Before the national grid, towns were powered from a local generating station. Big, relatively efficient boilers ran continuously, while smaller units were fired up for heavy usage periods. A human operator would watch the load and (literally) open the steam valves or bring reserve units in and out of operation to match output to demand.

From the 1930s the UK National Grid allowed demand to be pooled amongst different regions, which smoothed the peaks and troughs but increased the scale and magnitude of the balancing operation. Early attempts to influence users included government control of factory working hours and consumer incentives such as cheaper overnight tariffs.

Today, renewables can provide much of the UK’s electricity needs. Their output is roughly predictable by weather forecasting but cannot be controlled, so balancing demand against capacity matters more than ever. Done well, it reduces the amount of temporary, higher carbon generation, such as gas-fired capacity, brought into operation. Part of that balancing is now handled by the “smart grid”.

What is a smart grid?

The concept is not new; the term and the software-driven approach date from the early 2000s. Software and connectivity have since improved so that individual devices can adapt their load according to received price and carbon data, with so-called “aggregators” coordinating large fleets of them. The controlled devices tend to be relatively large electricity consumers such as electric vehicle (EV) chargers, heat pumps and domestic batteries. External control gives a more reliable response than simple tariff incentives. For example, an EV charger needed at 3am is far more likely to switch on if externally controlled than if a user must do it manually.

Cyber security is also central; an aggregator that can switch a fleet of heat pumps on or off is as much a potential risk to the grid as a benefit. The UK Government is consulting on a licensing regime for organisations that control domestic load, reflecting this.

IP aspects

In terms of the IP landscape, other than brand protection for the aggregators, activity falls into various areas.

• Regarding the devices themselves, patentable inventions can relate to controller circuitry, sensing techniques and device adaptations (for example, a heat pump that pre-heats a building, or a power consuming device with improved tolerance for externally controlled interruptions).
• Algorithm related inventions might involve scheduling logic to predict grid carbon emissions and capacity changes, to forecast domestic consumption and its reaction to control, and to reduce net power consumption when needed without adversely affecting comfort and operation. Virtual power plant technology, similar to the California example mentioned earlier, is attracting particularly heavy patent activity.
• Standards inevitably become involved when devices from multiple manufacturers must respond to signals from many aggregators. Standard essential patents (SEPs), fair, reasonable and non-discriminatory (FRAND) conditions and open protocols may determine where the value in these systems is captured.
• Finally, smart grid use generates a huge amount of data (for example, switching histories, household load profiles) which could itself be valuable. Here, data protection laws, trade secrets and database rights become relevant.

What comes next?

We can expect at least three main areas of development. Machine learning (ML) will play an increasing part in controlling devices and predicting grid loading. Vehicle-to-grid technology will allow EV batteries to store and return energy, having the advantage that EV batteries are often much higher capacity than domestic units. Peer-to-peer and local grid technology will allow local pooling of load and stored power, for example within a single residential street.

Will it make a difference?

Smart grid technology will not itself decarbonise the grid, but it can help the grid move towards decarbonisation by improving the usage of wind and solar while aiding the retirement of high carbon peak duty power plants. The scale of the benefit is debated. UCL modelling suggests smart appliances and EVs could cut UK peak load by around 9% by 2050, but other research warns that the carbon benefit can be negligible unless the way generation is dispatched also changes.

Useful links

• ElectronicsForU.com, 140,000 Home Batteries Supply 580 MW To California Grid: https://dycip.com/california-home-battery-grid
• UCL Discovery, assessing the benefits of demand-side flexibility in residential and transport sectors from an integrated energy systems perspective: https://dycip.com/ucl-demand-flexibility  
• University of Oxford Research Archive, how much do smart appliances reduce CO2 emissions? Assessing the environmental impact of domestic demand response technologies: https://dycip.com/oxford-smart-apps-co2