
"The project was sold on the basis of a thoroughly dishonest impact assessment and it's pressing ahead, despite lack of agreement on many aspects of the specification.
"It's a classic IT disaster in the making."
Beyond the choice of wireless over powerline for smart meter connectivity and the potential ineffectiveness of people to manage their consumption, the software development and project management of the programme may not be robust enough.
The costs associated with the IT behind the programme are likely to escalate and timescales will overrun, according to software engineer Martyn Thomas, a member of the IET Information Technology Policy Panel.
"It is a very large IT project, and the government's track record is not good," said Thomas. "The government usually overlooks the amount of business change."
Risk is another factor he believes will contribute to the project's demise. "There isn’t a properly constructed risk register with provisions to sort out problems if they should arise," he said.
Thomas, who previously worked on tax office IT, said: "One of the ways we managed projects in taxes was by identifying possible outcomes for the project and assigning risk to each."
Thomas is a strong proponent of formal methods, which he said will decrease project cost.
"Most of the costs in an IT project are the efforts in finding errors. We’ve known for 40 years that testing only shows the presence of errors, not their absence," he said.
According to Thomas, a formal methodology would reduce errors getting into the smart meter programme. Problems areas are identified early. "You save an awful lot of time."
The chance of failure is high. In his evidence to the Select Committee Andrew Ward, operations director at Scottish Power highlighted a project oversight which occurred at the company's US division.
"As part of the deployment [the US operations] rolled out 200,000 meters and had to replace 5,000 because they could not update communications over the wire," said Ward.
In other words, the team had failed to identify the potential flaw of meters not being remotely accessible and had to replace 2.5% of them at considerable cost. If this were to occur in the UK's roll-out, almost a million households would be affected.