<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=515828355457072&amp;ev=PageView&amp;noscript=1">

Data for good: How AI is the key to maximising renewable energy

By Noel Peatfield on 20/06/18 12:40
Author Image
Noel Peatfield

photovoltaic-2814504_1920

Our second post in the 'Data for Good' blog series takes a closer look at the ways artificial intelligence (AI) can be utilised to improve access to cleaner and renewable energy for neighbourhoods across the UK. Guest blogger and tech expert Noel Peatfield explains further...


Greater Manchester Community Renewables (GMCR) is run by volunteers to install community-owned solar energy. To date, they have generated over 110,000 kWh of clean electricity, prevented more than 40 tonnes of carbon dioxide emissions and saved over £2,500 on electricity bills.

All together, as of 31 March, there are 2,724 community installations and 774,914 domestic installations selling their surplus renewable electricity back to the grid wholesale. Regulators are now looking into making the changes needed to allow small generators to provide their power directly to consumers peer to peer (P2P).

Volunteer and Director of GMCR, Andrew Hunt, explains: "The margin at which we currently sell surplus back to the grid would be certainly be improved if we could supply consumers directly, which would help us to continue building new solar installations for local schools and community centres."

April saw the first P2P energy trade in the UK conducted via the Ofgem regulatory sandbox. This was done between residents of Hackney's Banister House Estate, using a P2P energy trading platform and their solar panel installation as part of Ofgem's Innovation Link. 

This P2P model would make it possible to supply people in local neighbourhoods with electricity from their very own renewable source, all at a competitive price. With domestic consumers having access to multiple suppliers via smart meters, this new data flow will open up new possibilities for AI and greener energy. 

The current system of Load Profiles used to keep costs down, whilst ensuring that the supply is as clean as possible, is used to classify users into groups depending on patterns of consumption.

By taking the usage data of customers, eight profile classes have been established to help forecast the amount of electricity required by the grid. The examples below represent the average consumption of Profile Class 1. Those in Profile Class 2 are offered electricity at a cheaper rate during off-peak times to help equalise supply and demand every 30 minutes throughout the year.

Capture

Capture2
Graphs courtesy of Elexon

  • Profile Class 1: Domestic Unrestricted Customers
  • Profile Class 2: Domestic Economy 7 Customers
  • Profile Class 3: Non-Domestic Unrestricted Customers
  • Profile Class 4: Non-Domestic Economy Customers
  • Profile Class 5: Non-Domestic Maximum Demand Customers with a Peak Load Factor between 20-30%
  • Profile Class 6: Non-Domestic Maximum Demand Customers with a Peak Load Factor between 30-40%
  • Profile Class 7: Non-Domestic Maximum Demand Customers with a Peak Load Factor over 40%

As you can see, there are only two profile classes for domestic customers, segmenting a high number of consumers with variable patterns of use. With multiple suppliers for each meter, electricity supply will become more distributed. Together with the intermittent nature of renewable energy, future systems that include P2P will need to be able to automatically react to changing circumstances in real time - and that's where AI comes in.

Individual usage data, cleverly combined with weather forecasting data, can build up enough information for a system of AI to understand when the best times are to switch a customer over from their default supplier to a cleaner energy source, all while getting the best price for both the buying and selling parties.

To help balance supply and demand with multiple power sources, including onsite renewable generators, battery storage company Powervault has trialled a system where a self-learning algorithm builds up a picture of a specific building's consumption patterns.

Data scientist Ioanna Armouti explains: "SmartSTOR is a machine learning algorithm which uses historical house consumption data, metadata and weather data to best predict the power demand of the household, and therefore optimise Powervault's behaviour to maximise energy efficiency and savings."

If the plans to allow multiple suppliers to provide single residencies become a reality, the way homes receive their power will change, and we'll see energy start to be provided in all round better and smarter ways. For example, when solar panels have fully-charged a battery in a home while the occupants are at work (and are therefore using little or no power at home), AI could facilitate the distribution and costs of supplying neighbouring buildings - eventually cutting down unused renewable power to almost zero.

Smart metering, multiple suppliers and storage all generating more detailed user data - more data than has ever been seen before in the retail energy market - will enable AI to create further substantial savings and a significantly greener environment.

Want to know how you can use AI to help your business work in better and smarter ways? Download our free AI for Sales guide:

Download the AI sales guide