Paul Micallef, global digital grid leader, and Tony Ward, UK & Ireland power and utilities leader, EY Energy networks, Energy retail, Policy, Policy & regulation, Smart metering, Opinion

Tony Ward and Paul Micallef argue that the costs of the smart meter rollout have been underestimated and that the 2020 deadline is effectively unattainable in the project’s current form.

The latest assessment of Britain’s smart meter rollout by the National Audit Office (NAO) made for stark reading – but did not come as a surprise to EY. In November, the NAO said: “The government’s original ambition of installing smart meters in every home by 2020 will not be met, and the cost of the rollout will likely escalate beyond initial expectations.”

Rocketing costs

We have consistently flagged that the smart meter rollout would cost more than expected.

As far back as 2009, when the government had initially estimated the cost of the rollout to be £7-9 billion, we publicly argued that the true cost was likely to be significantly higher, at £13.4 billion. We contended that the government was underestimating the scale of the additional technology and infrastructure required to support smart meters and was relying on an assumption of absolute efficiency. This was not a view shared by senior figures in the industry and government at the time.

We continue to believe that the ultimate cost of the rollout in its current form will be more than £13 billion, particularly because likely “early adopters” of smart meter technology will have been targeted early on by the energy suppliers charged with ­delivering the rollout, with remaining installations ­progressively more challenging to deliver.

This suggests an extra cost of up to £75 for every household will be required to deliver the programme as it is currently conceived.

Missed deadline

EY has become increasingly convinced that this hugely complex infrastructure project would also miss its deadline for completion.

• At the current installation rate of 1.25 million per quarter, it is likely to take six-and-a-half years to finish the domestic rollout – well past the end of 2020 deadline.

• To hit that deadline, the installation rate will need to average 3.3 million per quarter, almost three times the rate achieved in the latest quarter. As a result, costs will skyrocket.

As things stand, energy suppliers are under pressure to achieve the rollout deadline. They are obliged, under their licence conditions, to “take all reasonable steps” to install smart meters in all households and small businesses by the end of 2020. And they have already seen peers being fined substantial sums for missing interim targets.

Ofgem’s approach suggests it will take its enforcement obligations relating to the smart meter rollout very seriously. There is a serious risk of unintended consequences if the outcome is to raise the overall cost of the smart meter rollout significantly, while also increasing the likelihood of customers being misled in the drive to install more meters.

Time for a time-out

We believe that the time has now come for a sober evaluation of the likelihood and value of meeting the rollout’s original schedule and targets. Blindly continuing to attempt to hit the 2020 deadline is going to be increasingly uneconomic.

We still agree with the programme’s objective of delivering more accurate billing to customers across the country, but the quality of the installations achieved is of great importance in unlocking the potential benefits of the rollout in full. Good installations need to be safe, strengthen consumer trust in the industry, and prompt effective conversations about energy efficiency.

With all this is mind, now is the right point for a time-out.

We do not want to see challenges associated with the programme being made worse by making bad decisions in an ultimately fruitless attempt to meet the current deadline. There is a huge risk that doing so could fatally undermine the credibility of smart metering in the eyes of consumers, severely compromising the benefits that will ultimately flow from the rollout.

Finding a way forward

There are two options for the smart meter rollout that should now be considered:

• Relaxing the deadline for completion from the end of 2020 to the end of 2025. ­Maintaining the installation rates currently reached – 1.25 million per quarter – should make this comfortably achievable, even allowing for some fall-off as likely “early adopters” have been tapped.

• Ignoring the “tail” of properties – ­aiming, at least initially, for 70 per cent ­take-up instead of 100 per cent. According to the NAO’s report, energy suppliers have said they expect to replace 70-75 per cent of meters with smart meters by the end of 2020. The EU only ever stipulated that member nations should pursue 80 per cent smart meter penetration by 2020 (under Directive 2009/72/EC). The ­German economics ministry decided that even this 80 per cent level would be ­economically unreasonable for most German consumers after EY published research in 2013.

The aim here must be to make the rollout as cost-efficient as possible, minimising the overall expense of a programme that will ultimately be borne by consumers. If both measures were to be implemented together, we estimate the cost per household could fall by £73 or more, compared with the government’s latest assessment.

Between a rock and a hard place

There is no doubt that the recently implemented price cap is already putting significant pressure on energy suppliers’ margins. Seven energy suppliers failed in 2018, including Spark Energy and Extra Energy, with a combined 400,000 customers.

As the NAO noted, the costs of the smart meter rollout will increase further if SMETS1 meters have to be replaced, or if the cost of installing smart meters – which in 2017 was already 50 per cent higher than the government had expected – cannot be reduced.

Caught between a revenue cap and install costs that have stubbornly refused to come down (and could yet increase), it is not overly fanciful to suggest there could be more energy supplier failures in the coming year.


What to read next