Turning homes into virtual power plants

The potential for smart homes and devices to support the smooth operation of a low-carbon power grid whilst generating billions of pounds in savings has long been touted by the energy industry. But as Tom Grimwood reports, there is a raft of problems to solve before the dream of domestic flexibility can start becoming a reality.

Domestic flexibility is not an entirely new concept in the UK. After all, there have been teleswitch meters fitted in British homes since the 1980s.

But it nevertheless remains far less developed than industrial and commercial flexibility.

Half-hourly metering and time-of-use ­tariffs are already the norm for the country’s largest energy users. For years they have been able to lower their network charges by minimising their consumption during the three half-hours of peak demand each winter – also known as the triads.

Industrial and commercial demand-side response (DSR) regularly wins frequency response contracts and is now being bid into the balancing mechanism by Limejump and Flexitricity.

Despite the inability to secure the 15-year contracts available to new-build generation, it has also been relatively successful in the capacity auctions to date.

Domestic flexibility is much further behind.

Millions of households do now have half-hourly metering as a result of the smart meter rollout – yet very few are actually on time-of-use tariffs. Those that do exist are mostly glorified Economy 7 tariffs.

Only Octopus Energy’s Agile tariff offers half-hourly prices, and while it may be open to all customers, the company says it is effectively still in testing. So far, domestic flexibility has only really been used to provide grid services as part of trials and is completely absent from the capacity market.

This is hardly surprising. Flexibility from large energy users is, relatively speaking, the low-hanging fruit.

Households may have plenty of flexibility to offer collectively, but individual appliances can make very little difference on their own. It must therefore become cheap, quick and easy to combine them and operate them as so-called virtual power plants.

“My impression is that it’s on the verge of commercialisation,” says Rick Parfett, policy manager for the Association for Decentralised Energy. With the right changes, he believes domestic flexibility could “ramp up quite quickly between the next two and five years”.

But his list of changes is long.

Parfett says it will be important to be able to stack revenues from multiple services, including across local and national markets. Any exclusivity clauses that are not necessary must be removed, and the timings of contracts should be better aligned.

There need to be clear prioritisation rules, he adds: “Basically, that involves National Grid and the DNOs [distribution network operators] sitting down and establishing hierarchies of priority.

“If you’re providing a service that helps you solve a network stability issue, then that probably takes precedence over a ­commercial service.”

There also needs to be clarity over when and where flexibility will be needed: “The nature of demand-side response means that you need to talk to customers quite early on, gain their trust, aggregate everything up and get in a nice portfolio. That takes time.”

The first flexibility auctions by DNOs have sometimes been held just a few months in advance, which “isn’t anywhere near the time you need to start getting that flexibility, particularly if it’s in quite a specific location”.

Processes and systems, most of which were designed to work with a few hundred large generators, must be streamlined and upgraded to accommodate vast fleets of millions of assets.

They should also be standardised across DNOs: “It’s really minor things, but minor things when you’re trying to contract with individual domestic customers create quite a significant administrative burden.”

Parfett says the capacity market is one that is poorly suited to domestic flexibility at the moment: “They probably will need to reform things like minimum bid size and component reallocation. And if they really want domestic flexibility in there, they’re going to have to radically simplify things like the prequalification process.”

He says even experienced industrial and commercial aggregators struggle: “Anything that can be done to make that less clunky and manual and improve the process would be very welcomed by everyone.”

Market operators will also need a completely new mindset. They currently have the assurance of knowing exactly which assets are going to provide what service and can check them individually to make sure they can deliver.

“What happens when you’ve got a portfolio of distributed assets?” he asks. “If I’ve got 10,000 washing machines that I’ve aggregated up, I don’t know which ones are going to deliver, but I know by averages and by portfolio statistics that at any one time, 4,000 of the 10,000 will respond to a signal.”

They will instead have to put their faith in type-testing, monitoring and the law of large numbers: “As long as you’ve got good performance monitoring and the DSR provider is actually providing the service that you want it to, it doesn’t really matter what asset it comes from.

“Most of that is moving forward, but it takes a long time,” Parfett remarks. “They have got to change a lot of processes.”

He also highlights a number of regulatory changes that are being considered or implemented; for instance, the Balancing and Settlement Code (BSC) modification P344, which was approved by Ofgem in August last year.

The modification created a new type of BSC signatory called a virtual lead party, which will be able to form aggregated balancing market units without being the registered supplier for the sites they are aggregating or holding a supply licence at all.

Along with the Grid Code modification GC0097, which allowed sites to be aggregated across a wider geographical area, this should open up the valuable balancing market to virtual power plants. Code administrator Elexon began registration in March, but there are not yet any companies operating as virtual lead parties.

Then there is P379, which is currently undergoing assessment and seeks to further reform the supplier hub model by allowing consumers to buy and sell power through multiple parties by way of metering splitting.

“At the moment, you’re quite constrained in what you can offer,” says Parfett. “That opens up a range of new business models and it means you can do those things without needing permission from the main supplier.”

Daniel Mee, systems architect at the Energy Systems Catapult, says there is also lots of work to do to make different technologies and devices work together: “In my smart home, I want to balance my heating and my EV [electric vehicle] charging, using the battery that’s trickle charged over the day and maybe self-consuming from the solar panels on my roof.”

At the moment, he says “everything is being done as a vertical slice going back up to the cloud”.

Mee says this could create problems in the future: “The obvious example is that if the smart charger that’s charging your ­vehicle doesn’t have the hardware capability of talking to devices in your home but only upwards to the cloud, then in three or four years, trying to integrate across is going to be a lot more difficult than if you built that in at the beginning.

“If you look at the other industries like mobile telephony or the internet or any of those emergent digital areas, the lifecycle of the technology is much shorter,” he adds. Many consumers replace their mobile phone every few years.

“But I would proffer that chargers bolted to the house, your heating system in your house, that’s going to be a ten-year lifecycle, so your opportunity to iterate is lower. They tend to get replaced as a necessity when it breaks rather than as a lifestyle choice.”

One area that appears particularly promising is EV charging, which can easily be shifted to different parts of the day but will create significant extra costs if it takes place during peak hours.

Customers on Octopus’s Agile tariff can already smart charge their vehicles if they have the right equipment, and both itself and Ovo are offering their customers vehicle-to-grid charging on a trial basis.

Mee says EVs are a “great place to start” because they bring “obvious benefits to consumers straight away”. He says obtaining flexibility from EVs is “more urgent and more doable” than from appliances, adding: “If you learn the lessons right from that, you can maybe apply them to heat.”

It may be further behind, but Neal Coady, director of technology for home energy management at Centrica, says the adoption of low-carbon heating will also “naturally play into flexibility because you’ve got big loads there once you electrify heat to a greater or lesser degree.

“You’ve really got to start worrying about flexibility, otherwise you’re going to need a lot more power stations.”

The company recently received approval from the electricity system operator to provide frequency response using hundreds of smart water tanks installed in customers’ homes.

Coady says consumers are increasingly concerned about climate change and willing to make changes to help decarbonise the energy system, but the industry must make it easy and simple to do so: “You could buy a heat pump or a battery or solar panels and install them, but then you’re kind of on your own.

“People want to do something about decarbonising their energy, but when they start Googling and digging into it, they just get hit with a whole flood of stuff they don’t really understand.”

Both on stage and behind the scenes, there is a lot left to do.