Smart meters and untidy thinking - a blog from UKERC Co-Director Keith Bell

22 Jul 2015

Why does switching electricity take so long in Britain? And can smart meters help?

Earlier this month, the Competition and Markets Authority (CMA) published its latest interim report on competition in the energy sector, which concentrated on energy retail and the perception that consumers have paid significantly more for their energy than they needed to, and highlighting the ‘stickiness’ of customers who fail to switch retailers.

One of the things that makes it hard to foster competition around electricity and gas is that one retailer’s product is much the same as another’s. Voltage doesn’t have a colour except insofar as you are connected to the red, yellow or blue phases (sorry – engineers’ joke).  And it isn’t the retailer that determines quality of supply – the voltage, frequency, harmonic content and reliability of supply for electricity or the calorific value or pressure for gas – but the distribution network operator (DNO) – which consumers don’t get to choose.

All that is left in respect of the retailer are how much they charge are and whether they send out the correct bills. (OK, some retailers do offer options on ‘green’ energy).  Of course consumers need to be able to know what their bills are likely to be, and to have the option of switching if they don’t like what’s offered, and before their new retailer puts them on a worse deal.  But wouldn’t it help make the market more responsive to consumers if they could leave – or come back – much more quickly?

Both the CMA and Ofgem have recommended cutting the maximum time taken to switch from one retailer to another from 28 days to no more than a week, which would be more in line with other countries. There is currently something like a 40-step process to get through (though there is also a desire to give people a reasonable ‘cooling off period’ in which they can change their minds). Many of the steps are manual and include someone digging under the stairs to note the meter reading. Wouldn’t ‘smart meters’ help in this context?

DECC has mandated a roll-out of ‘smart meters’ to every household in Britain by 2020 at an estimated cost of £10.9 billion, justifying the expenditure on the basis of facilitating reductions in the energy use, mainly through consumer responses prompted by real-time displays of energy use. Understandably, in my view, many people are sceptical about the whole exercise if all it gives is ‘in-house displays’.

In other countries, ‘smart meters’ have been rolled out and claimed as a success. In Italy, they were justified as a means of reducing the cost of ‘non-technical losses’ such as theft or unpaid bills. On that basis, they were quick to deliver and enormously successful.

The clue to take from this is something implied by definitions not of ‘smart meters’, but ‘advanced metering infrastructure’ (AMI), a key feature of which  is a (secure) communication between utilities and customers. Such a communication is critical to providing information to consumers (or disconnecting reluctant bill-payers) and collecting time-stamped consumption data so that bills can be determined quickly and accurately, including for any time-of-use tariffs that might be in place. (Welcome, Economy 8?) However, that basic communication says nothing fixed about what the consumer might do with the information they receive from a utility or gather about their own consumption.

I confess that I haven’t followed Britain’s ‘smart meter’ programme closely but my feeling is that we have made two basic errors. The first is to fail to distinguish between ‘advanced metering’ and ‘smart stuff’. The second has been to make meters the responsibility of retailers rather than DNOs.

Focusing on the minimum specification for secure time-of-use billing and providing information to the DNO should enable them to manage the network better. As I understand it, the main part of the cost of the proposed new ‘smart meters’ is not the cost of the meter itself but of safely installing it: of the order of £100 out of around £150 per meter. The DNOs could fit the meters  – and forget about them, with little risk of rapid obsolescence – and manage the network.

Retailers, and anyone else with a bit of imagination, could then compete for the supply of devices easily installed by the consumer themselves on their side of the meter (and not within the meter itself). Consumers could choose what features they want, including none. The devices would be cheap and have minimal installation costs so retailers could, without too much difficulty, write off the costs when customers switch, if indeed it was the retailer that provided the device. (It could easily be the consumer himself or herself.)

As the availability of power becomes increasingly variable and uncertain, with growing volumes of wind and solar power, and electrification of heat and transport, the infrastructure and information will be in place to enable smarter grid management and the development of increasingly differentiated products (and cheaper tariffs) by retailers.

This might seem like pie in the sky. But we can already signs of what’s possible in the growing market for flexible demand among half-hourly metered electricity customers and the services offered by aggregators. So perhaps the future won’t be so dumb after all.