The UK smart meter transition: industry structure, market power, and interoperability

The UK government started an energy digital smart meter rollout in 2008, an “… £11bn scheme to put 53m devices in 30m homes and small businesses by 2020” to yield an estimated gross benefit of £16.7bn. Calling the rollout a disaster would be generous — it’s behind schedule, about £1bn over budget, and full of technical and economic problems that make a fantastic public policy case study in what not to do. Good media reporting is bringing needed attention to what’s happening, but the media discussion is missing a lot of what the economic value proposition of meters can, and should, be.

These implementation problems have real consequences for British consumers, and could slow or undermine the substantial economic and environmental benefits associated with energy network digitization. Since early 2017, diligent reporters at the Daily Telegraph have investigated the problems and cost of having to replace millions of meters shortly after installing them, of having the meters lose functionality if the consumer switches energy suppliers, and even if they then switch back, and of having average energy savings substantially lower than estimated and expected. Why and how is this happening?

The legislation authorizing the smart meter rollout directed energy suppliers to install meters for their retail customers. The organizational structure of the UK electricity industry yields some insights into why that was a poor design choice. Privatization of the nationalized vertically-integrated industry in 1990 created National Grid as the monopoly transmission network operator and 14 regional monopoly distribution network operators. Generation companies owned power plants, and could also serve as retail service providers in the 14 distribution regions. Retail markets were open and competitive, so non-generation-integrated energy suppliers could enter to compete against the generation-integrated energy suppliers, now known as the “big six”. The big six retain about an 82 percent market share in the retail market, which should not be a surprise, since allowing generators to be retail energy suppliers allows them to exercise vertical market power that acts as an entry cost for smaller firms. I gave a Beesley Lecture in London in November 2014, making precisely this market power point.

Given this structure, requiring electricity suppliers to install digital meters puts the technology choice in the hands of the “big six” vertically-integrated suppliers as well as their competitors. Take their economic incentive to differentiate themselves from their competitors, and add that to the government’s poorly-specified technology interoperability standard, and you get a failure of interoperability with meters not able to work across the different supplier communication networks. As described by network expert Nick Hunn,

Making the energy suppliers responsible for the meters was one mistake, but there was another one to come, which was in the specification of the meters.  An interesting by-product of asking the energy supply companies to be responsible for the meters was that they all wanted to add some of their own features to the meter specification.  So instead of accepting the approach of making it as simple as possible by following the engineering principle of “perfection is finally attained not when there is no longer anything to add, but when there is no longer anything to take away”, they all added their own favourite features.  It was a process akin to schoolboys competing to see who could pee highest up the wall.  I’ve documented the effect of that in my previous articles.  The result is that our meters are the most complex in the world, they cost four or five times what they should (which is all added to our energy bills) and the programme is running late.  Very, very late.

The idea initially was to have the first generation meters, SMETS1, installed sparsely and provide a sandbox in which suppliers could learn some things about design and deployment. Then the new, standard communication network (DCC) would connect all suppliers and all consumers and provide an information-rich interoperable system using meters with DCC-compliant standards called SMETS2. SMETS1 meters will not be able to operate on DCC, which was itself a poor architectural decision, but if they were only replacing a small number and they learned something it could be justified. But the complexities and delays in designing and building DCC, and the delay in deploying SMETS2 meters beyond their November 2016 target mean that, as Katie Morley reported in the Daily Telegraph in May 2017, up to 8 million SMETS1 meters may have to be replaced. Energy suppliers have been continuing to install them because of the regulatory mandate and harsh penalties of 10% of revenue they face if they don’t.

Morley doesn’t say this, but I don’t understand why the original SMETS1 meters couldn’t be updated through a firmware update, which is how a lot of technologies enable backward compatibility and how a lot of digital meters are designed to implement industry interoperability standards (and when I was on the GridWise Architecture Council this was how we advised firms and regulators to implement interoperability). Ofgem, the energy regulator, had a public consultation period in May 2018 on “maximising interoperability for first generation (SMETS1) smart meters,” but why didn’t they focus on that when they initially established the communications standard? They could have stipulated that the SMETS1 standard had to be extensible, to be able to be updated to future open-architecture interoperability standards. That would certainly involve some cost, but compared to the cost of installing and then replacing 8 million obsolete meters the software design cost would be smaller.

The smart meter rollout has been so disastrous that a group of MPs called the British Infrastructure Group of Parliamentarians (BIG) have called for an inquiry and review of the process. Their Not So Smart report provides a thorough catalogue of the economic, technical, and regulatory problems in the rollout, and they suggest some approaches to reduces the costs and deliver benefits from the digitization. I would like to see the media engage with and analyze the recommendations in this report; I am not close enough to the details to be able to do so.

One final point in the litany of “what not to do” that this case study highlights: all of the sources that I read in writing this post miss the point about the economics by fixating on how providing information to consumers will help them save energy and money. That value proposition only scratches the surface of what’s possible. The real value for consumers comes from being able to attach appliances and devices to the digital meter and automate choices that will help them save energy and money — more specifically, automating responses to price signals, to use transactive energy to change settings on appliances in the home without the consumer having to intervene or even to be home at all. Transactive energy and digitization empower consumers to avoid high prices and to have access to low prices, which is all-the-more important in a country that is moving so aggressively toward electric vehicles.

For this reason, Katie Morley’s recent article, Smart meters will let companies change cost of electricity every 30 minutes under ‘surge pricing’, is incorrectly scare-mongering. Surge pricing is a feature, not a bug, of transactive energy systems, because if the digital technology enables consumers to automate responses to the surge pricing by changing appliance settings, they can avoid the higher price by reducing their quantity demanded. And the flip side of high prices in peak periods is low prices in overnight periods. When consumers are on a fixed-price tariff they don’t see high prices, but they don’t see low prices either. Consider the EV owner who can plug in the car, and has pre-programmed a trigger price — when the price goes below £0.08 per kwh, charge the car! As the ways we use electricity become more diverse, the services we can choose can be differentiated and more valuable to each of us through such customization.

Deeper and more creative thinking about transactive energy, and how using digital technology to automate responses to price signals empowers consumers, can provide a vision that shows a way forward out of the smart meter rollout debacle.

Advertisements