Across the globe, energy systems are changing, creating unprecedented challenges for the organizations tasked with ensuring the lights stay on. In the UK, large fossil-fuelled power stations are being replaced by increasing levels of widely distributed wind and solar generation. This renewable power is clean and free at the point of use but it cannot always be relied upon. To date, the National Grid has managed this intermittency by keeping polluting power stations online to make up the difference but artificial intelligence (AI) offers an alternative approach.
What’s needed is a smart grid which can integrate renewable energy efficiently at scale without having to keep polluting power stations online to manage intermittency. This requires energy storage to act as a buffer, reducing demand when supply is too low or increasing it when it is too high. Most people associate energy storage with batteries, but the cheapest and cleanest type of energy storage comes from flexibility in our demand for energy.
This demand-side flexibility takes advantage of thermal or pumped energy stored in everyday equipment and processes, from an office air-con unit, supermarket fridge or industrial furnace through to water pumped and stored in a local reservoir. The electricity consumption patterns of these types of devices are not necessarily time-critical. Provided they operate within certain parameters – such as room temperature or water levels – they can be flexible about when they use energy.
This means that when electricity demand outstrips supply, instead of ramping up a fossil fuelled power station, certain types of equipment can defer their electricity use temporarily. And if the wind blows and too much electricity is being supplied instead of paying wind farms to turn off we can ask equipment to use more now instead of later. Making our demand for electricity “intelligent” in this way means we can provide vital capacity when and where it is most needed and pave the way for a cleaner, more affordable, and more secure energy system.
The key lies in unlocking and using demand-side flexibility so that consumers are a) not impacted and b) appropriately rewarded. Open Energi, (A UK based company) has been exploring how AI and machine learning (ML) techniques can be leveraged to orchestrate massive amounts of demand-side flexibility – from industrial equipment, co-generation and battery storage systems – towards the one goal of creating a smarter grid. They have spent the last six years working with some of the UK’s leading companies to manage their flexible demand in real-time and help balance electricity supply and demand UK-wide. In this time, they have connected to over 3,500 assets at over 350 sites, operating invisibly deep with business processes, to enable equipment to switch on and off in response to fluctuations in supply and demand.
Already, they are well on the way to realizing a smarter grid, but to unlock the full potential of demand-side flexibility, there is a need to adopt a portfolio level approach. AI and ML techniques are making this possible, enabling us to look across multiple assets on a customer site, and given all the operational parameters in place, make intelligent, real-time decisions to maximize their total flexibility and deliver the greatest value at any given moment in time.
For example, a supermarket may have solar panels on its roof and a battery installed on site, as well as flexibility inherent in its air-con and refrigeration systems. Using AI and ML means we can find creative ways to reschedule the power consumption of many assets in synchrony, helping the National Grid to balance the system while minimizing the cost of consuming that power for energy users.
A Lack of data is often an obstacle to progress but we collect between 10,000 and 25,000 messages per second relating to 30 different data points and perform tens of millions of switches per year. This data is forming the basis of a model which can look at a sequence of actions leading to the rescheduling of power consumption and make grid-scale predictions saying “this is what it would cost to take these actions”. The bleeding edge in deep reinforcement learning shows how, even with very large scale problems like this one, there are optimization techniques that can be used to minimize this cost beyond what traditional models would offer.
More rapid progress could be made across the industry if energy companies made more anonymised half-hourly power data available. It would enable companies working on smart grid technologies to validate these ideas quickly and cheaply. In the same vein, it would be a major breakthrough for balancing electricity supply and demand if energy companies made available APIs for reporting and accessing flexibility; it would allow companies like Open Energi to unlock enormous amounts of demand-side flexibility and put it to good use balancing, not just the grid but also helping to optimise the market positions of those same energy companies.
AI can help us to unlock this demand-side flexibility and build an electricity system fit for the future; one which cuts consumer bills integrates renewable energy efficiently and secures our energy supplies for generations to come.